The JI
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by VandenBussche, C. J.
Right arrow Articles by Hurley, C. K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by VandenBussche, C. J.
Right arrow Articles by Hurley, C. K.
The Journal of Immunology, 2006, 177: 5347-5357.
Copyright © 2006 by The American Association of Immunologists, Inc.

A Single Polymorphism Disrupts the Killer Ig-Like Receptor 2DL2/2DL3 D1 Domain1,2

Christopher J. VandenBussche*, Sivanesan Dakshanamurthy*, Phillip E. Posch{dagger} and Carolyn Katovich Hurley3,*

* Department of Oncology, Georgetown University Medical Center, Washington, DC 20057; and {dagger} Department of Microbiology and Immunology, Georgetown University Medical Center, Washington, DC 20057


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Genetic polymorphisms found in the killer Ig-like receptor (KIR), two domains, long cytoplasmic tail 2/3 (KIR2DL2/3) locus are responsible for the differential binding of KIR2DL2/3 allelic products with their HLA-C ligands and have been associated with the resolution of hepatitis C infection. In our study, a KIR CD3{zeta} fusion-binding assay did not detect any interaction between the KIR2DL2*004 extracellular domain and several putative KIR2DL2/3 ligands. To determine the amino acid polymorphism(s) responsible for the KIR2DL2*004 phenotype, we mutated the polymorphic residues of full-length KIR and expressed them in human Jurkat cells. Flow cytometry analysis failed to detect the surface expression of receptors containing a threonine at position 41 (T41), a polymorphism specific to KIR2DL2*004. Confocal microscopy showed that receptors containing T41 were retained inside the cell and had a perinuclear localization, possibly indicating that their extracellular domain was misfolded. Most KIR2DL2/3 alleles possess an arginine at position 41 (R41), and we predicted through molecular modeling and demonstrated by mutagenesis that R41 most likely interacts with the nearby residues Y77 and D47. Interaction between these residues would maintain C strand contact with the C' and F strands of the D1 domain beta-sheet. Furthermore, R41 and Y77 are conserved in the C and F strand amino acid alignments of Ig-like superfamily members, and may therefore be necessary for the structural integrity of other immune response proteins. Our data indicate that the extracellular T41 polymorphism encoded by the KIR2DL2*004 allele most likely results in misfolding of the D1 domain and complete intracellular retention of the receptor.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Killer Ig-like receptors (KIR)4 are found on the surface of NK cells and some T lymphocyte populations (1). A mature KIR consists of two or three extracellular domains that are similar to Ig superfamily C2-set members (2). The domains of most two-domain KIR (KIR2D) are named D1 and D2, while the three-domain KIR (KIR3D) have an additional N-terminal domain named D0. Stem and transmembrane regions may connect the extracellular domains to a relatively long intracellular tail, which contains ITIMs, providing these receptors (KIR2DL and KIR3DL) with an inhibitory function (3, 4). Other KIR contain a truncated intracellular tail and a charged transmembrane residue; at least some of these short-tailed KIR (KIR2DS and KIR3DS) associate with the DAP12 molecule and are thought to be activating receptors (5, 6). Although the ligands for some KIR are unknown, the known ligands for many KIR are HLA class I molecules. Functional studies have shown that the KIR2DL2 and KIR2DL3 bind members of the HLA-C1 subgroup, defined by the {alpha}-chain residues S77 and N80, while the KIR2DL1 binds members of the HLA-C2 subgroup, defined by the N77 and K80 residues (7, 8). Crystallographic structures indicate that the HLA-C molecules interact with loops in the KIR D1 and D2 domains (9, 10).

The KIR gene family has great genetic diversity, with more than a dozen members found in different combinations on chromosome 19 (11, 12). This haplotypic diversity results in an individual inheriting one, two, or no copies of each known KIR gene. The presence or absence of KIR genes most likely modulates an individual’s immune response and may be important in the context of HIV progression (13), cervical neoplasia (14), hepatitis C infection (15), and hemopoietic stem cell transplantation (16). In addition to haplotypic diversity, each KIR gene is polymorphic, with allelic products often differing by several amino acids (17). Polymorphic amino acids are distributed throughout the receptor proteins and in some instances change the specificity of the receptor. For instance, a single polymorphic amino acid residue abrogates KIR2DS2 binding to at least some members of the HLA-C1 subgroup (18).

In addition to creating receptors with differing specificities, it appears that the relatively rapid evolution of the KIR gene complex has also led to the creation of truncated or unexpressed KIR genes and alleles. In some instances, genetic changes that potentially result in a dysfunctional receptor are clearly distinguishable. The pseudogene KIR2DP1 contains a frameshift deletion that results in an early stop codon (19, 20). Most alleles of the KIR3DP1 pseudogene are missing large coding fragments (11). Except for one variant of KIR3DP1, transcripts of these pseudogenes have not been detected (21). Several KIR2DL4 and KIR2DS4 alleles contain frameshift mutations that prematurely truncate the encoded receptor product (22, 23, 24). In other instances, it is difficult to predict whether genetic polymorphism affects KIR structure or expression. Some variations in the KIR2DL5 promoter are associated with a lack of transcription (25). Only rare transcripts have been found for KIR3DL3, most likely because the KIR3DL3 promoter is heavily methylated (26). Finally, a single amino acid polymorphism is largely responsible for the intracellular retention of the KIR3DL1*004 allelic product (27). These observations complicate the prediction of KIR phenotype based on genotyping and indicate a closer examination of KIR polymorphisms.

The KIR2DL2 and KIR2DL3 genes are inherited as alleles at a single locus (28, 29) and are similar in sequence, varying by no more than six extracellular amino acids between any two allelic products. These polymorphisms may modify receptor function because KIR2DL2 and KIR2DL3 bind some ligands with different affinity (18). The KIR2DL2*004 extracellular domain sequence differs the most from other KIR2DL2 allelic products and possesses unique amino acid polymorphisms. Only described in one African-American family (30), the KIR2DL2*004 allele frequency has not been well evaluated. We were able to demonstrate its presence in a small (n = 33) study of unrelated African-American individuals. Therefore, we decided to investigate whether the novel polymorphisms found in the KIR2DL2*004 extracellular domain could influence the receptor phenotype.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosures
 References
 
DNA constructs

KIR2DL3*001-CD3{zeta} and KIR2DL1-CD3{zeta} constructs were provided by H. Reyburn (University of Cambridge, Cambridge, U.K.) (31). Site-directed mutagenesis was performed using QuikChange II (Stratagene), as per manufacturer’s instructions. To create full-length humanized KIR constructs, the extracellular and transmembrane portions of the KIR-CD3{zeta} constructs were fused by PCR to an intracellular portion of KIR2DL2 cDNA, provided by F. Borrego (National Institute of Allergy and Infectious Diseases, Rockville, MD). The KIR (exons 1 through partial exon 7) fragments of the KIR-CD3{zeta} constructs were amplified with the sense primer KIR-preF, 5'-cccactgcttactggcttat-3' and the antisense primer KIRE7FUSE-r, 5'-tgcaggtgtctgggg-3'. Partial exon 7 through exon 9 of KIR2DL2 was amplified with the sense primer KIRE7FUSE-f, 5'-taaccccagacacctgc-3', and the antisense primer KIRE9R-Stop, 5'-tcagggctcagcatttgg-3'. The two fragments were fused together with the sense primer KIR-preF and the antisense primer KIRE9RStop. For C-terminal V5-tagged constructs, the antisense primer KIRE9R-Tag, 5'-tgagggctcagcatttgg-3', was used in place of KIRE9R-Stop to replace the stop codon with a serine residue. All KIR constructs were inserted into pEF-DEST51 (Invitrogen Life Technologies) via the pCR8/GW/TOPO (Invitrogen Life Technologies) entry vector using Gateway Technology (Invitrogen Life Technologies). The HLA-Cw*0102 expression vector was obtained from The Institute of Physical and Chemical Research BioResource Center (32). All DNA constructs were prepared using the HiSpeed Plasmid Maxi Kit (Qiagen).

Cell lines, culture, and gene expression

The 721.221 cell lines with stable expression of HLA-Cw*0401, HLA-Cw*1202, or HLA-Cw*1403 (33) were obtained from the International Histocompatibility Working Group. A 721.221 cell line null for HLA-A, -B, and -C was a gift from F. Borrego (National Institute of Allergy and Infectious Diseases, Rockville, MD). BW5147 mouse thymoma line and Jurkat clone E6-1 were obtained from the American Type Culture Collection. All cells were cultured in RPMI 1640 supplemented with 2 mM L-glutamine, 10 mM HEPES, 1 mM sodium pyruvate, and 10% FBS. Gene transfection was performed with a Nucleofector II (Amaxa), as per manufacturer’s protocol, with Solution V being used with program S-16 for Jurkat transfection, U-15 for 721.221 transfection, and T-20 for BW5147 transfection. The 721.221 cells transfected with HLA-Cw*0102 were maintained in 700 µg/ml G-418, and cells with high HLA expression were sorted by flow cytometry to create a stable cell line.

