MCMV: Murine Cytomegalovirus; OPV: Orthopoxvirus; ECTV: Ectromelia Virus; HIV: Human Immunodeficiency Virus; HCV: Hepatitis C Virus; Mtb: Mycobacterium tuberculosis; ITIM: Immunoreceptor Tyrosine-based Inhibitory Motifs; KIR: Killer cell Immunologobulin-like Receptors; ITAM: Immunoreceptor Tyrosine-based Activation Motif

Natural killer (NK) cells are an important component of innate immunity, and serve as a crucial first line of defense against a diverse range of pathogens, such as viruses, bacteria and parasites [1-6]. Derived from the bone marrow, NK cells have wide tissue distribution, with high frequencies of mature NK cells in lung, liver, blood and spleen. They are generally found at low frequencies in lymph nodes and mostly with immature phenotype, however, following local stimulation, mature NK cells can accumulate in these sites [7,8]. NK cell responses are regulated by a balance of activating and inhibitory signals transmitted by cell surface receptors. Unlike B- and T-cell antigen receptors, NK cell receptors are encoded in the germ-line and do not undergo somatic recombination. Extensive work has been done to delineate the responses and functions of NK cells during pathogen infections. The best knowledge of NK cell function in the control of viral infections in vivo came from studies of murine cytomegalovirus (MCMV) infection. In the resistant C57BL/6 (B6) mice, NK cell activating receptor Ly49H recognizes m157-a viral encoded protein expressed on infected cell surface, which triggers the cytolytic function of NK cells [9,10]. The exact molecular mechanisms of NK cells during other viral infections are less clear. Further, much less is known about the responses and functions of NK cells during bacterial and parasitic infections. Moreover, NK cells are a heterogeneous population composed of cells that express overlapping subsets of activating and inhibitory receptors, studies have shown that the effector mechanisms that NK cells use may differ in different organs during pathogen infections [11-13]. Furthermore, accumulating evidences show that NK cells are a crucial link between innate and adaptive immune system via cytokine release and cell-cell interaction, which participate in the shaping of adaptive immune response [3,7,14]. This review highlights recent studies on the role of NK cells in the control of viruses, bacteria and parasites, and underscores the important role of NK cells playing in the defense against infectious diseases.

NK cell function is regulated by inhibitory and activating receptors [15,16]. NK cell inhibitory receptors recognize a wide variety of selfmolecules. Upon ligand binding, NK receptors having immunoreceptor tyrosine-based inhibitory motifs (ITIMs) in their cytoplasmic domain prevent NK cell effector function. The Ly49 inhibitory receptors in rodents and the inhibitory killer cell immunologobulin-like receptors (KIR) in primates recognize the MHC class I molecules, the CD94- NKG2A heterodimers in both rodents and primates recognize HLA-E (Qa-1b in mice), a nonclassical MHC class I molecule which binds peptides derived from leader segments of other class I proteins [17,18].

Activating receptors on NK cells can be broadly divided into those that recognize MHC-class-I-like ligands (human KIRs, murine Ly49 families, and CD94-NKG2s) and those that do not (NKG2D, 2B4, NKp30, NKp44 and NKp46) [19]. The activating receptors of the human KIR and murine Ly49 families are highly homologous to their inhibitory counterparts but have truncated cytoplasmic domains lacking ITIMs. These molecules associate with immunoreceptor tyrosine-based activation motif (ITAM)-bearing adapter molecules such as DAP12 and FεRIγ [20,21]. Other activating receptors, including NKG2D, associate with DAP10, an adapter protein containing an activating motif YXXM [22]. In rodent, an NKG2D isoform generated by alternative splicing can also associate with the DAP12 adapter protein in activated mouse NK cells [23,24].

