JOURNAL OF INTERFERON AND CYTOKINE RESEARCH 19:1135– 1144 (1999) Mary Ann Liebert, Inc.
Histamine Protects T Cells and Natural Killer Cells Against Oxidative Stress MARKUS HANSSON,1 SVANTE HERMODSSON,1 MATS BRUNE,2 ULF-HENRIK MELLQVIST,2 PETER NAREDI,3 ÅSA BETTEN,1 KURT R. GEHLSEN, 4 and KRISTOFFER HELLSTRAND 1
ABSTRACT Oxidative stress inflicted by monocytes/macrophages (MO) is recognized as an important immunosuppressive mechanism in human neoplastic disease. We report that two types of lymphocytes of relevance for protection against malignant cells, T cells and natural killer (NK) cells, became anergic to the T cell and NK cell activator interleukin-2 (IL-2) after exposure to MO-derived reactive oxygen metabolities and subsequently acquired features characteristic of apoptosis. The MO-induced anergy and apoptosis in T cells and NK cells were reversed by histamine, an inhibitor of reactive oxygen metabolite synthesis in MO. We propose that strategies to circumvent oxidative inhibition of lymphocytes may be of benefit in immunotherapy of neoplastic disease.
INTRODUCTION
C
(MO) lineage are frequently detected within neoplastic tumors and in surrounding tissue, but whether intratumoral and peritumoral MO are friends or foes in the host defense against tum or cells is a matter of controversy. (1) MO are endowed with an effective antitumor machinery and exert inter alia antibody-depende nt cellular cytotoxicity against tumor cells in vitro. Also, MO are antigen-presentin g cells (APC) and provide costim ulatory signals that are pivotal for the development of adaptive immunity against tum or cell antigens. (2) MO also can negatively influence the immune response against malignant cells. Two subsets of lymphocytes with protective activity against tumor cells, T cells and natural killer (NK) cells, are frequently detected within tumors. Functions of intratumoral or peritumoral T cells or NK cells are impaired as compared with NK/T cells in peripheral blood or in adjacent, nonmalignant tissue. The inhibition of lymphocytes in tumors has been attributed to changes in the expression of the CD3z chain, which is part of a critical signal transduction pathway in T cells and NK cells. CD3z is downmodulate d on intratumoral T cells and NK cells in several forms of hum an and experimental cancer. (3–7) That MO are responsible for inhibition of EL LS O F TH E M O N O C Y T E M A C RO PH A G E
tumor-infiltratin g lymphocytes is suggested by the findings that depletion of MO in vivo restores CD3z expression in intratumoral lymphocytes in mice (5 ) and that MO recovered from human malignant tum ors or tumors from experimental animals induce a pronounced downmodulati on of CD3z in lymphocytes ex vivo. (5,6) A main part of the MO-induced inhibition of signal transduction in intratumoral lymphocytes is mediated by hydrogen peroxide and other reactive oxygen metabolites (ROM) generated by MO (oxidative stress) (review ed in ref. 7). In vitro data lend further support to the notion that MO and, in particular, MO-inflicted oxidative stress profoundly suppress activities of lymphocytes with antitumor properties. Early studies revealed that the proliferation of lymphocytes in response to lectins was significantly reduced by MO-derived ROM.(8,9) More recently, it was dem onstrated that MO recovered from peripheral blood or from tum or tissue inhibit the tumor-killing activity of NK cells (1 0,11) and reduce CD3z expression in T cells. (5,6) Antioxidative agents, such as scavengers of ROM and inhibitors of ROM formation, reverse these MO-induced suppressive events, (5,8–11) suggesting that products of the oxidative metabolism of MO mediate the inhibition. The T cell-derived cytokine interleukin (IL-2) is a prototypic activator of T cell and NK cell functions. It expands the target
1 Department
of Virology, University of Göteborg, S-413 46 Göteborg, Sweden. of Hematology, University of Göteborg, S-413 45 Göteborg, Sweden. 3 Department of Surgery, Umeå University Hospital, S-901 85 Umeå, Sweden. 4 Maxim Pharmaceutica ls, Inc., San Diego, CA 92122. 2 Department
1135
1136
HANSSON ET AL.
cell spectrum of NK cells and induces the de novo appearance of certain markers for activation, transcription of cytokine genes, and cell cycle proliferation in both cell types. (1 2,13) Results obtained in tumor-bearing experimental animals support that activation of T cells (14) and NK cells (15) is pivotal for the antitum or properties of IL-2 treatment in vivo. Despite the impressive activating effects of IL-2 on lymphocytes with antitumor efficacy, the results of clinical trials using IL-2 in solid tumor diseases and in hematopoietic malignancies have not yet fulfilled expectations. Typically, T cells and NK cells in peripheral blood are strongly activated during IL-2 therapy, but the malignant tumor burden remains unchanged in the majority of treated patients (reviewed in ref. 16). What is the impact of MO-induced, ROM-mediated inhibition of T cells and NK cells on the therapeutic efficacy of IL2? To address this question, we reconstituted in vitro an environment of MO-inflicted oxidative stress, aiming to mimic the situation within tum ors. In this model, human T cells and NK cells are exposed to ROM generated spontaneously by autologous MO.(10,11) As a marker for IL-2-induced activation, we chose the CD69 (Leu-23) antigen, a cell surface activation antigen that is inducibly expressed after short-term exposure of mature T cells and NK cells to IL-2. (17 ) In addition, we studied the effects of antioxidative agents on the reactivity to IL-2 of T cells and NK cells exposed to MO-derived ROM. These antioxidative agents included histamine, an inhibitor of ROM synthesis in MO, diphenyleneionodo nium (DPI), an inhibitor of NADPH oxidase-dependen t formation of ROM in phagocytes, and catalase, a scavenger of ROM.
CD3e 1 /56 2 T cells (35%–40%), CD3e 2 /56 2 (5% –10%), and CD3e 1 /56 1 cells (1%–5%), as judged by flow cytometry. In som e experiments, dynabeads (Dynal A/S, Oslo, Norway) coated with anti-CD 56 were used to obtain purified lymphocyte preparations of T cells, as described in detail elsewhere. (9) The mixture of NK and T cells was exposed to autologous, elutriated MO in microplates. The lymphocytes (150,000 cells/well in 200 m l) were incubated with or without MO (150,000 cells/well) for 16 h at 37°C.
