Identification of HLA-E-specific alloreactive T lymphocytes: a cell subset that undergoes preferential expansion in mixed lymphocyte culture and displays a broad cytolytic activity against allogeneic cells

CD94-NKG2A recognition of human leukocyte antigen (HLA)-E bound to an HLA class I leader sequence

The recognition of human leukocyte antigen (HLA)-E by the heterodimeric CD94-NKG2 natural killer (NK) receptor family is a central innate mechanism by which NK cells monitor the expression of other HLA molecules, yet the structural basis of this highly specific interaction is unclear. Here, we describe the crystal structure of CD94-NKG2A in complex with HLA-E bound to a peptide derived from the leader sequence of HLA-G. The CD94 subunit dominated the interaction with HLA-E, whereas the NKG2A subunit was more peripheral to the interface. Moreover, the invariant CD94 subunit dominated the peptide-mediated contacts, albeit with poor surface and chemical complementarity. This unusual binding mode was consistent with mutagenesis data at the CD94-NKG2A-HLA-E interface. There were few conformational changes in either CD94-NKG2A or HLA-E upon ligation, and such a "lock and key" interaction is typical of innate receptor-ligand interactions. Nevertheless, the structure also provided insight into how this interaction can be modulated by subtle changes in the peptide ligand or by the pairing of CD94 with other members of the NKG2 family. Differences in the docking strategies used by the NKG2D and CD94-NKG2A receptors provided a basis for understanding the promiscuous nature of ligand recognition by NKG2D compared with the fidelity of the CD94-NKG2 receptors.

Cytomegalovirus immune evasion

Human cytomegalovirus (HCMV) has become a paradigm for viral immune evasion due to its unique multitude of immune-modulatory strategies. HCMV modulates the innate as well as adaptive immune response at every step of its life cycle. It dampens the induction of antiviral interferon-induced genes by several mechanisms. Further striking is the multitude of genes and strategies devoted to modulating and escaping the cellular immune response. Several genes are independently capable of inhibiting antigen presentation to cytolytic T cells by downregulating MHC class I. Recent data revealed an astounding variety of methods in triggering or inhibiting activatory and inhibitory receptors found on NK cells, NKT cells, T cells as well as auxiliary cells of the immune system. The multitude and complexity of these mechanisms is fascinating and continues to reveal novel insights into the host-pathogen interaction and novel cell biological and immunological concepts.

Gliadin regulates the NK-dendritic cell cross-talk by HLA-E surface stabilization

We analyzed the autologous NK cell interaction with gliadin-presenting dendritic cells. Gliadin is the known Ag priming the celiac disease (CD) pathogenesis. We demonstrate that gliadin prevents immature dendritic cells (iDCs) elimination by NK cells. Furthermore, cooperation between human NK cells-iDCs and T cells increases IFN-gamma production of anti-gliadin immune response. Gliadin fractions were analyzed for their capability to stabilize HLA-E molecules. The alpha and omega fractions conferred the protection from NK cell lysis to iDCs and increased their HLA-E expression. Gliadin pancreatic enzyme digest and a peptide derived from gliadin alpha increased HLA-E levels on murine RMA-S/HLA-E-transfected cells. Analysis of HLA-E expression in the small intestinal mucosa of gluten-containing diet celiac patients and organ culture experiments confirmed the in vitro data.

Expression and release of soluble HLA-E is an immunoregulatory feature of endothelial cell activation

Human leukocyte antigen (HLA)-E belongs, with HLA-G and HLA-F, to the non-classic major histocompatibility complex (MHC) class I (Ib) molecules, broadly defined by a limited polymorphism and a restricted pattern of cellular expression. In contrast to HLA-G, the expression and function of HLA-E and HLA-F in physiologic and pathologic processes remain poorly established. In the present study, we show that HLA-E protein expression in normal human nonlymphoid organs is mainly restricted to endothelial cells (ECs). HLA-E is also basally expressed by B and T lymphocytes, natural killer (NK) cells and by macrophages. We demonstrate that tumor necrosis factor alpha (TNFalpha), interleukin-1beta (IL-1beta), and interferon gamma (IFNgamma) up-regulate the cell-surface expression of HLA-E on ECs in vitro and induce the release of soluble HLA-E (sHLA-E). HLA-E up-regulation protects IFN-gamma-activated ECs from NK-mediated cell lysis, while sHLA-E protects bystander cells. Finally, sHLA-E is not detected in normal sera, and increased serum levels correlate with disease activity in patients with antineutrophil cytoplasmic antibody-associated systemic vasculitis. Thus, HLA-E expression and release of sHLA-E are features of EC activation and emphasize immunoregulatory functions of the endothelium. The present identification of soluble HLA-E molecules may have important implications in understanding the pathogenesis of immune-mediated vascular diseases and for the diagnosis and monitoring of patients.