Activation assay

Murine BW5147 cells were cotransfected with the appropriate KIR CD3 fusion vector and GFP expression vector. Eighteen hours posttransfection, transfectants were analyzed by flow cytometry to determine the percentage of GFP-positive transfectants. Effector cells and target cells (721.221) were resuspended in growth medium. Unless otherwise noted, the growth medium contained 5 ng/ml PMA to improve detection of weak activation. The BW5147 cells did not secrete IL-2 nonspecifically in the presence of a wide range of PMA. Cells were aliquoted in triplicate into a 96-well round-bottom plate at a 10:1 E:T cell ratio (1.5 x 106 cells/ml) and a total volume of 300 µl. After 48 h, the cell supernatant was harvested and the concentration of IL-2 was measured in duplicate by mouse IL-2 ELISA (BD Pharmingen), following the manufacturer’s protocol. IL-2 secretion was normalized using GFP expression (IL-2 concentration/percentage of GFP-positive transfectants). The IL-2 ELISA was tested against human IL-2 and found to be specific for mouse IL-2.

Flow cytometry

Abs used for flow cytometry were FITC-conjugated anti-V5 (Invitrogen Life Technologies), KIR2DL2/3-specific PE-conjugated CD158b (Beckman Coulter), and KIR2DL2/3-specific FITC-conjugated NKAT2 (BD Pharmingen). For intracellular Ag detection, cells were permeabilized with Cytofix/Cytoperm (BD Pharmingen) and blocked before staining with Image-iT FX (Invitrogen Life Technologies). When noted, cells were cotransfected with pmaxFP-Green (Amaxa) and gated on the green signal to increase detection sensitivity. Cells were stained 18 h posttransfection and analyzed on a BD Biosciences FACSort with FCS Express 2 software (De Novo Software).

Confocal microscopy

Cells were fixed and permeabilized 18 h posttransfection with Cytofix/Cytoperm (BD Pharmingen), blocked with Image-iT FX (Invitrogen Life Technologies), and stained with rabbit polyclonal anti-V5 (eBioscience) and Texas Red-conjugated goat anti-rabbit IgG (Invitrogen Life Technologies). Prepared cells were visualized in 35-mm glass-bottom dishes (World Precision Instruments) using a Fluoview FV300 confocal laser-scanning unit on an Olympus IX-70 microscope with a 60 x 1.4 NA PlanApo lens. Excitation light for the green and red images was provided with a blue Argon laser (488 nm) and a green helium-neon laser (543 nm), respectively. Green and red images were collected separately to avoid channel bleed through and merged using the Fluoview software (Optical Analysis). The presented images are single-slice images through each cell at approximately the greatest diameter. Laser intensity and detection sensitivity were minimally altered between samples; however, because expression of the tagged protein was transient, microscopy was used solely for protein localization and not used to measure relative levels of protein between samples.

Molecular modeling

Structural models were based primarily on the x-ray structure of KIR2DL2 (Brookhaven Protein Data Bank (PDB): 2DL2) (34). The x-ray structures of KIR2DL3 (PDB: 1B6U) and KIR2DS2 (PDB: 1M4K) were also examined (35, 36). KIR2DL2(R41) and KIR2DL2(T41) were energy minimized using the consistent valence force field (CFF91) with the default partial atomic charges available in Discover version 3.0 of Insight II (Accelrys). The cutoff for nonbonded interaction energies was set to {infty} (no cutoff). To avoid unrealistic movements of the protein caused by computational artifacts, the structures were relaxed gradually. The dielectric constant was set at {epsilon} = 4 to account for the dielectric shielding found in proteins. Each minimization was conducted in two steps, first using steepest descent minimization for 200 cycles and then using conjugate gradient minimization until the average gradient fell below 0.01 kcal/mol.

Molecular dynamics (MD)

Using the energy-minimized structure of KIR2DL2 as the initial model, 3-ns MD simulations with a distant-dependent dielectric constant were conducted by using the SANDER module of the AMBER7.0 simulation package (University of California) with the PARM98 force-field parameter. The SHAKE algorithm (37) was used to keep all bonds involving hydrogen atoms rigid. Weak coupling temperature and pressure coupling algorithms (38) were used to maintain constant temperature and pressure, respectively. MD simulations were performed using 0.001 picosecond time steps with temperature set at 300 K. Electrostatic interactions were calculated with the Ewald particle mesh method (39) with a dielectric constant at 1Rij and a nonbonded cutoff of 12 Å for the real part of electrostatic interactions and for van der Waals’ interactions. Structural analyses were done using the SYBYL 7.0 (Tripos) molecular modeling program.

Population study

Blood samples from 33 unrelated African-Americans were obtained with consent. Genomic DNA was isolated using either a QIAamp DNA Blood Mini Kit (Qiagen) or a Pel-Freez DNA Isolation Kit (Pel-Freez Clinical Systems), following the manufacturer’s protocol. KIR2DL2-specific amplification was performed using the sense primer 2DL2-E4F (5'-gagtccacagaaaaccttcc-3') and the antisense primer 2DL2-E5R (5'-gccctgcagagaacctaca-3'). The amplicon contained most of exons 4 and 5, as well as the intervening intron. The PCR (50 µl) contained genomic DNA (200–500 ng), TaqDNA polymerase High Fidelity (Invitrogen Life Technologies), 1x high-fidelity buffer (Invitrogen Life Technologies), 2 mM MgSO4, 200 µM dNTP, 0.4 µM each primer, and 2.5% DMSO. Amplification was performed on a PTC-225 (MJ Research) thermal cycler and consisted of an initial denaturation (95°C, 5 min), followed by a step-down PCR that consisted of five cycles of denaturation (95°C, 30 s), annealing (62°C, 45 s), and extension (68°C, 2 min); then 30 cycles of denaturation (95°C, 30 s), annealing (59°C, 45 s), and extension (68°C, 2 min); and a final product extension (68°C, 10 min) for one cycle. The KIR2DL4-specific PCR amplification has been described previously (22). KIR2DL4 was amplified in all KIR2DL2-negative samples to verify DNA quality. Reactions were purified using Microcon-100 (Millipore) and digested for 2 h with the restriction enzyme DraII. The digests were electrophoresed on a 2.5% agarose gel. The KIR2DL2*004 allele, as indicated by a 181-bp fragment, was confirmed by sequencing. To segregate unresolved alleles, the KIR2DL2 amplicon was cloned into the pCR2.1 TOPO TA vector (Invitrogen Life Technologies) for sequencing. Sequencing was performed using BigDye Terminator Ready Reaction mix (Applied Biosystems), according to the manufacturer’s protocol, on an Applied Biosystems Prism 377 DNA Sequencer. The data were analyzed with Sequencher software (Gene Codes). Use of human subjects was approved by the local institutional review board.

Phylogenic analysis

Construction of the phylogenetic tree was done using PAUP version 4 (Sinauer Associates). An unrooted phylogenetic tree was constructed using the neighbor-joining method, based on the amino acid alignments of the KIR2DL2/3 allelic products’ extracellular domains.

Immunoprecipitation and Western blotting

For immunoprecipitation, Jurkat cells were transiently transfected with the appropriate construct. Eighteen hours later, 2 x 107 cells were washed in PBS and then lysed in radioimmunoprecipitation assay buffer containing protease inhibitors (Protease Inhibitor Set III; EMD Biosciences). Cell lysates were split for immunoprecipitation with either 2 µg of anti-V5 or 4 µg of CD158b Ab with protein GPlus-agarose (EMD Biosciences). Immunoprecipitated protein was denatured in peptide-N-glycosidase (PNGase) F digestion buffer (New England Biolabs) and then split for mock treatment or digestion with PNGase F, according to the manufacturer’s instructions. The protein samples were reduced and denatured in 2x Laemmli buffer and run on 10% polyacrylamide Tris-HCl Ready gels (Bio-Rad). The gel protein was blotted to nitrocellulose, and the membrane was blocked for 2 h using 5% (w/v) nonfat dry milk. Blots were probed with a 1/5,000 dilution of anti-V5 (Invitrogen Life Technologies) primary Ab and a 1/20,000 dilution of secondary Ab directed against the mouse {kappa} L chain (Serotec). Protein bands were detected using ECL detection (Amersham Biosciences).


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosures
 References
 
The KIR2DL2*004 allele is present in the African-American population

KIR2DL2*004 was originally described in an African-American family (30), but its frequency has not been well defined. To determine the approximate frequency of KIR2DL2*004 in the African-American population, a fragment (~2000 bp) containing exon 4 and most of exon 5 was amplified from genomic DNA for a small (n = 33) population of unrelated individuals. KIR2DL2 was present in 17 samples (~52%), comparable to the frequency of KIR2DL2 in similar populations (40, 41, 42). The KIR2DL2 amplicons were screened using a DraII restriction digest site, initially believed unique to the KIR2DL2*004 allele, which results in a 181-bp fragment (data not shown). Two samples were indicated as containing KIR2DL2*004 and were sequenced. One of the samples contained KIR2DL2*004 (~6% of KIR2DL2-positive individuals), and the other sample contained a previously unreported KIR2DL2 allelic variant (PMID: DQ145175). The KIR2DL2*004 allele is therefore present in the African-American population as a whole and is not restricted to a particular family.

