Clin. exp. Immunol. (1991) 85, 90-97
ADONIS 0009910491002002
Induction of anti-mycobacterial and anti-listerial activity of human monocytes requires different activation signals G. ZERLAUTH*, M. M. EIBLt & J. W. MANNHALTER*t *Department of Immunological Research, Immuno, Vienna, and Institute of Immunology, University of Vienna, Vienna, Austria (Acceptedfor publication 23 January 1991)
SUMMARY The requirements for activation of anti-mycobacterial and anti-listerial activity of human monocytes were investigated. Human monocytes could be activated to display enhanced anti-mycobacterial activity by a 24-h treatment with lipopolysaccharide. The mediator induced by this treatment was identified as being tumour necrosis factor-alpha (TNF-a). Addition of recombinant TNF-a (rTNF-a) to the cultures of human monocytes for 24 h yielded comparable results (minimal dose required for induction of anti-mycobacterial activity, 10 U/ml). Addition of anti-TNF-oc antibody completely abrogated the effect. A similar treatment protocol failed to activate enhanced anti-listerial activity. To trigger anti-listerial activity, sequential treatment of human monocytes with rTNF-oc and IL-2 was required. Treatment of monocytes with 10 U/ml rTNF-a for 24 h followed by incubation in the presence of 200 U/ml of IL-2 for an additional 24 h yielded a reduction of listerial growth which was moderate but statistically significant (P < 0-00 1). The activation of monocytes observed with rTNFa/IL-2 treatment was (i) dependent on both cytokines; (ii) sequence dependent (i.e. when IL-2 was added prior to rTNF-a, no effect was observed); and (iii) absent in cells treated with one cytokine only. Enhancement of anti-listerial activity by sequential use of cytokines was not accompanied by an increase in oxidative burst, which indicated that oxidative mechanisms were not the reason for the observed Listeria monocytogenes growth restriction. Further support for this hypothesis was obtained after interferon-gamma treatment of human monocytes which led to an augmented PMAinducible release of active oxygen radicals, but was not paralleled by growth restriction of L. monocytogenes. Our results indicate that TNF-ca plays a crucial role in the activation of monocytes for growth restriction of intracellular microbes. Activation of human monocytes to restrict the growth of the facultative intracellular bacteria Mycobacterium avium intracellulare and L. monocytogenes, however, follows different patterns, the initial trigger in both cases being provided by TNF-a-induced signals. Keywords Listeria monocytogenes Mycobacterium avium intracellulare cytokines human monocytes, anti-bacterial activity of some facultative intracellular pathogens, these cells eventually become activated and can interfere with the growth of the intracellular microbes (Hahn & Kaufmann, 1981). Acquired resistance against Mycobacteria is mediated by specific T lymphocytes and expressed by activated macrophages (Hahn & Kaufmann, 1981). However, the observation that mice with severe combined immunodeficiency (SCID) mutation can partially reduce listerial growth strongly argues for the additional involvement of a T cell-independent mechanism in the activation of anti-bacterial activity of cells of the monocyte/ macrophage lineage (Bancroft et al., 1987). The purpose of this work was to analyse the role of cells of the monocyte/macrophage lineage in anti-bacterial activity with special emphasis on the signals required for its induction. This was achieved by activation of human monocytes with cyto-
INTRODUCTION The host immune system is a barrier that must be breached or avoided by the invading micro-organism. This necessity is even more pronounced if the micro-organisms choose monocytes or macrophages as their preferred habitat. Within these phagocytic cells, the invaders face an environment that is designed to kill micro-organisms. Micro-organisms, however, have developed different strategies to evade or withstand the lethal attack within the monocytes/macrophages. Although monocytes and macrophages are initially permissive for intracellular multiplication of Correspondence: Dr J. W. Mannhalter, Immuno AG, Department of Immunological Research, Industriestrasse 131, A- 1220 Vienna, Austria.
