Activation of natural killer cells by the mAb YTA-1 that recognizes leukocyte function-associated antigen-1

The mAb YTA-1, which brightly stains CD3^–CD16⁺ large granularlymphocytes (LGL)/natural killer (NK) cells and CD8⁺ T cellsby immunofluorescence, is specific for leukocyte function-associatedantigen (LFA)-1. Some mAbs recognizing the LFA-1 chain (CD11a)or LFA-1ß chain (CD18) inhibited the binding of YTA-1to peripheral blood mononuclear cells. YTA-1 mAb could be chemicallycross-linked to 170 and 96 kDa molecules, whose molecular weightscorrespond to those of LFA-1 and ß respectively.YTA-1bound to COS-7 cells co-transfected with CD11a and CD18 cDNAs,but not to untransfected cells. Reactivities of YTA-1 to K562cells transfected with LFA-1 and ß(CD11a/CD18) cDNAsand to CHO cells transfected with Mac-1 (CD11b/CD18) or p150,95 (CD11c/CD18) cDNAs strongly suggest that YTA-1 recognizeseither LFA-1 or an epitope formed by a combination of LFA-1and ß. Treatment of fresh CD3^–CD16⁺ LGL withYTA-1 augmented cytolytic activity and induced proliferation.F(ab')₂ fragments of YTA-1 augmented NK cytotoxicity, indicatingthat the NK activating signal was transmitted through LFA-1without involvement of Fc receptor III. In contrast, the othermAbs against LFA-1 could not activate NK cells. These resultscollectively indicate that YTA-1 recognizes a unique epitopeof LFA-1, which is involved in activation of fresh NK cells.     

Dysregulated expression of the IL-2 receptor {beta}-chain abrogates development of NK cells and Thy-1+ dendritic epidermal cells in transgenic mice

The IL-2 receptor ß-chain (IL-2Rß), a specificity-determiningsubunlt In the IL-2R complex with a restricted tissue distributionpattern, Is essential for signal transductlon. Our previousstudies demonstrate that the continuous treatment of mice withanti-IL-2Rß) resulted in the complete disappearanceof NK cells and Thy-1⁺ dendritic epidermal cells (Thy-1⁺ dEC),suggesting that signals through IL-2Rß are criticallyinvolved in development of these lymphocyte subsets. However,these lymphocyte subsets are reported to be apparently unaffectedIn the IL-2-deficient mice. To further examine the biologicalroles of the IL-2Rß, transgenic mice carrying theIL-2Rß transgene were generated. In these mice, highlevels of the cell surface expression of the IL-2Rßwere observed in essentially all hematopoietic lineage cells,and CD4⁺ T cells as well as CD8⁺ T cells showed vigorous cellproliferation upon IL-2 stimulation. Surprisingly, NK cellsmarked with a high expression of NK1.1 in the spleen and Thy-1⁺dEC in the skin were completely absent in transgenic mice. However,the development of other lymphocyte subsets Including conventionalßTCR ⁺ cells, TCR⁺ cells and B cells remained apparentlyintact. From these observations together with previous dataon IL-2-deficlent mice, we speculate that factors, other thanIL-2 that utilizes the IL-2Rß as its functional receptorsubunlt, may have a vital role in the development of NK cellsand Thy-1⁺ dEC. Implications for possible In vivo functionsof over-expressed IL-2Rß are discussed.     

Cytokine production by CD16 CD56bright natural killer cells in the human early pregnancy decidua

