Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 Jan 15.
Published in final edited form as: J Immunol. 2014 Dec 5;194(2):781–789. doi: 10.4049/jimmunol.1402542

The interaction of KIR3DL1*001 with HLA class I molecules is dependent upon molecular microarchitecture within the Bw4 epitope

Philippa M Saunders *,#, Julian P Vivian †,$,#, Nikola Baschuk ‡, Travis Beddoe †, Jacqueline Widjaja *, Geraldine M O'Connor §, Corinne Hitchen †, Phillip Pymm †, Daniel M Andrews ‡, Stephanie Gras †, Daniel W McVicar §, Jamie Rossjohn †,#,$,2, Andrew G Brooks *,2
PMCID: PMC4282956  NIHMSID: NIHMS641465  PMID: 25480565

Abstract

The Killer cell Immunoglobulin-like Receptor 3DL1 (KIR3DL1) inhibits activation of Natural Killer (NK) cells upon interaction with Human Leukocyte Antigen (HLA) class I molecules such as HLA-B*57:01 which contain the Bw4 epitope spanning residues 77-83 (e.g. NLRIALR) and not with HLA allomorphs that possess the Bw6 motif (e.g. HLA-B*08:01), which differ at residues 77, 80, 81, 82 and 83. Although Bw4 residues Ile80 and Arg83 directly interact with KIR3DL1*001, their precise role in determining KIR3DL1-HLA-Bw4 specificity remains unclear. Recognition of HLA-B*57:01 by either KIR3DL1+ NK cells or the NK cell line YTS transfected with KIR3DL1*001 was impaired by mutation of residues 80 and 83 of HLA-B*57:01 to the corresponding amino acids within the Bw6 motif. Conversely, the simultaneous introduction of three Bw4 residues at positions 80, 82 and 83 into HLA-B*08:01 conferred an interaction with KIR3DL1*001. Structural analysis of HLA-B*57:01, HLA-B*08:01 and mutants of each bearing substitutions at positions 80 and 83 revealed that Ile80 and Arg83 within the Bw4 motif constrain the conformation of Glu76, primarily through a salt-bridge between Arg83 and Glu76. This salt-bridge was absent in HLA-Bw6 molecules as well as position 83 mutants of HLA-B*57:01. Mutation of the Bw4 residue Ile80 also disrupted this salt-bridge, providing further insight into the role that position 80 plays in mediating KIR3DL1 recognition. Thus the strict conformation of HLA-Bw4 allotypes, held in place by the Glu76-Arg83 interaction, facilitates KIR3DL1 binding whereas Bw6 allotypes present a platform on the α1 helix that is less permissive for KIR3DL1 binding.

INTRODUCTION

Natural Killer (NK3) cell activation is regulated through a delicate balance of stimulatory and inhibitory signals, ensuring rapid responses to transformed or virus-infected cells without the need for prior sensitisation (1). In humans, the Killer cell Immunoglobulin (Ig)-like Receptors (KIR) are a key component of this immune surveillance system, where the interaction between inhibitory KIR and Human Leukocyte Antigen (HLA) class I molecules presenting self-peptides plays a role both in the acquisition of NK cell function during development and in target cell recognition.

Inhibitory KIR receptors can be divided into those with two (2D) or three (3D) Ig-like extracellular domains and possess a long (L), immunoreceptor tyrosine-based inhibitory motif (ITIM)-containing cytoplasmic tail (2). The two Ig-like domains (D1 and D2) of KIR2DL1 and KIR2DL2/3 interact with HLA-C molecules, docking towards the C-terminal end of the peptide-binding groove (3-5). The dimorphism among HLA-C molecules at position 80 is recognised by KIR2DL1 and KIR2DL2 and dictates their reactivity with C2 or C1 molecules, respectively (6, 7). KIR3DL1 interacts exclusively with HLA class I molecules that contain the Bw4 epitope, which is present within ~33% of HLA-B allotypes and ~20% of HLA-A allotypes (8).

The Bw4 motif itself is defined by five residues within the α1 helix (77, 80, 81, 82 and 83), which serologically distinguish it from the Bw6 epitope found in the remaining HLA-B allotypes (9). HLA-Bw4 molecules such as HLA-B*57:01 or HLA-B*15:13 are characterised by the presence of Asn77, Ile80, Ala81, Leu82 and Arg83 (10). While residues 82 and 83 of the Bw4 sequence are conserved, the remaining residues vary to create up to eight different Bw4 motifs (9-11). In contrast, molecules which possess the Bw6 epitope, like HLA-B*08:01 and HLA-B*15:02 (Ser77, Asn80, Leu81, Arg82 and Gly83), are unable to inhibit activation of KIR3DL1+ NK cells (10).

Recent structural studies demonstrated that KIR3DL1*001 interacted with HLA-B*57:01 bound to the endogenous peptide LSSPVTKSF (LF9) via two discontinuous sites (12). The D0 domain clamped around the HLA class I molecule and interacted with highly conserved residues on two loops of the α1 domain (residues 14-18 and 88-92). In addition, the D1 and D2 domains bound over the peptide-binding cleft, with the D2 domain contacting resides of limited polymorphism within the α2 helix (142-151) (12). Residues within the Bw4 motif interacted primarily with the D1 domain. However, despite being the primary determinant of HLA specificity, the interface appeared suboptimal, lacking both charge and shape complementarity (12). Indeed, structural analysis of KIR3DL1*001 bound to HLA-B*57:01/LF9 suggested that the specificity of KIR3DL1 for HLA-Bw4 molecules may lie with key residues both within and proximal to the Bw4 epitope and/or in the conformation of the α1 helix itself. This hypothesis is supported by the observation that while mutations to residues within the D1 domain of KIR3DL1*001 that contacted HLA-B*57:01/LF9 failed to perturb binding, alanine substitutions in HLA-B*57:01 (at positions 76, 79, 80 and 83) markedly impaired the interaction (12).

Despite the structural data, the basis for the specificity of KIR3DL1 for Bw4+ allotypes remains unclear. Indeed the contribution of residues 80, 82 and 83 to recognition by KIR3DL1 varies as a function of HLA polymorphism. For example mutation of Leu82 to its Bw6 counterpart arginine, impacts KIR3DL1 recognition of the Bw4 molecule HLA-B*51:01 yet the same mutation had little impact on recognition of HLA-B*15:13 (8). Further, polymorphisms within KIR3DL1 itself also impact the interaction between the receptor and its ligands (8, 13, 14). Therefore, to ascertain how individual Bw4 residues influence the interaction with KIR3DL1, residues 80, 82 and 83 of the Bw4 and Bw6 epitopes were mutated in HLA-B*57:01(Bw4) and HLA-B*08:01(Bw6). The effect of individual and multiple mutations was tested both via functional and biochemical assays and then further clarified by determining the crystal structures of HLA-B*57:01 and HLA-B*08:01 bearing individual mutations at position 80 or in which residues 80, 82 and 83 had been swapped to the corresponding residues from the Bw6 or Bw4 allotypes respectively. The data suggest that the Bw4 “specificity” of KIR3DL1 may not be due so much to the engagement of a particular interface on HLA-Bw4 molecules as to an inability of KIR3DL1 to accommodate residues that define the HLA-Bw6 motif.

