Markers and Function of Human NK Cells in Normal and Pathological Conditions

Cytometry Part B (Clinical Cytometry) 92B:100 114 (2017) Review Markers and Function of Human NK Cells in Normal and Pathological Conditions Genny Del Zotto, 1 Emanuela Marcenaro, 2,3 Paola Vacca, 2,4 Simona Sivori, 2,3 Daniela Pende, 4 Mariella Della Chiesa, 2 Francesca Moretta, 5,6 Tiziano Ingegnere, 7 Maria Cristina Mingari, 2,4 Alessandro Moretta, 2,3 and Lorenzo Moretta 7 * 1 Istituto Giannina Gaslini, Genova, Italy 2 Department of Experimental Medicine, University of Genova, Genova, Italy 3 Centro di Eccellenza per la Ricerca Biomedica CEBR, Genova, Italy 4 U.O. Immunology IRCCS AOU San Martino-IST, Genova, Italy 5 Department of Internal Medicine, University of Verona, Verona, Italy 6 Ospedale Sacro Cuore Negrar, Verona, Italy 7 Department of Immunology, IRCCS Bambino Gesu Children s Hospital, Rome, Italy Natural killer (NK) cells, the most important effectors of the innate lymphoid cells (ILCs), play a fundamental role in tumor immune-surveillance, defense against viruses and, in general, in innate immune responses. NK cell activation is mediated by several activating receptors and co-receptors able to recognize ligands on virus-infected or tumor cells. To prevent healthy cells from auto-aggression, NK cells are provided with strong inhibitory receptors (KIRs and NKG2A) which recognize HLA class I molecules on target cells and, sensing their level of expression, allow killing of targets underexpressing HLA-class I. In vivo, NK cell-mediated anti-tumor function may be suppressed by tumor or tumor-associated cells via inhibitory soluble factors/cytokines or the engagement of the so called immune-check point molecules (e.g., PD1-PDL1). The study of these immune check-points is now offering new important opportunities for the therapy of cancer. In haemopoietic stem cell transplantation, alloreactive NK cells (i.e., those that express KIRs, which do not recognize HLA class I molecules on patient cells), derived from HSC of haploidentical donors, are able to kill leukemia blasts and patient s DC, thus preventing both tumor relapses and graft-versus-host disease. A clear correlation exists between size of the alloreactive NK cell population and clinical outcome. Thus, in view of the recent major advances in cancer therapy based on immuno-mediated mechanisms, the phenotypic analysis of cells and molecules involved in these mechanisms plays an increasingly major role. VC 2017 International Clinical Cytometry Society Key terms: natural killer cells; innate lymphoid cells; immune check-points; haploidentical stem cell transplantation; viral infections; immunophenotyping; flow cytometry; cell biology; cancer; bone marrow transplant; immune monitoring; immunotherapy How to cite this article: Del Zotto G, Marcenaro E, Vacca P, Sivori S, Pende D, Della Chiesa M, Moretta F, Ingegnere T, Mingari MC, Moretta, A, and Moretta L. Markers and Function of Human NK Cells in Normal and Pathological Conditions. Cytometry Part B 2017; 92B: 100 114. *Correspondence to: Prof. Lorenzo Moretta, MD, Director, Immunology Area, Director Department Laboratories, Pediatric Hospital Bambino Gesu, VialeSanPaolo,15,00146Rome,ItalyEmail:lorenzo.moretta@opbg.net Grant sponsor: Associazione Italiana Ricerca sul Cancro (AIRC); Grant numbers: special project 5x1000 n8. 9962, AIRC IG 2014 Id. 15704, AIRC IG 2014 Id 15283, AIRC IG 2015 Id 16764. Grant sponsor: Progetto di Ricerca di Ateneo 2014. Received 4 November 2016; Revised 22 December 2016; Accepted 29 December 2016 Conflict of interest disclosure Moretta A. is a founder and shareholder of Innate-Pharma (Marseille, France). The remaining authors declare no competing financial interests. Published online 5 January 2017 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/cyto.b.21508 VC 2017 International Clinical Cytometry Society

MARKERS AND FUNCTION OF HUMAN NK CELLS 101 INTRODUCTION NK cells are major players of innate immune defenses (1,2). They belong to the rapidly growing family of innate lymphoid cells (ILC) (3 6). ILC, different from T or B lymphocytes, do not express receptors encoded by rearranging genes. While the other members of the ILC family have been discovered only in recent years, due mostly to their prevalent localization in mucosal tissues and in secondary lymphoid organs, NK cells are known since over 40 years. They were originally detected in spleen, peripheral blood (PB) and bone marrow (BM), however, more recent studies revealed their presence also in lymph nodes, thymus, uterus and liver. NK cells rapidly migrate to inflamed tissues or secondary lymphoid organs, where they play a defensive role against pathogens. They were originally identified on the basis of their ability to kill tumor cells in the absence of previous activation, and were shown to be involved in immunosurveillance against tumors and in preventing tumor metastases. Activated NK cells produce a number of chemokines and cytokines that mediate and/or amplify inflammatory responses. In addition, NK cells may greatly regulate the downstream immune responses, influencing both the quality and the strength of adaptive immune defenses (1,7). The molecular mechanism by which NK cells discriminate between virus infected or tumor cells and normal cells remained obscure for many years after their discovery (8,9). Only in early 90, following the missing self hypothesis (10) and thanks to the identification of MHC-class I specific inhibitory receptors (11) and of an array of activating receptors (12), the mechanisms involved in NK cell activation and inactivation has been clarified. Recognition of self MHCclass I molecules by specific inhibitory receptors was identified as the most important mechanism to prevent the NK-mediated attack to normal cells. INHIBITORY AND ACTIVATING NK RECEPTORS The prototypes of HLA-specific inhibitory receptors were identified in humans by Alessandro Moretta and coworkers (11,13 15). Originally termed p58.1 and p58.2, these members of the Ig superfamily, were shown to be expressed on subsets of PB NK cells, variable in size in different individuals. They were found to recognize allotypic determinants of HLA-C molecules, shared by many HLA-C alleles. Other members of such receptors were subsequently identified and molecularly characterized (16,17). They were collectively termed Killer Ig-like receptors (KIRs) and shared the property of binding different groups of HLA-class I alleles, also including HLA-B and HLA-A alleles. Another HLA-class I- specific inhibitory receptor, the CD94/NKG2A, was found to recognize HLA-E (18,19). Notably, under normal conditions, HLA-E is expressed at the cell surface upon binding peptides derived from the leader sequence of most HLA-class I molecules. Thus, the levels of expression of HLA-E is proportional to the amount of HLA-class I molecules in a given cell. Accordingly, CD94/NKG2A receptors sense the overall expression of HLA-class I molecules on cells. Notably, CD94/NKG2A is expressed relatively early during peripheral NK cell maturation, while KIRs appear only at later stages, although co-expression of CD94/NKG2A and KIRs is detectable in a subset of maturing NK cells. The substantial differences in the