KIR-CD3{zeta} activation assay demonstrates failure of the KIR2DL2*004 extracellular domain to interact with putative HLA ligands

The KIR2DL2*004 extracellular domain varies from other KIR2DL2/3 allelic products by between 3 and 6 aa (Fig. 1). Compared with KIR2DL3*001, the KIR2DL2*004 allelic product differs by only 3 aa in this region; however, each polymorphic residue contributes an ionic charge difference between the allelic products. To determine whether these differences modify the ligand recognition properties of KIR2DL2*004, BW5147 effector cells were transfected with 2DL3*001-CD3{zeta}, 2DL2*004-CD3{zeta}, and 2DL1-CD3{zeta} fusions. The KIR-CD3{zeta} fusion proteins stimulate effector cell transfectants to produce IL-2 in the presence of the activating ligands, thus allowing receptor activation to be measured by assay for IL-2. In the presence of 721.221 target cells bearing no surface HLA-C, minimal IL-2 was detected from all transfectants (Fig. 2). This indicates that effector cells were not nonspecifically activated in the absence of ligand. Against target cells expressing HLA-Cw*0401, a ligand for KIR2DL1, but not KIR2DL2/3, IL-2 could be detected from cells transfected with 2DL1-CD3{zeta}, but not 2DL3*001-CD3{zeta} or 2DL2*004-CD3{zeta} (Fig. 2). This result suggests that the 2DL1-CD3{zeta} fusion functioned as expected and that the 2DL3*001-CD3{zeta} and 2DL2*004-CD3{zeta} fusions did not interact with HLA-Cw*0401. Against target cells expressing HLA-Cw*0102, Cw*1202, and Cw*1403, all putative KIR2DL2/3 ligands, activation was seen in cells transfected with 2DL3*001-CD3{zeta}, but not 2DL2*004-CD3{zeta} or 2DL1-CD3{zeta} (Fig. 2).


Figure 1
View larger version (13K):
[in this window]
[in a new window]
  		FIGURE 1. The KIR2DL2*004 extracellular domain contains novel amino acid polymorphisms that differentiate it from the extracellular domains of other KIR2DL2/3 allelic products. A, Extracellular amino acid differences that distinguish KIR2DL2 and KIR2DL3 allelic products from the KIR2DL3*001 allelic product. This alignment of all known KIR2DL2/3 extracellular coding regions (D1 = domain 1, D2 = domain 2, S = stem) is based on data obtained from the Immuno Polymorphism Database (43 ). Numbering of positions is defined by the first residue of the mature full-length protein. The alignments of the KIR2DL3*001 and KIR2DL2*004 extracellular domains examined in this study are in boldface. Classification of KIR2DL2 vs KIR2DL3 is based on differences in the length of the intracellular tail encoded by exon 9. B, Phylogram based on the extracellular amino acid alignments of KIR2DL2/3 allelic products. A neighbor-joining tree was used to generate distances between sequences, represented by the phylogram branch lengths.

 

Figure 2
View larger version (16K):
[in this window]
[in a new window]
  		FIGURE 2. Effector cell IL-2 secretion indicates activation of KIR-CD3{zeta} fusion receptors by ligand. BW5147 effector cells were transfected with 2DL3*001-CD3{zeta}, 2DL2*004-CD3{zeta}, and 2DL1-CD3{zeta} fusion receptors. Negative control effector cells were transfected with an empty vector. Transfectants were incubated with target cells expressing putative KIR2DL2/3 ligands HLA-Cw*0102, Cw*1202, or Cw*1403, a putative KIR2DL1 ligand HLA-Cw*0401, or no HLA-C (.221). Cells were cotransfected with a GFP reporter vector, and IL-2 secretion levels (pg/ml) were measured and then normalized by dividing by the posttransfection percentage of GFP-positive effector cells (assay performed in triplicate).

 
The absence of 2DL2*004-CD3{zeta} activation in the presence of three putative ligands was unexpected and may have had several causes: the 2DL2*004-CD3{zeta} protein may not be present on the cell surface because 1) it was not produced, 2) it was produced, but failed to localize to the cell surface, 3) it was quickly degraded, or 4) the 2DL2*004-CD3{zeta} protein was present on the cell surface, yet did not interact properly with its putative ligands. The 2DL3*001-CD3{zeta}, but not 2DL2*004-CD3{zeta} transfectants could be surface stained with two commercially available KIR2DL2/3-specific mAbs (NKAT2, data not shown; CD158b, discussed below) suggesting that the 2DL2*004-CD3{zeta} protein was not present on the cell surface (scenarios 1–3). However, the fusion may have been present on the cell surface (scenario 4) if the KIR2DL2*004 extracellular domain was misfolded, because this would have disrupted both the Ab and ligand binding regions of the receptor.

Full-length humanized KIR2DL2*004 is produced, but retained within the cell

To determine whether full-length KIR is expressed at the surface, the CD3{zeta} domain was replaced with the continuing portion of KIR2DL2 taken from KIR2DL2*001 cDNA, creating the full-length humanized KIR constructs 2DL3*001F and 2DL2*004F. To test these constructs in a human system, they were transfected into Jurkat cells with a GFP expression vector. After 16 h, the cells were stained with KIR2DL2/3-specific mAb (CD158b). Flow cytometry analysis was gated on GFP-positive cells to increase detection sensitivity. Cells transfected with 2DL3*001F demonstrated high levels of CD158b staining, while cells transfected with 2DL2*004F did not (Fig. 3A). When transfection and staining were performed in triplicate, the CD158b staining of 2DL2*004F transfectants was not statistically greater than staining of GFP-only transfectants (Fig. 3B; mean fluorescence intensity (MFI) 6.1 ± 0.2 vs 5.4 ± 0.7; p = 0.28 by Student’s t test). These data indicate that either the 2DL2*004F protein was not present on the cell surface (scenarios 1–3) or that polymorphism may affect recognition of 2DL2*004F by CD158b (scenario 4).


Figure 3
View larger version (20K):
[in this window]
[in a new window]
  		FIGURE 3. The full-length humanized KIR2DL2*004 receptor (2DL2*004F) cannot be detected by KIR2DL2/3-specific CD158b surface staining. A, Jurkat cells cotransfected with GFP and 2DL3*001F or 2DL2*004F were surface stained with CD158b; negative control cells were transfected with GFP alone. B, The MFI of 2DL2*004F transfectants was compared with that of negative control transfectants (three replicates; 6.1 ± 0.2 vs 5.4 ± 0.7; p = 0.28 by Student’s t test). C, Cells transfected with C-terminally V5-tagged constructs were stained at the surface with CD158b and then intracellularly stained with anti-V5 mAb; mock-transfected cells were used as negative control.

 
To determine whether 2DL2*004F was being produced as a stable protein, constructs with C-terminal V5 tags were created (2DL3*001F-V5 and 2DL2*004F-V5). Jurkat cells were transiently transfected with these constructs and were surface stained with CD158b and intracellularly stained with V5-specific mAb. Flow cytometry analysis of 2DL3*001F-V5 transfectants indicated that most V5-stained cells also demonstrated CD158b staining (Fig. 3C). For cells transfected with 2DL2*004F-V5, V5-specific staining could be seen, but CD158b staining was absent. Positive intracellular V5 staining of 2DL2*004-V5 transfectants demonstrates that the 2DL2*004F-V5 protein can be produced and is stable 18 h after transfection. However, the lack of cell surface staining suggests either a disrupted epitope (scenario 4) or retention of the receptor within the cell (scenario 2).

To distinguish between these two possibilities, the cellular localization of 2DL2*004F-V5 was determined by confocal microscopy. Jurkat cells were transfected with either 2DL3*001F-V5 or 2DL2*004F-V5 and then stained with green DiO (left panels, Fig. 4), which stains lipid membranes such as the plasma membrane. A red V5-specific (center panels, Fig. 4) fluorescent stain was used to localize the V5-tagged proteins. In cells transfected with 2DL3*001F-V5, colocalization of the V5-specific staining with the lipophilic staining of the plasma membrane was visualized as a yellow color (right panels, Fig. 4, A and B). This is the result of 2DL3*001F-V5 protein accumulation at the plasma membrane. V5 staining of the 2DL2*004F-V5 transfectants was perinuclear and did not colocalize with the plasma membrane stain, suggesting retention of the tagged protein within the endoplasmic reticulum (Fig. 4, C and D). Therefore, it is likely that one or some combination of the three polymorphic amino acids in the extracellular domain of KIR2DL2*004 causes intracellular retention (scenario 4). This intracellular retention is responsible for the inability to surface-stain proteins containing the KIR2DL2*004 extracellular domain (Fig. 3).


Figure 4
View larger version (23K):
[in this window]
[in a new window]
  		FIGURE 4. Confocal microscopy indicates that 2DL2*004F-V5 is retained within the cell. Jurkat cells were transiently transfected with 2DL3*001F-V5 (A and B) or 2DL2*004F-V5 (C and D). Staining of transfectants was performed with lipophilic DiO staining (left panels) and V5-specific staining (middle panels). Merged images (right panels) reveal perinuclear retention of 2DL2*004F-V5, while 2DL3*001F-V5 colocalizes with the plasma membrane (yellow color).

 
The KIR2DL2*004 phenotype is solely caused by the T41 polymorphism

KIR2DL2*004 differs from KIR2DL3*001 at three positions: 41(R->T), 148(R->C), and 167(G->D) (Fig. 1A). To determine which amino acids were responsible for the lack of surface expression, mutations were cumulatively introduced into the 2DL3*001F construct to create the 2DL3*001(R148C)F, 2DL3*001(R148C,R41T)F, and 2DL3*001(R148C,R41T,G167D)F (i.e., 2DL2*004F) constructs. These constructs were cotransfected with a GFP expression vector into Jurkat cells and then surface stained with CD158b. Cells transfected with the 2DL3*001F and 2DL3*001(R148C)F constructs could be stained with CD158b, but cells transfected with the 2DL3*001(R148C,R41T)F and 2DL2*004F constructs could not be stained (Fig. 5A). Receptors containing T41 could not be stained, suggesting that T41 had some role in disrupting surface expression of these receptors.


