β2 -Microglobulin-deficient NK cells show increased sensitivity to MHC class I-mediated inhibition,but self tolerance does not depend upon target cell expression of H-2Kb and Db heavy chains

Mice lacking β2 -microglobulin (β₂ m− mice) express greatly reduced levels of MHC class I molecules, and cells from β₂ m− mice are therefore highly sensitive NK cells. However, NK cells from β₂ m− mice fail to kill β₂ m− normal cells, showing that they are self tolerant. In a first attempt to understand better the basis of this tolerance, we have analyzed more extensively the target cell specificity of β₂ m− NK cells. In a comparison between several MHC class I-deficient and positive target cell pairs for sensitivity to β₂ m− NK cells, we made the following observations: First, β₂ m− NK cells displayed a close to normal ability to kill a panel of MHC class I-deficient tumor cells, despite their nonresponsiveness to β₂ m− concanavalin A (Con A)-activated T cell blasts. Secondly, β₂ m− NK cells were highly sensitive to MHC class I-mediated inhibition, in fact more so than β₂ m+ NK cells. Third β₂ m− NK cells were not only tolerant to β₂ m− Con A blasts but also to Con A blasts from H-2Kb − /Db − double deficient mice in vitro. We conclude that NK cell tolerance against MHC class I-deficient targets is restricted to nontransformed cells and independent of target cell expression of MHC class I free heavy chains. The enhanced ability of β₂ m− NK cells to distinguish between MHC class I-negative and -positive target cells may be explained by increased expression of Ly49 receptors, as described previously. However, the mechanisms for enhanced inhibition by MHC class I molecules appear to be unrelated to self tolerance in β₂ m− mice, which may instead operate through mechanisms involving triggering pathways.     

Expression of the Ly49A gene in murine natural killer cell clones is predominantly but not exclusively mono-allelic

The Ly49 family of natural killer (NK) cell receptors are major histocompatibility complex (MHC) class I-specific inhibitory receptors that are distributed to overlapping NK cell subsets. Extending earlier studies of polyclonal NK cell populations, we have employed an analysis of short-term NK cell clones from Ly49A heterozygous mice, to demonstrate that the Ly49A gene is usually expressed from one or the other allele in each Ly49A⁺ cell. However, we also detected a small percentage of clones that expressed both Ly49A alleles. The possibility that the colonies exhibiting bi-allelic Ly49A gene expression had been inoculated with more than one cell was ruled out by parallel analysis of clones isolated from mixtures of NK cells from Ly49A homozygous mice. The frequency of bi-allelic Ly49A⁺ clones suggested that the two Ly49A alleles in an NK cell are chosen for expression independently. These data are consistent with the proposal that mono-allelic Ly49A gene expression may arise as a consequence of a stochastic Ly49 gene activation mechanism. Analysis of Ly49A⁺ clones from MHC-different mice demonstrated that class I-deficient mice harbored a greater number of bi-allelic Ly49A⁺ cells than did H-2^d mice, which express a Ly49A ligand. Although the numbers were insufficiently large for a clear assignment, H-2^b mice may harbor an intermediate number of bi-allelic Ly49A⁺ NK cells. The effects of MHC expression on the prevalence of bi-allelic Ly49A⁺ cells suggest that an MHC-dependent education process modifies the Ly49 repertoire.     

Differential reactivity of residual CD8+ T lymphocytes in TAP1 and β2-microglobulin mutant mice

TAP1 -/- and β2-microglobulin (β2m) -/- mice (H-2^b background) express very low levels of major histocompatibility complex (MHC) class I molecules on the cell surface. Consequently these mice have low numbers of mature CD8⁺ T lymphocytes. However, TAP1 -/- mice have significantly higher numbers of CD8⁺ T cells than β2m -/- mice. Alloreactive CD8⁺ cytotoxic T lymphocyte (CTL) responses were also stronger in TAP1 -/- mice than in β2m -/- mice. Alloreactive CTL generated in TAP1 -/- and β2m -/- mice cross-react with H-2^b-expressing cells. Surprisingly, such cross-reactivity was stronger with alloreactive CTL from β2m -/- mice than with similar cells from TAP1 -/- mice. The β2m -/- mice also responded more strongly when primed with and tested against cells expressing normal levels of H-2^b MHC class I molecules. Such H-2^b-reactive CD8⁺ CTL from β2m -/- mice but not from TAP1 -/- mice also reacted with TAP1 -/- and TAP2-deficient RMA-S cells. In contrast, H-2^b-reactive CD8⁺ CTL from neither β2m -/- mice nor TAP1 -/- mice killed β2m -/- cells. In line with these results, β2m -/- mice also responded when primed and tested against TAP1 -/- cells. We conclude that the reactivity of residual CD8⁺ T cells differs between TAP1 -/- and β2m -/- mice. The MHC class I-deficient phenotype of TAP1 -/- and β2m -/- mice is not equivalent: class I expression differs between the two mouse lines with regard to quality as well as quantity. We propose that the differences observed in numbers of CD8⁺ T cells, their ability to react with alloantigens and their cross-reactivity with normal H-2^b class I are caused by differences in the expression of MHC class I ligands on selecting cells in the thymus.     

