Negative regulation of rat natural killer cell activity by major histocompatibility complex class I recognition

The cytolytic activity of human and mouse natural killer (NK) cells is negatively regulated by self major histocompatibility complex (MHC) class I molecules on potential target cells. In the rat, protection by RT1 class I gene products has so far not been formally shown although the complex effects of foreign and self RT1 genes on polyclonal NK cell activity suggest that MHC recognition can have both stimulatory and inhibitory effects. Here we report that the expression of self-MHC class I molecules on target cells strongly inhibits lysis by a long term NK cell line derived from LEW (RT1^l) rats and by LEW NK cells activated by short-term culture in the presence of interleukin-2. This was demonstrated with mouse-rat hybridoma target cells expressing different rat MHC alleles and with mouse tumor target cells transfected with classical (RT1.A^l) and nonclassical (RT1.C^l) rat MHC class I genes. With hybridoma target cells, the strongest reduction in lysis as compared to the parental mouse myeloma line was observed when “self” (LEW) MHC was expressed, while hybridomas expressing other MHC alleles showed less and variable reduction. Transfection of RT1.A^l protected both L-929 fibroblasts and P815 mastocytoma cells from lysis by the NK cell line, while RT1.C^l only protected P815 cells, indicating that additional target cell properties regulate rat NK cell activity.     

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.     

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.     

Homozygosity at the major histocompatibility complex is required for optimal immunogenicity of bone marrow cell allografts in irradiated rats

Hemopoietic histocompatibility ( Hh ) Genes associated with the H-2 region control the antigenicity of hemopoietic cell grafts in the mouse. We have tested for similar genes in rats. Wistar Furth (WF, RTI ) or Lewis (LEW RTI ¹) bone marrow cell grafts did not proliferate in spleens of lethally irradiated (WFxLEW) Fl hybrid rats as assessed by measuring the incorporation of 5-iodo-2'deoxyuridine—¹²⁵I (IUdR) into recipient spleens 5 days after transplantation. In contrast, (WFxLEW)Fl hybrid marrow cells grew well in both WF and LEW parental strain hosts. (WFxDA)Fl or (WFxLEW)Fl hybrid rats were backcrossed to WF parental strain rats to produce progeny, either homozygous, or heterozygous for the MHC. The RTl type of 46 individual backcross progeny was determined using a 5 day mixed-lymphocyte reaction (MLR). Correlation between RTl type and growth of marrow grafts of individual backcross rats was determined bt using each rat as a bone marrow donor for irradiated LEW hosts. Marrow grafts from rats heterozygous for RTl were accepted in all 25 cases, whereas, grafts from 19 of 21 homozygous donors were rejected by the LEW hosts. Thus, homozygosity, for Hh determinants in or near the RTl region appears to be necessary for optimal immunogenicity of bone marrow allografts.     

The ability of H-2Dd molecule to affect natural resistance to hemopoietic allografts is an intrinsic property shared by Ddm1 but not Ld

F₁ hybrid resistance (HR) to parental bone marrow grafts is mediated by natural killer (NK) cells, and thought to be controlled by the non-class I hemopoietic histocompatibility (Hh) genes linked to the major histocompatibility complex (MHC). However, as in the in vitro NK cytotoxicity against hemopoietic targets, expression of certain class I MHC molecules does affect HR, although mechanisms underlying such an effect are not understood. In this study, we examine the relevance of the “self/non-self” property of class I molecules and the molecular domains responsible for this function. H-2^b/Hh-1^b lymphoma cells were transfected with class 1 H-2D^d or L^d gene, and its effect on the Hh-1 phenotype was examined by testing the transfectant's ability to competitively inhibit the in vivo rejection of parental H-2^b/Hh-1^b bone marrow grafts by irradiated F₁ hybrid hosts. Multiple independent clones of transfectants show that the genomic or cDNA of the D^d gene, but not of L^d, renders the Hh-1^b-positive cells incapable of inhibiting HR in F₁ mice, although both genes belong to the same region of the same haplotype. The same effect could be observed not only in H-2^b/d F₁ mice for which D^d and L^d are self, but also in H-2^b/k F₁ mice for which both D^d and L^d are non-self. Thus, this function of the D^d molecule is an intrinsic property, not necessarily related to its self/non-self characteristic relative to the effector cells. Furthermore, given the nature of the assay used in this study, the results favor a “target interference” model as the underlying mechanism of the D^d effect. To locate the relevant domain(s) of the D^d molecule, mutant D^dm1 gene was tested and found to have the same effect as the non-mutant D^d. D^dm1 is a hybrid molecule between D^d and L^d, sharing with D^d only the α1 domain and a portion of the α2 domain. The two N-terminal domains of D^dm1 differ from those of D^d by three amino acid substitutions, two of which affect the molecules' peptide-binding properties.     

