- Email: [email protected]
Investigation of the role of macrophages and endogenous interferon-γ in natural resistance of mice against Legionella pneumophila infection
183
FEMS MicrobiologyImmunology89 (1992) 183-192 © 1992 Federation of European MicrobiologicalSocieties 0920-8534/92/$05.00 Published by Elsevier
FEMSIM 00203
Investigation of the role of macrophages and endogenous interferon-,/in natural resistance of mice against Legionella pneumophila infection H i r o n o b u Fujio 1, Shin-ichi Y o s h i d a 1, H i r o s h i M i y a m o t o and Yasuo Mizuguchi 1
1, M a s a o
Mitsuyama 2
1Department of Microbiology, School of Medicine, Uniuersityof Occupational and Enuironmental Health, Kitakyushu, and e Department of Bacteriology, School of Medicine, Niigata University, Niigata, Japan
Received 29 November 1991 Accepted 31 December 1991 Key words: Legionella pneumophila; Interferon-y; Macrophage; Natural resistance
1. S U M M A R Y Mice are highly resistant to Legionella pneumophila infection. To study the natural resistance, we used A / J and C 5 7 B L / 6 mice which have macrophages permissive and non-permissive for the intracellular growth of L. pneumophila, respectively. The LDs0 for A / J and C 5 7 B L / 6 were 2.7 × 107 and 7.2 x 107 CFU, respectively, indicating that the difference in macrophage ability to regulate the bacterial growth had some effect on susceptibility to L. pneumophila. There was no difference between both strains in elimination of the bacteria from the blood stream within 5 h after infection. When mice were challenged intravenously with a sublethal dose (4 x 10 6 CFU), the bacterial burden in the liver at day 1 was significantly higher in A / J than in C57BL/6. The bacteria, thereafter, were elimiCorrespondence to: H. Fujio, Department of Microbiology, School of Medicine, Universityof Occupational and Environmental Health, Kitakyushu, 807 Japan.
nated rapidly from the liver at a similar rate in both strains. Elimination of the bacteria from the spleen and lungs was also delayed in A / J as compared to C57BL/6. Naive spleen cells of both strains in vitro could produce a large amount of interferon-y (IFN-y) one day after they were stimulated with formalin-killed L. pneumophila. When anti-murine IFN-y monoclonal antibody was administered, the bacterial burden in liver, spleen and lungs significantly increased in A / J , and also in C 5 7 B L / 6 to some extent. We suggest that the innate macrophages' ability to regulate the intracellular bacterial growth and the endogenous IFN-y produced in a very early phase play a critical role in murine natural resistance against L. pneumophila infection.
2. I N T R O D U C T I O N Legionella pneumophila is a facultative intracellular bacterium which was first described as an important causative agent for acute pneumonia
184 [1]. The bacteria can grow in human monocytes [2] and alveolar macrophages [3]. The bacteria inhibit phagosome-lysosome fusion [4] and acidification of phagosomes [5] of human monocytes, and inhibit polymorphonuclear leukocyte function by their exoproducts [6]. These characteristics of L. pneumophila are thought to be pathogenic factors for the intracellular bacterial growth and outbreak of Legionnaires' disease. In rodents there is a species difference in susceptibility to L. pneurnophila infection [7]. Guinea pigs were highly susceptible to this pathogen and a 50% lethal dose (LDs0) was 7.6 x 10 4 CFU/animal by intraperitoneal injection. The bacteria could proliferate in peritoneal macrophages of guinea pigs, resulting in the death of macrophages [7,8]. On the other hand, many strains of mice were highly resistant to L. pneumophila. For example, the LDs0 value was 6.7 x 10 7 CFU/animal for BALB/c mice. The bacteria could not proliferate in the macrophages of many strains of mice, including the C57BL/6 strain. It was thought that the difference in abilities of macrophages to suppress the intracellular bacterial growth resulted in the species difference between guinea pigs and mice. Recently, Yamamoto et al. reported that L. pneumophila could proliferate well in thioglycollate-elicited peritoneal macrophages of A / J mouse strain [9] and that interferon-y (IFN-T) activated the macrophages to inhibit the intracellular bacterial growth [10]. It has been shown that IFN-T activated human monocytes [11] and human alveolar macrophages [12] to inhibit the intracellular multiplication of L. pneurnophila. In the present study, the natural resistance of mice against L. pneumophila infection by using A / J and C57BL/6 strains was examined. The 50% lethal dose of L. pneumophila for A / J strain was 2.7 x 107 CFU per animal, suggesting that A / J mice are resistant to the bacteria despite the fact that their macrophages are permissive for intracellular bacterial growth in vitro. Then it was demonstrated that endogenous IFN-T enhanced the protection from a very early stage of Legionella infection in A / J mice.
3. MATERIALS AND METHODS
3.1. Bacteria L. pneumophila Philadelphia-1 (a clinical isolate, ATCC33152) was donated by the Centers for Disease Control, Atlanta, GA, in 1980. The bacteria were passaged once in mice before they were used in this study, i.e., they were inoculated intravenously (i.v.) into A / J mice. Fresh isolates were obtained from the spleen on day 3 postinoculation and were grown once on charcoal-yeast extract (CYE) agar plates. CYE agar plates, whose pH was adjusted to 6.9 with 1 M KOH, contained a legionella agar base (Difco Laboratories, Detroit, MI) supplemented with (per liter of deionized water) 0.4 g of L-cysteine and 0.25 g of ferric pyrophosphate. The bacteria were stored at - 8 0 ° C until use. Formalin-killed L. pneumophila was prepared as described previously [13]. Briefly, the bacteria were cultured on CYE agar plates for 2 days and harvested with phosphate-buffered saline (PBS). After being washed several times, the bacterial cells were treated with 0.5% formalin in PBS for 24 h and then washed twice. We confirmed that all bacteria were killed by this treatment. They were stored at 4°C until use. 3.2. Animals Male mice of A / J , C57BL/6, and ddY strains were purchased from Shizuoka Experimental Animals, Hamamatsu, Japan. Age-matched 8- to 12-weeks old mice were used. 3.3. Preparation of macrophage monolayers Resident alveolar or peritoneal macrophages were collected from the lung or peritoneal lavage fluid of normal mice, respectively. The macrophage concentration was adjusted to approximately 5 x l0 s cells (alveolar macrophages) or 1 X 10 6 cells (peritoneal macrophages) per ml in RPMI 1640 medium containing 10% (v/v) newborn calf serum (GIBCO Laboratories, Grand Island, NY) and 100 U of penicillin G, and 0.2 ml of the cell suspension was placed into each well of 96-well tissue culture plate (Falcon no. 3072; Becton Dickinson Labware, Oxnard, CA). The
185 plates were incubated for 1.5 h at 37°C in a C O 2 incubator and then washed with sterile PBS to remove nonadherent cells. The adherent cells were incubated with RPMI 1640 containing newborn calf serum (10%) without antibiotics for 18 h.
3.4. Assay of intracellular growth of L. pneumophila Macrophages in each well were infected with bacterial cells (the bacterium to macrophage ratio was approximately 1 : 1) for 1.5 h, and nonphagocytosed bacteria were washed out. After 0, 24, 48, or 72 h of incubation, culture media was discarded, 0.2 ml of sterile distilled water was added to each well, and adherent cells were scraped from the bottoms of the wells with a rubber policeman. After serial 10-fold dilution of the suspension, the number of L. pneumophila CFU in each single well was determined on CYE agar.
3.5. Determination of 50% lethal dose A / J and C 5 7 B L / 6 mice (one group consisted of five animals) were injected i.v. with 0.2 ml of serial 3.17-fold dilutions of the bacterial suspensions (1.5 x 108 CFU, 4.7 x 107 CFU, and 1.5 x 107 CFU), and the total numbers of deaths were counted until 10 days after the infection. The 50% lethal dose (LDs0) was calculated by the Reed-Muench method.
3.6. Bacterial clearance from the bloodstream in mice After mice were injected i.v. with the bacteria (4.0 X 106 CFU), 0.05 ml of blood was obtained, at 1 min, 1, 3 and 5 h intervals, from the retroorbital plexus by capillary tubes as previously described [14]. After appropriate serial dilutions in PBS, viable cells in the blood were counted on CYE agar plates.
3. 7. Determination of numbers of viable L. pneumophila in the organs Mice were injected i.v. with a sublethal dose (3.0 x 105 ~ 4.0 x 106 CFU) of L. pneumophila. The numbers of viable L. pneumophila in livers, spleens and lungs of the infected mice were counted on CYE agar plates after appropriate
dilution of homogenized suspensions. The elimination rate (ER) from these organs was calculated using the following formula: E R m - n = [Mean Log10 CFU in the organs at m days posti n f e c t i o n - Mean Log10 CFU in the organs at n days post i n f e c t i o n ] / [ n - m ] . Some A / J mice were transferred i.v. with 0.5 ml of C5-containing serum from ddY mice 1 h before infection and the bacterial clearance from their organs were examined.
3.8. IFN-7 production from spleen cells in uitro Dispersed spleen cell suspensions were prepared by teasing the spleen in a sterile RPMI 1640 medium. The cells were then washed three times. The unfractionated spleen cell suspensions, at a concentration of 10 7 cells per ml, were added to 24-well plastic plates (Falcon, Oxnard, CA), using 1 ml of RPMI 1640 medium containing 10% of fetal calf serum and 100 U of penicillin G and 100/xg of streptomycin per ml. The cell cultures, with or without 108 per well of formalin-killed L. pneumophila, were incubated for 18 h at 37°C in a CO 2 incubator. After incubation, the culture supernatants were collected by centrifugation.
3.9. IFN-y assay The IFN-7 was quantitated by enzyme immunoassay (EIA) as described previously [15]. Rat anti-mouse IFN-7 monoclonal antibody (LEE Biomolecular Research Inc., San Diego, CA) was used as a primary antibody. Rabbit anti-mouse IFN-y serum was used as a secondary antibody, followed by incubation with peroxidase-conjugated goat anti-rabbit immunoglobulin G (Cappel, Organon Teknika Corp., West Chester, PA). All EIAs were run with standard mouse IFN-7. The IFN-7 activity was expressed as international units per milliliter. Recombinant mouse IFN-7 and rabbit anti-IFN-7 serum were generous gifts from the Research Institute, Daiichi Seiyaku Co. Ltd., Tokyo, Japan.
3.10. In viuo administration of anti-mouse IFN-y antibody Rat immunoglobulin G1 monoclonal antibody (mAb) against natural mouse IFN-y was purified
186 by using D E A E Affi-Gel Blue Column chromatography [16] from the ascites fluid in Pristane-primed B A L B / c nude mice injected with hybridoma R4-6A2 [17]. Purified mAb (1 mg) had a neutralizing titer of 2 X 105 U against recombinant murine I F N - y [18]. We injected into each mouse a single intravenous injection of 3 0 / z g of anti-IFN- 7 mAb in 0.2 ml of PBS 2 h before infection with L. pneumophila. Purified normal rat globulin used as a control was purchased from Cappel, Organon Teknika Corp., West Chester, PA.
3.10. Statistics Data were analysed using the Student's t-test. A P value of < 0.05 was significant.
100 -
80 t~
]
N J , 1.5X107 and
I
B6,1.5X107
//
z{z B6, 4.7X10 7
1 5X108
g,
60
e-
~
4o
a.
A/J, 4.7X107 20 1.5 110 8
Days after Infection
Fig. 2. The survival curve of the mice injected i.v. with three different doses of L. pneumophila. A/J mice (o) or C57BL/6 mice (B/6, e) were injected with 1.5× 10 7 CFU, 4.7)<10 7 CFU, or 1.5 x 108 CFU, respectively. Each group consisted of five animals.
4. R E S U L T S
4.1. Strain differences in intracellular L. pneumophila growth between A / J and C 5 7 B L / 6 macrophages The intracellular growth of L. pneumophila in resident alveolar and peritoneal macrophages of both strains was compared (Fig. 1). The bacteria proliferated more than 10-fold in resident peri-
toneal macrophages of A / J mice, but bacterial growth was hardly detected in C 5 7 B L / 6 macrophages. In A / J alveolar macrophages the bacteria proliferated several-fold after 2 days of in vitro phagocytosis, in contrast to a decrease in bacterial numbers in C 5 7 B L / 6 alveolar macrophages.
4.2. Strain differences in the susceptibilities of A / J and C57BL / 6 mice to L. pneumophila Peritoneal 2
mecrophages
*
• ,
Alveolar
macrophages
2. A/J
'o
'
.J .2,
3
.3 0
1
2
3
Days of in vitro culture
Fig. 1. Differences in intracellular L. pneumophila growth in A/J (©) and C57BL/6 (e) macrophages. Bacterial burdens in cultured macrophages are expressed as logt0 increases in CFU from day 0 to day 3 of in vitro culture. The data are means + standard deviations of representative data from three individual experiments in which four mice were used in each. The asterisks * and ** indicate P<0.05 and P<0.01, respectively, significant differences between A/J and C57BL/6.
In order to compare the susceptibilities of A / J and C 5 7 B L / 6 mice to L. pneumophila, mice were injected i.v. with different doses of the bacteria. Their survival was observed for 10 days after injection (Fig. 2). When mice were injected with 4.7 x 107 CFU of L. pneumophila, the difference of the survival rate between the two strains was observed, i.e., 80% of C 5 7 B L / 6 mice survived in comparison with no survivors in A / J mice. All mice surviving on day 3 after inoculation did not die thereafter. The LDs0 values of the bacteria for A / J and C 5 7 B L / 6 mice were 2.7 × 107 CFU and 7.2 × 10 7 CFU, respectively,
4.3. Bacterial clearance of L. pneumophila from the bloodstream of A / J and C57BL / 6 mice Clearance rates of bacteria from the bloodstream were compared in the two strains (Fig. 3). Blood samples were obtained 1 min, 1, 3, and 5 h
187 5.5.
i'°[
M.
4.5'
o
o
.J
3.5
30
Hours after infection Fig. 3. Bacterial clearance from the bloodstream of A / J (©) and C57BL/6 (e) mice..After the mice were injected i.v. with
injected with a bacterial dose far below the LDs0 (3.0 × 105 CFU), the number of bacteria in the organs of both A / J and C 5 7 B L / 6 decreased at a similar and considerably rapid rate by day 1 after infection (data not shown). Because A / J mice are C5-deficient, they were passively transferred with C5-containing serum of ddY mice in order to examine the possibility that C5 plays some roles in the resistance. A / J mice which were injected i.v. with 0.5 ml of C5-containing serum per mouse 1 h before infection did not improve their ability to eliminate the bacteria in the livers, spleens and lungs (data not shown).
4× 10 6 CFU of L. pneumophila, blood samples were collected from the retro-orbital plexus and the bacterial CFU/ml were counted. The data are means± standard deviations of representative data from three individual experiments in which three mice were used in each.
Liver
u
after the bacterial injection. Bacterial counts fell by half after 1 h, and by 90% after 5 h in both A / J and C 5 7 B L / 6 mice, indicating that bacterial uptake by organs within 5 h were similar in the two strains.
4.4. Strain differences in the growth of L. pneumophila in the livers, spleens, and lungs of infected mice Mice of A / J and C 5 7 B L / 6 strains were injected i.v. with a sublethal dose (4.0 × 106 CFU) of bacteria, and CFU in the organs were determined at intervals for comparison (Fig. 4). At one day after infection, the bacterial burden in the liver was significantly higher in A / J than in C 5 7 B L / 6 ( P < 0 . 0 5 ) . Thereafter, the bacteria count in the livers of both strains decreased at a rapid and similar rate until the bacterium became undetectable ( < 102 CFU per organ). E R 1 - 3 was 1.28 for A / J and 1.00 for C 5 7 B L / 6 . In the spleen, the elimination rate of the bacteria was slower in A / J ( E R 0 - 3 = 0.63) than in C 5 7 B L / 6 ( E R 0 - 3 = 1.83) mice, although the increase of the bacterial number was not seen in A / J . The bacterial elimination in lungs of A / J and C 5 7 B L / 6 were similar to that in the spleen (data not shown). When both strains of mice were
(J o
& 0 ..J
1
-
i
-
1
i
2
.
i
3
•
i
4
•
i
•
i
5
•
i
6
7
Days after Infection 6
Spleen
1
2
3
4
5
6
7
Days after Infection Fig. 4. Differences in the growth of L. pneumophila in the livers and spleens of A/J ((3) and C57BL/6 (e) mice after the mice were injected i.v. with 4 × 10 6 CFU of the bacteria. The data are means+ standard deviations for three animals. The asterisks * and ** indicate P < 0.05 and P < 0.01, respectively, significant differences between A/J and C57BL/6.
188
4.5. In vitro production of IFN-T in response to killed bacteria To compare the abilities of IFN-y production in A / J and C57BL/6 strains, spleen cell cultures of non-immunized mice were stimulated in vitro with formalin-killed L. pneumophila for 18 h and IFN-y activity in the culture supernatant fluids was assayed. The amount of IFN-T produced by splenocytes of A / J was 467.9 _+ 169.1 I U / m l and that of C57BL/6 was 655.5 +__63.0 I U / m l (mean +standard deviations for four animals). Although the amount of IFN-y production by A / J splenocytes is somewhat smaller than that of C57BL/6, the difference was not significant (P = 0.067). 4.6. The effect of anti-IFN-7 mAb administration on the resistance of mice against L. pneumophila infection Mice of either strain were injected i.v. with rat anti-mouse IFN-y mAb (30 ~g) or normal rat globulin (30 tzg) 2 h before infection with a sublethal dose (4 × 106 CFU for A / J and 1 x 107 CFU for C57BL/6) of L. pneumophila. The number of L. pneumophila cells in livers, spleens and lungs of treated mice was determined on day 1, 4
Spleen
Liver
..
6
Anti-IFNy
s
4- ~
4
.J
Spleen
Liver
5
~
B
AnI-iIFN-7
:1
........... ::::
Days after infection Fig. 6. The effect of anti-IFN-7 m A b administration on the resistance of C 5 7 B L / 6 mice against L. pneumophila infection. C 5 7 B L / 6 mice were pretreated with anti-IFN-7 m A b (e) or normal rat globulin (©), and injected i.v. with 1 × 107 C F U of L. pneumophila. T h e bacterial growth in the organs was determined at intervals. The data are m e a n s 5: standard deviations for three animals. * indicates P < 0.05 with respect to control mice.
and 7 of infection in A / J (Fig. 5), or on day 1, 3 and 5 of infection in C57BL/6 (Fig. 6). The bacterial number in the organs of A / J mice pretreated with anti-IFN-y mAb significantly increased as compared with those of control mice until day 4 of infection. The number of bacteria in the organs of C57BL/6 mice pretreated with anti-IFN-y mAb increased significantly on day 1 of infection. Data of the bacterial elimination from lungs of both strains adminstered with antiIFN-y mAb or normal rat globulin showed similar tendencies to those seen in the spleens (data not shown).
3"
2
0
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2-
I 2 3 4 5 6 7 ,
.
,
-
,
-
,
-
,
-
,
"
,
Days after infection Fig. 5. T h e effect of anti-IFN-7 m A b administration on the resistance of A / J mice against L. pneumophila infection. A / J mice were pretreated with anti-IFN-3, m A b (e) or normal rat globulin (©), and injected i.v. with 4 x 106 C F U of L. pneumophila. T h e bacterial growth in the organs was determined at intervals. The data are m e a n s + s t a n d a r d deviations for three animals. * indicates P < 0.05 with respect to control mice; * * indicates P < 0.01.
5. DISCUSSION As far as we know, there is no report which has evaluated the role of macrophages and endogenous IFN-y in natural resistance of mice against L. pneumophila infection. Our results made clear the following: strain differences in the macrophage ability to inhibit the intracellular bacterial growth (Fig. 1) had some effect on their susceptibilities to the infection (Fig. 2); the LD50 value of A / J mice (2.7 × 107 CFU) was greater
189
than the LDs0 for guinea pigs (7.6 x 10 4 CFU) whose macrophages also permit the intracellular bacterial growth [7]; even though the A / J macrophages permitted intracellular bacterial growth in vitro, the bacteria could be eliminated in the liver and spleen (Fig. 4); and there was a significant increase of bacterial burden in the organs of A / J (Fig. 5) and C57BL/6 (Fig. 6) when mice were pretreated with anti-IFN-7 mAb suggesting that endogenous IFN-7 acts as an important protective factor in a very early phase of Legionella infection in mice. Splenocytes of A / J and C57BL/6 mice could produce a large amount of IFN-7 within 18 h of formalin-killed L. pneumophila stimulation in vitro, i.e. 467.9 ___169.1 for A / J and 655.5 + 63.0 for C57BL/6 (IU/ml). Although the amount of IFN-7 production by A / J splenocytes was somewhat smaller than that of C57BL/6, the difference was not significant (P = 0.067). Both strains of mice seem to be able to produce just enough IFN-7 for augmentation of the protection in the very early phase of infection with L. pneumophila. Results of studies using anti-IFN-7 mAb showed that the pretreatment with anti-IFN-7 was effective only on day 1 of infection in C57BL/6 mice (P values for both livers and spleens were < 0.05, see Fig. 6), while in A / J mice it was effective until day 4 of infection (both P values < 0.01, see Fig. 5). From the above, it is indicated that endogenous IFN-7 had a relatively small influence on the resistance of C57BL/6 mice whose macrophages are innately non-permissive for the intracellular bacterial growth. It is known that IFN-7, by activating macrophages, plays an important role in host defense against many bacteria [reviewed by Murray, 19]. IFN-7 was produced in the early phase of listerial infection in mice [20]. Blanchard et al. reported that in vitro production of IFN-7 from mouse splenocytes could be detected as early as 6 h after stimulation with the L. pneumophila antigen, continued to increase for up to 24 h after stimulation and thereafter almost completely ceased [13]. The same group showed that macrophages of A / J mice pretreated with IFN-y could restrict the intracellular growth of L. pneumophila in vitro [10].
The natural resistance of A / J mice to L. pneumophila will be well understood by comparing it with that of guinea pigs. In spite of the evidence that macrophages of both A / J mice and guinea pigs permitted the intracellular growth of L. pneumophila [8,9], LDs0 of L. pneumophila for A / J mice was 2.7 x 107 CFU which was still higher than 7.6 x 104 CFU of guinea pigs [7]. The proliferative response of spleen cells against Legionella vaccine did not significantly increase until day 6 post-infection in immunized guinea pigs [21], while splenocytes from even non-immunized A / J mice proliferated to a reasonable extent in response to the antigen [10]. After several hours of culture with Legionella antigen, A / J splenocytes could produce IFN-7 in vitro, while macrophage activating factor (IFN-7) could not be produced by normal guinea pig splenocytes by in vitro antigen stimulation. This factor could be produced in guinea pigs after 4 days of the infection with L. pneumophila [21]. From the above comparison, a rapid T cell response and rapid IFN-7 production are thought to be the reasons why A / J are very resistant to Legionella infection. It was suggested that mouse natural killer (NK) cells were responsible for in vitro production of IFN-7 in response to L. pneumophila [22] or L. monocytogenes [23] antigen. Dunn et al. showed in vivo that early elimination of NK cells by treatment with rabbit antiserum to asialo GM1 resulted in exacerbation of listerial infection in mice [24]. We, however, could not observe the role of NK cells in L. pneumophila infection by injecting antiserum to asialo GM1 in vivo (unpublished data). In addition, administration of antiL3T4 (CD4) mAb or anti-Lyt 2.2 (CD8) mAb [25], or combined administration of these three antibodies (including anti-asialo GM1) could not influence the murine resistance against L. pneumophila (unpublished data). Cell population responsible for the production of IFN-7 in the course of infection is yet to be determined. Although Yamamoto et al. have reported that L. pneumophila did not grow in resident peritoneal macrophages of A / J mice within two days after in vitro phagocytosis [9], we could observe the bacterial growth in these macrophages. The
190
discrepancy would be due to the difference of L.
pneumophila strains used. In addition, they did not determine bacterial growth at day 3 of culture, at which we found the significant intracellular bacterial growth. In the murine protection system against L. pneumophila, there are two critical factors: the ability of innate macrophages to inhibit the intracellular bacterial growth, and the enhancement of resistance by IFN-y . Although the question of the identity of the cell population responsible for endogenous IFN-3, production in vivo remains, we insist that these two factors are important in the natural resistance.
ACKNOWLEDGEMENTS This study was supported by Grant-in Aid for Scientific Research 03670229 from the Ministry of Education, Science and Culture, Japan.
[8]
[9]
[10]
[11]
[12]
[13]
REFERENCES [14] [1] Fraser, D.W., Tsai, T.R., Orenstein, W., Parkin, W.E., Beecham, H.J., Sharrar, R.G., Harris, J., Mallison, G.F., Martin, S.M., McDade, J.E., Shepard, C.C. and Brachman, P.S. (1977) Legionnaires' disease, description of an epidemic of pneumonia. N. Engl. J. Med. 297, 1189-1197. [2] Horwitz, M.A. and Silverstein, S.C. (1980) Legionnaires' disease bacterium (Legionella pneumophila) multiplies intracellularly in human monocytes. J. Clin. Invest. 66, 441-450. [3] Nash, T.W., Libby, D.M. and Horwitz, M.A. (1984) Interaction between the Legionnaires' disease bacterium (Legionella pneumophila) and human alveolar macrophages. J. Clin. Invest. 74, 771-782. [4] Horwitz, M.A. (1983) The Legionnaires' disease bacterium (Legionella pneumophila) inhibits phagosomelysosome fusion in human monocytes. J. Exp. Med. 158, 2108-2126. [5] Horwitz, M.A. and Max_field, F.R. (1984) Legionella pneumophila inhibits acidification of its phagosome in human monocytes. J. Cell. Biol. 99, 1936-1943. [6] Sahney, N.N., Lambe, B.C., Summersgill, J.T. and Miller, R.D. (1990) Inhibition of polymorphonuclear leukocyte function by Legionella pneumophila exoproducts. Microb. Pathogen. 9, 117-125. [7] Yoshida, S. and Mizuguchi, Y. (1986) Multiplication of Legionella pneumophila Philadelphia-1 in cultured peri-
[15]
[16]
[17]
[18]
[19]
[20]
toneal macrophages and its correlation to susceptibility of animals. Can. J. Microbiol. 33, 438-442. Yoshida, S., Mizuguchi, Y., Nikaido, Y., Mitsuyama, M. and Nomoto, K. (1987) Fate of Legionella pneumophila Philadelphia-1 strain in resident, elicited, activated and immune peritoneal macrophages of guinea pigs. Infect. Immun. 55, 2477-2482. Yamamoto, Y., Klein, T.W., Newton, C.A., Widen, R. and Friedman, H. (1988) Growth of Legionella pneumophila in thioglycolate-elicited peritoneal macrophages from A / J mice. Infect. Immun. 56, 370-375. Klein, T.W., Blanchard, D.K., Yamamoto, Y., Newton, C., Widen, R. and Friedman, H. (1988) Role of cytokines in resistance to infection with Legionella pneumophila. Adv. Biosci. 68, 259-265. Byrd, T.F. and Horwitz, M.A. (1989) Interferon gammaactivated human monocytes downregulate transferrin receptors and inhibit the intracellular multiplication of Legionella pneumophila by limiting the availability of iron. J. Clin. Invest. 83, 1457-1465. Nash, T.W., Libby, D.M. and Horwitz, M.A. (1988) IFN3,-activated human alveolar macrophages inhibit the intracellular multiplication of Legionella pneumophila. J. Immunol. 140, 3978-3981. Blanchard, D.K., Klein, T.W., Friedman, H. and Stewart II, W.E. (1985) Kinetics and characterization of interferon production by murine spleen cells stimulated with Legionella pneumophila antigens. Infect. Immun. 49, 719723. Yoshida, S., Ogawa, M. and Mizuguchi, Y. (1985) Relation of capsular materials and colony opacity to virulence of Vibrio vulnificus. Infect. Immun. 47, 446-451. Igarashi, K., Mitsuyama, M., Muramori, K., Tsukada, H. and Nomoto, K. (1990) Interleukin-l-induced promotion of T-cell differentiation in mice immunized with killed Listeria monocytogenes. Infect. Immun. 58, 3973-3979. Bruck, C., Portelle, D., Glineur, C. and Bollen, A. (1982) One-step purification of mouse monoclonal antibodies from ascitic fluid by DEAE Affi-Gel Blue chromatography. J. Immunol. Methods 53, 313-319. Havell, E.A. (1986) Purification and further characterization of anti-murine interferon-3, monoclonal neutralizing antibody. J. Interferon Res. 6, 489-497. Nakane, A., Minagawa, T., Kohanawa, M., Chen, Y., Sato, H., Moriama, M. and Tsuruoka, N. (1989) Interactions between endogenous gamma interferon and tumor necrosis factor in host resistance against primary and secondary Listeria monocytogenes infections. Infect. Immun. 57, 3331-3337. Murray, H.W. (1988) Interferon-gamma, the activated macrophages and host defense against microbial challenge. Ann. Intern. Med. 108, 595-608. Nakane, A., Numata, A., Asano, M., Kohanawa, M., Chen, Y. and Minagawa, T. (1990) Evidence that endogenous gamma interferon is produced early in Listeria monocytogenes infection. Infect. Immun. 58, 2386-2388.
191 [21] Nikaido, Y., Yoshida, S., Goto, Y., Mizuguchi, Y. and Kuroiwa, A. (1989) Macrophage-activating T-cell factor(s) produced in an early phase of Legionella pneumophila infection in guinea pigs. Infect. Immun. 57, 3458-3465. [22] Blanchard, D.K., Friedman, H., Stewart II, W.E., Klein, T.W. and Djeu, J.Y. (1988) Role of gamma interferon in induction of natural killer activity by Legionella pneumophila in vitro and in an experimental murine infection model. Infect. Immun. 56, 1187-1193. [23] Wherry, J.C., Schreiber, R.D. and Unanue, E.R. (1991) Regulation of gamma interferon production by natural
killer cells in scid mice: roles of tumor necrosis factor and bacterial stimuli. Infect. Immun. 59, 1709-1715. [24] Dunn, P.L. and North, R.J. (1991) Early gamma interferon production by natural killer cells is important in defense against murine listeriosis. Infect. Immun. 59, 2892-2900. [25] Sasaki, T., Mieno, M., Udono, H., Yamaguchi, K., Usui, T., Hara, K., Shiku, H. and Nakayama, E. (1989) Roles of CD4 + and CD8 ÷ cells, and the effect of administration of recombinant murine interferon-~, in listerial infection. J. Exp. Med. 171, 1141-1154.
FEMS MicrobiologyImmunology89 (1992) 183-192 © 1992 Federation of European MicrobiologicalSocieties 0920-8534/92/$05.00 Published by Elsevier
FEMSIM 00203
Investigation of the role of macrophages and endogenous interferon-,/in natural resistance of mice against Legionella pneumophila infection H i r o n o b u Fujio 1, Shin-ichi Y o s h i d a 1, H i r o s h i M i y a m o t o and Yasuo Mizuguchi 1
1, M a s a o
Mitsuyama 2
1Department of Microbiology, School of Medicine, Uniuersityof Occupational and Enuironmental Health, Kitakyushu, and e Department of Bacteriology, School of Medicine, Niigata University, Niigata, Japan
Received 29 November 1991 Accepted 31 December 1991 Key words: Legionella pneumophila; Interferon-y; Macrophage; Natural resistance
1. S U M M A R Y Mice are highly resistant to Legionella pneumophila infection. To study the natural resistance, we used A / J and C 5 7 B L / 6 mice which have macrophages permissive and non-permissive for the intracellular growth of L. pneumophila, respectively. The LDs0 for A / J and C 5 7 B L / 6 were 2.7 × 107 and 7.2 x 107 CFU, respectively, indicating that the difference in macrophage ability to regulate the bacterial growth had some effect on susceptibility to L. pneumophila. There was no difference between both strains in elimination of the bacteria from the blood stream within 5 h after infection. When mice were challenged intravenously with a sublethal dose (4 x 10 6 CFU), the bacterial burden in the liver at day 1 was significantly higher in A / J than in C57BL/6. The bacteria, thereafter, were elimiCorrespondence to: H. Fujio, Department of Microbiology, School of Medicine, Universityof Occupational and Environmental Health, Kitakyushu, 807 Japan.
nated rapidly from the liver at a similar rate in both strains. Elimination of the bacteria from the spleen and lungs was also delayed in A / J as compared to C57BL/6. Naive spleen cells of both strains in vitro could produce a large amount of interferon-y (IFN-y) one day after they were stimulated with formalin-killed L. pneumophila. When anti-murine IFN-y monoclonal antibody was administered, the bacterial burden in liver, spleen and lungs significantly increased in A / J , and also in C 5 7 B L / 6 to some extent. We suggest that the innate macrophages' ability to regulate the intracellular bacterial growth and the endogenous IFN-y produced in a very early phase play a critical role in murine natural resistance against L. pneumophila infection.
2. I N T R O D U C T I O N Legionella pneumophila is a facultative intracellular bacterium which was first described as an important causative agent for acute pneumonia
184 [1]. The bacteria can grow in human monocytes [2] and alveolar macrophages [3]. The bacteria inhibit phagosome-lysosome fusion [4] and acidification of phagosomes [5] of human monocytes, and inhibit polymorphonuclear leukocyte function by their exoproducts [6]. These characteristics of L. pneumophila are thought to be pathogenic factors for the intracellular bacterial growth and outbreak of Legionnaires' disease. In rodents there is a species difference in susceptibility to L. pneurnophila infection [7]. Guinea pigs were highly susceptible to this pathogen and a 50% lethal dose (LDs0) was 7.6 x 10 4 CFU/animal by intraperitoneal injection. The bacteria could proliferate in peritoneal macrophages of guinea pigs, resulting in the death of macrophages [7,8]. On the other hand, many strains of mice were highly resistant to L. pneumophila. For example, the LDs0 value was 6.7 x 10 7 CFU/animal for BALB/c mice. The bacteria could not proliferate in the macrophages of many strains of mice, including the C57BL/6 strain. It was thought that the difference in abilities of macrophages to suppress the intracellular bacterial growth resulted in the species difference between guinea pigs and mice. Recently, Yamamoto et al. reported that L. pneumophila could proliferate well in thioglycollate-elicited peritoneal macrophages of A / J mouse strain [9] and that interferon-y (IFN-T) activated the macrophages to inhibit the intracellular bacterial growth [10]. It has been shown that IFN-T activated human monocytes [11] and human alveolar macrophages [12] to inhibit the intracellular multiplication of L. pneurnophila. In the present study, the natural resistance of mice against L. pneumophila infection by using A / J and C57BL/6 strains was examined. The 50% lethal dose of L. pneumophila for A / J strain was 2.7 x 107 CFU per animal, suggesting that A / J mice are resistant to the bacteria despite the fact that their macrophages are permissive for intracellular bacterial growth in vitro. Then it was demonstrated that endogenous IFN-T enhanced the protection from a very early stage of Legionella infection in A / J mice.
3. MATERIALS AND METHODS
3.1. Bacteria L. pneumophila Philadelphia-1 (a clinical isolate, ATCC33152) was donated by the Centers for Disease Control, Atlanta, GA, in 1980. The bacteria were passaged once in mice before they were used in this study, i.e., they were inoculated intravenously (i.v.) into A / J mice. Fresh isolates were obtained from the spleen on day 3 postinoculation and were grown once on charcoal-yeast extract (CYE) agar plates. CYE agar plates, whose pH was adjusted to 6.9 with 1 M KOH, contained a legionella agar base (Difco Laboratories, Detroit, MI) supplemented with (per liter of deionized water) 0.4 g of L-cysteine and 0.25 g of ferric pyrophosphate. The bacteria were stored at - 8 0 ° C until use. Formalin-killed L. pneumophila was prepared as described previously [13]. Briefly, the bacteria were cultured on CYE agar plates for 2 days and harvested with phosphate-buffered saline (PBS). After being washed several times, the bacterial cells were treated with 0.5% formalin in PBS for 24 h and then washed twice. We confirmed that all bacteria were killed by this treatment. They were stored at 4°C until use. 3.2. Animals Male mice of A / J , C57BL/6, and ddY strains were purchased from Shizuoka Experimental Animals, Hamamatsu, Japan. Age-matched 8- to 12-weeks old mice were used. 3.3. Preparation of macrophage monolayers Resident alveolar or peritoneal macrophages were collected from the lung or peritoneal lavage fluid of normal mice, respectively. The macrophage concentration was adjusted to approximately 5 x l0 s cells (alveolar macrophages) or 1 X 10 6 cells (peritoneal macrophages) per ml in RPMI 1640 medium containing 10% (v/v) newborn calf serum (GIBCO Laboratories, Grand Island, NY) and 100 U of penicillin G, and 0.2 ml of the cell suspension was placed into each well of 96-well tissue culture plate (Falcon no. 3072; Becton Dickinson Labware, Oxnard, CA). The
185 plates were incubated for 1.5 h at 37°C in a C O 2 incubator and then washed with sterile PBS to remove nonadherent cells. The adherent cells were incubated with RPMI 1640 containing newborn calf serum (10%) without antibiotics for 18 h.
3.4. Assay of intracellular growth of L. pneumophila Macrophages in each well were infected with bacterial cells (the bacterium to macrophage ratio was approximately 1 : 1) for 1.5 h, and nonphagocytosed bacteria were washed out. After 0, 24, 48, or 72 h of incubation, culture media was discarded, 0.2 ml of sterile distilled water was added to each well, and adherent cells were scraped from the bottoms of the wells with a rubber policeman. After serial 10-fold dilution of the suspension, the number of L. pneumophila CFU in each single well was determined on CYE agar.
3.5. Determination of 50% lethal dose A / J and C 5 7 B L / 6 mice (one group consisted of five animals) were injected i.v. with 0.2 ml of serial 3.17-fold dilutions of the bacterial suspensions (1.5 x 108 CFU, 4.7 x 107 CFU, and 1.5 x 107 CFU), and the total numbers of deaths were counted until 10 days after the infection. The 50% lethal dose (LDs0) was calculated by the Reed-Muench method.
3.6. Bacterial clearance from the bloodstream in mice After mice were injected i.v. with the bacteria (4.0 X 106 CFU), 0.05 ml of blood was obtained, at 1 min, 1, 3 and 5 h intervals, from the retroorbital plexus by capillary tubes as previously described [14]. After appropriate serial dilutions in PBS, viable cells in the blood were counted on CYE agar plates.
3. 7. Determination of numbers of viable L. pneumophila in the organs Mice were injected i.v. with a sublethal dose (3.0 x 105 ~ 4.0 x 106 CFU) of L. pneumophila. The numbers of viable L. pneumophila in livers, spleens and lungs of the infected mice were counted on CYE agar plates after appropriate
dilution of homogenized suspensions. The elimination rate (ER) from these organs was calculated using the following formula: E R m - n = [Mean Log10 CFU in the organs at m days posti n f e c t i o n - Mean Log10 CFU in the organs at n days post i n f e c t i o n ] / [ n - m ] . Some A / J mice were transferred i.v. with 0.5 ml of C5-containing serum from ddY mice 1 h before infection and the bacterial clearance from their organs were examined.
3.8. IFN-7 production from spleen cells in uitro Dispersed spleen cell suspensions were prepared by teasing the spleen in a sterile RPMI 1640 medium. The cells were then washed three times. The unfractionated spleen cell suspensions, at a concentration of 10 7 cells per ml, were added to 24-well plastic plates (Falcon, Oxnard, CA), using 1 ml of RPMI 1640 medium containing 10% of fetal calf serum and 100 U of penicillin G and 100/xg of streptomycin per ml. The cell cultures, with or without 108 per well of formalin-killed L. pneumophila, were incubated for 18 h at 37°C in a CO 2 incubator. After incubation, the culture supernatants were collected by centrifugation.
3.9. IFN-y assay The IFN-7 was quantitated by enzyme immunoassay (EIA) as described previously [15]. Rat anti-mouse IFN-7 monoclonal antibody (LEE Biomolecular Research Inc., San Diego, CA) was used as a primary antibody. Rabbit anti-mouse IFN-y serum was used as a secondary antibody, followed by incubation with peroxidase-conjugated goat anti-rabbit immunoglobulin G (Cappel, Organon Teknika Corp., West Chester, PA). All EIAs were run with standard mouse IFN-7. The IFN-7 activity was expressed as international units per milliliter. Recombinant mouse IFN-7 and rabbit anti-IFN-7 serum were generous gifts from the Research Institute, Daiichi Seiyaku Co. Ltd., Tokyo, Japan.
3.10. In viuo administration of anti-mouse IFN-y antibody Rat immunoglobulin G1 monoclonal antibody (mAb) against natural mouse IFN-y was purified
186 by using D E A E Affi-Gel Blue Column chromatography [16] from the ascites fluid in Pristane-primed B A L B / c nude mice injected with hybridoma R4-6A2 [17]. Purified mAb (1 mg) had a neutralizing titer of 2 X 105 U against recombinant murine I F N - y [18]. We injected into each mouse a single intravenous injection of 3 0 / z g of anti-IFN- 7 mAb in 0.2 ml of PBS 2 h before infection with L. pneumophila. Purified normal rat globulin used as a control was purchased from Cappel, Organon Teknika Corp., West Chester, PA.
3.10. Statistics Data were analysed using the Student's t-test. A P value of < 0.05 was significant.
100 -
80 t~
]
N J , 1.5X107 and
I
B6,1.5X107
//
z{z B6, 4.7X10 7
1 5X108
g,
60
e-
~
4o
a.
A/J, 4.7X107 20 1.5 110 8
Days after Infection
Fig. 2. The survival curve of the mice injected i.v. with three different doses of L. pneumophila. A/J mice (o) or C57BL/6 mice (B/6, e) were injected with 1.5× 10 7 CFU, 4.7)<10 7 CFU, or 1.5 x 108 CFU, respectively. Each group consisted of five animals.
4. R E S U L T S
4.1. Strain differences in intracellular L. pneumophila growth between A / J and C 5 7 B L / 6 macrophages The intracellular growth of L. pneumophila in resident alveolar and peritoneal macrophages of both strains was compared (Fig. 1). The bacteria proliferated more than 10-fold in resident peri-
toneal macrophages of A / J mice, but bacterial growth was hardly detected in C 5 7 B L / 6 macrophages. In A / J alveolar macrophages the bacteria proliferated several-fold after 2 days of in vitro phagocytosis, in contrast to a decrease in bacterial numbers in C 5 7 B L / 6 alveolar macrophages.
4.2. Strain differences in the susceptibilities of A / J and C57BL / 6 mice to L. pneumophila Peritoneal 2
mecrophages
*
• ,
Alveolar
macrophages
2. A/J
'o
'
.J .2,
3
.3 0
1
2
3
Days of in vitro culture
Fig. 1. Differences in intracellular L. pneumophila growth in A/J (©) and C57BL/6 (e) macrophages. Bacterial burdens in cultured macrophages are expressed as logt0 increases in CFU from day 0 to day 3 of in vitro culture. The data are means + standard deviations of representative data from three individual experiments in which four mice were used in each. The asterisks * and ** indicate P<0.05 and P<0.01, respectively, significant differences between A/J and C57BL/6.
In order to compare the susceptibilities of A / J and C 5 7 B L / 6 mice to L. pneumophila, mice were injected i.v. with different doses of the bacteria. Their survival was observed for 10 days after injection (Fig. 2). When mice were injected with 4.7 x 107 CFU of L. pneumophila, the difference of the survival rate between the two strains was observed, i.e., 80% of C 5 7 B L / 6 mice survived in comparison with no survivors in A / J mice. All mice surviving on day 3 after inoculation did not die thereafter. The LDs0 values of the bacteria for A / J and C 5 7 B L / 6 mice were 2.7 × 107 CFU and 7.2 × 10 7 CFU, respectively,
4.3. Bacterial clearance of L. pneumophila from the bloodstream of A / J and C57BL / 6 mice Clearance rates of bacteria from the bloodstream were compared in the two strains (Fig. 3). Blood samples were obtained 1 min, 1, 3, and 5 h
187 5.5.
i'°[
M.
4.5'
o
o
.J
3.5
30
Hours after infection Fig. 3. Bacterial clearance from the bloodstream of A / J (©) and C57BL/6 (e) mice..After the mice were injected i.v. with
injected with a bacterial dose far below the LDs0 (3.0 × 105 CFU), the number of bacteria in the organs of both A / J and C 5 7 B L / 6 decreased at a similar and considerably rapid rate by day 1 after infection (data not shown). Because A / J mice are C5-deficient, they were passively transferred with C5-containing serum of ddY mice in order to examine the possibility that C5 plays some roles in the resistance. A / J mice which were injected i.v. with 0.5 ml of C5-containing serum per mouse 1 h before infection did not improve their ability to eliminate the bacteria in the livers, spleens and lungs (data not shown).
4× 10 6 CFU of L. pneumophila, blood samples were collected from the retro-orbital plexus and the bacterial CFU/ml were counted. The data are means± standard deviations of representative data from three individual experiments in which three mice were used in each.
Liver
u
after the bacterial injection. Bacterial counts fell by half after 1 h, and by 90% after 5 h in both A / J and C 5 7 B L / 6 mice, indicating that bacterial uptake by organs within 5 h were similar in the two strains.
4.4. Strain differences in the growth of L. pneumophila in the livers, spleens, and lungs of infected mice Mice of A / J and C 5 7 B L / 6 strains were injected i.v. with a sublethal dose (4.0 × 106 CFU) of bacteria, and CFU in the organs were determined at intervals for comparison (Fig. 4). At one day after infection, the bacterial burden in the liver was significantly higher in A / J than in C 5 7 B L / 6 ( P < 0 . 0 5 ) . Thereafter, the bacteria count in the livers of both strains decreased at a rapid and similar rate until the bacterium became undetectable ( < 102 CFU per organ). E R 1 - 3 was 1.28 for A / J and 1.00 for C 5 7 B L / 6 . In the spleen, the elimination rate of the bacteria was slower in A / J ( E R 0 - 3 = 0.63) than in C 5 7 B L / 6 ( E R 0 - 3 = 1.83) mice, although the increase of the bacterial number was not seen in A / J . The bacterial elimination in lungs of A / J and C 5 7 B L / 6 were similar to that in the spleen (data not shown). When both strains of mice were
(J o
& 0 ..J
1
-
i
-
1
i
2
.
i
3
•
i
4
•
i
•
i
5
•
i
6
7
Days after Infection 6
Spleen
1
2
3
4
5
6
7
Days after Infection Fig. 4. Differences in the growth of L. pneumophila in the livers and spleens of A/J ((3) and C57BL/6 (e) mice after the mice were injected i.v. with 4 × 10 6 CFU of the bacteria. The data are means+ standard deviations for three animals. The asterisks * and ** indicate P < 0.05 and P < 0.01, respectively, significant differences between A/J and C57BL/6.
188
4.5. In vitro production of IFN-T in response to killed bacteria To compare the abilities of IFN-y production in A / J and C57BL/6 strains, spleen cell cultures of non-immunized mice were stimulated in vitro with formalin-killed L. pneumophila for 18 h and IFN-y activity in the culture supernatant fluids was assayed. The amount of IFN-T produced by splenocytes of A / J was 467.9 _+ 169.1 I U / m l and that of C57BL/6 was 655.5 +__63.0 I U / m l (mean +standard deviations for four animals). Although the amount of IFN-y production by A / J splenocytes is somewhat smaller than that of C57BL/6, the difference was not significant (P = 0.067). 4.6. The effect of anti-IFN-7 mAb administration on the resistance of mice against L. pneumophila infection Mice of either strain were injected i.v. with rat anti-mouse IFN-y mAb (30 ~g) or normal rat globulin (30 tzg) 2 h before infection with a sublethal dose (4 × 106 CFU for A / J and 1 x 107 CFU for C57BL/6) of L. pneumophila. The number of L. pneumophila cells in livers, spleens and lungs of treated mice was determined on day 1, 4
Spleen
Liver
..
6
Anti-IFNy
s
4- ~
4
.J
Spleen
Liver
5
~
B
AnI-iIFN-7
:1
........... ::::
Days after infection Fig. 6. The effect of anti-IFN-7 m A b administration on the resistance of C 5 7 B L / 6 mice against L. pneumophila infection. C 5 7 B L / 6 mice were pretreated with anti-IFN-7 m A b (e) or normal rat globulin (©), and injected i.v. with 1 × 107 C F U of L. pneumophila. T h e bacterial growth in the organs was determined at intervals. The data are m e a n s 5: standard deviations for three animals. * indicates P < 0.05 with respect to control mice.
and 7 of infection in A / J (Fig. 5), or on day 1, 3 and 5 of infection in C57BL/6 (Fig. 6). The bacterial number in the organs of A / J mice pretreated with anti-IFN-y mAb significantly increased as compared with those of control mice until day 4 of infection. The number of bacteria in the organs of C57BL/6 mice pretreated with anti-IFN-y mAb increased significantly on day 1 of infection. Data of the bacterial elimination from lungs of both strains adminstered with antiIFN-y mAb or normal rat globulin showed similar tendencies to those seen in the spleens (data not shown).
3"
2
0
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2-
I 2 3 4 5 6 7 ,
.
,
-
,
-
,
-
,
-
,
"
,
Days after infection Fig. 5. T h e effect of anti-IFN-7 m A b administration on the resistance of A / J mice against L. pneumophila infection. A / J mice were pretreated with anti-IFN-3, m A b (e) or normal rat globulin (©), and injected i.v. with 4 x 106 C F U of L. pneumophila. T h e bacterial growth in the organs was determined at intervals. The data are m e a n s + s t a n d a r d deviations for three animals. * indicates P < 0.05 with respect to control mice; * * indicates P < 0.01.
5. DISCUSSION As far as we know, there is no report which has evaluated the role of macrophages and endogenous IFN-y in natural resistance of mice against L. pneumophila infection. Our results made clear the following: strain differences in the macrophage ability to inhibit the intracellular bacterial growth (Fig. 1) had some effect on their susceptibilities to the infection (Fig. 2); the LD50 value of A / J mice (2.7 × 107 CFU) was greater
189
than the LDs0 for guinea pigs (7.6 x 10 4 CFU) whose macrophages also permit the intracellular bacterial growth [7]; even though the A / J macrophages permitted intracellular bacterial growth in vitro, the bacteria could be eliminated in the liver and spleen (Fig. 4); and there was a significant increase of bacterial burden in the organs of A / J (Fig. 5) and C57BL/6 (Fig. 6) when mice were pretreated with anti-IFN-7 mAb suggesting that endogenous IFN-7 acts as an important protective factor in a very early phase of Legionella infection in mice. Splenocytes of A / J and C57BL/6 mice could produce a large amount of IFN-7 within 18 h of formalin-killed L. pneumophila stimulation in vitro, i.e. 467.9 ___169.1 for A / J and 655.5 + 63.0 for C57BL/6 (IU/ml). Although the amount of IFN-7 production by A / J splenocytes was somewhat smaller than that of C57BL/6, the difference was not significant (P = 0.067). Both strains of mice seem to be able to produce just enough IFN-7 for augmentation of the protection in the very early phase of infection with L. pneumophila. Results of studies using anti-IFN-7 mAb showed that the pretreatment with anti-IFN-7 was effective only on day 1 of infection in C57BL/6 mice (P values for both livers and spleens were < 0.05, see Fig. 6), while in A / J mice it was effective until day 4 of infection (both P values < 0.01, see Fig. 5). From the above, it is indicated that endogenous IFN-7 had a relatively small influence on the resistance of C57BL/6 mice whose macrophages are innately non-permissive for the intracellular bacterial growth. It is known that IFN-7, by activating macrophages, plays an important role in host defense against many bacteria [reviewed by Murray, 19]. IFN-7 was produced in the early phase of listerial infection in mice [20]. Blanchard et al. reported that in vitro production of IFN-7 from mouse splenocytes could be detected as early as 6 h after stimulation with the L. pneumophila antigen, continued to increase for up to 24 h after stimulation and thereafter almost completely ceased [13]. The same group showed that macrophages of A / J mice pretreated with IFN-y could restrict the intracellular growth of L. pneumophila in vitro [10].
The natural resistance of A / J mice to L. pneumophila will be well understood by comparing it with that of guinea pigs. In spite of the evidence that macrophages of both A / J mice and guinea pigs permitted the intracellular growth of L. pneumophila [8,9], LDs0 of L. pneumophila for A / J mice was 2.7 x 107 CFU which was still higher than 7.6 x 104 CFU of guinea pigs [7]. The proliferative response of spleen cells against Legionella vaccine did not significantly increase until day 6 post-infection in immunized guinea pigs [21], while splenocytes from even non-immunized A / J mice proliferated to a reasonable extent in response to the antigen [10]. After several hours of culture with Legionella antigen, A / J splenocytes could produce IFN-7 in vitro, while macrophage activating factor (IFN-7) could not be produced by normal guinea pig splenocytes by in vitro antigen stimulation. This factor could be produced in guinea pigs after 4 days of the infection with L. pneumophila [21]. From the above comparison, a rapid T cell response and rapid IFN-7 production are thought to be the reasons why A / J are very resistant to Legionella infection. It was suggested that mouse natural killer (NK) cells were responsible for in vitro production of IFN-7 in response to L. pneumophila [22] or L. monocytogenes [23] antigen. Dunn et al. showed in vivo that early elimination of NK cells by treatment with rabbit antiserum to asialo GM1 resulted in exacerbation of listerial infection in mice [24]. We, however, could not observe the role of NK cells in L. pneumophila infection by injecting antiserum to asialo GM1 in vivo (unpublished data). In addition, administration of antiL3T4 (CD4) mAb or anti-Lyt 2.2 (CD8) mAb [25], or combined administration of these three antibodies (including anti-asialo GM1) could not influence the murine resistance against L. pneumophila (unpublished data). Cell population responsible for the production of IFN-7 in the course of infection is yet to be determined. Although Yamamoto et al. have reported that L. pneumophila did not grow in resident peritoneal macrophages of A / J mice within two days after in vitro phagocytosis [9], we could observe the bacterial growth in these macrophages. The
190
discrepancy would be due to the difference of L.
pneumophila strains used. In addition, they did not determine bacterial growth at day 3 of culture, at which we found the significant intracellular bacterial growth. In the murine protection system against L. pneumophila, there are two critical factors: the ability of innate macrophages to inhibit the intracellular bacterial growth, and the enhancement of resistance by IFN-y . Although the question of the identity of the cell population responsible for endogenous IFN-3, production in vivo remains, we insist that these two factors are important in the natural resistance.
ACKNOWLEDGEMENTS This study was supported by Grant-in Aid for Scientific Research 03670229 from the Ministry of Education, Science and Culture, Japan.
[8]
[9]
[10]
[11]
[12]
[13]
REFERENCES [14] [1] Fraser, D.W., Tsai, T.R., Orenstein, W., Parkin, W.E., Beecham, H.J., Sharrar, R.G., Harris, J., Mallison, G.F., Martin, S.M., McDade, J.E., Shepard, C.C. and Brachman, P.S. (1977) Legionnaires' disease, description of an epidemic of pneumonia. N. Engl. J. Med. 297, 1189-1197. [2] Horwitz, M.A. and Silverstein, S.C. (1980) Legionnaires' disease bacterium (Legionella pneumophila) multiplies intracellularly in human monocytes. J. Clin. Invest. 66, 441-450. [3] Nash, T.W., Libby, D.M. and Horwitz, M.A. (1984) Interaction between the Legionnaires' disease bacterium (Legionella pneumophila) and human alveolar macrophages. J. Clin. Invest. 74, 771-782. [4] Horwitz, M.A. (1983) The Legionnaires' disease bacterium (Legionella pneumophila) inhibits phagosomelysosome fusion in human monocytes. J. Exp. Med. 158, 2108-2126. [5] Horwitz, M.A. and Max_field, F.R. (1984) Legionella pneumophila inhibits acidification of its phagosome in human monocytes. J. Cell. Biol. 99, 1936-1943. [6] Sahney, N.N., Lambe, B.C., Summersgill, J.T. and Miller, R.D. (1990) Inhibition of polymorphonuclear leukocyte function by Legionella pneumophila exoproducts. Microb. Pathogen. 9, 117-125. [7] Yoshida, S. and Mizuguchi, Y. (1986) Multiplication of Legionella pneumophila Philadelphia-1 in cultured peri-
[15]
[16]
[17]
[18]
[19]
[20]
toneal macrophages and its correlation to susceptibility of animals. Can. J. Microbiol. 33, 438-442. Yoshida, S., Mizuguchi, Y., Nikaido, Y., Mitsuyama, M. and Nomoto, K. (1987) Fate of Legionella pneumophila Philadelphia-1 strain in resident, elicited, activated and immune peritoneal macrophages of guinea pigs. Infect. Immun. 55, 2477-2482. Yamamoto, Y., Klein, T.W., Newton, C.A., Widen, R. and Friedman, H. (1988) Growth of Legionella pneumophila in thioglycolate-elicited peritoneal macrophages from A / J mice. Infect. Immun. 56, 370-375. Klein, T.W., Blanchard, D.K., Yamamoto, Y., Newton, C., Widen, R. and Friedman, H. (1988) Role of cytokines in resistance to infection with Legionella pneumophila. Adv. Biosci. 68, 259-265. Byrd, T.F. and Horwitz, M.A. (1989) Interferon gammaactivated human monocytes downregulate transferrin receptors and inhibit the intracellular multiplication of Legionella pneumophila by limiting the availability of iron. J. Clin. Invest. 83, 1457-1465. Nash, T.W., Libby, D.M. and Horwitz, M.A. (1988) IFN3,-activated human alveolar macrophages inhibit the intracellular multiplication of Legionella pneumophila. J. Immunol. 140, 3978-3981. Blanchard, D.K., Klein, T.W., Friedman, H. and Stewart II, W.E. (1985) Kinetics and characterization of interferon production by murine spleen cells stimulated with Legionella pneumophila antigens. Infect. Immun. 49, 719723. Yoshida, S., Ogawa, M. and Mizuguchi, Y. (1985) Relation of capsular materials and colony opacity to virulence of Vibrio vulnificus. Infect. Immun. 47, 446-451. Igarashi, K., Mitsuyama, M., Muramori, K., Tsukada, H. and Nomoto, K. (1990) Interleukin-l-induced promotion of T-cell differentiation in mice immunized with killed Listeria monocytogenes. Infect. Immun. 58, 3973-3979. Bruck, C., Portelle, D., Glineur, C. and Bollen, A. (1982) One-step purification of mouse monoclonal antibodies from ascitic fluid by DEAE Affi-Gel Blue chromatography. J. Immunol. Methods 53, 313-319. Havell, E.A. (1986) Purification and further characterization of anti-murine interferon-3, monoclonal neutralizing antibody. J. Interferon Res. 6, 489-497. Nakane, A., Minagawa, T., Kohanawa, M., Chen, Y., Sato, H., Moriama, M. and Tsuruoka, N. (1989) Interactions between endogenous gamma interferon and tumor necrosis factor in host resistance against primary and secondary Listeria monocytogenes infections. Infect. Immun. 57, 3331-3337. Murray, H.W. (1988) Interferon-gamma, the activated macrophages and host defense against microbial challenge. Ann. Intern. Med. 108, 595-608. Nakane, A., Numata, A., Asano, M., Kohanawa, M., Chen, Y. and Minagawa, T. (1990) Evidence that endogenous gamma interferon is produced early in Listeria monocytogenes infection. Infect. Immun. 58, 2386-2388.
191 [21] Nikaido, Y., Yoshida, S., Goto, Y., Mizuguchi, Y. and Kuroiwa, A. (1989) Macrophage-activating T-cell factor(s) produced in an early phase of Legionella pneumophila infection in guinea pigs. Infect. Immun. 57, 3458-3465. [22] Blanchard, D.K., Friedman, H., Stewart II, W.E., Klein, T.W. and Djeu, J.Y. (1988) Role of gamma interferon in induction of natural killer activity by Legionella pneumophila in vitro and in an experimental murine infection model. Infect. Immun. 56, 1187-1193. [23] Wherry, J.C., Schreiber, R.D. and Unanue, E.R. (1991) Regulation of gamma interferon production by natural
killer cells in scid mice: roles of tumor necrosis factor and bacterial stimuli. Infect. Immun. 59, 1709-1715. [24] Dunn, P.L. and North, R.J. (1991) Early gamma interferon production by natural killer cells is important in defense against murine listeriosis. Infect. Immun. 59, 2892-2900. [25] Sasaki, T., Mieno, M., Udono, H., Yamaguchi, K., Usui, T., Hara, K., Shiku, H. and Nakayama, E. (1989) Roles of CD4 + and CD8 ÷ cells, and the effect of administration of recombinant murine interferon-~, in listerial infection. J. Exp. Med. 171, 1141-1154.