Lack of F1 anti-parental resistance in H-2b/d F1 hybrids devoid of β2-microglobulin

F₁ hybrid mice often reject parental hematopoietic grafts, a phenomenon known as hybrid resistance. Hybrid resistance is mediated by natural killer (NK) cells and although the molecular interactions responsible for this phenomenon are largely unknown, one hypothesis suggests that parental cells are rejected because they fail to express a complete set of host major histocompatibility complex (MHC) class I molecules. Inherent in this theory is that NK cells in the F₁ hybrid are instructed by self MHC class I molecules to form an NK cell repertoire capable of reacting against cells lacking these self MHC class I molecules. Here, we show that C57BL/6 x DBA/2 mice (H-2^b/d) devoid of β₂-microglobulin (β₂m) are incapable of rejecting β₂m?/? parental C57BL/6 cells (H-2^b) both in vivo and in vitro. From this, we conclude that the development of an NK cell repertoire, at least in F₁ mice of the H-2^b/d haplotype, requires expression of MHC class I molecules complexed with β₂m.     

Location-specific regulation of transgenic Ly49A receptors by major histocompatibility complex class I molecules

Inhibitory receptors expressed on natural killer (NK) cells and T cells specific for major histocompatibility complex (MHC) class I are believed to prevent these cells from responding to normal self tissues. To understand the regulation and function of Ly49 receptor molecules in vivo, we used the CD2 promoter to target Ly49A expression to all thymocytes, T cells, and NK cells. In animals expressing its MHC class I ligand, H-2D^d or H-2D^k, there was a large decrease in the expression of Ly49A on thymocytes, peripheral T cells, and NK1.1⁺ cells. The extent of the down-regulation of Ly49A was dependent on the expression of the MHC ligand for Ly49A and on the site where the cells were located. The level of expression of endogenous Ly49A was similarly found to be dependent upon the organ where the cells resided. Data from bone marrow chimeras indicated that most cell types may regulate Ly49A expression, but the efficacy to regulate receptor expression may vary depending on the cell type.     

Reduced expression of major histocompatibility complex class I free heavy chains and enhanced sensitivity to natural killer cells after incubation of human lymphoid lines with β2-microglobulin

Enhancement of major histocompatibility complex (MHC) class I expression leads to protection from recognition by natural killer (NK) cells in several systems. MHC class I gene products can be expressed in different forms at the cell surface - for example as “empty” β₂-microglobulin (β₂m)-associated heterodimers or free heavy chains. To study the role of different class I heavy chain forms in NK target interactions, we have used lymphoblastoid target cell lines preincubated with β₂m. This was found to shift the equilibrium between β₂m-associated and nonassociated - heavy chains in favor of the former. In parallel, there was a significant increase in NK sensitivity. The recognition of MHC class I-deficient cell lines was not affected by β₂m, arguing against a general nonspecific effect of fern on NK sensitivity. Our data indicate that protection against NK recognition correlates with target cell expression of free heavy chains (i.e. devoid of β₂m) rather than with expression of complexes.     