90
Anti-bacterial activity of human monocytes kine(s) or with endotoxin and subsequent challenge of the monocytes with different facultative intracellular microbes, Listeria monocytogenes (strain EGD) or Mycobacterium avium intracellulare (strain Boone). Since these bacteria are capable of surviving in non-activated monocytes, they provide a wellsuited system for studying activation of monocytes for enhanced anti-bacterial activity. The results presented here indicate that human monocytes can be activated in vitro to interfere with the growth of M. avium intracellulare upon treatment with recombinant tumour necrosis factor-alpha (rTNF-a). Anti-listerial activity was observed only after sequential treatment with rTNF-a plus IL-2. These results indicate that TNF-ca is an important activator of human monocytes for enhanced antibacterial activity. Furthermore, the data suggest that different activation patterns are required for induction of anti-listerial or anti-mycobacterial activity. MATERIALS AND METHODS
Materials RPMI 1640, L-glutamine and fetal calf serum (FCS) were purchased from Flow Laboratories, Irvine, UK. Interleukin-2 (IL-2) was from Biotest, Frankfurt, Germany. rTNF-a was obtained from Genzyme, Boston, MA. Anti-human TNF-a monoclonal antibody (MoAb) was from The Accurate Chemical & Scientific Corporation, New York, NY. This MoAb (IgG I subclass) reacts with human natural and recombinant TNF-c (company brochure). Phorbol myristate acetate (PMA), nitroblue tetrazolium (NBT, grade III), saponin, and interferongamma (IFN-y) were purchased from Sigma, Deisenhofen, Germany. DMSO was from Merck, Munich, Germany. Lipopolysaccharide (LPS) SW Salmonella typhosa 0901 was purchased from Biotest, Frankfurt, Germany. Phytohemagglutinin (PHA, HA16) was from Wellcome, Dartford, UK. Cells were cultivated in Costar six-well plates (Costar, Badhoevedorp, The Netherlands). 7H9 broth and 7H 11 agar plates as well as tryptic soy broth and tryptic soy agar plates were prepared from powdered media purchased from Difco Biologicals, Detroit, MI. Isolation and activation of human peripheral blood monocytes Human peripheral blood mononuclear cells (MNC) were drawn from the venous blood of healthy volunteers and isolated by buoyant density gradient centrifugation (B0yum, 1968). The MNC from the interphase were aspirated, washed with saline and suspended in RPMI 1640 supplemented with 2 mM L-glutamine and 15% pooled, heat-inactivated human AB serum (complete medium). Antibiotics were omitted from the culture media to prevent possible growth-inhibiting effects in later killing experiments (Cole & Brastoff, 1975). Monocytes were isolated by adherence to plastic as described previously (Mannhalter, Borth & Eibl, 1986). Briefly, 2 x 107 mononuclear cells in I ml of complete medium were added to each well of sixwell culture plates; after 90 min, the non-adherent cells were discarded and the adherent cells were further incubated for 24 h in 1 ml/well of complete medium at 37°C in a humidified atmosphere containing 5% CO2. After this incubation period, > 93 % of the adherent population of cells stained positive in the non-specific esterase assay (Koski, Poplack & Blaese, 1976) and viability was 95% or better as assessed by trypan blue exclusion. For activation, the adherent monolayers were cultured in the
91
presence or absence of cytokines (rTNF-a, IL-2, IFN-y) for a
given period of time. The cytokines used throughout this study were tested for their possible direct influence on survival of the bacteria. Although the cytokines were not directly in contact with the bacteria, as the treated monocytes were washed
extensively prior to addition of bacteria, we grew the bacteria in medium containing cytokine concentrations that were similar to those used for monocyte modulation. The results demonstrated that none of the cytokines had any effect on bacterial growth
(data not shown). Preparation of PHA blasts MNC were deprived ofmonocytes by three rounds of adherence to plastic. The non-adherent cells were stimulated by addition of 1 mitogenic unit of PHA/2 x 106 MNC for 48-72 h. The resulting PHA blasts were added to adherent monocytes at a ratio of five blasts/monocyte for 24 h. Bacteria and bacterial counts M. avium intracellulare strain Boone (MAIS complex 14, ATCC 35761) was grown in 7H9 broth to a concentration of approximately 1 x 108 bacteria/ml. This number of micro-organisms was generally obtained within 12-14 days. The micro-organisms were harvested, washed and frozen in liquid nitrogen. After thawing, the number of viable bacteria was checked by plating serial dilutions on 7H II agar plates and counting bacterial colonies after 2-3 weeks. During the entire incubation period the agar plates were kept in sealed plastic bags to avoid drying. Infection of adherent monocytes with M. avium intracellulare was performed according to Rook et al. (1986). Briefly, adherent monocytes were rinsed extensively with warm RPMI 1640 supplemented with 5% human AB serum. The number of adherent cells per single well of the six-well culture plate was determined by assessing the number of monocytes in representative wells of each treatment (scraping off the monocytes with a rubber policeman and counting the cells in a Coulter counter). In this context, it is important to note that the treatment of monocytes with LPS, rTNF-a, IL-2 or IFN-y had no effect on the number of monocytes recovered/well. As generally observed in an outbred population, the total number of monocytes isolated varied with the individual donors. For infection, a number of M. avium intracellulare equivalent to the number of monocytes determined in the representative well was added to the remaining culture wells of one given type