Figure 5
View larger version (28K):
[in this window]
[in a new window]
  		FIGURE 5. The R41T polymorphism abrogates receptor surface detection by the KIR2DL2/3-specific CD158b mAb. A, Jurkat cells cotransfected with GFP and 2DL3*001F or sequentially mutated 2DL3*001F constructs were stained at the surface with CD158b. Cells cotransfected with GFP and 2DL3*001F with and without the R41T mutation (B) or 2DL2*004F with and without the T41R mutation (C) were stained at the surface with CD158b. Cells transfected with GFP alone were used as a negative control. D, Cells transfected with C-terminally V5-tagged constructs were stained at the surface with CD158b and then intracellularly stained with anti-V5 mAb; mock-transfected cells were used as negative control. E and F, BW5147 effector cells were transfected with 2DL3*001(R41T)-CD3{zeta} and 2DL2*004(T41R)-CD3{zeta} fusion receptors. Negative control effector cells were transfected with an empty vector. Transfectants were incubated in the presence (E) or absence (F) of PMA with target cells expressing putative KIR2DL2/3 ligands HLA-Cw*0102, Cw*1202, or Cw*1403, or expressing no HLA-C (.221). Cells were cotransfected with a GFP reporter vector, and IL-2 secretion levels (pg/ml) were measured and then normalized by dividing by the posttransfection percentage of GFP-positive effector cells (assay performed in triplicate). G, Effector cells used in F were surface stained with PE-conjugated CD158b and analyzed by flow cytometry.

 
To establish whether T41 alone was sufficient to disrupt surface staining, the polymorphism was introduced into the 2DL3*001F construct, which prevented its surface staining (Fig. 5B). These data eliminate the possibility of C148 acting together with T41 to disrupt surface expression. To exclude the prospect of D167 contributing to the KIR2DL2*004 phenotype, a T41R mutation was introduced into the 2DL2*004F construct. Detection by CD158b surface staining was possible (Fig. 5C), leaving T41 as the polymorphism solely responsible for the lack of surface expression.

To verify that 2DL3*001(R41T)F was being produced, the 2DL3*001(R41T)F and 2DL2*004(T41R)F constructs were created with C-terminal V5 tags and transfected into Jurkat cells. After CD158b surface staining and V5-specific intracellular staining, flow cytometry analysis indicated positive V5 staining for both 2DL3*001(R41T)F and 2DL2*004(T41R)F transfectants, indicating that both proteins were produced. However, no CD158b surface staining was detectable for cells transfected with the 2DL3*001(R41T)F construct. Therefore, it is unlikely that the absence of staining for 2DL3*001(R41T)F can be attributed to a lack of protein production.

Because R41 appeared to be a critical amino acid, it was most likely responsible for the absence of 2DL2*004-CD3{zeta} activation in the presence of putative KIR2DL2/3 ligands (Fig. 2). To investigate this possibility, two mutated constructs were created: 2DL3*001(R41T)-CD3{zeta} and 2DL2*004(T41R)-CD3{zeta}. When these constructs were expressed in effector cells and then tested against the same ligands, 2DL2*004(T41R)-CD3{zeta} transfectants were activated by putative KIR2DL2/3 ligands, but no such activation was seen in effector cells transfected with 2DL3*001(R41T)-CD3{zeta} (Fig. 5E). This is probably because KIR-CD3{zeta} fusions containing the T41 polymorphism are not present on the cell surface for interaction with ligand. The activation of 2DL2*004(T41R)-CD3{zeta} transfectants suggests that a KIR with both the C148 and D167 polymorphisms can still interact with ligand from the cell surface. When the assay was performed in the absence of PMA, the 2DL2*004(T41R)-CD3{zeta} transfectants did not show significant activation against HLA-Cw*0102 or HLA-Cw*1403 ligands, but were still activated against HLA-Cw*1202 (Fig. 5F). Because the 2DL2*004(T41R)-CD3{zeta} transfectants consistently secreted less IL-2 than the 2DL3*001-CD3{zeta} transfectants, it is possible that the other polymorphisms, 148(R->C) and/or 167(G->D), may affect receptor-binding affinity. However, because the T41 polymorphism alone causes the loss of receptor surface expression, the binding affinity difference was not further explored.

To prove that T41 alone results in the intracellular retention of these receptors, the C-terminal V5-tagged constructs 2DL3*001(R41T)F-V5 and 2DL2*004(T41R)F-V5 were transfected into Jurkat cells for examination by confocal microscopy. In cells transfected with 2DL3*001(R41T)F-V5 (Fig. 6, A and B), the protein (red stain) was distributed intracellularly, similar to what was previously seen in cells transfected with 2DL2*004F-V5 (Fig. 4, C and D). In cells transfected with 2DL2*004(T41R)F-V5 (Fig. 6, C and D), the protein colocalized with the plasma membrane (yellow stain), similar to 2DL3*001F-V5 (Fig. 4, A and B). These data are consistent with the threonine residue at codon 41 being solely responsible for the intracellular confinement of the proteins.


Figure 6
View larger version (22K):
[in this window]
[in a new window]
  		FIGURE 6. The introduction of T41 causes intracellular retention of 2DL3*001F. Jurkat cells were transiently transfected with 2DL3*001(R41T)F-V5 (A and B) or 2DL2*004(T41R)F-V5 (C and D). Staining of transfectants was performed with lipophilic DiO staining (left panels) and anti-V5 staining (middle panels). Merged images (right panels) reveal perinuclear retention of 2DL3*001(R41T)F-V5, while 2DL2*004(T41R)F-V5 colocalizes with the plasma membrane (yellow color).

 
There are several reasons that the T41 polymorphism may result in the receptor being intracellular confined. The polymorphic amino acid may disrupt interaction of the newly synthesized receptor with molecular chaperones that may be required for the proper folding of the receptor. KIR molecules have the ability to multimerize (44), and it is possible that T41 disrupts a homophilic interaction that is required for KIR surface expression. It is unknown whether KIR2DL2/3 must bind other proteins to be carried to the cell surface; T41 could possibly block such binding. Finally, R41 may form critical interactions with adjacent amino acid residues. If unable to form similar interactions, T41 may prevent the receptor from folding properly or maintaining its proper conformation once folded.

Molecular modeling predicts that R41 may interact with Y77, D47, and D72

KIR2DL2 and KIR2DL3 have two domains (D1 and D2) that are folded into a secondary structure of nine beta strands (A, A', B, C, C', D, E, F, and G) (34). Because the perinuclear distribution of T41-containing receptors is indicative of protein misfolding, molecular modeling was used to examine the importance of residue 41 to KIR extracellular domain stability. The crystal structure of a KIR2DL2*001 extracellular domain (34) was energy minimized with and without a R41T mutation (Fig. 7, B and A, respectively), and the possible interactions of residue 41 with adjacent residues were analyzed. A hydrogen bond interaction between the R41 and the Y77 side chains, present in the C and F strands of domain D1, respectively, may maintain the contact between these two adjacent beta strands. Although R41 is not properly positioned to make a traditional cation-{pi} interaction with Y77, a splayed cation-{pi} conformation cannot be ruled out (45). The R41 C{delta} atom could possibly provide a favorable hydrophobic contact with Y77. A salt bridge between R41 and D72 (near the F strand) may further stabilize the interaction between the C and F strands.


Figure 7
View larger version (52K):
[in this window]
[in a new window]
  		FIGURE 7. Molecular modeling predicts that T41 results in extracellular domain instability. KIR2DL2(R41) (A) and KIR2DL2(T41) (B) energy-minimized structures were generated for the previously solved KIR2DL2 crystal structure (PDB: 2DL2). Yellow represents carbon atoms, red represents oxygen atoms, and blue represents nitrogen atoms. The shaded coloring surrounding each atom represents the molecular surface. The {Delta}E values indicate the energy change of each structure after energy minimization. C, A wider perspective of the KIR crystal structure region of interest. D, MD simulations of the KIR2DL2(T41) structure indicate a significant movement (>3Å RMSD) of shown residues over 3 ns. Yellow residues are shown in their initial positions, with the overlaid white residues indicating positions in the MD-simulated structure.

 
Residue 41 may also help maintain contact between the C and C' strands, because R41 most likely forms a hydrogen bond with the backbone oxygen of F45 (on the C' strand). This interaction would be present in both R41 and T41 structures. In addition, R41 may form a water-mediated hydrogen bond contact with residue D47 (present on the C' strand). However, a salt bridge between D47 and R41 cannot be ruled out, due to the relatively low resolution (3.0 Å) of the KIR2DL2 crystal structure. Thus, either indirect interaction through water or a salt bridge between D47 and R41 would stabilize contact between the C and C' strands.