Major histocompatibility complex genes determine natural killer cell tolerance

Murine natural killer (NK) cell subsets, as defined by expression of members of the Ly49 gene family, discriminate target cells expressing different major histocompatibility complex (MHC) class I alleles. For example, Ly49A⁺ NK cells lyse H-2^b but not H-2^d tumor target cells. The specificity arises because D^d on target cells binds to Ly49A, transducing an inhibitory signal into the Ly49A⁺ NK cells. The capacity of NK cells to discriminate allelic class I determinants raises a key issue: are NK cells self-tolerant, and if so what are the mechanisms that lead to self-tolerance? As previously reported, potentially autoaggressive Ly49A⁺ NK cells are not clonally deleted in H-2^b mice. However, IL-2- cultured Ly49A⁺ effector cells from H-2^b mice exhibit reduced lysis of H-2^b (self) concanavalin A blast target cells, compared to Ly49A⁺ effector cells from H-2^d mice. Possible mechanisms accounting for this self-tolerance are addressed in this report. Self-tolerance was not due to anergy of the cells, because the Ly49A⁺ effector cells from both types of mice lysed β2-microglobulin-deficient target cells efficiently and equivalently. These results also suggest that tolerance results from inhibition mediated by β2m-dependent H-2^b class I molecules. Significantly, blockade of Ly49A on Ly49A⁺ effector cells from H-2^b mice did not restore lysis of H-2^b target cells, suggesting that inhibition is not mediated through the Ly49A receptor. Additional experiments suggest that inhibition is also not mediated primarily through the Ly49C receptor. These results suggest that Ly49A⁺ effector cells from H-2^b mice, unlike those from H-2^d mice, express inhibitory receptors specific for H-2^b molecules that are distinct from Ly49A and Ly49C.     

Glycosylation contributes to variability in expression of murine cytomegalovirus m157 and enhances stability of interaction with the NK‐cell receptor Ly49H

NK cell‐mediated resistance to murine cytomegalovirus (MCMV) is controlled by allelic Ly49 receptors, including activating Ly49H (C57BL/6 strain) and inhibitory Ly49I (129 strain), which specifically recognize MCMV m157, a glycosylphosphatidylinositol‐linked protein with homology to MHC class I. Although the Ly49 receptors retain significant homology to classic carbohydrate‐binding lectins, the role of glycosylation in ligand binding is unclear. Herein, we show that m157 is expressed in multiple, differentially N‐glycosylated isoforms in m157‐transduced or MCMV‐infected cells. We used site‐directed mutagenesis to express single and combinatorial asparagine (N)‐to‐glutamine (Q) mutations at N178, N187, N213, and N267 in myeloid and fibroblast cell lines. Progressive loss of N‐linked glycans led to a significant reduction of total cellular m157 abundance, although all variably glycosylated m157 isoforms were expressed at the cell surface and retained the capacity to activate Ly49H^B6 and Ly49I¹²⁹ reporter cells and Ly49H⁺ NK cells. However, the complete lack of N‐linked glycans on m157 destabilized the m157‐Ly49H interaction and prevented physical transfer of m157 to Ly49H‐expressing cells. Thus, glycosylation on m157 enhances expression and binding to Ly49H, factors that may impact the interaction between NK cells and MCMV in vivo where receptor–ligand interactions are more limiting.     

An allele-specific,stochastic gene expression process controls the expression of multiple Ly49family genes and generates a diverse,MHC-specific NK cell receptor repertoire

Mouse NK cells express MHC class I-specific inhibitory Ly49 receptors. Since these receptors display distinct ligand specificities and are clonally distributed, their expression generates a diverse NK cell receptor repertoire specific for MHC class I molecules. We have previously found that the D ^d (or D^k )-specific Ly49A receptor is usually expressed from a single allele. However, a small fraction of short-term NK cell clones expressed both Ly49A alleles, suggesting that the two Ly49A alleles are independently and randomly expressed. Here we show that the genes for two additional Ly49 receptors (Ly49C and Ly49G2) are also expressed in a (predominantly) mono-allelic fashion. Since single NK cells can co-express multiple Ly49 receptors, we also investigated whether mono-allelic expression from within the tightly linked Ly49 gene cluster is coordinate or independent. Our clonal analysis suggests that the expression of alleles of distinct Ly49 genes is not coordinate. Thus Ly49 alleles are apparently independently and randomly chosen for stable expression, a process that directly restricts the number of Ly49 receptors expressed per single NK cell. We propose that the Ly49 receptor repertoire specific for MHC class I is generated by an allele-specific, stochastic gene expression process that acts on the entire Ly49 gene cluster.     