of experiment. For this purpose, M. avium intracellulare were thawed, appropriately diluted in RPMI 1640/5% AB serum and sonicated to produce a uniform suspension. Adherent monocytes were incubated in the presence of M. avium intracellulare for 4 days at 37°C in a humidified atmosphere containing 5% CO2. Growth of M. avium intracellulare was assessed by quantifying the number of viable bacteria recovered from lysed monocytes at day 0 and at 4 days post-infection. For the determination of the number of surviving M. avium intracellulare strictly standardized procedures were followed. Each well of a six-well plate received saponin to yield a final saponin concentration of 0 20/0. This saponin concentration did not interfere with M. atium intracellulare viability (data not shown). After 10 min of saponin lysis, the lysing solution of each well was transferred to a plastic tube and sonicated briefly to ensure dispersion of the bacilli. The mixture was serially diluted in 10-fold steps in ice-cold distilled water and plated in triplicate to 7H 11 agar plates. CFU were
92
G. Zerlauth, M. M. Eibl & J. W. Mannhalter
assessed 12-19 days later with a colony counter (BioFoss, Foss Electric, The Chantry, UK). These counts were converted to colony-forming units (CFU)/ml of lysate, where I ml of lysate was the product of an equal number of monocytes in a given experiment. Listeria monocytogenes strain EGD (ATCC 15313) was grown in tryptic soy broth to logarithmic phase, harvested, washed and frozen at -70'C. After thawing, the number of viable bacteria was checked by plating serial dilutions on tryptic soy agar plates and counting bacterial colonies at the end of 24-48 h. The infection of adherent monocytes with L. monocytogenes was performed according to Czuprynski, Campbell & Henson (1983) with minor modifications. Briefly, adherent monocytes were rinsed extensively with warm HBSS supplemented with 0-1 % gelatin (G-HBSS) and 4% FCS. The number of adherent cells per single well of the six-well culture plate was determined by assessing the number of monocytes in representative wells of each treatment (scraping off the monocytes with a rubber policeman and counting the cells in a Coulter counter). For infection, a number of L. monocytogenes equivalent to the number of monocytes determined in the representative wells was added to the remaining culture wells of one given type of treatment. For this purpose, L. monocytogenes were thawed, diluted in G-HBSS/4% FCS, sonicated to produce a uniform suspension and used for infection of the monocytes. The cell culture plates were centrifuged for 5 min at 400 g to bring the bacteria into close contact with the monocytes and then incubated at 37 C in a humidified atmosphere containing 5% CO2. Growth of L. monocytogenes was assessed by quantifying the number of bacteria that were recovered from lysed monocytes immediately after infection and 90 min post-infection. To lyse monocytes, saponin was added to the wells at a final saponin concentration of 0-2%. This saponin concentration did not interfere with L. monocytogenes viability (data not shown). The contents of the saponin-treated wells were serially diluted in ice-cold distilled water and plated to tryptic soy agar plates. CFU were assessed 24-48 h later with a colony counter. These counts were converted to CFU/ml of lysate, where I ml of lysate was the product of an equal number of monocytes in a given experiment. It should be noted that we refrained from using antibiotics as an attempt to prevent growth of extracellular bacteria. This technique has been widely used despite the possibility that antibiotics might enter the macrophages and affect the measurement of intracellular killing (Cole & Brastoff, 1975). In particular, increased pinocytosis following activation of monocytes by cytokines could seriously bias the results if antibiotics were present after the activation stimulus was given. We therefore did not use antibiotics at any stage of the
experiment.
was removed, and the wells were washed thoroughly with methanol to remove unreduced NBT dye. The wells were airdried and the formazan was solubilized by the addition of 120 pl/well 2 M KOH and 140 p1/well DMSO. Complete formazan solubilization was achieved by rapid mixing with a multi-channel pipette. Then the optical density (OD) at 690 nm was determined with a plate reader. Wells that had been incubated with NBT solution but without cells were subjected to the same procedures as the sample wells and served as blanks. A standard curve was established by reducing known quantities of NBT in KOH and DMSO. This allowed direct deduction of nmoles NBT reduced per well from OD values.
Statistical analysis All experiments were done in triplicate and repeated at least three times with cells from different donors. Results are expressed as mean values + s.d. The test for significance used in all cases was Student's t-test. RESULTS Treatment with LPS enhances anti-mycobacterial, but not antilisterial, activity of human monocytes A variety of bacterial endotoxins are capable of inducing resistance to infection upon challenge with bacteria (Blackwood, Lin & Rowe, 1987), fungi (Brummer & Stevens, 1987) and parasites (Wirth & Kierszenbaum, 1988). We addressed the question of whether in vitro treatment of human monocytes with LPS from S. typhosa (1 ng LPS/ml, 24 h) activates these cells to restrict growth of the facultative intracellular bacteria M. avium intracellulare or L. monocytogenes. As illustrated in Fig. 1 a, LPS treatment effectively reduced growth of M. avium intracellulare. In contrast to the results obtained with M. avium intracellulare, LPS-treated monocytes failed to interfere with the growth of L. monocytogenes (Fig. I b).