Chemiluminescence analysis The spontaneous extracellular generation of superoxide anion in elutriated MO was measured by isoluminal-enhan ced chemiluminesc ence as described in detail elsewhere. (10,18 )
Assays of apoptosis Apoptosis was monitored by flow cytometry. In these experiments, we gated nonviable T cells or NK cells after exposure to MO, as described elsewhere. The gate was set to com prise lymphocytes with reduced forward scatter and increased right angle scatter characteristic of apoptosis. (11) Two additional methods were used to determ ine apoptosis in T cells and NK cells: analysis of DNA strand breaks by terminal deoxynucleotidyl transferase-med iated bromolated dUTP nick end-labeling of DNA fragments (TUNEL assay) (as described in ref. 19) and analysis of reduced intracellular glutathione (as described in ref. 20).
Detection of surface antigens MATERIALS AND METHODS Separation of mononuclear cells (MNC) Peripheral venous blood was obtained as freshly prepared leukopacks from healthy blood donors at the Blood Centre, Sahlgren’ s Hospital, Göteborg, Sweden. The blood (65 ml) was mixed with 92.5 ml Iscove’ s medium, 35 ml 6% Dextran (Kabi Pharmacia, Stockholm, Sweden), and 7.5 ml ACD (Baxter, Deerfield, IL). After incubation for 15 min at ambient tem perature, the supernatant was carefully layered onto Ficoll-Hypaque (Lymphoprep, Nyegaard, Norway). The MNC were collected at the interface after centrifugation at 380g for 15 min, washed twice in phosphate-buffere d saline (PBS), and resuspended in Iscove’ s medium supplemented with 10% hum an AB 1 serum . During all further separation of cells, the cell suspensions were kept in siliconized test tubes (Vacuette, Greiner, Stockholm). The cells were further separated into lymphocytes and MO using the countercurrent centrifugal elutriation (CCE) technique, as described in detail elsewhere. (10,11) Briefly, the MNC were resuspended in elutriation buffer containing 0.5% bovine serum albumin (BSA) and 0.1% EDTA in buffered NaCl and fed into a Beckman J2-21 ultracentrifuge with a JE-6B rotor at 2100 rpm . A fraction with . 90% MO was obtained at a flow rate of 18 ml/min. A lym phocyte fraction enriched for NK cells (CD3e 2 /56 1 phenotype) and T cells (CD3e 1 /56 2 ) was recovered at flow rates of 14–15 ml/min. This fraction contained 3% MO and consisted of CD3e 2 /56 1 NK cells (45% –50%).
One million cells were incubated with appropriate fluorescein isothiocyanate (FITC) conjugated and phycoerythrin (PE)conjugated monoclonal antibodies (mAb) (Becton Dickinson, Stockholm, Sweden) (10 m l/10 6 cells) on ice for 30 min. The cells were washed twice in PBS and resuspended in 500 m l sterile filtrated PBS and analyzed by flow cytometry on a FACSort with a Lysys II software program (Becton Dickinson). Lymphocytes were gated on the basis of forward and right angle scatter. The flow rate was adjusted to , 1000 cells 3 sec 2 1 , and at least 5 3 103 cells were analyzed for each sample, unless otherwise stated.
Compounds Human recombinan t IL-2 (Genzym e, Stockholm, Sweden), histamine dihydrochlo ride (Sigma Chemicals, Stockholm , Sweden), ranitidine hydrochlori de (Glaxo, Mölndal, Sweden), AH20399AA (Glaxo), diphenylion odonium (Sigma Chemicals, St. Louis, MO), catalase (Boehringer Mannheim , Mannheim, Germany), and a Fas ligand inhibitor (com prising the extracellula r domain of hum an Fas, amino acids 1–154, fused to the Fc portion of hum an IgG 1) (Kamiya Biomedical Company, Seattle, WA) were used. MNC treated with culture medium served as controls. All compound s were readily dissolved in culture medium . FITC-conjugated and PE-conju gated mAb against CD3 e , CD4, CD8, CD56, and CD69 were purchased from Becton Dickinson. Reagents and media were regularly checked for the presence of endotoxin using the limulus amebocyte assay.
1137
HISTAMINE AND OXIDATIVE STRESS
Statistics For statistical analysis, the Mann-Whitney U-test was used.
RESULTS Expression of CD69 activation antigen in T cells and NK cells We determined the expression of CD69 on T cells and NK cells incubated in the presence or absence of MO. The follow ing lymphocyte subsets were studied: CD3e 1 (all T cells), CD3e 1 /4 1 (T helper cells), CD3e 1 /8 1 (T suppressor/ cytotoxic cells), and CD3e 2 /8 2 /56 1 (NK cells). In concurrence with earlier studies, (17) T cells with CD3e , CD4, or CD8 phenotype acquired the CD69 activation antigen when treated with IL-2 (100 U/ml) in the absence of MO. CD69 appeared at 2–4 h after the onset of treatment with IL-2 and reached a maximum at 16–24 h. The effect of IL-2 on CD69 expression was dose dependent at final IL-2 concentrations of 1–100 U/ml (data not shown). The IL-2-induced expression of CD69 in these subsets of T cells was significantly reduced by the addition of MO. No significant qualitative differences were observed between CD41 cells and CD81 T cells as regards the induction of CD69 by IL-2 or the inhibition of the IL-2 response by MO (Table 1). The presence of histamine (50 m M) did not significantly alter the expression of CD69 in either subset of control or IL-2 treated T cells incubated without MO. However, histamine counteracted the MO-induced inhibition of IL-2-induced acquisition of CD69 in T cells. Thus, histamine restored the expression of CD69 to the level observed in the absence of MO (Table 1). Figure 1 (A–D) shows a histogram of the IL-2 induced expression of CD69 in gated CD3e 1 lym phocytes incubated with and without MO and treated with or without histamine.
T A B LE 1.