modality of HLAclass I recognition by NK cells at different stages of maturation that express different receptors have important effects on NK cell function. These would be true not only in NK-mediated killing of virus-infected or tumor cells, but also in the setting of allogeneic, haploidentical hemopoietic stem cell transplantation (see below). Importantly, also activating forms of KIRs were discovered (20,21). Activating KIRs display a high degree of sequence homology, in their extracellular Ig domains with the corresponding inhibitory KIRs. Despite these similarities, the majority of activating KIRs do not have a binding specificity similar to their inhibitory counterpart. So far, this has been clearly proven only for KIR2DS1. While the receptor mediated NK cell inhibition represents an important fail-safe mechanism to avoid inappropriate attack to healthy cells, it also implies the existence of on signals, inducing NK cell triggering upon interaction with potential target cells (e.g., tumor cells). Unravelling the mechanisms of human NK cell triggering following interaction with target cells initiated when Sivori et al. identified NKp46, a surface receptor capable of inducing NK cell activation which is involved in the detection and killing of different tumors (22,23). Other two receptors were subsequently identified, namely NKp44 (24,25) and NKp30 (26). Although the molecular characterization of these receptors revealed major differences in their molecular structures, they were collectively referred to as natural cytotoxicity receptors (NCR) (12). In most instances, NK-mediated killing of different tumors involved one or more NCR. Another important receptor mediating NK cell activation is NKG2D, while others, including 2B4 (27), NTBA (28,29), DNAM-1 (30), CD59 (31), and NKp80 (32) primarily display the function of coreceptors, that is are capable of amplifying the effect of NCRs or NKG2D (Fig. 1). In addition, NK cells may express Toll-like receptors (TLRs) that, after interaction with bacterial or viral products and in the presence of proinflammatory cytokines, can induce NK cell activation (33 35). Thus, it appears that NK cells do not express a master receptor like T or B lymphocytes, but their activation depends on the engagement of multiple receptors and the strength of NK cell activation reflects the number of triggering receptors involved and their surface density. In addition, it is evident that NK cell activation requires that the activating NK receptors recognize ligands expressed on target cells. Most of these ligands have now been identified (36 38). In general, they are absent or expressed in small amounts in normal cells (with the remarkable exception of haemopoietic cells), while they are expressed de novo or upregulated in stressed normal

102 DEL ZOTTO ET AL. FIG. 1. Inhibitory and activating NK cell receptors and their ligands. The main NK cell receptors and their respective ligands are depicted. cells and, in particular, in virus-infected or tumor cells. Thus, the expression or lack thereof of ligands for activating receptors represents another important checkpoint for NK cell activation (39). Certain particular activating receptors, such as the HLA-E-specific NKG2C, are expressed only at late stages of NK cell maturation and may be correlated with CMV infection (see below). Notably, HLA-class I-specific inhibitory receptors including KIR and CD94/NKG2A are also expressed by some cytolytic T lymphocytes (40). These T cells were shown to be present only in some individuals and to represent oligoclonal or monoclonal cell expansions (41). Subsequent studies revealed that they recognize by their abtcr HLA-E molecules (42) that are known to bind self and CMV-derived peptides, corresponding to the leader sequences of various HLA-class I alleles or of the UL-40 protein of the human CMV (43 45). A number of other receptors expressed by NK cells are listed in Figure 1. Some of these molecules are expressed only at given stages of differentiation or by NK subsets. For example, CCR7 is expressed by CD56bright NK cells whereas CXCR1, CD16 and KIRs are expressed on CD56dim NK cells. Other markers, such as 2B4, NTBA, CD2, DNAM-1, NKG2D, NKp80, are not restricted to a specific NK cell subset. Of major interest is CD16, a low affinity Fcg receptor, binding to IgG complexes and mediating the antibody-dependent cell-mediated cytotoxicity (ADCC). The bright expression of CD16 is restricted to mature, KIR1 cytolytic NK cells. DIFFERENTIATION OF NK CELLS AND OTHER ILCS NK cells derive from haemopoietic stem cells (HSC), which give rise to all blood cells. Their stages of differentiation are characterized by the sequential acquisition (or loss) of markers and functional properties. NK cell differentiation can be analyzed in vitro by culturing HSC under suitable conditions including mixture of cytokines and/or stromal cells. HSC are usually enriched on the basis of the surface expression of CD34 antigen (that is lost during HSC differentiation towards the various haemopoietic cell lineages). Notably, some stages of NK cell differentiation can be identified in the PB of patients who received HSC transplantation to cure acute

MARKERS AND FUNCTION OF HUMAN NK CELLS 103 FIG. 2. Stages of human ILC development. Model of ILC differentiation pathway based on findings obtained in murine models. Common lymphoid progenitor (CLP) is capable to give rise to both adaptive (T and B cells) and innate (cytotoxic- and helper-ilcs) lymphocytes. From common ILC precursor (CILP) originate: 1) cytotoxic ILC (i.e., NK cells) through a NK cell-precursor (NKp); 2) common helper-ilc precursor (CHILP). The CHILP would further differentiate towards mature helper ILC trough specific ILC precursors (ILC2p and ILC3p). Transcription factors defining/driving the different ILC subsets and signature cytokines are indicated. For sake of simplicity, in this figure, CD56bright and CD56dim NK cells have been placed in the same column. Indeed, it is now well established that CD56dim derive from CD56bright NK cells and that further maturation steps have been identified on the basis of the expression of different markers/receptors (see Fig. 3). The boxes (right) show the most informative markers for each cell population. leukemia or other severe, non-malignant disorders. Similar to other haemopoietic cells, NK cell differentiation is induced by cytokines, transcription factors (TF) activation and cellular interactions. The common lymphoid precursor (CLP) expresses the ID2 TF, while differentiation towards NK cells is guided by Eomesodermin (Eomes) and T-box transcription factor T-bet (Tbx2I) TF (46). Tbx2I is required also for the differentiation of the interferon-g producers, non cytolytic, ILC1 (47,48). On the other hand, ILC2 differentiation requires GATA3 while ILC3 require RORgt TF(49).Figure 2 shows a tentative scheme of NK and ILC differentiation based on recent data. Notably, however, the plasticity observed in different experimental systems would suggest the occurrence of more complex interactions among different ILC and their precursors (50). In any case, mature NK cells and ILC may now be identified according to given cell surface phenotypes and functional properties (such as cytolytic activity and/or production of defined cytokines). Remarkably, some surface molecules/receptors display different functional capabilities at different maturation stages. For example, CD161 functions as an activating receptor during early stages of NK/ILC differentiation, leading to the production of cytokines such as IL-8 (51). On the other hand, 2B4 that functions as an activating co-receptor in mature NK cells (27) displays a strong inhibitory activity at early stages of NK cell maturation (52). This opposite effect is due to the late appearance of the SAP intracytoplasmic protein, crucial for the generation of triggering signals via 2B4. Notably, deletion or critical mutations of SAP are responsible of the lifethreatening immunodeficiency X-linked lymphoproliferative disease (XLP) (53). Regarding the cytokines required for NK cell development, a central role is played by IL-15 which acts in

