HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Su, R.-C. Articles by Miller, R. G. Search for Related Content PubMed PubMed Citation Articles by Su, R.-C. Articles by Miller, R. G.

Ly-49C^B6 NK Inhibitory Receptor Recognizes Peptide-Receptive H-2K^{b 1}

Ruey-Chyi Su^*, Sam Kam-Pun Kung^2,*, Elizabeth T. Silver, Suzanne Lemieux, Kevin P. Kane and Richard G. Miller^3,*

^* Ontario Cancer Institute, Department of Medical Biophysics, University of Toronto, Toronto, Canada; Department of Immunology, University of Alberta, Edmonton, Alberta, Canada; and Human Health Research Center, INRS-Institut Armand-Frappier, Université du Québec, Laval, Québec, Canada

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
NK-mediated cytotoxicity involves two families of receptors: activating receptors that trigger lysis of the target cells being recognized and inhibitory receptors specific primarily for MHC I on the target cell surface that can override the activating signal. MHC I molecules on the cell surface can be classified into molecules made stable by the binding of peptide with high affinity or unstable molecules potentially capable of binding high affinity peptide (hence, peptide receptive) and being converted into stable molecules. It has been previously shown that the Ly-49A inhibitory receptor recognizes stable D^d molecules. We show in this study that the inhibitory receptor Ly-49C^B6 recognizes peptide-receptive K^b molecules, but does not recognize K^b molecules once they have bound high affinity peptide.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
It is widely accepted that NK cells recognize and lyse target cells through the interplay of two families of receptors (1, 2, 3, 4, 5). Activating receptors can trigger lysis of the target cell being recognized. The activating signal, however, can be overridden by a negative signal from an inhibitory receptor when the latter interacts with its ligand (if present) on the target cells. The ligand(s) for activating receptors is not yet clearly defined, but the only ligands identified to date for inhibitory receptors are associated with MHC I alleles (1, 2, 3, 4, 5). Ly-49A has been shown to bind to H-2D^d and D^k, and can inhibit the lysis of target cells expressing these molecules (6, 7). The same appears true for Ly-49C. Thus, COS-7 cells transfected with Ly-49C receptor can bind to H-2^b, H-2^s, H-2^d, and H-2^k cell lines in cell-cell adhesion assays (8), and the presence of Ly-49C has been shown to be responsible for the resistance to lysis of K^b-expressing target cells by 5E6⁺ (NZB x B6)F₁ NK cells in cytotoxicity assays (9).

Most MHC I molecules on the normal cell surface exist as ternary complexes, each composed of a properly folded heavy chain () containing the peptide-binding groove, a noncovalently associated ß₂-microglobulin (ß₂m),⁴ and a peptide (p) (10). The stability of the complex depends primarily on the binding affinity of the peptide. The ternary complex p_H-ß₂m (associated with high affinity peptide, p_H) is most stable, with a t_1/2 of 10 h or more (10). The complex p_L-ß₂m, in which the peptide is either too long or lacks the proper binding motif and thus binds with low affinity (therefore, p_L), is also present and is less stable (11). Three other forms of MHC I, all unstable, ß₂m, p-, and (perhaps in decreasing order of stability), can also be found on the cell surface (10, 12). Collectively, these have a t_1/2 of 30 min or less (10).

MHC I molecules capable of binding exogenously added peptides are present on the surface of both normal and TAP-deficient (e.g., RAM-S) cells, and are conventionally referred to as empty MHC I molecules. It is essentially unknowable, however, to what extent these molecules are truly empty, i.e., contain solvent in their binding groove (e.g., ß₂m or ) vs a weakly bound peptide (e.g., p_L-ß₂m) (13). Nonetheless, the exogenously added peptide is most likely to bind to ß₂m and/or displace the p_L in the p_L-ß₂m complex (after the p_L dissociates), instead of -chain. Binding to -chain alone is probably unlikely, as it is thought to be highly unstable on the cell surface at 37°C (10). We, in this work, refer to all forms of MHC I that can bind exogenous peptide of high affinity as peptide-receptive MHC I (13). This is an operational definition. Peptide-receptive MHC I molecules are most likely the ß₂m binary complex and the p_L-ß₂m ternary complex, but may also include alone. Approximately 10% of D^b molecules expressed on the surface of EL-4 tumor cells are peptide receptive (14).

Several studies have focused on defining the regions of MHC I involved in the recognition by NK inhibitory receptors. By performing exon-shuffling and point-mutation experiments on human MHC I, Storkus et al. showed that the 1-2 region of the -chain appears to be critical in determining the specificity of MHC I (HLA-A and HLA-B) as an inhibitory ligand (15), and that the amino acids in the peptide-binding groove of MHC I are important in conferring NK resistance (16). For mouse MHC I, Karlhofer et al. (6) and Sundbäck et al. (17) have mapped the determinant recognized by the inhibitory receptor Ly-49A to the 2 region of the D^d molecule.

Several studies have addressed the question of whether the presence of peptide in the peptide-binding groove within the 1-2 region is critical in recognition by an NK inhibitory receptor. Storkus et al. (18), using C1R cells that are normally lysed by human NK cells, found that they could be protected from lysis by transfection of certain HLA-A or HLA-B MHC I, and that protection was reversed by the addition of peptide that could bind to the protective MHC I. Chadwick and Miller (19) and Chadwick et al. (20) found that normal nontransformed lymphoblasts could be lysed by syngeneic mouse NK cells in the presence of peptide specific for their MHC I. They tested nine different peptides specific for K^b, D^b, K^d, L^d, or D^d. All of these peptides could sensitize a normal lymphoblast to be lysed if they could bind to it. Using a similar system, Su et al. (21) found that the acquisition of sensitivity to lysis correlated with the disappearance of peptide-receptive MHC I. When lymphoblasts made sensitive to NK lysis by being pulsed with peptide were incubated in the absence of peptide, they lost their sensitivity to lysis as they reacquired peptide-receptive MHC I, a process that took about 90 min. When production of new peptide-receptive MHC I was inhibited (by inhibiting transport of new cell surface protein through the trans-Golgi), lymphoblasts remained sensitive to NK lysis. These experiments led to the hypothesis that there are NK inhibitory receptors that recognize peptide-receptive MHC I. The hypothesis would gain considerable credibility if one could actually identify an NK inhibitory receptor reactive against peptide-receptive MHC I.

Ly-49A is the best-characterized NK inhibitory receptor. Results of Correa and Raulet (22) and Orihuela et al. (23) show that Ly-49A recognizes not the peptide-receptive, but the peptide-bound D^d molecule. These groups transfected TAP-deficient RMA-S (22) or LKD8 (23) cells with D^d and showed that the protection of D^d-transfected cells from lysis mediated by Ly-49A⁺ B6 NK cells requires peptide binding to the D^d molecule. The extent of protection correlated with the extent to which the added peptide stabilized D^d expression. Both groups suggested that the role of peptide was to promote the assembly and cell surface expression of MHC I and that there was no peptide specificity in Ly-49A recognition of D^d. Su et al. (21), using Ly-49A⁺ and Ly-49A^- BALB/c (H-2^d) NK cells, investigated the effect of pulsing BALB/c lymphoblasts with a high affinity D^d-binding peptide. The results obtained were consistent with those just summarized, i.e., Ly-49A⁺ cells did not lyse D^d+ normal lymphoblasts either before or after they were pulsed with a high affinity D^d-binding peptide. However, pulsing the lymphoblasts with high affinity K^d peptide sensitized these D^d+ normal lymphoblasts to be lysed by Ly-49A⁺ cells, whether or not they were also pulsed with D^d peptide, thus implying the existence of an inhibitory receptor reactive against peptide-receptive K^d, and also implying that it is the total balance of inhibitory and stimulatory signals that determines whether there is lysis.