Fine tuning of natural killer cell specificity and maintenance of self tolerance in MHC class I-deficient mice

TAP1 −/−, β2-microglobulin (β2m) −/− and TAP1/β2m −/− mice all express low but quantitatively different levels of MHC class I molecules. Using these mice, we have addressed questions relating to the fine tuning of natural killer (NK) cell specificity and maintenance of self tolerance in the NK cell system. NK cells from B6 wild-type mice killed target cells from TAP1 −/−, β2m −/− and TAP1/β2m −/− mice in vivo and rejected bone marrow grafts from the same mice in vivo at equivalent levels. NK cells from TAP1 −/−, β2m −/− mice did not kill target cells or reject bone marrow grafts from TAP1/β2m −/− mice. NK cells in all MHC class I-deficient mice were tolerant to autologous MHC class I-deficient cells, as revealed by in vitro cytotoxicity assays using NK cell effectors activated with the interferon-inducing agent Tilorone, or by in vivo bone marrow graft experiments. However, the self-tolerant state of MHC class I-deficient NK cells was broken by in vitro stimulation with IL-2 for 4 days. Under these conditions, NK cells from the MHC class I-deficient mice killed autologous MHC class I-deficient cells while MHC class I-positive targets were spared. The C-type lectin inhibitory receptor Ly49C has a specificity for H-2K^b and is expressed on a subset of NK1.1⁺ cells in B6 mice. Wild-type and all MHC class I-deficient mice had similar numbers of Ly49C-positive NK1.1⁺ cells. However, Ly49C expression was markedly down-regulated on NK1.1⁺ cells from B6 mice, as compared to TAP1 −/−, β2m −/− and TAP1/β2m −/− mice. In vitro stimulation of NK cells with IL-2 for 4 days did not significantly change this pattern. The present results are discussed in relation to the role of MHC class I molecules and Ly49 receptors in shaping the NK cell repertoire and raise new questions about maintenance of self tolerance in the NK cell system.     

Major histocompatibility complex-controlled protective influences on experimental autoimmune encephalomyelitis are peptide specific

The myelin basic protein (MBP) peptide 63–88-induced experimental autoimmune encephalomyelitis (EAE) and its associated T cell cytokine profile are influenced by the rat major histocompatibility complex (MHC). There is an allele-specific protective influence of the MHC class I region, whereas the MHC class II region display either disease-protective or -promoting effects. To investigate if the MHC-associated protection is dependent on certain combinations of MBP peptide and MHC molecules, we have now used another peptide (MBP 89–101). A broader and different set of rat MHC alleles were associated with EAE induced with MBP 89–101 as compared to MBP 63–88. All EAE-susceptible strains mounted peptide-specific strong T helper (Th) 1-like immune responses in vitro. Immunization of rats with an extended peptide (MBP 87–110) induced EAE associated with the same MHC haplotypes as the 89–101 peptide, except in LEW.1N (RT1ⁿ) rats which were relatively resistant. Only this strain responded with additional Th2-like and transforming growth factor-β responses to the peptide in vitro. In vivo depletion of CD8⁺ cells aggravated the disease in this strain. We conclude that both MHC-controlled promoting and protective influences on EAE are dependent on certain MHC/MBP peptide combinations, and that the 87–110 region of MBP contains a major MHC-associated encephalitogenic epitope in the rat.     

Genes in two major histocompatibility complex class I regions control selection, phenotype, and function of a rat Ly-49 natural killer cell subset.