rTNF-a-treated human monocytes display enhanced anti-mycobacterial activity Since human monocytes and macrophages produce TNF-oa in response to LPS (Kornbluth et al., 1989), we investigated whether LPS-mediated production of TNF-ac might lead to a stimulation of the monocytes for enhanced anti-mycobacterial activity in an autocrine fashion. TNF-a was considered a likely modulator, since it has been shown to be involved in the activation of monocytes against tumour cells (Kornbluth, 12
-
(a)
10
-
(b)
0~ ~,
0 8-
6
Quantitative NBT reduction PMA-induced production of oxygen radicals was measured by reduction of NBT to insoluble formazan. The assay was performed according to Rook et al. (1985), but modified to allow for measurement of cells in suspension (Rainard, 1986). Briefly, 2-5 x 105 monocytes in 100 p1 RPMI 1640 were added to microtitre wells. The cells were then mixed with 100 p1 NBT with or without PMA to give final concentrations of 10 pg/ml PMA and I mg/ml NBT. After incubation at 37'C for 30 min the microtitre wells were centrifuged for 10 min at 120 g, medium
02 -3 Control
LPS
Control
LPS
Fig. 1. Treatment with lipopolysaccharide (LPS) enhances anti-mycobacterial but not anti-listerial activity of human monocytes. Data of representative experiments are expressed as CFU/ml+s.d. (a) after a 4-day killing assay with M. avium intracellulare-infected monocytes and (b) after a 90-min killing assay with L. monocytogenes-infected monocytes. CFU at day 0 were 1-26 x 106/ml in (a) and 5-35 x 105/ml in (b).
93
Anti-bacterial activity of human monocytes 3.5 3*0
2*5 r0
0-0 2
2-0
x
0
.
I-*5 11
0-20I
1-0 _
Control 0
10
50
100 500 1000 10000 rTNF-a (U/ml) Fig. 2. rTNF-a augments anti-mycobacterial activity of monocytes in a dose-dependent manner. Adherent monocytes were treated with the indicated concentrations of rTNF-a for 24 h, infected with M. avium intracellulare and processed for CFU determination 4 days postinfection. Data of a representative experiment are expressed as CFU/ ml+s.d. 4 days post-infection. CFU at day 0 were 2-21 x 106/ml.
Gregory & Edgington, 1988), parasites (Wirth & Kierszenbaum, 1988) and bacteria (Bermudez & Young, 1988; Bancroft et al., 1989). We therefore tested whether rTNF-a was able to replace LPS as a modulator for enhanced anti-mycobacterial (and possibly anti-listerial) activity. Adherent monocytes were exposed to various concentrations of rTNF-a by using a 24-h incubation period. Figure 2 shows that anti-mycobacterial activity increased in a dose-dependent manner, reached a maximum at 500 U/ml rTNF-a, and decreased thereafter. To confirm that the augmented anti-mycobacterial activity was attributable to TNF-a, we attempted to block the effect by using anti-TNF antibody. rTNF-a was pre-incubated for 15 min at 37°C with a 50% excess of anti-TNF-a MoAb and subsequently used for monocyte pretreatment (Bermudez & Young, 1988). Incubation of monocytes with anti-TNF antibody completely abolished the enhanced anti-mycobacterial activity of human monocytes (Fig. 3). In contrast, L. monocytogenes infection could not be interfered with by rTNF-ax. Concentrations ranging from 10 U/ ml to 10 000 U/ml did not influence the anti-listerial capacity of the treated monocytes (number of surviving Listeria in the presence of non-treated monocytes: 1 45 + 0-14 x 106 CFU/ml; in the presence of monocytes treated with: 100 U rTNF-ac, 1-48+0-15 x 106 CFU/ml; 500 U rTNF-a, 1-53+0-03x 106 CFU/ml; 1000 U rTNF-a, 1-42+0-15 x 106 CFU/ml; 104 U rTNF-ax 1-34 + 0-05 x 106 CFU/ml; n 6, differences not significant). =
Sequential application of TNF and IL-2 is required to enhance anti-listerial activity of human monocytes Cooperation of lymphokines has been shown to be involved in the induction of resistance to infection with different micro-
TNF
TN F- anti -TNF
Fig. 3. TNF-mediated activation of anti-mycobacterial activity is blocked by an antibody to TNF-a. Control, untreated monocytes; TNF: rTNF-a (100 U/ml) was added to monocytes for 24 h; TNF/anti-TNF: rTNF-a was incubated in the presence of anti-TNF-a antibody for 30 min prior to addition to the monocytes. Data of a representative experiment are expressed as CFU/ml + s.d. 4 days post-infection. CFU/ ml at day 0 were 1-9 x 106/ml.