The inhibitory NK receptors (mainly KIR, Ly49, CD94/NKG2A) recognizing self-MHC class I play a predominant role in NK cell tolerance to self. The ”missing self hypothesis” states that NK cells scrutinize target cell MHC expression and only respond to target cells lacking the expression of MHC I [25]. The inhibitory receptor/MHC I interaction allows them to gain full functionality by directly activating or modifying other signals for developing functional NK cells. Recent works support that NK cell education operates as a rheostat, tuned by a quantitative influence on individual NK cells by the MHC class I alleles present in vivo [26]. The inhibitory input that an individual NK cell receives during education quantitatively tunes the responsiveness of individual NK cells. The higher the inhibitory input during education, the more likely it is that it will pass the threshold to respond (i.e. degranulation and IFN-γ expression) under given conditions of stimulation [26]. The role of activating receptors in human NK cell education has also been discovered recently. Fauriat C et al. found that NK cells expressing KIR2DS1 are hyporesponsive against cellular targets from donors homozygous for HLA-C2, the ligand of KIR2DS1. These results suggest that the education of human NK cells via activating receptors is a mechanism to secure tolerance that complements education via inhibitory receptors [20].

The effector function of NK cells is regulated by the signaling balance of activating and inhibitory receptors. When activating receptor/ligand interactions predominate over inhibitory receptor/ligand signals, NK cells are activated and target cells would be lysed. On the other hand, NK cells function is inhibited when activating receptor/ligand signals are weaker than inhibitory receptor/ligand signals [27-29].

NK cell effector functions are stimulated through direct contact with activated dendritic cells (DCs) in vivo [30]. In fact, DC/NK-cell interaction is bi-directional and complex, as it could result not only NK cell activation but also in DC maturation or apoptosis, depending on the activation status of both players [31]. Upon stimulation by various pathogens or by TLR ligands, DCs secrete several cytokines such as IL- 12, IL-18, IL-15 and type I interferons (IFN-I) which trigger NK cell effector function. DC-derived IL-12 stimulates IFN-γ production by NK cells in different systems [31-33]. Il-18 is known to synergize with IL-12 to induce IFN-γ production by NK cells [31]. DCs use IL-15Rα to present IL-15 in trans to NK cells, which is important to trigger NK cell proliferation [34-36]. Lastly, NK cell activity is controlled by IFN-I at various levels. NK cell function cannot be elicited in IFNAR-deficient mice after viral or bacterial infections or injection of TLR ligands [36,37]. IFN-I can be produced by all DC subsets. More recently, studies from Diefenbach’s group have shown that system failure of mononuclear phagocytes such as DCs to produce IFN-I in response to microbial stimulation in germ-free mice results in defective NK cell responses [38]. In addition, optimal NK cell activation by DCs also requires direct cell-cell contacts [30,32]. On the other hand, activated NK cells can induce DC maturation, while immature DCs are susceptible to NK cell mediated cytolysis while mature DCs are protected [31,39]. The NK cell activating receptor NKp30 appears to play a central role in DC maturation or apoptosis induced by NK cells [33,40,41].

NK cells play an important role in the control of some, but not all, mouse and human viruses. NK cells can respond to infection either directly or indirectly. They respond directly by recognizing virus-infected cells, and indirectly by interacting with DCs, which express Toll-like receptors (TLRs) and secrete cytokines in response to encounter with pathogens. NK cells use two main effector mechanisms to control viral infections: the secretion of IFN-γ and direct lysis of infected cells through the granule-exocytosis pathway.