We have generated a monoclonal antibody (STOK2) which reacts with an inhibitory MHC receptor on a subset of alloreactive NK cells in the rat. This receptor, termed the STOK2 antigen (Ag), belongs to the Ly-49 family of lectin-like molecules and displays specificity for the classical MHC class I molecule RT1-A1c of PVG rats. Here, we have investigated the influence of the MHC on the selection, phenotype and function of the STOK2+ NK subset in a panel of MHC congenic and intra-MHC recombinant strains. STOK2 receptor density was influenced by the presence of its classical MHC I ligand RT1-A1c, as evidenced by a reduction of STOK2 Ag on the surface of NK cells from RT1-A1c+, as compared with RT1-A1c-, strains. In addition, a role for nonclassical MHC I RT1-C/E/M alleles in the selection of the STOK2 Ag was demonstrated. The relative number of STOK2+ NK cells was fivefold higher in rats expressing the RT1-C/E/M(av1) as compared with those expressing the RT1-C/E/M(u) class Ib haplotype. The STOK2 ligand RT1-A1c inhibited cytotoxicity of STOK2+ NK cells regardless of effector cell MHC haplotype. Allospecificity of STOK2+ NK cells varied markedly with effector cell MHC, however, and suggested that inhibitory MHC I receptors apart from STOK2 were variably co-expressed by these cells. These data provide evidence for the MHC-dependent regulation of the allospecific repertoire within a subset of potentially autoreactive Ly-49+ rat NK cells.     

Variable immune cell frequencies in peripheral blood of LEW.1AR1‐iddm rats over time compared to other congenic LEW strains

The LEW.1AR1‐iddm rat is an animal model of human type 1 diabetes (T1D), which arose through a spontaneous mutation within the major histocompatibility complex (MHC)‐congenic background strain LEW.1AR1. The LEW.1AR1‐iddm rat is characterized by two phenotypes: diabetes development with a diabetes incidence of 60% and a variable T cell frequency in peripheral blood. In this study the immune cell repertoire of LEW.1AR1‐iddm rats was analysed over time from days 30 to 90 of life and compared to the background strain LEW.1AR1 and the LEW rat strain as well as the LEW.1WR1 rat strain. The LEW.1AR1‐iddm rats are characterized by a high variability of CD3⁺, CD4⁺ and CD8⁺ T cell frequencies in peripheral blood over time, and the frequency is unique for each animal. The variability within the frequencies resulted in changes of the CD4⁺ : CD8⁺ T cell ratio. The other three rat strains studied were characterized by a stable but nevertheless strain‐specific T cell frequency resulting in a specific CD4⁺ : CD8⁺ T cell ratio. The frequency of natural killer (NK) cells and B cells in LEW.1AR1‐iddm rats was increased, with a higher variability compared to the other strains. Only monocytes showed no differences in frequency and variability between all strains studied. These variabilities of immune cell frequencies in the LEW.1AR1‐iddm rats might lead to imbalances between autoreactive and regulatory T cells in peripheral blood as a prerequisite for diabetes development.     

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.     

Ly-49-independent inhibition of natural killer cell-mediated cytotoxicity by a soluble major histocompatibility complex class I molecule

Natural killer (NK) cells lyse their targets in a non-major histocompatibility complex (MHC)-restricted manner, and the cytotoxicity can be inhibited by a number of MHC class I allele products, suggesting that NK cells may have a “positive receptor” that recognizes the target and a “negative receptort” that receives an inhibitory signal from class I. Since negative receptors could also exert their effect by masking a positive ligand, we have determined whether there may be a direct interaction between class I and an NK surface receptor by measuring cytotoxicity in the presence of a soluble class I molecule, K^d. Soluble K^d at micromolar concentrations could efficiently block NK cell cytotoxicity, suggesting that class I has a direct effect on cytotoxicity, rather than masking another target cell ligand. Inhibition required that K^d be at least divalent, probably because of its higher affinity or its ability to cross-link the NK surface receptor. In addition, the effect was independent of the peptide used to load K^d, and there was inhibition of cytotoxicity of NK cells derived from either H-2^d or H-2^b mice. Finally, depletion of NK cells expressing Ly-49 had no effect on the specific inhibition by K^d, raising the possibility that NK cells are endowed with additional negative receptors besides Ly-49. Taken together, these results suggest that there may be a family of NK receptors recognizing different class I alleles, which can receive negative signals by directly binding to class I on the target cell surface.     

