monocytogenes by Oxidative Agents Generated by Neutrophils

Vol. 55, No. 12

INFECTION AND IMMUNITY, Dec. 1987, p. 3197-3203 0019-9567/87/123197-07$02.00/0 Copyright C) 1987, American Society for Microbiology

Relationship of Bacterial Growth Phase to Killing of Listeria monocytogenes by Oxidative Agents Generated by Neutrophils and Enzyme Systems R. BORTOLUSSI,l* C. M. J. E. VANDENBROUCKE-GRAULS,2 B. S. VAN ASBECK,2 AND J. VERHOEF2 Departments of Pediatrics and Microbiology, Dalhousie University, Halifax, Nova Scotia, Canada,' and Department of Clinical Bacteriology, University Hospital, Utrecht, The Netherlands2 Received 6 March 1987/Accepted 1 September 1987

Listeria monocytogenes, a gram-positive motile bacterium which can cause severe bacterial infection in humans, is considered to be pathogenic by virtue of its ability to resist intracellular killing. Since the mechanism of intracellular survival is poorly understood, we assessed the sensitivity of L. monocytogenes to several potent antibacterial products. Phorbol myristate acetate (PMA)-stimulated polymorphonuclear cells (PMNs) produced extracellular antibacterial products which were inhibited completely by catalase, suggesting a role for oxidative agents in this process. L. monocytogenes in logarithmic (log) growth phase resisted PMA-stimulated PMN extracellular products significantly more than L. monocytogenes in stationary (stat) growth phase or Escherichia coli (three strains) in either phase of growth. The role of oxidative agents was studied further by using xanthine oxidase-xanthine, glucose oxidase-glucose, and myeloperoxidase enzyme systems to generate hydroxyl radical (. OH), hydrogen peroxide (H202), and hypochlorous acid (OCIf), respectively. L. monocytogenes in log phase resisted the antibacterial products of these enzyme systems under conditions which produced superoxide (02-) and H202 at concentrations similar to those produced extracellularly by PMA-stimulated PMNs, while stat-growth-phase L. monocytogenes and E. coli in either phase of growth were susceptible. Antibacterial activity could be blocked or inhibited by exogenous catalase (for all oxygen radical-generating systems), mannitol, or desferoxamine (for xanthine oxidase-xanthine) and alanine (for myeloperoxidase), suggesting that OH and OCI were responsible for this activity. Log-phase L. monocytogenes had 2.5-fold higher bacteria-associated catalase activity, as compared with stat-phase L. monocytogenes. These experiments, therefore, suggest that log-phase L. monocytogenes resists oxidative antibacterial agents by producing sufficient catalase to inactivate these products. This may contribute to the ability of L. monocytogenes to survive intracellularly.

within phagocytic cells are poorly understood. The organism is readily phagocytosed (5, 38), triggers a burst of oxidative activity, and is killed by PMNs in vitro (5). The role of macrophages, however, is crucial in listeriosis in vivo since the organism appears to be viable within such cells for days (27, 29, 31, 34). In vitro studies have failed to clarify the interaction of the organism with these cells. In some studies, phagocytosis and killing appeared to occur, while in other studies, the organism survived (9, 10, 17, 33). We have assessed the sensitivity of L. monocytogenes to phorbol myristate acetate (PMA)-induced PMN extracellular products and to oxidative products of xanthine oxidase (XO)-xanthine (primarily OH), glucose oxidase (GO)glucose (primarily H202), and GO-glucose-MPO (primarily OCI-). The organism was tested at an early phase of replication (logarithmic [log] phase) and at a late phase (stationary [stat] phase) since it was thought that the phase of bacterial growth might affect its ability to survive within the phagocytic cell. In these experiments, we have found that L. monocytogenes resists oxidative radical species, particularly OH, during the log phase of growth. The importance of this and of its survival in vivo is discussed.

Phagocytic cells, including polymorphonuclear cells (PMNs), monocytes, and macrophages, kill most microorganisms within minutes of ingestion by delivering potent antibacterial products into the phagosome. Among the most powerful of these products is a group of oxidizing agents that are produced by phagocytes when they encounter microorganisms or other stimuli. The first of these to appear, superoxide (02-), is the product of univalent reduction of oxygen. Although 02 has minimal antibacterial activity, the products derived from it, hydrogen peroxide (H202) and hydroxyl radical ( OH), have potent antimicrobial activity, either alone or in combination (2, 20, 28). In contrast to monocytes and macrophages, which contain little myeloperoxidase (MPO), PMNs contain high concentrations of MPO, a heme enzyme present in the azurophilic granules of PMNs which is excreted into the phagosome-lysosome during phagocytosis (8, 15, 20). The predominant product of MPO, in the presence of H202 and halide, is hypochlorous acid (HOCI), a potent antimicrobial oxidant (2, 20). Listeria monocytogenes, the causative agent of listeriosis, is a short, motile, gram-positive rod which survives intracellularly for at least the first 2 to 4 days of infection (22, 34). To survive, facultative intracellular bacteria, such as L. monocytogenes, must avoid the lethal effects of oxygen metabolites. To do so, organisms must either resist the antimicrobial activity of such agents or fail to stimulate their production. The mechanisms which permit L. monocytogenes to survive *

MATERIALS AND METHODS Bacteria. L. monocytogenes type 4b (strain 15U), originally isolated from the blood of an infant with early-onset sepsis (12, 32), was used as a representative of virulent L. monocytogenes. This strain has been studied in detail previously (5, 18). Three strains of Escherichia coli serotypes,

Corresponding author. 3197

3198

BORTOLUSSI ET AL.