Using a cell sorter, CD16^–CD56^bright natural killer (NK)cells were sorted from decidual mononuclear cells at an earlystage of pregnancy. These cells were examined by the reversetranscrlptase-polymerase chain reaction (RT-PCR) method fortheir expression of mRNA coding for the following 12 cytokines:IL-1ß, IL-2, IL-3, IL-4, IL-5, IL-6, granulocyte colony-stimulatingfactor (G-CSF), granulocyte-macrophage colony-stimulating factor(GM-CSF), macrophage colonystimulating factor (M-CSF), tumornecrosis factor- (TNF-), interferond- (IFN-), and leukemia inhibitoryfactor (LIF). Although mRNA coding for every cytokine was detectedin decidual mononuclear cells, mRNAs coding for only G-CSF,GM-CSF, M-CSF, TNF-, IFN-, and LIF were detected In CD16^–CD56^brightNK cells. Also, the supernatant of CD16^–CD56^bright NKcell cultures was found to contain G-CSF, GM-CSF, M-CSF, TNF-,IFN-, and LIF. These findings indicate that CD16^–CD56^brightNK cells produce many different cytokines and that these cytokinesmay play an important role in a successful pregnancy.     

Combination chemo-immunotherapy: kinetics of in vivo and in vitro generation of natural killer cells and lymphokine-activated killer cells in the rat.

Rats received a single high dose of cyclophosphamide (Cy) (150 mg/kg), followed 48 h later (on day 0) by immunization with a T cell-dependent soluble antigen, ovalbumin in Freund's complete adjuvant (FCA). The effect of this treatment on lymphoid cell subpopulations in the spleen, natural killer (NK) cell and interleukin-2 (IL-2) induced lymphokine-activated killer (LAK) cell activity was examined. Cy (with and without ovalbumin) caused a large relative increase (by day 14) in splenic OX8+, OX19- cells with NK morphology. A marked relative increase in fresh NK cell activity was noted after Cy + ovalbumin, but not consistently after Cy alone. Elevated NK activity was Cy dose- and time-dependent, was evident within 7 days post Cy/ovalbumin and persisted for at least 28 days. Pooled splenic mononuclear cells (MNC), obtained 14 days after Cy/ovalbumin, lost all cytolytic activity against YAC-1 cells when cultured in the absence of human recombinant IL-2 (rIL-2). In contrast, similarly maintained cells from normal rats displayed NK activity higher than normal 'fresh' levels. Upon culture in medium containing 500 U/ml rIL-2, however, 'augmented' NK activity was equivalent, on a per-cell basis, in both normal and Cy/ovalbumin-pretreated groups. LAK activity generated in vitro (i.e. against NK-resistant target cells) was significantly lower in the latter group, and the overall yield of cells was reduced. By day 21 after Cy/ovalbumin, augmented NK activity was significantly greater than controls, on a per-cell and total culture yield basis. Moreover, LAK activity was now similar between groups. It is concluded that the chemotherapy/immunization protocol which we have used can greatly enhance NK activity in vivo and that these cells are responsive to induction of LAK activity by IL-2 in vitro.     

In vivo boosting of lung natural killer and lymphokine-activated killer cell activity by interleukin-2: comparison of systemic, intrapleural and inhalation routes.

Natural killer (NK) cells are thought to play a role in host defence against malignancy and infection, in immunoregulation and as precursor cells in a generation of lymphokine-activated killer (LAK) cells which can lyse NK-resistant tumour cells. As the lung is a major site for malignancy and infection and as there are large numbers of lymphoid cells including NK cells in the interstitial compartment of the lung, we evaluated the capacity of interleukin-2 (IL-2), a lymphokine capable of augmenting NK activity in vitro, to augment lung NK cell activity in vivo, using different routes of IL-2 administration. We compared both systemic (i.v. and i.p.) and local (intrapleural and inhalation) routes of IL-2 administration (50,000 U/daily for 5 days) using CBA mice, assessing NK and LAK cell activity in the spleen (systemic) and in the lung. The target cells used for these studies were the YAC-1 (NK-sensitive) and P815, NO36 and HA56 (NK-resistant, LAK-sensitive) cell lines. Splenic NK activity was increased by 1.4-1.9-fold for i.v./i.p., respectively, compared with controls with both systemic routes of administration, and lung NK activity was increased 3.2-fold and 3.8-fold (i.v./i.p, respectively, P less than 0.05), to levels which were comparable to systemic (splenic) NK activity following the same therapy. Intrapleural IL-2 administration similarly enhanced lung NK activity (3.3-fold) and splenic NK activity (1.3-fold; P less than 0.05 versus controls for both). Surprisingly, inhaled IL-2 suppressed both splenic and lung NK cell activity (84 +/- 8% and 78 +/- 10% suppression, respectively, P less than 0.05). LAK cell activity was also enhanced in the lung by 1.8-8-fold in response to i.v., i.p. and intrapleural IL-2, whereas inhaled IL-2 was ineffective in generating LAK cell activity. These results suggest that the systemic and intrapleural administration of IL-2 effectively boost pulmonary NK and LAK activity whereas inhalation of IL-2 does not. Thus, in clinical situations where boosting of local lung NK or LAK cell activity is desired, these routes of IL-2 administration may be effective.     

