Department of Hepatology, Division of Medicine, Imperial College London, UK
Address correspondence to: Division of Medicine Imperial College London St Mary's Hospital Campus South Wharf Road London W2 1PG Tel: 44 207 594 9024 Fax: 44 207 724 9369 E-mail: skhakoo@imperial.ac.uk
February 2011
Summary
Natural Killer cell functions are regulated by combinations of activating and inhibitory receptors, derived from a number of different gene families. This review focuses on receptors for MHC class I, which include the killer cell immunoglobulin-like receptors (KIR) and the CD94:NKG2 family of receptors. In particular the KIR are diverse and rapidly co-evolving with their classical MHC class I ligands. Thus NK cells are part of the innate immune system that are continuing to adapt to the challenges of pathogens.
Introduction
NK cells are an important component of the innate immune system, which participate in the early immune defence against intracellular pathogens and tumour transformation. They were originally defined by their ability to spontaneously eliminate rare cells lacking expression of class I Major Histocompatibility Complex (MHC class I) self molecules, a process commonly referred to as "missing self" recognition (Biron et al., 1999; Purdy and Campbell, 2009).
Upon activation, NK cells can mediate direct cytotoxicity or secrete cytokines and chemokine that modulate subsequent steps in the adaptive immune response. These functions are regulated by the combination of signals from activating and inhibitory receptors (Lanier, 1998). The MHC class I receptors are particularly important for NK cells to discriminate "self" (healthy cells) from "altered-self" (infected- and transformed-cells) or "missing self". MHC class I receptor:ligand interactions can induce inhibitory signals that counteract activating receptor signals and lead to NK cell inhibition. In contrast down-regulation or loss of MHC class I expression, during viral infection or carcinogenesis, shifts the balance towards NK cell activation and target cell destruction by removing this inhibitory signal. Thus in health NK cells are tolerant towards host cells, but in disease this tolerance can be readily broken.
MHC class I receptors on NK cells
MHC class I receptors on NK cells can be either inhibitory or activating. The inhibitory receptors for MHC class I regulate NK cell function by generating a tonic inhibitory signal as hypothesized in the "missing-self" model (Ljunggren and Karre, 1990). The role of the activating receptors for MHC class I appears less clear, but genetic studies have implicated them in recognition of virally infected cells.
Several inhibitory receptors have been identified, but there are two main families involved in NK regulation by MHC class I: the Killer cell Immunoglobulin-like Receptors (KIR) and the C-type lectin-like CD94/NKG2A heterodimers. KIR interact with the classical MHC class Ia (HLA-A, -B and -C) while CD94/NKG2A recognizes the non-classical MHC class Ib, HLA-E. Both synergize and permit NK cells to sense and respond to changes in MHC class I expression. These receptors are expressed in a combinatorial fashion on NK cells to generate an NK cell repertoire. The importance of this is gradually being realised. Expression of an MHC class I inhibitory receptor appears to confer additional reactivity on these NK cells, a phenomenon originally termed "licensing" (Kim et al., 2005). Thus NK cells without inhibitory receptors for MHC class I are thought to be relatively hypofunctional, although in specific scenarios, these cells can become important for viral eradication as shown by studies in murine CMV infection (Orr et al., 2010). Furthermore in disease states in which a specific inhibitory receptor may be beneficial, such as KIR2DL3 in hepatitis C virus infection, then individuals with more beneficial repertoires may be more likely to clear infection (Alter et al., 2010).
In addition to these key receptors, members of the Ig-Like Transcripts (ILTs) family, which are genetically related to KIR (Wilson et al., 2000), can also recognize MHC class I. For example, LILRB1 (LIR-1, ILT-2), which binds a broad range of MHC class I molecules including HLA-G (Vitale et al., 1999), is able to inhibit the NK cell line NK92 (Kirwan and Burshtyn, 2005). It is also expressed in a variegated fashion on NK cells (Davidson et al., 2010). However it does not appear to play a part in NK cell education and it has yet to be demonstrated that this gene family play a major role in inhibiting NK cells in vivo (Yawata et al., 2008).