E X P RE SSIO N
OF
The expression of CD69 on unstimulated, viable CD56 1 NK cells (mean 21%) was significantly higher than that of CD41 (1%) or CD81 T cells (9% ). The majority (mean . 90% ) of NK cells acquired CD69 after incubation with IL-2 in the absence of MO, with a time and dose dependence similar to that observed for T cells. IL-2 only weakly induced CD69 on NK cells when these cells were incubated with MO (Table 1). Experiments were carried out in which the inhibition of reactivity to IL-2 in CD31 T cells was compared with that of CD56 1 NK cells when these cells were exposed to different amounts of MO. Although the MO-induced inhibition of IL-2 reactivity varied between experiments, a significant anergy to IL-2 was usually observed at an MO/lym phocyte ratio of 1:4, and the degree of inhibition was similar in T cells and NK cells (data not shown). Histamine restored the IL-2-induced expression of CD69 to the level observed in the absence of MO (Table 1). Figure 1 (E–H) shows contour plots of the expression of CD56 and CD69 in gated, viable lymphocytes incubated with and without MO and treated with histamine or IL-2. In accordance with earlier studies, (11) MO induced the disappearance of CD56 on otherwise unstimulated NK cells, and this downm odulation of CD56 was prevented by histamine (Fig. 1E, F). The effect of histamine on IL-2-induced expression of CD69 in T cells and NK cells was dose dependent at final histamine concentrations of 0.1–50 m M, with an ED 50 of approximately 2 m M (Fig. 2), similar to that previously reported for histamine H 2 -receptor-media ted effects on leukocyte functions. (10) Further, histamine-induce d reversal of the T cell and NK cell anergy to IL-2 was completely antagonized by ranitidine, an antagonist at histamine H 2 receptors used at concentrations equimolar or 10-fold lower than histam ine (Fig. 2). To exclude nonspecific effects of ranitidine, we used a ranitidine analog, AH20399AA (C 13 H 22 O 4 ), in which the thioether group of ranitidine has been replaced by an ether, thereby reducing its affinity for H 2 . 50-fold. (10) AH20399AA did not block the histamine effect (data not shown).
CD69 1 A C TIV A TIO N A N TIG EN
IN
T C EL LS
AND
NK C ELLS
% activated (CD691 ) lymphocytes with respective phenotype a Treatment Control Histam ine IL-2 IL-2 1 histamine
CD3e 2 2 27 27 6
6
6
6
1
0.3 0.3 4 4
CD3e 1 1 MO 2 3 10 31 6
6
6
6
0.2 0.6 3c 4e
CD41 1 1 21 24 6
6
6
6
0.3 0.3 5 4d
1
CD41 MO
1 2 7 30 6
6
6
6
0.3 0.4 3d 5f
CD81 9 9 33 33 6
6
6
6
CD81 MO
b
1
2 2 5 3
4 9 11 41 6
6
6
6
1 2 5d 5f
CD561 1 MO
CD561 21 19 94 94 6
6
6
6
3 3 1 1
19 23 40 91 6
6
6
6
6 4 10 c 2g
a Lymphocytes and MO were recovered from peripheral blood mononuclear cells by centrifugal elutriation. (11) The lym phocytes (150,000 cells/well in a total volum e of 200 m l) were incubated with or without MO (150,000 cells/well) and were concomitantly treated with IL-2 (100 U/ml), histamine (50 m M), or culture medium (control) for 16 h at 37°C. CD69 expression was monitored in indicated subsets of lym phocytes by use of flow cytometry in gates com prising all viable lymphocytes. The data show the mean percentage 6 SEM of CD69 1 cells with respective phenotype recovered from up to 11 blood donors. b A significant fraction of NK cells carry the CD8 antigen. (12) In studies of CD81 T cells, CD81 NK cells were depleted by use of anti-CD -56-coated beads, as described elsewhere. (11) Similar results were obtained in experiments in which three-color analysis of CD3e 1 /8 1 /56 2 T cells was performed (data not shown). c p , 0.01, d p , 0.05 vs. cells treated with IL-2 without MO (Mann-Whitney U-test). e p , 0.001, f p , 0.05, g p , 0.01 vs. MO-containing cells treated with IL-2 without histamine.
HANSSON ET AL.
Role of ROM MO can reduce molecular oxygen and generate ROM (respiratory burst) both spontaneously and in response to certain soluble or particulate stimuli (reviewed in ref. 21). Experim ents were performed in which DPI, an inhibitor of NADPH oxidase activity in MO,(2 2) was added to mixtures of lymphocytes and MO in studies of the acquisition of CD69 on T cells and NK cells in response to IL-2. DPI significantly reversed the MOinduced inhibition of T cells (Fig. 3) and NK cells (not shown). MO also produce reactive nitrogen interm ediates, of which nitric oxide (NO) is the ultimate effector molecule, and DPI is an inhibitor also of NO synthase (NOS).(22) To study whether NO induction in MO contributed to the observed T cell and NK cell anergy to IL-2, we used an NOS inhibitor, N-monomethyl- L arginine ( L -NMMA). This com pound, used at concentrations sufficient to inhibit NO synthesis in MO,(11) did not affect the MO-induced suppression of T cells and NK cells (data not shown). Catalase, a scavenger of hydrogen peroxide, (2 1) significantly reversed the MO-induced inhibition of IL-2-induced CD69 expression in T cells and NK cells at concentrations exceeding 50 U/ml, whereas superoxide dism utase, a scavenger of superoxide anion, (21) was ineffective at concentrations sufficient to scavenge . 90% of superoxide anion (200 U/ml) (data not shown).
Apoptotic cell death in T cells and NK cells
FIG. 1. Anergy to IL-2 in CD3e 1 T cells and CD56 1 NK cells: reversal by histamine. Lymphocytes and MO were incubated in microplates and treated with IL-2 or histamine or both as described in Table 1. Cells were labeled with PE-conjugated mAb against CD3e or CD56 and an FITC-labeled mAb against CD69. CD3e 1 or CD56 1 lym phocytes were gated, and the relative fluorescence intensity and the percentage of cells stained with anti-CD69 were determined over 50,000 events. (A, B, C, D) Histogram of CD69 (abscissa) and the number of viable CD3e 1 lymphocytes (ordinate). (E, F, G, H) Contour plot of CD69 (abscissa) and CD56 (ordinate). Bold numbers represent the percentage of respective cell type with CD69 1 phenotype. (A) CD3e 1 lymphocytes 1 MO. (B) CD3e 1 lymphocytes 1 MO 1 histamine (50 m M). (C) CD3e 1 lym phocytes 1 MO 1 IL-2. (D) CD3e 1 lymphocytes 1 MO 1 IL-2 1 histamine. (E) CD56 1 lymphocytes 1 MO. (F) CD56 1 lymphocytes 1 MO 1 histamine (50 m M). (G) CD561 lym phocytes 1 MO 1 IL-2. (H) CD56 1 lym phocytes 1 MO 1 IL-2 1 histamine.