P-glycoprotein-mediated multidrug resistance and lymphokine-activated killer cell susceptibility in ovarian carcinoma

The sensitivity of tumor cells to lysis by natural killer (NK) and interleukin-2 (IL-2)-activated killer (LAK) cells was studied in three ovarian carcinoma cell lines (2780.9S, SKOV-3, and CHOAUXB1), four multidrug-resistant (MDR) variants, and a melphalan-resistant line. The antitumor activity of LAK cells was evaluated both by⁵¹Cr release and by conjugate formation assays. Four of four P-glycoprotein-positive (P-gp⁺) MDR ovarian carcinoma cell line variants were lysed by human LAK cells to a greater extent than were their drug-sensitive counterparts. In contrast, a melphalan-resistant ovarian carcinoma cell line that does not overexpress P-gp (P-gp^–) did not exhibit an increased susceptibility to LAK cells relative to its parental cell line. Two of the four P-gp⁺ MDR ovarian carcinoma cell line variants were tested for human NK cell susceptibility and this was found to be unchanged or decreased. The P-gp⁺ MDR ovarian carcinoma cell line 2780.AD645 showed a higher frequency of tumor cell binding to LAK cells than did the drug-sensitive parental line. A monoclonal antibody (mAb) against a cell surface epitope of P-gp, MRK16, used at 1 g/ml, enhanced the LAK susceptibility of P-gp⁺ MDR ovarian carcinoma cell lines. However, when incubation with 10 g/ml MRK-16 antibody (Ab) was followed by 12.5 g/ml F(ab)₂ goat anti-mouse (GAM) immunoglobulin (Ig), the increased LAK susceptibility of P-gp⁺ MDR cell lines was inhibited. These data strongly suggest that P-glycoprotein-positive MDR ovarian carcinoma cells not only are targets for LAK cells, but are more sensitive than their drug-sensitive parental lines. This is in contrast to their susceptibility to NK cells, which is low to start with and remains unchanged or even decreased in MDR cells. It is postulated here that P-gp or associated changes result in a greater frequency of effector-target cell binding, leading to increased LAK cell cytotoxicity.     

IL-4 producing CD4+ TCR{alpha} {beta}int liver lymphocytes: influence of thymus, {beta}2-microglobulin and NK1.1 expression

The present report describes developmental, phenotypic and functionalfeatures of unconventional CD4⁺ TCRß lymphocytes.In C57BL/6 mice, the majority of liver lymphocytes expressingintermediate intensity of TCRß (TCRß^int)are CD4⁺NK1.1⁺ and express a highly restricted TCR V_ßrepertoire, dominated by V_ß8 with some contributionby V^ß7 and V^ß2. Although these cells expressthe CD4 co-receptor, they are present in H2-l Aß (Aß)^+/–gene disruption mutants but are markedly reduced in ß₂-microglobulin(ß₂m)^–/– mutant mice and hence are ß₂mdependent. Thymocytes expressing the CD4⁺NK1.1⁺ TCR ßphenotype are also (ß₂m) contingent, suggesting thatthese two T lymphocyte populations are related. The CD4⁺NK1.1⁺TCRßlymphocytes in liver and thymus share several markers such asLFA-1⁺, CD44⁺, CD5⁺, LECAM-1^– and IL-2Rßa. TheCD4⁺NK1.1 ⁺ TCRß^int liver lymphocytes were not detectedin athymic nuinu mice. We conclude that ß₂m expressionis crucial for development of the CD4⁺NK1.1⁺ TCRß^intliver lymphocytes and that thymus plays a major role. CD4⁺ TCRß^intliver lymphocytes were also identified in NK1.1⁺ mouse strains,there lacking the NK1.1 marker. We assume that the NK1.1 moleculeis a characteristic marker of the CD4⁺TCR"^int liver lymphocytesin NK1.1⁺ mouse strains, although its expression is not obligatoryfor their development. The liver lymphocytes from +₂m^+/–,but not from +₂m^–/–mice are potent IL-4 producersin response to CD3 or TCRß engagement and the IL-4production by liver lymphocytes was markedly reduced by treatmentwith anti-NK1.1 mAb. We conclude that the CD4+NK1.1⁺ TCRß^intliver lymphocytes are capable of producing IL-4 in responseto TCR stimulation.     