The KIR2DL3 (PDB: 1B6U) and KIR2DS2 (PDB: 1M4K) receptors have also been crystallized; both have great sequence similarity to KIR2DL2, and the KIR2DS2 crystal was solved at a high resolution (2.30 Å) (35, 36). Therefore, the region of interest was also examined in these structures. For KIR2DL3 and KIR2DS2, R41 is likely to form a hydrogen bond with Y77 and also with the backbone oxygen of F45, as was seen for the KIR2DL2 structure. The D72 residue is more distant from R41 in both the KIR2DL3 and KIR2DS2 structures (O—H...N being 4.81 and 5.33 Å, respectively) compared with the KIR2DL2 structure (O—H...N being 3.11 Å), which may indicate that a R41-D72 interaction is not crucial for the D1 domain. In the KIR2DS2 structure, a water molecule most likely interacts with R41 (O...H—N being 2.02 and 2.21 Å), but is far away (O—H...O is 3.61 Å) from D47, thus making any water-mediated contact between R41 and D47 unlikely. Moreover, the distance between R41 and D47 is too great to allow for hydrogen bonding. For KIR2DL3, no water molecule is located near R41, but this may be due to the structure’s low resolution (3.0 Å) relative to that of KIR2DS2.

If in the KIR2DL2 structure a threonine is substituted for arginine at codon 41 (Fig. 7B), the residue is uncharged and cannot form a salt bridge with D72 or D47. T41 also lacks the C{delta} atom, and therefore also lacks any favorable hydrophobic interaction between R41-C{delta} and Y77. The mutant structure was examined using a single degree of freedom by considering the three possible C{alpha}-Cbeta rotamers of T41. For one rotamer, the methyl group creates either a steric clash or a hydrophobic interaction with Y77. In either case, this structure would lack a hydrogen bond between T41 and Y77. For the second rotamer, the threonine side chain is not lengthy enough to form a hydrogen bond with Y77, as is predicted in the arginine side chain. For the third rotamer, hydrogen may have a hydrophobic interaction with Y77, but this structure would lack a hydrogen bond between T41 and Y77. Therefore, all three T41 rotamers are predicted to be less stable than the R41 structure. Energy calculations predicted that the KIR2DL2(R41) structure is stable with a {Delta}G of –89.8 kcal/mol, whereas the predicted {Delta}G for the KIR2DL2(T41) structure is +13.2 kcal/mol. This is most likely due to the absence of interaction of R41 with Y77, D72, and/or D47. Molecular surface analysis of the KIR2DL2(T41) structure (Fig. 7B) shows a comparatively blank space in the region between strands, indicative of decreased interactions between the C strand and the C' and F strands.

In addition to the structural analyses, MD simulations were used to study the influence of KIR2DL2(T41) on the folding of the D1 domain. Three nanoseconds of MD simulations were conducted for KIR2DL2(T41) with a distant dependent dielectric constant. The root mean square deviation between the initial and simulated positions of the C-{alpha} atoms in the interconnecting strand region fluctuates around an average value of 3 Å, with some 3.5 Å deviations (Fig. 7D). It is evident in the simulated structure that residues on the adjacent strands moved away from each other. By comparison, MD simulations generated from the KIR2DL2(R41) structure did not demonstrate significant residue movement. In the final MD structure, the beta-sheet remains intact due to the interaction of residues along the C and F strands. The strand movement found in the KIR2DL2(T41) simulation suggests that R41, compared with T41, provides a stronger interaction between the C and F strands. This interaction may be critical for the completion of the D1 domain beta-sheet structure. If KIR2DL2(T41) lacks a strong enough interaction between the C and F strands, the D1 domain may not fold into its regular beta-sheet conformation, resulting in disruption of molten globule formation. In addition, the water-mediated contact between D47 and residue 41 was disrupted in the final structure, because water maintained a hydrogen bond with T41, while the D47 residue moved away from the site. Therefore, there may also be a weakened interaction between the C and C' strands of the D1 domain. Predicting the exact mechanism of D1 domain instability by current MD simulation methods is difficult because the inaccuracies in the model and force-field parameters are large relative to the few kcal/mol that are needed to make the difference between a native and unfolded structure.

Mutation studies indicate that Y77 and D47 are critical residues

If the interaction of residues D72, D47, and/or Y77 with R41 is important for maintaining KIR extracellular domain structure, elimination of interacting chemical groups should result in a phenotype similar to the R41T mutation. Therefore, 2DL3*001F-V5 was mutated to remove both the Y77 cation-{pi} and hydrogen bond interactions (Y77L), only the Y77 hydrogen bond interaction (Y77F), the D72 salt bridge (D72L), or any interaction with residue 47 (D47L). Jurkat cells were transiently transfected with each construct, and surface expression levels were determined. Transfectants were surface stained with CD158b mAb and intracellularly stained with V5-specific mAb. Flow cytometry analysis was performed by gating only on V5-positive cells, and the MFI of CD158b staining was measured (Fig. 8, A and B). Both mutations to residue Y77 resulted in a significant decrease in CD158b staining, with the Y77L mutation causing a complete lack of CD158b surface staining. Decreased CD158b staining may result from epitope perturbation or a diminished amount of protein at the cell surface. The Y77F mutation lacks only a single hydroxyl group, and this small change significantly decreased the amount of CD158b staining. The Y77F mutant protein could still be detected at the cell surface, possibly due to the phenyl ring of phenylalanine forming a cation-{pi} interaction with R41. These interactions are absent in the Y77L mutant protein, which could not be stained at the cell surface. This suggests that a cation-{pi} interaction may exist in the wild-type structure, or at very least maintains some level of surface expression in the absence of hydrogen bonding between R41 and Y77. Mutation of the D72 residue did not significantly decrease CD158b staining, indicating that the potential salt bridge between D72 and R41 is not critical to receptor surface expression nor epitope integrity. The D47L mutant could be detected at the cell surface, albeit at a significantly decreased level compared with the wild-type receptor; therefore, an interaction between D47 and R41 may be important.


Figure 8
View larger version (19K):
[in this window]
[in a new window]
  		FIGURE 8. Mutagenesis of amino acids that potentially interact with R41 results in decreased KIR2DL2/3-specific CD158b immunoprecipitation and surface staining. Jurkat cells were transfected with the V5-tagged construct 2DL3*001F-V5 or the same construct with the mutation R41T, Y77F, Y77L, D47L, or D72L. Cells were surface stained with CD158b and intracellularly stained with V5-specific mAb. A, The histograms show the intensity of CD158b staining of V5-positive cells, with surface-unstained cells as negative controls. B, The mean (three replicates) CD158b/V5 MFI of each transfectant group was compared with that of the 2DL3*001F transfectants for statistical significance (***, p < 0.001 by Student’s t test). C, The wild-type and mutant receptors were immunoprecipitated by a V5-specific Ab and subjected to overnight PNGase F digestion (+) or mock treatment (–). D, The wild-type and R41T mutant receptor were immunoprecipitated by the CD158b or V5-specific Abs without PNGase F treatment. E, The wild-type and mutant receptors were immunoprecipitated by the CD158b or V5-specific Abs and treated with PNGase F overnight. The specific band of interest is ~45 kDa and corresponds with the PNGase F-digested bands in C. F, Biochemical sketch of R41 and potentially relevant surrounding amino acid residues, based on the 2DL2 (PDB: 1EFX) crystal structure. Dotted lines indicate most likely the critical interactions and distances between residues. Dashed lines indicate the distance between nearby residues less likely to interact critically with R41.

 
To provide further evidence that the intracellular receptor’s D1 domain is misfolded, both wild-type and mutant receptors were expressed in Jurkat cells and then immunoprecipitated with the CD158b and V5-specific Abs. Immunoprecipitated protein was run on a protein gel and detected using the V5-specific Ab. When immunoprecipitated by the V5-specific Ab, the wild-type receptor appeared as two specific bands, a strong and diffuse band of the expected size (58 kDa) and a faint and sharp lower molecular mass band (~50 kDa). Immunoprecipitation of the R41T, Y77F, and Y77L mutant receptors using the V5-specific Ab yielded only the lower molecular mass band, suggesting that this band represents the immature receptor that has not finished posttranslational modification (Fig. 8C, untreated lanes). This would be consistent with the mutant receptors being trapped in the endoplasmic reticulum because they would not reach the Golgi apparatus for further glycosylation. N-glycanase (PNGase F) digestions of the immunoprecipitated receptors were performed to show that the molecular mass difference is due to the addition of sugar molecules to the wild-type receptor. All PNGase F-digested immunoprecipitates could be visualized as a single specific band of ~45 kDa (Fig. 8C, treated lanes). Untreated immunoprecipitates of the D72L receptor resulted in two specific bands, similar to those of the wild-type receptor. These data are consistent with the CD158b surface-staining data (Fig. 8, A and B) and suggest that D72 is not a critical residue for maintenance of the D1 domain. Untreated immunoprecipitates of the D47L receptor also resulted in the same bands, although the 58-kDa band was faint. Because D47L could be weakly detected by CD158b surface staining (Fig. 8B), this suggests a correlation between CD158b surface staining and the amount of receptor glycosylation. It is unlikely that receptor glycosylation state is directly responsible for CD158 binding, because immunoprecipitation of the wild-type receptor using the CD158b Ab continues to yield the ~50-kDa band (Fig. 8D).

The CD158b Ab is able to immunoprecipitate the wild-type receptor, but is unable to immunoprecipitate the R41T mutant receptor (Fig. 8E). Therefore, the R41T receptor is not only trapped intracellularly, but also cannot bind the CD158b Ab well enough to be immunoprecipitated. The CD158b epitope may be disrupted in this receptor, or the receptor may be unavailable for immunoprecipitation because it is aggregated within the cell; both possibilities support that the receptor is not properly folded. The Y77L and D47L mutant receptors demonstrate a similar phenotype as the R41T receptor, while the Y77F and D72L receptors can only be weakly immunoprecipitated by CD158b (Fig. 8E). It is possible that D47L may also be weakly immunoprecipitated by CD158b, but was below the limit of detection. The D72L was expected to be immunoprecipitated by CD158b at a greater level; it is possible that the D72L structure is mildly altered such that it can be glycosylated and stained by CD158b Ab, but not efficiently immunoprecipitated by the same Ab.