In this study, we have investigated whether another well-characterized NK inhibitory receptor, Ly-49C, recognizes peptide-bound or peptide-receptive K^b. We have used both a syngeneic experimental system (21) and an in vitro hybrid-resistance model (19) in which inhibitory function of Ly-49A and 5E6 Ags has been demonstrated (9). 5E6 mAb stains B6 LAK cells expressing Ly-49C, and/or Ly-49I receptors (24). Most of our study has used 5E6⁺ and 5E6^- B6 LAK subpopulations as effector cells, but we also show that Ly-49C^-I⁺ (5E6⁺4LO3311^-) B6 LAK cells are not inhibited in the presence of either peptide-receptive or peptide-bound K^b. We conclude from this study that the Ly-49C^B6 NK inhibitory receptor recognizes the peptide-receptive form of the K^b molecule.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Mice

Normal C57BL/6 (B6, H-2^b), BALB/c (H-2^d), and (BALB/c x B6, H-2^b/d)F₁ mice were purchased from The Jackson Laboratory (Bar Harbor, ME). C57BL/6 (H-2^b) athymic nude mice were purchased from Taconic (Germantown, NY). (BALB/c x B6, H-2^b/d)F₁ athymic nude mice were purchased from The Jackson Laboratory. D^b-/- B6 mice (D^b-K^b+) and K^b-/- B6 mice (D^b+K^b-) have been previously described (25, 26). They were generous gifts from Dr. F. Lemonnier (Pasteur Institute, Paris, France), and were bred in our animal facility. All mice were kept in a specific pathogen-free environment. In most experiments, 6–10-wk-old female mice were used (although either sex gave similar results).

NK generation

The method used for producing activated NK cells (LAK cells) was identical to that used previously (19, 21, 27). Briefly, 2 x 10⁶ nylon wool nonadherent spleen cells from B6 athymic nude mice were cultured at 37°C for 3 to 4 days in 5 ml -MEM supplemented with 10% FCS, 50 µM 2-ME, and 10 mM HEPES buffer (hereafter referred to as CM), containing 500 U/ml mouse rIL-2. In some experiments, as specified, (BALB/c x B6)F₁ athymic nude mice were used. Mouse rIL-2 was obtained as a supernatant from a cell line transfected with the IL-2 gene (28). These cells were cultured in 25-cm² flasks at 37°C in a 10% CO₂ in air incubator. Yields typically exceeded 1200 U/ml of rIL-2.

Target cell generation

Target cells were B6, BALB/c, or (BALB/c x B6)F₁ Con A (ICN Pharmaceuticals Canada, Montreal, Quebec) blast cells produced by incubating 10⁷ splenocytes for 3 days in 10 ml CM supplemented with Con A (2 µg/ml). On day 3, Con A blast cells were harvested on lympholyte M (Cedarlane Laboratory, Hornby, ON, Canada) and ⁵¹Cr labeled by incubating about 6 x 10⁶ cells for 90 min at 37°C with 360 µCi Na⁵¹CrO₄ (NEN Life Science, Boston, MA) in 150 µl of PBS containing 67% FCS. They were then washed three times with CM containing 1% FCS, to remove nonincorporated Na⁵¹CrO₄.

MHC class I-binding peptides

The effect of MHC class I-binding peptides on normal lymphoblast sensitivity to NK lysis was assessed by pulsing lymphoblasts with the experimental peptide (at a concentration of 100 ng/ml in CM, unless stated otherwise) for 45 min at 4°C before the assay. Peptides utilized were K^b-restricted epitopes of chicken OVA, SIINFEKL (OVAp_258–265) (29), and vesicular stomatitis virus NP, RGYVYQGL (VSVp_52–59) (30), a D^b-restricted epitope of influenza nucleoprotein, ASNENMETM (Flu-NP_366–374) (31), and a D^d-restricted epitope of HIV gp160, RGPGRAFVTI (HIVp_318–327) (32). Chicken OVA, SIINFEKL(OVAp_258–265), and its derivative, biotinylated OVA peptide, SIINFEK(bio)L were prepared by the Ontario Cancer Institute Biotechnology Laboratory, using an Applied Biosystems peptide synthesizer (Applied Biosystems, Foster City, CA). Both VSVp_52–59 and Flu-NP peptides (>90% purity) were generous gifts from Dr. B. H. Barber (University of Toronto). HIVp (>90% purity) was a gift from Dr. D. Williams (University of Toronto). OVAp and VSVp peptides are natural ligands for K^b and bind to K^b with high affinities (29, 30, 31, 33). HIVp peptide is a natural ligand for D^d and binds with high affinity (32).

Cytotoxicity assay

Methods for measuring lytic activity were identical to those used previously (20, 21, 34). After three washes, ⁵¹Cr -labeled Con A lymphoblasts were incubated with peptide in 3 ml of CM for 45 min at 4°C and washed again before being used in a 4-h ⁵¹Cr release assay performed in 96-well V-bottom microtiter plates using 2000 targets/well, dispensed in 100-µl aliquots and effectors at an E:T ratio as indicated or at 30:1, also added in 100-µl aliquots. For experiments in which preincubation of NK cells with 5E6 mAb was required, the preincubation was done at 4°C for 30–45 min while preparing target cells for the assay. For experiments in which preincubation of target cells with 5F1, Y-3, or 25D1.16 mAb (5 µg/ml/2 x 10⁵ cells) was required, the preincubation was done at 4°C for 20–30 min. Before the addition of mAb to target/effector cells, mAb were preincubated with soluble protein A (2 µg per 10 µg of mAb used; Sigma, St. Louis, MO) and soluble protein A/G mix (2 µg per 10 µg of mAb used; ICN Biomedicals, Aurora, OH) for 30 min on ice. The mAb remained in the assay mixture during the 4-h ⁵¹Cr release assay. Specific lysis was calculated as percentage of specific lysis = (E - S)/(T - S) x 100, in which each value represents the mean ± SEM of five replicates. E is the experimental mean of ⁵¹Cr released; S, the amount of ⁵¹Cr released when the target cells were cultured in medium alone; and T, the total amount of ⁵¹Cr released in the presence of 2% acetic acid. Dialyzed FCS (12-kDa cutoff) was regularly used in place of regular FCS during the ⁵¹Cr labeling, pulsing, and assay stages (14, 35).