A population of CD62Llow NK1.1− CD4+ T cells that resembles NK1.1+ CD4+ T cells

In MHC class II−/− C57BL/6 (II−/−) mouse spleen, a small population of CD4⁺ T cells is present of which NK1.1⁺ CD4⁺ (NK) T cells comprise 40 to 45 %. We report here that many of the NK1.1⁻ CD4⁺ T cells derived from II−/− mice are also NK T cells. They produce large amounts of IL-4 in response to anti-CD3 ligation and do so without any requirement for the presence of IL-4 in the priming culture, a property characteristic of NK T cells. Their IFN-γ production is large and is enhanced by IL-12. In addition, II−/− NK1.1⁻ CD4⁺ T cells produce IL-4 as a result of culture with L cells expressing murine CD1 (L-CD1). We report that CD49b, a component of integrin VLA-2, is expressed on the majority of both NK1.1⁺ and NK1.1⁻ NK T cells. NK1.1⁻ NK T cells also exist in wild-type C57BL/6 mice. Evidence supporting this is that Vβ8 usage by CD62L^low NK1.1⁻ CD4⁺ T cells was ∼ 5 % higher than that by CD62L^high CD4⁺ T cells in wild-type mice in keeping with the estimated proportion of NK1.1⁻ NK T cells in the CD62L^low population. CD62L^low CD4⁺ T cells from β2-m−/− mice, which lack NK T cells, showed no increase in Vβ8 usage. When activated by anti-CD3 or L-CD1, CD62L^low NK1.1⁻ CD4⁺ T cells from conventional but not β2-m−/− and CD1−/− mice produce IL-4 in a manner indistinguishable from II−/− NK1.1⁻ CD4⁺ T cells. NK1.1⁻ NK T cells in normal mouse spleens are approximately as numerous as NK1.1⁺ NK T cells.     

Viral infection transiently reverses activation receptor‐mediated NK cell hyporesponsiveness in an MHC class I‐independent mechanism

Continuous engagement of the Ly49H activating receptor with its ligand (m157) in a transgenic mouse expressing m157 (m157‐Tg) results in hyporesponsiveness of Ly49H⁺ NK cells. The same interaction, during murine cytomegalovirus (MCMV) infection, leads to activation of Ly49H⁺ NK cells. MCMV infection results in decreased MHC class I (MHC‐I) expression on the infected cell as well as inflammatory responses, both of which do not take place in the uninfected m157‐Tg mouse, potentially allowing for activation of NK cells in the context of MCMV infection. In this study, we demonstrated that viral infection transiently reverses activation receptor‐mediated NK cell hyporesponsiveness in an MHC‐I‐independent mechanism. Furthermore, Ly49H⁺ NK cells in an MHC‐I‐deficient environment remained hyporesponsive in the context of m157 expression, even when mature WT splenocytes were transferred into m157‐Tg mice in an MHC‐I‐deficient environment. However, the administration of cytokines TNF‐α, IL‐12, and IFN‐β resulted in a partial recovery from activation receptor‐induced hyporesponsiveness. Thus, the release of the aforementioned cytokines during MCMV infection and not the downregulation of MHC‐I expression appears to be responsible for partial resolution of Ly49H receptor‐induced NK cell hyporesponsiveness.     

α1 / α2 domains of H-2Dd,but not H-2Ld,induce “missing self” reactivity in vivo – No effect of H-2Ld on protection against NK cells expressing the inhibitory receptor Ly49G2

Introduction of the MHC class I transgene H-2D^d on C57BL / 6 (B6) background conveys NK cell-mediated “missing self” reactivity against transgene-negative cells, and down-regulates expression of the inhibitory receptors Ly49A and Ly49G2 in NK cells. We here present an analysis of transgenic mice expressing chimeric H-2D^d / L^d MHC class I transgenes, and show that the α1 / α2 domains of H-2D^d were necessary and sufficient to induce “missing self” recognition and to down-modulate Ly49A and Ly49G2 receptors. In contrast, transgenes containing the α1 / α2 domains of H-2L^d induced none of these changes, suggesting that not all MHC class I alleles in a host necessarily take part in NK cell education. The lack of effect of the α1 / α2 domains of H-2L^d on NK cell specificity was surprising, considering that both H-2L^d and H-2D^d have been reported to interact with Ly49G2. Therefore, the role of H-2L^d for protection against NK cells expressing Ly49G2 was re-investigated in a transfection system. In contradiction to earlier reports, we show that H-2D^d, but not H-2L^d, abolished killing by sorted Ly49G2⁺ NK cells, indicating that H-2L^d does not inhibit NK cells via the Ly49G2 receptor.     