CD94/NKG2A and KIR molecules have adopted two different recognition strategies. To a large extent CD94/NKG2A ignore MHC class I diversity by recognizing HLA-E. This non-polymorphic MHC class I molecule binds leader peptide sequences derived from classical MHC-A, -B and -C molecules and also from HLA-G (Llano et al., 1998). In contrast the KIR family embrace the diversity of MHC class I through direct recognition of polymorphic determinants. This strategy leads to a highly variable and polymorphic KIR system with diversity comparable to that of MHC class I (Valiante et al., 1997a).
KIR structure and signalling function
The KIR family (assigned the designation of CD158) is a member of the immunoglobulin superfamily that comprises 15 expressed receptors (KIRDL1-5B, KIR3DL1-3, KIR2DS1-5 and KIR3DS1): which can be either inhibitory or activating (Table 1). All KIR are type I transmembrane glycoproteins formed from either two (KIR2D) or three (KIR3D) extracellular Ig-like domains, a stem region, a transmembrane region and a cytoplasmic tail. Depending on the length of the cytoplasmic tail KIR can be subdivided into long-tailed and short-tailed receptors. In general these structural characteristics correlate with their function. Long-tailed KIR are generally inhibitory and short-tailed KIR are activating (Vilches and Parham, 2002). An exception to this rule is the receptor KIR2DL4 which has a long intracytoplasmic tail but stimulates potent cytokine production, although only minimal cytotoxicity (Rajagopalan et al., 2006).
Inhibitory KIR contain one or two Immunoreceptor Tyrosine-based Inhibitory Motifs (ITIMs; V/I/LxYxxL/V), which are require for NK cell inhibition via recruitment of the protein tyrosine phosphatases SHP-1 and SHP-2. SHP-1/2 activation leads to the suppression of activating receptor signals (Long, 2008). Activating KIR possess a positively charged residue (usually arginine) in the transmembrane region, which facilitate the association with accessory molecules, DAP12 or FcεRIγ (KIR2DL4) and NK cell activation (cytotoxicity and/or cytokine production) (MacFarlane and Campbell, 2006). The exception, KIR2DL4 contains both ITIMs and a positively charged residue (lysine), which facilitates the association with FcεRIγ and the induction of the activating signals (Kikuchi-Maki et al., 2005).
The KIR genes can be divided into six lineages based on phylogenetic analysis. This has allowed "tracking" of the KIR across species and given insights into the evolution of the KIR gene families. Lineage I KIR have two extracellular domains in the D0D2 conformation; lineage II KIR are specific for MHC-A and -B; lineage III KIR bind HLA-C; lineage V KIR are related to the human KIR framework gene KIR3DL3; and lineage IV and VI KIR are expansions specific to the rhesus macaque and new world monkeys respectively.
KIR locus and diversity
Genotyping of individuals for specific KIR genes demonstrated an unexpected diversity in gene content amongst the population (Uhrberg et al., 1997). The genotype of these individuals correlated with expression of KIR genes thus demonstrating that this genetic diversity would be important for NK cell function. This seminal study started a detailed investigation into KIR genetics. Sequencing of two KIR haplotypes from a single individual showed that the KIR are encoded by a compact family of genes which occupy about 150 kb of the Leukocyte Receptor Complex (LRC) on chromosome 19q13.4 (Wilson et al., 2000; Wende et al., 1999). The locus is flanked by the LILR and the FCAR genes and contains up to 17 KIR genes and pseudogenes (Kelley et al., 2005). Because the KIR genes have high sequence homology to each other (90-95%) and are closely distributed within the LRC, they have been proposed to evolve by non-allelic homologous recombination (Carrington and Cullen, 2004). This mechanism could explain the expansion and contraction of the KIR locus and provide a basis for the substantial diversity observed (Hsu et al., 2002a; Hsu et al., 2002b; Shilling et al., 2002; Uhrberg et al., 2002; Whang et al., 2005; Martin et al., 2003).