Organisms (Belosevic et al., 1988). Whereas our results with M. avium intracellulare indicate that TNF-a on its own has an activation potential for enhanced anti-mycobacterial activity, defence against Listeria was not enhanced. However, antilisterial activity of human peripheral blood monocytes could be augmented when the cells were treated with IO U/ml rTNF-a for 24 h followed by subsequent treatment with 200 U/ml IL-2 for an additional 24 h (Table 1). This type of treatment will be referred to as rTNF-a/IL-2 treatment. Although Listeria growth was never reduced more than 30% in cytokine-treated monocytes as compared with control monocytes, the effect was highly reproducible and the difference was statistically significant
(P
ADONIS 0009910491002002
Induction of anti-mycobacterial and anti-listerial activity of human monocytes requires different activation signals G. ZERLAUTH*, M. M. EIBLt & J. W. MANNHALTER*t *Department of Immunological Research, Immuno, Vienna, and Institute of Immunology, University of Vienna, Vienna, Austria (Acceptedfor publication 23 January 1991)
SUMMARY The requirements for activation of anti-mycobacterial and anti-listerial activity of human monocytes were investigated. Human monocytes could be activated to display enhanced anti-mycobacterial activity by a 24-h treatment with lipopolysaccharide. The mediator induced by this treatment was identified as being tumour necrosis factor-alpha (TNF-a). Addition of recombinant TNF-a (rTNF-a) to the cultures of human monocytes for 24 h yielded comparable results (minimal dose required for induction of anti-mycobacterial activity, 10 U/ml). Addition of anti-TNF-oc antibody completely abrogated the effect. A similar treatment protocol failed to activate enhanced anti-listerial activity. To trigger anti-listerial activity, sequential treatment of human monocytes with rTNF-oc and IL-2 was required. Treatment of monocytes with 10 U/ml rTNF-a for 24 h followed by incubation in the presence of 200 U/ml of IL-2 for an additional 24 h yielded a reduction of listerial growth which was moderate but statistically significant (P < 0-00 1). The activation of monocytes observed with rTNFa/IL-2 treatment was (i) dependent on both cytokines; (ii) sequence dependent (i.e. when IL-2 was added prior to rTNF-a, no effect was observed); and (iii) absent in cells treated with one cytokine only. Enhancement of anti-listerial activity by sequential use of cytokines was not accompanied by an increase in oxidative burst, which indicated that oxidative mechanisms were not the reason for the observed Listeria monocytogenes growth restriction. Further support for this hypothesis was obtained after interferon-gamma treatment of human monocytes which led to an augmented PMAinducible release of active oxygen radicals, but was not paralleled by growth restriction of L. monocytogenes. Our results indicate that TNF-ca plays a crucial role in the activation of monocytes for growth restriction of intracellular microbes. Activation of human monocytes to restrict the growth of the facultative intracellular bacteria Mycobacterium avium intracellulare and L. monocytogenes, however, follows different patterns, the initial trigger in both cases being provided by TNF-a-induced signals. Keywords Listeria monocytogenes Mycobacterium avium intracellulare cytokines human monocytes, anti-bacterial activity of some facultative intracellular pathogens, these cells eventually become activated and can interfere with the growth of the intracellular microbes (Hahn & Kaufmann, 1981). Acquired resistance against Mycobacteria is mediated by specific T lymphocytes and expressed by activated macrophages (Hahn & Kaufmann, 1981). However, the observation that mice with severe combined immunodeficiency (SCID) mutation can partially reduce listerial growth strongly argues for the additional involvement of a T cell-independent mechanism in the activation of anti-bacterial activity of cells of the monocyte/ macrophage lineage (Bancroft et al., 1987). The purpose of this work was to analyse the role of cells of the monocyte/macrophage lineage in anti-bacterial activity with special emphasis on the signals required for its induction. This was achieved by activation of human monocytes with cyto-
INTRODUCTION The host immune system is a barrier that must be breached or avoided by the invading micro-organism. This necessity is even more pronounced if the micro-organisms choose monocytes or macrophages as their preferred habitat. Within these phagocytic cells, the invaders face an environment that is designed to kill micro-organisms. Micro-organisms, however, have developed different strategies to evade or withstand the lethal attack within the monocytes/macrophages. Although monocytes and macrophages are initially permissive for intracellular multiplication of Correspondence: Dr J. W. Mannhalter, Immuno AG, Department of Immunological Research, Industriestrasse 131, A- 1220 Vienna, Austria.