104 DEL ZOTTO ET AL. combination with IL-7, Flt3-L and SCF (52). Remarkably, human HSC cannot respond directly to IL-15 because they do not express CD122 (the common b-chain of the IL-2/IL-15 receptor) and CD132 (the common g chain receptor). Both chains are expressed by HSC upon exposure to SCF and Flt3-L, thus becoming responsive to IL- 15. Late stages of NK cell differentiation require the synergistic effect of IL-15 and IL-21 (54). Finally, other cytokines, including IL-8 and MIP-1a, favor the preferential expansion of cells of the NK cell lineage, primarily by inhibiting the myeloid cell development (51). Although NK cell development from HSC or CLP occurs primarily in the BM, it is now clear that this process may also take place in the periphery (in both lymphoid and non lymphoid tissues) (55). For example, CD341 cells undergoing in vitro differentiation towards NK cells have been detected in tonsils, mucosaassociated lymphoid structures, thymus (56) and PB. Similarly, precursors capable of differentiating towards other ILC were detected in tonsil, decidua and mucosaassociated lymphoid structures (57,58). These data suggest that HSC, CLP or both, migrating from BM to various peripheral tissues, undergo differentiation towards NK or ILC in loco. This may allow a more prompt response to the need of generating effector cells in response to infection or other danger signals occurring in peripheral tissues. Thus, the presence of both mature ILC (or NK cells) and their precursors in tissues may allow both a prompt response to infection and a rapid generation of other effector cells in situ (50). THE PROCESS OF NK CELL LICENSING AND SELF-TOLERANCE KIR genes (located on chromosome 19) and HLA genes (on chromosome 6) are inherited independently. Accordingly, KIRs may be expressed in the absence of their HLA ligands (59). Extensive clonal analysis revealed that the large majority, but not all, NK cell clones expressed inhibitory receptors for self HLA-class I molecules. The few NK cells lacking such self-reactive inhibitory receptors were virtually anergic (60). This suggested the existence of a mechanism by which, during NK cell maturation, only those cells expressing at least one self-reactive inhibitory KIR are licensed upon interaction with the relevant HLA-allele. Those lacking such receptors would not be licensed, remaining hyporesponsive. As a result of this education process the KIR repertoire results largely influenced by self HLAclass l. I alleles (61,62). On the other hand, the activating KIRs display an opposite functional behavior, since the specific recognition of a self-allele by NK cells leads to NK cell hyporesponsiveness (63). These mechanisms are important in the NK cell reconstitution occurring in haplo-hsct in which NK cells originated from HSC of the donor acquire a KIR repertoire of donor and not of recipient type. This result may appear contradictory. However, can be explained by the very high numbers of donor cells infused, leading to a BM microenvironment that is predominantly of donor type (64). PERIPHERAL NK CELLS SUBSETS IDENTIFIED ON THE BASIS OF SURFACE MARKERS Most peripheral NK cells are composed of mature cells. However, they are rather heterogeneous both in terms of phenotypic and functional characteristics. Two major subsets were first identified on the basis of the surface expression of CD56 molecules: CD56 bright and CD56 dim (55,65). Until recently, cytokine production was thought to be restricted to the CD56 bright cells, while cytolytic activity was associated to the CD56 dim subset. However, it has been shown that the cytolytic CD56 dim cells contain high amounts of mrna for IFN-g and produce high levels of this and other cytokines upon receptor-induced NK cell triggering (66,67). The two subsets display a different tissue distribution. While CD56 dim are largely predominant in PB where they represent about 90%, CD56 bright are more frequent in certain tissues and secondary lymphoid organs. In agreement with their different cytolytic potential, CD56 dim express high levels of cytolytic granules, containing perforin and granzymes that are poorly expressed in CD56 bright cells. These cells release cytokines, relatively late (>16 h) after cell triggering. This is primarily consequent to cytokine-induced cell activation. Functional differences between the two NK subsets are related, at least in part, to the differential expression of cytokine receptors (IL-2, IL-15, and IL-7 receptors) and of functional NK receptors. Different experimental evidences have now established that CD56 bright represent an early stage of peripheral NK cell maturation, further progressing to CD56 dim cells. Notably, telomeres are longer in CD56 bright than in CD56 dim cells (68). In addition, analysis of mice with a humanized immune system revealed that infused CD34 1 cells give rise to CD56 bright first and subsequently to CD56 dim NK cells that also become KIR1 (69). Likewise, in haplo-hsct, the first wave of lymphoid cells, derived from donor s HSC (2 3 weeks after HSCT), is entirely composed of CD56 bright cells expressing the HLA-E-specific CD94/NKG2A inhibitory receptors. CD56 dim, that may also express KIR, are detectable only 4 6 weeks later. Cells at later stages of NK cell differentiation in addition to KIR also express high levels of CD16 (FcgR) and CD57, lose the expression of CD94/NKG2A and are characterized by a low proliferative capacity (70,71). Differences in tissue distribution and homing capability of CD56bright and CD56dim cell population are dependent on the chemokine receptors expressed. Thus, the CD56 dim subset expresses chemerin-r, CXCR1 and CX3CR1 and respond to the chemokines, IL-8 and CX3CL1 (fractalkine), determining their migration towards inflammatory peripheral tissues (65,72). In contrast, CD56bright, expressing CCR7 (and CD62L) are attracted by CCL19 and CCL21 produced by cells present in secondary lymphoid organs. Interestingly, CD56 dim cells can acquire CCR7 by trogocytosis, upon

