Zbl. Bakt. Hyg., I. Abt. Orig. A 255,145,...149 (1983)
Proteins and Antibody in Lavage Fluid of Guinea Pigs with Legionella pneumophila Pneumonia D.W. GUMP, G. S. DAVIS, W. C. WINN jr., and H. N. BEATY
Departments of Medicine and Pathology, University of Vermont College of Medicine, Burlington, Vermont 05405, USA
Abstract Guinea pigs were infected in an inhalation facility that limited an aerosol of L. pneumophila to the snout, as previously reported in detail (Davis et al., 1982). Individual animals were sacrificedfor study either immediately after exposure, at 16 hours, at days one through seven, or at 11 days. Bronchoalveolar lavage was carried out to obtain fluid to study the following: total protein, albumin and immunoglobulin G (IgG) concentrations, and the titer of antibody to L. pneumophila. Antibody also was measured in serum obtained at the time of sacrifice. Concentrations of total protein, albumin, and IgG in lavage fluids peaked 2 days after exposure and correlated with the appearance of maximal numbers of polymorphonuclear cells in the lung. Presumably, this increased protein resulted from exudation of serum across the alveolarcapillary membrane, which loses its integrity secondary to pneumonia. However, the ratio of IgGjalbumin was elevated in animals studied 11 days after exposure even though the concentration of albumin was normal by this time. One possible explanation for this observation is that IgG was being produced in the lung. Antibody in lavage fluid was detected 7, and 11 days post-exposure, and might be important in the recovery of guinea pigs from this infection.
Keywords: Legionnaires' disease, Legionella pneumophila, bronchoaiveloar lavage, antibody, guinea pig. Introduction
We have developed an aerosol model of Legionnaires' disease in the guinea pig which simulates many aspects of the human disease caused by 1. pneumopbila (Davis et aI., 1982). In this model the Burlington 1 strain of 1. pneumophila, a representative of serogroup 1, produced infection in 100% and mortality in 56% of the animals exposed. Animals are exposed to a nose-only inhalation facility for 30 minutes and the lungs contained from 103 to 104 1. pneumophila immediately after exposure. Initially, a diffuse, patchy pneumonia developed, which eventually coalesced and consolidated. Rapid initial intrapulmonary growth suggested that resident lung defense mechanisms are quite ineffective against 1. pneumophila, and that recruited or immunospecific defenses may be more critical in the outcome of 10 Zbl. Bakt. Hyg., I. Abt. Orig. A 255
D.W.Gump, G.S.Davis, W.C.Winn jr., and H.N.Beaty
infection. This paper presents data on total protein, IgG, and albumin concentrations in lavage fluids as well as the appearance of antibody in lavage fluid and serum.
Methods Bronchoalueolar lavage Animals were lightly anesthetized with pentobarbital, the trachea was exposed in the neck, a polyethylene catheter tied in place in the trachea, the chest opened by midsternal incision, and the animal sacrificed by right ventricular puncture and exsanguination. Serum was frozen for quantitation of antibody. Lavage was performed by the instillation and immediate withdrawal of 10 boluses of 10 ml of phosphate buffered saline solution (PBS), ph 7.4, at room temperature. From 40 to 96 ml of lavage fluid was recovered from each animal and kept at 5 °C during further manipulations. Cells were separated from the fluid by centrifugation at 200 g for 15 min and the supernatant was frozen at - 30°C for future analysis. Concentration Procedure
All lavage fluids were filtered through a 0.45 micropore filter to remove mucus and bacteria. A 5 ml aliquot was reserved for protein and albumin determination and the remainder was concentrated, 12 to 53 fold by pressure filtration through an Amicon YM-I0 membrane (Amicon Corp., Lexington, MA). Total protein concentrations were determined on unconcentrated lavage fluids using the technique of chemiluminescence (Ward, Owens, and Rennie, 1980). Albumin and IgG concentrations were quantitated by the radial immunodiffusion technique of Mancini et al. (1965). Affinity-purified guinea pig albumin and IgG were obtained from Cappel Laboratories, Cochranville, MD, as were the antisera directed against them. Preliminary experiments were carried out to determine optimum concentrations of antiserum and incubation times. Subsequently, plates were incubated at room temperature for 48 and 96 h for albumin and IgG determinations respectively. Albumin concentrations were determined on unconcentrated lavage fluids, but all determinations of IgG concentrations were performed on concentrated lavages. Antibody activity against the infecting strain of L. pneumophila was measured in sera and concentrated lavage fluids using the technic of microagglutination (MA) as developed by Farshy, Klein, and Feeley (1978). Sera were tested at an initial 1:2 dilution, whereas lavage fluids were assayed undiluted and at serial two-fold dilutions. Seroconversion was defined as a 4-fold increase in titer. Lavage fluids with a titer of 1: 2 or greater were considered positive for antibody.