0156:K?, 01:K1, and 078:K80 (strains F8, 699, and Wy, respectively), were used as representatives of gram-negative organisms. Strains F8 and 699 were originally isolated from the stool of normal subjects, while strain Wy was isolated from the blood of a subject with sepsis. All bacteria were cultured at 37°C in Mueller-Hinton broth supplemented with 0.3% glucose (Difco Laboratories, Detroit, Mich.) for 20 h (the stat phase of growth) and for 4 h (the log phase). For the stat-phase culture, 10 p.l of an overnight broth culture of the organism was inoculated into 10 ml of fresh Mueller-Hinton broth. For the log-phase culture, 1 ml of overnight L. monocytogenes broth or 100 p.1 of E. coli broth was inoculated into 10 ml of fresh MuellerHinton broth. After centrifugation at 4,000 x g for 10 min at 4°C, the organisms were washed twice in Hanks balanced salt solution (HBSS; GIBCO, Europe POV, Breda, The Netherlands) to an approximate concentration of 108 CFU per ml, which was estimated spectrophotometrically. The actual number of CFU per milliliter was determined by the pour-plate technique with serial dilutions and molten brain heart infusion agar (Difco). Phagocytic cell isolation. PMNs and mononuclear leukocytes from healthy donors were isolated by a modification of the method developed by Boyum (6), as described elsewhere (36). Briefly, heparinized venous blood (10 U of heparin per ml of blood) was settled by gravity in 6% dextran (molecular weight, 70,000; Pharmacia, Uppsala, Sweden). The leukocyte-rich plasma was withdrawn and centrifuged at 160 x g for 10 min, and the pellet was suspended in Eagle minimal essential medium and layered on Ficoll-Paque (Pharmacia). After centrifugation for 35 min at 160 x g, the mononuclear cells and PMNs were removed separately. The PMNs were washed twice in HBSS containing 0.1% gelatin, counted, and differentiated. The final PMN pellet was adjusted to a concentration of 107 PMNs per ml (the percentage of mononuclear cells was approximately 3%). Mononuclear cells were suspended in complete RPMI 1640 (supplemented with 20 mM HEPES [N-2-hydroxyethylpiperazine-N'-2ethanesulfonic acid], 2 g of bicarbonate per liter, 100 IU of penicillin per ml, and 100 g of streptomycin per ml) with 5% fetal calf serum, and 5 x 107 cells were added to plastic petri dishes (144 by 140 mm) precoated with autologous plasma. The plates were incubated for 1 h at 37°C in a 5% C02 humidified atmosphere. Nonadherent cells were removed by gentle washing with HBSS containing 0.1% gelatin. The remaining adhering cells were removed with a rubber policeman after a 15-min incubation with ice-cold HBSS without Ca2' and Mg2' and containing 3 mM EDTA and 0.1% bovine serum albumin. Adherent cells (monocytes) were detached from the plastic, washed, and suspended in RPMI with 5% fetal calf serum. Cells were stored overnight at 4°C. For testing, cells were washed once in HBSS containing 0.1% gelatin, counted, and then suspended at a concentration of 107 viable monocytes per ml. About 80% of these cells were monocytes (90% viability as assessed by trypan blue

dye exclusion). PMA antibacterial assay. PMNs and monocytes were stimulated by PMA (Sigma Chemical Co., St. Louis, Mo.) to produce extracellular antibacterial products in an assay modified from one described by Passo and Weiss (25). Stator log-phase bacteria (25 p.l of a 10-5 dilution of stock organisms, approximately 250 CFU, suspended in HBSS), PMA solution (36 ng of PMA in 25 ,ul of HBSS), and HBSS (25 ,ul) were added to U-bottom microtiter plates. The reaction was triggered by the addition of 25 p.l of PMNs or monocytes which had been washed and suspended in HBSS

INFECT. IMMUN.