Natural killer cell activity in lymphocytic choriomeningitis virus-infected {beta}2-microglobulin-deficient mice

We have investigated the induction and role of natural killer(NK) activity in lymphocytic choriomeningitis virus (LCMV)-infectedß₂-microglobulin-deficient (ß₂m^–)mice. We demonstrate that LCMV infection is more effective thanpolyinosinic:poiycytidylic acid (poly I:C) at stimulating NKactivity in ß₂m^– .In addition, ß₂m^–NK cells respond poorly to in vitro treatment with IL-12. Thetarget specificity of the virally induced NK cells is similarto that previously reported for chemically induced ß₂m^–NK cells. In both cases they can lyse YAC-1 tumor cells butare unable to kill ß₂m^– orß₂m⁺ T cellblasts. We have also found that the time course of inductionof NK and cytotoxic T lymphocyte (CTL) activity by LCMV in ß₂m^–mice is delayed compared with normal mice. Maximal NK and CTLactivity is attained at day 8 and 10 post-infection respectivelyin ß₂m^– compared with day 4 and 6—8 inB6 mice. Whereas normal mice die {small tilde}7 days followingintracranial infection with LCMV, the course of disease in ß₂^–mmice is protracted and characterized by a marked loss of bodyweight. We show that although the CD4⁺ CTL response in thesemice is intimately involved in mediating weight loss, the virus-inducedNK cells do not appear to play a role in the disease.     

慢性乙肝患者杀伤性免疫细胞功能的研究

通过对４４例病毒性肝炎患者Ｔ细胞亚群，ＮＫ细胞活性与ＬＡＫ细胞活性的观察，探讨了在慢性乙型肝炎病毒复制与非复制状态下的杀伤性细胞活性。结果表明：在乙肝病毒的高复制状态下，ＣＤ８＾＋细胞数增加，ＣＤ４＾＋／ＣＤ８＾＋比例显著下降；ＮＫ细胞活性与ＬＡＫ细胞活性也明显低下，且在ＨＢｅＡｇ与ＨＢＶＤＮＡ阳性组中，ＮＫ活性与ＬＡＫ活性的改变与ＨＢｅＡｇ的Ｐ／Ｎ值变化呈显著负相关，而ＮＫ活性与ＬＡＫ活性变化则     

Impairment of natural killer (NK) cytotoxic activity in hepatitis C virus (HCV) infection

In the present study, we evaluated the NK cell cytotoxic activity in a group of HCV-infected individuals. Although the number of NK cells present in the peripheral blood of the HCV-infected patients was comparable to non-infected individuals, spontaneous NK cytotoxicity was four-fold lower (P< 0.001) than in normal donors. This functional impairment was not overcome by depletion of adherent or B cells, and it was partially restored by short-term (18 h) stimulation with IL-2. However, long-term stimulation (72 h) with this lymphokine induced activated killer cell (LAK) activity comparable to normal controls. The reduction in NK cytotoxic response does not seem to be due to soluble suppressive factors, since incubation of normal peripheral blood mononuclear cells (PBMC) with infectious HCV serums for a 4-h period does not affect NK spontaneous cytotoxic activity. Successful in vitro infection of PBMC with HCV infectious serum also resulted in an impairment of NK cytotoxicity, suggesting that altered NK function is associated with HCV infection and may be responsible, at least in part, for the chronicity of the infection.     