METHODS

Mutation of HLA class I molecules

Full length HLA-B*57:01 and HLA-B*08:01 cDNAs were cloned into pcDNA3.1(−) vectors for transfection into 721.221 cells. A series of mutants bearing reciprocal changes in the Bw4 and Bw6 motifs were generated via site directed mutagenesis essentially as described previously (12). An analogous set of mutations were introduced into plasmids which direct expression of the extracellular domains of HLA-B*57:01 and HLA-B*08:01 in E.coli. (15, 16).

Transfection of Cell Lines

721.221 (221) cells are a human B-lymphoblastoid cell line lacking endogenous expression of the classical HLA-A, -B and -C alleles (17). Transfection of 721.221 cells with either HLA-B*57:01 or HLA-B*08:01 mutants was performed via electroporation at 200V and 975μF with stable transfectants selected with 0.5mg/ml geneticin (Life Technologies). The CD56+ CD3− NK cell line YTS (18, 19) was transfected with a FLAG-tagged KIR3DL1*001 construct in the pEF6/V5-His-TOPO vector (12) via electroporation and stable transfectants selected by culture in the presence of 10μg/ml blasticidin (Life Technologies). All cells were cultured in RPMI media with 10% fetal bovine serum plus supplements.

Intracellular Cytokine Staining of PBMC

IFNγ production by KIR3DL1+ NK cells was assessed essentially as previously described (8). Briefly, 5×105 PBMC were incubated with 5x105 target cells (221 transfectants) in the presence of 2500 U IL-2 for 14 hours, with GolgiPlug (BD Biosciences) added one hour into the incubation. Cells were then washed and stained with anti-CD56-APC (BD Bioscience), anti-CD3-PECy7 (eBioscience) and anti-NKB1-FITC (anti-KIR3DL1; BD Bioscience) prior to fixation with 2% paraformaldehyde. Following permeablisation with 0.2% saponin, cells were stained with anti-IFNγ-PerCPCy5.5 (eBioscience) and analysed by flow cytometry. Data was analysed with FlowJo software, with the percentage of IFNγ producing KIR3DL1+ NK cells (CD56+, CD3−) normalised to the maximal IFNγ output when incubated with the parental 721.221 cell line (or 221.B0801 in the analysis of HLA-B*08:01 mutants).

KIR3DL1 sequencing

RNA was isolated from 1×107 PBMC and converted to cDNA, priming with oligo dT. KIR3D transcripts were amplified using the primers Fw: 5’ CTGTCTGCCTGGCCCAGCGCTGTG 3’ and Rv: 5’TCATGGCAGGAGACAACTTTGG 3’ (94°C for one minute, 69°C for 45 seconds, 72°C for 90 seconds; 35 cycles). PCR fragments were cloned into the pGEM-T easy vector (Promega, Wisconsin) and sequenced using T7 and SP6 primers. Sequences were aligned against published KIR3DL1 sequences and identification of polymorphic alleles verified using the Immuno Polymorphisms Database (IPD) - KIR database (20).

Chromium Release Assays

Cytotoxicity was assessed essentially as described previously (21). Target cells were incubated with 150 μCi [51Cr]/106 cells in serum-free complete DMEM for 60–75 minutes, washed and then co-cultured in complete medium with effector cells for four hours. Cytotoxicity was calculated as 100 × (experimental cpm- spontaneous release)/(total cpm- spontaneous cpm).

Protein Expression and purification

Plasmids encoding the ectodomains of HLA-B*08:01, HLA-B*57:01 (or mutant class I heavy chains) and β2-microglobulin were expressed in E. coli and purified as previously described (15, 22). Soluble KIR3DL1*001 was generated using a baculovirus expression system and purified from the culture supernatant of infected Hi5 cells as described previously (12).

Surface Plasmon Resonance

The interactions between KIR3DL1*001 and the HLA-B*57:01 and HLA-B*08:01 mutants were analysed by surface plasmon resonance (SPR) using a BIAcore 3000 (GE Healthcare) as described previously (12). All experiments were performed at 298 K in a buffer containing 10 mM HEPES pH 7.4, 300 mM NaCl and 0.005% surfactant P20 (P20-HBS). The anti-HLA class I antibody W6/32 (23) was immobilized on adjacent flow cells of a CM5 Sensorchip (GE Healthcare) by amine coupling to a surface density of ~ 1000 resonance units. The respective HLA class I molecules were captured by the immobilized W6/32. An adjacent flow cell to which HLA was not added was activated and quenched in the same manner and served as a control cell. KIR3DL1*001 was serially diluted into P20-HBS to concentrations from 300 – 0.5 μM. KIR3DL1*001 was injected simultaneously over the test and control surfaces at a flow rate of 5 μl/min. Measurements were taken in duplicate.

Crystallization and data collection

Complexes of HLA-B*57:01.I80N and HLA-B*57:01.I80N.L82R.R83G were prepared with the LSSPVTKSF (LF9) peptide. Complexes of HLA-B*08:01.N80I and HLA-B*08:01.N80I.R82L.G83R were prepared with the FLRGRAYGL (FLR) peptide (16). Before crystallisation, all complexes were concentrated to 10 mg/ml in 10 mM Tris pH 8.0. Crystals of each HLA complex were obtained at 294 K by the hanging-drop vapour-diffusion method from reservoir solutions comprising 12-20 % PEG 4000, 0.2 M ammonium acetate and 0.1 M tri-sodium citrate pH 5.4–5.6. Prior to data collection, crystals were equilibrated in reservoir solution with 10% glycerol added as a cryoprotectant and then flash-cooled in a stream of liquid nitrogen at 100 K. X-ray diffraction data were recorded on a Quantum-315 CCD detector at the MX2 beamline of the Australian Synchrotron. The data were integrated and scaled using MOSFLM and SCALA from the CCP4 program suite (24). Details of the data processing statistics are given in Table 1.

Table 1.

Data collection and refinement statistics

B*0801.N80I B*0801.N80I.R82L.G83R B*5701.I80N B*5701.I80N.L82R.R83G
Data collection statistics
Temperature (K) 100 100 100 100
X-ray source MX2 Australian synchrotron MX2 Australian synchrotron MX2 Australian synchrotron MX2 Australian synchrotron
Space group P212121 P212121 P21 P212121
Cell Dimensions a =50.6 , b = 81.0, c = 107.0 a =50.3, b =80.8, c =106.0 a =61.3, b = 49.3, c = 71.0, β = 99.3° a =50.4 , b = 81.1, c = 108.5
Resolution Å 50 - 1.8 (1.90 - 1.80) 50 - 2.0 (2.1 - 2.0) 50 - 1.80 (1.90 - 1.80) 45 -2.15 (2.23 - 2.15)
Total no. observations 162356 (22323) 116464 (10636) 168965 (25284) 240737 (34827)
No. unique observations 40745 (5787) 29877 (2799) 40340 (5841) 25112 (3600)
Multiplicity 4.0 (3.9) 3.9 (3.8) 4.2 (4.3) 9.6 (9.7)
Data completeness (%) 98.3 (97.3) 99.6 (99.6) 98.7 (98.3) 100 (100)
1/σI 13.6 (5.4) 16 (2.9) 6.2 (2.3) 12.1 (3.1)
R merge 1 0.08 (0.346) 0.09 (0.45) 0.13 (0.49) 0.15 (0.81)
Refinement statistics
Non-hydrogen atoms
Protein 3179 3140 3101 3112
Water 461 359 305 237
R factor 2 0.170 0.178 0.184 0.189
R free 2 0.207 0.229 0.219 0.231
r.m.s.d. from ideality
Bond lengths (Å) 0.01 0.01 0.01 0.01
Bond angles (°) 1.03 1.05 1.06 1.08
Ramachandran plot
Favoured regions (%) 96.5 97.1 96.5 95.8
Allowed regions (%) 3.2 2.7 3.5 4.0
B-factors (Å2)
Average main chain 17.7 27.5 33.1 29.7
Average side chain 27.5 34.0 41.8 36.8
Average water 36.2 38.7 56.8 44.1
Open in a new tab
1

Remerge = Σhkl Σj |Ihkl,j - < Ihkl > | / Σhkl Σj Ihkl,j.