NK cells are critical for control of the mouse pathogen MCMV [42-44]. In resistant B6 mice, NK cells recognize MCMV-infected cells mainly by the activating receptor Ly49H [16]. MCMV encodes an MHC class I-like glycoprotein m157, which is expressed on the surface of infected cells and is directly ligated by Ly49H [9,10]. Upon recognition of MCMV-infected cells via Ly49H, NK cells secrete cytokines, such as IFN-γ and TNF-α, and kill infected cells by the release of lytic granules that contain perforin and granzymes. Moreover, during a later phase of infection, Ly49H⁺ NK cells selectively proliferate, expand from 50% of all NK cells to 90% of all NK cells [45]. Mutant virus with m157 deletion abolishes Ly49H⁺ NK cell activation after viral infection, gains virulence in B6 mice [46]. Further, Ly49H deficient mice in B6 background are susceptible to MCMV infection, while remain resistant to Leishmania major and ectromelia virus infection [47]. Mouse strains are either resistant or susceptible to MCMV infection. BALB/c mice and FVB/N mice that do not express Ly49H on NK cell surface are susceptible to MCMV infection. Ly49HB6 bacterial artificial chromosome transgenes transfer virus resistance in BALB/c and FVB/N mice, and the resistance is correlated with the protein expression level of Ly49H on NK cells [48]. These studies demonstrate Ly49H as a pivotal factor in resistance to MCMV.

New Zealand White (NZW) mice are resistant to MCMV infection, and the resistance is also dependent on NK cells as NK cell depletion results in increased virus titers in multiple organs [49]. New Zealand Black (NZB) mice are susceptible to MCMV infection even though NZW and NZB NK cells display comparable levels of Ly49-related (anti-Ly49H mAb 3D10+) receptors. However, the NK cell mediatedresistance in NZW mice is independent of Ly49H and m157 because NZW NK cells Ly49H and other receptors do not bind MCMVencoded m157. Backcross NZW mice with MCMV-sensitive NZB mouse strains show resistance is mediated by multiple gene products, which remain to be identified [49]. These results indicate that NK cell control of viral infection in NZW genetic system differ from Ly49Hdependent, dominant control previously characterized for the B6 and BALB/c genetic system.

NK cells are also required for the resistance to MCMV in Ma/ My mice, Ma/My mice lack the Ly49H receptor, another activating receptor Ly49P encoded on the NK cells in Ma/My mice recognizes the MHC class I molecule H2-Dk in complex with the viral protein m04 [50]. Different from Ly49H in B6 mice that directly recognize m157 protein encoded by MCMV, Ly49P recognizes MCMV-infected cells requiring H2-Dk peptide-binding platform. These findings reveal the importance of activating NK cell receptor-MHC class I-peptide complex interactions in the recognition of viral infections.

Another MCMV-susceptible mouse strain 129/J mice lack Ly49H but express two highly related molecules: an inhibitory Ly49I receptor and an activating Ly49U [51]. MCMV m157 binds to the inhibitory Ly49I receptor, preventing NK cell control of MCMV [9]. MCMV encoded m157 may have evolved to provide a selective advantage for the virus by engaging an inhibitory NK cell receptor to prevent NK cell activation. It is intriguing to note that the Ly49I alleles from B6 and 129/J mice both bind selectively to H-2K^d, whereas the Ly49I receptor in B6 mice fails to bind m157. Meanwhile, the activating Ly49H receptor does not recognize any known H-2 ligand, but binds to m157 with high affinity. The observations that m157 binds to both an inhibitory and a highly homologous activating Ly49 receptor suggest the possibility that Ly49H evolved from the inhibitory receptor as a consequence of selection by the pathogen [9,52].

Another NK cell activating receptor NKG2D also plays an important role in the NK cell mediated-protection during MCMV infection. NKG2D interacts with three different cellular ligands: the retinoic acid early inducible (RAE)-1 family of proteins [53]; the minor histocompatibility antigen H60 [54,55], and the murine UL16-binding protein-like transcript (MULT)-1 glycoprotein [56,57], all of which are regulated by MCMV. The MCMV m145-encoded glocoprotein regulates MULT-1 by preventing plasma membrane residence of MULT-1 [58]. The MCMV m155 protein specifically down-regulates H60 without affecting expression of RAE-1 and MULT-1, an MCMV mutant virus lacking m155 is severely attenuated in BALB/c mice [59]. The MCMV m152-encoded gp40 glycoprotein specifically down-regulates the cell surface expression of all RAE-1 proteins, an m152 deletion mutant virus is less virulent in vivo compared with the wild-type virus [60,61].