Earlier studies have revealed that oxidative stress induces apoptotic cell death in many types of mammalian cells, including lymphocytes (reviewed in ref. 23). To further study the fate of T cells and NK cells after contact with MO, we used flow cytometry to monitor apoptosis in lymphocytes. It was found that a large fraction of T cells and NK cells acquired reduced forward scatter and increased right angle scatter characteristic of apoptosis after overnight incubation with MO. Analyses of DNA fragm entation and depleted intracellula r glutathione were employed to confirm apoptosis in these cells. DNA fragm entation was analyzed in gated lymphocyte subsets. Lymphocytes and MO from up to nine blood donors were incubated in microplate s and treated with culture medium (control) or histam ine (Table 1). Cells were labeled with PEconjugated mAb against CD3 e and DNA strand breaks in lym phocytes were analyzed by term inal deoxynucleo tidyl transferase-med iated bromolated dUTP nick end labeling of DNA fragments (TUNEL assay). (19 ) The mean percentage of control CD3 e 1 lymphocyte s with fragmented DNA in the absence of MO was 3% 6 1% (mean 6 SEM of nine donors) and was not significantl y changed by histamine, IL-2, or histamine 1 IL-2. The percentage of CD3 e 1 T cells with fragm ented DNA in the mixture of lym phocytes and MO was 57% 6 7%, 9% 6 2%, 68% 6 9%, and 7% 6 3% after treatment with culture medium, histam ine, IL-2, or histamine 1 IL-2, respectivel y. The MO-induced DNA fragm entation in T cells and the protective effect of histamine were statisticall y significant (p , 0.01). Intracellular glutathione is assumed to be a defense mechanism against oxidative damage, (2 4) and oxidatively induced apoptotic cell death is accompanied by depletion of intracellular stores of glutathione. (2 4,25) We observed a striking reduction of intracellular glutathione in T cells and in NK cells after incubation with MO (Fig. 4). The depletion of glutathione was
1139
HISTAMINE AND OXIDATIVE STRESS barely detectable at 4 h after the start of incubation with MO and reached a maxim um at 16–24 h (data not shown). NK cells were more sensitive to the MO-induced triggering of apoptosis were than T cells. Thus, a mean of 63% of CD56 1 NK cells and 39% of CD3e 1 T cells displayed morphologic features characteristic of apoptosis after contact with MO, and this difference attained statistical significance. A mean of 45% –55% of CD41 or CD81 /56 2 cells were apoptotic after contact with MO, and the propensity of MO-induced apoptotic cell death was similar in these T cell subsets (Table 2). The frequency of T cells or NK cells carrying CD69 after activation with IL-2 was similar in apoptotic and viable lym phocytes (data not shown), suggesting that induction of CD69 expression occurred prior to the development of apoptosis. The MO-induced apoptosis in NK cells and T cells was prevented by histamine but unaffected by concomitant treatment with IL-2 (Table 2 and Fig. 4). The protective effect of histamine against MO-induced apoptosis was mimicked by catalase (not shown) and by DPI (Fig. 3) and completely reversed by ranitidine but not by AH20399AA (Fig. 2) (data not shown). By reversing the anergy to IL-2 and concomitantly preventing lymphocyte apoptosis, histam ine significantly increased the num ber of viable T cells and NK cells incubated with MO. A calculation in which both the expression of CD69 and the degree of apoptotic cell death (on the basis on apoptotic morphology) were considered revealed that IL-2 alone augmented the number of viable activated NK cells 1.8-fold in the presence of MO (as compared with untreated NK cells incubated with MO). The corresponding figure for NK cells treated with histamine (50 m M) 1 IL-2 was 11.2-fold (p , 0.001 vs. IL-2, Mann-Whitney U-test). In CD3e 1 T cells admixed with MO,
IL-2 increased the number of activated T cells 5.9-fold over control, whereas histamine 1 IL-2 increased the number of activated cells 61.2-fold (p , 0.001 vs. IL-2, Mann-Whitney Utest).
Role of Fas/FasL interaction The Fas ligand (CD95L) triggers apoptosis in many cell types after interaction with the Fas receptor (CD95), which is expressed on inter alia T cells (26 ) and NK cells. (27 ) To evaluate the role of FasL/Fas interactions for the observed oxidatively induced apoptosis, we used a Fas ligand inhibitor, which com prises the extracellular domain of human Fas (am ino acids 1–154), fused to the Fc portion of human IgG1. This Fas/FcIgG fusion protein, used at a concentration (20 m g/ml) sufficient to reduce FasL-mediated, activation-induced apoptosis in T cells by . 60% ,(26) did not affect the MO-induced anergy to IL-2 or the MO-induced apoptosis in T cells or NK cells (Table 3).
ROM formation in MO From these data, we conclude that histamine, catalase, and DPI prevent oxidative inhibition of T cells and NK cells and, thereby, allow appropriate activation of these cells by IL-2 also in the presence of ROM-generating MO. The ROM-induced inhibition was spontaneous, that is, expressed in the absence of exogenous ROM-triggering stimuli, and it is likely that the ROM-induced anergy to IL-2 and apoptosis were dependent on the spontaneous extracellular release of ROM from adjacent MO. Earlier studies have revealed that histam ine and DPI inhibit stimulus-triggere d ROM formation in MO,(10,22) but the
FIG. 2. Histamine-indu ced protection of CD3e 1 T cells and CD56 1 NK cells against oxidative inhibition: dose-response and antagonism by ranitidine. Lymphocytes and MO were recovered from peripheral blood as described in the legend to Figure 1. A mixture of MO and lymphocytes was treated with culture medium (control, open circles) or the histam ine H 2 -receptor antagonist ranitidine (closed circles) and histamine (at indicated final concentrations) for 16 h. Ranitidine was used at concentrations equimolar to histamine. After incubation, lym phocytes were labeled with antibodies to CD3e , CD56, and CD69. Data show IL2-induced (100 U/ml) CD69 expression in viable T cells (CD3 e 1 ) and NK cells (CD56 1 ) (left) and the percentage of T cells and NK cells with reduced forward and increased side angle scatter characteristic of apoptosis (right). The results are representative of three similar experiments.