MARKERS AND FUNCTION OF HUMAN NK CELLS 105 interaction with CCR71 cells (e.g., dendritic cells) and acquire the ability to migrate to lymph nodes (73,74). Regarding the interaction between NK and DC, their crosstalk may be an important mechanism to regulate downstream adaptive responses. Indeed, It has been shown that NK cells are capable of killing immature DC that have failed to undergo proper maturation and express low amounts of HLA-class I molecules (1,7,75,76). Thus, thanks to this NK-DC interaction taking place at the interface between natural and adaptive immunity, NK cells can exert a quality control of DC (the so called DC editing) and eliminate cells unfit for proper antigen presentation and induction of useful TH1 responses (77,78). NK CELLS AND VIRAL INFECTIONS Although NK cells have been discovered because of their ability to kill tumor cells, it is conceivable that their evolution and functional specialization has been primarily dictated by infectious agents. Indeed, NK cells play a major role in cellular defenses in primary virusinfections, before an effective adaptive response is generated (79). This NK cell function is particularly important to limit the aggressiveness of cytopathic viruses, allowing virus-specific cytolytic T lymphocytes to be properly expanded and generate sufficient cell numbers. Notably, two main mechanisms by which NK cells kill virus-infected cells have been identified. (1) Certain viruses, primarily herpes viruses, induce a profound downregulation of HLA-class I antigens. In this case, NK cells are not blocked by the interactions between HLAclass I and inhibitory-r (19,80,81). In cells-infected by many different viruses, viral peptides may bind to and alter the tridimensional structure of HLA-class I alleles to such an extent that KIR are not properly engaged, resulting in the lack of inhibition and killing of infected cells. In both cases, it should be recalled that virus infected cells express high levels of ligands recognized by activating NK receptors and induce NK cell triggering. In addition to these two mechanisms leading to killing of virus-infected cells and requiring cell-to-cell interaction, other means by which NK cells can limit virus spread is the production of IFN-g, interfering with viral replication (82), or the activation of a TLR-mediated anti-viral response (33). A particular scenario occurs in the case of CMV infection/reactivation, in patients receiving a HLAhaploidentical HSCT to cure high risk leukemias (see below). In these patients, an expansion of NKG2C 1 NK cells, accompanied by a rapid NK cell maturation, can be detected (71,83 85). Although, NKG2C could not be identified as a CMV-specific receptor, NKG2C1 NK cells have been shown to display the features of memory-like cells undergoing expansion following re-exposure to CMV (84). Indeed, in healthy HCMV 1 donors, a fraction of HCMV-induced NKG2C 1 NK cells expressing CD57 show an adaptive signature consisting in broad epigenetic modifications likely responsible for the downregulated expression of given signaling molecules that is the adaptor proteins FceRg and EAT-2 and the tyrosine kinase Syk, observed in this subset (Fig. 3B). This distinctive molecular signature can modify certain functional capabilities of the adaptive NK cell subset. In particular, the lack of FceRg may favor a more efficient killing via ADCC of opsonized HCMV-infected targets (86,87). Notably, in rare cases of NKG2C2/2 individuals, a noticeable expansion of cells expressing activating KIRs could be detected (88). Although the molecular mechanism underlying these peculiar NK responses remain to be elucidated, these data suggest that NK cells expressing NKG2C or activating KIRs may play a relevant role in immune responses against CMV and display features common to T-cell-mediated, adaptive responses. Recently, a further novel NK cell subset has been identified in HCMV 1 individuals. This subpopulation, called PD-1 1 NK cells is mainly composed of fully mature NK cells and displays impaired capacity to kill PD-L 1 tumor cells (see below), due to the expression on their surface of high levels of PD-1 that functions as an important immune checkpoint (89,90) together with low levels of NCRs. The conditions that lead to the generation of this subset and its role in health and disease, in particular in patients with advanced cancers, are still unknown (91,92). EFFECT OF THE TUMOR MICROENVIRONMENT ON THE NK CELL PHENOTYPE AND FUNCTION Despite the fact that NK cells display a potent cytolytic activity against tumor cells in vitro, this functional capability may be strongly impaired by the tumor microenvironment. Because of the low or absent HLA-class I expression, certain tumors are virtually undetectable by tumor-specific cytolytic T cells. However, provided that these tumors express ligands for activating NK receptors, they are susceptible to NK-mediated killing. Notably, NK cells have also been shown to kill cancer stem cells (CSC) (93,94), this is indeed an important functional capability since CSC, responsible of tumor relapses, are resistant to chemo- and radio-therapy, due to their quiescent status. However, in spite of this, various attempts to adoptively transfer autologous or allogeneic NK cells in cancer patients resulted poorly effective. While an inefficient NK cell traffic to tumor sites may be involved, mechanisms of tumor resistance or escape from the NK-mediated attack certainly play a central role. Indeed, NK cells isolated from various tumor samples were shown to be hypofunctional. This functional impairment is due not only to the inhibitory activity of tumor cells, but also to that of different cell types present in the tumor microenvironment. For example, it has been shown that even fibroblasts of the tumor stroma can exert a potent suppressive effect on both NK cell proliferation and effector function. This inhibition is in part consequent to the release of inhibitory factors, including kinurenine, PGE-2 and cytokines such as TGFb, IL-10 and, at least in the case of acute myeloid leukemias, IL-1b (95 98). In addition, hypoxia further contributes to establish an inhibitory tumor

106 DEL ZOTTO ET AL. FIG. 3. Heterogeneity of CD56 dim peripheral NK cells and possible expression of PD-1 or NKG2C. (A) Peripheral blood NK cells are composed of CD56 bright CD16 dim/- NK cells (more immature and poorly represented in PB) and of CD56 dim CD16 1 NK cells (90% of PB NK cells) that comprise cells at different stages of maturation: maturing NKG2A 1 KIR 2, NKG2A 1 KIR 1, and mature NKG2A 2 KIR 1 NK cells. Among mature NK cells those expressing high levels of CD57 and CD16 represent terminally differentiated ones. (B) In the PB of HCMV 1 healthy donors a subset of terminally differentiated NK cells (CD56 dim KIR 1 NKG2A 2 CD16 high CD57 high ) brightly expressing the HLA-E-specific activating receptor NKG2C is detectable in variable proportions (upper panel). This HCMV-induced subset retains memory-like properties and can be further characterized by broad epigenetic modifications, decreased expression of given signaling molecules (i.e., FceRg, Syk, EAT-2) and transcription factors (i.e., PLZF) (see text). In HCMVhd this memory-like NKG2C 1 CD57 1 cell subset is virtually absent (lower panel). (C) In the majority of HCMV 1 healthy donors, an additional terminally differentiated NK cell subpopulation expressing the immune checkpoint inhibitory receptor PD-1 can be identified. This PD-1 1 NK cells subset is also characterized by a remarkably low expression of the NCRs NKp46 and