Effector and regulatory events during natural killer-dendritic cell interactions

The different cell types of the innate immune system can interact with each other and influence the quality and strength of an immune response. The cross talk between natural killer (NK) cells and myeloid dendritic cells (DCs) leads to NK cell activation and DC maturation. Activated NK cells are capable of killing DCs that fail to undergo proper maturation ('DC editing'). Encounters between NK cells and DCs occur in both inflamed peripheral tissues and lymph nodes, where both cell types are recruited by chemokines released in the early phases of inflammatory responses. Different NK cell subsets (CD56(bright)CD16(-) versus CD56(+)CD16(+)) differ in their homing capabilities. In particular, CD56(bright)CD16(-) NK cells largely predominate the lymph nodes. In addition, these two subsets display major functional differences in their cytolytic activity, cytokine production, and ability to undergo proliferation. NK cell functions are also greatly influenced by the presence of polarizing cytokines such as interleukin (IL)-12 and IL-4. The cytokine microenvironment reflects the presence of different cell types that secrete such cytokines in response to microbial products acting on different Toll-like receptors (TLRs). Moreover, NK cells themselves can respond directly to microbial products by means of TLR3 and TLR9. Thus, it appears that the final outcome of a response to microbial infection may greatly vary as a result of the interactions occurring between different pathogen-derived products and different cell types of the innate immunity system. These interactions also determine the quality and strength of the subsequent adaptive responses. Remarkably, NK cells appear to play a key role in this complex network.

Homozygous status for HLA-E*0103 confers protection from acute graft-versus-host disease and transplant-related mortality in HLA-matched sibling hematopoietic stem cell transplantation

BACKGROUND

The posttransplant period following hematopoietic stem cell transplantation (HSCT) is potentially high risk for developing survival-compromising complications, many of which are known to be under the control of immunogenetic factors. Given the dual role of human leukocyte antigen (HLA)-E molecules in innate and adaptive immune processes, we analyzed the impact of HLA-E polymorphism in genoidentical HSCT setting.

METHODS

We analyzed 187 HLA-genoidentical sibling pairs for HLA-E polymorphism. To explore its potential association with the incidence of acute and chronic graft versus host disease (aGVHD, cGVHD), severe infections, risk for transplant-related mortality (TRM), and overall survival, HLA-E locus was genotyped by a polymerase chain-reaction-sequence-specific primer (PCR-SSP) strategy.

RESULTS

Multivariate analysis, taking into account the patient-, donor- and transplant-related factors, showed that the incidence of aGVHD and TRM at day 180 were low when the genotype was HLA-E*0103/E*0103, either in the donor or in the recipient, the pairs being identical for HLA-E alleles (hazard ratio [HR]=0.71, P=0.009; and HR=0.42, P=0.04, respectively). We also found a trend towards association between E*0103 homozygosity and improved survival (P=0.05). There was no association between HLA-E polymorphism and incidence of severe infections.

CONCLUSIONS

These data suggest that the homozygous state for HLA-E*0103 allele behaves as a protective genetic factor against aGVHD and TRM and likely contributes to improved survival in HLA-genoidentical bone marrow transplantation.