1140
HANSSON ET AL.
FIG. 3. Protection of CD3e 1 T cells against oxidative inhibition by DPI. Lymphocytes and MO were recovered from peripheral blood as described in the legend to Figure 1. A mixture of MO and lymphocytes was treated with culture medium (control, open circles) or IL-2 (100 U/ml) (closed circles) for 16 h. After incubation, lymphocytes were labeled with antibodies to CD3e and CD69. Data show CD69 expression in viable T cells (CD3 e 1 ) (left) and the percentage of T cells with reduced forward and increased side angle scatter characteristic of apoptosis (right). Similar results were obtained when CD69 expression was examined in CD56 1 NK cells incubated with MO: 29.5% (control) or 79.0% (DPI 1000 nM) of NK cells acquired the CD69 antigen in response to IL-2. DPI did not increase the IL-2-induced expression of CD69 in T cells or NK cells incubated in the absence of MO (not shown). The results are representative of three similar experiments.
effects of these agents on the constitutive extracellular release of ROM are not known. The assess whether DPI and histamine inhibited the spontaneous release of ROM from MO, we used a chemilumines cence assay that specifically quantifies extracellular ROM (superoxide anion). A more than 4-fold reduction of released extracellular superoxide anion was observed at a DPI concentration of 10 nM, and similar results were obtained in three experiments using MO recovered from three blood donors. Similarly, histamine (at 50 m M) inhibited the concentration of extracellular superoxide anion in this model more than 5-fold. This effect of histam ine was completely antagonized by ranitidine, used at concentrations equimolar to histamine (data not shown).
FIG. 4. Protection of T cells and NK cells from MO-induced apoptosis by histam ine. Lymphocytes and MO were incubated in microplates and treated with culture medium (control) or histamine as described in Table 1. Cells were labeled with PEconjugated mAb against CD3e or CD56 and 5-chlorom ethylfluorescein diacetate (Cell Tracker Green CMFDA, Molecular Probes, Leiden, The Netherlands), as described in detail elsewhere. (20) Bold numbers represent the percentage of CD3e 1 T cells or CD56 1 NK cells with depleted intracellular glutathione. The results are representative of five similar experiments. (A) CD3e 1 lymphocytes 1 MO. (B) CD3e 1 lymphocytes 1 MO 1 histamine (50 m M). (C) CD3e 1 lymphocytes 1 MO 1 IL-2. (D) CD3e 1 lymphocytes 1 MO 1 IL-2 1 histam ine. (E) CD56 1 lymphocytes 1 MO. (F) CD56 1 lymphocytes 1 MO 1 histamine (50 m M). (G) CD56 1 lymphocytes 1 MO 1 IL-2. (H) CD56 1 lym phocytes 1 MO 1 IL-2 1 histamine.
1141
HISTAMINE AND OXIDATIVE STRESS T A B LE 2.
MO-IN D U C ED A PO PT O SIS
IN
T C E LLS
AND
NK C ELL S : R EV ER S A L
BY
H IS TA M IN E
% apoptotic lymphocytes with respective phenotype a CD3e
Treatment Control Histamine IL-2 IL-2 1 histamine
39 7 42 13
1
1 6 6 6 6
MO 7 1c 6 3f
CD41 53 9 53 16
1 6 6 6 6
CD81
MO 6 5d 4 9g
46 6 48 8 6
6
6
6
1
MO 14 2e 13 2g
CD56 1 63 12 69 18 6
6
6
6
1
MO 8b 2d 6 2h
a Lymphocytes and MO were recovered from peripheral blood and labeled with anti-CD3 e , anti-CD4, anti-CD8, and anti-CD 56 as described in Table 1. Apoptosis was measured by flow cytometry using a gate for lymphocytes with reduced forw ard scatter and increased right angle scatter. The data show the mean apoptotic cells with respective phenotype 6 SEM. The percentage of lymphocytes with apoptotic features when incubated without MO was 7 6 2 (CD3e 1 ), 14 6 8 (CD41 ), 5 6 2 (CD8 1 ), and 14 6 3 (CD56 1 ) (mean 6 SEM) and was not significantly altered by histamine, IL-2, or histamine 1 IL-2. Cells were recovered from up to 11 blood donors. b p , 0.05 vs. CD3e 1 cells. cp , 0.001, d p , 0.01, e p , 0.05 vs. control. f p , 0.01, g p , 0.05, h p , 0.001 vs. cells treated without histam ine.