Flow cytometry/FACS analysis

To measure the disappearance of peptide-receptive K^b using OVAp-_K-bio, day 3 B6 Con A-activated lymphoblasts were prepulsed with increasing concentrations of nonlabeled OVAp for 30–45 min to convert peptide-receptive K^b to peptide-bound K^b. Unbound OVAp was removed and the cells were then incubated at 4°C for 45 min with biotinylated OVAp peptide (OVAp-_K-bio, 1 µg/ml). The binding of biotinylated OVAp was visualized with R-PE-conjugated streptavidin (SA-PE; Sigma, St. Louis, MO), which binds to biotin, and analyzed using a FACScan flow cytometer and LYSIS II program (Becton Dickinson). The maximum level of peptide-receptive K^b was measured by omitting the addition of OVAp during the peptide-pulsing procedure. The relative level of peptide-receptive K^b was calculated as the fraction of the mean fluorescent intensity (MFI) detected when a particular OVAp concentration was used over the MFI of the maximum level of peptide-receptive K^b and is plotted against the OVAp concentrations used. Alternatively, the presence of peptide-receptive K^b molecule on the cell surface can be measured with OVAp binding (see Fig. 2, B and C), which can then be visualized using FITC-conjugated 25D1.16 mAb (36). 25D1.16 mAb was purified from the supernatant of 25-D1.16 hybridoma (reclone 21) culture, a generous gift from Dr. R. Germain (National Institute of Health, Bethesda, MD). 25D1.16 mAb binds specifically to OVAp-K^b complex, and not VSVp-K^b complex or any other peptide-K^b complex (36). The level of peptide-receptive K^b on the cell surface was titrated by pulsing the B6 Con A blasts with increasing concentrations of VSVp for 30–45 min and then washing away unbound VSVp. The remaining peptide-receptive K^b molecules were detected by pulsing with OVAp (1 µg/ml/10⁶ cells for 30–45 min) and FITC-conjugated 25D1.16 mAb, which detects the OVAp-K^b complexes. The maximum level of peptide-receptive K^b was measured when no VSVp was used, assuming that OVAp would bind to all of the peptide-receptive K^b molecules. In separate kinetics experiments (unpublished), it was found that 1 µg of OVAp/ml/10⁶ cells produced half-maximum binding within 15 min and saturation binding within 45 min. The relative level of peptide-receptive K^b is calculated as the fraction of the MFI detected when a particular VSVp concentration is used over the MFI of the maximum level of peptide-receptive K^b and is plotted against the concentrations of VSVp used.

View larger version (13K): [in this window] [in a new window]   FIGURE 2. Detection of OVAp-Kb complexes on the B6 Con A blasts using 25D1.16 mAb that specifically recognizes OVAp-Kb complexes on the cell surface. A, B6 Con A blasts pulsed with OVAp (100 ng/ml/106 cells) stained positive by 25D1.16 mAb, suggesting that OVAp-Kb complexes are present. B, These OVAp-pulsed B6 Con A blasts, after being cultured at 37°C for 2 h in the absence of OVAp, were still stained positive by 25D1.16 mAb. C, Furthermore, the B6 Con A blasts, when cultured at 37°C for 2 h in the presence of OVAp (100 ng/ml/106 cells), remained positive for 25D1.16 mAb staining.

Day 3 or day 4 B6 LAK cultures were harvested and resuspended in 1% BSA/PBS (10⁷ cells/ml). The cells were then incubated with 5E6 mAb (4 µg/10⁶ cells; PharMingen, San Diego, CA), at 4°C on a rotator for 45 min. 5E6 mAb recognizes the NK inhibitory receptors, Ly-49C and I (24). Stained cells were washed with cold 1% BSA/PBS and then sorted (Coulter Epics V, Palo Alto, CA) into 5E6⁺ and 5E6^- subpopulations. Ly-49C⁺ (5E6⁺4LO3311⁺) and Ly-49C^-I⁺ (5E6⁺4LO3311^-) B6 LAK subpopulations were isolated by sorting 5E6⁺ B6 LAK cells (day 5 or day 6) stained with biotinylated 4LO3311 mAb (24) (1 µg/10⁶ cells, specific for Ly-49C). (BALB/c x B6)F₁ LAK cells were sorted into 5E6^-Ly-49A⁺ and 5E6⁺Ly-49A^- subpopulations using FITC-labeled 5E6 (4 µg/10⁶ cells), and biotinylated JR9.318 mAb (4 µg/10⁶ cells). JR9-318 mAb binds specifically to Ly-49A (37); the hybridoma was obtained from Dr. D. Raulet with permission of Dr. J. Roland (Pasteur Institute, Paris, France). Sorted cells were cultured in 5 ml CM containing 500 U/ml mouse rIL-2 (28) for an additional 2 to 3 days, as above.

Fab fragment generation

For Fab fragment generation, 8 mg of affinity-purified anti-K^b mAb (5F1) was resuspended in 1 ml of digestion buffer (20 mM phosphate, 20 mM cysteine-HCl, 10 mM EDTA-Na₄, pH 7) and then digested with 0.5 ml of immobilized papain (Pierce, Rockford, IL) for 10 h in a shaker at 37°C at high speed. The reaction was stopped by adding 1.5 ml of 10 mM Tris-HCl (pH 7.5) to the digestion mixture, as suggested by the manufacturer’s protocol (Pierce). After a centrifugation at 10,000 rpm for 5 min, the supernatant was collected and mixed with protein A/G-Sepharose beads (ICN Biomedicals, Aurora, OH) to remove undigested Ab and Fc fragments. The purity and the binding activity of Fab fragments were checked by 10% SDS-PAGE and flow cytometry, respectively.

COS-7 cell expression and cell-cell adhesion assay

Ly-49C^B6 and Ly-49I^B6 cDNAs were inserted into the multiple cloning site in PCI-neo mammalian expression vectors (Promega, Madison, WI). Transfection of COS-7 cells (gift from Dr. M. B. Wheeler, University of Toronto) with Ly-49 cDNA was conducted using lipofectamine (Life Technologies, Gaithersburg, MD), following the manufacturer’s instructions. Briefly, 1 day before transfection, COS-7 cells were seeded in six-well plates (2 x 10⁵ cells in 2 ml of 10% CM per well) and incubated at 37°C for sufficient time (18–24 h) to reach 50–80% confluence. On the day of transfection, 2 µg of DNA was incubated with 10 µl of lipofectamine reagent diluted in 750 µl of DMEM (free of serum and antibiotics) at room temperature for 30 min. COS-7 cells were washed once with 2 ml of DMEM (DMEM H21, free of serum and antibiotics) before the addition of transfecting reagent (DNA-lipofectamine complexes). At the end of a 5-h incubation at 37°C, transfecting reagent was replaced with 2 ml DMEM supplemented with 20% FBS. One day following transfection, COS-7 cells were trypsinized and transferred to 24-well plates at 0.5 x 10⁵ cells/well. Three days posttransfection, day 3 Con A blasts (6 x 10⁶ cells) were labeled for 90 min at 37°C with 360 µCi Na⁵¹CrO₄ (NEN Life Science Products, Boston, MA) in 150 µl of PBS containing 67% FCS. They were then washed four times with CM containing 1% FCS, to remove nonincorporated Na⁵¹CrO₄. These target cells were incubated with OVAp (1 µg/ml/10⁶ cells), mAb (5F1, or Y-3, 50 µg/ml/10⁶), or 1% CM alone on ice for 45 min and tested for adhesion to COS-7 cells by addition to the wells at 4 x 10⁵ cells/well in 0.5 ml. As indicated in Fig. 3C, COS-7 cells were incubated with 5E6 mAb (20 µg/well/0.2 ml) at room temperature for 30 min before the addition of target cells. Plates were centrifuged for 5 min at 700 rpm and incubated for 2 h at 37°C in a CO₂ incubator. At the end of incubation, plates were washed five times with prewarmed media and then photographed. Bound Con A blasts were lysed with 500 µl/well of 10% Triton X-100 (Sigma), and the radioactivity was determined by subjecting 250 µl/well of cell lysate to gamma counting. Cell binding was calculated as percentage of cells bound = (E/T) x 100, in which each value represents the mean ± SEM of quadruplicate wells. E is experimental mean of the radioactivity in the lysate; and T is the total amount of radioactivity in the 4 x 10⁵ Con A blasts added to each well.