90
Anti-bacterial activity of human monocytes kine(s) or with endotoxin and subsequent challenge of the monocytes with different facultative intracellular microbes, Listeria monocytogenes (strain EGD) or Mycobacterium avium intracellulare (strain Boone). Since these bacteria are capable of surviving in non-activated monocytes, they provide a wellsuited system for studying activation of monocytes for enhanced anti-bacterial activity. The results presented here indicate that human monocytes can be activated in vitro to interfere with the growth of M. avium intracellulare upon treatment with recombinant tumour necrosis factor-alpha (rTNF-a). Anti-listerial activity was observed only after sequential treatment with rTNF-a plus IL-2. These results indicate that TNF-ca is an important activator of human monocytes for enhanced antibacterial activity. Furthermore, the data suggest that different activation patterns are required for induction of anti-listerial or anti-mycobacterial activity. MATERIALS AND METHODS
Materials RPMI 1640, L-glutamine and fetal calf serum (FCS) were purchased from Flow Laboratories, Irvine, UK. Interleukin-2 (IL-2) was from Biotest, Frankfurt, Germany. rTNF-a was obtained from Genzyme, Boston, MA. Anti-human TNF-a monoclonal antibody (MoAb) was from The Accurate Chemical & Scientific Corporation, New York, NY. This MoAb (IgG I subclass) reacts with human natural and recombinant TNF-c (company brochure). Phorbol myristate acetate (PMA), nitroblue tetrazolium (NBT, grade III), saponin, and interferongamma (IFN-y) were purchased from Sigma, Deisenhofen, Germany. DMSO was from Merck, Munich, Germany. Lipopolysaccharide (LPS) SW Salmonella typhosa 0901 was purchased from Biotest, Frankfurt, Germany. Phytohemagglutinin (PHA, HA16) was from Wellcome, Dartford, UK. Cells were cultivated in Costar six-well plates (Costar, Badhoevedorp, The Netherlands). 7H9 broth and 7H 11 agar plates as well as tryptic soy broth and tryptic soy agar plates were prepared from powdered media purchased from Difco Biologicals, Detroit, MI. Isolation and activation of human peripheral blood monocytes Human peripheral blood mononuclear cells (MNC) were drawn from the venous blood of healthy volunteers and isolated by buoyant density gradient centrifugation (B0yum, 1968). The MNC from the interphase were aspirated, washed with saline and suspended in RPMI 1640 supplemented with 2 mM L-glutamine and 15% pooled, heat-inactivated human AB serum (complete medium). Antibiotics were omitted from the culture media to prevent possible growth-inhibiting effects in later killing experiments (Cole & Brastoff, 1975). Monocytes were isolated by adherence to plastic as described previously (Mannhalter, Borth & Eibl, 1986). Briefly, 2 x 107 mononuclear cells in I ml of complete medium were added to each well of sixwell culture plates; after 90 min, the non-adherent cells were discarded and the adherent cells were further incubated for 24 h in 1 ml/well of complete medium at 37°C in a humidified atmosphere containing 5% CO2. After this incubation period, > 93 % of the adherent population of cells stained positive in the non-specific esterase assay (Koski, Poplack & Blaese, 1976) and viability was 95% or better as assessed by trypan blue exclusion. For activation, the adherent monolayers were cultured in the
91
presence or absence of cytokines (rTNF-a, IL-2, IFN-y) for a
given period of time. The cytokines used throughout this study were tested for their possible direct influence on survival of the bacteria. Although the cytokines were not directly in contact with the bacteria, as the treated monocytes were washed
extensively prior to addition of bacteria, we grew the bacteria in medium containing cytokine concentrations that were similar to those used for monocyte modulation. The results demonstrated that none of the cytokines had any effect on bacterial growth
(data not shown). Preparation of PHA blasts MNC were deprived ofmonocytes by three rounds of adherence to plastic. The non-adherent cells were stimulated by addition of 1 mitogenic unit of PHA/2 x 106 MNC for 48-72 h. The resulting PHA blasts were added to adherent monocytes at a ratio of five blasts/monocyte for 24 h. Bacteria and bacterial counts M. avium intracellulare strain Boone (MAIS complex 14, ATCC 35761) was grown in 7H9 broth to a concentration of approximately 1 x 108 bacteria/ml. This number of micro-organisms was generally obtained within 12-14 days. The micro-organisms were harvested, washed and frozen in liquid nitrogen. After thawing, the number of viable bacteria was checked by plating serial dilutions on 7H II agar plates and counting bacterial colonies after 2-3 weeks. During the entire incubation period the agar plates were kept in sealed plastic bags to avoid drying. Infection of adherent monocytes with M. avium intracellulare was performed according to Rook et al. (1986). Briefly, adherent monocytes were rinsed extensively with warm RPMI 1640 supplemented with 5% human AB serum. The number of adherent cells per single well of the six-well culture plate was determined by assessing the number of monocytes in representative wells of each treatment (scraping off the monocytes with a rubber policeman and counting the cells in a Coulter counter). In this context, it is important to note that the treatment of monocytes with LPS, rTNF-a, IL-2 or IFN-y had no effect on the number of monocytes recovered/well. As generally observed in an outbred population, the total number of monocytes isolated varied with the individual donors. For infection, a number of M. avium intracellulare equivalent to the number of monocytes determined in the representative well was added to the remaining culture wells of one given type