just before the assay. Studies were done at 37°C with quadruplicate wells. Microtiter plates were constantly agitated during the incubation phase (Martin Vibrators, lis, Utrecht, The Netherlands). The reaction was stopped after 60 min by the addition of 150 p.1 of ice-cold water, and the number of viable bacteria was determined by the pour-plate technique with brain heart infusion agar. Results were expressed as percent antibacterial activity by using the following formula: [(PM - PP)/PM] x 100 = AB, where PP indicates the number of CFU after a 60-min incubation in wells containing PMNs, bacteria, HBSS, and PMA solution, PM indicates the number of CFU in control wells containing PMNs, bacteria, and HBSS but no PMA, and AB is the percent antibacterial activity. In preliminary experiments, we had established that no significant reduction in CFU occurred if PMA was not included in the mixture or if heat-treated PMNs were used in otherwise complete suspensions. In some experiments, HBSS containing alanine, acetyl-L-alanine (Fluka AG, Buchs SG, Switzerland), or catalase (Sigma) were used in place of HBSS alone. XO antibacterial assay. XO (Sigma) was used to generate oxidative products, with xanthine (5 mM; Sigma) as substrate in an assay similar to one described previously (3). Xanthine was solubilized by alkalinizing (to pH 12) a suspension of 10 ml of HBSS containing 7.6 mg of xanthine and then adjusting the pH to 7.4 with 1 N HCl just before testing. Bacteria (25 p.1), the xanthine solution (25 p.l of the 5 mM solution), and HBSS (25 p.l) were added to quadruplicate U-bottom microtiter plates as described for the PMA antibacterial assay. The reaction was triggered by the addition of 25 p.l of XO (usual final concentration, 64 mU/ml) and terminated by adding 150 p.l of ice-cold water. Results were expressed as the percentage of inoculated bacteria that remained viable after 60 min of incubation at 37°C. Preliminary experiments had established that inactivated XO (56°C for 30 min) had no antibacterial activity when it was used in an otherwise complete system. In some experiments, the oxidative radical scavenger thiourea (Sigma) or the iron chelator desferoxamine (CIBA-GEIGY N.V., Arnheim, The Netherlands) was used. GO antibacterial assay. GO (Sigma) was used to generate H202 by using glucose as the substrate for the enzyme, as described by Rosen and Klebanoff (30). As with previous assays, all tests were done in quadruplicate in U-bottom microtiter plates. The final mixture contained bacteria, glucose (0.01 M), NaCl (0.1 M), Na2SO4 (0.01 M), acetate buffer (0.04 M; pH 4.5), and gelatin (0.005%). The reaction was started by the addition of 25 p.1 of GO. Results were expressed as the percentage of the inoculated bacteria that remained viable after a 60-min incubation at 37°C with agitation. Inactivated GO (boiled for 30 min) had no antibacterial activity. MPO antibacterial assay. MPO (kindly supplied by D. Roos, Amsterdam, The Netherlands) was used to study the antibacterial effects of oxidative products generated by MPO in solutions containing H202, as described by Rosen (30). GO was used as a generator of H202. The final concentration of GO (6 mU/ml) used in this assay had little antibacterial activity by itself. MPO had been prepared from human leukocytes as described by Agner (1) and modified by the method of Bakhenist et al. (4), and its activity had been determined by D. Roos (4). The enzyme was stored at -70°C at a concentration of 1.67 ,uM, thawed, and diluted in glucose-acetate buffer just before testing. The MPO antibacterial assay was performed as the other assays were, by

KILLING OF L. MONOCYTOGENES AND BACTERIAL GROWTH PHASE

VOL. 55, 1987

C-,

-J cc

LLJ C-,

m .-

100 90 80 70 60 50 40 30 20 10 -

-H

IT

A ON Log L. monocytogenes

ON Log

E. cofi FIG. 1. Extracellular antibacterial activity by PMA-stimulated PMNs. L. monocytogenes and E. coli in log (Log) and stat (ON) phases of growth were incubated in microtiter wells containing 106 PMNs and 36 ng of PMA. The antibacterial activity is expressed as a percentage and was calculated as described in the text after 60 min of incubation at 37°C. *, P < 0.001 for log- versus stat-phase L. monocytogenes. Standard errors are indicated by the vertical bars (n = 14).

using U-bottom microtiter plates inoculated in quadruplicate. Results were expressed as the percentage of the inoculated bacteria which were viable after a 30-min incubation at 37°C. GO in the absence of MPO and MPO in the absence of GO had no antibacterial activity in this system. Measurement of superoxide and H202 and catalase activity. Superoxide was measured spectrophotometrically by using a superoxide dismutase (SOD)-controlled ferricytochrome c assay (35). Catalase (75 U/ml) was added to the reaction mixture to prevent secondary reduction of ferricytochrome c by H202 and OH- (23). H202 was measured by using a phenol red colorometric assay described by Pick and Keisari (26). For the catalase assay, catalase standards or washed bacteria (they were washed in water twice and then resuspended to the original volume) were added to H202 solution (2 mM) and then left at room temperature for 15 min. The mixture was sedimented (3000 x g for 15 min), and 25 ,u1 of supernatant was used to determine the concentration of H202 remaining. Purified catalase standards had a linear relationship to H202 consumption at concentrations between 0.16 and 5 IU/ml. Catalase activity in the bacterial suspension was expressed as IU per 100 ,ug of dry weight of bacteria. RESULTS Extracellular antibacterial activity of PMA-stimulated cells. PMA-stimulated phagocytic cells produce an assortment of extracellular products which may affect the viability of microorganisms (2, 25). Several lines of evidence strongly suggest that the observed effects were due to extracellular, not intracellular, events. Studies by our group have shown that phagocytosis of L. monocytogenes 15U occurs only if serum is present (5). No serum is present in the PMA antibacterial assay. In addition, no significant antibacterial effect was observed in the present study if PMA was absent from the cell-bacteria mixture. The effect of PMA-stimulated PMNs or monocytes was therefore attributed to the release of antibacterial products from these cells into the superna-