NKp30 (Fig. C, purple dots). microenvironment (99). However, we have shown that NK cells isolated from malignant pleural effusions are not anergic and, upon IL-2 activation, they could efficiently kill autologous tumor cells. This finding has relevant implications for a possible NK-based adoptive immunotherapy in patients with primary or metastatic tumors of the pleural cavity (100). Another important mechanism that further increases tumor resistance to immune responses is the induction of inhibitory loops involving other cell types that, once recruited to the tumor site, become polarized to a suppressor phenotype. These cells include regulatory T cells (Treg) (101), myeloid-derived suppressor cells (MDSC) (102) and M2 macrophages (103). Importantly, in many instances, the compromised NK cell function reflects modulation of activating receptors involved in tumor cell recognition and killing. While this phenomenon may be consequent to shedding from tumor cells of ligands of such activating receptors, soluble inhibitory factors and hypoxia appear to play a more important role. In this context, recent data have revealed that NKp30 and DNAM-1 expression on NK cells derived from peritoneal fluid (PF) of patients affected by seropapillary ovarian carcinoma was significantly lower on tumorassociated NK cells than in NK cells isolated from PB (104,105). The impaired expression of NKp30 was related to the chronic interactions between this activating NK receptor and one of its ligand (termed B7-H6) (37), present in soluble form in the ascitic fluid (Fig. 4). Regarding the effect of hypoxia, it is noteworthy that, while it profoundly inhibits the NK-mediated tumor cell

MARKERS AND FUNCTION OF HUMAN NK CELLS 107 FIG. 4. Impaired expression of activating receptors in tumor associated NK cells. In patients affected by seropapillary histotype ovarian carcinoma, the expression of the activating receptors NKp30 and DNAM-1 in tumor-associated NK cells is substantially reduced as compared to autologous PB- NK cells, whereas the expression of NKp46 is unchanged. This impaired expression is thought to be due to the chronic interaction between activating NK receptors and their ligands. In addition, in most cases, high proportions of PD-1 1 NK cells are detectable. killing, has no effect on ADCC, because both the expression and the function of CD16 are not affected. Because the efficacy of the immunotherapy with tumor-specific antibodies is largely dependent on ADCC, this notion may be particularly relevant for clinical applications. A novel additional and particularly important mechanism leading to NK cell inactivation involves the interaction between the PD-1 inhibitory receptor expressed by highly mature NK cells and its ligand(s) PD-L1 and 2 on tumor cells. PD-1 was originally discovered on cytolytic T cells and was found to exert a sharp inhibitory effect on their anti-tumor activity. We have recently shown that a portion of PB-NK cells in patients with ovarian carcinoma is brightly positive for PD-1 (91). Analysis of cells isolated from the ascitic fluid revealed an even higher proportion of PD-1 1 NK cells, suggesting their possible induction/expansion in the tumor environment. Functional analysis showed that purified PD-11 cells did not kill PD-L11 tumor cells. However, the cytolytic activity could be restored by mab-mediated masking of PD-1. It is conceivable that restoration of NK cell function may play an important role in the therapeutic effect of anti-pd-1 therapy, especially in tumors that are poorly susceptible to T cell-mediated immunity (91,106). In this context, PD-L1 has been detected on neuroblastoma tumor cells (107), thus offering an important clue for

108 DEL ZOTTO ET AL. FIG. 5. Haplo HSC: donor selection on the basis of the presence and of the size of alloreactive NK subset. (A) In haplo-hsct, donor NK cells are alloreactive versus the recipient if a KIR-L (HLA alleles) is present in the donor and absent in the recipient (i.e., KIR-L mismatch), and the inhibitory KIR (ikir) specific for the mismatched KIR-L is expressed in the donor KIR gene profile. The alloreactive NK cells express this ikir as the only HLA-specific inhibitory receptor. (B) the alloreactive NK cell subset (represented by green cells) can be variably represented in different donors and the donor selection is based on the size of this alloreactive population (i.e., haplo-donor A in this figure). (C) the size of the alloreactive NK cell subset (green) can be assessed by cytofluorimetric analysis using appropriate combinations of anti-kirs and anti-nkg2a mabs. the therapeutic use of anti-pd-1 (or anti-pd-l1) mabs in this otherwise fatal pediatric tumor. NK CELLS IN THE THERAPY OF HIGH RISK LEUKEMIAS: THE HAPLO-IDENTICAL HAEMOPOIETIC STEM CELL TRANSPLANTATION SETTING As discussed above, in a self -environment, NK cells may kill only cells with impaired expression of HLAclass I molecules. On the other hand, in an allogeneic setting, if a mismatch occurs between KIR and HLAclass I molecules, NK cells may kill allogeneic cells (13,15) provided that these express ligands of activating NK receptors (12). This situation may occur in allogeneic HSCT, in particular in haploidentical HSCT. Allogeneic HSCT is the best therapeutic option for the therapy of high risk leukemia. However compatible donors can be found only for 70% of patients. The prognosis of the remaining patients, in the absence of HSCT, is invariably particularly poor. For these patients, however, a family member identical for one HLA haplotype and mismatched for the other may be immediately available, especially for pediatric patients. Because the incompatibility at three HLA loci would cause severe/fatal GvHD, an extensive T cell depletion is strictly required, in association with the infusion of high doses of purified CD341 cells. Thanks to this protocol, haplo-hsct resulted in low percentages (10%) of mild (grade I or, rarely grade II) GvHD and high rates of engraftment, in the absence of post-transplant immunosuppression (108,109). These pioneering clinical studies reported a potent graft versus leukemia (GvL) effect in adult patients with AML (108). In addition, it was shown that NK cells derived from donor CD341 cells were responsible for the anti-leukemia effect. However, it was also evident that the positive effect occurred only in those donor/patients pairs in which NK cells were alloreactive (i.e., expressing KIR specific for HLA-class I alleles absent in the recipient) (Fig. 5A). Similar results were also obtained in pediatric patients, affected by high risk ALL. Importantly, in haplo-hsct, the GvL effect was