DISCUSSION The object of this study was to determine the role of MO, and in particular MO-derived ROM, in the activation of hum an T cells and NK cells by IL-2. In short, our data demonstrate that autologous MO induce anergy to the T cell and NK cellactivating properties of IL-2, followed by fragmentation of DNA and apoptosis in CD41 and CD81 T cells and in CD56 1 NK cells. The lym phocyte anergy and the triggering of apoptosis occurred independently of Fas/FasL interactions and were mediated by MO-derived ROM, as inhibitors of ROM synthesis (histamine and DPI) and a scavenger of hydrogen peroxide (catalase) prevented the MO-induced anergy and apoptosis. Earlier studies have revealed that cytokine-induced activation of NK cell cytotoxicity against appropriate target cells is inhibited by MO-derived ROM.(10) The results in this study confirm and extend these observations by demonstrating that the acquisition of CD69 on the surface of NK cells in response to IL-2 is similarly inhibited by MO-derived ROM. Also, our data show that IL-2-induced activation of T cell subsets is inhibited by MO-derived ROM in an apparently similar fashion. An advantage with the study of CD69 is that the expression of this activation antigen can be monitored in phenotypically distinct
T A B LE 3. Treatment Control DPI Fas/Fc-IgG
F A S /F A S L-I N D EP EN D EN T A N ER G Y
Viable CD31 /CD69 1 (gated events) a 89 701 113
lymphocyte subsets, and it was possible to compare the sensitivity to oxidatively induced inhibition in subsets of T cells and in NK cells. These comparisons revealed that all subsets of T cells were anergic to IL-2 after exposure to MO, and no significant differences between CD4 and CD8 cells were detected. Also, the degree of MO-induced anergy to IL-2 was similar in CD31 T cells and CD56 1 NK cells. However, the propensity of MO-induced apoptotic cell death was significantly higher in NK cells than in T cells. These results extend the previous finding that low er concentrations of exogenous hydrogen peroxide are required to induce apoptosis in NK cells than in T cells. (11) The mechanism s underlying this difference should be studied further, but it seems clear that NK cells are more prone to apoptosis induced not only by oxidative stress but also by treatment with inhibitors of DNA repair or g -irradiation. (11) Thus, the difference in MO-induced cell death among T cells and NK cells observed in this study probably reflects a higher tendency of NK cells to enter programmed cell death regardless of the apoptosis-inducing stimulus rather than a specific difference in acquiring apoptotic characteristics when exposed to oxidative stress. Which species of ROM mediate the inhibition of lymphocyte reactivity and trigger apoptosis? Our data do not precisely
AND
A PO PT O S IS
IN
T C E LLS
AND
NK C ELL S
Viable CD56 1 /69 1 (gated events)
Apoptotic CD31 /CD69 1 (%) b
Apoptotic CD561 /CD69 1 (%)
36 1248 26
52 7 57
71 13 68
a Lymphocytes and MO were recovered from peripheral blood and labeled with anti-CD 3e , anti-CD56 , and anti-CD69 as described in Table 1 and analyzed for respective phenotype by flow cytometry. DPI was used at 100 nM and Fas/Fc-IgG at 20 m l. b Apoptosis was measured by flow cytometry using a gate for lymphocytes with reduced forward scatter and increased right angle scatter. Similar results were obtained in three separate experiments.
1142 define the mediator responsible for the inhibition, but it seems safe to conclude that ROM downstream of superoxide anion play a major role, as an inhibitor of NADPH oxidase activity and a scavenger of hydrogen peroxide reversed the MO-induced inhibition, whereas a scavenger of superoxide anion did not. Hydrogen peroxide is converted by myeloperoxidas e (MPO) to a variety of toxic radicals, including hypohalous acids, (2 1) and in further studies, we have show n that scavengers specific for MPO-derived radicals prevent the MO-induced inhibition of T cells and NK cells as effectively as scavengers of hydrogen peroxide (Å Betten et al., manuscript in preparation), suggesting a predom inant role in the inhibition for MPO-dependent species of ROM. By use of a Fas ligand inhibitor, we explored the role of the FasL/Fas interaction for the MO-induced lym phocyte apoptosis and the anergy to IL-2. Earlier studies have revealed that activation-induced , Fas-dependent cell death in T cells is reversible by a variety of antioxidants, which reportedly act by inhibiting transcriptional activation of FasL.(28) Our finding that the MO-induced suppressive events were not reversible by the Fas/Fc-IgG fusion protein is in agreem ent with the results of a recent study demonstrating that apoptosis in T cells induced by exogenous hydrogen peroxide occurs independently of FasL/ Fas interaction. (29) Several investigators report that the tumor environm ent is frequently subjected to MO-induced oxidative stress, (3–7) and our results suggest that IL-2 is a weak activator of lymphocytes with antitumor efficacy in such an environment. Animal studies support that MO-dependen t inhibitory signals may be important for the antitumor efficacy of IL-2 in vivo Bottazzi et al.(30) introduced a gene encoding for an MO-attracting protein (monocyte chem otactic protein type 1, MCP-1) in B16 murine melanom a cells. When inoculated into mice, these MCP-1 expressing clones produced melanom a tumors with a high intratumoral content of MO. MCP-1-transfected tumors were considerably less susceptible to IL-2 treatment than were MCP-1-negative control tumors, (30) but the role of oxidative inhibition of lymphocytes for the inhibitory effect of intratumoral MO was not disclosed. A more direct indication of a role for ROM-mediated inhibition of IL-2 reactivity in vivo was provided by Yim et al. (31) These authors combined IL-2 with Nacetyl-cysteine (NAc-cys), an agent known to increase intracellular glutathione levels and, thereby, to protect lymphocytes against oxidative stress. (24,25) Mice carrying an IL-2-refractory subcutaneous tum or responded to combined treatment with IL2 and oral NAc-cys. (31) In addition, treatment with histamine or specific H 2 -receptor agonists has been reported to improve the antitumor efficacy of IL-2 in rats with an established, syngeneic prostate adenocarcinoma (32) and in mice carrying YAC1 lymphoma (33) or B16 melanoma. (33,34) In human metastatic melanom a, several investigators report that T cells are exposed to ROM, as judged from a reduction of CD3z expression in T cells and NK cells recovered from peripheral blood (35) or from tumor-involved lymph nodes (36) and the appearance of apoptotic lymphocytes in tum or tissue. (37) A reduced z chain expression in peripheral blood T cells was associated with a significantly shorter survival with a trend toward faster growing tum ors. (35) Kono et al. (6) ascribed the decreased z chain expression in T cells and in NK cells to ROM
HANSSON ET AL. generated by tumor-infiltratin g MO, as MO recovered from metastatic hum an lymph nodes induced a pronounced reduction of z chain expression in T cells and in NK cells in vitro, an event that was reversible by a scavenger of ROM. These results imply that melanoma could be an attractive candidate for therapy with IL-2 together with an antioxidant agent, but, as yet, only limited inform ation is available about the outcome of such treatm ent. We have used histamine dihydrochloride, at doses sufficient to saturate H 2 R, as a supplement to IL-2 in human metastatic melanom a, (38) and the putative efficacy of this treatment is currently being investigated in two phase III trials. Recent data indicate that lymphocytes from acute myelogenous leukemia (AML) patients also are subjected to oxidative stress. T cells and NK cells in the peripheral blood of patients with AML at various stages of disease display a pronounced reduction of CD3z expression. (39) Furthermore, treatment of AML patients with histamine and IL-2 has been found to normalize decreased CD3z expression in peripheral blood T cells and NK cells during therapy (J. Sjöberg et al., unpublished observations). In a phase II trial, histamine and IL-2 were given to AML patients in remission with the aim of preventing or delaying relapse of leukemia. (40,41 ) The potential benefit of histamine and IL-2 to prolong the duration of remission in AML is currently being investigated in a phase III trial.