Expression and release of HLA-E by melanoma cells and melanocytes: potential impact on the response of cytotoxic effector cells

HLA-E are nonclassical MHC molecules with poorly characterized tissue distribution and functions. Because of their capacity to bind the inhibitory receptor, CD94/NKG2A, expressed by NK cells and CTL, HLA-E molecules might play an important role in immunomodulation. In particular, expression of HLA-E might favor tumor cell escape from CTL and NK immunosurveillance. To address the potential role of HLA-E in melanoma immunobiology, we assessed the expression of these molecules ex vivo in human melanoma biopsies and in melanoma and melanocyte cell lines. Melanoma cell lines expressed no or low surface, but significant intracellular levels of HLA-E. We also report for the first time that some of them produced a soluble form of this molecule. IFN-gamma significantly increased the surface expression of HLA-E and the shedding of soluble HLA-E by these cells, in a metalloproteinase-dependent fashion. In contrast, melanocyte cell lines constitutively expressed HLA-E molecules that were detectable both at the cell surface and in the soluble form, at levels that were poorly affected by IFN-gamma treatment. On tumor sections, a majority of tumor cells of primary, but a low proportion of metastatic melanomas (30-70 and 10-20%, respectively), expressed HLA-E. Finally, HLA-E expression at the cell surface of melanoma cells decreased their susceptibility to CTL lysis. These data demonstrate that HLA-E expression and shedding are normal features of melanocytes, which are conserved in melanoma cells of primary tumors, but become dependent on IFN-gamma induction after metastasis. The biological significance of these findings warrants further investigation.

Effector Activity of Decidual CD8+ T Lymphocytes in Early Human Pregnancy1

Although CD8+ T lymphocytes are present in human decidua throughout pregnancy, albeit as a minor population in early pregnancy, their role in normal pregnancy is largely unknown. The present study aimed to characterize their effector phenotype, including cytolytic activity, cytokine profile, and capacity to affect placental invasion. CD8+ lymphocytes were positively selected from normal early pregnancy decidua (7-14 wks gestational age). Decidual CD8+ T lymphocytes were studied using standard and redirected chromium release assays to investigate natural killer cell-sensitive cytotoxicity and cytotoxicity that requires T-cell receptor signal transduction respectively, multiplex cytokine analysis to analyze cytokine production, and a placental explant invasion model to assess the effect of soluble products of decidual CD8+ T lymphocytes on trophoblast invasion. Decidual CD8+ T lymphocytes exhibited cytolytic ability against P815 target cells (mean % Specific Chromium Release at effector:target ratio of 32:1 [SCR(32)] of 32.7 +/- 5.8) and against K562 target cells (mean SCR(32) of 20.3 +/- 0.5). Phytohemagglutinin-P (PHA-P)-stimulated decidual CD8(+) T lymphocytes produced high levels of both interferon gamma and interleukin (IL) 8, and low levels of granulocyte-macrophage colony-stimulating factor (CSF2), IL1B, IL2, IL6, IL10, IL12, and tumor necrosis factor; these did not vary with gestational age. IL4 was undetectable. Decidual CD8+ T lymphocyte supernatants increased the capacity of extravillous trophoblast cells to invade through Matrigel compared with the PHA-P control. These findings suggest that decidual CD8+ T cells can display cytolytic activity, do not evoke a predominant local intrauterine Th2 type cytokine environment, and may act to regulate invasion of extravillous trophoblast cells into the uterus, a crucial process for normal uteroplacental development.

The involvement of HLA-E and -F in pregnancy

The human MHC class I genes, HLA-E, -F and -G are referred to as non-classical or class Ib genes and are distinguished from their close relatives (the classical HLA class I genes) by expression patterns and low allelic polymorphism. To date, most studies that relate these molecules to the immunology of pregnancy have concerned only HLA-G. However, recent advances have suggested potential unique roles as well for HLA-E and HLA-F in pregnancy. A notable advance was the observation that all three proteins are expressed on the surface of extravillous trophoblast that has invaded the maternal decidua. Given this expression site, possibly the only cell type in human development where this occurs, it is logical to hypothesize that all three antigens, each with its own unique receptor-ligand interaction(s), contribute collectively to enable the growth of the developing child. In this review, we examine and discuss the accumulated data on expression and function of HLA-E and HLA-F and attempt to relate what is known to the involvement of HLA-E and -F in human pregnancy.