MARKERS AND FUNCTION OF HUMAN NK CELLS 109 independent from GvHD (caused by T cells). Thus, alloreactive NK cells attack leukemia blasts which express ligands for activating NK receptors that are not expressed by normal tissues (with the important exception of haemopoietic cells). Notably, this particular tissue distribution of the ligand explains why donor alloreactive NK cells may cause only low rates of GvHD and graft rejection. Indeed, they kill patient DC (preventing activation of contaminating donor T cells, resulting in GvHD) and patient T cells (preventing rejection of the graft) residual after the conditioning regimen (110). Because donor s alloreactive NK cells are central in preventing leukemia relapses, GvHD, and graft rejection, it was essential to determine whether alloreactive NK cells were present in the donor and to assess the size of the alloreactive population in different potential donors (Fig. 5B). Indeed, since in haplo-hsct two or more donors may be available, it is important to select the most suitable one. Another important information was to assess whether alloreactive NK cells were generated in the patient from donor CD341 cells and could acquire full functional capability. It has been possible to answer these questions and to identify the alloreactive NK subset thanks to the use of suitable combinations of anti- KIR mabs in multicolor cytofluorimetric analysis (111). More recently, it became evident that certain activating KIRs contribute to the anti-leukemia effect and that their presence correlates with a positive clinical outcome (112). Therefore, the selection of mabs capable of discriminating between inhibitory and activating KIRs allowed a more precise definition of the size of the alloreactive NK subset (Fig. 5C). Additional criteria for donor selection have been added recently. These include: (a) the presence of a KIR B haplotype (i.e., containing genes encoding for activating KIR) (113,114); (b) mothers have been shown to be better donors than fathers (115); (c) higher numbers of peripheral blood lymphocytes. These additional criteria revealed to be useful particularly in the absence of alloreactive NK cells. The survival rate in the case of NK alloreactivity was over 70% for ALL and over 40% for AML. Although these results could be considered a major step forward, in the cure of high-risk leukemias, both transplant-related mortality and leukemia relapses, particularly in AML, clearly implied the need of further improvement. In this context, it should be considered that KIR1 NK cell maturation from CD341 precursors requires 6 8 weeks, resulting in delayed GvL effect and consequent risk of early leukemia relapses. To reduce this risk, a novel transplant approach, has been introduced foreseeing the simultaneous infusion of donor CD341 and mature KIR1 NK cells. This approach is based on the removal of TCRab T cells and of CD191 B-lymphocytes (to prevent GvL and to reduce the risk of B cell malignancies, related to the immune-compromised conditions) (116,117). Importantly, other donor mononuclear cell types, with potential anti-leukemia activity, are given to the patients. Among these cells, of particular interest are TCRgd1 T cells in view of their potent anti-tumor activity due to the expression of DNAM-1 and NKG2D activating NK receptors, in addition to the TCRgd, that recognizes phosphoantigens present on leukemia cells (118 120). The overall survival rate, also including cases with no NK alloreactivity, is now over 70% for both ALL and AML (>100 patients studied, Locatelli, manuscript in preparation). This further major improvement, particularly in AML, strongly suggests a TCRab1, CD191 - depleted graft as the first therapeutic option in children with high risk leukemias. In addition to the treatment of leukemias, this novel graft manipulation revealed to be highly successful in the cure of severe non-malignant disorders such as primary immunodeficiencies and other genetic diseases involving the haemopoietic system (121). Another important finding has been the demonstration that NK cell maturation from transplanted CD341 cells is greatly accelerated by CMV infection. In addition, in case of CMV infection/reactivation, a selective expansion of NKG2C1 cells has been detected (88). NKG2C1 NK cells are highly mature, KIR1 cells that may not only counter CMV infection, but also play a substantial role in the anti-leukemia effect since this is mediated by KIR1 mature NK cells. As discussed above, ILC play an important role in both innate defenses against pathogens, in lymphoid tissue organization/reconstitution and in tissue homeostasis (3). Accordingly, in view of the major tissue damages caused by the conditioning regimen, it is particularly important to define the process and the timing of ILC reconstitution after HSCT. In this context remarkable differences have been detected in the capability of CD341 cells from different sources in the generation of NK cells or ILC3 (122). MONOCLONAL ANTIBODIES TO CHECKPOINT INHIBITORS AND CAR-NK CELLS IN TUMOR IMMUNOTHERAPY In the field of tumor immunotherapy, the combination treatment with PD-1 blockers and mab to a second checkpoint receptor, CTLA-4, have proven effective only for some patients, suggesting the need for combining with other checkpoint blockers (90). In this context NK cells as well as subpopulations of T cells express a series of additional checkpoint inhibitors including KIR and NKG2A that may be used for immunotherapeutic intervention. For example, human antibodies specific for KIR2D (123) that have been previously proved to confer alloreactivity to NK cells against normal or leukemic targets (124,125) have been used alone or in combination with Retuximab as anti-lymphoma treatment thus illustrating the potential efficacy of combining a tumortargeting therapy with an NK-cell agonist (126). Preclinical studies showed improved autologous NK-mediated killing of multiple myeloma (MM) cells after mabmediated KIR blockade and phase I clinical trial data in patients with MM showed a good tolerability of the antibody (124). Remarkably, the same anti-kir mab was recently used in combination with anti-pd-1 to treat patients carrying solid tumors. Preliminary results in

110 DEL ZOTTO ET AL. patients with advanced platinum-refractory squamous cell carcinoma of head and neck (SCCHN) demonstrated a clinical benefit, with encouraging response rates that were deep and durable in some patients (NCT01714739). Another important novel example of immune checkpoint blocker is represented by an anti- NKG2A antibody that was originally isolated (127) to prevent the generation of inhibitory signals in NK cells exposed to HLA-E expressing targets. Most neoplastic cells express HLA-E, thus, in hematological cancers, anti- NKG2A antibodies could exploit the anti-leukemic activity mediated by the early wave of NKG2A 6 NK cells appearing 2 3 weeks after hematopoietic stem cell transplants. Alternatively this antibody could be used to pretreat NK cells in adoptive immunotherapy approaches from matched or mismatched family donors (after chemotherapy) in the treatment of acute leukemias (128). A phase I/II clinical trial with the anti- NKG2A mab in HNSCC patients is ongoing (NCT02331875) and the combination treatment with anti-nkg2a and anti-pd-1 antibodies has been also successfully analyzed in preclinical models in which an increased frequency of complete tumor regression was measured as compared to monotherapies with anti- NKG2A or anti-pd-1 alone (Sola et al. Meeting abstract at American Association for Cancer Research (AACR) 107th Annual Meeting 2016, DOI: 10.1158/1538 7445.AM2016-2342). Another promising approach of immunotherapy is the adoptive transfer of chimeric antigen receptor (CAR)-engineered NK cells. NK-mediated cytotoxicity can be efficiently and specifically redirected