View larger version (30K): [in this window] [in a new window]   FIGURE 3. Lysis of B6 Con A blasts pulsed with Kb-specific peptide correlated inversely with the presence of peptide-receptive Kb molecules on the target cell surface. A, The decrement in the expression of peptide-receptive Kb on the target cell surface (squares, right y-axis) correlated with the increasing sensitivity of these target cells to lysis by 5E6+ B6 LAK cells (, left y-axis). The presence of peptide-receptive Kb molecule on the cell surface was measured by OVAp-K-bio binding, as described in Fig. 1. The relative level of peptide-receptive Kb was calculated as the fraction of the MFI detected when a particular OVAp concentration was used over the MFI detected when OVAp was omitted during the peptide-pulsing procedure (right y-axis). The specific lysis of these OVAp-pulsed target cells by 5E6+ B6 LAK cells () or 5E6- B6 LAK cells () was evaluated using a 4-h 51Cr release assay. B, The specific lysis (left y-axis) of VSVp-pulsed Con A blasts by Ly-49C+ B6 LAK cells () increased proportionally to the decrement of peptide-receptive Kb on the target cell surface (, right y-axis). The disappearance of peptide-receptive Kb from the cell surface had no effect on the resistance of these B6 Con A blasts to lysis by Ly-49C-I+ B6 LAK cells (). The susceptibility of the VSVp-pulsed B6 Con A blasts to lysis by either Ly-49C+ () or Ly-49C-I+ B6 LAK () was examined with the 51Cr release assay. The presence of peptide-receptive Kb molecule on the cell surface was measured with OVAp binding, as described in C. C, The level of peptide-receptive Kb on the cell surface was titrated by pulsing the B6 Con A blasts with increasing concentrations of VSVp for 30–45 min and then washing away unbound VSVp. The remaining peptide-receptive Kb was then measured by pulsing with OVAp and labeling with FITC-conjugated 25D1.16 mAb. The level of peptide-receptive Kb that remained on the cell surface was inversely proportional to the amount of VSVp used in the pulsing procedure. The maximum level of peptide-receptive Kb was measured when no VSVp was used, assuming that OVAp would bind to all of the peptide-receptive Kb molecules. The relative level of peptide-receptive Kb was calculated as the fraction of the MFI detected when a particular VSVp concentration was used over MFI of the maximum level of peptide-receptive Kb and is plotted in B (, right y-axis).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

The mAb 5E6 identifies an inhibitory receptor that recognizes H-2K^b (9). We performed experiments to identify which form of K^b this receptor was recognizing, in particular whether it recognizes peptide-receptive K^b. LAK cells from B6 nude mice (lack T cells) or normal B6 mice were sorted into 5E6⁺ and 5E6^- populations and used as effectors. Targets were Con A-activated B6 blasts (B6 Con A blasts) either unpulsed or pulsed with CM containing MHC I-specific peptide before being used as target cells in a 4-h ⁵¹Cr release assay. As shown in Table I, B6 Con A blasts were resistant to lysis by either 5E6⁺ or 5E6^- B6 LAK cells (lines 1 and 8). They became significantly susceptible to lysis by 5E6⁺ B6 LAK cells when pulsed with K^b-specific peptides (OVAp or VSVp) (Table I, lines 2 and 3). Furthermore, the masking of 5E6 Ags on 5E6⁺ B6 NK cells resulted in the lysis of unpulsed syngeneic Con A blasts (line 4). K^b-peptide-pulsed B6 Con A blasts remained resistant to lysis by 5E6^- B6 LAK cells (lines 9 and 10), and the addition of 5E6 mAb had little effect (line 11). These results are consistent with there being an inhibitory receptor present in the 5E6⁺ population and absent in the 5E6^- population that recognizes peptide-receptive K^b.

None of the 5E6^- groups showed lysis above background. This is consistent with there being no inhibitory receptor for peptide-receptive K^b in the 5E6^- NK subpopulation. To verify that the 5E6^- LAK had normal cytotoxic function, we tested both 5E6⁺ and 5E6^- B6 LAK cells for their ability to lyse D^b-peptide (Flu-NP_366–374)-pulsed B6 Con A blasts. They were lysed by both subpopulations (Table I, lines 15–18). This observation verifies that the 5E6^- LAK cells are active, and implies the presence of an undescribed inhibitory receptor present on at least some LAK cells in both the 5E6⁺ and 5E6^- subpopulations, a receptor that recognizes peptide-receptive D^b molecules (see Discussion).

In summary, reagents (mAb or peptide) that could bind to peptide-receptive K^b molecules or 5E6 Ag rendered B6 Con A blasts susceptible to lysis by 5E6⁺ B6 LAK cells, but had no effect on their lysis by 5E6^- B6 LAK cells.

The presence of peptide-receptive K^b molecules on B6 Con A blasts correlates with the resistance to lysis by 5E6⁺ B6 LAK cells

We have previously shown that peptide-receptive MHC I molecules can be regenerated when peptide-pulsed Con A blasts are incubated for 90 min or longer at 37°C in CM that contains no free exogenous peptide (21). We therefore tested whether the reappearance of peptide-receptive K^b molecules on the target cell surface would restore the resistance of these peptide-pulsed B6 Con A blasts to lysis by 5E6⁺ B6 LAK cells. As shown in Fig. 1A, B6 Con A blasts pulsed overnight with the K^b-specific peptide, OVAp, had no detectable peptide-receptive K^b molecules on the cell surface. When these OVAp-pulsed B6 Con A blasts were incubated at 37°C for 2 h in CM that contained no free exogenous peptide, peptide-receptive K^b molecules could be detected (Fig. 1B). However, no measurable peptide-receptive K^b molecules were observed if OVAp was included in the culture during that 2-h incubation (Fig. 1C). The cell surface expression of peptide-receptive K^b molecule was measured using an OVAp peptide in which the lysine (K) at position 7 was biotinylated (OVAp-_K-bio). This lysine side chain is known to be one of the CTL epitopes on OVAp (41, 42), and has been shown to protrude from the peptide-binding groove toward the solvent (43, 44). We found that this peptide binds specifically to K^b and can be readily detected by the addition of SA-PE (21).

View larger version (12K): [in this window] [in a new window]   FIGURE 1. Detection of peptide-receptive Kb molecules on the cell surface. Peptide-receptive Kb molecule on the cell surface can be detected by pulsing the cells with a high concentration of OVAp-K-bio (1 µg/ml/106 cells). OVAp-K-bio binds specifically to peptide-receptive Kb molecules on the cell surface and can be detected with SA-PE shown in the graph as dark solid lines. The staining control was done by omitting the addition of OVAp-K-bio (gray line). A, B6 Con A blasts pulsed overnight with OVAp showed no staining of peptide-receptive Kb on their cell surface. B, The expression of peptide-receptive Kb molecules was detected when these OVAp-pulsed B6 Con A blasts were incubated for 2 h at 37°C in the absence of Kb-specific peptide. C, The regeneration of peptide-receptive Kb molecule on the cell surface was not detected when these same OVAp-pulsed B6 Con A blasts were incubated at 37°C in the presence of OVAp for 2 h.