MHC-dependent shaping of the inhibitory Ly49 receptor repertoire on NK cells: evidence for a regulated sequential model.

Engagement of MHC class I-specific inhibitory receptors regulates natural killer (NK) cell development and function. Using both new and previously characterized anti-Ly49 monoclonal antibodies, we comprehensively determined expression and co-expression frequencies of four Ly49 receptors by NK cells from MHC-congenic, MHC class I-deficient, and Ly49A-transgenic mice to study mechanisms that shape the inhibitory Ly49 repertoire. All Ly49 receptors were expressed on partially overlapping subsets. Significantly, in the absence of class I MHC, several receptor pairs were co-expressed more frequently than predicted from a purely random expression model, indicating that biases independent of MHC class I underlie receptor co-expression in some cases. MHC interactions were found to inhibit Ly49 co-expression variably depending on the MHC allele and the receptor pair examined. These data extend previous evidence that interactions with MHC shape the repertoire. It was previously proposed that developing NK cells express Ly49 receptors sequentially and cumulatively, until self-MHC specific receptors are expressed and inhibit new receptor expression. Fulfilling a major prediction of this model, we found that class I recognition by a Ly49A transgene expressed by all developing NK cells equivalently inhibited expression of endogenous self-specific and nonself-specific Ly49 receptors.     

The Qa-1b molecule binds to a large subpopulation of murine NK cells

Recent studies on human NK cells have demonstrated that the NK cell CD94/NKG2 receptors bind to the nonclassical MHC class I molecule HLA-E. A functional CD94/NKG2 complex has not yet been identified in rodents, but cDNA encoding rat and mouse CD94 and NKG2 have recently been cloned, suggesting that CD94/NKG2 receptors may exist in species other than man. The mouse nonclassical MHC class I molecule Qa-1 shares several features with HLA-E. This suggests that Qa-1 may be similarly recognized by murine NK cells. To study the ability of Qa-1 to bind to murine NK cells, we have produced a soluble tetrameric form of Qa-1^b . In the present study, we demonstrate that Qa-1^b tetramers distinctly bind to a large subset of fresh or IL-2-activated NK1.1⁺ /CD3⁻ splenocytes independently of the expression of Ly49 inhibitory receptors. Binding occurs whether NK cells have evolved in an MHC class I-expressing or in an MHC class I-deficient environment. Our data suggest the existence of a Qa-1-recognizing structure on a large subpopulation of murine NK cells that may be similar to the human CD94/NKG2 heterodimeric complex.     

Control of thymus-independent intestinal intraepithelial lymphocytes by β2-microglobulin

Murine intestinal intraepithelial lymphocytes (i-IEL) comprise thymusdependent cells such as T cell receptor (TcR) α/β CD8α/β⁺ i-IEL, as well as thymus-independent ones such as TcRα/β CD8α/α⁺ and TcRγ/δ CD8α/α⁺ i-IEL. Whilst the development of the CD8α/β expressing i-IEL is strictly contingent on major histocompatibility complex (MHC) class I surface expression, that of CD8α/α i-IEL appears largely MHC class I independent. We have used β2-microglobulin (β2m)^?/? mutant mice lacking surface-expressed MHC class I and TcRα/β CD8α/β⁺ i-IEL to analyze the potential impact of MHC class I on regional activation of thymus-independent i-IEL. To analyze the role of TcRγ/δ i-IEL in regional cell interactions, these mice were treated with the anti-TcRγ/δ mAb, GL3. Whilst numbers of TcRα/β CD8α/α i-IEL were markedly reduced in βm^+/? mice, those of TcRγ/δ i-IEL were elevated. Administration of GL3 in vivo caused TcR down-modulation and functional inactivation of TcRγ/δ i-IEL in β2m^+/? mice. In contrast, TcR expression and functional activities of TcRγ/δ i-IEL from β2m^?/? mice were not impaired by GL3 treatment. The TcRα/β CD8β i-IEL from β2m^?/? mice were expanded and functionally activated as a consequence of TcRγ/δ engagement. The TcRγ/δ i-IEL and TcRα/β CD8α/α⁺ i-IEL from athymic nu/nu mice which express MHC class I, but lack TcRα/β CD8α/β⁺ i-IEL, responded to TcRγ/δ engagement as those from the β2m^+/? controls. In addition, the TcRγ/δ i-IEL from TcRβ^?/? and TcRβ^+/? mutants were equally affected by GL3. We conclude that the absence of β2m renders TcRγ/δ i-IEL resistant to TcR-mediated inactivation and promotes activation of TcRα/β CD8β^? i-IEL. The activation of TcRγ/δ i-IEL seems to be directly controlled by β2m/MHC class I expression and independent from TcRα/β CD8β⁺ i-IEL. Regulation of self-reactive thymus-independent i-IEL through β2m/MHC class I may contribute to control of autoreactive immune responses in the intestine.     