3199

tant. When equal numbers of monocytes and PMNs were compared, we found a significantly higher antibacterial activity (P < 0.02 by the two-sided Student t test) with PMNs. For both cell preparations, antibacterial activity was greatest with organisms in the stat phase of growth. Antibacterial activity for monocytes was 5 + 8 and 9 ± 3 (mean ± the standard error of the mean) for log- and stat-phase organisms, respectively (P > 0.05, n = 4). The antibacterial activity was studied in detail with PMN preparations. The antibacterial activity of PMNs for L. monocytogenes in the log phase of growth was significantly less than for L. monocytogenes in the stat phase (11.9 ± 7.2 versus 55.9 ± 3.5 [mean ± the standard error of the mean]; P < 0.001, n = 14, by the two-sided Student t test) (Fig. 1). In contrast, no significant difference was found between phases of growth when three E. coli strains were used. Antibacterial activity of PMA-stimulated PMNs against L. monocytogenes was significantly less (P < 0.001) than such activity for log-phase E. coli. The antibacterial activity of PMA-stimulated PMNs could be almost completely blocked by catalase and partially blocked by SOD and alanine (Table 1). The addition of NaI enhanced the antibacterial activity in this system. Because products which affect oxidative mechanisms for the killing of bacteria had a marked effect on the antibacterial activity in our system, we measured the production of H202 and superoxide by PMA-stimulated PMNs and monocytes. PMNs produced 1.5 M H202 per min per 106 PMNs, and monocytes produced 0.15 M H202 per min per 106 cells, as calculated over the first 15 min of PMA-stimulated activity. GO antibacterial activity. The antibacterial activity of GO-generated H202 was assessed for stat- and log-phase L. monocytogenes over a wide range of GO concentrations. In this system, H202 production ranged from 0.95 to 18 M H202 per min (for 6 and 750 mU of GO, respectively). Viable L. monocytogenes were decreased at concentrations of GO over 30 mU/ml (Fig. 2). H202 production at these concentrations was 7.0 ,uM/min or more. The antibacterial effects of GO-generated H202 on log- and stat-phase L. monocytogenes were similar over a wide range of GO concentrations. Since 6 mU of GO per min produced H202 at a rate similar to that produced by PMA-stimulated PMNs and monocytes,

(b)

(a) 100

-J 2

-J C.) 0

zD0-'

z

.O 0

6

30

mU GO/ml

150

750

LM

E. co/i 6 mU GO/ml

FIG. 2. GO antibacterial activity. (a) L. monocytogenes in the log (Log) or stat (ON) phase of growth was incubated in microtiter wells containing various concentrations of GO. The percentage of the inoculum which was viable after 60 min was calculated. (b) L. monocytogenes (LM) and E. coli strains in log (L) and stat ( Vii) phases of growth were incubated in microtiter wells containing 6 mU of GO per ml. Greater than 70% viability was observed at this concentration for both bacteria. Standard errors are indicated by the vertical bars (n = 6 to 10).

BORTOLUSSI ET AL.

3200

INFECT. IMMUN.

TABLE 1. Effect of various inhibitors and stimulators on antibactericidal activity CFU (% inhibition or stimulationa) with the following assay: PMAb God MPOC

Modifying agent (concn)

++

+

Control Catalase (800 U/ml) SOD (125 U/ml) Thiourea (25 mM) Desferoxamine (25 mM) Alanine (1 mM) Acetyl-L-alanine (1 mM) Nal (1 ,uM)

xOe

350 745 (90) 404 (25)

795

NDf ND 540 (46) 810 (0) 125 (+ 140)

282 700 (100) ND ND ND 450 (41) 270 (0) 140 (+ 135)

695

210 520 (100) 200 (0) 215 (0) 180 (0) ND ND ND

510

35 150 (100) 40 (5) 83 (59) 97 (75) ND ND ND

117

a Inhibition was calculated by using the following formula: [1 - (ABi/ABc)] x 100, where AB, is the antibacterial activity in the presence of an inhibitor and ABc is the antibacterial activity in the presence of a control solution. Stimulation was calculated by using the following formula: ABs/ABc x 100, where ABs is the antibacterial activity in the presence of the stimulator (Nal) and ABc is the antibacterial activity in the absence of Nal. b PMA antibacterial assay, as described in the text, using 36 ng of PMA (+) or no PMA (-) and 2.5 x 105 PMNs in a final volume of 100 ,ul. c MPO antibacterial assay, as described in the text, using 0.6 mU of GO in a final volume of 100 p.1 containing 0.03 nM MPO (+) or no MPO (-). d GO antibacterial assay, as described in the text, using 3.0 mU of GO (+) or no GO (-) in a final volume of 100 pd. e XO antibacterial assay, as described in the text, using 6.4 mU of XO (±) or no XO (-) in a final volume of 100 ,ul.

f ND, Not done.

we selected this concentration of GO to compare its antibacterial effect on log- and stat-phase L. monocytogenes and E. coli. There was little antibacterial effect on either L. monocytogenes or E. coli (Fig. 2). In addition, the phase of growth made little difference in this system. Antibacterial activity of GO-generated H202 was completely blocked by catalase and unaffected by SOD (Table 1). MPO antibacterial activity. As described above, GO at concentrations of 6 mU/ml produced concentrations of H202 that were minimally antibacterial and were similar to those generated by PMA-stimulated PMNs. When purified MPO was added to this mixture, a marked enhancement of antibacterial activity was found for all strains tested. Complete antibacterial activity against L. monocytogenes was seen at 0.17 nM or higher MPO, while no effect was seen at 0.006 nM (Fig. 3). At an MPO concentration of 0.03 nM, there was little effect on log-phase L. monocytogenes (78.0% survival) but more of an antibacterial effect on stat-phase organisms (40% survival; P < 0.0002 versus that of log-phase organisms (b)

(a) 100 80 -J C.,

60 -

0 z

z

40

40

-

a-e

a-

ON "