2

Rfactor = Σhkl||Fo - |Fc||/Σhkl|Fo| for all data excluding the 5% that comprised the Rfree used for cross-validation.

Structure determination and refinement

Phases for each of the complexes were determined by molecular replacement as implemented in PHASER (25). The search probe for the HLA-B*57:01.I80N and HLA-B*57:01.I80N.L82R.R83G complexes was the structure of the native HLA-B*57:01/LF9 (PDB code 2RFX (15)) with the peptide removed. The search probe for the HLA-B*08:01.N80I and HLA-B*08:01.N80I.R82L.G83R complexes was the structure of the native HLA-B*08:01/FLR (PDB code 1M05 (16)) with the peptide removed. Refinement of the models was carried out in PHENIX (26) with iterative rounds of manual building in COOT (27). Solvent molecules were added with COOT and the structure validated with MOLPROBITY (28). The final refinement values are summarized in Table 1. Coordinates and structure factors are deposited in the PDB (http://www.rcsb.org/pdb/home/home.do) under accession codes HLA-B*57:01I80N.L82R.R83G: 3X11; HLA-B*57:01I80N: 3X12; HLA-B*08:01N80I: 3X13; and HLA-B*08:01N80I.R82L.G83R: 3X14.

RESULTS

Cooperative roles for residues 80, 82 and 83 in HLA-Bw4 discrimination by KIR3DL1+ NK cells

HLA-B*57:01(Bw4) and HLA-B*08:01(Bw6) differ by 33 amino acids, yet the residues that contact the KIR3DL1*001 D0 and D2 domains (residues 14-18/88-92 and 142-151 respectively) in HLA-B*57:01 are identical in both HLA molecules (Figure 1). To examine the basis of KIR3DL1 specificity for Bw4 allotypes, a series of reciprocal mutations spanning residues 77-83 were generated between HLA-B*57:01 and HLA-B*08:01. These were then transfected into the HLA class I-deficient cell line 721.221 (221) and their surface expression assessed with the HLA class I-specific mAb W6/32 by flow cytometry. HLA-B*57:01 molecules bearing individual N77S, I80N or R83G mutations or those with multiple mutations (I80N and R83G or I80N, L82R and R83G) were all expressed at similar levels to wild type HLA-B*57:01, whereas cells transfected with HLA-B*57:01.L82R consistently displayed slightly higher surface expression than the other cell lines (Figure 2A). In addition, the transfectants were also stained with the Bw4-specific mAb RM7.9.63. Notably the N77S and I80N mutations did not affect antibody binding whilst mutation of residues 82 and 83 diminished anti-Bw4 antibody recognition either in isolation or in combination with other mutations, indicating that these two positions impact the conformation of the Bw4 epitope.

Figure 1. Differences between HLA-B*57:01(Bw4) and HLA-B*08:01(Bw6).

Figure 1

Open in a new tab

A. Conservation of residues between HLA-B*57:01 and HLA-B*08:01. Differences are coloured orange. Positions that are bound by KIR3DL1*001 are coloured green. Positions 80 and 83, which differ between HLA-B*57:01 and HLA-B*08:01 and are bound by KIR3DL1*001, are coloured blue. B. Partial sequence alignments of HLA-B*57:01 and HLA-B*08:01. The Bw4/Bw6 motif is highlighted in grey and HLA-B*57:01 regions involved in KIR3DL1*001 binding are noted.

Figure 2. KIR3DL1+ NK cell recognition of HLA class I molecules is influenced by residues 80 and 83.

Figure 2

Open in a new tab

A. Surface expression of mutant HLA-B*57:01 molecules transfected into 721.221 (221) cells was assessed by flow cytometry, staining with either the pan class I antibody W6/32 (black histograms) or with a specific anti-Bw4 antibody (dashed histograms). Filled, grey histograms represent background staining with a secondary anti-mouse IgG-FITC alone. B. PBMCs were isolated from healthy donors and incubated overnight with equal numbers of 221 or 221.B5701 mutant transfectants in the presence of Brefeldin A. Cells were assayed for the production of IFNγ, gating on CD56+, CD3− NK cells expressing KIR3DL1. IFNγ production was normalized against the maximum observed upon incubation with parental, untransfected 221 cells. Data were averaged from five KIR3DL1*001+ donors with the standard error of the mean represented and analysed via a one-way ANOVA, applying a Tukey's Multiple Comparison Test (only values significant compared to inhibition obtained with wild type HLA-B*57:01 are depicted). C. Mutations were introduced into the Bw6 molecule HLA-B*08:01 and transfected into 721.221 cells. Their surface expression was assessed by flow cytometry staining with the pan class I antibody W6/32 and anti-Bw4. D. PBMC were incubated overnight with equal numbers of 221.B0801 transfectants in the presence of Brefeldin A and cells then assayed for the production of IFNγ, gating on CD56+, CD3− NK cells expressing KIR3DL1. IFNγ production was normalized against the maximum observed upon incubation with 221.B0801 cells. The standard error of the mean is depicted from the five KIR3DL1*001+ donors and was analysed using a one-way ANOVA with Tukey test (depicting only differences compared to wild type HLA-B*08:01).

NK cell recognition of both wild type and mutant HLA-B*57:01 molecules was assessed by examining the IFNγ production by KIR3DL1+ NK cells following culture in the presence or absence of parental 221 cells or 221 transfectants. Given that KIR3DL1 is highly polymorphic and that these polymorphisms can impact on the interaction with Bw4+ HLA allotypes (13, 14, 29), five donors with KIR3DL1*001/x genotypes were used for analysis. Culture of PBMC with 221 cells stimulated a robust IFNγ response from KIR3DL1+ NK cells (average 16±4%). IFNγ production by KIR3DL1+ NK cells in the presence of 221 cells was taken as the maximal production and, as expected, the expression of HLA-B*57:01 on target cells strongly inhibited this response (Figure 2B). Analysis of the response to 221 cells expressing mutant HLA-B*57:01 molecules showed that mutation of individual residues had only a modest impact on IFNγ production, although cells expressing the I80N and R83G mutants tended to show less extensive inhibition than those expressing wild-type HLA-B*57:01. Successive mutations to the Bw4 motif, in which two or three Bw6 residues were introduced, significantly augmented NK cell activation as demonstrated by increased production of IFNγ by KIR3DL1+ NK cells when compared to cells cultured with 221.B5701.