ACKNOWLEDGMENTS We are indebted to Marie-Louise Landelius for expert technical assistance and to Claes Dahlgren, Bo Nilsson, Örjan Strannegård, and Elisabeth Wallhult for critically reading the manuscript. This work was supported by grants from the Swedish Cancer Society, the Swedish Medical Research Council, and Maxim Pharmaceutic als, Inc., San Diego, CA. K.H. is a research fellow of the Swedish Cancer Society.
REFERENCES 1. MANTOVANI, A., BOTTAZZI, B., COLOTTA, F., SOZZANI, S., and RUCO, L. (1992). The origin and function of tumor-associated macrophages. Immunol. Today 13, 265–270. 2. ARMSTRONG, T.D., PULASKI, B.A., and OSTRAND-ROSENBERG, S. (1998). Tumor antigen presentation. Cancer Immunol. Immunother. 46, 70–74. 3. MATSUDA, M., PETERSSON, M., LENKEL, R., TAUPIN, J.L., MAGNUSSON, I., MELLSTEDT, H., ANDERSON, P., and KIESSLING, R. (1995). Alterations in the signal-transduc ing molecules of T cells and NK cells in colorectal tumor-infiltrati ng, gut mucosal and peripheral lymphocytes: correlation with the stage of the disease. Int. J. Cancer 61, 765–772. 4. Mizoguchi, H., O’Shea, J.J., Longo, D.L., Loeffler, C.M., McVicar, D.W., and Ochoa, A.C. (1992). Alterations in signal transduction molecules in T lymphocytes from tumor-bearing mice. Science 258, 1795–1798. 5. OTSUJI, M., KIMURA, Y., AOE, T., OKAMOTO, Y., and SAITO, T. (1996). Oxidative stress by tumor-derived macrophages suppresses the expression of CD3 zeta chain of T-cell receptor complex and antigen-specific T-cell responses. Proc. Natl. Acad. Sci. USA 93, 13119–13124. 6. KONO, K., SALAZAR-ONFRAY, F., PETERSSON, M., HANS-
HISTAMINE AND OXIDATIVE STRESS
7.
8.
9.
10.
11.
12. 13. 14.
15.
16.
17.
18.
19.
20.
21. 22.
SON, J., MASUCCI, G., WASSERMAN, K., NAKAZAWA, T., ANDERSON, P., and KIESSLING, R. (1996). Hydrogen peroxide secreted by tumor-derived macrophages downmodulates signaltransducing zeta molecules and inhibits tumor-specific T cell and natural killer cell-mediated cytotoxicity. Eur. J. Immunol. 26, 1308– 1313. KIESSLING, R., KONO, K., PETERSSON, M., and WASSERMAN, K. (1996). Immunosuppr ession in human tumor-host interaction: role of cytokines and alterations in signal-transduc ing molecules. Springer Semin. Immunopathol . 18, 227–242. DOHLSTEN, M., LINDEN, O., HEDLUND, G., SJÖGKEN, H.O., DIAMANTSTEIN, T., and CARLSSON, R. (1986). Synergistic action of gam ma interferon and catalase to reverse the suppressive effect of peritonea l macrophage s on concanavalin A-induced lymphocyt e proliferation. Scand. J. Immunol. 24, 49–58. LINDEN, O., DOHLSTEN, M., BOKETOFT, A., HEDLUND, G., and SJÖGREN, H.-O. (1987). Catalase and lipopolysaccharid e enhance proliferation in the rat mixed lymphocyte reaction. Scand. J. Immunol. 26, 223–228. HELLSTRAND, K., ASEA, A., DAHLGREN, C., and HERMODSSON, S. (1994). Histaminergic regulation of NK cells. Role of monocyte-deriv ed reactive oxygen metabolites. J. Immunol. 153, 4940– 4947. HANSSON, M., ASEA, A., ERSSON, U., HERMODSSON, S., and HELLSTRAND, K. (1996). Induction of apoptosis in NK cells by monocyte-deriv ed reactive oxygen metabolites. J. Immunol. 156, 42–47. TRINCHIERI, G. (1989). Biology of natural killer cells. Adv. Immunol. 47, 187–376. HAWKINS, M.J. (1993). Interleukin-2 antitumor and effector cell responses. Semin. Oncol. 20, 52–59. BECKER, J.C., PANCOOK, J.D., GILLIES, S.D., FURUKAWA, K., and REISFELD, R.A. (1996). T cell-mediated eradication of murine metastatic melanoma induced by targeted interleukin 2 therapy. J. Exp. Med. 183, 2361–2366. BRUNDA, M.J., BELLA NTONI, D., and SULICH, V. (1987). In vivo anti-tumor activity of combinations of interferon alpha and interleukin-2 in a murine model. Correlation of efficacy with the induction of cytotoxic cells resembling natural killer cells. Int. J. Cancer 40, 365–371. WHITESIDE, T.L., VUJANOVIC, N.L., and HERBERMAN, R.B. (1998). Natural killer cells and tumor therapy. Curr. Top. Microbiol. Immunol. 230, 221–244. LANIER, L.L., BUCK, D.W., RHODES, L., DING, A., EVANS, E., BARNEY, C., and PHILLIPS, J.H. (1998). Interleukin 2 activation of natural killer cells rapidly induces the expression and phosphorylation of the Leu-23 activation antigen. J. Exp. Med. 167, 1572– 1585. LUNDQVIST, H., and DAHLGREN, C. (1996). Isoluminol-enhanced chemiluminesce nce: a sensitive method to study the release of superoxide anion from human neutrophils. Free Radic. Biol. Med. 20, 785–792. ALLEN, R.T., HUNTER, W.J., and AGRAWAL, D.K. (1997). Morphological and biochemical characterization and analysis of apoptosis. J. Pharmacol. Toxicol. Methods 37, 215–228. BARHOUMI, R., BOWEN, J.A., STEIN, L.S., ECHOLS, J., and BURGHARDT, R.C. (1993). Concurrent analysis of intracellular glutathione content and gap junctional intercellular communication. Cytometry 14, 747–756. KLEBANOFF, S.J. (1982). Oxygen dependent defence mechanisms of phagocytes. Adv. Host Def. Mech. 1, 111–151. MIESEL, R., KURPISZ, M., and KROGER, H. (1996). Suppression of inflammatory arthritis by simultaneous inhibition of nitric oxide synthase and NADPH oxidase. Free. Radic. Biol. Med. 20, 75–81.
