Correlation of glycosphingolipids and sialic acid in YAC-1 lymphoma variants with their sensitivity to natural killer-cell-mediated lysis

Exploiting NK Cell Surveillance Pathways for Cancer Therapy

Natural killer (NK) cells can evoke potent anti-tumour activity. This function is largely mediated through a battery of specialised cell-surface receptors which probe the tissue microenvironment for changes in surface and secretory phenotypes that may alert to the presence of infection or malignancy. These receptors have the potential to arouse the robust cytotoxic and cytokine-secreting functions of NK cells and so must be tightly regulated to prevent autoimmunity. However, such functions also hold great promise for clinical intervention. In this review, we highlight some of the latest breakthroughs in fundamental NK cell receptor biology that have illuminated our understanding of the molecular strategies NK cells employ to perceive malignant cells from normal healthy cells. Moreover, we highlight how these sophisticated tumour recognition strategies are being harnessed for cancer immunotherapies in the clinic.

Tailoring Natural Killer cell immunotherapy to the tumour microenvironment

Natural killer (NK) cells are cytotoxic and cytokine-secreting cells that can mediate potent anti-tumour activity. Accumulating evidence indicates that NK cell functions are severely compromised within the confines of the tumour microenvironment thus impairing the efficacy and development of NK cell-based therapies. Here we review the various cellular and molecular pathways that tumours have supplanted to evade NK cell surveillance. We highlight novel strategies designed to alleviate or circumvent the immunosuppressive conditions of the tumour microenvironment in order to emancipate NK cell function and stifle the inexorable growth and metastasis of malignant cells.

Glycocalyx engineering reveals a Siglec-based mechanism for NK cell immunoevasion

The increase of cell surface sialic acid is a characteristic shared by many tumor types. A correlation between hypersialylation and immunoprotection has been observed, but few hypotheses have provided a mechanistic understanding of this immunosuppressive phenomenon. Here, we show that increasing sialylated glycans on cancer cells inhibits human natural killer (NK) cell activation through the recruitment of sialic acid-binding immunoglobulin-like lectin 7 (Siglec-7). Key to these findings was the use of glycopolymers end-functionalized with phospholipids, which enable the introduction of synthetically defined glycans onto cancer cell surfaces. Remodeling the sialylation status of cancer cells affected the susceptibility to NK cell cytotoxicity via Siglec-7 engagement in a variety of tumor types. These results support a model in which hypersialylation offers a selective advantage to tumor cells under pressure from NK immunosurveillance by increasing Siglec ligands. We also exploited this finding to protect allogeneic and xenogeneic primary cells from NK-mediated killing, suggesting the potential of Siglecs as therapeutic targets in cell transplant therapy.

Sialylation of 3-methylcholanthrene-induced fibrosarcoma determines antitumor immune responses during immunoediting

Sialylation of tumor cells is involved in various aspects of their malignancy (proliferation, motility, invasion, and metastasis); however, its effect on the process of immunoediting that affects tumor cell immunogenicity has not been studied. We have shown that in mice with impaired immunoediting, such as in IL-1α(-/-) and IFNγ(-/-) mice, 3-methylcholanthrene-induced fibrosarcoma cells are immunogenic and concomitantly bear low levels of surface sialylation, whereas tumor cells derived from wild type mice are nonimmunogenic and bear higher levels of surface sialylation. To study immune mechanisms whose interaction with tumor cells involves surface sialic acid residues, we used highly sialylated 3-methylcholanthrene-induced nonimmunogenic fibrosarcoma cell lines from wild type mice, which were treated with sialidase to mimic immunogenic tumor cell variants. In vivo and in vitro experiments revealed that desialylation of tumor cells reduced their growth and induced cytotoxicity by NK cells. Moreover, sialidase-treated tumor cells better activated NK cells for IFN-γ secretion. The NKG2D-activating receptor on NK cells was shown to be involved in interactions with desialylated ligands on tumor cells, the nature of which is still not known. Thus, the degree of sialylation on tumor cells, which is selected during the process of immunoediting, has possibly evolved as an important mechanism of tumor cells with low intrinsic immunogenicity or select for tumor cells that can evade the immune system or subvert its function. When immunoediting is impaired, such as in IFN-γ(-/-) and IL-1α(-/-) mice, the overt tumor consists of desialylayed tumor cells that interact better with immunosurveillance cells.

Phaseolus vulgaris leukoagglutinating lectin-binding reactivity in human diffuse large B-cell lymphoma and its relevance to the patient's clinical outcome: lectin histochemistry and lectin blot analysis

Many reports have suggested a strong correlation between certain lectin-binding patterns and biological behavior in various tumors. To clarify a relationship between lectin-binding reactivity and survival of patients with diffuse large B-cell lymphoma (B-DLCL), 57 cases with B-DLCL were analyzed by lectin histochemistry and lectin blot method with or without treatment of neuraminidase or acidic hydrolytic conditions. B-DLCL cases were grouped into three types based on the data on lectin-binding reactivity under neuraminidase-treated or untreated conditions: (i) Group A (non-reactive type); (ii) Group B (sialylated type); and (iii) Group C (non-sialylated type). Among various lectins, Phaseolus vulgaris-L (L-PHA) binding reactivity showed that the survival of patients with Group A + B or Group B was significantly shorter than that of patients with Group C. Lectin blot analysis revealed failure of L-PHA-binding to 32 kd and 29 kd glycoproteins, which may be attributable to the masking of L-PHA-binding sites by sialylation or the lack of L-PHA-binding sites, leading to the short survival of patients with B-DLCL. L-PHA-binding reactivity may be a useful marker for the evaluation of survival of patients with B-DLCL.

The significance of hypersialylation of dipeptidyl peptidase IV (CD26) in the inhibition of its activity by Tat and other cationic peptides. CD26: a subverted adhesion molecule for HIV peptide binding

The functionality of DPP-IV, purified from human placenta and isolated from CD4+/CD26+ T cells of noninfected and HIV-1-infected individuals, was investigated as to its ability to bind certain specific peptides. Using isoelectric focusing and the specificity of substrate-impregnated overlay membranes, we found that DPP-IV from term placenta and from T cells of HIV-infected individuals was significantly more sialylated compared with enzyme isozyme patterns of other tissues. We report here that (1) the number of isoforms of DPP-IV and extent of sialylation are critical to function and peptide binding; (2) the number of sialylated isoforms isolated from PBMCs increases significantly with age greater than 40 years; (3) hypersialylation by extreme anionic isoforms is highly associated with HIV infection and pathognomonic to remaining CD4+ cells in overt AIDS; and (4) highly sialylated DPP-IV is more significantly inhibited by Tat and cationic peptides. We conclude that hypersialylation of DPP-IV modifies surface charge of the CD26 antigen, promoting binding of HIV peptides through their cationic domains to the sialic acid residues of DPP-IV, and that certain HIV moieties are likely to engage this phenomenon as an auxiliary adhesion mechanism to fuse with cells. Furthermore, as a consequence of this occurrence, DPP-IV enzymatic activity can be significantly reduced, competitively.