On the Heterologous Interaction between β2-Microglobulin and the Heavy Chain of Rat Major Histocompatibility Complex Class 1 Antigens

The heterologous interaction between β₂-microglobulin (β₂m) and rat major histocompatibility complex (MHC) (RT1) antigens was measured in a two-step binding assay consisting of binding of radiolabelled β₂-m to RT1 antigens and immunoprecipitation of β₂m-RT1 antigen complexes with RT1 antisera. The effects of varying the concentrations of the three reactants involved were studied. The molecular events taking place in the two steps were analysed by gel chromatography. The β₂m-RT1 antigen complex had the apparent size of albumin and reacted completely with specific alloantisera. RT1 antigens prepared from Wistar/Furth (RT1^u) and Brown Norway (RT1ⁿ), respectively, both effectively bound heterologous β₂m. The times for association and dissociation, respectively, at 37°C, were of the same order, but dissociation was slightly slower. Association was markedly temperature-dependent and was considerably slower at low temperatures. All these processes were slower for RT1^u than for RT1^u antigens. The association constant for the interaction between RT1^u antigens and ¹²⁵I-human β₂m was estimated by Scatchard analysis to be about 10⁹ M^-1. Contribution to the heterologous interaction by products from various rat MHC subloci (A, B, and C) was investigated by the introduction of sublocus-specific antisera in step 2. The reaction apparently involved neither class 2 antigens (sublocus B) nor the presumed rat Qa homologue (sublocus C). Classical class 1 antigens (suhlocus A) clearly contributed to the binding. However, a monoclonal antibody against products from rat MHC class 1 genes only precipitated less than half of the RTI antigen-complexed β₂m. Thus, at least two RT1^u class 1 alloantigen molecules seem to participate in the reaction. This, in turn, indicates that the rat genome may contain multiple class 1 genes, an is the case for most other mammals investigated.     

Cellular basis of graft-versus-host reactions

The biologic basis of Graft-Versus-Host Disease (GVHD) is presented as an extremely complex immunopathologic syndrome that involves interaction between many different donor and host cell types. A model of acute lethal GVHD was employed where adult unirradiated (DA X LEW)F₁ rats were injected with LEW spleen and lymph node cells. Controls received the same dose of syngeneic cells. At intervals from 2 to 21 days after cell injection, GVHD and control animals were killed and nonadherent cell suspensions prepared from their lymph nodes, spleen and peripheral blood. Cell suspensions were treated with LEW-anti-DA-alloantiserum or normal LEW serum and then analyzed for sIgM⁺ (B cells), W 3/13⁺ (T cells), and IgG-Fc receptors (FcR). Evidence is discussed for the selective removal of host cells with the alloantiserum. In addition, the level of naturally cytolytic (NK/NC) cells was assessed by adding GVHD and control nonadherent lymphoid cells to heterologous lymphoma and sarcoma target cells. Evidence is presented that during acute GVHD, in this parental →F₁ combination, there is an early increase within most compartments of donor as well as host W 3/13⁺ and W 3/13⁺FcR⁺ cells. NK/NC cells are increased as well at day 7. During middle stages of acute GVHD, host sIgM⁺ cells predominate. Late-stage acute GVHD rats contain few donor and host W 3/13⁺FcR⁺, and NK/NC cells but many null cells most of which are FcR^?. The importance of unraveling the nature of donor- and host-cell interactions occurring during acute GVHD, which result in rats whose lymphoid tissues are severely depleted of all nonadherent lymphoid cells but FcR^? null cells, is discussed.     

Allorecognition of isolated, denatured chains of class I and class II major histocompatibility complex molecules. Evidence for an important role for indirect allorecognition in transplantation.

Classical RT1-A class I and RT1-B class II major histocompatibility complex (MHC) molecules were purified from DA (RT1avl) spleens, and the individual chains separated and purified by preparative polyacrylamide gel electrophoresis in sodium dodecyl sulfate. LEW (RT1l) rats were immunized with the pure class I heavy chain, the RT1-B alpha chain and the RT1-B beta chain with the aim of priming to indirect allorecognition (i.e. after processing and presentation of DA MHC chains on LEW antigen-presenting cells) in the absence of any priming to direct allorecognition (i.e. to whole, undenatured, dimeric DA MHC molecules). LEW rats immunized with each of the three DA MHC chains produced alloantibodies to these chains, suggesting that indirect allorecognition did occur, because of the requirement for cognate recognition of B cells by T helper cells. This also demonstrated polymorphism of all three chains between the DA and LEW strains. The antibodies to the isolated, denatured MHC chains did not react to the whole MHC molecules on DA cells, with the possible exception of very weak reactions in some class I heavy chain-immunized rats. DA skin grafts placed on LEW recipients immunized with each of the DA MHC chains were rejected in an accelerated fashion. Following DA skin grafting, there was an accelerated production of antibodies to whole, undenatured class I MHC molecules, even in the LEW rats preimmunized with RT1-B alpha and RT1-B beta chains. These data suggest that indirect allorecognition can play an important role in the effector mechanisms of allograft rejection, and demonstrate T helper priming as one possible mechanism whereby this might be effective.     