Structural basis for a major histocompatibility complex class Ib-restricted T cell response

In contrast to antigen-specific immunity orchestrated by major histocompatibility complex (MHC) class Ia molecules, the ancestrally related nonclassical MHC class Ib molecules generally mediate innate immune responses. Here we have demonstrated the structural basis by which the MHC class Ib molecule HLA-E mediates an adaptive MHC-restricted cytotoxic T lymphocyte response to human cytomegalovirus. Highly constrained by host genetics, the response showed notable fine specificity for position 8 of the viral peptide, which is the sole discriminator of self versus nonself. Despite the evolutionary divergence of MHC class Ia and class Ib molecules, the structure of the T cell receptor-MHC class Ib complex was very similar to that of conventional T cell receptor-MHC class Ia complexes. These results emphasize the evolutionary 'ambiguity' of HLA-E, which not only interacts with innate immune receptors but also has the functional capacity to mediate virus-specific cytotoxic T lymphocyte responses during adaptive immunity.

Identification of effector-memory CMV-specific T lymphocytes that kill CMV-infected target cells in an HLA-E-restricted fashion

HLA-E-restricted T cells represent a minor cytolytic T lymphocyte (CTL) population characterized by the surface expression of HLA class I-specific inhibitory receptors and by the capability of killing a large panel of allogeneic target cells (therefore named NK-CTL). Here we show that this subset of T cells is present in a sizeable fraction in the peripheral blood of human cytomegalovirus (HCMV)-seropositive healthy individuals. We provide evidence that NK-CTL recognize in an HLA-E-restricted fashion a naturally processed CMV-derived peptide in the transporter associated with antigen processing (TAP)-2(-/-) UL40+ RMA-S cell transfectants. Moreover, we show that they recognize and kill HCMV-infected target cells. NK-CTL are characterized by the CD8beta(dull) (CD45RA+)(CD28-)(CD27-)(CCR7-)(CD56+) surface phenotype, thus suggesting that they belong to the effector-memory cell compartment. Consistent with the effector-memory phenotype, they promptly produce IFN-gamma, but not IL-2, upon interaction with the specific HCMV UL40-derived peptide. Our data suggest that HLA-E-restricted CTL may represent an additional effector cell type involved in defenses against HCMV, a virus which escapes the control exerted by conventional CTL or NK cells.

Human cytolytic T lymphocytes expressing HLA class-I-specific inhibitory receptors

MHC class-1-specific inhibitory receptors were originally described in NK cells, in which they represent an important fail-safe mechanism that induces NK cell tolerance to normal self cells. These inhibitory NK receptors (iNKRs) were subsequently found expressed on different T cell subsets, primarily CD8(+) cytolytic T lymphocytes (CTLs), in which they can inhibit T cell receptor mediated functions. Some iNKR(+) CTLs are HLA-E-restricted, represent oligo- or monoclonal expansions, and can play a defensive role in viral infections. Although T cell activation, in the presence of certain cytokines, can induce the expression of the CD94-NKG2A heterodimeric receptor, the mechanism leading to the expression of killer immunoglobulin-like receptors (KIRs) is still unknown. The expression of iNKRs in T cells might contribute to the prevention of apoptotic cell death, thus allowing their survival and clonal expansion in vivo. In addition, iNKR(+) T cells might contribute to peripheral self-tolerance.

Potential Role of Natural Killer Cell Receptor-Expressing Cells in Immunotherapy for Leukemia

Natural killer cell receptor (NKR)-expressing cells have cytolytic activity against leukemic cells, and solid tumor cells escape from T-cell recognition because of the low expression levels of class I HLA molecules in both allogeneic and autologous settings. This characteristic feature of NK cell recognition of target cells in contrast with that of T-cells provides a strategy to overcome tolerance in the tumor-bearing host. Furthermore, inhibitory NKR-expressing cells may have cytolytic activity and immunoregulatory functions. Several methods can be used to expand NKR-expressing cells for adoptive immunotherapy for leukemia and other malignant diseases. We review recent developments in the biology and clinical application of NKR-expressing cells, such as NK cells, lymphokine-activated killer cells, cytokine-induced killer cells, NKT cells, and other NKR-expressing cells.