Results Table 1 gives the mean values for protein, albumin, and IgG concentrations, all of which peaked two days post-exposure. Albumin concentration always was less than 50% of the protein concentration, and IgG concentration never exceeded 3% of the protein concentration. Also shown is the ratio of IgG to albumin concentrations at each time period. It is evident that this ratio rose rapidly, and at two and three days post-exposure was significantly higher than the ratio for the unexposed control animals. The very high mean IgG concentration and IgG/albumin ratio at 11 days is due to one animal with a high IgG concentration (10.8 mcg/ml), but a
Proteins and Antibody in Lavage Fluid of Guinea Pigs with L. pneumophila Pneumonia 147 Table 1. Protein (means in mcg/ml) Present in Lavage Fluid Time Post Exposure
Ratio IgG/Albumin a
Control" a hour 16 hours 1 day 2 days 3 days 4 days 6 days 7 days 11 days
8 8 3 14 5 6 2 2 4 2
77.3 82.6 77.1 149.5 428.8 224.0 150.0 187.5 121.8 306.3
26.1 22.7 23.6 50.6 123.3 69.8 45.0 51.7 58.8 27.8
0.6 0.6 0.5 1.7 9.5 5.9 2.2 3.2 3.6 6.1
0.02 0.03 0.02 0.03 0.08 c 0.09 c 0.05 0.06 0.06 0.21
Data shown as mean ratio of IgG concentration/albumin concentration at each time. Animals not exposed to infectious aerosol. c Values significantly different from unexposed control (IX = 0.05) by one-way analysis of variance with the Newman-Keuls test for differences between groups. a
normal albumin concentration (30.5 mcg/ml). These determinations were repeated and were consistent. Least-squares linear regression correlation coefficients showed significant relationships between protein and albumin concentrations (r = 0.83), protein and IgG concentrations (r = 0.75), and albumin and IgG concentrations (r = 0.79). No significant relationships were evident between these proteins and the numbers of total cells or bacteria recovered by lavage (data on cells and bacteria not shown) . Albumin concentration and the percentage of neutrophiIs in lavage fluid peaked simultaneously 2-3 days post-exposure and were significantly correlated (r = 0.64). Table 2 depicts the antibody data for all animals which were positive. No antibody was detected until six days post-exposure, and all animals were sero-negative prior to exposure. By six days post-exposure all animals had demonstrable serum
Table 2. Relationship of Lavage and Serum Antibodies
Days Post Exposure
Reciprocal MA titers Lavage Serum
54-11 54-17 53-7 53-18 54-3 54-10 53-2 53-9
6 6 7 7 7 7 11 11
< 1 (28)a
Concentration factor of lavage fluid.
1 2 8 16 2 64 16
(28) (21) (34) (39) (39) (37) (53)
4 32 64 64 256 128 512 256
D.W.Gump, G.S.Davis, w.e. Winn jr., and H.N.Beaty
antibody, and by seven days all animals demonstrated antibody in lavage fluid as well. It is interesting that the animal (53-2) with the highest titer present in lavage fluid, also had the highest ratio of IgG/albumin (0.35) of any animal. Discussion We have utilized the guinea pig exposed to an aerosol of L. pneumopbila as a model of Legionnaires' disease to study the pathogenesis of this infection. We previously published results indicating that 100% of animal exposed to such aerosols become ill and seroconvert by one week (Davis et al., 1982). Thus, L. pneumopbila infection is a potent stimulus for the formation of serum antibody in this model. In the present publication we have shown that antibody also appears in lavage fluid about one week after infection. Protein, albumin and IgG concentrations peaked 2-3 days after aerosol exposure. We have shown previously (Davis et al., in press) that maximal numbers of polymorphonuclear leukocytes were present in lavage fluid 2-3 days after infection, and presumably reflect the marked degree of inflammation present at that time. We hypothesize that the increased concentration of proteins in lavage fluid originate from the blood and reflect a breakdown in the integrity of the alveolarcapillary membrane permeability barrier. Hill et al. (1980) showed that albumin in lavage fluid accurately reflected the degree of inflammation caused in dog lungs by the instillation of sheep red blood cells. Similarly, Pennington and Kuchmy (1980) demonstrated the rapid appearance of antibodies derived from the bloodstream in bronchial fluids of animals with pseudomonas pneumonia. We observed an approximate four-fold increase in the IgG/albumin ratio from baseline levels by the second post-exposure day. This increase in the ratio of IgG/albumin from 0.02 to 0.08 is consistent with increased alveolar-capillary permeability since the ratio of IgG/albumin in normal guinea pig serum is approximately 0.3 (Altman and Katz, 1979; Reid, 1958). However, even at a time when the pneumonia demonstrated the maximum inflammatory response as reflected by histology (Davis et al., 1982), albumin concentrations, and appearance of polymorphonuclear cells in the lavage fluid (Davis et al., in press) the IgG/albumin ratio was still considerably below that of serum. It is noteworthy that the elevated IgG/albumin ratio persisted 11 days after exposure, at a time when albumin concentrations had returned to normal levels. This may suggest that local production of IgG was occurring or could reflect continuing alteration in the permeability of the alveolar-capillary membrane despite the normal albumin concentrations. Local production of antibody within the lung has been demonstrated by a number of investigators (Johnson, Chapman, and Ward, 1979), but has not been studied in L. pneumophia infection. Additional experiments would be required to document local production of antibody definitively. One approach would be to expose guinea pigs to an aerosol of heat-killed L. pneumophila organisms, which we have previously shown does not result in the development of serum antibodies (Davis et al., 1982). Antibody occurring in the lavage fluids of such animals would considerably strengthen the argument for the local production of antibody. The fact that antibody is detectable in lavage fluid as early as it is in serum, suggests that this antibody may have an important role in controlling the infection within the lung. It is known that serum antibody potentiates phagocytosis of L.