20-

. .r

.006

.03 nM MPO

.17 .85

LM E. coll .03 nM MPO monocytogenes in the

FIG. 3. MPO antibacterial activity. (a) L. log (Log) or stat (ON) phase of growth was incubated in microtiter wells containing various concentrations of MPO in the presence of 6 mU of GO per ml. (b) L. monocytogenes (LM) and E. coli strains in log (Oii) and stat ( ED ) phases of growth were incubated in microtiter wells containing 6.0 mU of GO per ml and 0.03 nM MPO. Viability, expressed as a percentage of the inoculum, was determined after 30 min. *, P < 0.0002 for log- versus stat-phase L. monocytogenes. Standard errors are indicated by the vertical bars (n = 6 to 8).

by the two-sided Student t test). At this concentration of MPO, all three E. coli strains were markedly affected (percent survival, 31, 0.5, and 0 for the log phase and 23, 2, and 1 for the stat phase for strains F8, WY, and 699, respectively). There was no significant difference in susceptibility between log- or stat-phase E. coli (P > 0.05). In another experiment with E. coli F8, the antibacterial activity of 0.03 nM MPO was markedly increased at a higher concentration of GO (30 mU/ml, to 3% survival from 20%) and decreased at a lower concentration of GO (1.5 mU/ml, 100% survival), showing that the effects of the products of MPO-H202 increased with increasing concentrations of H202. Catalase and alanine, but not acetyl-L-alanine, had inhibitory effects on MPO antibacterial activity, while NaI increased its effects (Table 1). These observations support the interpretation that a hypohalide anion is the product which is antibacterial for the following reasons: (i) its activity is directly related to the concentration of H202; (ii) its activity is inhibited by catalase (which destroys H202) and alanine (which scavenges OCl-); and (iii) it is enhanced in the presence of iodine, allowing OI- formation (01- is a hypohalide product which is more potent than OCI-). XO antibacterial activity. XO-xanthine products had markedly different effects on log- and stat-phase L. monocytogenes (Fig. 4). At 64 mU of XO per ml, 94% survival for log-phase L. monocytogenes was seen versus 34% for statphase organisms (P < 0.000002). Since we found that PMAstimulated PMNs generated 25.6 nM superoxide per min per 106 PMNs and 64 mU of XO generated a similar concentration (20 nM/min), we used this concentration of XO to compare L. monocytogenes with three E. coli strains. The oxidative products of PMA-stimulated PMNs and of XOxanthine are different since XO-xanthine results in the formation of 02-, H202, and OH, while PMA-stimulated PMNs also produce hypohalide (20, 28, 30, 39). At this concentration of XO, E. coli strains in log and stat phases were susceptible to the oxidative products (51, 65, and 43% survival for the log phase and 41, 31, and 45% for the stat phase for strains F8, Wy, and 699, respectively). In contrast to previous PMA, GO, and MPO antibacterial assays in which no difference was seen between log- and stat-phase E. coli, there was a difference between phases in the XO antibacterial assay (53 + 10% overall survival for log-phase E. coli and 39 + 6.5% for stat-phase E. coli [P < 0.02]). -

Log-phase E. coli was more susceptib ile than log-phase L. monocytogenes (53 versus 94% survivial, respectively [P < 0.00002]). Stat-phase L. monocytogene.s was similar to statphase E. coli in susceptibility to XO'-xanthine. Catalase, thiourea, and, to a small degree, SOD iinhibited the antibacterial effect of the XO reaction (Table 1). Catalase activity. L. monocytogenes Eand E. coli in log and stat phases of growth were compare d for the ability to consume H202 to assess the catalase .activity of these two organisms. Catalase activity was lower ifor both organisms in the stat phase of growth, with 4.5 anid 4.0 U/100 pug dry weight of L. monocytogenes and E. colii, respectively (Table 2). Catalase activity was highest for L. nnonocytogenes in the log phase of growth, almost twofold higher than that for log-phase E. coli and 2.6-fold higher th,an that for stat-phase L. monocytogenes. DISCUSSION L. monocytogenes has long been reg arie asanintraco lular pathogen because of its ability to s within phagocytic cells in vivo. The mechanisms Steigbigel been poorly understood with in vitro t4 et al. (33) in 1974 found no difference between L. monocytogenes and other bacterial species, inc]luding E. coli, Staphylococcus aureus, and Salmonella ty,phimurium in their abilities to survive within phagocytic c ells, both PMNs and monocytes. Similarly, in 1983 Czupry,nski et al. (9) found that L. monocytogenes was killed as effiiciently as E. coliand S. typhimurium by human mononucle ar cell suspensions Glass-adherent mononuclear cells (moinocytes) reduced the number of viable cell-associated L. moonocytogenes dunng the first 2 h of incubation, but continLued culturing of the infected monocytes resulted in the repl[ication of the surviving L. monocytogenes. Both groups organisms cultured overnight for their ststudiest In contrast Harrington-Fowler et al. (16) found tha loguphase L monot cytogenes resisted the antibacterial a.ctivity of peritoneal macrophages, as compared with S. alireus. Stat-phase organisms were not studied in these exp(eriments. Several lines of evidence have sugggested that oxidative L of -L. antibacterial factors play a role in the virulenceoxidae monocytogenes: (i) Catalase-negative Listeria strains are

purvive chniquest

esthe thatlen

(b)

(a) 100

TI ' Log

80

a

-i

g

8

6008 60 z

60-

0

40-

o

12.8

3201


VOL. 55, 1987

64

320

1600

LM

E. coi

64 mU XO/mI FIG. 4. XO antibacterial activity. (a) L. monocytogenes in the log (Log) or stat (ON) phase of growth was incubated in microtiter wells containing various concentrations of:XO in the presence of 5 mu

XO/ml

mM xanthine. (b) L. monocytogenes (LM) (O) and stat ( vTLDI) phases of growth were incubated in microtiter wells containing 64 mU of XO per ml and 0 b.5 mM xanthine. *, P < 0.00002 for log- versus stat-phase L. momnocytogenes. Standard errors are indicated by the vertical bars (n = 6).