Coordinated Expression of Ig-Like Inhibitory MHC Class I Receptors and Acquisition of Cytotoxic Function in Human CD8+T Cells

MHC class I-specific inhibitory receptors are expressed by a subset of memory-phenotype CD8(+) T cells. Similar to NK cells, MHC class I-specific inhibitory receptors might subserve on T cells an important negative control that participates to the prevention of autologous damage. We analyzed here human CD8(+) T cells that express the Ig-like MHC class I-specific inhibitory receptors: killer cell Ig-like receptor (KIR) and CD85j. The cell surface expression of Ig-like inhibitory MHC class I receptors was found to correlate with an advanced stage of CD8(+) T cell maturation as evidenced by the reduced proliferative potential of KIR(+) and CD85j(+) T cells associated with their high intracytoplasmic perforin content. This concomitant regulation might represent a safety mechanism to control potentially harmful cytolytic CD8(+) T cells, by raising their activation threshold. Yet, KIR(+) and CD85j(+) T cells present distinct features. KIR(+)CD8(+) T cells are poor IFN-gamma producers upon TCR engagement. In addition, KIR are barely detectable at the surface of virus-specific T cells during the course of CMV or HIV-1 infection. By contrast, CD85j(+)CD8(+) T cells produce IFN-gamma upon TCR triggering, and represent a large fraction of virus-specific T cells. Thus, the cell surface expression of Ig-like inhibitory MHC class I receptors is associated with T cell engagement into various stages of the cytolytic differentiation pathway, and the cell surface expression of CD85j or KIR witnesses to the history of qualitatively and/or quantitatively distinct T cell activation events.

Immunobiology of Natural Killer Lymphocytes in Transplantation

Natural killer (NK) lymphocytes are powerful effector cells of the peripheral immune system. NK cell functions are controlled by the expression of a variety of cell surface receptors with either inhibitory or activating roles. The genetic and functional diversity of this repertoire of receptors and the role of human leukocyte antigen class I histocompatibility molecules as a major group of NK receptor ligands endows NK cells with an innate alloreactive capacity. Early studies of experimental bone marrow transplantation revealed an important role for NK cells in the rejection of allogeneic grafts and contributed significantly to our understanding of NK cell behavior. Both animal models and in vitro studies have since implicated NK cells as contributors to the pathology of clinical transplantation. However, recent clinical studies have demonstrated the potential benefits of exploiting NK cell alloreactivity in mismatched hematopoietic stem cell transplantation for particular types of acute leukemia. Future investigations of NK cell alloreactive functions will undoubtedly reveal additional roles and potential therapeutic applications of this fundamental cell type in clinical transplantation.

Pig-tailed macaques (Macaca nemestrina) possess six MHC-E families that are conserved among macaque species: implication for their binding to natural killer receptor variants

MHC loci encode highly polymorphic molecules involved in the presentation of self and non-self peptides to cells of the adaptive and innate immune systems. Although variable, MHC-E genes are well conserved among primates and provide signals to natural killer cells. In this study, we sequenced and analyzed MHC-E alleles of pig-tailed macaque (Macaca nemestrina), a nonhuman primate used for HIV pathogenesis and vaccine studies. Among a group of seven macaques, the characterization of eight Mane-E alleles revealed an increased number of polymorphic sites compared with human HLA-E alleles. Phylogenetic analyses of MHC-E alleles from pig-tailed macaque, rhesus monkey (Macaca mulatta) and cynomolgus macaque (Macaca fascicularis) demonstrated that the three macaque species shared six families of macaque MHC-E alleles and indicated that these families existed in the common ancestor 5.5 million years ago. Polymorphic Mane-E sites were not concentrated within the peptide-binding pockets, but were distributed throughout the entire ORF. The peptide-binding domain of Mane-E is similar to its human analogue, and peptide substrates theoretically capable of binding to Mane-E molecules were found in the leader sequence of classical Mane-A and -B molecules. Additionally, the polymorphic amino acids located in the alpha(1) and alpha(2) domains of Mane-E molecules have side chains expected to be oriented toward solvent and away from the peptide-binding groove, suggesting that some of them (positions 19, 73, 79 and 145) might be available for interaction with polymorphic receptors of natural killer cells.

HLA-E-restricted recognition of human cytomegalovirus by a subset of cytolytic T lymphocytes