To further examine how the level of peptide-receptive K^b expression on the target cell surface correlates with the sensitivity of the target cell to lysis by 5E6⁺ and 5E6^- B6 LAK populations, B6 Con A blasts were pulsed with increasing concentrations of OVAp. The level of the remaining peptide-receptive K^b molecules on the cell surface after OVAp pulsing was measured by pulsing these cells with 1 µg/10⁶ cells/ml of OVAp-_K-bio, as described above. As shown in Fig. 3A, the relative level of remaining peptide-receptive K^b molecules decreased as the concentration of OVAp used in pulsing these cells was increased. The sensitivity of these OVAp-pulsed B6 Con A blasts to lysis mediated by 5E6⁺ B6 NK cells was also evaluated, and found to be inversely correlated with the relative level of peptide-receptive K^b molecules on the cell surface after the OVAp-pulsing procedure (Fig. 3A). Thus, as the OVAp concentration used in pulsing the target cells was increased, the specific lysis of these target cells also increased, reaching a plateau around 1 ng/ml of OVAp. The resistance of these target cells to lysis by 5E6^- B6 LAK cells was independent of their exposure to K^b-binding peptide (Fig. 3A, ). We conclude that the sensitivity of the target cells to lysis by 5E6⁺ B6 LAK cells is inversely proportional to the expression of peptide-receptive K^b molecule on the cell surface.

The Ly-49C NK inhibitory receptor recognizes the peptide-receptive form of the K^b molecule

The mAb 5E6 recognizes two members of the Ly-49 family, Ly-49C and Ly-49I (24), whereas the mAb 4LO3311 binds specifically to Ly-49C (45). To determine whether Ly-49C or Ly-49I or both recognize the peptide-receptive form of K^b, we sorted Ly-49C⁺ (i.e., 5E6⁺4LO3311⁺) and Ly-49C^-I⁺ (i.e., 5E6⁺4LO3311^-) B6 LAK populations and used them as effectors in ⁵¹Cr release assays. B6 Con A blasts pulsed with K^b-specific peptide (OVAp or VSVp) became significantly susceptible to lysis by Ly-49C⁺ B6 LAK cells (Table III, lines 2 and 3), but remained relatively resistant to lysis by Ly-49C^-I⁺ B6 LAK cells (lines 8 and 9). Masking of peptide-bound K^b molecules on B6 Con A blasts with 5F1 Fab fragments in the presence of soluble protein A and protein G did not alter their resistance to lysis by either Ly-49C⁺ or Ly-49C^-I⁺ B6 LAK cells (lines 4 and 10). However, the use of 5E6 mAb to block Ly-49C and Ly-49I led to the lysis of B6 Con A blasts by Ly-49C⁺, but not by Ly-49C^-I⁺ B6 LAK cells (lines 6 and 12). Taken together, the results indicate that Ly-49C recognizes the peptide-receptive form of K^b molecules and not the stable ternary K^b complex containing a high affinity peptide. The data, however, do not enable one to determine the ligand identified by Ly-49I or even whether it functions as an inhibitory receptor.

Ly-49A^F1 recognizes peptide-bound D^d on BALB/c and (BALB/c x B6)F₁ cells

Our results indicate that Ly-49C^B6 recognizes the peptide-receptive form of the K^b molecule in a syngeneic system. Using a system of hybrid resistance, Yu et al. have shown that the ligand for Ly-49C on (NZB x B6, H-2^d/b)F₁ NK cells is the K^b molecule expressed on B6 (H-2^b) Con A blasts, and the ligand for Ly-49A on the same F₁ NK cells is the D^d molecule expressed on BALB/c (H-2^d) Con A blasts (9). In this study, we first confirm their findings and then test whether Ly-49C and Ly-49A are recognizing peptide-receptive or peptide-bound forms of K^b and D^d, respectively. In concordance with Yu et al., B6 Con A blasts were resistant to lysis by (B6 x BALB/c)F₁ 5E6⁺Ly-49A^- LAK cells (Table IV, line 1), but were susceptible to lysis by 5E6^-Ly-49A⁺ F₁ LAK cells (line 8). This was interpreted by Yu et al. as meaning that the presence of K^b on B6 Con A blasts prevented the lysis mediated by 5E6⁺Ly-49A^- F₁ LAK cells, and the absence of D^d rendered B6 Con A blasts sensitive to lysis by 5E6^-Ly-49A⁺ F₁ LAK (9). On the other hand, BALB/c Con A blasts were found to be resistant to lysis by 5E6^-Ly-49A⁺ F₁ LAK cells (line 11) because the presence of the D^d molecule could inhibit Ly-49A-expressing LAK cells, whereas the absence of the K^b molecule on BALB/c Con A blasts made them sensitive to lysis by 5E6⁺Ly-49A^- F₁ LAK cells (line 4).

The D^d-specific peptide results are in agreement with published studies using D^d-transfected RMA-S cells and our previous study using the BALB/c syngeneic system showing that Ly-49A recognizes peptide-bound D^d (21, 22, 23). The same HIVp-pulsed BALB/c or F₁ Con A blasts became more susceptible to lysis by 5E6⁺Ly-49A^- F₁ LAK (lines 5 and 7), suggesting that there is an unidentified inhibitory receptor expressed on the Ly-49A^- NK subpopulation that recognizes peptide-receptive D^d.

Ly-49C^B6 binds to B6 Con A blasts expressing peptide-receptive K^b molecules

Brennan et al. (8) have shown that Ly-49C^B6 expressed on COS cells (by transfection) mediates cell-cell adhesion by binding to H-2^b or H-2^d on the cell lines tested. In this study, this same assay was used to test whether Ly-49C^B6 binds to the peptide-receptive K^b molecule or the p_H-ß₂m K^b ternary complex. As shown in Fig. 4A, COS-7 cells transfected with Ly-49C^B6 bound Con A blasts from D^b-/-K^b+/+ B6 mice, which express both peptide-receptive K^b and the p_H-ß₂m K^b ternary complex (a). The same Con A blasts, pulsed with OVAp (K^b-specific peptide) to remove peptide-receptive K^b, bound poorly to COS-7 cells transfected with Ly-49C^B6 (b). Exogenous OVAp was left in the wells throughout the whole assay. Only 7 ± 1% of OVAp-pulsed Con A blasts added bound to the Ly-49C^B6, compared with 56 ± 6% of nonpeptide-pulsed Con A blasts (Fig. 4B). Con A blasts from D^b+/+K^b-/- B6 mice (which express only H-2D^b) were used as a negative control and did not bind to COS-7 cells transfected with Ly-49C^B6 whether or not OVAp was present (Fig. 4A, c and d). It has been shown previously that Ly-49I^B6 does not bind to cells expressing H-2^b molecules using this assay system. COS-7 cells transfected with Ly-49I^B6 were, therefore, used as a negative control for cell-cell adhesion (Fig. 4A, e–h). As shown on Fig. 4A, COS-7 cells transfected with Ly-49I^B6 were bound by significantly fewer D^b-/-K^b+/+ B6 Con A blasts than Ly-49C^B6-expressing COS-7 cells (a vs e). The presence of OVAp did not have a major effect on the number of D^b-/-K^b+/+ B6 Con A blasts binding to COS-7 cells expressing Ly-49I^B6 (Fig. 4A, e and f). There was no visible binding of D^b+/+K^b-/- B6 Con A blasts to Ly-49I^B6-expressing COS-7 cells, regardless of the presence or absence of OVAp (Fig. 4A, g and h). Furthermore, none of the B6 Con A blasts tested bound to COS-7 cells treated with the same transfecting reagent from which DNA was omitted (Fig. 4A, i–l).