Surface expression of β2-microglobulin-associated thymus-leukemia antigen is independent of TAP2

Mouse thymus-leukemia antigen (TL), like other major histocompatibility complex (MHC) class I-b antigens, displays signs of a specialized function. It is normally expressed at high levels on immature thymocytes and at moderate levels on gut epithelium and activated mature T cells. A promoter/enhancer region unique among class I genes accounts for this narrow range of tissue distribution. Like most other class I molecules, TL is dependent upon endogenous β2-microglobulin (β2m) for transport to the surface. However, here we show that unlike most other MHC class I molecules, TL is expressed efficiently in the absence of functional transporter associated with antigen processing subunit 2 (TAP2). A putative fourth TL^a gene cloned from A.SL1 cells was expressed in RMA and RMA-S cells. In bulk transformants, TL expression is higher in TAP2^? RMA-S cells than in wild-type RMA cells, and is not elevated by incubation at reduced temperatures or exposure to exogenous β2m. Analysis of immunoprecipitasted molecules by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate indicates that TL is processed normally in RMA-S cells and is associated with β2m both intracellularly and at the cell surface. However, TL heavy chains expressed on the cell surface in the absence of TAP2 are cleaved to a predominant 38 kDa fragment, presumably the result of an altered conformation that renders TL more susceptible to proteolysis. These results suggest that while TL may normally acquire TAP2-dependent peptides, this class I-b molecule does not require them for efficient export to, and stable expression at the cell surface.     

Quantifying the reduction in accessibility of the inhibitory NK cell receptor Ly49A caused by binding MHC class I proteins in cis

Murine natural killer (NK) cells are inhibited by target cell MHC class I molecules via Ly49 receptors. However, Ly49 receptors can be made inaccessible to target cell MHC class I by a cis interaction with its MHC class I ligand within the NK cell membrane. It has recently been demonstrated that MHC class I proteins transfer from the target cells to the NK cell. Here, we establish that the number of transferred MHC class I proteins is proportional to the number of Ly49A receptors at the NK cell surface. Ly49A+ NK cells from mice expressing the Ly49A ligand H-2D(d) showed a 90% reduction in Ly49A accessibility compared to Ly49A+ NK cells from H-2D(d)-negative mice. The reduction was caused both by lower expression of Ly49A and interactions in cis between Ly49A and H-2D(d) at the NK cell surface. Approximately 75% of the Ly49A receptors on H-2D(d)-expressing NK cells were occupied in cis with endogenous H-2D(d) and only 25% were free to interact with H-2D(d) molecules in trans. Thus, H-2D(d) ligands control Ly49A receptor accessibility through interactions both in cis and in trans.     

Differential Expression of the Ly49GB6, but Not the Ly49GBALB,Receptor Isoform during Natural Killer Cell Reconstitution after Hematopoietic Stem Cell Transplantation