Perforin dependence of natural killer cell-mediated tumor control in vivo

Adaptive immune surveillance by T cells against infections and tumors depends on the presence of antigenic peptides presented by major histocompatibility complex (MHC) molecules. If antigenic tumor-specific peptides or MHC class I molecules are absent, the adaptive T cell immune response fails. Natural killer (NK) cells seem to complement the specific T cells by recognizing target cells lacking MHC class I (e.g. RMA-S). The role of perforin, which is crucially involved in T cell and NK cell-mediated target cell lysis, was evaluated in mice lacking perforin with respect to their capacity to eliminate a syngeneic lymphoid tumor. Here, we show that growth of MHC class I^? RMA-S tumor cells in unprimed mice was controlled by NK cells through perforin-dependent cytotoxicity.     

Natural killer activity in (C57BL/6 × DBA/2)F1 hybrids undergoing acute and chronic graft-vs.-host reaction

The present findings demonstrate that a total i. v. transfer of 100 × 10⁶ C57BL/6 (B6) parental spleen cells into untreated (C57BL/6 × DBA/2)F₁ hybrids (B6D2F₁) resulted in acute runting, which was associated with a significantly elevated graft-vs.-host (GVH) index over a one-month period following GVH induction. Furthermore, this B6-induced acute GVH disease was associated with a marked depression of natural killer (NK) cell activity (spleen and peripheral blood) (with or without addition of mouse fibroblast interferon), which correlated with lymphoid cell hypocellularity, prominent splenic extramedullary hematopoiesis (EMH), and parallel depressions of both concanavalin A- and lipopolysaccharide-induced mitogenesis. Significantly increased killing by antibody-dependent cellular cytotoxicity of antibody-coated chicken red blood cells, as well as increased T cell killing of the NK-insensitive cell line P815 (as compared to the significantly decreased killing of the NK-sensitive cell line YAC-1) was also observed in the spleens of this 100 × 10⁶ B6-injected F₁ group. In marked contrast to this 100 × 10⁶ B6-injected acute GVH group, untreated mice injected i. v. with the same or greater numbers of parental DBA/2 spleen cells (100 × 10⁶-150 × 10⁶ DBA/2 spleen cells) exhibited a milder and more chronic form of GVH disease, which was not associated with a significant decrease of NK activity. It was of considerable interest that a total i. v. transfer of 50 × 10⁶ B6 spleen cells (i.e. one-half of that required to produce acute GVH, markedly depressed NK, and prominent splenic EMH) into B6D2F₁ hybrids also resulted in a more chronic form of GVH disease, but was associated with significantly increased levels of NK activity at two weeks post GVH induction.     

H-2K RESTRICTION OF THE T CELL-MEDIATED LYSIS OF A CHEMICALLY-INDUCED BALB/c FIBROSARCOMA

Previous work has shown that a cytotoxic T lymphocyte (CTL) immune response of syngeneic mice immunized with a chemically-induced BALB/c (H-2^d) fibrosarcoma was directed against an individual tumour-associated antigen. To see whether this reaction was restricted by products of the major histocompatibility complex (MHC), anti-H-2 alloantisera to K or D antigens were used to interfere with the CTL-mediated immune response. Antisera to K^d but not to D^d antigens inhibited the lytic activity of CTL against fibrosarcoma cells. In addition, the study of the CTL response in F₁→ P antitumour immunized chimeric mice showed that antitumour cytotoxicity developed only when F₁ and parental host shared the K^d region. Both experiments strongly indicate that recognition of the individual tumour-associated antigen of the BALB/c fibrosarcoma is restricted by the products of H-2K^d genes.     

Evidence for a similar or common mechanism for natural killer cell activity and resistance to hemopoietic grafts.