Interleukin-2 induced killer cell activity against Epstein-Barr virus-immortalized human B cells

Interleukin-2 (IL-2) activated killer (LAK) cells, generated in vitro by treating peripheral blood lymphocytes (PBL) with human IL-2, are able to lyse a wide variety of target cells without restriction by major histocompatibility complex (MHC) molecules. Earlier observations from this and other laboratories indicated that patients with Epstein-Barr virus (EBV) induced infectious mononucleosis, a self-limiting viral disease, have high EBV-nonspecific natural killer (NK) cell activity. Since the effect of LAK cells on EBV-immortalized B lymphocytes has not yet been studied, we decided to investigate LAK cell activity against autologous and heterologous B lymphocytes immortralized in vitro by EBV and other EBV genome-positive and -negative targets of malignant origin. LAK activity was determined by ⁵¹Chromium release assay. The results obtained show that LAK activity was not specific for EBV and was not MHC-restricted. Results of experiments using NK cell reactive monoclonal antibodies suggest that the cytotoxicity is due predominantly to activated NK cells. Our observations suggest that LAK cells may be very effective for immunotherapy in patients with chronic or progressive EBV infections and EBV-induced lymphoproliferative diseases.     

Interleukin-2 modulates the expression of lymphocyte function-associated antigen-one (LFA-1) and p150,95 during the generation of lymphokine-activated killer (LAK) cells

Lymphocyte function-associated antigen-one (LFA-1), Mac1 and p150,95 represent a family of heterodimeric cell surface molecules with a common beta subunit and distinct alpha subunits. LFA-1 is known to be functionally important in cell-cell interactions between immune cells. In the present study, a mouse monoclonal antibody (mAb), RH1-38, which recognizes an epitope on the beta-chain of LFA-1 was used to study the function and expression of LFA-1 on lymphokine-activated killer (LAK) cells. This mAb has been shown previously to block, in the absence of complement, cytolytic activity mediated by natural killer (NK) cells, cytotoxic T lymphocytes (CTL), and a monocyte-like cell (phorbol diester-stimulated HL-60 cells). LAK cells were generated by culturing in vitro human peripheral blood lymphocytes (PBL) in the presence of human recombinant interleukin-2 (rIL-2), and cytotoxic activity was measured by a 51Cr-release assay using the human NK-resistant Daudi cell line. Addition of RH1-38 ascites supernatant, purified RH1-38 mAb, or F(ab')2 fragment of RH1-38 markedly reduced (>80) LAK cytolytic activity, whereas NS-1 (parent hybridoma) ascites supernatant, normal mouse IgG, and monoclonal anti-HLA had no effect on LAK-mediated killing. Equivalent inhibition of NK and CTL activity by purified RH1-38 required 10-100-fold more antibody. Appreciable inhibition occurred if the mAb was added up to 2 hr after LAK cells were mixed with targets. Indirect immunofluorescence flow cytometry and immunoprecipitation studies revealed that LFA-1 and p150,95 expression were dramatically enhanced in PBL populations cultured with rIL-2 compared with PBL cultured without rIL-2; Daudi cells expressed no detectable LFA-1 family heterodimers. Time-course experiments demonstrated that during culture of PBL in the presence of rIL-2, development of enhanced expression of LFA-1 and p150,95 correlated closely with LAK cytolytic activity. These studies (i) demonstrate that LFA-1 and/or p150,95 are functionally important effector cell surface molecules expressed by LAK cells and that some homology to NK and CTL mechanisms of cell-mediated lysis may exist; and (ii) suggest that enhanced LFA-1 and/or p150,95 expression are important for development of the fully differentiated LAK effector cell in the presence of rIL-2.     