Natural killer (NK)-cytotoxic T lymphocytes (CTL) are a subset of CD8(+) cytolytic T lymphocytes that express human leukocyte antigen (HLA) class I-specific inhibitory receptors. They are detectable as monoclonal expansions in the blood of cytomegalovirus (CMV)-seropositive individuals displaying particular HLA-Cw allotypes. Similar to NK cells, they are capable of killing various allogeneic tumor cell lines, a function referred to as "NK-like activity." The mechanism underlying this unusual functional property has recently been clarified. Via their T-cell receptor, NK-CTL recognize the nonclassical HLA class I molecule HLA-E, which is characterized by a limited polymorphism and by the ability to bind peptides derived from the leader sequence of various HLA class I alleles as well as from few viral proteins. The analysis of the T-cell receptor avidity revealed that NK-CTL recognize with high avidity a CMV UL40-derived peptide. The HLA-E-restricted recognition of CMV by NK-CTL may represent an important immunologic strategy in defenses against this virus. Indeed, unlike conventional CTL, NK-CTL mediated lysis is apparently not affected by the downregulation of major histocompatibility complex class I that occurs during CMV infection. Because the CMV UL40-derived peptide is identical to the one generated from the leader sequence of various HLA-Cw alleles, NK-CTL are also able to display an "HLA-E-dependent alloreactivity" against allogeneic target cells expressing appropriate HLA-Cw alleles. This broad ability to recognize and kill allogeneic cells may pose serious problems in transplantation.

The raft marker GM1 identifies functional subsets of granular lymphocytes in patients with CD3+ lymphoproliferative disease of granular lymphocytes

The raft marker GM1 is expressed at very low levels at the plasma membrane of resting T cells (GM1dull). In vitro T-cell activation induces synthesis of this lipid, which is then expressed at very high levels (GM1bright) at the membrane of activated/effector cells. By flow cytometry and confocal microscopy, we analyzed the expression and organization of GM1 in a series of 15 patients with CD3+ lymphoproliferative disease of granular lymphocytes (LDGL). We found that GM1bright GL were detectable in fresh blood samples obtained in all LDGL patients, although the range of brightly stained cells was extremely variable. This distinctive in vivo pattern has never been shown in T lymphocytes from healthy individuals or in patients with different chronic T or B lymphoproliferative disorders or active infectious diseases. The low number of cycling cells detected in LDGL patients was always included within the GM1bright GL population. Interestingly, GM1bright GL were demonstrated to contain a higher amount of IFN-gamma as compared to GM1dull GL. These findings allow to distinguish subsets of GL at different levels of activation within the monoclonal CD3+ population. The GM1bright GL subset is likely to be responsible for the renewing of GL and thus for maintaining chronic proliferation.

The inhibitory NK cell receptor CD94/NKG2A and the activating receptor CD94/NKG2C bind the top of HLA-E through mostly shared but partly distinct sets of HLA-E residues

The human non-classical MHC class I molecule HLA-E is a ligand for both an inhibitory NK cell receptor (CD94/NKG2A) and an activating receptor (CD94/NKG2C). To identify HLA-E surface recognized by both receptors, especially to determine if both receptors recognize the same epitope, we made a series of individually Ala-substituted HLA-E proteins and analyzed their binding to CD94/NKG2A orCD94/NKG2C. Eight HLA-E mutations that significantly impaired HLA-E binding to CD94/NKG2A are all found in the top of alpha1/alpha2 domain of HLA-E. These results suggest that CD94/NKG2A binds a HLA-E surface equivalent to a NKG2D binding site on MICA. Of the eight mutations that impaired HLA-E binding to CD94/NKG2A, six significantly impaired HLA-E binding to CD94/NKG2C suggesting that CD94/NKG2C also binds a similar surface of HLA-E. Unexpectedly, the two HLA-E mutations (D69A and H155A) selectively abrogated HLA-E binding to CD94/NKG2A, not largely affected CD94/NKG2C. These results indicate that a mostly shared, but partly distinct set of HLA-E residues is discriminated by the two receptors.

Qa-1, a nonclassical class I histocompatibility molecule with roles in innate and adaptive immunity

Qa-1, a nonclassical class I histocompatibility molecule expressed in mice, predominantly assembles with a single nonameric peptide, Qdm, derived from the signal sequence of certain class Ia molecules. The Qa-1/Qdm complex is the primary ligand for CD94/NKG2A inhibitory receptors expressed on a major fraction of natural killer (NK) cells. Cells become susceptible to killing by NK cells under conditions where surface expression of the Qa-1/Qdm inhibitory ligand is reduced. The CD94/NKG2 "missing-self" recognition system serves as mechanism for removing cells that have abnormalities in the intracellular machinery required for assembly and expression of class I-peptides complexes, as a consequence of viral infection, for example. Despite its highly focused peptide-binding specificity, Qa-1 also has a capacity to act as an antigen-presentation molecule for CD8+ T cells. It appears that a small subpopulation of these T cells undergoes positive selection by interaction with Qa-1 in the thymus, and they maintain their specificity for Qa-1 after maturation. The role of these unusual T cells in adaptive immune responses remains to be defined.