1143 23. BUTTKE, T.M., and SANDSTROM, P.A. (1994). Oxidative stress as a mediator of apoptosis. Immunol. Today 15, 7–10. 24. DROGE, W., SCHULZE-OSTHOFF, K., MIHM, S., GALTER, D., SCHENK, H., ECK, H.P., ROTH, S., and GMUNDER, H. (1994). Functions of glutathione and glutathione disulfide in immunology and immunopatholo gy. FASEB J 8, 1131– 1138. 25. SLATER, A.F., STEFAN, C., NOBEL, I., VAN DEN DOBBELSTEEN, D.J., and ORRENIUS, S. (1995). Signalling mechanisms and oxidative stress in apoptosis. Toxicol. Lett. 82–83, 149–153. 26. ALDERSON, M.R., TOUGH, T.W., DAVIS-SMITH, BRADDY, S., FALK, B. SCHOOLEY, K.A., GOODWIN, R.G., SMITH, C.A., RAMSDELL, F., and LYNCH, D.H. Fas ligand mediates activation-induced cell death in human T lymphocytes. J. Exp. Med. 181, 71–77. 27. MEDVEDEV, A.E., JOHNSEN, A.C., HAUX, J., STEINKJER, B., EGEBERG, K., LYNCH, D.H., SUNDAN, A., and ESPEVIK, T. Regulation of Fas and Fas-ligand expression in NK cells by cytokines and the involvement of Fas-ligand in NK/LAK cell-mediated cytotoxicity. Cytokine 9, 394–404. 28. BAUER, M.K.A., VOGT, M., LOS, M., SIEGEL, J., WESSELBORG, S., and SCHULZE-OSTHOFF, K. (1998). Role of reactive oxygen intermediates in activation-induced CD95 (APO-1/Fas) ligand expression. J. Biol. Chem. 273, 8048– 8055. 29. DUMONT, A., HEHNER, S.P., HOFMANN, T.G., UEFFING, M., DROGE, W., and SCHMITZ, M.L. (1999). Hydrogen peroxide-induced apoptosis is CD95-independ ent, requires the release of mitochondria-der ived reactive oxygen species and the activation of NF-kappaB. Oncogene 18, 747–757. 30. BOTTAZZI, B., WALTER, S., GOVINO, D., COLOTTA F., and MANTOVANI, A. (1992). Monocyte chemotactic cytokine gene transfer modulates macrophage infiltration, growth and susceptibility to IL-2 therapy of a murine melanoma. J. Immunol. 148, 1280– 1285. 31. YIM, C.Y., HIBBS, J.B., McGREGOR, J.R., GALINSKY, R.E., and SAMLOWSKI, W.E. (1994). Use of N-acetyl cysteine to increase intracellular glutathione during the induction of antitumor responses by IL-2. J. Immunol. 152, 5796– 5805. 32. JOHANSSON, S., LANDSTRÖM, M., HELLSTRAND, K., and HENRIKSSON, R. (1998). The response of Dunning R3327 prostatic adenocarcinom a to IL-2, histamine and radiation. Br. J. Cancer 77, 1213– 1219. 33. ASEA, A., HERMODSSON, S., and HELLSTRAND K. (1996). Histaminergic regulation of natural killer-cell-med iated clearance of tumor cells in mice. Scand. J. Immunol. 43, 9–15. 34. HELLSTRAND, K., ASEA, A., and HERMODSSON, S. (1990). Role of histamine in natural killer cell-mediated resistance against tumor cells. J. Immunol. 145, 4365– 4370. 35. ZEA, A.H., CURTI, B.D., LONGO, D.L., ALVORD, N.G., STROBL, S.L., MIZOGUCHI, H., CREEKMORE, S.P., OSHEA, J.J., POWERS, G.C., and URBA, W.J. Alterations in T cell receptor and signal transduction molecules in melanoma patients. Clin. Cancer Res. 1, 1327– 1335. 36. RABINOWICH, H., BANKS, M., REICHERT, T.E., LOGAN, T.F., KIRKWOOD, J.M., and WHITESIDE, T.L. (1996). Expression and activity of signaling molecules in T lymphocytes obtained from patients with metastatic melanoma before and after interleukin 2 therapy. Clin. Cancer Res. 2, 1263– 1274. 37. WHITESIDE, T. (1998). Immune cells in the tumor microenvironment. Mechanisms responsible for functional and signaling defects. Adv. Exp. Med. Biol. 451, 167–171. 38. HELLSTRAND, K., NAREDI, P., LINDNER, P., LUNDHOLM, K., RUDENSTAM, C.M., HERMODSSON, S., ASZTELY, M., and HAFSTROM, L. (1994). Histamine in immunotherapy of advanced melanoma: a pilot study. Cancer Immunol. Immunother. 39, 416–419.
1144 39. BUGGINS, A.G., HIRST, W.J., PAGLIUCA, A., and MUFTI, G.J. (1998). Variable expression of CD3-zeta and associated protein tyrosine kinases in lymphocytes from patients with myeloid malignancies. Br. J. Haematol. 100, 784–792. 40. HELLSTRAND, K., MELLGVIST, U.H., WALLHULT, E., CARNESKOG, J., KIMBY, E., CELSING, F., and BRUNE, M. (1997). Histamine and interleukin-2 in acute myelogenous leukemia. Leuk. Lymphoma 27, 429–438. 41. WALLHULT, E., MELLGVIST, U.H., CELSING, F., HELLSTRAND, K., NILSSON, B., SIMONSSON, B., and BRUNE, M. (1998). Safety and feasibility of IL-2 plus histamine dihydrochloride in AML patients in first or subsequent complete remission. Blood 92, 2533.
HANSSON ET AL. Address reprint requests to: Dr. Kristoffer Hellstrand Departm ent of Virology University of Göteborg S-413 46 Göeborg Sweden Tel: 46 31 342 4731 Fax: 46 31 342 4960 E-mail: kristoffer.hellstran
[email protected] u.su Received 14 April 1999/Accepted 3 June 1999