CC chemokines induce the generation of killer cells from CD56+ cells

We describe here that members of the CC chemokines exhibit biological activities other than chemotaxis. Macrophage inflammatory protein (MIP)-1α, MIP-1β, monocyte chemoattractant protein-1 and RANTES, but not interleukin (IL)-8, induce the generation of cytolytic cells, designated here as CHAK (CC chemokine-activated killer) cells to distinguish them from IL-2-activated (LAK) cells. Like IL-2, CC chemokines can induce the proliferation and activation of killer cells. While incubating CC chemokines with CD4⁺ or CD8⁺ cells did not generate CHAK activity, all CC chemokines were capable of inducing CHAK activity upon incubating with CD56⁺ cells, suggesting that the primary effectors are NK cells. However, the presence of other cell types, such as CD4⁺ or CD8⁺, are necessary to induce the proliferation of CD56⁺ cells. Confirming the involvement of T cell-derived factors in inducing the proliferation of these cells, anti-IL-2 and anti-interferon-γ, but not anti-IL-1β, anti-tumor necrosis factor-α, anti-IL-8, or anti-granulocyte/monocyte-colony-stimulating factor inhibited RANTES-induced proliferation of nylon wool column-nonadherent cells. Our results may have important clinical applications for the utilization of CHAK cells in the treatment of cancer and immunodeficient patients.     

rIL-4 differentially regulates rIL-2-induced murine NK and LAK killing in CD8+ and CD8- precursor cell subsets

Interleukin 4 (IL-4) and IL-2 have complementary or synergistic roles in many aspects of lymphocyte development. IL-2 supports the induction of cytolytic activity in cytotoxic T lymphocyte (CTL), natural killer (NK), and lymphokine-activated killer (LAK) cells. IL-4 has also been shown to support CTL and LAK in primary murine spleen cell culture. This report demonstrates that IL-4 selectively down-regulates IL-2 inducible murine CD8- precursors of NK cells. For maximal regulatory effect it is necessary to add IL-4 to cultures before 40 h. Enrichment for NK1.1+ cells failed to recover precursor cells which are down-regulated in overnight cultures or can be cultivated in vitro to yield NK cytolytic activity. Furthermore, phenotypic analysis of effector cells demonstrated a marked inhibition of development of NK1.1+ cells in cultures containing IL-4 plus IL-2 versus IL-2 alone. Thus, it appears that IL-4 down-regulates the precursors of murine NK cells by inhibiting proliferation and/or development. In addition, we show that IL-2-induced murine LAK activity mediated by CD8- precursor cells is unaffected by IL-4, while CD8(+)-derived LAK cells are up-regulated by co-culture with IL-4 and IL-2. Analysis of these data relative to reports documenting down-regulation of human LAK by IL-4 suggests that in vitro cultured, IL-2-activated murine NK cells are the correlates to what are commonly described as human LAK cells. The discrepancy may stem from differences in the characteristics of target cells used in the murine versus the human systems. These results clarify the conflicting reports on the effect of IL-4 on killing activity.     

Proliferation of alloreactive human natural killer cells independent of specific allogeneic stimulation

This study establishes that natural killer (NK) cells cytolytlcfor B-lymphoblastoid cell lines (B-LCL) expressing NK-deflnedalloantigerts can be stimulated to proliferate in culture independentlyof allogeneic stimulation. NK cells proliferate following co-culturewith a -lrradlated malignant melanoma cell line (MM-170) andIL-2-condltioned medium. The cultured NK cells from some donorsshowed a high level of cytotoxicity against NK-1⁺ B-LCL andthis corresponded with a high precursor frequency (46–64%)determined from limiting dilution analysis. Alloreactive NKcells proliferated in cultures containing autologous activatedT cells, demonstrating that alloantlgens were not essentialto stimulate proliferation. B-LCL expressing NK-1 or NK-2 orneither of these alloantigens stimulated proliferation of NKcells cytolytlc for NK-1⁺ B-LCL. Studies using metabollcallyinactivated B-LCL confirmed that stimulation was not alloantlgendependent. The results demonstrate that recognition of alloantlgensby NK cells, sufficient to trigger their lytlc program, is notrequired and indeed is not sufficient to confer a stimulatorysignal for proliferation of alloreactive NK cells.     