Haplotypic diversity
The number of putatively expressed KIR genes usually ranges from 7 to 12, depending primarily on the presence or absence of activating KIR loci (Wilson et al., 2000; Uhrberg et al, 2002; Witt et al., 1999). This variation in gene content is one component of KIR diversity. Despite this extreme variability some systematic features are conserved in the organisation of the KIR locus. Four KIR genes, KIR3DL3, KIR3DP1, KIR2DL4 and KIR3DL2, are found in all individuals and have been named framework loci (Bashirova et al., 2006). KIR3DL3 and KIR3DL2 define the ends of the KIR-gene region and KIR3DP1-KIR2DL4, the middle. Regions of genetic variability are located between KIR3DL3 and KIR3DP1, and between KIR2DL4 and KIR3DL2 (Wilson et al, 2000; Martin et al., 2000).
Two distinct forms of haplotype, termed A and B, can be distinguished on the basis of gene content. Haplotype A has a fixed gene content (KIR2DL1, KIR2DL3, KIR2DL4, KIR2DS4, KIR3DL1, KIR3DL2, KIR3DP1 and KIR3DL3) (Uhrberg et al., 1997) and fewer genes than B haplotype but the most functionally relevant distinction between these two haplotypes is the number of activating receptors. Haplotype A contains only a single activating KIR gene, KIR2DS4, whereas haplotype B contains various combinations of KIR2DS1, KIR2DS2, KIR2DS3, KIR2DS5, KIR3DS1 and KIR2DS4. Furthermore, the KIR2DS4 gene has a null allele with a population frequency of about 84% (Maxwell et al., 2002), thus some homozygous A haplotype individuals don't express any activating KIR (Hsu et al., 2002b). Although A haplotypes are fixed in term of the number and type of genes present, they show extensive allelic variation at several of the genes. In contrast to A haplotype, B haplotype displays a much greater variety of gene contents. Based on segregation analysis, more than 20 different B haplotypes have been described (Hsu et al., 2002a; Yawata et al., 2002a). These haplotypes contain various combinations of KIR genes, including several activating KIR but there is a very high linkage disequilibrium (LD) between many pairs of genes, as KIR2DL1/KIR2DL3 or KIR3DL1/KIR3DL2 alleles (Uhrberg et al., 1997; Shilling et al., 2002; Witt et al., 1999; Norman et al., 2002; Norman et al., 2001; Toneva et al., 2001; Crum et al., 2000).
However despite the broad categorizations, there are several exceptions to these simple rules. For instance despite the fact that KIR3DL1 (inhibitory) and KIR3DS1 (activating) segregate as alleles of a single locus in the vast majority of individuals, haplotypes have been described in which they occur on the same chromosome (Martin et al., 2008; Norman et al., 2009). Similarly some KIR haplotypes have fewer than expected KIR. For instance a recently described haplotype contains only 3 KIR genes KIR3DL3, KIR2DS1 and KIR3DL2 (Traherne et al., 2010). Thus the rules for the KIR locus appear unusually flexible, perhaps due to the combination of a high sequence homology between the genes, and an overlap in function of these receptors with the well conserved CD94:NKG2 family of receptors. Furthermore some genes (KIR2DS3 and KIR2DS5) appear to occur in two different chromosomal locations. This has led to the splitting of the locus into two separate sections a centromeric section (KIR3DL3-KIR3DP1) and a telomeric section (KIR2DL4-KIR3DL2) which appear to have diversified independently (Pyo et al., 2010). Sequence analysis of a number of KIR haplotypes shows that allelic diversity of the centromeric section is predominant in the A haplotypes, but it is in the telomeric section of the B haplotypes that allelic diversity is most noticeable.