View larger version (49K): [in this window] [in a new window]   FIGURE 4. Cell-cell binding mediated by Ly-49CB6 and Ly-49IB6. A, Ly-49CB6 cDNA (a–d), Ly-49IB6 (e–h) cDNA, or blank (i–l) were transiently expressed in COS-7 cells and tested for binding to 51Cr-labeled Con A blasts derived from Db-/-Kb+/+ B6 mice in the presence (b, f, and j) or absence (a, e, and i) of exogenous OVAp, or from Db+/+Kb-/- B6 mice in the presence (d, h, and l) or absence (c, g, and k) of OVAp. A total of 4 x 105 cells was incubated in each well (24-well plate) for 2 h at 37°C. Unbound cells were removed and plates were photographed. B, Cell binding was quantitated for individual plate wells by lysing bound cells with Triton X-100. 51Cr radioactivity in the cell lysate was measured by gamma counting, as described in Materials and Methods. Results are represented as the mean percentage of input cells bound and the SEM from quadruplicate wells. Results in A and B are from one experiment and are representative of three independent experiments. C, Cell-cell adhesion could be prevented by the binding of Y-3 or 5E6 mAb to targets or Ly-49CB6-transfected COS-7 cells, respectively. Results are from one experiment and are representative of three independent experiments.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
Our results demonstrate that the presence of peptide-receptive K^b on the target cell surface prevents lysis mediated by 5E6⁺ B6 LAK cells (Tables I and II). Binding of high affinity K^b-specific peptide converts most, if not all, peptide-receptive K^b molecules on the target cell surface to the form of p_H-ß₂m K^b ternary complexes that are no longer peptide receptive (or only slightly so). This prevents the inhibitory recognition mediated by 5E6 Ags and results in the lysis of previously NK-resistant target cells by 5E6⁺ B6 LAK cells. The disappearance of peptide-receptive K^b from the target cell surface can be measured and was found to correlate with the increased sensitivity of these target cells to lysis by 5E6⁺ B6 LAK cells ( Figs. 1–3). Furthermore, when peptide-receptive K^b molecules were masked by Y-3 mAb, which binds to both peptide-receptive and p_H-ß₂m K^b ternary complexes (38, 39), the target cells became susceptible to lysis by 5E6⁺ B6 LAK cells, but remained resistant when only p_H-ß₂m K^b ternary complexes were masked using 5F1 mAb. This argues against the involvement of p_H-ß₂m K^b ternary complexes in inhibitory recognition (Table I).

NK recognition involves both activation and inhibitory receptors. It is possible (although rendered unlikely by the results just summarized) that pulsing normal B6 Con A blasts with K^b-specific peptide creates a target structure that activates the 5E6⁺ B6 LAK cells to kill the target cells rather than masking an inhibitory ligand as proposed above. In this case, the susceptibility of the target cells to lysis by 5E6⁺ B6 LAK cells would directly correlate with the formation of peptide-ß₂m K^b ternary complexes. To directly test this possibility, target cells expressing both peptide-receptive K^b molecules and OVAp-ß₂m K^b complexes were used. Cells carrying large numbers of OVAp-ß₂m K^b complexes were generated by culturing B6 Con A blasts overnight with OVAp. Peptide-receptive K^b molecules were regenerated on OVAp-pulsed B6 Con A blasts by culturing these cells at 37°C for 2 h in the absence of exogenous peptide. The 2-h time period should be sufficient to allow newly synthesized peptide-receptive K^b molecules to restore equilibrium values on the cell surface (21). Despite the presence of OVAp-ß₂m K^b complexes on the cell surface, as shown by 25D1.16 mAb staining (Fig. 2, B and C), the cells remained resistant to lysis by 5E6⁺ B6 LAK cells if peptide-receptive K^b molecules were expressed on the cell surface (Table II). Furthermore, the blocking of OVAp-K^b complexes with 25D1.16 mAb did not alter the resistance of the B6 Con A blasts that expressed peptide-receptive K^b or the sensitivity of the B6 Con A blasts that expressed no peptide-receptive K^b to lysis by 5E6⁺ B6 LAK cells (Table II). Taken together, these observations support our hypothesis that pulsing the B6 Con A blasts with OVAp masked the inhibitory ligand (the peptide-receptive K^b) from being recognized by 5E6 Ags on B6 LAK cells instead of creating a target structure.

The 5E6 mAb has been shown to bind to both Ly-49C and Ly-49I B6 NK inhibitory receptors (24). Our data (from both the functional and binding assays) show that the NK inhibitory receptor that recognizes the peptide-receptive K^b molecule is Ly-49C^B6, and not Ly-49I^B6. We have not yet examined whether Ly-49C^BALB/c recognizes peptide-receptive K^b. The 4LO3311 mAb, specific for Ly-49C, was used in this study to sort Ly-49C⁺ (4LO3311⁺5E6⁺) and Ly-49C^-I⁺ (4LO3311^-5E6⁺) B6 LAK subpopulations, which were then used in peptide-pulsing experiments. B6 Con A blasts pulsed with K^b-specific peptide (OVAp or VSVp) were susceptible to lysis by Ly-49C⁺ B6 LAK cells either in the presence or the absence of 5F1 mAb, while B6 Con A blasts pulsed with media alone remained resistant under the same conditions (Table III, lines 1–5). Furthermore, the presence of 5E6 mAb in the assay enhanced the lysis of B6 Con A blasts by Ly-49C⁺ B6 LAK cells, as expected (Table III, line 6). These observations suggest that peptide-receptive K^b is the ligand for Ly-49C^B6, and that blockage of Ly-49C-mediated inhibitory signal by masking Ly-49C^B6 with 5E6 mAb or masking peptide-receptive K^b with high affinity K^b-specific peptide or mAb leads to lysis of target cells, as observed (Table I, lines 4 and 5; Table III, line 6). In support of our functional data, we showed that Ly-49C^B6 expressed on the surface of COS-7 cells can bind to D^b-/-K^b+/+ B6 Con A blasts and that such binding could be blocked by the presence of K^b-specific peptide, OVAp (Fig. 4A, a and b). Again, the binding data strongly suggest that peptide-receptive K^b, and not the p_H-ß₂m K^b ternary complex, is the ligand for Ly-49C^B6 in inhibitory recognition.

On the basis of the predicted amino acid sequence, Ly-49I^B6 contains an immunoreceptor tyrosine-based inhibitory motif in its intracellular domain and is therefore predicted to be an NK inhibitory receptor (24, 46). We observed (Fig. 4) weak binding of B6 Con A blasts to COS-7 cells transfected with Ly-49I^B6 that required the presence of both K^b (B6 D^b-/-K^b+/+ bound, B6 D^b+/+K^b-/- did not) and Ly-49I^B6 (no binding to nontransfected controls). Furthermore, binding was marginally higher after pulsing with high affinity K^b-specific peptide (OVAp). These data suggest that Ly-49I^B6 recognizes stable K^b-peptide complexes. However, this recognition did not lead to inhibition in our experimental system. Masking of either Ly-49I^B6 or p_H-K^b with appropriate mAbs (Table III, lines 12 and 10, respectively) did not lead to lysis. It may be that the Ly-49I^B6-K^b interaction is of too low affinity to trigger inhibition or that the mAbs tested do not effectively block the functional sites in this particular interaction. The role of Ly-49I^B6 in NK recognition remains unclear.