Additionally, a second series of transfected cell lines was generated in which residues in the Bw6+ allotype HLA-B*08:01 were replaced with the corresponding residues present in Bw4 allotypes, namely, N80I, R82L or G83R and a triple mutant (N80I.R82L.G83R). Flow cytometric analyses demonstrated that wild type and mutant HLA-B*08:01 molecules were expressed at similar levels on transfected 221 cells (Figure 2C). Further, while HLA-B*08:01.N80I was not recognised by the anti-Bw4 antibody, the R82L and G83R mutants did confer weak reactivity. As expected, the triple mutant showed similar reactivity to HLA-B*57:01, suggesting successful reengineering of the Bw4 epitope (Figure 2C).

To assess the ability of these HLA-B*08:01 mutants to inhibit IFNγ production by KIR3DL1 expressing NK cells, transfected 221 target cells were cultured with PBMC and IFNγ production by NK cells was examined by flow cytometry. Notably, the IFNγ response in the presence of the wild type Bw6 molecule was less than that generated in response to the HLA class I-deficient 221 parental cell alone (7.6±1.8%) most likely reflecting the impact of CD94-NKG2A recognition of HLA-E, since residues 3-11 of the leader sequence of HLA-B*08:01 have been shown to promote this interaction (30-32). Consequently the maximal IFNγ output was normalised to that seen in the presence of HLA-B*08:01 (Figure 2D).

The transfection of 221 cells with HLA-B*08:01 bearing three Bw4 residues (N80I.R82L.G83R) was sufficient to inhibit IFNγ production by KIR3DL1+ NK cells (Figure 2D). This inhibition was as robust as that achieved with wild type HLA-B*57:01, suggesting that no further sequence alterations outside of the Bw4 motif are required for HLA-B molecules from distinct serological groups to interact with KIR3DL1. Substitution of residue 82 in HLA-B*08:01 had no detectable effect upon IFNγ production by KIR3DL1+ NK cells, whilst mutation of N80I or G83R had a modest impact, although this difference was not as pronounced as the partial protection previously observed to be conferred by HLA-B*15:02G83R (10). That no single mutation fully converted HLA-B*08:01 into an inhibitory ligand suggested that while these positions each contributed to the interaction with KIR3DL1, an Ile80, Leu82 or Arg83 substitution alone was insufficient for fully functional recognition by KIR3DL1.

Ile80 and Arg83 are important for functional KIR3DL1*001 recognition of HLA-B*57:01

Since the primary NK cells were obtained from donors with KIR3DL1*001/x genotypes, they had the potential to express additional KIR3DL1 alleles together with other receptors (such as CD94-NKG2A and LILRB1) all of which may impact recognition of target cells expressing wild-type and mutant HLA-B*57:01 proteins. Consequently, the specificity of KIR3DL1*001 was further assessed by transfecting the NK cell line YTS with FLAG-tagged KIR3DL1*001 and comparing the capacity of YTS or YTS.KIR3LDL1*001 cells to lyse 221 or 221 cells transfected with wild type and mutant HLA-B*57:01 (Figure 3). As expected, parental 221 cells were lysed by both YTS and YTS.KIR3DL1*001 cells whereas the expression of wild type HLA-B*57:01 protected 221 cells from lysis by YTS.KIR3DL1*001 but not the parental YTS cells (Figure 3B). Similarly HLA-B*57:01 molecules bearing the N77S or L82R mutations inhibited lysis by YTS.KIR3DL1*001 but not YTS cells, suggesting that these residues were not critical for the interaction with KIR3DL1*001. Mutation of Ile80 to asparagine or Arg83 to glycine, however, essentially abrogated KIR3DL1*001-mediated protection. Furthermore, cells expressing HLA-B*57:01 with multiple mutations were equally sensitive to lysis by both parental and KIR3DL1*001-transfected YTS cells. These results suggested that there is a hierarchy within the HLA-B*57:01 Bw4 epitope where Ile80 and Arg83 are more critical for recognition of HLA-B*57:01 by KIR3DL1*001 than either Asn77 or Leu82.

Figure 3. R83G and I80N mutations to HLA-B*57:01 confer susceptibility to lysis by NK cell lines expressing KIR3DL1*001.

Figure 3

Open in a new tab

A. YTS cells were transfected with Flag-tagged KIR3DL1*001 and the surface expression examined via staining with anti-NKB1-FITC. B. 221 transfectants labelled with 150 μCi [51Cr]/106 cells were incubated with YTS cells or YTS cells transfected with KIR3DL1*001. Experiments were performed twice in triplicate with one data set shown. Open circles represent percentage specific lysis by YTS cells and filled circles lysis by YTS.KIR3DL1*001 cells at the indicated effector to target ratio with the standard error of the mean depicted.

KIR3DL1*001 binding is sensitive to single Bw4 to Bw6 mutations

To further quantify the impact of mutations within the Bw4 motif on the interaction with KIR3DL1*001, soluble HLA-B*57:01 molecules bearing Bw4 to Bw6 mutations (I80N, L82R, R83G, I80N.L82R, I80N.R83G and I80N.L82R.R83G) were refolded with the peptide LSSPVTKSF (LF9) derived from constant region of Igκ and their interaction with recombinant KIR3DL1*001 assessed via SPR. Wild type HLA-B*57:01/LF9 bound with the highest affinity to KIR3DL1*001 (KD of 28μM), a value consistent with our previously published data (12). While HLA-B*57:01.L82R/LF9 bound KIR3DL1*001 with an affinity of 35μM (Table 2), the mutation of Ile80 to asparagine dramatically reduced binding to HLA-B*57:01.I80N/LF9 and no interaction at all could be observed between HLA-B*57:01.R83G/LF9 or indeed any mutant which contained two or more residues of the Bw6 epitope. These data are in full agreement with the impact of the mutations in the YTS-based cytotoxicity experiments.

Table 2.

Steady state KDfor KIR3DL1*001 interactions with HLA class I proteins

HLA Mutant Affinity (|iM)
HLA-B*57:01 Wild type 28.02 ± 3.66
I80N >300
L82R 35.51 ± 8.34
R83G NB
I80N.L82R NB
I80N.R83G NB
I80N.L82R.R83G NB
HLA-B*08:01 Wild type NB
N80I NB
R82L >300
G83R NB
N80I.R82L.G83R 41.86 ± 7.61
Open in a new tab

NB- no detectable interaction

The capacity of KIR3DL1*001 to bind either HLA-B*08:01 or a series of mutants carrying Bw6 to Bw4 mutations was similarly assessed. HLA-B*08:01 as well as HLA-B*08:01.N80I, -R82L, -G83R and the triple mutant, -N80I.R82L.G83R, were refolded with the peptide FLRGRAYGL (FLR; from the EBNA3 protein of EBV) (16). Consistent with KIR3DL1 being specific for HLA class I allotypes of the Bw4 serotype, no interaction between HLA-B*08:01/FLR and KIR3DL1*001 was observed (Table 2). In agreement with the functional assays, the introduction of all three mutations (N80I.R82L.G83R) allowed for KIR3DL1*001 binding with a KD of 40.05±5.6μM. No binding was observed between KIR3DL1*001 and HLA-B*08:01/FLR bearing mutations at residues 80 or 83 and similarly there was little interaction with HLA-B*08:01/FLR bearing the R82L mutation. In conjunction with the YTS experiments, these data confirm the importance of Ile80 and Arg83 for binding of KIR3DL1*001 and raise the potential for a degree of cooperation in creating its ligand.