by recombinant CARs consisting of an extracellular single-chain variable fragment (scfv), which recognizes a cancer antigen and which is coupled to intracellular signaling domains that activate NK cytotoxicity upon antigen recognition. In contrast to the numerous pre-clinical studies and clinical trials regarding CAR-modified T cells, little is known about CARengineered NK cells (129). Pre-clinical data have been reported for CAR-modified primary human NK cells redirected against CD19 (130,131), CD20 (132), and CD244 (133), and for CAR-expressing NK-92 cells targeting a wider range of tumor antigens (134,135), but only few clinical studies using CAR-modified NK cells have started very recently (NCT00995137, NCT01974479, NCT02944162, NCT02892695, NCT02742727, NCT02839954). CONCLUDING REMARKS In conclusion, all these studies indicate how cells of the innate immunity particularly NK cells, may provide important tools to cure otherwise fatal malignancies. In this context, the generation of mabs against NK cell surface antigens allowed not only to identify a large array of receptors responsible for NK cell activation and inactivation, but also to define novel immune checkpoints (39) and functional NK subsets (136). The results of these studies could be successfully translated to clinical applications. Indeed, they allowed to define the NK alloreactivity and to identify alloreactive NK subsets. This possibility has a major impact in the identification of the best possible donor in haploidentical HSCT (111,137) and in the follow-up after HSCT. In addition, the development of anti-kir mabs and anti-nkg2a mabs offered an important tool for a novel therapeutic approach. Thus, stable mab-mediated KIR blocking confers alloreactivity to all KIR1 NK cells whereas all NKG2A 1 KIR - NK cells are rendered alloreactive after treatment with anti-nkg2a antibodies. A further progress will be the adoptive administration of (expanded) autologous NK cells treated with anti-kir or anti- NKG2A mabs. This approach may be applied also to solid tumors, possibly in combination therapies together with mab to checkpoint inhibitors such as anti-pd1 to minimize the effect of the inhibitory tumor microenvironment. Taken together, it is evident that these new therapies may represent important tools for major advances in the cure of cancer. LITERATURE CITED 1. Moretta A. Natural killer cells and dendritic cells: rendezvous in abused tissues. Nat Rev Immunol 2002;2:957 964. 2. Vivier E, Raulet DH, Moretta A, Caligiuri MA, Zitvogel L, Lanier LL, Yokoyama WM, Ugolini S. Innate or adaptive immunity? The example of natural killer cells. Science 2011;331:44 49. 3. Montaldo E, Vacca P, Vitale C, Moretta F, Locatelli F, Mingari MC, Moretta L. Human innate lymphoid cells. Immunol Lett 2016;179:2 8. 4. Cortez VS, Robinette ML, Colonna M. Innate lymphoid cells: new insights into function and development. Curr Opin Immunol 2015; 32:71 77. 5. Artis D, Spits H. The biology of innate lymphoid cells. Nature 2015; 517:293 301. 6. Vivier E, van de Pavert SA, Cooper MD, Belz GT. The evolution of innate lymphoid cells. Nat Immunol 2016;17:790 794. 7. Ferlazzo G, Tsang ML, Moretta L, Melioli G, Steinman RM, Munz C. Human dendritic cells activate resting natural killer (NK) cells and are recognized via the NKp30 receptor by activated NK cells. J Exp Med 2002;195:343 351. 8. Rosenberg EB, Herberman RB, Levine PH, Halterman RH, McCoy JL, Wunderlich JR. Lymphocyte cytotoxicity reactions to leukemia-associated antigens in identical twins. Int J Cancer 1972;9:648 658. 9. Karre K, Ljunggren HG, Piontek G, Kiessling R. Selective rejection of H-2-deficient lymphoma variants suggests alternative immune defence strategy. Nature 1986;319:675 678. 10. Ljunggren HG, Karre K. In search of the missing self : MHC molecules and NK cell recognition. Immunol Today 1990;11:237 244. 11. Moretta A, Bottino C, Vitale M, Pende D, Biassoni R, Mingari MC, Moretta L. Receptors for HLA class-i molecules in human natural killer cells. Annu Rev Immunol 1996;14:619 648. 12. Moretta A, Bottino C, Vitale M, Pende D, Cantoni C, Mingari MC, Biassoni R, Moretta L. Activating receptors and coreceptors involved in human natural killer cell-mediated cytolysis. Annu Rev Immunol 2001;19:197 223. 13. Moretta A, Tambussi G, Bottino C, Tripodi G, Merli A, Ciccone E, Pantaleo G, Moretta L. A novel surface antigen expressed by a subset of human CD3- CD161 natural killer cells. Role in cell activation and regulation of cytolytic function. J Exp Med 1990;171:695 714. 14. Moretta A, Bottino C, Pende D, Tripodi G, Tambussi G, Viale O, Orengo A, Barbaresi M, Merli A, Ciccone E, et al. Identification of four subsets of human CD3-CD161 natural killer (NK) cells by the expression of clonally distributed functional surface molecules: correlation between subset assignment of NK clones and ability to mediate specific alloantigen recognition. J Exp Med 1990;172:1589 1598. 15. Moretta A, Vitale M, Bottino C, Orengo AM, Morelli L, Augugliaro R, Barbaresi M, Ciccone E, Moretta L. P58 molecules as putative receptors for major histocompatibility complex (MHC) class I molecules in human natural killer (NK) cells. Anti-p58 antibodies reconstitute lysis of MHC class I-protected cells in NK clones displaying different specificities. J Exp Med 1993;178:597 604.

MARKERS AND FUNCTION OF HUMAN NK CELLS 111 16. Colonna M, Samaridis J. Cloning of immunoglobulin-superfamily members associated with HLA-C and HLA-B recognition by human natural killer cells. Science 1995;268:405 408. 17. Wagtmann N, Biassoni R, Cantoni C, Verdiani S, Malnati MS, Vitale M, Bottino C, Moretta L, Moretta A, Long EO. Molecular clones of the p58 NK cell receptor reveal immunoglobulin-related molecules with diversity in both the extra- and intracellular domains. Immunity 1995;2:439 449. 18. Braud VM, Allan DS, O Callaghan CA, Soderstrom K, D Andrea A, Ogg GS, Lazetic S, Young NT, Bell JI, Phillips JH, et al. HLA-E binds to natural killer cell receptors CD94/NKG2A, B and C. Nature 1998; 391:795 799. 19. Lopez-Botet M, Angulo A, Guma M. Natural killer cell receptors for major histocompatibility complex class I and related molecules in cytomegalovirus infection. Tissue antigens 2004;63:195 203. 20. Moretta A, Sivori S, Vitale M, Pende D, Morelli L, Augugliaro R, Bottino C, Moretta L. Existence of both inhibitory (p58) and activatory (p50) receptors for HLA-C molecules in human natural killer cells. J Exp Med 1995;182:875 884. 21. Biassoni R, Cantoni C, Falco M, Verdiani S, Bottino C, Vitale M, Conte R, Poggi A, Moretta A, Moretta L. The human leukocyte antigen (HLA)-C-specific activatory or inhibitory natural killer cell receptors display highly homologous extracellular domains but differ in their transmembrane and intracytoplasmic portions. J Exp Med 1996;183:645 650. 22. Sivori S, Vitale M, Morelli L, Sanseverino L, Augugliaro R, Bottino C, Moretta L, Moretta A. p46, a novel natural killer cell-specific surface molecule that mediates cell activation. J Exp Med 1997;186:1129 1136. 23. Pessino A, Sivori S, Bottino C, Malaspina A, Morelli L, Moretta L, Biassoni R, Moretta A. Molecular cloning of NKp46: a novel member of the immunoglobulin superfamily involved in triggering of natural cytotoxicity. J Exp Med 1998;188:953 960. 