Glycophorin A protects K562 cells from natural killer cell attack. Role of oligosaccharides

Glycophorin A is a protein with an abundant glycosylation (60% carbohydrate by weight), and studies have suggested that resistance of target cells to natural killing may be correlated with the level of glycophorin A expression. To assess the role of glycophorin A and of its carbohydrates in sensitivity to lysis by natural killer (NK) cells, the glycoprotein was inserted into the membrane of K562 target cells using electropulsation. Peripheral blood lymphocytes were used as effector cells. When glycophorin A was inserted into the membrane, the level of resistance to NK cell attack increased with the number of glycophorin A molecules electroinserted. The resistance to lysis was not due to a defect in target cell-effector cell binding. Electroassociation of glycophorin A did not cause a decrease in the expression of either "positive signals" for NK cells (such as CD71, CD15, and CD32 antigens) or cellular adhesion molecules (CD18, CD29, CD54, and CD58). Furthermore, electroinsertion of glycophorin A did not trigger any "negative signals," such as class I HLA antigen expression. Finally, it was shown that the sialic acid and O-linked oligosaccharides of glycophorin A did not play any role in its effect against NK cells. Conversely, the unique N-linked oligosaccharide was shown to be essential for resistance to occur.

Serum gangliosides as endogenous immunomodulators

Gangliosides suppress various immune activities in vitro and in vivo. Their level is significantly elevated in tumors and atherosclerotic aorta tissue, as well as in the sera of patients with tumors or atherosclerosis. Here, Lev Bergelson suggests that ganglioside-induced immunomodulation might be involved in atherogenesis and carcinogenesis, and describes a hypothesis that cites gangliosides as a factor interfering with the clearance of low-density lipoproteins (LDLs) and promoting the formation of atherosclerotic plaques.

Oligosaccharide ligands for NKR-P1 protein activate NK cells and cytotoxicity

A diversity of high-affinity oligosaccharide ligands are identified for NKR-P1, a membrane protein on natural killer (NK) cells which contains an extracellular Ca(2+)-dependent lectin domain. Interactions of such oligosaccharides on the target cell surface with NKR-P1 on the killer cell surface are crucial both for target cell recognition and for delivery of stimulatory or inhibitory signals linked to the NK cytolytic machinery. NK-resistant tumour cells are rendered susceptible by preincubation with liposomes expressing NKR-P1 ligands, suggesting that purging of tumour or virally infected cells in vivo may be a therapeutic possibility.

Tumour necrosis factor-alpha induces an increase in susceptibility of human glioblastoma U87-MG cells to natural killer cell-mediated lysis

The mechanism by which tumour necrosis factor (TNF)-alpha increases the susceptibility of U87-MG human glioblastoma cells to lysis by natural killer (NK) cells was studied. Treatment with TNF-alpha (100 units ml-1) for 48 h enhanced the susceptibility of tumour cells to lysis by NK cells. Increased susceptibility to lysis was associated with enhanced expression of intercellular adhesion molecule 1 (ICAM-1) and HLA class I antigen. Antisense ICAM-1 oligonucleotide inhibited lysis by NK cells of TNF-alpha-treated tumour cells. In contrast, acid treatment following TNF-alpha treatment increased lysis by NK cells. These findings indicate that TNF-alpha treatment of glioblastoma cells increased their susceptibility to lysis by NK cells, since ICAM-1 up-regulation would have more profound effects on NK susceptibility than would HLA class I antigen up-regulation.

CDw75 antigen expression in human gastric carcinoma and adjacent mucosa

BACKGROUND

The amount of sialic acid correlates with the invasiveness and metastasizing potential of several human tumors. The CDw75 epitope is a sialylated carbohydrate determinant generated by the beta-galactosyl alpha 2,6-sialyltransferase, which can be viewed as a target for identifying biologically aggressive tumors.

METHODS

The authors performed an immunohistochemical study of CDw75 epitope expression in 87 cases of gastric carcinoma and adjacent mucosa and 331 metastases (329 lymph node metastases and 2 liver metastases) with the monoclonal antibody HH2.

RESULTS

Normal-appearing mucosa, foci of intestinal metaplasia, and foveolar hyperplasia, adjacent to the carcinomas, were mainly nonimmunoreactive. Only a few parietal cells of the body mucosa were stained with HH2. Two of 12 cases with dysplasia showed CDw75 antigen expression in dysplastic glands. Forty-one cases (47.1%) were immunoreactive for CDw75 antigen in the primary tumors or metastases. A very close relationship was found between the expression of the antigen in primary tumors and their respective metastases. The expression of the antigen was correlated with an infiltrative growth pattern, lymphatic invasiveness, and aneuploidy. All but two immunoreactive cases had lymph node metastases or lymphatic permeation. No relationship was found between CDw75 antigen expression and the morphologic types of gastric carcinoma, amount of lymphoid infiltrate, vascular invasion, and penetration of the gastric wall.

CONCLUSIONS

The authors conclude that CDw75 antigen expression can be used as a marker of malignant transformation of gastric epithelium and is a good indicator of the biologic aggressiveness of gastric carcinoma.

Involvement of histo-blood-group antigens in the susceptibility of colon carcinoma cells to natural killer-mediated cytotoxicity

The susceptibility to natural-killer-cell lysis and expression of histo-blood-group antigens of 2 clones from a rat colon adenocarcinoma, of variants derived from them and of 17 human colon carcinoma cell lines were assessed in an attempt to determine if the major glycosidic tissue antigens of epithelial cells could influence the NK susceptibility of tumor target cells of epithelial origin. The rat REGb clone, which is relatively NK-sensitive, expressed higher levels of precursor structures T and Tn and lower levels of H antigenic determinants than the PROb clone, which displays higher resistance to NK-cell lysis. Cell variants were obtained from these 2 clones; it appeared that whether the cell variants were selected on the basis of expression of a blood-group antigenic determinant or on the basis of altered susceptibility to NK-cell lysis, there was a link between increased resistance and higher expression of cell-surface A and H histo-blood-group antigens, or conversely, between increased sensitivity and higher expression of precursor structures. Similar conclusions were obtained upon study of the human cell lines, since a significant correlation was found between the level of expression of T or Tn antigens and sensitivity to NK-cell lysis. A significant relationship was found between the expression of Lewis antigens and increased resistance to NK-cell-mediated cytotoxicity.

Studies on the susceptibility to NK-mediated lysis and the simultaneous expression of various surface molecules in anthracyclin-treated K562 cells and in four K562 cell clones

Target molecules for NK cells are unknown. Numerous studies have proposed putative target molecules, but have examined their role in the modulation of sensitivity to NK-mediated lysis one independently of each other. We examined the simultaneous expression of various surface molecules and the susceptibility of K562 cells to NK attack. We have previously shown that adriamycin (40 nM) and aclacinomycin (15 nM) can induce, in vitro, an increase of glycophorin A (GPA) on K562 cells, a modulation of transferrin receptor (TfR) and CD15 antigen expression and a significant resistance of cells to NK-mediated lysis. In the present work, Fc gamma receptor II (CD32) expression at the K562 cell membrane was clearly decreased after aclacinomycin-treatment but was unaltered by adriamycin-treatment. Four K562 cell clones were studied. Two clones (F and G) expressed a higher level of CD32 at the membrane (62% and 70% of erythrocyte antibody (EA) rosettes respectively) and two clones (9 and 19) expressed lower a level (18% and 7% EA rosettes respectively) than the original population (43%). The sensitivity to lysis by NK cells was increased in clones F, G and 9 but decreased in clone 19 (without alteration in the binding capacity). Relationships between the sensitivity to NK attack and the levels of simultaneous expression of CD32, TfR, CD15, glycophorin A (GPA) and MHC class I monomorphic antigens were studied. In addition, the presence at the membrane of some cellular adhesion molecules (CD54, CD58, CD29, CD18, CD56) was examined in anthracyclin-treated cells and in the four clones. The difference in the sensitivity of target cells to NK attack is not strictly related to variation of one or other of these molecules. Our previous and present data suggest that the resistance of K562 cells to NK cells may correlate with the level of erythroid maturation at the cell membrane, involving simultaneous variations in expression of several molecules such as a decrease of TfR, CD15 and CD32 and an increase of GPA.

Enhanced CMP-NeuAc:Gal beta 1,4GlcNAc-R alpha 2,6 sialyltransferase activity of human colon cancer xenografts in athymic nude mice and of xenograft-derived cell lines

The activity of an alpha 2,6 sialyltransferase acting on N-acetyllactosaminic sequences (alpha 2,6 ST E.C. 2.4.99.1) has previously been found to be increased in 90% of human colon cancer specimens. In the present study, the alpha 2,6 ST activity of 6 human colon cancer cell lines grown in culture was compared with that expressed by the corresponding nude mice xenografts and by the cell lines derived from the xenografts. We found that xenografts of COLO 205, HT-29, SW 620, SW 948 and SW 948 FL (a non-adherent sub-line of SW 948) cells express an alpha 2,6 ST activity much higher than that of the in vitro-grown cells. SW 48 cells grown either in culture or as xenografts lack the enzyme activity. All the xenograft-derived cell lines except HT-29 retained the increased alpha 2,6 ST activity at least for the first 6 passages. Those derived from SW 948 xenografts showed an enrichment of round, non-adherent cells, strongly reactive with the NeuAc alpha 2,6 Gal/GalNAc-specific lectin from Sambucus nigra (SNA), thus indicating that a selection of these cells has occurred.

Structural studies of gangliosides from the YAC-1 mouse lymphoma cell line by immunological detection and fast atom bombardment mass spectrometry

YAC-1 cells were propagated in bioreactors in 1 l and 7.5 l volumes. The cells were metabolically labelled with D-[1-14C]galactose and D-[1-14C]glucosamine. The ganglioside fraction, purified by DEAE-Sepharose and silica gel column chromatography, showed on thin layer chromatography four major bands with mobilities between GM1 and GD1a. Gangliosides, obtained by further purification steps including high performance liquid chromatography on silica gel 60 columns with a gradient system of isopropanol:hexane:water, and preparative high performance TLC were characterized by (1) immunostaining of corresponding asialogangliosides obtained by mild acid hydrolysis and neuraminidase treatment and (2) fast atom bombardment mass spectrometry of native and permethylated samples and methylation analysis of GM1b ganglioside. As well as small amounts of GM2 and GM1, the major gangliosides found in the complex mixture were GM1b and GalNAc-GM1b. The structural heterogeneity of these gangliosides was caused by (a) substitution of the ceramide moiety by fatty acids of different chain length and degree of unsaturation (C16:0, C24:0, C24:1) and (b) N-substitution of the sialic acid moieties with either acetyl or glycolyl groups. Disialogangliosides were detected only in low amounts and will be the subject of further investigation. A polyclonal chicken antiserum was raised against IVNeuAc-GgOse5Cer. The antiserum was highly specific for gangliosides (IVNeuAc and IVNeuGc) and asialogangliosides with a GgOse5Cer backbone. No cross-reaction with GM1b or GgOse4Cer was observed.

Glycosphingolipid expression on murine L1-fibrosarcoma cells: analysis of clonal in vivo and in vitro selected sublines with different lung colonisation potential

The patterns of acidic and neutral glycosphingolipids (GSLs) were examined in a syngeneic tumour system in Balb/c mice consisting of closely related cell lines with different colonisation potentials directed to the murine lungs (in vivo selected highly metastatic sublines of L1-fibrosarcoma cells and their WGA-resistant mutants with low metastatic potential). GSLs were analysed by high-performance thin-layer chromatography and structurally identified by fast atom bombardment mass spectrometry combined with compositional analyses and exo-glycosidase digestion. The results suggest that highly metastatic sublines L1-LM and L1-LM12 derived by in vivo selection from mouse fibrosarcoma cells (cell line L1) exhibit a drastic increase of polar ganglioside expression and a restriction to globo-series GSLs. Contrasting with this the low metastatic mutant cells (L1-LM13WGA) express a reduced portion of acidic GSLs and exhibit a shift to less polar ganglioside components. Total cellular and plasma membrane-integrated GSLs were demonstrated to exhibit largely identical patterns. Concomitant with a significant decrease in LacCer expression a substantial reduction of GM2 and a complete lack of GM3 expression can be assigned to the highly metastatic sublines of L1-cells. On the other hand, the more polar gangliosides GM1a and, to an even greater extent, GD1a (exceeding 70% of total gangliosides) accumulate on L1-LM and their clonal sublines. The shift to acidic GSLs of higher polarity is less pronounced on the low metastatic WGA-resistant mutant cells (L1-LM13WGA) showing a preponderance of GM1a. The portion of GD1a within the fractions of acidic GSLs does not correspond to the cellular activities of CMP-NeuAc/GM1 (alpha 2-3) sialyltransferase measured for high and low metastatic cell variants. Total sialic acid content of the various cell lines differs, but is not associated with the metastatic potential. Gangliosides on L1-cells exhibit a significant substitution of N-glycolyl for N-acetylneuraminic acid (13%) compared to their metastatic sublines and to mutant cells (less than 1%). A conversion of surface exposed GD1a to GM1a on membranes of metastatic cells by in situ treatment with Vibrio cholerae sialidase is associated with a significant reduction of tumour cell colonisation directed to the murine lungs.

Galactosylated glycan expression and macrophage sensitivity of Lewis lung tumor cells with different metastatic phenotype

Biochemical and cytochemical analysis of Lewis lung tumor variants revealed that the low metastatic cells contained more galactose/N-acetylgalactosamine residues in a high-molecular-mass (15-20 kDa) mixed N- and O-glycan fraction than the highly metastatic ones. It was also found that the highly metastatic variant was less sensitive to macrophage cytotoxicity in vitro. The cytotoxicity against the low metastatic target cells was inhibited by asialofetuin (10-20 microM), and, to a small degree--and at much higher concentration--by lactose, while galactose and other monosaccharides were ineffective. We suppose that complex galactosylated tumor cell membrane glycans could play a role in the antitumoral cytotoxicity of macrophages.

The restrictive role of sialic acid in antigen presentation to a subset of human peripheral CD4+ T lymphocytes that requires antigen-presenting dendritic cells

Dendritic cells from human epidermis, i.e., Langerhans cells (LC), are more potent antigen-presenting cells (APC) than APC from peripheral blood in proliferative in vitro responses of helper T lymphocytes to various soluble antigens. Analysis of antigen recognition by CD4+ T lymphocyte clones indicated that this increased potency of LC as APC results from a preferential requirement for LC of part of the T cell population. These T cells show a long-lasting restoration of antigen responsiveness to peripheral blood APC after antigen-specific restimulation in vitro, indicating that restrictive antigen recognition concerns T cells that are not fully differentiated. A similar restrictive antigen recognition was observed by treatment of the T cells or the APC with neuraminidase. This restoration of responsiveness is associated with the occurrence of nonspecific cell clustering between T cells and APC. These results suggest that the selective requirement for APC is regulated by the function of adhesion molecules that are functionally blocked by sialic acid groups on immature peripheral T cells but that are readily available on peripheral T cells at a later stage of differentiation.

Suppression of human natural killer cell activity by amino sugars

In the present study we investigated the effect of amino sugars on human natural killer (NK) activity against K562, a human myeloid leukemia cell line, and Molt-4, a human T lymphoma cell line. The presence of amino sugars such as D-mannosamine, D-galactosamine, and D-glucosamine [6-25 mM (in the case of D-mannosamine, 1.5-12.5 mM)] in a 4-hr chromium-51 (Cr) release assay significantly inhibited NK activity of large granular lymphocytes (LGL) without affecting effector cell viability or spontaneous release from target cells. Sugars with acetylated amino residues (N-acetyl-D-mannosamine, N-acetyl-D-galactosamine, and N-acetyl-D-glucosamine) showed much smaller NK inhibition. Among the amino sugars tested, D-mannosamine was the strongest suppressor. When either LGL or K562 cells were pretreated with amino sugars and used in the 4-hr 51Cr release assay, only the pretreatment of effector cells resulted in the reduction of NK activity. The binding capacity of LGL to K562 cells, determined by a conjugate assay, was not reduced by the amino sugars enough to explain the strong inhibition of NK activity by these amino sugars, although some inhibitory effect on the binding of LGL to K562 cells was observed in some cases. In contrast, the polarization of the effector cell cytoskeleton, one of the energy-dependent steps, was significantly impaired. The cellular ATP level of LGL was also significantly reduced and the reduction of cellular ATP correlated well with the degree of the inhibition of NK cytotoxicity. These results suggest that the suppression of NK activity by amino sugars is due to the reduction of the ATP-based energy supply of the effector cells and that amino sugars, especially D-mannosamine, should be recognized as potent suppressors of natural cell-mediated immunity.

Elevated NK-mediated lysis of Raji and Daudi cells carrying fixed iC3b fragments

Raji and Daudi cells were opsonized with C3b, iC3b, and C3d fragments by using purified complement components. The sensitivity of C3-opsonized cells to lysis mediated by low density blood lymphocytes was studied. Raji and Daudi cells carrying C3b or C3d fragments were lysed with similar efficiencies as the nonopsonized cells. The presence of iC3b on the target surface imposed elevated NK sensitivity. The iC3b-mediated enhancement of NK lysis was inhibited when iC3b fragments or rabbit anti-human C3 antibodies were included into the lytic assays. These results indicate that the iC3b fragments fixed on the targets bind to the CR3 on the lymphocytes. Results obtained in immobilized conjugate-lytic assays showed that iC3b-opsonized targets interact more readily with the lymphocytes. This was reflected by the elevated proportion of lymphocytes that were bound to the iC3b-carrying targets. The proportions of conjugates in which target damage occurred were similar with the control and with the iC3b-carrying cells. It seems therefore that opsonization of targets with iC3b leads to recruitment of effector lymphocytes due to contact with their CR3. However, once the effector-target contact is established, the triggering of lytic function does not seem to be influenced by the iC3b/CR3 bridge.

Prostaglandin biosynthetic capacity of hepatomas with different growth rates

1. Prostaglandin synthesis from [14C]arachidonate by microsomal fractions was measured with preparations from rat liver and from hepatomas of different growth rates. The highest rates of synthesis were observed with microsomal preparations from the rapidly growing hepatoma HTC. 2. Assay of endogenous levels of prostaglandins E2, F2 alpha and thromboxane B2 also indicated high levels in solid tumors of the HTC line. 3. With HTC cells in culture it was necessary to incubate in the absence of serum in order to detect prostaglandin synthesis. 4. The data indicated that, while prostaglandin synthesis was elevated in HTC cells, the synthesis of prostaglandins by a series of hepatomas was not closely correlated with the growth rates of the tumors.

The relationship between MHC antigen expression and metastasis

From the studies summarized here a complex picture of the role played by MHC products in determining tumorigenicity and metastasis is emerging. In order to be able to understand this relationship better, it is necessary to consider several factors. 1. Each tumor system or neoplastic tissue is unique, and its behavior reflects the influence of cell-specific characteristics, as well as its ability to modulate other cells and tissues--including cells belonging to the immune system--and also to be modulated by other cells and soluble factors. 2. Since metastasis formation is a multistep process in which only small subpopulations of tumor cells with complex and defined phenotypes are able to colonize secondary tissues, elimination of even one single phenotypic component of this structured process can easily reverse the metastatic capacity of the cells. Acquisition of metastatic ability, on the other hand, would be a more difficult task, since any new characteristic expressed by the cells or induced experimentally, such as gene transfection or results of IFN treatment, must be expressed in a temporal manner and in concert with other cellular characteristics. Therefore, an experimental protocol measuring a specific element in determining metastasis can easily produce conflicting results, depending on the type of cells and genetic background of the host studied. 3. The level of specific MHC products on tumor cells is one among many other cell characteristics that may determine the metastatic potential of cells. Moreover, each of the class 1 MHC products, and the relationship among them, including other than the classical K, L, or D products (Brickell et al., 1983), should be regarded as independent entities, with possible different regulatory roles in cell-cell recognition, in a general sense, which may be involved in determining invasiveness and homing as well as recognition by the immune system. 4. Both specific T-cell and nonspecific natural mediated immunity (which is much less understood) are involved in the selection of the metastatic cell population. 5. Immunogenicity of tumors is not necessarily determined by high levels of MHC antigen expression; it is also dependent on the level of TSA. Thus, immunoselection mediated by T lymphocytes during metastasis formation could be directed against both MHC and TSA antigens. Therefore, low expression of MHC antigens by metastatic cells as a result of immunoselection is not always observed.(ABSTRACT TRUNCATED AT 400 WORDS)

Biology of natural killer cells

Studies of cytotoxicity by human lymphocytes revealed not only that both allogeneic and syngeneic tumor cells were lysed in a non-MHC-restricted fashion, but also that lymphocytes from normal donors were often cytotoxic. Lymphocytes from any healthy donor, as well as peripheral blood and spleen lymphocytes from several experimental animals, in the absence of known or deliberate sensitization, were found to be spontaneously cytotoxic in vitro for some normal fresh cells, most cultured cell lines, immature hematopoietic cells, and tumor cells. This type of nonadaptive, non-MHC-restricted cellmediated cytotoxicity was defined as “natural” cytotoxicity, and the effector cells mediating natural cytotoxicity were functionally defined as natural killer (NK) cells. The existence of NK cells has prompted a reinterpretation of both the studies of specific cytotoxicity against spontaneous human tumors and the theory of immune surveillance, at least in its most restrictive interpretation. Unlike cytotoxic T cells, NK cells cannot be demonstrated to have clonally distributed specificity, restriction for MHC products at the target cell surface, or immunological memory. NK cells cannot yet be formally assigned to a single lineage based on the definitive identification of a stem cell, a distinct anatomical location of maturation, or unique genotypic rearrangements.

Increased susceptibility of periodate-treated tumour cells to human but not to mouse or rat natural killer cells

Treatment of intact cells in the cold with low concentrations (1 mM) of sodium meta periodate (PI) selectively oxidizes the surface-exposed sialic acid residues to the corresponding aldehydes. Such treated tumour cells show greatly enhanced sensitivity to lysis by fresh human NK cells but not to mouse or rat NK cells. Reduction of the PI-treated cells with sodium borohydride (NaBH4) reduced their NK sensitivity to that of untreated cells. In target conjugate formation assays PI-treated tumour cells displayed a higher binding capacity than control cells or PI+NaBH4-treated cells to both mouse and human effector cells. Neuraminidase treatment of K562 and Molt-4 increased target susceptibility to human NK cells but not to mouse, whereas the susceptibility of Yac-1 cells was left unchanged using both human and mouse effector cells. The same pattern of reactivity is shown in the target binding assay. These findings indicate that subtle molecular changes in the surface-exposed carbohydrates of target cells might have a fundamental impact on their sensitivity to lysis by NK cells from certain species, and that in cross species effector-target combinations a higher binding capacity is not sufficient for increased lysis to occur.

Natural killer cells discriminate between high mannose- and complex-type asparagine-linked oligosaccharides

Chinese hamster ovary cell lines with different types of N-linked oligosaccharides were tested as targets for control and lymphokine treated natural killer (NK) cells. The targets tested were parent cells, Lec1 mutants and Lec4 mutants. Due to an apparent defect in GlcNAc transferase V, Lec4 cells produce complex-type N-linked oligosaccharides devoid of GlcNAc beta(1-6) linked branches. Lec1 cells form only high mannose-type N-linked oligosaccharides because they lack GlcNAc transferase I activity. Lec1 cells are very sensitive to lysis by beta-interferon treated human NK cells, but both parent and Lec4 cells are resistant to NK lysis. The ability to discriminate between parent and Lec1 targets was demonstrated with untreated control effectors as well as those which were pretreated with either beta-interferon, gamma-interferon or interleukin-2. Both control and lymphokine-boosted NK cells exhibit much greater lytic activity against targets having only high mannose-type N-linked oligosaccharides. Five oligosaccharide structures resembling those found on N-linked glycoproteins were tested for their ability to block NK lysis of Lec1 targets. Only the high mannose-type glycopeptide from 7S soybean glycoprotein was inhibitory in the mu molar range. At the same concentration, none of the complex-type oligosaccharides had any effect on lytic activity. The results suggest that a high mannose-type N-linked oligosaccharides is recognized at some step in NK cell-mediated lysis.

Association of adriamycin-induced resistance to NK-mediated lysis with sialic acid level and immunological reactivity of transferrin receptors and glycophorin A

Adriamycin (ADM) can increase sialic acid content in K 562 cells and reduce their susceptibility to NK-mediated lysis. In this report, hypothetical relationship between this resistance and augmentation in sialylation has been investigated. Variations in the time of exposure to ADM showed that 12 hours were sufficient to cause maximal recruitment of benzidine-positive cells, growth inhibition and resistance to NK-mediated lysis. On the contrary, the membrane sialic acid density seemed stable and 24 hours of drug exposure were necessary to observe a clear rise in sialic acid. Neuraminidase treatment of control and ADM-treated K 562 cells was associated with an obvious enhancement in their susceptibility to NK-mediated lysis which can be explained by an increase in the target-effector binding ability as assessed by a direct conjugate-forming cell assay. However, the neuraminidase treatment did not modify the sensitivity difference to lysis between untreated and ADM-treated cells. As compared to control the reactivity of ADM-treated cells was higher with an antiglycophorin A (GPA) MAb and lower with an antitransferrin receptor (TFR) MAb. Kinetic studies suggested that GPA expression is a better index of ADM-induced resistance to NK-mediated lysis than TFR expression. In addition, neuraminidase treatment showed that TFR and GPA modulations induced by ADM can be correlated with sialylation alterations.

Unmasking of cryptic natural killer (NK) cell recognition sites on chronic lymphocytic leukemia lymphocytes

The sensitivity of chronic lymphocytic leukemia (CLL) lymphocytes to attack by natural killer (NK) cells has remained questionable. To clarify this issue, freshly isolated lymphocytes of 37 patients with B-CLL, five with WDLL and two with HCL, were tested with a standard cytotoxicity assay with NK cells from normal donors. All these targets were resistant to cytolysis by the effectors. Freeze-fracture analysis of CLL cell plasma membranes revealed that they have a larger number of intramembranous particles (IMP) associated with the external leaflet (E-face) than have normal lymphocytes. Unlike other neoplastic cells, exposure of CLL lymphocytes to phorbol esters or treatment with neuraminidase did not render them vulnerable to attack by NK cells, nor did 5 days of culture have an effect. Incubation of CLL lymphocytes with anti-Ig-mu (24-72 hr) or with 0.1% pepsin (15 min) resulted in 15% and 27% cytolysis, respectively. B-lymphocytes from the blood of healthy donors were not killed when treated similarly: These data establish that freshly isolated B-CLL lymphocytes are resistant to NK cytolysis but that in contrast to normal B-cells, they possess cryptic NK-recognition structures, which may be uncovered by surface modulation.

Changes in glycoprotein fucosylation in a concanavalin A-resistant variant of a human leukemia cell line (K562)

Previous studies from this and other laboratories have shown that variants of tumor cell lines can be selected for resistance to the lytic action of natural killer (NK) cells. One of these (K562-Clone I), when made resistant to the toxic effects of Concanavalin A (Con A-R1), regained its sensitivity to NK. Comparing the plasma membranes of Clone I and Con A-R1, we observed 1) a very similar electrophoretic pattern of membrane glycoproteins identified by binding to the lectins Con A, WGA, PNA, and SBA; 2) an increase in binding of Ulex europaeus lectin to a group of glycoproteins from Con A-R1 that were sensitive to treatment with fucosidase and N-glycanase and that had a diffuse mobility ranging in apparent molecular weight from 30 to 200 kDa; and 3) a marked decrease in binding of an antibody reactive with the lactoneofucopentaose III antigen (Lewis x). This constellation of changes is an unusual pattern to follow Con A resistance and may point to a pathway of glycosylation that a leukemic cell might use to modify its recognition by the NK mechanism.

Reversible susceptibility to natural cell-mediated cytotoxicity observed in altered BHK cells resistant to HVJ (Sendai virus) infection

Altered baby hamster kidney (BHK-R) cells which were subcultured in the continuous presence of HVJ (hemagglutinating virus of Japan--the Sendai strain of parainfluenza 1 virus) showed a high susceptibility to natural cell-mediated cytotoxicity, although BHK-R cells are not transiently or persistently infected with HVJ but contain the restricted amount of sialic acid. By repeated subcultivation of BHK-R cells in growth medium free of HVJ, the sensitivity to natural killer cytotoxicity decreased to the level of normal BHK cells with a counter increase of cellular sialic acid, and the subsequent treatment of the cells with neuraminidase caused a loss of proper sialic acid residues, once again resulting in a significant enhancement of lysis by natural killer cells. In the BHK-R cell system which exhibits a reversible resistance to the interferon action, the enhancing effect induced by interferon on target cell susceptibility to natural killer activity became more pronounced in accord with the recovery of sensitivity to the antiviral action of interferon.

The effect of tunicamycin on target cell susceptibility to natural killer cell cytotoxicity

Several sets of data indicate the possibility that carbohydrate moieties on the target cell are important structures in natural killer (NK) cell-mediated lysis. Striking changes in the NK susceptibility of targets can be induced in several systems involving in vitro differentiation of tumour cell lines. The effect on target cells of the glycosylation inhibitor tunicamycin, which acts by blocking the dolichol-dependent asparagine-linked glycosylation pathway was investigated. Using several different tumour cell lines we can conclude that: asparagine-linked carbohydrate chains do not contribute directly to NK susceptibility, induced differentiation may or may not be linked with a change in NK susceptibility, and secondary changes caused by tunicamycin treatment may lead to alterations in the gangliosides, a finding that is positively correlated with decreased NK susceptibility.

Specific inhibition of human natural killer cell-mediated cytotoxicity by sialic acid and sialo-oligosaccharides

We have tried to identify carbohydrate structures involved in recognition and/or lysis of K562 target cells by human natural killer (NK) cells. Inhibition studies were performed with mono-, di- and trisaccharides, and with glycopeptides and glycoproteins of known carbohydrate composition. When tested with various monosaccharides, lysis of K562 cells was inhibited only by N-acetylneuraminic acid (NeuAc). Di- and trisaccharides and glycopeptides containing NeuAc or N-glycolylneuraminic acid (NeuGc) all inhibited NK cell-mediated lysis. Among the non-sialylated carbohydrates tested, only Gal beta(1----3)GalNAcol was effective. The inhibitory capacity of sialylated compounds appeared to be dependent on the linkage type of the sialic acid residue; carbohydrates containing alpha(2----6)-linked sialic acids were more potent inhibitors than their alpha(2----3) isomers. Also the sugar to which the sialic acid residue was attached was of importance, NeuAc alpha(2----6)GalNAcol being more effective than NeuAc alpha(2----6)Gal beta 1----R (where R = glucose or oligosaccharide-peptide). Sialylated compounds and free sialic acid had minor or no effects on cell-mediated cytotoxicity by allo-sensitized cytotoxic T lymphocytes. The conjugation of target cells and NK effector cells was not inhibited by carbohydrates that effectively blocked the cytolytic response. These results may indicate that cell-surface carbohydrates containing alpha(2----6)-linked sialic acid are crucial structures in a post-binding event in NK-cell-mediated lysis.

Target cell specificity of human natural killer cells. II. Apparent change with activation

The activity of natural killer (NK) cells can be augmented by incubation with interferons, or with other compounds, such as staphylococcal protein A, which stimulate interferon production. In the experiments described here we compared the patterns of lytic activity of human lymphoid effector cells, before (NK-B) and after (NK-A) short-term activation. Target cells used were K562, Clone I (a partially NK-resistant K562 variant), and those that had been preincubated with neuraminidase or trypsin. The results obtained include the following: proteases, but not neuraminidase, decreased lysis by NK-B, but not by NK-A, of both K562 and Clone I in the standard Cr-release assay. In the single cell assay, trypsin minimally decreased conjugate formation and the fraction of bound cells that were killed by NK-B, but did not reverse the increased lytic efficiency of NK-A. In the cold-target competition assay, Clone I, which does not compete as well with K562 for NK-B, did so equally well for NK-A. Trypsinized targets, regardless of their equal sensitivity to lysis by NK-A, were not as active competitors for NK-A. We conclude that the most reasonable interpretation is that K562 cells bear surface structures which can induce release of lytic mediators from NK-A under conditions that are not sufficient to stimulate NK-B. Although it appears that NK-A may respond to a smaller number of the same target molecules recognized by NK-B, the process must be better defined at the molecular level to exclude the possibility that there are qualitative differences between the proposed recognition structures for these two states of NK activity.

Loss of intercalated membrane particles by treatment with phorbols

Because brief exposure to phorbol esters renders normal cells vulnerable to deformation and cytolysis by lymphocytes, it was postulated that these tumor promoters might cause a hitherto unrecognized physical alteration in membrane architecture. To investigate this possibility, four tissue culture cell lines (K-562 erythroleukemia cells, melanoma cells, N1121 adult fibroblasts, and normal fetal fibroblasts) and three blood cell types (lymphocytes, monocytes, and platelets) were subjected to freeze-fracture analysis before and after brief treatment with phorbol myristate acetate. Phorbol myristate acetate caused a 50% reduction of intramembranous particles associated with the external leaflet (E face) of the plasma membrane of every cell except platelets. In contrast, no change in size or number of intramembranous particles associated with the protoplasmic membrane leaflet (P face) was evident. Since the platelet membrane is known to be turned "inside out," as regards the partition coefficient of the intramembranous particles, the disparity between the results obtained with platelets and other cells may serve to determine the nature of intramembranous particles affected by phorbols. Also, since phorbols affect primarily glycolipids and/or glycoproteins anchored in the external membrane leaflet, these findings may provide a useful tool for future exploration of membrane structure.

Heterogeneity in the surface properties of B16 melanoma cells from sublines with differing metastatic potential detected via two-polymer aqueous-phase partition

When mixed in aqueous solution at low concentrations, the neutral polymers dextran and poly(ethylene glycol) (PEG) rapidly form a two-phase system, consisting of a dextran-enriched lower phase and a PEG-enriched upper phase. Two B16 mouse melanoma cell lines, B16-F1 (low lung colonizing capability) and B16-F10 (high lung colonizing capability) were found to partition differentially into the upper phase in a variety of two-phase systems. Upper-phase partition depends primarily on either hydrophilic (i.e., surface charge density) or hydrophobic (i.e., affinity for the hydrocarbon chain of a PEG-fatty acid ester) cell surface properties, depending on the system used. In single-step partition studies, cells of the B16-F10 subline displayed a greater preference than B16-F1 cells for the upper phase in the hydrophilic system and less preference in systems sensitive to hydrophobic properties. Countercurrent distribution (CCD) experiments, performed with [125I]deoxyuridine DNA-labelled cells, were consistent with single-step partition results. These CCD results demonstrated that B16-F10 cells exhibited greater DNA synthesis than B16-F1 cells and that considerable heterogeneity, in both hydrophobic and hydrophilic surface properties, was present in subpopulations of cells of both sublines. The data also showed considerable enrichment of 125I-specific cell activity in certain sections of the distributions, indicating that differences in cellular DNA synthesis are reflected in the surface properties to which partition is sensitive.

Lactosamine type asialooligosaccharide recognition in NK cytotoxicity

Inhibition of pig NK cell activity by asialooligosaccharides (aOS) isolated from human serum glycoproteins was investigated. Train-tennary aOS (aOSIII) of ceruloplasmin was found to be the most potent inhibitor up to the concentration 0.1 micrograms/ml, which is in agreement with its highly specific binding to NK-activity-enriched pig lymphocytes (with a morphology similar to human large granular lymphocytes (LGL]. Only lectins with the specificity to Gal(beta 1----4)GlcNAc or Gal(beta 1----3)GalNAc structures exhibited inhibition of NK cytotoxicity. F(ab)2 fragments of rabbit antibodies against pig spleen membrane lectin cross-reacting with the pig liver membrane lectin completely inhibited NK activity when preincubated with the effectors or present in the incubation mixture during the assay. These data suggest that lectin receptors on cells of pig NK-activity-enriched fraction specific for aOSIII and antigenically related to membrane lectins isolated from pig spleen and liver, are involved in the NK recognition of several xenogeneic targets.

The organization of cell populations within lymph nodes: their origin, life history and functional relationships

The normal lymph node comprises a superficial cortex, a deep cortex or paracortex and a medulla. In each of these regions there are three kinds of spaces: an intralymphatic space, an intravascular space and an extravascular space or interstitium. Both the vascular endothelium and the lymphatic endothelium are specialized in these different regions. The cell types in lymph nodes comprise lymphoid cells, accessory or non-lymphoid cells and stromal cells, and within these cell types a number of different sub-types can now be identified by means of enzyme- and immunocytochemistry. Based predominantly on experimental studies, the origin, migratory patterns, localization, inter-relationships and interactions between these various cells are reviewed.

Natural killer activity against cultured human neural tumor and fetal brain cells

Ten human neural tumor lines and three established from normal human brain were analyzed for sensitivities to natural killer (NK) cytolysis. Compared to MOLT-4, fetal brain cells were sensitive, but those from adult brain and eight of ten neural tumor cell lines demonstrated marked NK resistance. The frequencies of target-binding cells (TBC) and single-cell lysis of glioma cells bound within tumor cell conjugates demonstrated that the resistance of two lines was explained either by a decrease in the frequencies of TBC or reduced ability of bound NK cells to lyse the tumor cell conjugates. A third resistant line demonstrated decreases in both TBC and tumor cell conjugate lysis. Two glioma lines with less NK resistance had greater frequencies of TBC or conjugate lysis than the resistant lines. Thus, NK resistance can result from decreased recognition of targets, diminished NK lysis of bound targets, or a combination of both.

Carbohydrate structure in tumor immunity

Researchers have endeavored to define surface alterations associated with neoplasia for at least 25 years. In comparisons of normal tissues with animal and human tumors, cultured cells before and after transformation with oncogenic agents, tumorigenic and nontumorigenic transformed cells, metastatic and nonmetastatic tumor cells, high- and low-metastatic variants, and tumor cells before and after induction of differentiation to a less malignant phenotype, a consistent finding has been some form of alteration in surface carbohydrate structures. These changes in glycolipids, glycoproteins and glycosaminoglycans are reviewed, and their structures are illustrated. Both nucleotide sugar biosynthesis and glycosyltransferase changes have been associated with these alterations. In some cases, alterations in transformed cells were related to growth, rather than transformation. In others, the altered glycoconjugates are truly tumor-associated. There is evidence that cell surface glycoconjugates may function in growth control. Altered carbohydrate structures could also serve as receptors for growth promoting factors and be directly responsible for altered growth control. Recent studies with monoclonal antibodies indicate that the vast majority of antibodies recognizing tumor-associated antigens are detecting altered carbohydrate structures. Mechanisms by which the immune system can recognize these carbohydrate structures are considered, and immune recognition of tumor-associated carbohydrate structural alterations is explored. A number of these hypotheses relating to alterations in glycosylation, growth control, and tumor immunity deserve further investigation.

Asparagine-linked oligosaccharides in murine tumor cells: comparison of a WGA-resistant (WGAr) nonmetastatic mutant and a related WGA-sensitive (WGAs) metastatic line

MDW40, a wheat germ agglutinin-resistant (WGAr) mutant of the highly metastatic tumor cell line called MDAY-D2, is restricted to local growth at the subcutaneous site of inoculation. The WGAr tumor cells acquire metastatic ability by fusing spontaneously with a normal host cell followed by chromosome segregation, a process accompanied by reversion of the WGAr phenotype (i.e., WGAs). Since lectin-resistant mutant cell lines often have oligosaccharide alterations that may affect membrane function and consequently metastatic capacity, we compared the major Asn-linked glycopeptides in WGAr and WGAs cell lines. [2-3H]mannose-labeled glycopeptides were separated into four fractions on a DEAE-cellulose column and then further fractionated on a concanavalin A-Sepharose column. Glycopeptide structures were determined by: (a) sequential exoglycosidase digestion followed by chromatography on lectin/agarose and Bio-Gel P-4 columns and (b) proton nuclear magnetic resonance analysis. The metastatic WGAs cells had a sialylated poly-N-acetyllactosamine-containing glycopeptide which was absent in the nonmetastatic mutant cell line. Unique to the mutant was a neutral triantennary class of glycopeptide lacking sialic acid and galactose; the WGAr lesion therefore appeared to be a premature truncation of the antennae of the poly-N-acetyllactosamine-containing glycopeptide found in the WGAs cells. High mannose glycopeptides containing five to nine mannose residues constituted a major class in both WGAr and WGAs cells. Lysates of both wild-type and mutant cells had similar levels of galactosyltransferase activity capable of adding galactose to the N-acetylglucosamine-terminated glycopeptide isolated from mutant cells; the basis of the WGAr lesion remains to be determined.

Persistent infection with mouse hepatitis virus 3 in mouse lymphoid cell lines

The sensitivity of mice to mouse hepatitis virus 3 (MHV3) varies according to strain, age, and immune status of the animals. In semisusceptible strains, mice surviving the acute phase of infection develop a chronic disease characterized by the occurrence of paralysis, virus persistence, and immunodeficiency. Persistent MHV3 infections established in vitro in YAC and RDM -4 mouse lymphoid cell lines were characterized by virus production, presence of cytoplasmic viral antigens, and cell lysis. The occurrence of cell "crisis" in YAC cells was manifested by a sharp increase in cell lysis and in the number of fluorescent cells and, concomitantly, by a marked decrease in virus titers. A relationship was observed among the percentage of fluorescent cells, cell lysis, and virus yield and was modulated by renewal of culture media, change in temperature, or inhibition of cellular RNA synthesis. Cell cloning and antibody treatment experiments indicated that viral transmission was performed by viral infection of newly permissive cells produced by the division of uninfected cells in the culture and not by transmission of viral information by infected dividing cells. The biological and biochemical properties of MHV3 variants derived from persistently infected YAC lymphoid cells were characterized. Thermosensitivity and thermolability of cloned viruses originating from persistently infected YAC cells, as well as parent virus suspensions, were studied. A similar heterogeneity was observed when YAC-derived cloned substrains (YAC-MHV3) were compared with parent-derived cloned viruses, indicating that no selection of temperature-sensitive mutants was induced in persistently infected YAC cells. However, the capacity of MHV3 to induce a lethal acute disease when injected into susceptible mice was lost very rapidly. The absence of pathogenicity was related to the induction of a subclinical infection which elicited defense mechanisms. These data suggest, therefore, that MHV3 replication in lymphoid cell lines leads to induction or selection of variants which maintain pathogenicity in vitro but display reduced pathogenic effects in vivo.

Neoplastic cells as targets of spontaneously cytotoxic lymphocytes: studies with natural killer-like cell lines

Native natural killer (NK) cells comprise a heterogeneous family of lymphocytes distributed among several organs, which display spontaneous cytotoxic reactions directed against a broad range of tumor targets. In these studies, murine cell lines have been established in vitro following the selective expansion of bone marrow- and spleen-derived killer progenitors in culture medium supplemented with interleukin-2. Several clones of independent origin have been characterized in order to determine the extent of their phenotypic and functional diversity. With few exceptions most of them were found to be highly effective in lysing a variety of tumor cell lines, to share common cell surface alloantigens, lectin-binding receptors, and cytochemical markers. The presence of prominent azurophilic cytoplasmic granules is the most characteristic ultrastructural feature of these cells. In attempting to elucidate the nature of membrane components specifically recognized by NK cells we compared several isogenic tumor cell variants selected on the basis of their differential NK susceptibility, immunogenicity, metastatic potential or resistance to cytotoxic plant lectins. Sialylated glycoconjugates exposed on the external face of the tumor cell membrane appear to be essential determinants in the interaction between NK cells and their targets. Permanent cell lines retaining most of the functional attributes of endogenous NK cells may prove instrumental in understanding their role during tumor progression.

Gamma-interferon (IFN-gamma) produced during effector and target interactions renders target cells less susceptible to NK-cell-mediated lysis

Human mitomycin C-treated PBL were mixed with cells of an NK sensitive hybrid cell line (PUTKO-I). A fraction of tumor cells survived this treatment and could be recovered from the cultures. These surviving cells were completely NK-resistant and this property persisted for 2-3 weeks after cultivation in fresh medium. Treatment of a clone (C13) of PUTKO-I with PBL-PUTKO mixed lymphocyte-tumor-cell culture (MLTC) supernatants resulted in a marked reduction in NK sensitivity after 8-12 h of treatment. The kinetics of induction of NK resistance by MLTC supernatants was similar to that of purified IFN-gamma and was faster than for IFN-alpha. The active component in the supernatants was characterized as a mixture of IFN-gamma and IFN-alpha based on neutralization of activity with specific antisera. The role of mycoplasma contamination was investigated and it was found that cell lines free of detectable mycoplasma stimulated production of NK-protective activity by PBL and this activity was neutralized by anti-IFN-gamma serum. Separation of PBL on discontinuous Percoll gradients demonstrated a correlation between the NK activity of cell fractions and their ability to produce IFN in response to tumor cells. Taken together, the selection-dependent variations in NK sensitivity, the kinetics of IFN production and induction of resistance suggest that tumor cells may be able to escape elimination by NK cells due to protection by IFN produced by the effector-cell-containing population.

Modulation of K562 cells with sodium butyrate. Association of impaired NK susceptibility with sialic acid and analysis of other parameters

Neuraminidase treatment of parental and butyrate-induced K562 tumor cells was associated with an increase in natural killer (NK) susceptibility of these target cells. The degree of enhancement with neuraminidase was significantly greater for the NK-resistant (NRR) butyrate-differentiated K562 cells so that the relative difference between the parental NK-sensitive (NKS) K562 line and its induced NKR variants, in terms of NK sensitivity, was no longer five- or six-fold but only two-fold. The predominant reason for the altered NK susceptibilities of the target cells after neuraminidase treatment was an increase in the target-cell-binding ability of these cells as assessed by a direct conjugate-forming cell assay using Percoll-enriched NK cells and cold target competition assays. The enhancement did not appear to be due simply to an increased membrane-membrane attraction caused by a reduction of net negative cell surface charges since protamine sulphate, a positively charged molecule, had no effect on NK activity. Compared with the NKS parental K562 tumor cells, the NKR butyrate-induced cells had 3.6- to 4.0-fold higher sialo-transferase activities and were associated with significantly greater amounts of cell surface sialic acid detected both in sialyl glycoproteins (2.2- to 2.9-fold higher) and particularly within ganglioside extracts (6.2- to 13.6-fold higher). In conformity with the marked neuraminidase enhancement of NK-mediated cytolysis of the butyrate-induced targets, these NKR cells were associated with significantly enhanced levels of neuraminidase-accessible sialic acid compared to the NKS parental K562 cell line. Other parameters such as sensitivity to superoxide radicals, intrinsic superoxide dismutase levels, altered membrane repair mechanisms and transferrin competition, were not significantly different between the NKS and NKR target phenotypes. Sugar inhibition studies demonstrated an enhanced inhibition against the butyrate-induced cells with a variety of neutral sugars. The degree of inhibition with phosphorylated sugars was comparable between the parental and induced K562 tumor target cells and is consistent with our previous findings showing that these hexose phosphates may be inhibiting cytolysis at a step independent of target-cell recognition.

Relationship between expression of herpes simplex virus glycoproteins and susceptibility of target cells to human natural killer activity

Cells normally insensitive to human natural killer (NK) activity were rendered susceptible by infection with HSV-1. The cytotoxic effector cell was a nonadherent, non-T, non-B lymphocyte. Antibody plus complement treatment, using a monoclonal antibody that recognizes an antigen present on NK cells, removed much of the cytotoxic activity, and a density gradient fraction enriched for NK cells yielded cells of increased virus-specific cytotoxicity. It was concluded that the effector cell active against infected targets possessed characteristics of an NK cell. Blockage of viral protein synthesis during infection inhibited development of increased susceptibility of infected targets to NK activity. When targets were infected with a mutant virus unable to produce viral glycoprotein C (gC), NK activity against these targets was reduced approximately 30% compared with activity against targets infected with wild-type virus. Similarly, activity against targets infected in the presence of 2-deoxyglucose (2dG), which prevents cell surface expression of viral glycoprotein B (gB), was also reduced approximately 30%. An approximately 60% reduction in activity was seen against targets infected with mutant virus in the presence of 2dG; these targets express gD, but neither gB nor gC. When cells expressing various combinations of HSV-1 glycoproteins were used as both labeled targets and cold target competitors, it was found that the susceptibility of a particular target to NK activity was paralleled by its ability to act as a cold target competitor. This indicates that targets with decreased sensitivity to NK cells were less able to bind NK effectors. Further, the amount of interferon produced in co-cultures of NK effectors and infected target cells did not directly correlate with the amount of NK activity generated, and interferon pretreatment of effectors did not decrease virus-specific cytotoxicity. The present results suggest that HSV-1 glycoproteins expressed at the surface of infected targets may act as recognition structures for NK cells.