Using the same adhesion assay and COS cells transfected with the Ly-49C^B6 construct, Ly-49C^B6 was shown to bind to cells expressing H-2^b, H-2^s, H-2^k, or H-2^d molecules (8). Despite the binding of Ly-49C^B6, H-2^d cells that do not express K^b molecules are highly sensitive to lysis by 5E6⁺ B6 LAK cells (9). A possible explanation may be that Ly-49C^B6 has different affinities for H-2^b, H-2^s, H-2^d, and H-2^k. Kane (47) showed that a higher surface density of D^k than D^d is required to have equivalent binding to Ly-49A^B6 (47). Furthermore, mAb against D^d or K^d cannot completely block the binding of Ly-49C^B6 (48).

Ly-49J, Ly-49K, and Ly-49N are all encoded in the B6 genome, are closely related to Ly-49C^B6 (49), and may also recognize peptide-receptive K^b. From the predicted amino acid sequences, the presumed epitope for 4LO3311 mAb (Lemieux, unpublished) is absent, but the epitope for 5E6 mAb may be present on Ly-49J, Ly-49K, and/or Ly-49N. Therefore, it is possible that these Ly-49 members are expressed on either or both the 5E6⁺4LO3311⁺ and 5E6⁺4LO3311^- NK subsets. Ly-49J has an immunoreceptor tyrosine-based inhibitory motif and may be implicated in the lysis of B6 Con A blasts pulsed with K^b-specific peptide. However, further study is required to determine whether Ly-49I, Ly-49J, Ly-49K, and/or Ly-49N recognize peptide-receptive MHC I.

A caveat should be added regarding our use of anti-MHC I mAbs in lysis-blocking studies. Neither 5F1 (recognizes mostly p_H-K^b ternary complexes) nor 25D1.16 (recognizes OVAp-K^b) was able to block the delivery of an inhibitory signal to Ly-49C^B6. Our interpretation above was that these mAbs do not block because they do not recognize peptide-receptive K^b. It may be that neither mAb blocks the epitope on the K^b molecule recognized by Ly-49C^B6. Thus, the anti-D^d mAb 34-5-8S blocks D^d signaling to Ly-49A, but other anti-D^d mAbs that bind to different sites do not block (23). Furthermore, the effect of using mAb to block recognition elements (e.g., 5E6 Ag, K^b molecule) was not always optimal, and appeared to vary with the state of activation of the LAK cultures used. As shown in Table III (line 6), the use of 5E6 mAb in this experiment resulted in suboptimal blockage of 5E6 Ags, compared with the use of K^b-specific peptide (lines 2 and 3) in blocking inhibitory recognition. However, our argument for peptide-receptive K^b being the ligand for Ly-49C^B6 does not depend exclusively on our interpretation of the mAb-blocking studies. In particular, binding of peptide to K^b with high affinity does block signaling.

Our data strongly support the conclusion that Ly-49A^F1 and Ly-49C^B6 are NK inhibitory receptors that recognize different forms of MHC I (p_H-D^d for Ly-49A^F1 and peptide-receptive K^b for Ly-49C^B6). Our data of this study and a previous study (21) also provide direct evidence for the existence of three new NK inhibitory receptors that recognize peptide-receptive MHC I: 1) Both 5E6⁺ and 5E6^- B6 NK cells lyse B6 Con A blasts pulsed with high affinity D^b-specific peptide (Table I, lines 15–18), implying the existence of a novel inhibitory receptor recognizing peptide-receptive D^b. 2) 5E6⁺Ly-49A^- (B6 x BALB/c)F₁ NK cells lyse both F₁ and BALB/c Con A blasts pulsed with high affinity D^d-binding peptide (Table IV, lines 5 and 7), implying the existence of a novel inhibitory receptor recognizing peptide-receptive D^d. Su et al. found a similar receptor studying lysis of BALB/c Con A blasts by Ly-49A^- BALB/c NK cells (21). 3) BALB/c NK cells, either unfractionated or sorted into Ly-49A⁺ and Ly-49A^- subpopulations, lyse BALB/c Con A blasts pulsed with high affinity K^d-binding peptide (21), implying the existence of a novel inhibitory receptor recognizing peptide-receptive K^d. Whether these unidentified receptors are members of the Ly-49 family, mouse homologues of the human killer inhibitory receptor family (yet to be found), the CD94/NKG2 family, or some completely novel structure is unknown at the present time.

It appears that the presence of one NK inhibitory signal may not be sufficient to prevent lysis if a second is removed. Thus, the observation that B6 Con A blasts pulsed with D^b-specific peptide were sensitive to lysis by either 5E6⁺ or 5E6^- syngeneic NK subsets suggests that the inhibitory signal mediated by 5E6 Ags in recognition of peptide-receptive K^b is not dominant. It is in concordance with the hypothesis that the outcome of NK recognition is dependent on the balance of inhibitory and stimulatory signals (1, 2, 3, 4, 5). The binding of K^b-specific peptide can remove inhibitory ligand recognized by 5E6 Ags, resulting in the reduction of the total inhibitory signal. By the same token, the binding of D^b-specific peptide can also reduce the inhibitory signal in NK cells that bear receptor(s) recognizing peptide-receptive D^b, resulting in the activation of the NK cells. Single NK cells have been shown to coexpress multiple Ly-49 receptors in a stochastic manner; in particular, 5E6⁺ and 5E6^- B6 NK subsets express similar profiles of other Ly-49 members (50). Therefore, it is likely that both 5E6⁺ and 5E6^- NK subsets have an equal representation of NK inhibitory receptor(s) recognizing peptide-receptive D^b. Although NK inhibitory receptors recognizing Qa1 (i.e., the mouse homologues of human NKG2 and CD94 (51)) might also be equally expressed on both 5E6⁺ and 5E6^- NK subsets, the addition of D^b-specific peptide having an effect on the recognition of Qa1 is unlikely, as there is no evidence that a D^b-specific peptide such as Flu-NP can bind to nonclassical MHC I, like Qa1 (29, 30, 31).

It is not clear how an NK inhibitory receptor can distinguish between peptide-receptive and p_H-ß₂m MHC I molecules. A 30-min pulse with high affinity peptide at a concentration of 1 µg/ml/10⁶ cells (Fig. 3) was sufficient to load all peptide-receptive MHC I. We found previously (21) that if export of newly synthesized MHC I was blocked, no detectable new peptide-receptive MHC I appeared on the cell surface over at least 4 h, implying that peptide-bound MHC I is extremely stable. We hypothesize that during the peptide pulse, all empty MHC I and all MHC I containing low affinity peptide are converted to p_H-ß₂m ternary stable form. It is likely that the conformation of the stable, peptide-bound MHC I molecule differs from that of the possible peptide-receptive precursors. It could well be that NK inhibitory receptors can distinguish between these conformations.