Overall Structures of the HLA-B*57:01 and HLA-B*08:01 mutants

In order to better understand how residues 80, 82 and 83 impacted on the conformation of the α1 helix and specifically the KIR3DL1*001 binding site in HLA-B allotypes, the crystal structures of HLA-B*57:01.I80N, the HLA-B*57:01 triple mutant (I80N.L82R.R83G), HLA-B*08:01.N80I and the HLA-B*08:01 triple mutant (N80I.R82L.G83R) were determined to high resolution (Table 1). The structures of HLA-B*57:01 mutants were each solved bound to the LF9 peptide to enable direct comparison with the HLA-B*57:01/LF9 structure, in the binary state, and to the KIR3DL1*001/HLA-B*57:01/LF9 complex (12, 15). The HLA-B*08:01 mutant structures were solved in complex with the FLR peptide which has been the subject of extensive structural characterisation (16, 33). The introduction of Bw4/Bw6 swap mutations did not impact on the overall fold of the peptide-binding cleft (r.m.s.d. < 0.6Å), implying that any effects of the mutations were attributable to discrete changes surrounding the sites of the mutations. Analysis of these structures, however, revealed subtle changes that impacted on the architecture of the Bw4 motif and of the peptide-binding cleft that would likely impact KIR3DL1 binding.

The Bw4 motif of HLA-B*57:01 comprises a central isoleucine at position 80 flanked by the residues Glu76 and Arg83, which form a salt-bridge (Figure 4A and C). Framing this central motif are Arg79 and Leu82 that sit distal to the peptide-binding groove. By comparison, the Bw6 motif in HLA-B*08:01 has a central Asn80 and an arginine at position 82 (Figure 4A). Further, in HLA-B*08:01 the glycine at position 83 prevents a salt-bridging interaction to Glu76, which results in a repositioning of the acidic side chain (Figure 4B).

Figure 4. Analysis of the conformation of the Bw4 motif.

Figure 4

Open in a new tab

A. Conformation of the Bw4 and Bw6 motifs. HLA-B*08:01 (green) superposed onto HLA-B*57:01 (grey cartoon representation with yellow side chains). The Bw4 motif is characterised by a central isoleucine at position 80, a salt-bridge between Glu76 and Arg83 and a wider inter-helical spacing than Bw6. B. Overlay of HLA-B*57:01 (yellow) with HLA- B*57:01.I80N (cyan) and their relation to L166 of KIR3DL1 (orange). The mutation replicates the conformation observed in HLA-B*08:01. Upon mutation, the salt-bridge between Glu76 and Arg83 is lost. Further, the position of Tyr84 is altered, with a coincident narrowing of the spacing between the α1 and α2 helices. C. KIR3DL1*001 in relation to the Bw4 motif. Residues of the Bw4 motif are coloured yellow. The salt-bridge between Glu76 and Arg83 is indicated. Residues Leu166 and His278 of KIR3DL1*001 that interact with the Bw4 motif are coloured orange. D. Conformations of Glu76 in HLA-B*08:01 structures (blue, dark blue and magenta) and in HLA-B*57:01.I80N.L82R.R83G (green). The lack of constraints by an interaction with Arg83 is coincident with conformations that sterically clash with Leu166 of KIR3DL1*001 (orange). These conformations are overlaid with the conformation observed in the Bw4 motif (yellow).

In the context of HLA-B*57:01, Leu82 does not contact KIR3DL1*001 and points away from the peptide-binding cleft (Figure 4C). Comparison of available structures shows that position 82 is inherently flexible, with no correlation found between the conformation of residue 82 and the Bw4 motif. Indeed, substitutions at position 82 of HLA-B*57:01 did not affect KIR3DL1*001 affinity, as measured by SPR (Table 2). Taken together, this suggests no major role for position 82 in the context of KIR3DL1 binding to HLA-B*57:01/LF9.

Residue 80 influences the architecture of the peptide-binding groove

Position 80 appears to impact upon the inter-helical spacing of the α1 and α2 helices about the E and F pockets of the HLA class I molecules, a spacing observed to be a defining difference between the Bw4 and Bw6 motifs (Figure 4 and Figure 5). Analysis of all available crystal structures of HLA-Bw4 and HLA-Bw6 molecules was undertaken with inter-helical distance measured between the Cα groups at positions 80 and 146 (the intersection point with an axis of inertia of the D2 domain of KIR3DL1). This distance for Bw4 structures ranged from 12.5 Å to 13.2 Å (avg 12.8 Å) whereas the spacing in Bw6 allotypes was invariably narrower, ranging from 11.5 Å to 12.2 Å (avg 11.8 Å) (Figure 5). The inter-helical spacing was impacted following mutations to either the Bw4 or Bw6 motif. Introduction of asparagine into HLA-B*57:01 at position 80 resulted in Bw6-like inter-helical distances of 12.2 Å (HLA-B*57:01.I80N) and 12.1 Å (HLA-B*57:01.I80N.L82R.R83G). Likewise, Bw4-like spacing was observed when Ile80 was introduced into HLA-B*08:01 (12.8 Å for HLA-B*08:01.N80I and 12.9 Å for HLA-B*08:01.N80I.R82L.G83R) (Figure 5). This spacing of the α1 and α2 helices was coincident with the position of Tyr84 and its interaction with the Cγ2 group of Ile80 (Figure 4B). Upon mutation of Ile80 to asparagine, Tyr84 shifts 0.7 Å with a corresponding narrowing of the F pocket. These findings highlight the central role of position 80 in influencing the architecture of the peptide-binding groove, which may also impact upon the docking of KIR3DL1.

Figure 5. Inter-helical spacing for HLA-Bw4 and HLA-Bw6.

Figure 5

Open in a new tab

Distances between the α1 and α2 helices measured between positions 80 and 146. The dichotomy observed between Bw4 and Bw6 is mirrored in the mutational swaps in HLA-B*57:01 and HLA-B*08:01.

The Arg83-Glu76 salt-bridge is critical to the conformation of the Bw4 epitope

The contribution of the key residues Ile80 and Arg83, previously shown to contact KIR3DL1, were considered to determine if loss of KIR3DL1 recognition upon mutation to Bw6 was solely due to the removal of these interactions (Figure 4C). Ile80 contacted Leu166 on KIR3DL1*001, yet this contact would likely be maintained in HLA-B*57:01.I80N as the Asn80Nδ2 group overlaid with that of Ile80Cδ1 in HLA-B*57:01 (Figure 4B). Unlike Ile80, Arg83 does not interact with peptide and substitution for Gly, its Bw6 equivalent, is not likely to impact peptide binding. However Arg83 interacts with His278 on KIR3DL1*001 (Figure 4C), a contact naturally lost by substitution to Gly83. Nevertheless, when His278 and Leu166, of KIR3DL1*001 were mutated to alanine no loss of affinity for HLA-B*57:01 was observed (12). This suggested that the importance of Ile80 and Arg83 extends beyond directly binding KIR3DL1*001 and likely encompasses maintaining the scaffold of the Bw4 motif.