TABLE 2. Catalase activity of L. monocytogenes and E. coli Catalase Organism Growth phase [dry (U/100 1Lg Organism

wt])

L. monocytogenes

E. coli

Log

11.5

Stat

4.5

Log

6.3

Stat

4.0

less virulent in vivo than most catalase-positive strains (37); (ii) PMNs, which produce more oxidative radicals than monocytes do, have more antilisterial activity than monocytes (10, 14); and (iii) L. monocytogenes-immune, antigenelicited peritoneal macrophages produce more oxidative products and are also more antilisterial than Listeria antigenelicited but nonimmune macrophages (10, 14). However, it is argued that nonoxidative mechanisms are also important, since inhibitors of oxidative antimicrobial systems fail to completely alter the killing of Listeria spp. within phagocytic cells (9, 14) and because there is a poor correlation between antilisterial activity and the peroxidase content of peritoneal macrophages (16). In these latter studies, oxidative products were not assessed directly. The lack of any effect of inhibitors of oxidative products in two of these studies may be explained by the failure of such inhibitors to penetrate the cell. The results of the present experiment suggest that the survival of L. monocytogenes within phagocytic cells is facilitated by its ability to resist oxidative products. The failure of earlier studies to demonstrate this effect may have been due to differences in technique, since only organisms in the log phase of growth clearly demonstrate a resistance to oxdtvprucs In our first experiment, PMA-stimulated PMNs were found to produce extracellular factors that were antibacterial to L monocytogenes in the stat phase of growth and to E. coli in stat and log phases but were less antibacterial for log-phase Listeria. This effect was independent of phagocytosis, which requires serum (5, 9, 10). Oxidative products were clearly implicated as the antibacterial agents for this system since antibacterial activity was inhibited with catalase (H202 inactivation) and alanine (which scavenges OCI-). The enhancement of antibacterial activity by the addition of Nal to the medium is also consistent with this interpretation, since the product of H202 and MPO in the presence of iodine (OI-) has more antibacterial activity than OCl (20, 25). The oxidative radicals implicated in the oxidationmediated antimicrobial system were studied further by using cell-free enzyme systems at concentrations of enzyme (GO and XO) which simulated the superoxide and H202 production by PMA-stimulated PMNs. GO, which produces H202 only, had little antibacterial activity. In addition, E. coli did not differ from L. monocytogenes in this assay. Thus, the effects of PMA-stimulated PMNs could not be explained by inhibition by H202 by itself but, in light of the complete When purified catalase, were likely due to products of H202. MPO was added to the GO reaction medium, a marked enhancement of the antibacterial activity was seen which was dependent on the concentration of both MPO and GO-generated H202. Catalase and alanine inhibited this reaction, while NaI increased its effects, observations that support the interpretation that a hypohalide anion is antibacterial for L. monocytogenes. In this system, log-phase L. monocytogenes was more resistant than stat-phase L. mono-

3202

INFECT. IMMUN.

BORTOLUSSI ET AL.

cytogenes or E. coli in either phase of growth. Since the presence of MPO within phagosomes depends on the fusion of MPO-containing primary granules to the phagosome, it is likely that antilisterial activity would decrease if such fusion failed to occur, as suggested by de Heer et al. (11), or if the MPO concentration was low. Monocytes, especially macrophages, have low concentrations of MPO and less H202generating capacity than PMNs (2, 20). The observation that monocytes and macrophages have poor antilisterial activity, as compared with PMNs, therefore supports the interpretation of a role for MPO in PMNs. Compared with nonactivated macrophages, activated cells have greater H202- and OH-generating' capacity (13, 19). Such cells appear to have increased antilisterial activity as well. Oxidation-dependent antimicrobial activity was assessed further by using XO which, in the presence of xanthine, produces superoxide, H202, and OH (3). As described above, H202 and superoxide had little antilisterial activity, but in the presence of ferric ion, they interact by the Haber-Weiss reaction (2, 3, 20), producing OH, an oxidant which is antibacterial for a number of microbial species (20). OH is produced by phagocytic cells during the oxidative burst induced by'phagocytosis or by PMA stimulation (39). Since its production is independent of MPO, it provides an alternate antimicrobial mechanism for cells with low MPO concentrations or from individuals with MPO deficiency. XO-xanthine antibacterial activity was demonstrated for L. monocytogenes in stat phase and for E. coli in log or stat phase but was poor for log-phase L. monocytogenes over a wide range of enzyme concentrations. That the antimicrobial activity was due to OH is suggested by the inhibition with both catalase, which removes the substrate for OH production, and thiourea, a scavenger of OH. SOD lowered antilisterial activity only slightly. Although OH is not produced if superoxide is absent (due to SOD), the concentration of H202 (the product of SOD) is increased. We speculate that increased H202 compensated for lower concentrations of OH when SOD was present, resulting in a negligible decrease in overall antilisterial activity. Thus, L. monocytogenes in the log phase of growth resists at least two oxidative products, OCl- and OH, at concentrations which are potentially present in phagocytic cells. We showed that catalase is present in higher concentrations in log-phase L. monocytogenes as compared with stat-phase L. monocytogenes and log- or stat-phase E. coli. Indeed, catalase has already been implicated as a virulence factor for S. aureus (21) with high- and low-catalase-producing strains. The relevance of these observations to Listeria infection in vivo is unknown. We speculate that natural infection involves organisms in both log and stat phases of growth. Such organisms are readily phagocytosed by PMNs, monocytes, and macrophages resident in the liver and spleen. Those organisms in the stat phase of growth may be quickly killed by all three phagocytic cells since they are susceptible to concentrations of OH and 0Cl- which can be easily achieved. Organisms in the log phase of growth are also taken up by phagocytic cells but are more likely to resist intracellular killing since the ability of some phagocytic cells (macrophages) to produce OCl- and OH is limited. The organism may thus survive and replicate in these cells. In the normal host, intracellular L. monocytogenes are killed 3 to 4 days after initial exposure to the organism. At this time, production of alpha/beta and gamma interferon are markedly increased (17). Endogenous interferon appears to be essential for the resolution of L. monocytogenes infection in such animals (7). Considerably augmented production of oxygen