Inhibitory natural killer (NK) cell receptors specific for major histocompatibility complex class I (MHC-I) molecules include Ly49 receptors in mice and killer immunoglobulin-like receptors (KIR) in humans. The “licensing” or “arming” models imply that engagement of these receptors to self MHC-I molecules during NK cell development educates NK cells to be more responsive to cancer and viral infection. We recently reported that hematopoietic stem cell transplantation (HSCT) induced rapid and preferential expansion of functionally competent Ly49G⁺, but not other Ly49 family, NK cells independent of NK cell licensing via Ly49–MHC-I interactions. We now extend these studies to evaluate expression of the two Ly49G receptor isoforms Ly49G^B6 and Ly49G^BALB, using mice with different MHC-I haplotypes that express one or both of the isoforms. NK cells from CB6F₁ (H-2^bxd) hybrid mice express two different alleles for Ly49G receptor, Ly49G^B6 and Ly49G^BALB. We found that CB6F₁ mice had more Ly49G^B6+ NK cells than Ly49^BALB+ NK cells, and that only Ly49G^B6+ NK cells increased in relative numbers and in Ly49G mean fluorescence intensity values after HSCT similar to the B6 parental strain. We further observed that Ly49G⁺ NK cells in BALB/c (H-2^d) and BALB.B (H-2^b) mice, which have the same background genes, recover slowly after HSCT, in contrast to Ly49G⁺ NK cells in B6 (H-2^b) recipients. The difference in expression of Ly49G^B6 relative to Ly49G^BALB was linked to differences in the activity of the Pro1 promoter between the two alleles. Thus, we conclude that the Ly49G^B6 receptor dominates Ly49G expression on NK cells after HSCT in strains in which that allele is expressed. The data suggest that Ly49 allelic polymorphism within a particular Ly49 family member can differentially affect NK cell recovery after HSCT depending on the background genes of the recipient, not on the MHC-I haplotype.     

Impaired NK‐cell education diminishes resistance to murine CMV infection

Ly49G2 (G2⁺) NK cells mediate murine (M)CMV resistance in MHC D^k‐expressing mice. Bone marrow transplantation (BMT) studies revealed that G2⁺ NK cell‐mediated MCMV resistance requires D^k in both hematopoietic and nonhematopoietic cells. As a Ly49G2 ligand, D^k in both cell lineages may contribute to lysis of virus‐infected cells. Alternatively, cellular differences in self‐MHC D^k may have affected NK‐cell education, and consequently NK cell‐mediated viral clearance. We investigated the D^k‐licensing effect on BM‐derived NK cells in BMT recipients by analyzing cytokines, cytotoxicity and MCMV resistance. In BMT recipients with lineage‐restricted D^k, G2⁺ NK‐cell reactivity and cytotoxicity was diminished in comparison to BMT recipients with self‐MHC in all cells. Reduced G2⁺ NK‐mediated MCMV resistance in BMT recipients with lineage‐restricted self‐MHC indicates that licensing of G2⁺ NK cells is related to NK‐cell reactivity and viral control. Titrating donor BM with self‐MHC‐bearing hematopoietic cells, as well as adoptive transfer of mature G2⁺ NK cells into BMT recipients with self‐MHC in non‐hematopoietic cells only, enhanced NK‐cell licensing and rescued MCMV resistance. This disparate self‐MHC NK‐cell education model would suggest that inadequately licensed NK cells corresponded to inefficient viral sensing and clearance.     

Inhibition of natural killer cell-mediated bone marrow graft rejection by allogeneic major histocompatibility complex class I,but not class II molecules

The role of major histocompatibility complex (MHC) class I and class II molecules in natural killer (NK) cell-mediated rejection of allogeneic, semi-syngeneic and MHC-matched bone marrow grafts was investigated. The use of β₂-microglobulin (β₂m) -/- and β₂m +/- mice as bone marrow donors to MHC-mismatched recipients allowed an analysis of whether the presence of semi-syngeneic and allogeneic MHC class I gene products would be triggering, protective or neutral, in relation to NK cell-mediated rejection. Loss of β₂m did not allow H-2^b bone marrow cells to escape from NK cell-mediated rejection in allogeneic (BALB/c) or semi-allogeneic (H-2D^d transgenic C57BL/6) mice. On the contrary, it led to stronger rejection, as reflected by the inability of a larger bone marrow cell inoculum to overcome rejection by the H-2-mismatched recipients. In H-2-matched recipients, loss of β₂m in the graft led to a switch from engraftment to rejection. At the recipient level, loss of β₂m led to loss of the capability to reject H-2-matched β₂m-deficient as well as allogeneic grafts. When MHC class II-deficient mice were used as donors, the response was the same as that against donors of normal MHC phenotype: allogeneic and semi-syngeneic grafts were rejected by NK cells, while syngeneic grafts were accepted. These data suggest a model in which allogeneic class I molecules on the target cell offer partial protection, while certain syngeneic class I molecules give full protection from NK cell-mediated rejection of bone marrow cells. There was no evidence for a role of MHC class II molecules in this system.