The distribution of A and B haplotypes varies widely between distinct ethnic groups. The A and B haplotype frequencies are relatively even in Caucasian populations (Uhrberg et al., 1997; Hsu et al., 2002b). However the A haplotype dominates in the Korean, Japanese and Han Chinese populations with an approximate 75% frequency (Whang et al., 2005; Yawata et al., 2002b; Jiang et al., 2005) as compared to the Australian Aborigines, who have a very low frequency of the A haplotypes of about 13% (Toneva et al., 2001). These differences may reflect both founder effects and selection by pathogens and may account for some variation in worldwide disease susceptibility.
Allelic polymorphism
Point mutation and homologous recombination generate allelic polymorphism (Table 1) (Norman et al., 2009; Shilling et al., 1998). This allelic polymorphism gives an additional dimension to KIR diversity in that unrelated individuals are unlikely to have identical KIR alleles, similar to the situation for MHC diversity (Gardiner et al., 2001). Allelic polymorphism has been described for all the inhibitory KIR genes and names for alleles at several of the most polymorphic loci have been specified based on nomenclature used for HLA loci (Shilling et al., 2002). This polymorphism significantly influences their ligand affinities and levels of cell surface expression. For example, distinct alleles of KIR3DL1, one of the most polymorphic KIR genes encode molecules that appear to be expressed at different levels on the surface of NK cells (Gardiner et al., 2001; Yawata et al., 2006; Pando et al., 2003). Moreover this allelic variability can occur at positions encoding residues that affect interaction with HLA class I (Boyington et al., 2000; Fan et al., 2001) and influences both the binding affinity and the inhibitory capacity. Similarly the genes KIR2DL2 and KIR2DL3 also segregate as alleles of a single locus and although they have broadly similar MHC class I specificity, bind their ligands with substantially different avidities (Moesta et al., 2008). The synergy of haplotype variability and, allelic polymorphism has generated substantial diversity across both individual populations, but also across different ethnic groups (Rajalingam et al., 2001). This diversity is likely driven by both encounters with pathogens, but also by reproductive fitness.
Table 1: KIR nomenclature, lineages and ligands (IPD - KIR Database; Cadavid and Lun, 2009)
CD nomenclature
*KIR2DL5 gene is duplicated and encoded by two separate loci within the LRC gene cluster.
KIR recognition and peptide selectivity
Individual KIR recognize distinct subsets of the classical human MHC class I allotypes. This binding specificity is determined both by residues of the MHC class I and those of the peptide bound by the MHC class I molecule. Inhibitory KIR are able to recognize all the known HLA-C allotypes (C1 and C2 subgroup) and some subsets of HLA-A and HLA-B allotypes. KIR2DL1 binds HLA-C2 allotypes, which all have a lysine at position 80 (Colonna et al., 1993). KIR2DL2 and KIR2DL3, which segregate as alleles of the same locus, bind mainly HLA-C1 allotypes (with an asparagine at position 80), some HLA-C2 allotypes and a few HLA-B allotypes which have an asparagine at position 80 and also a valine at position 76 (Moesta et al., 2008; Wagtmann et al., 1995; Pende et al., 2009). KIR3DL1 recognize the "Bw4" motif present in 40% of the known HLA-B allotypes and in some HLA-A allotypes, with a higher affinity for the Bw4 motifs containing an isoleucine at position 80 (Cella et al., 1994; Gumperz et al., 1995). KIR3DL2 is only known to bind HLA-A3 and HLA-A11 allotypes whilst ligands for KIR2DL5 and KIR3DL3 have not yet been identified.