The ability to generate antigen-specific memory responses is classically regarded as a hallmark of the cells of the adaptive immune system. However, recent studies have shown the NK cells also have adaptive immune features–can be long-lived and generate enhanced responses on secondary encounter with antigen [62-66]. During MCMV infection in B6 mice, Ly49H⁺ NK cells preferentially proliferate. Following viral clearance, those effector Ly49H⁺ NK cells undergo a contraction phase resulting in a long-lived memory NK cell pool. These memory NK cells express higher level of Ly49H, KLRG1, CD43 and Ly6C, and a decreased expression of CD27 which indicate that they are more mature than naïve NK cells. More importantly, when adoptively transferred into MCMV-susceptible newborn mice, the memory NK cells significantly protect the newborn mice compared with an equivalent number of naïve NK cells [66]. The memory NK cells generated during MCMV infection clearly demonstrate protective properties which can be parallel with that of the memory CD8 T cells. Understanding the mechanisms and the properties of NK cell memory during viral infections and vaccinations will have major implications on our approach to vaccination strategies for the generation of immunological memory against infectious pathogens.

Mousepox is another viral disease where an essential role of NK cells in the protection is well established. Mousepox is caused by the mouse Orthopoxvirus (OPV) ectromelia virus (ECTV). Mouse strains can be generally divided into two groups following ECTV infection. Resistant mouse strains such as B6 and 129/J; susceptible mouse strains including DBA/2, DBA/2J, A/J, and BALB/c. To understand the genetic mechanisms involved in mousepox resistance, Brownstein et al. infected a panel of B6 Χ DBA2/J recombinant inbred strains of mice and mapped four genes involved in the B6 mice resistance to mousepox (Rmp1–Rmp4) [67]. Rmp1 resides in the natural killer (NK) complex (NKC) on chromosome 6 [68]. In recent years, the mechanisms of NK cell-mediated resistance to mousepox are beginning to be uncovered. In resistant B6 mice, the resistance to mousepox requires the direct cytolytic function of NK cells, as well as their ability to boost the adaptive T cell responses [7]. Further, we identified NK cell activating receptor CD94-NKG2E specifically recognize ECTV-infected cells via the targets expressing a specific peptide(s) (either a viral peptide or a peptide derived from an infection-induced cellular protein) bound to the non classical MHC class I Qa-1^b molecular. Another NK cell activating receptor NKG2D acts as a co-stimulator to boost optimal NK cell cytolytic function [69]. However, the peptide(s) bound to the Qa-1^b molecular remain unidentified. Importantly, different from the recognition of MCMV by Ly49-activating receptors, the CD94-NKG2 system is exquisitely conserved between rodents and primates [70], the involvement of activating CD94-NKG2 receptors during other viral infections deserves further in depth investigations.

Another important observation in the mousepox model is that the resistant B6 mice gradually lose their natural resistance to mousepox as they age. Surprisingly, the main reason for the loss of resistance is not of intrinsically defective T cell responses. Instead, the primary reason is because of defects in NK cell responses in the aged mice. In the aged mice, the percentage of total and mature NK cells decreases in blood and spleen, after ECTV infection, the aged NK cells also have defects to migrate to the local draining lymph node to control systemic virus spread, which results in increased early virus replication and spread as well as high mortality rate in the aged mice [8]. These results highlight the importance of NK cells in the control of age-related infectious diseases [71]. Further studies are needed to fully identify NK cell functional changes during physiological ageing process and investigate the mechanisms underlying these changes.