species ( OH, 02-, and H202) occurs in monocytes in response to alpha/beta interferon in the murine model (19) and to gamma interferon in humans (24). Such oxygen species may then be sufficient by themselves or with other products (3, 13) to kill the organism. Although this description of intracellular Listeria pathogenesis is speculative, it is consistent with several independent in vitro and in vivo observations. It thus warrants further study and may provide insight into the pathogenesis of other facultative intracellular

microorganisms. ACKNOWLEDGMENTS We acknowledge the assistance of Mary Joyce and Paula lantorno in preparing this manuscript. This research was made possible by grant MT 7610 from the Medical Research Council of Canada. R. Bortolussi was supported by Medical Research Council development grant DG 208. LITERATURE CITED 1. Agner, K. 1958. Crystalline myeloperoxidase. Acta Chem. Scand. 12:89-94. 2. Babior, B. 1984. The respiratory burst of phagocytes. J. Clin.

Invest. 73:599-601. 3. Babior, B. M., J. T. Curnutte, and R. S. Kipnes. 1975. Biological defense mechanisms. Evidence for the participation of superoxide in bacterial killing by xanthine oxidase. J. Lab. Clin. Med. 85:235-244. 4. Bakhenist, A. R. J., R. Wever, T. Vulsma, H. Plat, and B. F. Van Gelder. 1978. Isolation procedure and some properties of myeloperoxidase from human leukocytes. Biochim. Biophys. Acta

5Z4:45-54. 5. Bortolussi, R., A. Issekutz, and G. Faulkner. 1986. Opsonization of Listeria monocytogenes type 4b by human adult and newborn sera. Infect. Immun. 52:493-498. 6. Boyum, A. 1968. Isolation of mononuclear cells and granulocytes from human blood. Scand. J. Clin. Lab. Invest. 21(Suppl.

97):77-89. 7. Buchmeier, N. A., and R. D. Schreiber. 1985. Requirement of endogenous interferon-gamma production for resolution of Listeria monocytogenes infection. Proc. Natl. Acad. Sci. USA 82:7404-7408. 8. Capo, C., C. Farnarier, A.-M. Benoliel, P. Bongrand, and R. Depieds. 1983. Dissociation between phagocytosis and phagosome-lysosome fusion. RES J. Reticuloendothel. Soc. 34:359369. 9. Czuprynski, C. J., P. A. Campbell, and P. M. Henson. 1983. Killing of Listeria monocytogenes by human neutrophils and monocytes, but not by monocyte-derived macrophages. RES J. Reticuloendothel. Soc. 34:29-44. 10. Davies, W. A. 1982. Kinetics of killing Listeria monocytogenes by macrophages: correlation of 3H-DNA release from labeled bacteria and changes in numbers of viable organisms by mathematical model. RES J. Reticuloendothel. Soc. 32:461-476. 11. de Heer, E., M. C. Kersten, C. Van der Meer, W. A. Linnemans, and J. M. Willers. 1980. Electron microscopic observations on the interaction of Listeria monocytogenes and peritoneal macrophages of normal mice. Lab. Invest. 43:449-454. 12. Evans, J. R., A. C. Allen, D. A. Stinson, R. Bortolussi, and L. J. Peddle. 1985. Perinatal listeriosis: report of an outbreak. Pediatr. Infect. Dis. 4:237-241. 13. Freund, M., and E. Pick. 1985. The mechanism of action of lymphokines. VIII. Lymphokine-enhanced spontaneous hydrogen peroxide production by macrophages. Immunology 54:3539. 14. Godfrey, R. W., and M. S. Wilder. 1984. Relationships between oxidative metabolism, macrophage activation, and antilisterial activity. J. Leukocyte Biol. 36:533-543. 15. Goren, M. B., C. L. Swendsen, J. Fiscus, and C. Miranti. 1984. Fluorescent markers for studying phagosome-lysosome fusion. J. Leukocyte Biol. 36:273-292.