The F-loop (G/A)XYXC motif is conserved among other Ig superfamily members

The KIR extracellular domains are similar to members of the C2-type Ig-like receptor family (2). Interestingly, amino acid alignments of other Ig superfamily members reveal a conserved F strand sequence motif of (G/A)XYXC, similar to the conserved KIR F strand LAGTYRCYGS sequence (46). Examination of these alignments also indicates the presence of an arginine or lysine residue in the C strand region for many Ig superfamily members, commonly in a WX(R/K) motif. The WX(R/K) tryptophan is conserved among many Ig superfamily members, but is not present in the KIR family. To determine whether the conservation of these residues had potential functional significance, molecular modeling of a representative subgroup of available Ig superfamily member crystal structures was performed. For human Ig H chain (PDB: 1OL0) (47) and murine Ig {lambda} L chain V domain (PDB: 1P4B) (48) structures (V-set Ig-like family members), a potential hydrogen bond exists between the F strand (G/A)XYXC motif tyrosine and a C strand arginine (positions Y94/R38 and Y104/R45, respectively). A similar potential hydrogen-bonding interaction also exists in a structure for the C2-set domain of human CD3{epsilon} (PDB: 1XIW) (49) between residues Y94 and K38.

Although the (G/A)XYXC tyrosine appears to form interactions between the F and C strands in the domains of several Ig superfamily members, this function is seemingly not conserved among all members. For instance, the C1-set {alpha}3 domain of an HLA-B structure (PDB: 1ZSD) (50) lacks a potential interaction between the F strand (G/A)XYXC and a C strand arginine or lysine. In addition, the structure for a human Fab H chain (PDB: 1L7I) (51) has no interaction present in the C domain, although its V domain contains a potential hydrogen bond between R38 and Y90. Therefore, although an Ig superfamily member may contain these conserved residues, an interaction between them should be suspected, but not assumed.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosures
 References
 
In this study, introduction of T41 resulted in the disruption of both KIR-CD3{zeta} fusion and full-length KIR proteins in mouse and human cell lines, respectively. These data indicate that R41 is required to preserve the basic biochemical integrity of the KIR2DL2/3 D1 domain. Examination of the D1 domain amino acid alignment for all described KIR alleles reveals a conserved LHREG sequence, with lysine being substituted for arginine in the KIR3DS1 and KIR3DL1 receptors. Because lysine and arginine possess similar biochemical properties, it is likely that this is a conserved D1 domain amino acid position. This study predicts that the interaction of R41 with Y77 and/or D47 is critical to proper D1 domain folding. The tyrosine residue is also conserved among all known KIR alleles and is present in the sequence LAGTYRCYGS, 34 residues C-terminal of the LH(R/K)EG sequence. These two conserved residues provide an interstrand interaction between the C and F strands of the D1 domain, and disruption of this interaction may prematurely terminate folding of the KIR extracellular domain. The D47 residue is not conserved among the KIR genes encoding a D1 domain; instead, the aspartic acid is either an asparagine or histidine for all known KIR2DS4 and KIR2DS5 allelic products, respectively. The influence that this polymorphism has on the KIR2DS4/5 structure is unknown.

The Y77F mutation maintains any potential cation-{pi} interaction of residue 77 with R41, while removing its ability to form hydrogen bonds (Fig. 8F). This was found to be disruptive to the KIR extracellular domain structure, but the receptor was still detectable on the cell surface at reduced levels. It is possible that a cation-{pi} interaction forms only in the Y77F mutant to partially rescue the structure in the absence of the hydrogen bond between R41 and Y77. Introduction of leucine at position 77 disrupts the ability of the residue to form a potential cation-{pi} interaction with R41 and resulted in a complete lack of cell surface receptor detection. These data strongly suggest that an interaction between Y77 and R41 is required for maintenance of the KIR D1 domain and that both a cation-{pi} interaction and a hydrogen bond can occur between these residues, although the hydrogen bond is preferred. These residues are also generally conserved among Ig superfamily members and in several members appear to form a hydrogen bond. However, no cation-{pi} interactions were found between the conserved residues, making this type of interaction unique to the KIR D1 domain. Therefore, the interaction between the conserved tyrosine and arginine/lysine residues seems more crucial in the KIR D1 domain compared with the other Ig-like domains.

In addition to the interaction between Y77 and R41, molecular modeling predicted a possible interaction between R41 and D72 and/or D47 (Fig. 7). A triad interaction consisting of a hydrogen bond, cation-{pi}, and salt-bridge interaction between R41, Y77, and either D72 or D47 would be similar to other amino acid triad interactions that have been reported (52, 53, 54). However, the D72 residue alone appears to make no contribution to the stability of the KIR D1 domain (Fig. 8). D72 has been described previously as having a role at the KIR-HLA interface, which would spatially preclude an interaction between residues D72 and R41 (10). When the D47 residue was mutated, the receptor could not be stained as well at the surface, tended not to become fully glycosylated, and was not immunoprecipitated by the CD158b Ab (Fig. 8). These data indicate that D47 is a critical D1 domain residue, and this may be due to its interaction with R41 through either a salt bridge or the neighboring water molecule. Furthermore, the electrostatic interaction between R41 and D47 may extend far enough to impact protein stability even in the absence of a hydrogen bond between the two residues.

Although the presence of R41T eliminates receptor surface expression, alternative splicing may permit the surface expression of a KIR2DL2*004 receptor that lacks the D1 domain. Alternatively spliced KIR transcripts lacking the D1 domain exon have been described, although it is unknown whether these transcripts are translated into functional receptors (55). The KIR2DL2*004 D2 domain is stable when paired with a stable D1 domain, as indicated by CD3{zeta} activation in 2DL2*004(T41R)-CD3{zeta}-transfected effector cells exposed to target cells expressing C1 subgroup HLA molecules (Fig. 5E). If a KIR2DL2*004 transcript lacking the D1 domain is produced and encodes a stable protein, the disruptive T41 residue may serve to eliminate the D1 domain from the cell surface while maintaining expression of the D2 domain.

This is the second single amino acid polymorphism that has been found to eliminate KIR surface expression. The KIR3DL1*004 allelic product is retained almost entirely intracellularly, primarily due to a leucine-to-serine polymorphism (L86S) in the KIR3DL1 D0 domain (27). It is thought that L86S corrupts a WSXPS-like motif that has also been indicated in the fate of the erythropoietin receptor (56). KIR2DL2/3 do not possess a D0 domain, but their D1 and D2 domains contain similar WSXPS-like motifs in the loop region between the F and G strands. Because the location of this motif is distant from the C strand R41 residue, it is likely that R41 helps form a critical folding motif distinct from the WSXPS-like motif. It is curious that these KIR alleles encode receptors with no apparent function, possibly the result of evolutionary selection given the biological relevance of KIR in the regulation of NK cell activation.

The KIR2DL2 and KIR2DL3 genes are inherited as alleles, but KIR2DL2 has greater binding affinity than KIR2DL3 for some C1 subgroup members of HLA-C molecules (18). This affinity difference might result in KIR2DL2 having increased inhibitory activity and affect an individual’s ability to mount an immune response. In individuals exposed to low doses of hepatitis C, homozygosity for both KIR2DL3 and the HLA-C1 subgroup was significantly associated with increased viral clearance (15). It was hypothesized that the weakly interacting KIR2DL3 may provide less inhibition and thus allows for increased NK activity against hepatitis C-infected cells. Because the T41 polymorphism replaces the surface expression of a strongly interacting KIR2DL2 receptor with a noninteracting variant, certain KIR2DL3/KIR2DL2(T41) heterozygous individuals may have an advantage equal to or greater than that of KIR2DL3 homozygous individuals during hepatitis C infection. The KIR2DL2/3 genotype of a donor may also affect the recipient’s outcome in some hemopoietic stem cell transplants (16). Therefore, clinical studies and treatments should be cautious for KIR2DL2/3 variants, such as T41, which unexpectedly alter receptor expression.

Although highly similar to one another, the KIR genes have widespread diversity that has modified their products’ functions. In some instances, variations between the genes have led to differences in ligand specificity, ligand affinity, and intracellular signaling. In other instances, variations have created potentially dysfunctional receptors. It has been suggested that the variation found in the weakly binding KIR2DL3 may have resulted from evolutionary pressures selecting for a loss in NK cell inhibition (57). Perhaps the same pressure favors genetic polymorphisms that result in the functional loss of KIR gene products. This variation together with recombinatorial events may allow for the rapid modulation of KIR during evolution, such that receptors with greater or lesser affinities for certain ligands are created when the need arises.


   Acknowledgments
 
These studies were conducted using the Tissue Culture, Flow Cytometry and Cell Sorting, Macromolecular Analysis, and Microscopy and Imaging Shared Resources of Lombardi Cancer Center. Computing time and support were in part provided by the Advanced Biomedical Computing Center at the National Cancer Institute (Frederick, MD). We thank Gil Palchik and Karen Creswell for providing excellent technical assistance.