Peptide binding is known to influence the conformation of the surface of class I molecules, as detected with mAbs and TCR (52, 53, 54). Using a system employing fluorescence resonance energy transfer, Catipovic et al. found that H-2K^b ß₂m heterodimers are in a relatively extended conformation, and that this conformation becomes more compact when peptide is bound (55). This is consistent with peptide-receptive MHC I molecules (p_L- ß₂m and/or ß₂m) having conformation(s) different from that of a p_H-ß₂m ternary complex. Using computer-modeling analysis of MHC I structures, Achour et al. postulated that peptide binding to D^d imposes a specific conformation, different from that of peptide-receptive D^d, a conformation required for recognition by Ly-49A (56). This specific conformation is not found on molecules such as D^b or K^b, which are not ligands for Ly-49A. It is also possible that p_H-ß₂m and peptide-receptive MHC I molecules may differ in their ability to associate with each other or with other cell surface molecules, and thus affect their ability to be recognized by NK inhibitory receptors.

Peptide-receptive MHC I on the cell surface has a t_1/2 of less than 30 min at 37°C (10). For this reason, the recognition of peptide-receptive MHC I by NK inhibitory receptors poses a potential advantage for the organism. During a viral infection, very few peptide-receptive MHC I are exported to the cell surface either because very large quantities of viral peptide inside the cell saturate MHC I or because the virus greatly reduces overall MHC I production such that few peptide-receptive MHC I (albeit perhaps a higher percentage of all MHC I) reach the cell surface. In the latter case, the disappearance of peptide-receptive MHC I may be detected earlier compared with the detection of the loss of p_H-ß₂m ternary MHC I because these have a t_1/2 of more than 10 h (10, 57). The process described in this work would allow NK cells to detect virally infected host cells many hours earlier than if detection required a complete loss of MHC I, and many days before the acquired immune system (B cells and T cells) is able to mount an effective response.

	Acknowledgments

 
We thank Dr. F. A. Lemonnier (Pasteur Institute, Paris, France) for a kind gift of K^b-/- and D^b-/- C57BL/6 mice (H-2^b) and Dr. M. B. Wheeler (University of Toronto, Toronto, Canada) for a gift of COS-7 cells.

	Footnotes

 
¹ This work was supported by research grants to R.G.M. and K.P.K. from the National Cancer Institute of Canada. R.-C.S. and S.K.-P.K. are research students of the National Cancer Institute of Canada and are supported with funds provided by the Canadian Cancer Society. K.P.K. is a Medical Research Council (Canada) scholar and AHFMR Senior Scholar.

² Current address: Department of Microbiology and Immunology and Medicine, University of California, Los Angeles, AIDS Institute, Los Angeles, CA 90025.

³ Address correspondence and reprint requests to Dr. Richard G. Miller, Department of Medical Biophysics, Ontario Cancer Institute, 610 University Avenue, Toronto, Ontario, Canada, M5G 2 M9. E-mail address:

⁴ Abbreviations used in this paper: ß₂m, ß₂-microglobulin; CM, complete medium; Flu-NP, D^b-restricted epitope of influenza nuclear protein; HIVp, D^d-restricted epitope of HIV gp160; LAK, IL-2-activated NK cell; MFI, mean fluorescent intensity; OVAp, K^b-restricted epitope of chicken OVA; OVAp-_K-bio, biotinylated (at lysine(K) amino acid) OVAp; p, peptide; p_H, peptide with high binding affinity; p_L, peptide with low binding affinity; SA-PE, R-PE-conjugated streptavidin; VSVp, K^b-restricted epitope of vesicular stomatitis virus nuclear protein.

Received for publication May 25, 1999. Accepted for publication September 7, 1999.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Yokoyama, Y. M.. 1998. Natural killer cell receptors. Curr. Opin. Immunol. 10:298.[Medline]
2. Moretta, L., E. Ciccone, A. Moretta, P. Höglund, C. Ohlen, K. Kärre. 1992. Allorecognition by NK cells: nonself or self. Immunol. Today 13:300.[Medline]
3. Yokoyama, W. M., W. E. Seaman. 1993. The Ly-49 and NKR-P1 gene families encoding lectin-like receptors on natural killer cells: the NK gene complex. Annu. Rev. Immunol. 11:613.[Medline]
4. Lanier, L. L.. 1998. Follow the leader: NK cell receptors for classical and nonclassical MHC class I. Cell 92:705.[Medline]
5. Lanier, L. L., J. H. Phillips. 1996. Inhibitory MHC class I receptors on NK cells and T cells. Immunol. Today 17:86.[Medline]
6. Karlhofer, F. M., R. K. Ribaudo, W. M. Yokoyama. 1992. MHC class I alloantigen specificity of Ly49A⁺ IL-2-activated natural killer cells. Nature 358:66.[Medline]
7. Held, W., D. Cado, D. H. Raulet. 1996. Transgenic expression of the Ly49A natural killer cell receptor confers class I major histocompatibility complex (MHC)-specific inhibition and prevents bone marrow allograft rejection. J. Exp. Med. 184:2037.[Abstract/Free Full Text]
8. Brennan, J., D. Mager, W. Jefferies, F. Takei. 1994. Expression of different members of the Ly-49 gene family defines distinct natural killer cell subsets and cell adhesion properties. J. Exp. Med. 180:2287.[Abstract/Free Full Text]
9. Yu, Y. Y. L., T. George, J. R. Dorfman, J. Roland, V. Kumar, M. Bennett. 1996. The role of Ly49A and 5E6 (Ly49C) molecules in hybrid resistance mediated by murine natural killer cells against normal T cell blasts. Immunity 4:67.[Medline]
10. Neefjes, J. J., L. Smith, M. Gehrmann, H. L. Plöegh. 1992. The fate of the three subunits of major histocompatibility complex class I molecules. Eur. J. Immunol. 22:1609.[Medline]
11. Ortiz-Navarrete, V., G. J. Hämmerling. 1991. Surface appearance and instability of empty H-2 class I molecules under physiological conditions. Proc. Natl. Acad. Sci. USA 88:3594.[Abstract/Free Full Text]
12. Burshtyn, D. N., B. H. Barber. 1993. Dynamics of peptide binding to purified antibody-bound H-2D^b and H-2D^b-ß₂m. J. Immunol. 151:3082.[Abstract]
13. Day, P. M., F. Esquivel, J. Lukszo, J. R. Bennink, J. W. Yewdell. 1995. Effect of TAP on the generation and intracellular trafficking of peptide-receptive major histocompatibility complex class I molecules. Immunity 2:137.[Medline]
14. Christinck, E. R., M. A. Luscher, B. H. Barber, D. B. Williams. 1991. Peptide binding to class I MHC on living cells and quantitation of complexes required for CTL lysis. Nature 352:67.[Medline]
15. Storkus, W. J., J. Alexander, J. A. Payne, P. Cresswell, J. R. Dawson. 1989. The 1/2 domains of class I HLA molecules confer resistance to natural killing. J. Immunol. 143:3853.[Abstract]
16. Storkus, W. J., R. D. Salter, J. Alexander, F. E. Ward, R. E. Ruiz, P. Cresswell, J. R. Dawson. 1991. Class I-induced resistance to natural killing: identification of nonpermissive residues in HLA-A2. Proc. Natl. Acad. Sci. USA 88:5989.[Abstract/Free Full Text]
17. Sundback, J., M. C. Nakamura, M. Waldenstrom, E. C. Niemi, W. E. Seama