Comparative analysis of NK- or NK-CTL-mediated lysis of immature or mature autologous dendritic cells

Natural killer (NK) cells have been shown to kill efficiently autologous immature dendritic cells (iDC), while sparing those undergone maturation. In this study we investigated the effect of the interaction between autologous DC and NK-cytolytic T lymphocytes (NK-CTL), a subset of HLA-E-restricted CD8(+) T cells that express HLA class I-specific inhibitory NK receptors. Although these cells share with NK cells various phenotypic and functional features (such as the capacity to lyse most allogeneic, NK-susceptible tumor cell lines), different from NK cells, NK-CTL failed to lyse autologous DC. However, after pulsing DC with a cytomegalovirus-derived, HLA-E-binding peptide recognized by NK-CTL, both iDC and mature DC became highly susceptible to lysis. On the other hand,the addition of the peptide resulted in the down-regulation of the NK-mediated lysis of the same autologous iDC. The capability of killing autologous DC, presenting a non-self, HLA-E-binding peptide, may represent a feedback mechanism by which NK-CTL down-regulate HLA-E-restricted responses to certain pathogens.

Differential effects of US2, US6 and US11 human cytomegalovirus proteins on HLA class Ia and HLA-E expression: impact on target susceptibility to NK cell subsets

We compared in an inducible expression system the individual effect of US2, US6 and US11 human cytomegalovirus (HCMV) proteins on HLA-E and HLA class Ia surface expression, assessing in parallel their influence on target susceptibility to NK cell clones. To this end, the RPMI 8866 B lymphoma cell line (HLA-A2, HLA-A3, HLA-B7, HLA-Cw7, HLA-E(R), HLA-E(G)) was stably cotransfected with the ecdysone receptor, together with the US sequences under the control of an ecdysone-inducible promoter. Biosynthesis of viral proteins was turned on by incubating transfectants with Ponasterone A. US6 down-regulated expression of all class I molecules, hampering target resistance to NK cell clones controlled by the CD94/NKG2A, KIR2DL2 and/or CD85j (ILT2 or LIR-1) inhibitory receptors. By contrast, US11 reduced the surface levels of class Ia molecules but preserved HLA-E; this rendered US11(+) cells sensitive to NK clones under the control of KIR2DL2 and/or CD85j, while their resistance to CD94/NKG2A(+)KIR2DL2(-) effector cells was maintained. US2 preserved as well HLA-E expression but selectively targeted class Ia molecules; in fact, HLA-A and HLA-C allotypes were down-modulated whereas HLA-B7 remained unaltered. US2(+) targets became sensitive to KIR2DL2(+) cells but remained resistant to CD94/NKG2A(+)CD85j(+) NK clones. The differential effects of US proteins on HLA class Ia and HLA-E likely reflect the evolutionary adaptation of HCMV to counteract NK-mediated surveillance.

HLA-E-restricted recognition of cytomegalovirus-derived peptides by human CD8+ cytolytic T lymphocytes

HLA-E-restricted T cell receptor alphabeta+ CD8+ cytolytic T lymphocytes (CTLs) exist as monoclonal expansions in the peripheral blood of some individuals. Here, we show that they recognize, with high avidity, peptides derived from the UL40 protein of different human cytomegalovirus (CMV) strains. Recognition results in the induction of cytotoxicity, IFN-gamma production and cell proliferation. Autologous cells pulsed with CMV-derived peptides become susceptible to lysis by HLA-E-restricted CTLs and induce their proliferation. The high avidity for CMV-derived peptides may explain how these cells are generated in vivo and suggest their possible role in the host defenses against CMV, a virus that evolved various mechanisms to down-regulate classical HLA class I molecules, thus escaping detection by conventional CTLs.