Indeed, both Ile80 and Arg83 shape the architecture of the Bw4 motif via interaction with Glu76. In the native HLA-B*57:01 structure, the orientation of Glu76 is stabilised via the salt-bridge with Arg83 (Figure 4B) (12). This interaction is principally replicated when arginine is substituted into position 83 in the HLA-B*08:01.N80I.R82L.G83R structure. Interestingly, in the HLA-B*57:01.I80N mutant, Asn80 H-bonds to the guanidinium group of Arg83 (Figure 4C), which consequently perturbs the Arg83-Glu76 salt bridge, again resulting in movement of Glu76 (Figure 4B). Similarly, in HLA-B*57:01.I80N.L82R.R83G and HLA-B*08:01, the presence of a glycine at position 83 removed the salt-bridging interaction and was coincident with movement of Glu76, including conformations that would sterically clash with Leu166 of KIR3DL1*001 (Figure 4D). Accordingly, positions 80 and 83 shape the positioning of Glu76 and thereby the face of the Bw4 motif presented to the KIR3DL1 receptor.

DISCUSSION

While previous studies have implicated residues 80, 83 and to a lesser extent 82 within the Bw4 motif in the creation of the ligand for KIR3DL1, the molecular basis for KIR3DL1 discrimination between Bw4 and Bw6 allotypes remained unclear. Indeed determination of the structure of KIR3DL1*001 bound to HLA-B*57:01/LF9 revealed that the KIR3DL1/Bw4 interface itself lacked both shape and charge complementarity (12). Nevertheless, it provided a structural foundation to evaluate the contributions of individual residues within the Bw4 motif to KIR3DL1 recognition. Consequently we determined the structures of HLA-B*57:01 and HLA-B*08:01 bearing mutations in these signature positions within the Bw4 and Bw6 motifs respectively and assessed their impact on the interaction in both binding and functional assays. A comparison of existing structures of HLA class I allotpyes possessing either the Bw4 or Bw6 motifs showed that there was a subtle difference in the spacing of the α-helices of the peptide-binding groove, with Bw4 allotypes being slightly wider as assessed by the distance between Ile80/Asn80 and Lys146. Notably, relatively small adjustments within the peptide-binding groove can have a profound impact on TCR recognition (33) and we suggest potentially on KIR3DL1 recognition too. An increase in the size of the F-pocket has previously been noted for HLA-B*08:02(Bw4) compared to HLA-B*08:01 with a corresponding broadening of the bound peptide repertoire (34). Interestingly, mutation of Ile80 to asparagine in HLA-B*57:01 was associated with decreased inter-helical spacing. Conversely the introduction of isoleucine into position 80 of HLA-B*08:01 increased this distance relative to that observed in HLA-B*08:01 itself. However, this altered spacing alone did not account for differences in KIR3DL1 reactivity as mutation of residue 80 in HLA-B*08:01 to isoleucine was insufficient to restore KIR3DL1*001-binding or inhibit activation of KIR3DL1+ NK cells.

The structures of the mutant HLA-B*57:01 and HLA-B*08:01 molecules also allowed for a more detailed analyses of the role of residues 80 and 83 within the Bw4 and Bw6 allotypes and their contribution to the microarchitecture above and beyond their impact on inter-helical spacing. Arg83 is a key feature of the Bw4 motif which, in the context of HLA-B*57:01, directly contacts the main chain of His278 of KIR3DL1*001. Additionally, it salt-bridges to Glu76, appearing to stabilise or constrain the positioning of this residue. Critically, the perturbation of the Bw4 motif through the introduction of an asparagine at residue 80 is accompanied by a repositioning of Glu76. The requirement for stabilisation of Glu76 may also account for the observation that alanine substitutions to the KIR3DL1*001 D1 domain had little or no effect on binding whereas changes to HLA-B*57:01/LF9 between residues 76 and 89 abrogated recognition (12).

While mutations within the Bw4 motif of HLA-B*57:01 resulted in a repositioning of Glu76, previous structural studies of HLA-B*27:05 showed that the sequence of the bound peptide could also impact the orientation of Glu76 (35). Furthermore, structural analyses of HLA-B27 allotypes bound to an identical peptide have shown that in HLA-B*27:09, the P8 side chain (Arg) can salt-bridge directly to Glu76 whereas in the context of HLA-B*27:05, which differs only at residue 116, this contact is absent (36). Thus while peptide residues can directly impact interactions with KIR3DL1 via direct molecular contacts, they may also act indirectly, impacting on the conformation of the α1-helix. Together these data provide an additional molecular mechanism for the importance of the penultimate residue in KIR3DL1 binding (12, 13, 37, 38).

Glu76 has also been shown to play a role in regulating KIR interactions with the Bw4 motif in non-human primates (39). Interestingly, in contrast to human KIR3DL1*001, recognition of Bw4 allotypes by Mamu-KIR3DL01 was impaired by the presence of a glutamate at position 76. Indeed ligands for Mamu-KIR3DL01 such as Mamu-B*65:01, have a glycine at this position and mutation to glutamate impaired recognition (39). It is nevertheless unclear whether this inability to tolerate a glutamate at position 76 stems from an altered positioning of the side chain relative to human Bw4 allotypes or that the mode of ligand recognition differs between Mamu-KIR3DL01 and KIR3DL1*001. Indeed KIR3DL1*001 and Mamu-KIR3DL01 are not strictly orthologous and vary significantly at positions 165-7, residues which are integral for Bw4 recognition in humans.

Finally it should be noted that the analyses of primary NK cells from KIR3DL1*001+ donors showed that residues 80, 82 and 83 acted more or less as a single functional recognition block. Specifically, multiple mutations within the Bw4 motif were required to significantly abrogate recognition of HLA-B*57:01 by KIR3DL1+ NK cells from donors who were KIR3DL1*001+. Conversely the introduction of isoleucine, leucine and arginine at positions 80, 82 and 83 of HLA-B*08:01 were all required to restore levels of inhibition which approximated that conferred by HLA-B*57:01. Thus, while a clear role for positions 80 and 83 was demonstrated in the more reductionist binding and transfection approach, it should be noted that, although KIR3DL1*001 displays high inhibitory capacity (40), the level of KIR3DL1*001 expression on YTS transfectants was low when compared to primary NK cells. Indeed this low level of expression may have resulted in sub-optimal inhibitory signalling and the capacity to discriminate functional effects stemming from the mutation of individual residues within the Bw4 region. In contrast the signalling threshold of primary NK cells appears tuned to discriminate broadly between Bw4 and Bw6 allotypes. This may be through co-expression of other receptors such as those of the LILR family and/or other alleles of KIR3DL1 which may differ with respect to fine specificity (13, 14, 41). Effectively this may create a degree of tolerance for both peptide variation and micropolymorphism within Bw4 allotype. However the extent of any such signalling cooperativity between inhibitory receptors awaits further study.

Acknowledgments

This work was supported by grants from the National Health and Medical Research Council and the Association for International Cancer Research (to A.G.B. and J.R.), and the Intramural Research Program of the National Institutes of Health, National Cancer Institute, National Institute of Allergy and Infectious Diseases, and federal funds from the Frederick National Laboratory for Cancer Research, National Institutes of Health, under Contract HHSN26120080001E. J.P.V. is an Australian Research Council Discovery Early Career Researcher Award Fellow, D.M.A. is the recipient of an NHMRC Career Development Award and J.R. is a National Health and Medical Research Council of Australia Fellow.

Abbreviations used in this article

KIR

killer cell immunoglobulin-like receptor

LF9

LSSPVTKSF peptide

FLR

FLRGRAYGL peptide

SPR

surface plasmon resonance

221

721.221 cell line

REFERENCES

  • 1.Moretta L, Moretta A. Unravelling natural killer cell function: triggering and inhibitory human NK receptors. Embo J. 2004;23:255–259. doi: 10.1038/sj.emboj.7600019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Burshtyn DN, Long EO. Regulation through inhibitory receptors: Lessons from natural killer cells. Trends Cell Biol. 1997;7:473–479. doi: 10.1016/S0962-8924(97)01167-7. [DOI] [PubMed] [Google Scholar]
  • 3.Fan QR, Long EO, Wiley DC. Crystal structure of the human natural killer cell inhibitory receptor KIR2DL1-HLA-Cw4 complex. Nat Immunol. 2001;2:452–460. doi: 10.1038/87766. [DOI] [PubMed] [Google Scholar]
  • 4.Finton KA, Strong RK. Structural insights into activation of antiviral NK cell responses. Immunological reviews. 2012;250:239–257. doi: 10.1111/j.1600-065X.2012.01168.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Boyington JC, Motyka SA, Schuck P, Brooks AG, Sun PD. Crystal structure of an NK cell immunoglobulin-like receptor in complex with its class I MHC ligand. Nature. 2000;405:537–543. doi: 10.1038/35014520. [DOI] [PubMed] [Google Scholar]
  • 6.Mandelboim O, Reyburn HT, Vales-Gomez M, Pazmany L, Colonna M, Borsellino G, Strominger JL. Protection from lysis by natural killer cells of group 1 and 2 specificity is mediated by residue 80 in human histocompatibility leukocyte antigen C alleles and also occurs with empty major histocompatibility complex molecules. The Journal of experimental medicine. 1996;184:913–922. doi: 10.1084/jem.184.3.913. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Colonna M, Brooks EG, Falco M, Ferrara GB, Strominger JL. Generation of allospecific natural killer cells by stimulation across a polymorphism of HLA-C. Science. 1993;260:1121–1124. doi: 10.1126/science.8493555. [DOI] [PubMed] [Google Scholar]
  • 8.Sanjanwala B, Draghi M, Norman PJ, Guethlein LA, Parham P. Polymorphic sites away from the Bw4 epitope that affect interaction of Bw4+ HLA-B with KIR3DL1. Journal of immunology. 2008;181:6293–6300. doi: 10.4049/jimmunol.181.9.6293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Gumperz JE, Litwin V, Phillips JH, Lanier LL, Parham P. The Bw4 public epitope of HLA-B molecules confers reactivity with natural killer cell clones that express NKB1, a putative HLA receptor. The Journal of experimental medicine. 1995;181:1133–1144. doi: 10.1084/jem.181.3.1133. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Gumperz JE, Barber LD, Valiante NM, Percival L, Phillips JH, Lanier LL, Parham P. Conserved and variable residues within the Bw4 motif of HLA-B make separable contributions to recognition by the NKB1 killer cell-inhibitory receptor. Journal of immunology. 1997;158:5237–5241. [PubMed] [Google Scholar]
  • 11.Voorter CE, van der Vlies S, Kik M, van den Berg-Loonen EM. Unexpected Bw4 and Bw6 reactivity patterns in new alleles. Tissue Antigens. 2000;56:363–370. doi: 10.1034/j.1399-0039.2000.560409.x. [DOI] [PubMed] [Google Scholar]
  • 12.Vivian JP, Duncan RC, Berry R, O'Connor GM, Reid HH, Beddoe T, Gras S, Saunders PM, Olshina MA, Widjaja JM, Harpur CM, Lin J, Maloveste SM, Price DA, Lafont BA, McVicar DW, Clements CS, Brooks AG, Rossjohn J. Killer cell immunoglobulin-like receptor 3DL1-mediated recognition of human leukocyte antigen B. Nature. 2011;479:401–405. doi: 10.1038/nature10517. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Thananchai H, Gillespie G, Martin MP, Bashirova A, Yawata N, Yawata M, Easterbrook P, McVicar DW, Maenaka K, Parham P, Carrington M, Dong T, Rowland-Jones S. Cutting Edge: Allele-specific and peptide-dependent interactions between KIR3DL1 and HLA-A and HLA-B. Journal of immunology. 2007;178:33–37. doi: 10.4049/jimmunol.178.1.33. [DOI] [PubMed] [Google Scholar]
  • 14.Carr WH, Pando MJ, Parham P. KIR3DL1 polymorphisms that affect NK cell inhibition by HLA-Bw4 ligand. Journal of immunology. 2005;175:5222–5229. doi: 10.4049/jimmunol.175.8.5222. [DOI] [PubMed] [Google Scholar]
  • 15.Chessman D, Kostenko L, Lethborg T, Purcell AW, Williamson NA, Chen Z, Kjer-Nielsen L, Mifsud NA, Tait BD, Holdsworth R, Almeida CA, Nolan D, Macdonald WA, Archbold JK, Kellerher AD, Marriott D, Mallal S, Bharadwaj M, Rossjohn J, McCluskey J. Human leukocyte antigen class I-restricted activation of CD8+ T cells provides the immunogenetic basis of a systemic drug hypersensitivity. Immunity. 2008;28:822–832. doi: 10.1016/j.immuni.2008.04.020. [DOI] [PubMed] [Google Scholar]
  • 16.Kjer-Nielsen L, Clements CS, Brooks AG, Purcell AW, Fontes MR, McCluskey J, Rossjohn J. The structure of HLA-B8 complexed to an immunodominant viral determinant: peptide-induced conformational changes and a mode of MHC class I dimerization. Journal of immunology. 2002;169:5153–5160. doi: 10.4049/jimmunol.169.9.5153. [DOI] [PubMed] [Google Scholar]
  • 17.Shimizu Y, DeMars R. Production of human cells expressing individual transferred HLA-A,-B,-C genes using an HLA-A,-B,-C null human cell line. Journal of immunology. 1989;142:3320–3328. [PubMed] [Google Scholar]
  • 18.Chen X, Allan DS, Krzewski K, Ge B, Kopcow H, Strominger JL. CD28-stimulated ERK2 phosphorylation is required for polarization of the microtubule organizing center and granules in YTS NK cells. Proceedings of the National Academy of Sciences of the United States of America. 2006;103:10346–10351. doi: 10.1073/pnas.0604236103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Drexler HG, Matsuo Y. Malignant hematopoietic cell lines: in vitro models for the study of natural killer cell leukemia-lymphoma. Leukemia. 2000;14:777–782. doi: 10.1038/sj.leu.2401778. [DOI] [PubMed] [Google Scholar]