   Disclosures
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
References
 
The authors have no financial conflict of interest.


   Footnotes
 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1 Funding from the Office of Naval Research N00014-04-1-0398 and N00014-00-1-0898 to the C. W. Bill Young Marrow Donor Recruitment and Research Program supported this research. Back

2 The views expressed in this article are those of the authors and do not reflect the official policy of the Department of Navy, the Department of Defense, or the U.S. Government. Back

3 Address correspondence and reprint requests to Dr. Carolyn Katovich Hurley, Research Building Room E404, Georgetown University Medical Center, 3970 Reservoir Road N.W., Washington, DC 20057. E-mail address: hurleyc{at}georgetown.edu Back

4 Abbreviations used in this paper: KIR, killer Ig-like receptor; 2D, two domains; 3D, three domains; 2DL, 2D, long cytoplasmic tail; 3DL, 3D, long cytoplasmic tail; 2DS, 2D, short cytoplasmic tail; 3DS, 3D, short cytoplasmic tail; MD, molecular dynamics; MFI, mean fluorescence intensity; PNGase, peptide-N-glycosidase. Back

Received for publication October 25, 2005. Accepted for publication July 28, 2006.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 

  1. Vilches, C., P. Parham. 2002. KIR: diverse, rapidly evolving receptors of innate and adaptive immunity. Annu. Rev. Immunol. 20: 217-251. [Medline]
  2. Samaridis, J., M. Colonna. 1997. Cloning of novel immunoglobulin superfamily receptors expressed on human myeloid and lymphoid cells: structural evidence for new stimulatory and inhibitory pathways. Eur. J. Immunol. 27: 660-665. [Medline]
  3. Burshtyn, D. N., A. M. Scharenberg, N. Wagtmann, S. Rajagopalan, K. Berrada, T. Yi, J. P. Kinet, E. O. Long. 1996. Recruitment of tyrosine phosphatase HCP by the killer cell inhibitor receptor. Immunity 4: 77-85. [Medline]
  4. Campbell, K. S., M. Dessing, M. Lopez-Botet, M. Cella, M. Colonna. 1996. Tyrosine phosphorylation of a human killer inhibitory receptor recruits protein tyrosine phosphatase 1C. J. Exp. Med. 184: 93-100. [Abstract/Free Full Text]
  5. Lanier, L. L., B. C. Corliss, J. Wu, C. Leong, J. H. Phillips. 1998. Immunoreceptor DAP12 bearing a tyrosine-based activation motif is involved in activating NK cells. Nature 391: 703-707. [Medline]
  6. Biassoni, R., C. Cantoni, M. Falco, S. Verdiani, C. Bottino, M. Vitale, R. Conte, A. Poggi, A. Moretta, L. Moretta. 1996. The human leukocyte antigen (HLA)-C-specific "activatory" or "inhibitory" natural killer cell receptors display highly homologous extracellular domains but differ in their transmembrane and intracytoplasmic portions. J. Exp. Med. 183: 645-650. [Abstract/Free Full Text]
  7. Colonna, M., G. Borsellino, M. Falco, G. B. Ferrara, J. L. Strominger. 1993. HLA-C is the inhibitory ligand that determines dominant resistance to lysis by NK1- and NK2-specific natural killer cells. Proc. Natl. Acad. Sci. USA 90: 12000-12004. [Abstract/Free Full Text]
  8. Colonna, M., T. Spies, J. L. Strominger, E. Ciccone, A. Moretta, L. Moretta, D. Pende, O. Viale. 1992. Alloantigen recognition by two human natural killer cell clones is associated with HLA-C or a closely linked gene. Proc. Natl. Acad. Sci. USA 89: 7983-7985. [Abstract/Free Full Text]
  9. Fan, Q. R., E. O. Long, D. C. Wiley. 2001. Crystal structure of the human natural killer cell inhibitory receptor KIR2DL1-HLA-Cw4 complex. Nat. Immunol. 2: 452-460. [Medline]
  10. Boyington, J. C., S. A. Motyka, P. Schuck, A. G. Brooks, P. D. Sun. 2000. Crystal structure of an NK cell immunoglobulin-like receptor in complex with its class I MHC ligand. Nature 405: 537-543. [Medline]
  11. Wilson, M. J., M. Torkar, A. Haude, S. Milne, T. Jones, D. Sheer, S. Beck, J. Trowsdale. 2000. Plasticity in the organization and sequences of human KIR/ILT gene families. Proc. Natl. Acad. Sci. USA 97: 4778-4783. [Abstract/Free Full Text]
  12. Uhrberg, M., N. M. Valiante, B. P. Shum, H. G. Shilling, K. Lienert-Weidenbach, B. Corliss, D. Tyan, L. L. Lanier, P. Parham. 1997. Human diversity in killer cell inhibitory receptor genes. Immunity 7: 753-763. [Medline]
  13. Martin, M. P., X. Gao, J. H. Lee, G. W. Nelson, R. Detels, J. J. Goedert, S. Buchbinder, K. Hoots, D. Vlahov, J. Trowsdale, et al 2002. Epistatic interaction between KIR3DS1 and HLA-B delays the progression to AIDS. Nat. Genet. 31: 429-434. [Medline]
  14. Carrington, M., S. Wang, M. P. Martin, X. Gao, M. Schiffman, J. Cheng, R. Herrero, A. C. Rodriguez, R. Kurman, R. Mortel, et al 2005. Hierarchy of resistance to cervical neoplasia mediated by combinations of killer immunoglobulin-like receptor and human leukocyte antigen loci. J. Exp. Med. 201: 1069-1075. [Abstract/Free Full Text]
  15. Khakoo, S. I., C. L. Thio, M. P. Martin, C. R. Brooks, X. Gao, J. Astemborski, J. Cheng, J. J. Goedert, D. Vlahov, M. Hilgartner, et al 2004. HLA and NK cell inhibitory receptor genes in resolving hepatitis C virus infection. Science 305: 872-874. [Abstract/Free Full Text]
  16. Hsu, K. C., C. A. Keever-Taylor, A. Wilton, C. Pinto, G. Heller, K. Arkun, R. J. O’Reilly, M. M. Horowitz, B. Dupont. 2005. Improved outcome in HLA-identical sibling hematopoietic stem-cell transplantation for acute myelogenous leukemia predicted by KIR and HLA genotypes. Blood 105: 4878-4884. [Abstract/Free Full Text]
  17. Marsh, S. G., P. Parham, B. DuPont, D. E. Geraghty, J. Trowsdale, D. Middleton, C. Vilches, M. Carrington, C. Witt, L. A. Guethlein, et al 2003. Killer-cell immunoglobulin-like receptor (KIR) nomenclature report, 2002. Tissue Antigens 62: 79-86. [Medline]
  18. Winter, C. C., J. E. Gumperz, P. Parham, E. O. Long, N. Wagtmann. 1998. Direct binding and functional transfer of NK cell inhibitory receptors reveal novel patterns of HLA-C allotype recognition. J. Immunol. 161: 571-577. [Abstract/Free Full Text]
  19. Martin, A. M., E. M. Freitas, C. S. Witt, F. T. Christiansen. 2000. The genomic organization and evolution of the natural killer immunoglobulin-like receptor (KIR) gene cluster. Immunogenetics 51: 268-280. [Medline]
  20. Vilches, C., R. Rajalingam, M. Uhrberg, C. M. Gardiner, N. T. Young, P. Parham. 2000. KIR2DL5, a novel killer-cell receptor with a D0–D2 configuration of Ig-like domains. J. Immunol. 164: 5797-5804. [Abstract/Free Full Text]
  21. Gomez-Lozano, N., E. Estefania, F. Williams, I. Halfpenny, D. Middleton, R. Solis, C. Vilches. 2005. The silent KIR3DP1 gene (CD158c) is transcribed and might encode a secreted receptor in a minority of humans, in whom the KIR3DP1, KIR2DL4 and KIR3DL1/KIR3DS1 genes are duplicated. Eur. J. Immunol. 35: 16-24. [Medline]
  22. Gedil, M. A., N. K. Steiner, C. K. Hurley. 2005. Genomic characterization of KIR2DL4 in families and unrelated individuals reveals extensive diversity in exon and intron sequences including a common frameshift variation occurring in several alleles. Tissue Antigens 65: 402-418. [Medline]
  23. Maxwell, L. D., A. Wallace, D. Middleton, M. D. Curran. 2002. A common KIR2DS4 deletion variant in the human that predicts a soluble KIR molecule analogous to the KIR1D molecule observed in the rhesus monkey. Tissue Antigens 60: 254-258. [Medline]
  24. Witt, C. S., J. M. Whiteway, H. S. Warren, A. Barden, M. Rogers, A. Martin, L. Beilin, F. T. Christiansen. 2002. Alleles of the KIR2DL4 receptor and their lack of association with pre-eclampsia. Eur. J. Immunol. 32: 18-29. [Medline]
  25. Vilches, C., C. M. Gardiner, P. Parham. 2000. Gene structure and promoter variation of expressed and nonexpressed variants of the KIR2DL5 gene. J. Immunol. 165: 6416-6421. [Abstract/Free Full Text]