Extensive evidence supports an important role for NK cells during influenza virus (flu) infection. Using anti-asialo GM1 or anti-NK1.1 antibody PK136 to deplete NK cells leads to increased morbidity and mortality in mice with flu infection [72,73]. During the early days of flu infection, NK cells in the lungs become activated and demonstrate high cytotoxic activity followed by production of anti-viral cytokine IFN-γ [72,74]. Furthermore, NKp46, an activating NK cell receptor that recognizes flu haemagglutinin (HA) is critical for NK cell mediatedprotection against lethal flu infection in mice [75,76].

In contrast, a recent study reported that increased survival in the NK cell depleted wild-type mice or IL-15^-/- mice which lack NK cells due to deficiency of IL-15, a cytokine important for NK cell maintenance was associated with significantly lower lung lesions as well as reduced pulmonary inflammation following lethal flu infection [77]. This study suggested that in some settings, NK cells might contribute to the pathogenesis of flu infection. The discrepancy of the results might be due to a difference in the challenge virus dose, mouse strains used in those studies. Moreover, even though the authors also used anti-asialo GM1 antibody which will not deplete NKT cells to exclude the role of NKT cells as anti-NK1.1 antibody also deplete NKT cells in B6 mice [77], further in depth studies need to be done to explore the role of NKT cells in the pulmonary inflammation during flu infection because NKT cells numbers are also greatly reduced in IL-15^-/- mice [78] and NKT cells have been shown to exacerbate pulmonary inflammation [79-81]. Nonetheless, NK cells clearly play an important role during flu infection. Further studies are required to fully dissect the function of NK cells in the defense against flu infection.

Human peripheral NK cells comprise 5-20% of peripheral blood mononuclear cells and can be divided into two distinct subsets characterized by their relative expression of the cell surface markers CD56 and CD16 [82,83]. The CD56^brightCD16- subset, which constitutes less than 10% of peripheral-blood NK cells, has robust cytokines production as IFN-γ, plays an important role in regulating the adaptive immune function but has weaker cytotoxicity compared to CD56^dimCD16⁺ subset, which is the terminal differentiated population with high cytolytic function [84,85]. In addition to the conventional NK cells, a population of CD56^-CD16⁺ NK cells have been found in human immunodeficiency virus (HIV) or hepatitis C virus (HCV) infected patients, which express a similar receptor profile to CD56^low NK cell, but have impaired cytolytic function and poor cytokine production upon target cells recognition [86-89].

NK cells are directly involved in HCV infection. In chronic HCV patients, the frequency of circulating NK cells is reduced compared with healthy controls, and is accompanied with subset skewing, with the enrichment of CD56bright subset and a terminally differentiated hypofunctional CD56^-CD16⁺ NK cell subset [90-95]. The dysfunctional CD56^- NK cells expanded during chronic HCV infection is associated with an impaired ability to respond to antiviral treatment with IFN-α and ribavirin [95]. Jinushi M et al. have demonstrated that NK cells from chronic HCV-infected donors (HCV-NK) were incapable of activating DCs, HCV-NK also showed higher expression of CD94/NKG2A and produced IL-10 and TGF-β when cultured with hepatic cells expressing HLA-E, a ligand for CD94/NKG2A [96]. Further, studies suggest that particular combinations of NK cell inhibitory receptor KIR2DL3 and its ligand HLA-C1 are associated with the clearance of HCV infection [97]. Moreover, lower frequencies of NKp30⁺NKp46⁺, CD161⁺, and NKG2D⁺ NK cells have been recently associated with the ability to clear HCV following acute infection [98]. Oliviero B et al. observed that different NK cell phenotypic and functional features were associated with different treatment outcome in chronic HCV patients [99]. In the patients achieved sustained virological response (SVR) with treatment of pegylated interferon α/ribavirin, their NK cells displayed higher perforin content, lower CD16 expression and higher proportion of CD56^dim/CD16^- cells compared with NK cells from non-responder. Moreover, SVR patients displayed higher natural and antibodydependent NK cell cytotoxicity [99]. These results indicate that early NK cell activation and robust cytolytic function are important in the rapid IFN-α induced elimination of HCV-infected hepatocytes [99].