  • 20.Robinson J, Waller MJ, Parham P, de Groot N, Bontrop R, Kennedy LJ, Stoehr P, Marsh SG. IMGT/HLA and IMGT/MHC: sequence databases for the study of the major histocompatibility complex. Nucleic Acids Res. 2003;31:311–314. doi: 10.1093/nar/gkg070. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Zappacosta F, Borrego F, Brooks AG, Parker KC, Coligan JE. Peptides isolated from HLA-Cw*0304 confer different degrees of protection from natural killer cell-mediated lysis. Proceedings of the National Academy of Sciences of the United States of America. 1997;94:6313–6318. doi: 10.1073/pnas.94.12.6313. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Clements CS, Kjer-Nielsen L, MacDonald WA, Brooks AG, Purcell AW, McCluskey J, Rossjohn J. The production, purification and crystallization of a soluble heterodimeric form of a highly selected T-cell receptor in its unliganded and liganded state. Acta Crystallogr D Biol Crystallogr. 2002;58:2131–2134. doi: 10.1107/s0907444902015482. [DOI] [PubMed] [Google Scholar]
  • 23.Parham P, Barnstable CJ, Bodmer WF. Use of a monoclonal antibody (W6/32) in structural studies of HLA-A,B,C, antigens. Journal of immunology. 1979;123:342–349. [PubMed] [Google Scholar]
  • 24.Collaborative Computational Project N. The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr. 1994;50:760–763. doi: 10.1107/S0907444994003112. [DOI] [PubMed] [Google Scholar]
  • 25.McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. Phaser crystallographic software. J Appl Crystallogr. 2007;40:658–674. doi: 10.1107/S0021889807021206. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Adams PD, Grosse-Kunstleve RW, Hung LW, Ioerger TR, McCoy AJ, Moriarty NW, Read RJ, Sacchettini JC, Sauter NK, Terwilliger TC. PHENIX: building new software for automated crystallographic structure determination. Acta Crystallogr D Biol Crystallogr. 2002;58:1948–1954. doi: 10.1107/s0907444902016657. [DOI] [PubMed] [Google Scholar]
  • 27.Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004;60:2126–2132. doi: 10.1107/S0907444904019158. [DOI] [PubMed] [Google Scholar]
  • 28.Davis IW, Leaver-Fay A, Chen VB, Block JN, Kapral GJ, Wang X, Murray LW, Arendall WB, 3rd, Snoeyink J, Richardson JS, Richardson DC. MolProbity: all-atom contacts and structure validation for proteins and nucleic acids. Nucleic Acids Res. 2007;35:W375–383. doi: 10.1093/nar/gkm216. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.O'Connor GM, Guinan KJ, Cunningham RT, Middleton D, Parham P, Gardiner CM. Functional polymorphism of the KIR3DL1/S1 receptor on human NK cells. Journal of immunology. 2007;178:235–241. doi: 10.4049/jimmunol.178.1.235. [DOI] [PubMed] [Google Scholar]
  • 30.Borrego F, Ulbrecht M, Weiss EH, Coligan JE, Brooks AG. Recognition of human histocompatibility leukocyte antigen (HLA)-E complexed with HLA class I signal sequence-derived peptides by CD94/NKG2 confers protection from natural killer cell-mediated lysis. The Journal of experimental medicine. 1998;187:813–818. doi: 10.1084/jem.187.5.813. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Petrie EJ, Clements CS, Lin J, Sullivan LC, Johnson D, Huyton T, Heroux A, Hoare HL, Beddoe T, Reid HH, Wilce MC, Brooks AG, Rossjohn J. CD94-NKG2A recognition of human leukocyte antigen (HLA)-E bound to an HLA class I leader sequence. The Journal of experimental medicine. 2008;205:725–735. doi: 10.1084/jem.20072525. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Sullivan LC, Clements CS, Beddoe T, Johnson D, Hoare HL, Lin J, Huyton T, Hopkins EJ, Reid HH, Wilce MC, Kabat J, Borrego F, Coligan JE, Rossjohn J, Brooks AG. The heterodimeric assembly of the CD94-NKG2 receptor family and implications for human leukocyte antigen-E recognition. Immunity. 2007;27:900–911. doi: 10.1016/j.immuni.2007.10.013. [DOI] [PubMed] [Google Scholar]
  • 33.Gras S, Burrows SR, Turner SJ, Sewell AK, McCluskey J, Rossjohn J. A structural voyage toward an understanding of the MHC-I-restricted immune response: lessons learned and much to be learned. Immunological reviews. 2012;250:61–81. doi: 10.1111/j.1600-065X.2012.01159.x. [DOI] [PubMed] [Google Scholar]
  • 34.Arnett KL, Huang W, Valiante NM, Barber LD, Parham P. The Bw4/Bw6 difference between HLA-B*0802 and HLA-B*0801 changes the peptides endogenously bound and the stimulation of alloreactive T cells. Immunogenetics. 1998;48:56–61. doi: 10.1007/s002510050400. [DOI] [PubMed] [Google Scholar]
  • 35.Stewart-Jones GB, di Gleria K, Kollnberger S, McMichael AJ, Jones EY, Bowness P. Crystal structures and KIR3DL1 recognition of three immunodominant viral peptides complexed to HLA-B*2705. Eur J Immunol. 2005;35:341–351. doi: 10.1002/eji.200425724. [DOI] [PubMed] [Google Scholar]
  • 36.Hulsmeyer M, Welfle K, Pohlmann T, Misselwitz R, Alexiev U, Welfle H, Saenger W, Uchanska-Ziegler B, Ziegler A. Thermodynamic and structural equivalence of two HLA-B27 subtypes complexed with a self-peptide. J Mol Biol. 2005;346:1367–1379. doi: 10.1016/j.jmb.2004.12.047. [DOI] [PubMed] [Google Scholar]
  • 37.Fadda L, O'Connor GM, Kumar S, Piechocka-Trocha A, Gardiner CM, Carrington M, McVicar DW, Altfeld M. Common HIV-1 peptide variants mediate differential binding of KIR3DL1 to HLA-Bw4 molecules. J Virol. 2011;85:5970–5974. doi: 10.1128/JVI.00412-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Brackenridge S, Evans EJ, Toebes M, Goonetilleke N, Liu MK, di Gleria K, Schumacher TN, Davis SJ, McMichael AJ, Gillespie GM. An early HIV mutation within an HLA-B*57-restricted T cell epitope abrogates binding to the killer inhibitory receptor 3DL1. J Virol. 2011;85:5415–5422. doi: 10.1128/JVI.00238-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Schafer JL, Colantonio AD, Neidermyer WJ, Dudley DM, Connole M, O'Connor DH, Evans DT. KIR3DL01 recognition of Bw4 ligands in the rhesus macaque: maintenance of Bw4 specificity since the divergence of apes and Old World monkeys. Journal of immunology. 2014;192:1907–1917. doi: 10.4049/jimmunol.1302883. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Yawata M, Yawata N, Draghi M, Little AM, Partheniou F, Parham P. Roles for HLA and KIR polymorphisms in natural killer cell repertoire selection and modulation of effector function. The Journal of experimental medicine. 2006;203:633–645. doi: 10.1084/jem.20051884. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.O'Connor GM, Vivian JP, Widjaja JM, Bridgeman JS, Gostick E, Lafont BA, Anderson SK, Price DA, Brooks AG, Rossjohn J, McVicar DW. Mutational and Structural Analysis of KIR3DL1 Reveals a Lineage-Defining Allotypic Dimorphism That Impacts Both HLA and Peptide Sensitivity. Journal of immunology. 2014;192:2875–2884. doi: 10.4049/jimmunol.1303142. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES