Depressed natural killer cell activity in children with protein--calorie malnutrition. II. Correction of the impaired activity after nutritional recovery

Natural killer (NK) cell activity and responsiveness to interferon (IFN) were measured in the peripheral blood of infants having kwashiorkor or marasmus and of nutritionally recovered malnourished children. Depression of NK activity in the peripheral blood lymphocytes (PBLs) of the malnourished children was noted, while normal levels of activity were observed in the nutritionally recovered infants. Addition of exogeneous interferon in vitro potentiated the NK levels of PBLs from well-nourished and nutritionally recovered infants, but had either a nonsignificant impact on cells from the marasmic infants or a suppressive effect on the cells from infants with kwashiorkor. The success of exogenous interferon to potentiate the NK levels of PBLs from nutritionally recovered infants suggests that nutritional repletion corrects the impaired cellular responsiveness in these patients.     

Some biological properties of the human amniotic membrane interferon.

Human amniotic interferon was investigated to define the species specificity of its antiviral action and to compare its anti-cellular and NK cell stimulating activities with those of other human interferons. The antiviral effect was titrated in bovine (RV-IAL) and monkey (VERO) cells. Amniotic interferon exhibited, in bovine cells, 5% of the activity seen in monkey cells, while alpha interferon displayed 200%. No effect was detected with either beta or gamma interferon in bovine cells. Daudi cells were exposed to different concentrations of various interferons and the cell numbers were determined. The anticellular effect of the amniotic interferon reached its peak on the third day of incubation. Results suggested a higher activity for alpha and gamma interferons and a lower activity for beta when compared to amniotic interferon. Using total mononuclear cells as effector cells and K 562 as target cells in a 51Cr release assay, it was demonstrated that low concentrations of amniotic interferon consistently stimulated NK cell activity in cells derived from several donors, the results indicating a higher level of activity with this interferon than with alpha and beta interferons.     

Elevated natural killer cell-mediated cytotoxicity, plasma interferon, and tumor cell rejection in mice persistently infected with lymphocytic choriomeningitis virus

To assess the effects of chronic virus infection on NK cells, the related phenomena of interferon (IFN) production, NK cell activation, and resistance to tumor implants were studied in mice persistently infected with lymphocytic choriomeningitis virus (LCMV). NK cells from these LCMV-carrier mice displayed augmented killing of the NK-sensitive YAC-1 target cell. They did not lyse the more resistant targets L-929 and P815, whereas NK cells from acutely infected mice efficiently lysed all three cell types. The plasma from LCMV-carrier mice contained an antiviral substance identified as IFN type I, based on species specificity, virus nonspecificity, resistance to pH 2, and sensitivity to antibody to type I IFN. IFN titers in plasma from LCMV-carrier mice were 32 to 64 U/ml, about 20-fold less than those in acutely infected mice. Both the IFN and NK cell levels continuously remained elevated in the LCMV carrier mice up to at least 6 months of age. IFN is known to activate NK cells and to induce their blastogenesis in vivo. As determined by centrifugal elutriation, large NK blast-size cells were isolated from the spleens of acutely infected mice, but not from either normal or LCMV-carrier mice, suggesting augmented NK cell-mediated lysis in the absence of enhanced proliferation. Poly inosinic-cytidylic acid induced high levels of NK cell-mediated cytotoxicity and blastogenesis in both control and LCMV-carrier mice, but IFN was induced to lower levels in carriers as compared with controls. Coincidental with augmented NK cell activity, the LCMV-carrier mice rejected intravenously injected 125IUdR-labeled tumor cells more efficiently than did normal mice. Thus, LCMV carrier mice have low levels of type I IFN, moderately augmented NK cell activity lasting for at least 6 months, and increased resistance to tumor cell implants. This indicates that augmented NK cell-mediated cytotoxicity can be maintained in vivo over prolonged periods of time in the presence of chronic low-level IFN stimulation.     

Stimulation of Aryl Hydrocarbon Hydroxylase Induction in Cell Cultures by Interferon

In cell cultures previously treated with homologous interferon, the magnitude of antiviral activity and the degree of stimulation of aryl hydrocarbon hydroxylase induction appear to be directly related. In a highly purified mouse interferon preparation, the factor stimulating hydroxylase induction and the factor directing antiviral activity are inactivated by heating to 70 C or by treating with trypsin. Also, both activities demonstrate species specificity.     

Effects of diethyldithiocarbamate, an inhibitor of interferon antiviral activity, upon human natural killer cells

In support of a postulated role of the Cu++-dependent enzyme, superoxide dismutase (SOD), in antiviral effects of interferon (IFN), a close correspondence was previously shown to exist between inactivation of cellular SOD and concomitant blockade of IFN antiviral activity in fibroblasts by the Cu++-chelating agent, diethyldithiocarbamate (DDC). To further define the extent of "anti-IFN" activity, we initiated studies of DDC effects on IFN stimulation in the NK cell system. Unexpectedly, DDC directly inhibited cytotoxicity mediated by unstimulated NK cells. Pronounced inactivation occurred rapidly (less than 30 min), but was spontaneously reversible in the absence of DDC. Neither cell viability nor lymphocyte binding to target cells was detectably affected. Preincubation of DDC with Cu++ or Zn++ failed to neutralize its inhibitory effects nor could function be restored in DDC-pretreated NK cells by subsequent addition of Cu++, Zn++, Mg++, or Ca++. DDC treatment that inactivated NK cells did not detectably alter lymphocyte SOD activity. Thus, inhibition was probably not attributable to chelating properties of DDC. N-ethyl maleimide (NEM) and para-( hydroxymercuri ) benzoic acid ( PMBA ), enzyme inhibitors that preferentially react with sulfhydryl groups, both inactivated NK cells in a time- and dose-dependent manner similar to that of DDC. Preincubation with the sulfhydryl compound, cysteine, neutralized in parallel fashion the capacity of NEM, PMBA , and DDC to inhibit NK cell activity. Thus, a previously unreported reactivity of DDC with sulfhydryl groups appeared to be the basis of inhibition. NK cells incubated 1 hr with IFN and subsequently cultured 17 to 23 hr without IFN were activated to an extent comparable to cells continuously incubated 18 to 24 hr with IFN. Exposure to IFN for 1 hr was therefore sufficient to commit NK cells to acquisition of a fully activated state. Whether preactivated by a 1-hr or 18- to 24-hr IFN treatment, activated NK cells retained the DDC-sensitive phenotype characteristic of fresh unstimulated NK cells. Thus, prolonged IFN treatment did not render NK cells resistant to DDC or preferentially activate a DDC-sensitive NK cell subset. An 18- to 24-hr incubation of DDC-pretreated cells in the continual presence of IFN resulted in the boosting of NK cell activity. However, the 1-hr IFN pulse treatment protocol was consistently ineffective in boosting when IFN was added just after DDC-pretreatment. These results strongly suggested that DDC temporarily rendered NK cells unresponsive to what, under normal circumstances, approximated an optimally potentiating IFN stimulus.(ABSTRACT TRUNCATED AT 400 WORDS)     

Restoration of TLR3-activated myeloid dendritic cell activity leads to improved natural killer cell function in chronic hepatitis B virus infection

There is increasing evidence that the function of NK cells in patients with chronic hepatitis B (CHB) infection is impaired. The underlying mechanism for the impaired NK cell function is still unknown. Since myeloid dendritic cells (mDC) are potent inducers of NK cells, we investigated the functional interaction of mDC and NK cells in CHB and the influence of antiviral therapy. Blood BDCA1(+) mDC and NK cells were isolated from 16 healthy controls or 39 CHB patients at baseline and during 6 months of antiviral therapy. After activation of mDC with poly(I · C) and gamma interferon (IFN-γ), mDC were cocultured with NK cells. Phenotype and function were analyzed in detail by flow cytometry and enzyme-linked immunosorbent assay. Our findings demonstrate that on poly(I · C)/IFN-γ-stimulated mDC from CHB patients, the expression of costimulatory molecules was enhanced, while cytokine production was reduced. In cocultures of poly(I · C)/IFN-γ-stimulated mDC and NK cells obtained from CHB patients, reduced mDC-induced NK cell activation (i.e., CD69 expression) and IFN-γ production compared to those in healthy individuals was observed. Antiviral therapy normalized mDC activity, since decreased expression of CD80 and CD86 on DC and of HLA-E on NK cells was observed, while poly(I · C)/IFN-γ-induced cytokine production by mDC was enhanced. In parallel, successful antiviral therapy resulted in improved mDC-induced NK cell activation and IFN-γ production. These data demonstrate that CHB patients display a diminished functional interaction between poly(I · C)/IFN-γ activated mDC and NK cells due to impaired mDC function, which can be partially restored by antiviral therapy. Enhancing this reciprocal interaction could reinforce the innate and thus the adaptive T cell response, and this may be an important step in achieving effective antiviral immunity.     

The role of the thymus in the maintenance of natural killer cells in vivo

This report describes a model for investigating the role of the thymus in regulating natural killer (NK) cell activity in vivo. Evidence is presented that the thymus can regulate NK cells, and that at least some NK cells can develop without thymic help. Marrow from thymectomized rats depleted of circulating T cells by thoracic duct cannulation was transplanted into rats without a thymus (1 degree ATX.BM). These 1 degree ATX.BM rats had NK cell levels above controls 3 months after reconstitution but markedly depressed NK cell levels by 9 months. When 1 degree ATX.BM marrow was used to reconstitute rats with or without a thymus, those without a thymus (2 degrees ATX.BM) exhibited low NK cell levels after 3 months, and a similar result was obtained when 2 degrees ATX.BM marrow was used to reconstitute 3 degrees ATX.BM rats. The low NK cell levels in 2 degrees and 3 degrees ATX.BM rats were due to a deficiency in spontaneously cytotoxic NK cells, as they had normal numbers of interferon-responsive pre-NK cells. Spleen cells from 2 degrees and 3 degrees ATX.BM rats produced less interferon than control spleen cells when cultured with P815 tumor cells in vitro. However, 2 degrees and 3 degrees ATX.BM rats had higher numbers of large granular lymphocytes than controls despite their low NK cell levels. In marked contrast to 2 degrees and 3 degrees ATX.BM rats, spleen cells from 4 degrees ATX.BM rats had higher levels of cytotoxicity and a higher frequency of both spontaneously cytotoxic and pre-NK cells than controls. The 4 degrees ATX.BM rats also had the highest frequency of large granular lymphocytes in the spleen.     

Antiviral effect of lymphokine-activated killer cells: characterization of effector cells mediating prophylaxis

Lymphokine-activated killer (LAK) cells generated by cultivation of C57BL/6 mouse spleen cells in the presence of recombinant interleukin-2 were transferred into natural killer (NK) cell-deficient suckling mouse recipients. These mice were then challenged with either murine cytomegalovirus (MCMV) or lymphocytic choriomeningitis (LCMV) and sacrificed 3 days later. No interleukin 2 infusions were given. Mice receiving as few as 5 x 10(5) LAK cells had several 100-fold decreases in spleen MCMV titers as compared with untreated mice. This treatment had no effect on spleen LCMV titers. The LAK cell cultures contained 10 to 17% NK 1.1+, 50 to 55% Lyt-2+, and 33 to 50% immunoglobulin D+ cells. Double fluorescence labeling and in vitro cytotoxicity assays with fluorescence-activated cell sorting revealed at least two mutually exclusive killer cell populations. NK 1.1+ LAK cells resembled freshly isolated activated NK cells with regard to target cell range (YAC-1 cell killing greater than L-929, P815, and EL-4 cell killing), large granular lymphocyte (LGL) morphology, and decreased ability to lyse interferon (IFN)-treated target cells. Lyt-2+ LAK cells lysed the targets mentioned above but at lower levels and without the differences in susceptibility mentioned above. These Lyt-2+ LAK cells also had a decreased ability to lyse IFN-treated targets, in contrast to classic cytotoxic T lymphocytes, which lyse IFN-treated targets far more efficiently than untreated targets. Purified populations of LAK cells obtained by fluorescence-activated cell sorting were used in the antiviral protection model. The results showed that protection against MCMV could be mediated by NK 1.1+, NK 1.1-, Lyt-2+, Lyt-2-, and IgD- populations but not by IgD+ cells. The five protective populations all had in common the LGL phenotype and cytotoxic activity in vitro. The IgD+ population did not contain LGLs, lyse target cells in vitro, or mediate an antiviral effect in vivo. These results suggest that LAK cells may be therapeutically useful against certain virus infections (MCMV) but not others (LCMV) and that despite their heterogeneity in antigenic phenotype and cytotoxic activity, their pattern of antiviral activity in vivo resembles that of NK cells, which protect against MCMV but not LCMV.     

Augmentation of mouse natural killer cell activity by LS 2616, a new immunomodulator

The quinoline-3-carboxamide LS 2616 administered to mice in drinking water increased spontaneous cytotoxicity against YAC-1 cells in a dose-dependent manner. The enhancement of spontaneous cytotoxicity was found to be mediated by NK cells, as judged by their lack of adherence to nylon wool columns, relative resistance to treatment with antibodies to Thy-1.2 and complement, and almost total abrogation after depletion of asialo-GM1+ cells. Enhancement of NK activity was evident after 2 days of treatment, was maximal after 4 days, and remained elevated during the 14-day exposure period studied. NK activity returned to control levels 4 days after cessation of treatment. NK activity was significantly increased in spleen, peripheral blood, lymph nodes, and bone marrow of LS 2616-treated mice, while activity in peritoneal exudate cells and thymus remained low. LS 2616 was able to elevate NK activity in several mouse strains studied, including mice homozygous for the beige gene. Serum interferon levels were not increased during treatment with LS 2616. Combined injection of the interferon inducer Poly I:C and LS 2616 did not increase NK activity above that of animals injected with Poly I:C alone. However, Poly I:C, in contrast to LS 2616, increased NK activity in peritoneal exudate cells. Studies at the single cell level revealed that LS 2616 increased NK activity by increasing the number of lytically active cells via recruitment of new target-binding cells and not by increasing the lytic activity of pre-existing binders.     

The role of natural killer cells and interferon in resistance to acute infection of mice with herpes simplex virus type 1

The relative roles of interferon (IFN) and natural killer (NK) cells in herpes simplex virus type 1 (HSV-1) infection of mice were examined. Adoptive transfer of adult mouse leukocytes into 4- to 6-day-old suckling mice protected the recipients from HSV-1 infection, as judged by viral titers in the spleen 2 days postinfection. Protection was mediated by several classes of leukocytes, including those depleted of NK cell activity by antibody to asialo GM1 and those depleted of macrophages by size separation. Mice receiving these leukocytes produced significantly higher levels of IFN 6 hr postinfection (early IFN) than did HSV-1-infected mice not receiving donor leukocytes. Antibody to IFN, under conditions that blocked early but not late IFN synthesis, greatly enhanced HSV-1 synthesis in mice receiving leukocytes and completely removed the protective effect mediated by leukocytes. High doses of anti-asialo GM1 blocked both NK cell activity and early IFN production and resulted in high titers of HSV-1. This effect on virus synthesis was not seen if mice were given antibody 1 day postinfection. Lower doses of anti-asialo GM1, which still depleted NK cell activity but had no effect on early IFN production, did not enhance HSV-1 synthesis. Depletion of NK cell activity with a low dose of antibody had no effect on the reduced HSV-1 synthesis resulting from prophylactic IFN treatment or on the enhanced HSV-1 synthesis resulting from antibody to IFN treatment. Thus, resistance to acute HSV-1 infection in mice correlates with early IFN production but not with NK cell activity, suggesting that NK cells are not major mediators of natural resistance in this model and that the antiviral effect of IFN is not mediated by NK cells.     

Immunological responses of aging Japanese A-bomb survivors

Immune response parameters were studied on 1341 A-bomb survivors residing in Hiroshima, Japan. Mononuclear cells were isolated from venous blood and tested for interleukin-2 production; lymphocytes were purified and tested for natural killer (NK) cell activity and interferon (IFN) production; and serum was tested for IFN and circulating immune complex (CIC) levels. Statistical analyses were performed for each type of assay using a linear models procedure including sex, age at the time of the bomb, radiation exposure, all the interaction variables, and the categorical variable day-of-assay in the model. The findings showed that (1) none of the immunologic variables were significantly affected by radiation exposure; (2) NK activity and CIC levels were positively associated with age; and (3) NK activity was on average higher for males than females. The data exemplify the difficulty in reaching firm conclusions concerning associations with radiation exposure when the dependent variable exhibits a large degree of interindividual and day-of-assay variability.     

The role of cell membrane in the antiviral effect of interferon

The mechanism of interferon action in human fibroblasts has been studied by use of both antisera to human fibroblast interferon and the antisera to the surface of human fibroblast cell. The anti-interferon serum completely neutralized the antiviral effect of human fibroblast interferon. Interferon antiserum prevented the intracellular antiviral state from developing when added to the medium of the cells in which interferon synthesis had already been induced by poly (I · C). This suggests that development of the antiviral state involves interferon interaction with the external part of the producing cell. Treatment with the serum directed against the surface of human fibroblast cells failed to inhibit the antiviral activity of human interferon in these cells. In addition, the effect of gangliosides on the antiviral activity of human interferon was studied and it was found that human interferon binds to gangliosides and that this interaction leads to inactivation of the antiviral effect of interferon. Pretreatment of human fibroblasts with gangliosides had no effect on the sensitivity of these cells to exogenous interferon.     

Lack of spontaneous and inducible natural killer cell activity in human bone marrow: presence of adherent suppressor cells

Human bone marrow cells collected from ribs of patients undergoing thoracotomy had low or no natural killer (NK) cell activity against K562 in a 4-hour chromium release assay. In vitro overnight treatment with interferon or interleukin 2 of bone marrow cells resulted in no induction or augmentation of NK cell activity. In the presence of adherent bone marrow cells interferon was unable to enhance NK cell activity of blood lymphocytes, although the baseline level of NK cell activity was not suppressed. These results suggest that adherent bone marrow cells regulate the development of active NK cells and that bone marrow components do not provide a favorable environment for the functional differentiation of NK cells.     

An interferon-induced increase in cyclic AMP levels precedes the establishment of the antiviral state.

Mouse interferon induces an increase in cyclic AMP levels in interferon-sensitive mouse Ly cells but not in interferon-insensitive human KB-3 cells. The interferon-induced elevation in cyclic AMP precedes the induction of antiviral activity. Although interferon stimulation of the adenylate cyclase activity of Ly cell plasma membranes has not been detected, interferon is effective in stimulating this activity in plasma membranes from rat thyroid cells.     

Relative Antiviral Resistance Induced in Homologous and Heterologous Cells by Cross-Reacting Interferons

Human cells incubated with human interferon become more resistant to vesicular stomatitis virus (VSV) than to Semliki Forest virus (SFV); monkey cells treated with monkey interferon become more resistant to SFV than to VSV. However, monkey cells incubated with human interferon developed relative antiviral activity identical to that induced by homologous interferon, and human cells developed characteristic human interferon-induced relative antiviral activity when exposed to monkey interferon. Therefore, cross-reacting interferons induce the relative antiviral activity characteristic of the interferon-treated cell rather than the cell of the interferon's origin. This relationship supports the hypothesis that interferon is not itself antiviral but rather induces cells to develop their own antiviral activity.     

Defective differentiation of natural killer cells in SJL mice. Role of the thymus

As previously shown, three distinct phenotypes exist in murine natural killer (NK) cell activity when it is evaluated by the endogenous levels of activity and the susceptibility to augmentation by interferon (IFN) and IFN inducers. The "low" phenotype has low levels of activity which can be poorly augmented by IFN, as in mice of SJL strain. The "inducible" phenotype exhibits low endogenous levels but can vigorously respond to IFN-mediated augmentation, as in A.SW strain. The "high" NK phenotype shows high levels of endogenous activity which can be augmented to still higher levels by IFN, as in B10.S mice. Since SJL mice with congenital absence of the thymus (nude) were of the inducible type, the effect of neonatal thymectomy was examined in the present study. Neonatal thymectomy was found to convert the low phenotype of SJL mice to the inducible, mimicking the effect of nu/nu genotype. Thymectomy as late as 25 days after birth was effective, but retransplantation of a syngeneic newborn or adult thymus, or thymocytes, failed to reverse the effect of thymectomy. The poor responsiveness of NK activity to IFN in SJL, therefore, is extrinsic to the NK cell lineage and is attributable to suppression or maturational block of NK cell differentiation by the thymus during the first few weeks of neonatal life. A series of experiments with bone marrow chimeras showed that the SJL recipients did not allow the expression of inducible or high phenotype by bone marrow progenitors from allogeneic donors with either phenotype. Therefore, the SJL recipients provide an environment which suppresses not only the development of IFN-sensitive NK cell precursors, but also the levels of endogenous NK cell activity. SJL bone marrow cells gave rise to NK activity of inducible phenotype in B10.S recipients, confirming the crucial role of the environment in which NK cell differentiation takes place.     

Natural killer cell cytotoxicity: role of calmodulin

The phenothiazine derivatives, fluphenazine and trifluoperazine which are known to bind to calmodulin and to inhibit its activity, abrogate the development of both spontaneous and interferon-enhanced cytotoxicity of mouse splenic lymphocytes enriched for NK cell activity. Phenothiazines also inhibit the rapid increase in cyclic GMP levels in interferon-treated splenic lymphocytes. Furthermore, treatment of mouse splenic lymphocytes with electrophoretically pure interferon, alpha/beta caused a marked decrease in the level of calmodulin within 1 to 4 hours. These results provide evidence that calmodulin may play a role in the development of NK cell cytotoxicity and that the effect of interferon on calmodulin may constitute part of the molecular mechanism of interferon action.     

Murine natural killer cells limit coxsackievirus B3 replication

Previous indirect evidence suggested that natural killer (NK) cells play a role in coxsackie virus B3 serotype 3, myocarditic variant (CVB3m)-induced myocarditis by limiting virus replication. In this study, we present direct evidence that NK cells can limit CVB3m replication both in vitro and in vivo. Virus titers are lowered in primary murine neonatal skin fibroblast (MNSF) cultures incubated with activated splenic large granular lymphocytes (LGL) taken from mice 3 days postinoculation of CVB3m, a time of maximal NK cell activity. The antiviral effect of this cell population is diminished by complement-mediated lysis with the use of anti-asialo GM1 antiserum but not with anti-Lyt-2 monoclonal antibody. Neither interferon nor anti-CVB3m-neutralizing antibody was detected in these cultures. Although activated LGL initiate lysis within CVB3m-infected MNSF in vitro within 3 hr of addition, they do not lyse uninfected MNSF cultures. CVB3m replication is required for expression of surface changes on MNSF that result in lysis by NK cells because cell cultures treated with compounds that prevent CVB3m replication are not killed by LGL. LGL also do not lyse MNSF cultures inoculated with UV-inactivated virus. Mice inoculated with activated LGL and subsequently challenged with CVB3m had reduced titers of virus in heart tissues in comparison to titers of CVB3m in heart tissues of mice not given LGL. The antiviral activity of the LGL preparation was abolished by prior treatment with anti-asialo GM1 antiserum plus complement but not by prior treatment with anti-Lyt-2 monoclonal antibody and complement. These data suggest that NK cells can specifically limit a nonenveloped virus infection by killing virus-infected cells.     

Induction of natural killer cell activity of thoracic duct lymphocytes by polyinosinic-polycytidylic acid (poly(I:C)) or interferon

Natural killer (NK) cell activity of thoracic duct lymphocytes (TDL) was examined in normal mice and in mice treated with polyinosinic-polycytidylic acid (poly(I:C) and interferon (IFN). TDL from mice treated with phosphate-buffered saline (PBS) expressed little or no NK cell activity against YAC-1 target cells at effector-to-target ratios of up to 200:1, even after in vitro treatment with murine L-cell IFN. In contrast, TDL from poly(I:C)- or IFN-treated mice expressed significant NK activity, which correlated with the significantly higher NK activity of splenocytes from these mice compared to the NK activity of splenocytes from PBS-treated mice. These data indicate that although TDL from normal mice express no detectable NK cell activity, NK cell activity can be induced in TDL by in vivo treatment with poly(I:C) or IFN.     

Tumor necrosis factor and lymphotoxin secretion by human natural killer cells leads to antiviral cytotoxicity

NK cells mediate their cytotoxicity against tumor cells through abroad array of cytotoxic and cytostatic proteins. We investigated whether specific proteins could also be identified that contributed to NK cell-mediated antiviral immunity. Human CD16+/CD3- NK cells were obtained by using FACS and subsequently cloned by using limiting dilution. These NK cell lines, which were cytotoxic against NK-sensitive tumor targets and virally infected cells, also generated supernatants that selectively killed vesicular stomatitis virus-infected cells while sparing noninfected cells. This soluble antiviral activity was completely neutralized by antibodies specific for TNF and lymphotoxin. Purified human rTNF also duplicated this specific cytotoxicity against vesicular stomatitis virus-infected cells, as well as against CMV-, Theiler's murine encephalomyelitis virus-, and HSV-infected cells. The degree of cytotoxicity varied for the different viruses and depended on the cell type infected. These results suggest that NK cells can mediate selective and direct cytotoxicity against virally infected cells by the secretion of TNF and lymphotoxin.