NK cells also involve in the control of HIV infection, through several pathways such as through secreting perforin and granzyme B to kill target cells; the Fas–FasL pathway to induce apoptosis of infected cells; production of cytokines to regulate immunity; and antibodydependent cell-mediated cytotoxicity (ADCC) to lyse infected cells [100]. NK cell surface receptors profile is associated with different clinical conditions in HIV patients. In patients with late stage HIV infection, their NK cells are found to contain fewer cytotoxic NK cells with higher expression of inhibitory receptor NKG2A [101]. Furthermore, certain HLA-B antigens have been associated with lack of progression to Acquired Immune Deficiency Syndrome (AIDS) after HIV infection [102]. Subsequent studies observed a protective effect of the combined presence of KIR3DS1 and HLA-Bw4-I80 alleles in patients with chronic HIV-1 infection [103,104]. Further analysis revealed that an increased NK cell KIR3DS1 count generated by copy number variants of KIR3DL1/DS1 was associated with a lower viral set point in the presence of their appropriate ligands. These results suggest that the relative amounts of these activating and inhibitory KIRs play a role in regulating the peripheral expansion of highly antiviral KIR3DS1⁺ NK cells, which may influence the disease progress following HIV^- 1 infection [105]. Recently, Apps R et al. have shown that increasing HLA-C expression was associated with protection against multiple outcomes in HIV infection, and the protective mechanisms conferred by higher HLA-C expression level occurs partly through an enhanced CTL-mediated immune response [106]. However, it remains to be determined whether higher HLA-C expression also affects NK cell responses in HIV infection.

The importance of NK cells in protection against bacterial infection has been controversial. It might depend on different infection dose, infection site and also the type of inflammatory response elicited by different pathogens. NK cells respond to bacterial infection by direct lysing of infected host cells or secreting cytokines to promote inflammation. In addition, NK cell derived IFN-γ can activate macrophages to destroy bacteria. Similarly, macrophages are stimulated to produce NK cell activating cytokines IL-12 and TNF-α when exposed to bacteria, and in turn activate the NK cells to kill these target pathogens [107-109].

In B6 mice infected with Mycobacterium tuberculosis (Mtb), the causative agent of human tuberculosis (TB), activated NK cells increase in the lung, and are capable of producing IFN-γ and perforin. However, in vivo depletion of NK cells has no influence on bacterial load within the lung [6]. These findings indicate that NK cells are activated during the early response to pulmonary TB and are a source of IFN-γ, even though removal of NK cells does no substantially alter the host control of bacterial burden, NK cells might reduce immunopathology or favor development of protective immune responses. In support of this, Feng C et al. found that NK cell-derived IFN-γ differentially regulated T-independent resistance and granulocyte function in Mtb infection [110]. Human NK cells lyse Mtb-infected monocytes and alveolar macrophages through the NKp46 receptor and NKG2D [111,112]. More recently, when exposed to autologous monocytes and gamma-irradiated Mtb, human NK cells are found to produce IL-22, which inhibits intracellular mycobacterial growth by enhancing phagolysosomal fusion [113]. Further, NK cells lyse Mtb-expanded CD4⁺ regulatory T cells (Tregs) in vitro [114]. Using B6 mice model first vaccinated with Bacillus Calmette-Guerin (BCG), followed by challenge with virulent Mtb, Dhiman R et al. showed that NK cells contribute to the efficacy of vaccination against Mtb infection by producing Il-22 and also by direct lysis of inhibitory Tregs that expanded after BCG immunization, induces optimal protective immunity through enhancing Ag-specific T cell responses after challenge with Mtb [14].

Innate immune responses are rapidly triggered upon Listeria monocytogenes (LM) infection and are essential for host survival [115,116]. NK cells represent an important source of IFN-γ at the infectious loci containing infected microphage and inducible nitric oxide synthase-producing DCs (TipDCs) [117-119]. However, NK might not involve in the clearance of LM [120]. As suggested in recent studies, CD8 T cells, rather than NK cells, play a more important role in immunity against LM [121]. In addition, type I IFN is found to hinder host control of LM [122-124]. NK cells are activated by type I IFN, so it is possible that NK cells worsen disease rather than protect from LM infection.

NK cells are also involved in control of other bacterial infection. Mice deficient in the RAG2 and common γ-chain genes (RAG^-/-γc^-/-) lack B, T and NK cells and serve as a good model to evaluate the function of these cells in immune defense against bacterial infection. RAG^-/-γc^-/- mice infected with Shigella flexneri have higher bacterial titers and compromised survival compared with RAG^-/- mice (lack B and T cells, but have NK cells), suggesting that IFN-γ produced by NK might limit infection [125]. IL-15^-/- mice deficient in NK cells are highly susceptible to pulmonary staphylococcal infection. In addition, WT mice depleted of NK cells are similarly susceptible to staphylococcal infection suggest a critical role for NK cells in host defense against pulmonary extracellular bacterial infection [126].

Malaria is a serious infectious disease, caused by infection with Plasmodium parasites. Immunity against P. falciparum, the most dangerous malaria-provoking agent, develops with age over the course of multiple infections and the humoral immunity is crucial for protection from severe disease. Interactions between NK cells, infected erythrocytes (iRBC) and other immune cells lead to specific NK cell responses to P. falciparum [6,127]. NK cells derived from malaria-naïve or infected individuals are shown having cytolytic activity against P. falciparum-iRBCs that is possibly mediated by Fas and Granzyme B [128,129]. Mavoungou E, et al. have shown that the interaction of Duffy binding-like (DBL)-1 alpha domain of P. falciparum-infected erythrocyte membrane protein-1 (PfEMP-1) and NK natural cytotoxicity receptor (NCR) NKP30 is the key recognition mechanism leading to parasite killing by NK cells [130]. Recently, another study indicates that presence of Hsp70 and absence of HLA-E on the membrane of iRBC prompt NK cell cytotoxicity toward infected host cells [131].

In addition to their up-regulation of CD69 and CD25 activation markers after contact with RBC infected by several different strains of P. falciparum, NK cells are one of the first cells to produce IFN-γ in response to P. falciparum infection [6,127]. IFN-γ production by human NK cells cultured with iRBCs is dependent on IL-12 and IL-18 derived from myeloid cells in cultures [6]. Moreover, the magnitude of IFN-γ released by NK cells is known to be heterogeneous among individuals, possibly influencing susceptibility to disease [127]. Cerebral malaria is a major, life-threatening complication of P. falciparum malaria, and has a high mortality rate [132]. IFN-γ is known to be potentially involved in the pathogenesis of cerebral malaria [133-135]. NK cell depletion resulted in significant protection against cerebral malaria, suggesting involvement of NK cell activity in its pathogenesis [136].

Recently, associations of killer immunoglobulin-like receptors (KIRs) on NK cells with the incidence of malaria pathogenesis have been characterized. Comparison of KIRs between Plasmodium-positive and Plasmodium-negative Melanesian individuals in the Solomon Islands revealed the trend of parasitic positive individuals to be KIR3DL1/ KIR3DS1 heterozygous in pair with KIR2DS4 nondeleted variants in a set of KIR genes in heritable as the AB genotypes [137]. Hirayasu K, et al. found that the combination of KIR2DL3 and its cognate HLA-C1 ligand was significantly associated with the development of cerebral malaria when compared with non-cerebral malaria, which indicates that the NK cell repertoire shaped by the KIR2DL3 -HLA-C1 interaction shows certain functional responses that facilitate development of cerebral malaria [138].

The eventual outcome of a specific pathogen infection is determined by the battle between host and invading pathogen. Some viruses have evolved strategies to evade NK cell mediated–immune surveillance (Figure 1).