Given the high sequence homology between the extracellular domains of some activating and inhibitory KIR (~99%), several activating KIR have been reported to bind the same HLA molecules as their inhibitory counterparts, although with significantly weaker affinity (Biassoni et al., 1997; Valés-Gómez et al., 1998; Stewart et al., 2005). Due to their low affinities, the activating KIR-HLA binding specificity is quite uncertain. Moreover the KIR-HLA affinities can be enhanced by specific peptides presented in the HLA molecules, as has been shown for KIR2DS1 interactions with Epstein-Barr virus-infected cells (Stewart et al., 2005). One potential model is that these receptors may bind specific viral peptides that have yet to be determined.
In addition to the MHC class I heavy chain, all inhibitory KIR tested to date have some degree of peptide selectivity (Boyington et al., 2000; Malnati et al., 1995; Rajagopalan and Long, 1997; Hansasuta et al., 2004). This appears to have a functional relevance in that NK cells expressing KIR2DL3 are exquisitely sensitive to the peptide bound by MHC class I. This is because peptides that stabilise MHC class I, but bind KIR weakly can antagonize the inhibition due to MHC class I:peptide complexes that bind KIR strongly (Fadda et al., 2010). This process appears to be more efficient than MHC class I downregulation in activating NK cells, and may be important for recognition of infected targets (Rajagopalan and Long, 2010).
KIR evolution
A comparison of the KIR genes in human and chimpanzees revealed unexpectedly rapid evolution of the KIR locus, in many ways exceeding the pace of their MHC class I ligand (Khakoo et al., 2000). This contrasts with the high conservation of the CD94:NKG2A system (Shum et al., 2002). Work in the higher primates has revealed that in these species the KIR genes have expanded substantially. Mice, which are the most frequently used immunological model for the immune system of man and his response to disease, do not have KIR as regulators of NK cell activity (Figure 1). Instead they have an expansion of the C-type lectin-like receptors, the Ly49 genes which also bind classical MHC class I molecules. These genes are related to the NKG2A family of receptors, and both these gene families in addition to the CD94 gene are found on murine chromosome 6 in a region designated the natural killer cell complex (NKC) (Vance et al., 1998). In mice CD94:NKG2A binds the non-classical MHC class I molecule Qa-1, which also binds MHC class I leader sequences. Thus comparison of humans and rodents has revealed two distinct evolutionary pathways for NK cell receptors: one leading to diversification of KIR and the other to diversification of Ly49. Both species have inhibitory NK cell receptors for classical class I molecules and both for a non-classical MHC class I molecule, although NKG2A in mouse and human are not strictly orthologous. Remnants of non-functional genes can be found in the alternate species: the KIR are represented by a gene on murine chromosome X and Ly49 is a pseudogene in humans (Kelley et al., 2005) (Figure 1).
Figure 1
Studies in other species have revealed the uniqueness of KIR in the simian primates. KIR genes have been found in species as diverse as cattle, horses, dogs and pinnipeds (Parham et al., 2010). These are thought to have derived from duplication of an ancestral KIR3D gene over 100 million years ago. This resulted in two genes: KIR3DL and KIR3DX (Guethlein et al., 2007). The KIR3DL gene is thought to have spawned the KIR genes of the primates, and the KIR3DX gene given rise to the multigene KIR family in cattle (Sambrook et al., 2006). KIR3DX is retained in humans, however it is a non-functional pseudogene in the LRC amongst the LILR gene. The adoption of different solutions to the issue of NK cell receptor variability is further illustrated by the expansion of the NKG2 family of genes in the prosimians (Averdam et al., 2009) and the observation that the pinnipeds (seals and sea lions) seem to cope with having only one functional KIR and one functional Ly49 gene (Hammond et al., 2009).