24. Vitale M, Bottino C, Sivori S, Sanseverino L, Castriconi R, Marcenaro E, Augugliaro R, Moretta L, Moretta A. NKp44, a novel triggering surface molecule specifically expressed by activated natural killer cells, is involved in non-major histocompatibility complex-restricted tumor cell lysis. J Exp Med 1998;187:2065 2072. 25. Cantoni C, Bottino C, Vitale M, Pessino A, Augugliaro R, Malaspina A, Parolini S, Moretta L, Moretta A, Biassoni R. NKp44, a triggering receptor involved in tumor cell lysis by activated human natural killer cells, is a novel member of the immunoglobulin superfamily. J Exp Med 1999;189:787 796. 26. Pende D, Parolini S, Pessino A, Sivori S, Augugliaro R, Morelli L, Marcenaro E, Accame L, Malaspina A, Biassoni R, et al. Identification and molecular characterization of NKp30, a novel triggering receptor involved in natural cytotoxicity mediated by human natural killer cells. J Exp Med 1999;190:1505 1516. 27. Sivori S, Parolini S, Falco M, Marcenaro E, Biassoni R, Bottino C, Moretta L, Moretta A. 2B4 functions as a co-receptor in human NK cell activation. Eur J Immunol 2000;30:787 793. 28. Bottino C, Falco M, Parolini S, Marcenaro E, Augugliaro R, Sivori S, Landi E, Biassoni R, Notarangelo LD, Moretta L, et al. NTB-A [correction of GNTB-A], a novel SH2D1A-associated surface molecule contributing to the inability of natural killer cells to kill Epstein-Barr virus-infected B cells in X-linked lymphoproliferative disease. J Exp Med 2001;194:235 246. 29. Falco M, Marcenaro E, Romeo E, Bellora F, Marras D, Vely F, Ferracci G, Moretta L, Moretta A, Bottino C. Homophilic interaction of NTBA, a member of the CD2 molecular family: induction of cytotoxicity and cytokine release in human NK cells. Eur J Immunol 2004;34:1663 1672. 30. Shibuya K, Lanier LL, Phillips JH, Ochs HD, Shimizu K, Nakayama E, Nakauchi H, Shibuya A. Physical and functional association of LFA-1 with DNAM-1 adhesion molecule. Immunity 1999;11:615 623. 31. Marcenaro E, Augugliaro R, Falco M, Castriconi R, Parolini S, Sivori S, Romeo E, Millo R, Moretta L, Bottino C, et al. CD59 is physically and functionally associated with natural cytotoxicity receptors and activates human NK cell-mediated cytotoxicity. Eur J Immunol 2003;33:3367 3376. 32. Vitale M, Falco M, Castriconi R, Parolini S, Zambello R, Semenzato G, Biassoni R, Bottino C, Moretta L, Moretta A. Identification of NKp80, a novel triggering molecule expressed by human NK cells. Eur J Immunol 2001;31:233 242. 33. Sivori S, Falco M, Della Chiesa M, Carlomagno S, Vitale M, Moretta L, Moretta A. CpG and double-stranded RNA trigger human NK cells by Toll-like receptors: induction of cytokine release and cytotoxicity against tumors and dendritic cells. Proc Natl Acad Sci U S A 2004; 101:10116 10121. 34. Marcenaro E, Ferranti B, Falco M, Moretta L, Moretta A. Human NK cells directly recognize Mycobacterium bovis via TLR2 and acquire the ability to kill monocyte-derived DC. Int Immunol 2008;20:1155 1167. 35. Sivori S, Falco M, Carlomagno S, Romeo E, Soldani C, Bensussan A, Viola A, Moretta L, Moretta A. A novel KIR-associated function: evidence that CpG DNA uptake and shuttling to early endosomes is mediated by KIR3DL2. Blood 2010;116:1637 1647. 36. Bauer S, Groh V, Wu J, Steinle A, Phillips JH, Lanier LL, Spies T. Activation of NK cells and T cells by NKG2D, a receptor for stressinducible MICA. Science 1999;285:727 729. 37. Brandt CS, Baratin M, Yi EC, Kennedy J, Gao Z, Fox B, Haldeman B, Ostrander CD, Kaifu T, Chabannon C, et al. The B7 family member B7-H6 is a tumor cell ligand for the activating natural killer cell receptor NKp30 in humans. J Exp Med 2009;206:1495 1503. 38. Bottino C, Castriconi R, Pende D, Rivera P, Nanni M, Carnemolla B, Cantoni C, Grassi J, Marcenaro S, Reymond N, et al. Identification of PVR (CD155) and Nectin-2 (CD112) as cell surface ligands for the human DNAM-1 (CD226) activating molecule. J Exp Med 2003; 198:557 567. 39. Moretta L, Bottino C, Pende D, Vitale M, Mingari MC, Moretta A. Different checkpoints in human NK-cell activation. Trends Immunol 2004;25:670 676. 40. Mingari MC, Vitale C, Cambiaggi A, Schiavetti F, Melioli G, Ferrini S, Poggi A. Cytolytic T lymphocytes displaying natural killer (NK)-like activity: expression of NK-related functional receptors for HLA class I molecules (p58 and CD94) and inhibitory effect on the TCR-mediated target cell lysis or lymphokine production. Int Immunol 1995; 7:697 703. 41. Mingari MC, Schiavetti F, Ponte M, Vitale C, Maggi E, Romagnani S, Demarest J, Pantaleo G, Fauci AS, Moretta L. Human CD81 T lymphocyte subsets that express HLA class I-specific inhibitory receptors represent oligoclonally or monoclonally expanded cell populations. Proc Natl Acad Sci U S A 1996;93:12433 12438. 42. Pietra G, Romagnani C, Falco M, Vitale M, Castriconi R, Pende D, Millo E, Anfossi S, Biassoni R, Moretta L, et al. The analysis of the natural killer-like activity of human cytolytic T lymphocytes revealed HLA-E as a novel target for TCR alpha/beta-mediated recognition. Eur J Immunol 2001;31:3687 3693. 43. Moretta L, Romagnani C, Pietra G, Moretta A, Mingari MC. NK-CTLs, a novel HLA-E-restricted T-cell subset. Trends Immunol 2003;24: 136 143. 44. Romagnani C, Pietra G, Falco M, Millo E, Mazzarino P, Biassoni R, Moretta A, Moretta L, Mingari MC. Identification of HLA-E-specific alloreactive T lymphocytes: a cell subset that undergoes preferential expansion in mixed lymphocyte culture and displays a broad cytolytic activity against allogeneic cells. Proc Natl Acad Sci U S A 2002; 99:11328 11333. 45. Pietra G, Romagnani C, Mazzarino P, Falco M, Millo E, Moretta A, Moretta L, Mingari MC. HLA-E-restricted recognition of cytomegalovirus-derived peptides by human CD81 cytolytic T lymphocytes. Proc Natl Acad Sci U S A 2003;100:10896 10901. 46. Luetke-Eversloh M, Killig M, Romagnani C. Signatures of human NK cell development and terminal differentiation. Front Immunol 2013; 4:499. 47. Klose CS, Flach M, Mohle L, Rogell L, Hoyler T, Ebert K, Fabiunke C, Pfeifer D, Sexl V, Fonseca-Pereira D, et al. Differentiation of type 1 ILCs from a common progenitor to all helper-like innate lymphoid cell lineages. Cell 2014;157:340 356. 48. Spits H, Bernink JH, Lanier L. NK cells and type 1 innate lymphoid cells: partners in host defense. Nat Immunol 2016;17:758 764. 49. Spits H, Artis D, Colonna M, Diefenbach A, Di Santo JP, Eberl G, Koyasu S, Locksley RM, McKenzie AN, Mebius RE, et al. Innate lymphoid cells a proposal for uniform nomenclature. Nat Rev Immunol 2013;13:145 149. 50. Montaldo E, Juelke K, Romagnani C. Group 3 innate lymphoid cells (ILC3s): Origin, differentiation, and plasticity in humans and mice. Eur J Immunol 2015;45:2171 2182. 51. Montaldo E, Vitale C, Cottalasso F, Conte R, Glatzer T, Ambrosini P, Moretta L, Mingari MC. Human NK cells at early stages of differentiation produce CXCL8 and express CD161 molecule that functions as an activating receptor. Blood 2012;119:3987 3996. 52. Sivori S, Falco M, Marcenaro E, Parolini S, Biassoni R, Bottino C, Moretta L, Moretta A. Early expression of triggering receptors and regulatory role of 2B4 in human natural killer cell precursors undergoing in vitro differentiation. Proc Natl Acad Sci U S A 2002;99: 4526 4531.