Figure 1: Strategies of viruses to evade NK cell-mediated immune surveillance. A) Upon recognition of virus-infected cells by activating NK cell receptors, NK cells secrete cytokines and kill infected cells to control viral infection. However, viruses have evolved strategies to evade NK cell immune surveillance, such as listed in B-E. B) Viruses encode their own MHC class I homolog to inhibit NK cell function through binding to NK cell inhibitory receptors. C) Viruses encode anti-inflammatory cytokine IL-10 mimicry to suppress NK cell function. D) Viruses encode soluble ligands binding to NK cell activating receptor NKG2D to block NK cell recognition of infected cells. E) Viruses directly infect NK cells resulting impaired NK cell function and NK cell apoptosis.

NK cell activation is inhibited by inhibitory signals provided through interaction of receptors on NK cells with self-MHC class I products. Some viruses encode decoy proteins for NK cell inhibitory receptors. Both human and mouse CMV encode their own MHC class I homolog UL18 [139] and m144 [140], respectively, to inhibit NK cell activity through binding to NK inhibitory receptors. In support of this, NK cells are more effective at restricting replication of a recombinant MCMV with m144 deletion than of wild-type MCMV [140].

Cytokines are potent immune response regulators, which play an important role in the outcome during pathogen infection. Several viruses encode cytokine mimicry to escape host immune surveillance. For example, Epstein-Barr virus (EBV) gene BCRF1 encodes a protein homologous to the cellular cytokine IL-10 (vIL-10) with antiinflammatory properties [141,142]; recent studies show that vIL-10 protects EVB-infected B cells from NK cell mediated elimination [143]. Besides EBV, primate cytomegaloviruses (CMVs), Orf poxvirus, and equine herpes virus type 2 (EHV-2) also encode homologues of human IL-10 [144-146].

It is a universal strategy for viruses to escape NK cell recognition by encoding viral proteins to disturb NK cell receptor-ligand interaction. During some viral infections, viruses escape NK-mediated immunosurveillance by interfering NK cells activation and increased shedding of soluble ligands. NKG2D usually is a target of virus for escaping NK cell immune recognition [147]. NKG2D ligands may be either induced or down modulated upon infection by several pathogens [148]. The viral protein UL16 secreted from HCMV binds to ULBP-1, ULBP-2, and MICB, which are members of NKG2D-ligand family, and blocks the activation of NK cells [149,150]. Zoonotic orthopoxviruses encode a protein that is secreted by infected cells and competitively binds to NKG2D, thus blocking the recognition of infected cells by NK cells [151]. HIV-1 escapes NGG2D-mediated immunosurveillance by modulation NKG2D ligands (NKG2DLs) expression and increased shedding of soluble NKG2DLs [152].

One of the viral immune-evasion tactics is to infect and kill immune cells, including NK cells. Influenza virus can directly infect NK cells, which results in the impairments of NK cell function and also induces marked apoptosis of NK cells [153,154]. A few viruses, for example, HIV, herpes simplex virus, and ECTV also possess the ability to directly infect NK cells [155-157].

Significant advances have been recently made in understanding the functions of NK cells in innate defense against infective pathogens. NK cell responses have the potential to mediate defense through both direct antiviral and immunoregulatory pathways. NK cell responses are regulated by a balance of activating and inhibitory signals transmitted by cell surface receptors. Other factors such as type I interferon, crosstalk of NK cells with DC and macrophages, and genetic background might also influence NK cell responses during certain pathogen infections. In the battle between host and invading pathogens, virus has also developed several strategies for evading NK cell elimination. Thus, how to reverse the immune suppression of NK cells and boost NK cell activity are crucial for fully usage of NK cells for prevention and treatment of emerging and severe pathogen infections. To achieve this goal, a better molecular definition of the mechanisms by which NK cells participate in innate immunity against these pathogens is required.