Genetic studies implicating specific combinations of KIR in infectious diseases imply that pathogens are a major driving force in KIR evolution. This follows naturally from the observation that natural killer cells are important in clearing viral infections. Pathogens can drive KIR selection both by a direct effect on specific KIR genes and also via an indirect selective pressure through driving the evolution of MHC Class I. This is well illustrated by the co-evolution of KIR and the MHC-C locus in the great apes. The KIR can be divided into lineages based on sequence homologies. The lineage III KIR have MHC-C ligands. The most divergent species from man with an MHC-C allele is the orangutan (Adams et al., 1999). In this species this locus is present in only about half the individuals. Nevertheless, in the orangutan and man's more closely related ancestors (the gorilla, the common chimpanzee and the pygmy chimpanzee) lineage III KIR have expanded, implying that MHC and the KIR are co-evolving (Abi-Rached et al., 2010b). Similarly in the old world monkeys the expansion of MHC-A/B locus has led to expansion of lineage II KIR, and in the hoolock gibbon loss of MHC-G corresponds to loss of the lineage I gene KIR2DL4, which has been shown to be relatively conserved amongst higher primates (Abi-Rached et al., 2010a; Parham et al., 2010).
Whilst humans and chimpanzees share MHC-A, -B and -C loci, they share relatively few KIR genes. These include the framework genes KIR3DL3, KIR2DL4 and KIR3DL2 (which in the chimpanzee is a chimera of human KIR3DL1 and KIR3DL2 called Pt-KIR3DL1/2), and the genes KIR2DL5 and KIR2DS4 (Khakoo et al., 2000). Humans have retained lineage III inhibitory KIR which bind strongly to HLA-C, however the activating KIR for HLA-C are low avidity. Conversely the common chimpanzee has retained high avidity activating and inhibitory for both group 1 and group 2 HLA-C allotypes. Furthermore, although the lineage II human KIR bind both HLA-A and -B allotypes, the most relevant interaction appears to be that of KIR3DL1 with HLA-B allotypes with the Bw4 serological motif. Although this KIR does bind HLA-A allotypes with the same serological motif. The binding of other KIR to HLA-A allotypes is less well documented. KIR3DL2 binds HLA-A3 and HLA-A11, although the avidity of this interaction is not well studied (Dohring et al., 1996; Pende et al., 1996; Valiante et al., 1997b). It has also been shown to bind tetramers of HLA-B27 homodimers (Kollnberger et al., 2007). KIR2DS4 also binds HLA-A11 (Graef et al., 2009). However the functional relevance of these interactions with HLA-A is not as well documented as for HLA-B: and KIR. Conversely Pt-KIR3DL2, binds both MHC-A and -B allotypes, and can demonstrably inhibit chimpanzee NK cells in a manner not restricted by the Bw4 serological motif, even though this motif is present on a number of chimpanzee MHC-B allotypes (Khakoo et al., 2002). Thus even where the most simple motifs for KIR binding on MHC appear relatively conserved between species there is substantial evidence for a more rapid evolution at the KIR locus.
Within different human populations there is a great diversity in the frequencies of individual KIR genes (Gonzalez-Galarza et al., 2011). The essential role for unequal crossing over in generating this diversity is illustrated by in depth study of the KIR3DL1/KIR3DS1 locus in which it has been demonstrated that this process can account for both duplication and deletion within the KIR gene complex (Norman et al., 2009). Further selection may occur on the basis of the interaction of KIR with its MHC class I ligands, to maintain a functional relationship and which may be fine tuned by the affinity of the KIR for its MHC class I ligand (Single et al., 2007). Thus in the Yucpa population there is a relatively high frequency of the "strong educating" HLA-C*07 allele, and correspondingly higher frequencies of KIR2DL3 alleles with low avidities for HLA-C (Gendzekhadze et al., 2009). This implies that evolutionary pressures have combined to ensure that inhibitory signals to NK cells can be easily overcome and so NK cells can be readily activated in response to pathogens.