VOL. 55, 1987


16. Harrington-Fowler, L., P. M. Henson, and M. S. Wilder. 1981. Fate of Listeria monocytogenes in resident and activated macrophages. Infect. Immun. 33:11-20. 17. Havell, E. A. 1986. Augmented induction of interferons during Listeria monocytogenes infection. J. Infect. Dis. 153:960-969. 18. Hawkins, A. E., R. Bortolussi, and A. C. Issekutz. 1984. In vitro and in vivo activity of various antibiotics against Listeria monocytogenes type 4b. Clin. Invest. Med. 7:335-341. 19. Ito, M., R. Karmali, and M. Krim. 1985. Effect of interferon on chemiluminescence and hydroxyl radical production in murine macrophage stimulated by PMA. Immunology 56:533-541. 20. Klebanoff, S. J. 1980. Oxygen metabolism and the toxic properties of phagocytes. Ann. Intern. Med. 93:480-489. 21. Mandel, G. L. 1975. Catalase, superoxide dismutase and virulence of Staphylococcus aureus; in vitro and in vivo studies with emphasis on staphylococcal leukocyte interaction. J. Clin. Invest. 55:561-566. 22. Mandel, T. E., and C. Cheers. 1980. Resistance and susceptibility of mice to bacterial infection: histopathology of listeriosis in resistant and susceptible strains. Infect. Immun. 30:851861. 23. Nasrallah, V. N. J., P. S. Shirley, Q. Myrvik, and M. Waite. 1983. The use of acetylated cytochrome-c in detecting superoxide anion. J. Immunol. 131:2104-2106. 24. Nathan, C. F., C. R. Horowitz, J. de la Harpe, S. Vadhan-Raj, S. A. Sherwin, H. F. Oettgen, and S. E. Krown. 1985. Administration of recombinant interferon-gamma to cancer patients enhances monocyte secretion of hydrogen peroxide. Proc. Natl. Acad. Sci. USA 82:8686-8690. 25. Passo, S. A., and S. J. Weiss. 1984. Oxidative mechanisms utilized by human neutrophils to destroy Escherichia coli. Blood 63:1361-1368. 26. Pick, E., and Y. Keisari. 1980. A simple colorimetric method for measurement of hydrogen peroxide produced by cells in culture. J. Immunol. Methods 38:161-170. 27. Pietrangeli, C. E., E. Skamene, P. J. Edelson, and P. A. L. Kongshavn. 1981. Measurement of 5'-nucleotidase in mouse peritoneal macrophages in listeriosis. Infect. Immun. 32:1206-1210. 28. Repine, J. E., J. W. Eaton, M. W. Anders, J. R. Hoidal, and R. B. Fox. 1979. Generation of hydroxyl radical by enzymes, chemicals, and human phagocytes in vitro. J. Clin. Invest. 64:1642-1651.

3203

29. Rhodes, J. M., J. Bennedsen, S. Olesen Larsen, S. Riisgaard, and J. V. Sparck. 1979. Correlation between in vivo and in vitro functional tests for activated macrophages. Infect. Immun. 23: 34-40. 30. Rosen, H., and S. J. Klebanoff. 1985. Oxidation of microbial iron-sulfur centers by the myeloperoxidase-H202-halide antimicrobial system. Infect. Immun. 47:613-618. 31. Sato, K. 1984. Enhancement of host resistance against Listeria infection by Lactobacillus casei: role of macrophages. Infect. Immun. 44:445-451. 32. Schlech, W. F., III, P. M. Lavigne, R. A. Bortolussi, A. C. Allen, E. V. Haldane, A. J. Wort, A. W. Hightower, S. E. Johnson, S. H. King, E. S. Nicholls, and C. V. Broome. 1983. Epidemic listeriosis-evidence for transmission by food. N. Engl. J. Med. 308:203-206. 33. Steigbigel, R. T., L. H. Lambert, Jr., and J. S. Remington. 1974. Phagocytic and antibacterial properties of normal human monocytes. J. Clin. Invest. 53:131-142. 34. Tatsukawa, K., M. Mitsuyama, K. Takeya, and K. Nomoto. 1979. Differing contribution of polymorphonuclear cells and macrophages to protection of mice against Listeria moonocytogenes and Pseudomonas aeruginosa. J. Gen. Microbiol. 115: 161-166. 35. Tauber, A. I. 1985. Phagocyte NADPH-oxidase, p. 25-30. In R. A. Greenwald (ed.), Handbook of methods for oxygen radical research. CRC Press, Inc., Boca Raton, Fla. 36. Verhoef, J., P. K. Peterson, and P. G. Quie. 1977. Kinetics of staphylococcal opsonization, attachment, ingestion and killing by human polymorphonuclear leukocytes: a quantitative assay using 3H-thymidine labeled bacteria. J. Immunol. Methods 14: 303-311. 37. Welch, D. F., C. P. Sword, S. Brehm, and D. Dusanic. 1979. Relationship between superoxide dismutase and pathogenic mechanisms of Listeria monocytogenes. Infect. Immun. 23:863-

872. 38. Wilson, C. B., V. Tsai, and J. S. Remington. 1980. Failure to trigger the oxidative metabolic burst by normal macrophages: possible mechanism for survival of intracellular pathogens. J. Exp. Med. 151:328-346. 39. Winterbourn, C. C. 1986. Myeloperoxidase as an effective inhibitor of hydroxyl radical production: implications for the oxidative reactions of neutrophils. J. Clin. Invest. 78:545-550.