  26. Trompeter, H. I., N. Gomez-Lozano, S. Santourlidis, B. Eisermann, P. Wernet, C. Vilches, M. Uhrberg. 2005. Three structurally and functionally divergent kinds of promoters regulate expression of clonally distributed killer cell Ig-like receptors (KIR), of KIR2DL4, and of KIR3DL3. J. Immunol. 174: 4135-4143. [Abstract/Free Full Text]
  27. Pando, M. J., C. M. Gardiner, M. Gleimer, K. L. McQueen, P. Parham. 2003. The protein made from a common allele of KIR3DL1 (3DL1*004) is poorly expressed at cell surfaces due to substitution at positions 86 in Ig domain 0 and 182 in Ig domain 1. J. Immunol. 171: 6640-6649. [Abstract/Free Full Text]
  28. Witt, C. S., C. Dewing, D. C. Sayer, M. Uhrberg, P. Parham, F. T. Christiansen. 1999. Population frequencies and putative haplotypes of the killer cell immunoglobulin-like receptor sequences and evidence for recombination. Transplantation 68: 1784-1789. [Medline]
  29. Hsu, K. C., S. Chida, D. E. Geraghty, B. DuPont. 2002. The killer cell immunoglobulin-like receptor (KIR) genomic region: gene-order, haplotypes and allelic polymorphism. Immunol. Rev. 190: 40-52. [Medline]
  30. Rajalingam, R., C. M. Gardiner, F. Canavez, C. Vilches, P. Parham. 2001. Identification of seventeen novel KIR variants: fourteen of them from two non-Caucasian donors. Tissue Antigens 57: 22-31. [Medline]
  31. Richardson, J., H. T. Reyburn, I. Luque, M. Vales-Gomez, J. L. Strominger. 2000. Definition of polymorphic residues on killer Ig-like receptor proteins which contribute to the HLA-C binding site. Eur. J. Immunol. 30: 1480-1485. [Medline]
  32. Akatsuka, Y., T. A. Goldberg, E. Kondo, E. G. Martin, Y. Obata, Y. Morishima, T. Takahashi, J. A. Hansen. 2002. Efficient cloning and expression of HLA class I cDNA in human B-lymphoblastoid cell lines. Tissue Antigens 59: 502-511. [Medline]
  33. Shimizu, Y., R. DeMars. 1989. Production of human cells expressing individual transferred HLA-A,-B,-C genes using an HLA-A,-B,-C null human cell line. J. Immunol. 142: 3320-3328. [Abstract]
  34. Snyder, G. A., A. G. Brooks, P. D. Sun. 1999. Crystal structure of the HLA-Cw3 allotype-specific killer cell inhibitory receptor KIR2DL2. Proc. Natl. Acad. Sci. USA 96: 3864-3869. [Abstract/Free Full Text]
  35. Maenaka, K., T. Juji, D. I. Stuart, E. Y. Jones. 1999. Crystal structure of the human p58 killer cell inhibitory receptor (KIR2DL3) specific for HLA-Cw3-related MHC class I. Structure 7: 391-398. [Medline]
  36. Saulquin, X., L. N. Gastinel, E. Vivier. 2003. Crystal structure of the human natural killer cell activating receptor KIR2DS2 (CD158j). J. Exp. Med. 197: 933-938. [Abstract/Free Full Text]
  37. Hanson, R. N., C. Y. Lee, C. J. Friel, R. Dilis, A. Hughes, E. R. DeSombre. 2003. Synthesis and evaluation of 17{alpha}-20E-21-(4-substituted phenyl)-19-norpregna-1,3,5(10),20-tetraene-3,17beta-diols as probes for the estrogen receptor {alpha} hormone binding domain. J. Med. Chem. 46: 2865-2876. [Medline]
  38. Berendsen, H. J. C., J. P. M. Postma, W. F. Vangunsteren, A. Dinola, J. R. Haak. 1984. Molecular-dynamics with coupling to an external bath. J. Chem. Phys. 81: 3684-3690.
  39. Darden, T., D. York, L. Pedersen. 1993. Particle mesh Ewald: an N.Log(N) method for Ewald sums in large systems. J. Chem. Phys. 98: 10089-10092.
  40. Norman, P. J., C. V. Carrington, M. Byng, L. D. Maxwell, M. D. Curran, H. A. Stephens, D. Chandanayingyong, D. H. Verity, K. Hameed, D. D. Ramdath, R. W. Vaughan. 2002. Natural killer cell immunoglobulin-like receptor (KIR) locus profiles in African and South Asian populations. Genes Immun. 3: 86-95. [Medline]
  41. Denis, L., J. Sivula, P. A. Gourraud, N. Kerdudou, R. Chout, C. Ricard, J. P. Moisan, K. Gagne, J. Partanen, J. D. Bignon. 2005. Genetic diversity of KIR natural killer cell markers in populations from France, Guadeloupe, Finland, Senegal and Reunion. Tissue Antigens 66: 267-276. [Medline]
  42. Cook, M. A., P. A. Moss, D. C. Briggs. 2003. The distribution of 13 killer-cell immunoglobulin-like receptor loci in UK blood donors from three ethnic groups. Eur. J. Immunogenet. 30: 213-221. [Medline]
  43. Robinson, J., M. J. Waller, P. Stoehr, S. G. Marsh. 2005. IPD: the Immuno Polymorphism Database. Nucleic Acids Res. 33: D523-D526. [Abstract/Free Full Text]
  44. Vales-Gomez, M., R. A. Erskine, M. P. Deacon, J. L. Strominger, H. T. Reyburn. 2001. The role of zinc in the binding of killer cell Ig-like receptors to class I MHC proteins. Proc. Natl. Acad. Sci. USA 98: 1734-1739. [Abstract/Free Full Text]
  45. Tatko, C. D., M. L. Waters. 2003. The geometry and efficacy of cation-{pi} interactions in a diagonal position of a designed beta-hairpin. Protein Sci. 12: 2443-2452. [Medline]
  46. Williams, A. F., A. N. Barclay. 1988. The immunoglobulin superfamily: domains for cell surface recognition. Annu. Rev. Immunol. 6: 381-405. [Medline]
  47. Dottorini, T., C. K. Vaughan, M. A. Walsh, P. LoSurdo, M. Sollazzo. 2004. Crystal structure of a human VH: requirements for maintaining a monomeric fragment. Biochemistry 43: 622-628. [Medline]
  48. Zahnd, C., S. Spinelli, B. Luginbuhl, P. Amstutz, C. Cambillau, A. Pluckthun. 2004. Directed in vitro evolution and crystallographic analysis of a peptide-binding single chain antibody fragment (scFv) with low picomolar affinity. J. Biol. Chem. 279: 18870-18877. [Abstract/Free Full Text]
  49. Arnett, K. L., S. C. Harrison, D. C. Wiley. 2004. Crystal structure of a human CD3-{epsilon}/{delta} dimer in complex with a UCHT1 single-chain antibody fragment. Proc. Natl. Acad. Sci. USA 101: 16268-16273. [Abstract/Free Full Text]
  50. Tynan, F. E., N. A. Borg, J. J. Miles, T. Beddoe, D. El Hassen, S. L. Silins, W. J. van Zuylen, A. W. Purcell, L. Kjer-Nielsen, J. McCluskey, et al 2005. High resolution structures of highly bulged viral epitopes bound to major histocompatibility complex class I: implications for T-cell receptor engagement and T-cell immunodominance. J. Biol. Chem. 280: 23900-23909. [Abstract/Free Full Text]
  51. Vajdos, F. F., C. W. Adams, T. N. Breece, L. G. Presta, A. M. de Vos, S. S. Sidhu. 2002. Comprehensive functional maps of the antigen-binding site of an anti-ErbB2 antibody obtained with shotgun scanning mutagenesis. J. Mol. Biol. 320: 415-428. [Medline]
  52. Ryu, K. S., J. I. Kim, S. J. Cho, D. Park, C. Park, H. K. Cheong, J. O. Lee, B. S. Choi. 2005. Structural insights into the monosaccharide specificity of Escherichia coli rhamnose mutarotase. J. Mol. Biol. 349: 153-162. [Medline]
  53. Slutsky, M. M., E. N. Marsh. 2004. Cation-{pi} interactions studied in a model coiled-coil peptide. Protein Sci. 13: 2244-2251. [Medline]
  54. Lee, W. S., C. H. Park, S. M. Byun. 2004. Streptomyces griseus trypsin is stabilized against autolysis by the cooperation of a salt bridge and cation-{pi} interaction. J. Biochem. 135: 93-99. [Abstract/Free Full Text]
  55. Dohring, C., J. Samaridis, M. Colonna. 1996. Alternatively spliced forms of human killer inhibitory receptors. Immunogenetics 44: 227-230. [Medline]
  56. Hilton, D. J., S. S. Watowich, L. Katz, H. F. Lodish. 1996. Saturation mutagenesis of the WSXWS motif of the erythropoietin receptor. J. Biol. Chem. 271: 4699-4708. [Abstract/Free Full Text]
  57. Parham, P.. 2004. Immunology: NK cells lose their inhibition. Science 305: 786-787. [Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 J. Immunol.Home page
J. Yu, G. Heller, J. Chewning, S. Kim, W. M. Yokoyama, and K. C. Hsu
Hierarchy of the Human Natural Killer Cell Response Is Determined by Class and Quantity of Inhibitory Receptors for Self-HLA-B and HLA-C Ligands
J. Immunol., November 1, 2007; 179(9): 5977 - 5989.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by VandenBussche, C. J.
Right arrow Articles by Hurley, C. K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by VandenBussche, C. J.
Right arrow Articles by Hurley, C. K.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS