JOURNAL OF BIOLUMINESCENCE AND CHEMILUMINESCENCE VOL 7,109-1 16 (1992)
Differences in t h e Respiratory Burst of Hum a n Polymorphonuclea r Leukocytes Induced by Virulent and Avirulent Legionella pneumophila Serogroup 1t Premavathy Rajagopalan-Levassew’*, Eric Dournon*, Jean-Louis Vilde’ and Jean-Jacques Pocidalo’ l Institut National de la Sant6 et de la Recherche Medicale, U 13, HBpital Claude Bernard, Paris VII, Universit6 Paris, France
* Hapital Raymond Poincare,
Two strains o f Legionella pneumophila of different virulence were examined f o r their influence on the metabolic oxidative activity of human polymorphonuclear leukocytes. The leukocytes exhibited decreased rates of oxygen consumption and diminished chemiluminescence activity following phagocytosis of a virulent strain o f L. pneumophila serogroup 1. In contrast, phagocytosis o f its multipassaged derivative rendered avirulent, was accompanied by increased rates of both oxygen consumption and chemiluminescence activity. Although no differences were observed in oxygen uptake induced by the virulent legionellae compared t o leukocytes a t rest, statistically significant differences were observed i n the chemiluminescence responses. These observations were n o t unexpected, since the luminol-enhanced chemiluminescence assay, is more sensitive than the oxygen uptake assay. In spite of decreased metabolic activity of PMN in the presence of virulent legionellae, electron microscope studies showed higher numbers of intracellular L. pneumophila than t he avirulent subtype. Thus, virulent and avirulent L. pneumophila can be differentiated on the basis o f oxygen consumption and chemiluminescence assays.
Keywords: Legionella pneumophila serogroup 1 ; virulence; chemiluminescence; oxygen consumption; polymorphonuclear leukocytes
Legionella pneumophilu serogroup 1, the etiologic agent of legionnaires’ disease (LD), is a facultative intracellular gram-negative bacillus. L. pneumophila-mononuclear phagocyte interactions have been extensively characterized by Horwitz and cowork-
ers (Horwitz., 1988). However, L. pneumophilapolymorphonuclear leukocyte (PMN) interactions remain to be elucidated. It has been reported that virulent legionellae are not killed following ingestion by PMN (Horwitz and Silverstein, 198 1) and that survival within PMN may be promoted by a cytotoxin (Friedman et al., 1982). The role of this
work was presented in part at the 23rd Interscience Conference on Antimicrobial Agents and Chemotherapy, October 1983, Las Vegas, USA. * Author for correspondence. 0884-3996/92/020109-08$05.00 6 1992 by John Wiley & Sons, Ltd.
Received 20 February 1991 Revised 5 November 1991
R. LEVASSEUR, E. DOURNON, J.-L. VlLDE AND J.-J. POCIDALO
toxin in the pathogenesis of LD is not completely understood, although it has been implicated in blocking the production of superoxide anions by PMN (Lochner et al., 1985; Saha et al., 1985). Phagocytosis and the bactericidal activity of PMN are classically associated with (i) and enhanced rate of oxygen consumption and an increase in hexose monophosphate shunt activity (Miller et al., 1972), and (ii) the production of highly reactive oxygen radicals (Welch, 1980). These radicals generated by phagocytic cells play an important role in the killing of bacteria and parasites (Gabig and Babior, 1981). To characterize the metabolic events associated with virulence of L. pneumophila, we investigated the oxidative metabolic activity of normal human PMN after exposure in uitro to virulent and avirulent L. pneumophila using oxygen consumption and chemiluminescence assays as tools. In addition, morphological observations using immunofluorescence and electron microscopy were used to determine whether differences exist in the ingestion of virulent and avirulent L. pneumophila.
BCYE plates at 35°C to mid-logarithmic phase. The bacteria were harvested in sterile distilled water and tested for viability and the absence of contaminating bacteria. The suspension was adjusted to lo9 bacteria per millilitre, divided into aliquots and stored at - 70°C until use. The LD,, was > 2 x lo9 bacteria per millilitre.
Killed bacteria were prepared by treating the above virulent and avirulent L. pneumophila for 45 min at 100°C in a water bath. The suspensions were then divided into aliquots and stored at - 70°C until use.
Sera from five healthy donors free of anti-L. pneumophila antibodies were pooled, filtered, and stored in aliquots at - 20°C until use.
MATERIALS AND METHODS Bacteria
L.pneumophila serogroup 1 (strain Paris CB 81-13) was isolated from the lung of a patient who died from legionnaires’ disease during a nosocomial outbreak in Paris, France. Crude lung homogenate was plated on buffered charcoal yeast extract agar supplemented with a-ketoglutarate (BCYE) and antibiotics. Plates were incubated at 35°C with 2.5% C 0 2 and 95% humidity. A single colony was taken after 72 h and plated onto BCYE plates, and bacteria were grown at 35°C to mid-logarithmic phase. The bacteria were harvested in sterile distilled water and tested for viability and the absence of contaminating bacteria. The suspension was then adjusted to lo9 bacteria per millilitre, divided into aliquots, and stored at - 70°C until use. The virulence of this strain, passaged twice in BCYE agar, was assessed by intraperitoneal infection of guinea pigs, as previously described (Dournon et al., 1986). The LD,, was 2 x lo5 bacteria per millilitre. The avirulent subtype was obtained by 76 passages of the virulent strain on BCYE over a sixmonth period (McDade and Shepard, 1979). A single colony from the 76th passage was grown on
Immune serum was obtained from a patient who had recently recovered from legionnaires’ disease and who was no longer receiving medication. The anti-l. pneumophila antibody titre, measured by indirect immunofluorescence, was 1/512. Immediately after collection, the serum was filtered (Millipore Corp., Bedford, MA) and stored in aliquots at - 70°C until use.
Zymosan (Sigma Chemical Co., St Louis, MI) was used as a control stimulant. It was suspended at 10 mg/ml in phosphate buffered saline (PBS, pH 7.2; 0.05 M with CaZ and Mg2 ions), washed twice in PBS, then homogenized in a potter homogenizer. The homogenized zymosan was opsonized with 50% pooled normal human serum for 30 min at 37°C. After 30min, it was rewashed twice with PBS, divided into aliquots and stored at -70°C until use. +
L. PNEUMOPHILA AND RESPIRATORY BURST OF HUMAN LEUCOCYTES
was initiated by the introduction of 200 to 250 pl of either (i) live or heat-killed virulent or avirulent L. Luminol (Sigma Chemical Co., St Louis, MI) was pneumophilu at a bacteria to PMN ratio of 10: 1, or initially dissolved in 0.2 ml of dimethyl sulphoxide (ii) opsonized zymosan (10% w/v). The final vol(Prolabo, RhGne-Poulenc, France) and further di- ume of the reaction mixture was between 2 and luted in PBS. Portions of 10T3M luminol in PBS 2.5ml. Oxygen consumption by the PMN was were stored at - 20°C until use. Working dilutions further recorded for another 15 min until a linear were made in PBS. slope was obtained. The results were expressed as a stimulation index: Luminol
Isolation of PMN
PMN were separated from blood using Mono-Poly Resolving Medium (Flow Laboratories Ltd, Irvine, Scotland, UK). This medium enables the resolution of both mononuclear and PMN leukocytes into two distinct bands in a single step. Both bands are free of contaminating erythrocytes (Ferrante and Thong, 1980). Briefly, 20 to 30ml of heparinized venous blood was obtained from healthy donors seronegative for L. pneumophila antibodies and used within two hours of collection. Whole blood (3ml) was carefully layered over 3.5 ml of the gradient in a 13 x 100 mm sterile tube and centrifuged at 300 x g for 30min in a swinging bucket rotor at room temperature. Two well-separated bands of mononuclear cells (upper band) and PMN (lower band) were obtained. The upper band along with plasma and platelets was carefully aspirated and discarded. The lower band was then recovered using a fine Pasteur pipette, suspended in three times the original volume with tissue culture medium RPMI 1640 (GIBCO, Paisley, Scotland) and centrifuged at 400 x g for 15 min. The pelleted cells were washed twice with RPMI 1640 and suspended in PBS. The percentage of PMN was determined by differential staining with MayGrunwald-Giemsa, and the concentration adjusted to 1 x lo7 PMN per millilitre.
pl of oxygen consumed by PMN after stimulation with bacteria or zymosan pl of oxygen consumed by PMN at rest (baseline oxygen consumption) Chemiluminescence ( C L)
CL was measured with a Packard Picolite 6100 luminometer which records light emission by PMN stimulated during oxidative metabolism. Photons of free oxygen radicals are amplified by luminol, a cyclic hydrazide (Allen et af., 1972; Cheson et af., 1976). Each reaction vial consisted of 9 x lo5 PMN in 450 pl PBS with 10% immune serum and lo-' M luminol. After 30 rnin of dark adaptation and temperature equilibration at 37"C, background CL was recorded for 5 min. Bacterial suspension or zymosan was added to initiate the metabolic burst, at a bacteria to PMN ratio of 10 :1 and a final zymosan concentration of 10% of the stock. CL was recorded every minute until a peak value and a decreasing slope were obtained; this process took about 15 to 20 min. CL was expressed as a CL index for 9 x lo5 PMN: maximum count per minute after stimulation with bacteria or zymosan mean counts per minute at rest (background CL)
Oxygen consumption assay Morphological studies
Oxygen consumption during phagocytosis was measured by means of a polarographic technique using a Clark electrode (Yellow Springs Instrument Co., INC, Yellow Springs, Ohio), as described by Crowley et al., (1975). Briefly, the phagocytic mixture in the glass reaction vial consisted of 1 x lo7 PMN/ml in PBS with 10% (v/v) immune serum. After five minutes of temperature equilibration at 37"C, baseline oxygen consumption was recorded for five minutes to establish linearity. Phagocytosis
(1) Immunofluorescence. Preparations identical to those used for the oxygen consumption assay were used for immunofluorescent staining. PMN and bacteria were incubated for 15 rnin with constant gentle agitation in a tube rotor (30 rpm) at 37"C, and 50 pl suspensions were cytocentrifuged (Shandon Cytospin 2 Centrifuge, Cheshire, UK) for 5 min at 500 rpm at room temperature. The slides were air-dried, fixed in 70% ethanol and stained
R. LEVASSEUR, E. DOURNON, J.-L. VILDE AND J.-J. POCIDALO
with FITC-conjugated rabbit anti-L. pneumophila sg 1 antibody (Centers for Disease Control, Atlanta, GA) for 18min, followed by a 2-min counterstaining with Evans blue (diluted 1/1000 in PBS). The number of PMN containing at least two fluorescing bacteria was scored, and a minimum of 200 cells were counted per preparation run in duplicate.
Assessment of the functional activity of PMN after contact with virulent and avirulent bacteria
After ingesting the virulent and avirulent bacteria, PMN were further stimulated with opsonized zymosan (10% dilution of stock), and the post-phagocytic metabolic burst was evaluated in the oxygen consumption assay, as described above. Viability was checked using a trypan blue dye exclusion test.
(2) Transmission electron microscopy (EM). For EM studies, cells were prepared as for the oxygen consumption assay, except that the bacteria to PMN ratio was 1OO:l. PMN and bacteria were Statistical analysis incubated at 37°C for 5 or 15min with constant agitation in a tube rotor. Cells were pelleted by cold A one-way analysis of variance and Student's t test centrifugation (4°C) with pre-chilled phosphate were used to test significance (p < 0.05). buffer (PB, pH 7.2, 0.1 M). Before fixation with gluteraldehyde, cells were washed twice in PB to remove excess serum, which otherwise precipitates with gluteraldehyde. PMN were then fixed for RESULTS ultrathin section preparation with 2.5% gluteraldehyde in PB for 20 min at 4°C. After washing twice with cold PB, cells were postfixed in 2% OsO,, Oxygen consumption assay (Table I) dehydrated in a graded ethanol series, critical-point dried and covered with epoxy resin. Ultrathin The oxygen consumption of normal PMN in the sections were viewed in a Philips transmission presence of live or heat-killed virulent bacteria electron microscope (Model EM 301). Electron mi- increased slightly over background, although the crographs (magnification x 22 000) were eva- difference was not statistically significant. In conluated for the number of bacteria ingested per trast, oxygen consumption by PMN in the presence PMN. Nine to 17 micrographs were scanned for of live or heat-killed avirulent L. pneumophila was each bacterial preparation and the mean number of significantly increased compared to P M N at rest 0, < 0.001). Oxygen consumption by PMN was bacteria ingested per PMN was determined.
Table 1. Oxygen consumption by PMN stimulated by virulent and avirulent L. pneumophila Stimulant
Live virulent bacteria ( n = 14) Live avirulent bacteria ( n = 14) Heat-killed virulent bacteria (n = 13) Heat-killed avirulent bacteria ( n = 9)
Oxygen consumption (pi of o,/h/i07 P M N ) ~
6.22& 2.92 5.99f 2.41
8.83f 4.72 16.84f 7.27