Proteins and Antibody in Lavage Fluid of Guinea Pigs with L.pneumophila Pneumonia 149
pneumophila by alveolar macrophages (Johnson, Pesanti, and Elliot, 1979; Davis and Winn, 1979), and may (Arko, Wong, and Feeley, 1979) or not (Horwitz and Silverstein, 1980) be bactericidal. Whether local antibody would be more effective in protecting animals from an aerosol challenge of viable 1. pneumophila bacteria than serum antibody, as has been demonstrated by Jakob and Green (1973) for Proteus mirabilis, is an intriguing but unanswered question. Acknowledgements: This work was supported by grant ROI-AI15903 from the National Institute of Allergy and Infectious Diseases. The authors wish to thank Peter Durda, Brooks Gump, and Nancy Kearney for their technical assistance, and Word Processing Center for preparing the manuscript.
References Altman, P. L. and D. D. Katz: Inbred and Genetically Defined Strains of Laboratory Animals, Part 2. Federation of American Societies for Experimental Biology, Bethesda, Maryland (1979) Arko, R. J., K. H. Wong, and j. C. Feeley: Immunological factors affecting the in vivo and in vitro survival of Legionnaires' disease bacterium. Ann. intern. Med. 90 (1979) 680-683 Davis, G. S. and W. C. Winn: Phagocytosis of Legionnaires' disease bacillus by human alveolar macrophages. Clin, Res. 27 (1979) 589 Davis, G.S., W. C. Winn, [r., D. W. Gump, J. c. Craighead, and H. N. Beaty: Legionnaires' pneumonia after aerosol exposure. Amer. Rev. resp. Dis. 126 (1982) 1050-1057 Davis, G.S., W. C. Winn, jr., D. W. Gump, and H. N. Beaty: The kinetics of early inflammatory events during experimental Legionella pneumophila pneumonia in guinea pigs. J. infect. Dis. (in press) Farshy, C. E., G. C. Klein, and j. C. Feeley: Detection of antibody to Legionnaires disease organism by micro agglutination and micro-enzyme linked immunosorbant assay tests. J. Clin. Microbiol, 1 (1978) 327-331 Hill, j. 0., D. E. Bice, D. L. Harris, B. A. Muggenburg, and D. G. Brownstein: Immunochemical analysis of bronchoalveolar lavage fluid: albumin as an indicator of local inflammation and as a tool measuring locally produced components in lung surface fluids. Amer. Rev. resp. Dis. 121 (1980) 73 Horwitz, M. A. and S. C. Silverstein: Interaction of the Legionnaires' disease bacterium (Legionella pneumophila) with human phagocytes. II Antibody promotes binding of L. pneumophila to monocytes but does not inhibit intracellular multiplication. J. expo Med. 153 (1980) 398-406 jakob, G. J. and G. M. Green: Immune enhancement of pulmonary bactericidal activity in murine virus pneumonia. J. Clin. Invest. 52 (1973) 2878-2884 johnson, K. J., W. E. Chapman, and P. A. Ward: Immunopathology of the lung. Amer. J. Path. 95 (1979) 795-840 johnson, W., E. Pesanti, and j. Elliott: Serospecificity and opsonic activity of antisera to Legionella pneumophila. Infect. Immun. 26 (1979) 698-704 Mancini, G., A. O. Carbonara, and j. F. Heremans: Immunochemical quantitation of antigens by single radial immunodiffusion. Immunochemistry 2 (1965) 235-254 Pennington, j.E. and D.Kuchmy: Mechanism for pulmonary protection by lipopolysac. charide pseudomonas vaccine. J. infect. Dis. 142 (1980) 191-198 Reid, M.E.: The Guinea Pig in Research. Human Factors Research Bureau, Inc., Washington, D. C. (1958) Ward, M. W. N., C. W. I. Owens, and M. J. Rennie: Nitrogen estimation in biological samples by use of chemiluminescence. Clin. Chern. 26 (1980) 1336-1339