of experiment. For this purpose, M. avium intracellulare were thawed, appropriately diluted in RPMI 1640/5% AB serum and sonicated to produce a uniform suspension. Adherent monocytes were incubated in the presence of M. avium intracellulare for 4 days at 37°C in a humidified atmosphere containing 5% CO2. Growth of M. avium intracellulare was assessed by quantifying the number of viable bacteria recovered from lysed monocytes at day 0 and at 4 days post-infection. For the determination of the number of surviving M. avium intracellulare strictly standardized procedures were followed. Each well of a six-well plate received saponin to yield a final saponin concentration of 0 20/0. This saponin concentration did not interfere with M. atium intracellulare viability (data not shown). After 10 min of saponin lysis, the lysing solution of each well was transferred to a plastic tube and sonicated briefly to ensure dispersion of the bacilli. The mixture was serially diluted in 10-fold steps in ice-cold distilled water and plated in triplicate to 7H 11 agar plates. CFU were
92
G. Zerlauth, M. M. Eibl & J. W. Mannhalter
assessed 12-19 days later with a colony counter (BioFoss, Foss Electric, The Chantry, UK). These counts were converted to colony-forming units (CFU)/ml of lysate, where I ml of lysate was the product of an equal number of monocytes in a given experiment. Listeria monocytogenes strain EGD (ATCC 15313) was grown in tryptic soy broth to logarithmic phase, harvested, washed and frozen at -70'C. After thawing, the number of viable bacteria was checked by plating serial dilutions on tryptic soy agar plates and counting bacterial colonies at the end of 24-48 h. The infection of adherent monocytes with L. monocytogenes was performed according to Czuprynski, Campbell & Henson (1983) with minor modifications. Briefly, adherent monocytes were rinsed extensively with warm HBSS supplemented with 0-1 % gelatin (G-HBSS) and 4% FCS. The number of adherent cells per single well of the six-well culture plate was determined by assessing the number of monocytes in representative wells of each treatment (scraping off the monocytes with a rubber policeman and counting the cells in a Coulter counter). For infection, a number of L. monocytogenes equivalent to the number of monocytes determined in the representative wells was added to the remaining culture wells of one given type of treatment. For this purpose, L. monocytogenes were thawed, diluted in G-HBSS/4% FCS, sonicated to produce a uniform suspension and used for infection of the monocytes. The cell culture plates were centrifuged for 5 min at 400 g to bring the bacteria into close contact with the monocytes and then incubated at 37 C in a humidified atmosphere containing 5% CO2. Growth of L. monocytogenes was assessed by quantifying the number of bacteria that were recovered from lysed monocytes immediately after infection and 90 min post-infection. To lyse monocytes, saponin was added to the wells at a final saponin concentration of 0-2%. This saponin concentration did not interfere with L. monocytogenes viability (data not shown). The contents of the saponin-treated wells were serially diluted in ice-cold distilled water and plated to tryptic soy agar plates. CFU were assessed 24-48 h later with a colony counter. These counts were converted to CFU/ml of lysate, where I ml of lysate was the product of an equal number of monocytes in a given experiment. It should be noted that we refrained from using antibiotics as an attempt to prevent growth of extracellular bacteria. This technique has been widely used despite the possibility that antibiotics might enter the macrophages and affect the measurement of intracellular killing (Cole & Brastoff, 1975). In particular, increased pinocytosis following activation of monocytes by cytokines could seriously bias the results if antibiotics were present after the activation stimulus was given. We therefore did not use antibiotics at any stage of the
experiment.
was removed, and the wells were washed thoroughly with methanol to remove unreduced NBT dye. The wells were airdried and the formazan was solubilized by the addition of 120 pl/well 2 M KOH and 140 p1/well DMSO. Complete formazan solubilization was achieved by rapid mixing with a multi-channel pipette. Then the optical density (OD) at 690 nm was determined with a plate reader. Wells that had been incubated with NBT solution but without cells were subjected to the same procedures as the sample wells and served as blanks. A standard curve was established by reducing known quantities of NBT in KOH and DMSO. This allowed direct deduction of nmoles NBT reduced per well from OD values.
Statistical analysis All experiments were done in triplicate and repeated at least three times with cells from different donors. Results are expressed as mean values + s.d. The test for significance used in all cases was Student's t-test. RESULTS Treatment with LPS enhances anti-mycobacterial, but not antilisterial, activity of human monocytes A variety of bacterial endotoxins are capable of inducing resistance to infection upon challenge with bacteria (Blackwood, Lin & Rowe, 1987), fungi (Brummer & Stevens, 1987) and parasites (Wirth & Kierszenbaum, 1988). We addressed the question of whether in vitro treatment of human monocytes with LPS from S. typhosa (1 ng LPS/ml, 24 h) activates these cells to restrict growth of the facultative intracellular bacteria M. avium intracellulare or L. monocytogenes. As illustrated in Fig. 1 a, LPS treatment effectively reduced growth of M. avium intracellulare. In contrast to the results obtained with M. avium intracellulare, LPS-treated monocytes failed to interfere with the growth of L. monocytogenes (Fig. I b).