Impact of KIR diversity on Human health and disease
The impact of KIR diversity on human health is well illustrated by disease association studies. Whilst infection is thought to be the major driving force for the evolution of KIR, there is substantial evidence that KIR diversity impacts a number of pregnancy associated disorders, including pre-eclampsia and recurrent spontaneous abortion (Moffett-King, 2002). During placentation the trophoblast burrows into the placenta, and natural killer cells appear to be important for this process. Analysis of maternal and foetal KIR and MHC class I genotypes demonstrate that if these result in greater foetal NK cell activation then pregnancy is more likely to be successful, due to improved placentation. Thus in cases where the KIR haplotype of the fetus has a preponderance of activating receptors, such as a type B KIR haplotype, then there I a lower probability of pre-eclampsia (Hiby et al., 2004). Conversely if the fetus has only one activating KIR, as is found in a group A KIR haplotype and the mother has a strong inhibitory MHC class I type for example two group 2 HLA-C alleles then there is a greater risk of pre-eclampsia, foetal growth retardation and recurrent spontaneous abortion (Hiby et al., 2010). This likely drives the evolution of the KIR locus towards a preponderance of activating receptors.
Early studies in infectious disease would concur with this evolutionary direction. In HIV infection the activating receptor KIR3DS1 and its HLA-B ligands Bw4 with isoleucine at position 80 (Bw480I) are associated with a slower progression to AIDS (Martin et al., 2002). This begs the question as to the persistence of the "inhibitory" A haplotypes in the human population. Further studies of HIV infection have revealed a second model in which a hierarchy of inhibitory interactions between the inhibitory receptor KIR3DL1 and its HLA-BBw4 ligands influences the progression to AIDS (Martin et al., 2007). Interestingly in this genetic analysis the most protective allele KIR3DL1*004 is one which is not expressed on the cell surface (Pando et al., 2003). This is a feature of other KIR alleles, including KIR2DL2*004 and a number of alleles of KIR2DS3 (VandenBussche et al., 2006; VandenBussche et al., 2009). This implies that care must be taken with the interpretation of these genetic analyses as the presence of the receptor and its ligand does not necessarily mean that there is a simple functional relationship.
The importance of inhibitory receptor:ligand interactions is probably best illustrated by consideration of hepatitis C virus (HCV) infection. This is a positive stranded RNA virus that has relatively little specific effects on MHC class I expression, and so may act as a more general template for understanding the role of NK cells in viral infection. Genetic analysis has shown that weaker inhibitory interactions are associated with a more beneficial outcome of HCV infection. It was originally shown that KIR2DL3 and its group 1 HLA-C ligands were associated with spontaneous resolution of HCV infection, which in the vast majority of individuals leads to chronic infection (Khakoo et al., 2004). Binding analysis reveals that KIR2DL3 is a weaker binder to HLA-C than its allele KIR2DL2, which has a similar MHC class I specificity (Winter et al., 1998). It is therefore thought that a weaker inhibitory receptor:ligand interaction can be more easily perturbed than a strong one and hence is more likely to lead to NK cell activation. This protection has been observed in other HCV exposed cohorts and disease settings, including in the clinically important treatment setting (Romero et al., 2008; Knapp et al. 2010; Vidal-Castiñeira et al., 2010). Furthermore, similar to HIV, it can be mapped to the allelic level, but in this case at the HLA-C locus. Thus the common group 1 HLA-C*07 alleles are not protective whereas most other group 1 HLA-C alleles are protective in combination with KIR2DL3 (Knapp et al., 2010). Thus it appears that there is a balancing selection on the KIR "A" and "B" haplotypes in humans which has permitted maintenance of both in the extant human populations (Gendzekhadze et al., 2009).
Conclusion
By virtue of the expression of KIR, natural killer cells are a branch of the innate immune system that are undergoing constant evolution in response to both pathogens and the challenge of successful reproduction. The KIR have diversified in response to MHC driven selective pressures, following exposure to pathogens. Due to the independent segregation of KIR and their MHC class I ligands, in any given individual some KIR may be redundant. Nevertheless, on a population level, by fine tuning natural killer activity these receptors are key regulators of the innate immune response to pathogens.