Two types of host reactivities not requiring immunization in the mouse and not mediated by T lymphocytes were compared: resistance of irradiated and nonirradiated F₁ hybrids to accept parental grafts of normal or malignant hemopoietic cells (Hh system), and the natural killer cell activity against mouse lymphomas (NK system). The effects of six independent variables known to influence resistance to marrow grafts were investigated in the NK system using YAC-1 lymphoma cells as targets. The following properties were shared: (a) maturation during the fourth week of life; (b) low sensitivity to acute total body irradiation; (c) dependence on the integrity of bone marrow as demonstrated by reduced reactivity in ⁸⁹Sr-treated mice; (d) suppression by a single injection of rabbit anti-mouse bone marrow serum; (e) suppression by a single injection of the anti-macrophage agents silica and i-carrageenan; and (f) suppression by multiple injections of parental spleen cells into F₁ mice. These positive correlations are particularly significant because most of the variables have either opposing or no effect on conventional immunity. F₁ mice rendered specifically unresponsive to parental marrow grafts, could retain NK cell activity, and genetically susceptible mice could be rendered hyporeactive in terms of NK cells, indicating that the specificities of YAC-1 and Hh-1 incompatible targets were different. It is extremely unlikely that this remarkable parallelism is fortuitous. These results indicate that either a very similar, or more likely a common mechanism is operative in the two cell-mediated natural reactivities: effector cells in the NK and Hh systems do not bear B or T lymphocyte markers but are nevertheless endowed with “specificity”. They are dependent for generation in vivo (presumably by maturation or by recruitment) on the interaction with nonlymphoid accessory cells not endowed with specificity, capable of also interacting in vitro with Thy-1-positive F₁ hybrid prekiller cells specific for parental targets. Because of thymus independence in vivo and apparent restriction to target cells of the hemopoietic system, these reactivities should be effective in the regulation of hemopoiesis and surveillance over leukemogenesis.     

Hybrid Resistance to EL-4 Lymphoma Cells

(C57BL/6 × DBA/2)F₁ hybrid (B6D2F₁) mice resist the growth of parental-strain (B6) EL-4 lmphoma cells inoculated intraperitoneally; that is, B6D2F₁ mice survive longer than B6 mice and do not develop uscites As compared with B6 mice, B6D2F₁ mice have higher levels of natural killer (NK) activity against ⁵¹Cr-labelled EL-4 cells in their lyphiod organs. B6D2F₁ mice treated with ⁸⁹Sr lose NK activity for certain lymphoma cell targets, e.g. YAC-1, but NK(EL-4) function is usually intact. However, ⁸⁹Sr-treated mice had hybrid resistance to EL-4 cell in vivo, as determined by survival times and the development of ascites. NK(EL-4) and NK-(YAC-1) activities were stimulated by irrudiated or unirradiated EL-4 cells, Corynebacterium parvum, or polyinosinic:polycytidylic acid (pl:pC) in spleens of normal B6D2F₁ mice, but NK(EL-4) activity was depressed Within 3 days by such treatment in B6D2F₁ mice previously injected with ⁸⁹Sr. Suppressor cells for NK(EL-4) but not for NK(YAC-1) effectors were easily detected in spleens of ⁸⁹Sr-treated mice ‘challenged’ with C. parvum. Thus, agents capable of Stimulating NK cell function in normal mice may lead to suppression of that activity in mice depleted of marrow-dependent cell function by ⁸⁹Sr. Spleen cells of ⁸⁹Sr-treated B6D2F₁ mice were also unable to generate anti-EL-4 cytotoxic T lymphocytes in a cell-mediated lympholysis system; this defect appeared also in to be mediated by suppressor cells. Lymphoid cell depleted by ⁸⁹Sr-induced marrow aplasia may have two functions in host defences against tumours (especially lymphomas): they may lyse tumour cells directly and they may ‘down-regulate’ suppressor cells capable of inhibiting other ‘natural’ or ‘induced’ immune functions.     

Attempts to immunize F1 hybrid mice against their parents

Attempts were made to immunize F₁ hybrid mice to their inbred parental strains by active immunization with either living or dead parental cells and by adoptive transfer of isogenic (F₁) lymph node cells to irradiated hybrids that subsequently received parental bone marrow. Cumulative mortality studies and erythrocyte serotyping revealed that many mice survived and retained their parental grafts for long periods of time. In most experiments reported here, ⁵⁹Fe uptake by erythrocytes and spleens of chimeras 1 week after marrow transplantation was used as a measure of erythropoiesis and thereby of success or failure of the marrow graft. Pre-treatment (`pre-immunization'') of hybrids with parental spleen cells produced no evidence of specific immunization. However, pre-treatment with spleen cells from parent 1 was detrimental to subsequently transplanted marrow from parent 2 and vice-versa. Adoptive transfer of viable F₁ lymph node cells failed to decrease growth of parental cells in irradiated F₁ hybrids. These findings indicate that the poor-growth phenomenon of particular parent—F₁ combinations cannot be explained in terms of classical immune rejection of parental cells by F₁ hybrids. Instead, growth of parental cells in the F₁ environment is inhibited or delayed.