rTNF-a-treated human monocytes display enhanced anti-mycobacterial activity Since human monocytes and macrophages produce TNF-oa in response to LPS (Kornbluth et al., 1989), we investigated whether LPS-mediated production of TNF-ac might lead to a stimulation of the monocytes for enhanced anti-mycobacterial activity in an autocrine fashion. TNF-a was considered a likely modulator, since it has been shown to be involved in the activation of monocytes against tumour cells (Kornbluth, 12
-
(a)
10
-
(b)
0~ ~,
0 8-
6
Quantitative NBT reduction PMA-induced production of oxygen radicals was measured by reduction of NBT to insoluble formazan. The assay was performed according to Rook et al. (1985), but modified to allow for measurement of cells in suspension (Rainard, 1986). Briefly, 2-5 x 105 monocytes in 100 p1 RPMI 1640 were added to microtitre wells. The cells were then mixed with 100 p1 NBT with or without PMA to give final concentrations of 10 pg/ml PMA and I mg/ml NBT. After incubation at 37'C for 30 min the microtitre wells were centrifuged for 10 min at 120 g, medium
02 -3 Control
LPS
Control
LPS
Fig. 1. Treatment with lipopolysaccharide (LPS) enhances anti-mycobacterial but not anti-listerial activity of human monocytes. Data of representative experiments are expressed as CFU/ml+s.d. (a) after a 4-day killing assay with M. avium intracellulare-infected monocytes and (b) after a 90-min killing assay with L. monocytogenes-infected monocytes. CFU at day 0 were 1-26 x 106/ml in (a) and 5-35 x 105/ml in (b).
93
Anti-bacterial activity of human monocytes 3.5 3*0
2*5 r0
0-0 2
2-0
x
0
.
I-*5 11
0-20I
1-0 _
Control 0
10
50
100 500 1000 10000 rTNF-a (U/ml) Fig. 2. rTNF-a augments anti-mycobacterial activity of monocytes in a dose-dependent manner. Adherent monocytes were treated with the indicated concentrations of rTNF-a for 24 h, infected with M. avium intracellulare and processed for CFU determination 4 days postinfection. Data of a representative experiment are expressed as CFU/ ml+s.d. 4 days post-infection. CFU at day 0 were 2-21 x 106/ml.
Gregory & Edgington, 1988), parasites (Wirth & Kierszenbaum, 1988) and bacteria (Bermudez & Young, 1988; Bancroft et al., 1989). We therefore tested whether rTNF-a was able to replace LPS as a modulator for enhanced anti-mycobacterial (and possibly anti-listerial) activity. Adherent monocytes were exposed to various concentrations of rTNF-a by using a 24-h incubation period. Figure 2 shows that anti-mycobacterial activity increased in a dose-dependent manner, reached a maximum at 500 U/ml rTNF-a, and decreased thereafter. To confirm that the augmented anti-mycobacterial activity was attributable to TNF-a, we attempted to block the effect by using anti-TNF antibody. rTNF-a was pre-incubated for 15 min at 37°C with a 50% excess of anti-TNF-a MoAb and subsequently used for monocyte pretreatment (Bermudez & Young, 1988). Incubation of monocytes with anti-TNF antibody completely abolished the enhanced anti-mycobacterial activity of human monocytes (Fig. 3). In contrast, L. monocytogenes infection could not be interfered with by rTNF-ax. Concentrations ranging from 10 U/ ml to 10 000 U/ml did not influence the anti-listerial capacity of the treated monocytes (number of surviving Listeria in the presence of non-treated monocytes: 1 45 + 0-14 x 106 CFU/ml; in the presence of monocytes treated with: 100 U rTNF-ac, 1-48+0-15 x 106 CFU/ml; 500 U rTNF-a, 1-53+0-03x 106 CFU/ml; 1000 U rTNF-a, 1-42+0-15 x 106 CFU/ml; 104 U rTNF-ax 1-34 + 0-05 x 106 CFU/ml; n 6, differences not significant). =
Sequential application of TNF and IL-2 is required to enhance anti-listerial activity of human monocytes Cooperation of lymphokines has been shown to be involved in the induction of resistance to infection with different micro-
TNF
TN F- anti -TNF
Fig. 3. TNF-mediated activation of anti-mycobacterial activity is blocked by an antibody to TNF-a. Control, untreated monocytes; TNF: rTNF-a (100 U/ml) was added to monocytes for 24 h; TNF/anti-TNF: rTNF-a was incubated in the presence of anti-TNF-a antibody for 30 min prior to addition to the monocytes. Data of a representative experiment are expressed as CFU/ml + s.d. 4 days post-infection. CFU/ ml at day 0 were 1-9 x 106/ml.
Organisms (Belosevic et al., 1988). Whereas our results with M. avium intracellulare indicate that TNF-a on its own has an activation potential for enhanced anti-mycobacterial activity, defence against Listeria was not enhanced. However, antilisterial activity of human peripheral blood monocytes could be augmented when the cells were treated with IO U/ml rTNF-a for 24 h followed by subsequent treatment with 200 U/ml IL-2 for an additional 24 h (Table 1). This type of treatment will be referred to as rTNF-a/IL-2 treatment. Although Listeria growth was never reduced more than 30% in cytokine-treated monocytes as compared with control monocytes, the effect was highly reproducible and the difference was statistically significant
(P