Soluble nonclassical HLA generated by the metalloproteinase pathway

Soluble human leukocyte antigens (HLA-A, -B, and -C) proteins can be generated by a membrane-bound metalloproteinase (MPase). The MPase-mediated pathway produces soluble nonconformed HLA proteins susceptible to further degradation, and also HLA proteins with high affinity peptides stable at physiologic temperatures. Accessibility of classical HLA to the MPase cleavage inversely correlates with stability of heavy chain (HC) interactions with beta2-microglobulin (beta(2)m). Whether a MPase is involved in release of soluble nonclassical HLA or CD1 proteins is unknown. We have investigated this question with transfectants expressing full-length HLA proteins. Native surface HLA-E and -G complexes, similar to HLA-A2, were unstable at low pH and dissociated giving rise to beta(2)m-free HC. Furthermore, HLA-E and -G proteins, similar to HLA-A2, were readily released from cell surface into supernatants as soluble 37-kilodalton beta(2)m-free HC. However, the stability of surface CD1d complexes was not affected by pH changes and no soluble CD1d was detected. Because beta(2)m-free CD1d HC were expressed on cells, the lack of cleaved soluble products cannot be explained by high stability of native complexes. Instead, absence of a CD1d-specific MPase in these cells or its impaired interactions with substrate HC may be responsible.

Analysis of HLA-E peptide-binding specificity and contact residues in bound peptide required for recognition by CD94/NKG2

The MHC class Ib molecule HLA-E is the primary ligand for CD94/NKG2A-inhibitory receptors expressed on NK cells, and there is also evidence for TCR-mediated recognition of this molecule. HLA-E preferentially assembles with a homologous set of peptides derived from the leader sequence of class Ia molecules, but its capacity to bind and present other peptides remains to be fully explored. The peptide-binding motif of HLA-E was investigated by folding HLA-E in vitro in the presence of peptide libraries derived from a nonameric leader peptide sequence randomized at individual anchor positions. A high degree of selectivity was observed at four of five total anchor positions, with preference for amino acids present in HLA-E-binding peptides from class Ia leader sequences. Selectivity was also observed at the nonanchor P5 position, with preference for positively charged amino acids, suggesting that electrostatic interactions involving the P5 side chain may facilitate assembly of HLA-E peptide complexes. The observed HLA-E peptide-binding motif was strikingly similar to that previously identified for the murine class Ib molecule, Qa-1. Experiments with HLA-E tetramers bearing peptides substituted at nonanchor positions demonstrated that P5 and P8 are primary contact residues for interaction with CD94/NKG2 receptors. A conservative replacement of Arg for Lys at P5 completely abrogated binding to CD94/NKG2. Despite conservation of peptide-binding specificity in HLA-E and Qa-1, cross-species tetramer-staining experiments demonstrated that the interaction surfaces on CD94/NKG2 and the class Ib ligands have diverged between primates and rodents.

Natural killer receptors in patients with lymphoproliferative diseases of granular lymphocytes

The most relevant criterion for recognizing chronic granular lymphocyte (GL) proliferations, defined as lymphoproliferative disease of GL (LDGL), has been historically based on the number of the proliferating cells displaying typical GL morphology. With the extensive development of immunological and molecular techniques, two major groups of LDGL have been recognized, one belonging to the T-cell and the other to the NK cell lineage. The recent definition of a series of receptors of NK cells (NKR) and the identification of the specific targets recognized has expanded our knowledge of the properties of these cells and the discrimination between functional reactive and pathological proliferations. Some of these receptors are expressed by GL of T-cell lineage, suggesting a possible involvement in the genesis of GL proliferation. Following an extensive description of NKR in humans, this review will summarize the recent data on phenotypic and functional characteristics of NKR expressed by proliferating GL in patients with LDGL, discussing their role in this disorder.

Natural killer receptor repertoires in transplantation

Natural killer (NK) lymphocytes are potent effector cells that are controlled by the expression of a variety of cell surface receptors with either inhibitory or activating functions. The genetic and functional diversity of this receptor repertoire and the role of HLA class I molecules as a major group of NK receptor ligands create an innate alloreactive capacity in this cell type. Both animal models and in vitro studies have implicated NK cells as contributors to the pathology of clinical transplantation. However, recent clinical studies have indicated the potential benefit of exploiting NK cell alloreactivity in mismatched haematopoietic stem cell transplantation. Further investigations of NK cell alloreactivity will undoubtedly reveal additional applications of this fundamental cell type in clinical transplantation.