Dissociation of natural killer and lymphocyte-activated killer cell lytic activities in human CD3− large granular lymphocytes

CD3^? large granular lymphocytes (LGL) are known to display natural killer cell (NK) activity without prior sensitization or restriction by major histocompatibility antigens. Upon short-term exposure to interleukin-2, NK cells were shown to acquire lymphocyte-activated killer cell (LAK) activity. The aim of this study was to analyze the characteristics of these lytic activities. Our data indicated that both NK and LAK activities were Ca²⁺ dependent; however, they could be dissociated by a Ca²⁺ channel blocker or a Ca²⁺ channel competitor agent. Moreover, NK activity was associated with granule exocytosis of lytic proteins spontaneously present in CD3^? LGL, the most likely candidate being the pore-forming protein perforin. By contrast, LAK activity was found to be dependent on de novo protein synthesis and distinct from granule exocytosis. Our results strongly suggest that NK and LAK activities could be defined as two distinct pathways involving different lytic mediators.     

Differential augmentation of in vivo natural killer cytotoxicity in normal primates with recombinant human interleukin-1 and granulocyte-macrophage colony-stimulating factor.

The effect of recombinant human interleukin-1 (IL-1) alpha, granulocyte-macrophage colony-stimulating factor (GM-CSF), and combined factor therapy (CFT) on Rhesus monkey peripheral blood natural killer (NK) activity in vivo was compared. During a 14-day treatment period, IL-1-treated animals demonstrated a 170% increase in NK activity against K562 target cells by day 4, reaching maximal levels (300%) at day 16, and returning to baseline by day 30. NK activity of GM-CSF-treated monkeys increased slightly (60-100%) during days 4-12, as did saline-treated monkeys, but returned to baseline values by day 16. A delayed increase in NK activity resulted after GM-CSF treatment, reaching a peak (260%) on day 23 and remaining elevated through day 39. CFT resulted in a bimodal response pattern, with two peaks of NK activity: one at day 16 and a second at day 39. The first peak of activity (223%) was significantly less than the activity attained with IL-1 alone; the second peak (300%) was of greater duration and occurred later than the peak observed in GM-CSF-treated monkeys. Unlike IL-1, GM-CSF treatment did not lead to a immediate stimulation of NK activity; augmentation was delayed by more than 7 days post treatment. CFT results suggest that GM-CSF reduced the direct NK response to IL-1; while IL-1 led to an enhanced delayed NK response. Therefore, IL-1 and GM-CSF augment NK activity through different but interrelated pathways.     

Pentoxifylline suppresses interleukin-2-mediated activation of immature human natural killer cells by inhibiting endogenous tumor necrosis factor-alpha secretion

We recently reported that immature human peripheral blood-derived natural killer (NK) cells, the free NK subset, can be activated by interleukin-2 (IL-2) to become killer cells and to undergo proliferation. Activation by IL-2 is dependent on endogenous secretion of tumor necrosis factor-alpha (TNF-) by the free cells. Because pentoxifylline (PTX) inhibits TNF- synthesis and secretion in monocytes, we hypothesized that PTX may also inhibit TNF- secretion by NK cells and thus would inhibit IL-2-mediated activation of free cells. The free NK cells were separated from purified NK cells by flow cytometry and cell sorting of non-target binding cells. IL-2-mediated secretion of TNF- by the free cells was inhibited by PTX. In the presence of PTX, IL-2-mediated activation of free cells into cytotoxic function, proliferation, and recruitment of binder and killer cells was markedly inhibited. Also, PTX inhibited IL-2-triggered upregulation of the expression of CD69, CD25, ICAM-1, and p75TNF-R on the cell surface. These findings demonstrate that PTX has a marked suppression on IL-2-mediated activation of immature free NK cells and that the suppression is due, in large part, to PTX-mediated inhibition of endogenous TNF- secretion. The implication of these findings in the clinical use of PTX for therapy is discussed.