INFECTION AND IMMUNITY, Mar. 1979, p. 737-742
Vol. 23, No. 3
0019-9567/79/03-0737/06$02.00/0
Interaction of Polymorphonuclear Leukocytes with Smooth and Rough Strains of Brucella abortus DONALD L. KREUTZERt LAWRENCE A. DREYFUS, AND DONALD C. ROBERTSON* Department of Microbiology, University of Kansas, Lawrence, Kansas 66045 Received for publication 8 December 1978
The bactericidal activity of guinea pig and human polymorphonuclear leukocytes (PMNs) against a smooth-intermediate strain (45/0) and a rough strain (45/20) of Brucella abortus has been examined. After incubation for 120 min, guinea pig PMNs incubated with either the smooth strain 45/0 or the rough strain 45/20 exhibited no bactericidal activity against the former and caused only a 34% decrease in viability of the latter. Human PMNs were more bactericidal than guinea pig PMNs to both strains; however, the killing of strain 45/20 by human PMNs was less than that observed in control experiments with S. aureus strain 502A. Both strains of B. abortus readily associated with guinea pig and human PMNs, and the bacteria were apparently ingested without stimulation of the hexose monophosphate pathway. Lysates (10 pg/ml, pH 5.5), prepared from guinea pig or human granules, were not particularly toxic to either strain unless supplemented with H202 and a halide (I- or Cl-). An oxygen-dependent killing system appeared to be lethal against both strains of B. abortus, with I- being more active than Cl- in the presence of H202 and granule lysate. The data suggest that degranulation after ingestion of Brucella by phagocytes does not occur due to the lack of a proper stimulus or possibly the baccilli actively inhibit the degranulation process thereby protecting the microbe from killing systems normally effective against extracellular parasites.
Facultative intracellular parasites (e.g., Mycobacterium tuberculosis, Brucella, and Listeria monocytogenes) survive conditions within phagocytic cells where most pathogens are killed. Their survival is likely influenced by the surface properties of bacilli, but it is not clear as to which surface components can be considered virulence factors except for the sulfolipids of the tubercle bacilli which inhibit degranulation after ingestion by monocytes (8, 9). Virulence factors of intracellular parasites play at least two possible roles: (i) provide protection from the hostile environment within the phagolysosome and (ii) inhibit or not stimulate the sequence of events after ingestion which would normally be bactericidal. It is difficult to compare virulence factors important in the pathogenesis of Mycobacteria with Brucella, the latter being a gram-negative rod with very different surface properties. Therefore, our approach has been to compare the surface macromolecules of two closely related strains of Brucella abortus that differ significantly in virulence and infectivity. The rough strain, B. abortus 45/20, was derived from the t Present address: Department of Pathology, University of Connecticut Health Center, Farmington, CT 06032.
smooth-intermediate strain, 45/0 (14). The chemical composition of cell walls of both strains has been examined with respect to total lipids, proteins, peptidoglycan content, and composition, as well as lipopolysaccharides (LPS) (14, 15). The only significant difference was in the LPS composition; that is, when the two strains were extracted by the Westphal procedure, LPS was found in the phenolic and aqueous phases of B. abortus strain 45/0, but only in the aqueous phase of B. abortus strain 45/20. The phenolsoluble LPS was the only fraction toxic to mice and contained low levels of heptose and dideoxyaldoses, which may suggest a unique core region and 0-specific side chain (15). It is equivocal whether the phenolic-phase LPS can be considered a virulence factor; however, it is the only surface difference observed in comparison studies of the two strains. We chose to examine the interaction of smooth and rough strains of B. abortus with polymorphonuclear leukocytes (PMNs) for two reasons: (i) the PMN is likely the first phagocyte which the bacteria encounter, especially if infection occurs through a cut or the conjunctival route, and (ii) it was of interest to ascertain whether the surface differences detected in B. 737
738
KREUTZER, DREYFUS, AND ROBERTSON
INFECT. IMMUN.
Assay of HMP activity. For the assay of hexose abortus strain 45/0 and 45/20 were related to enhanced intracellular survival of one strain over monophosphate (HMP) activity, the reaction mixture contained a portion of the PMN suspension, 3 x 107; the other. bacteria, 3 X 109 colony-forming units; homologous MATERLALS AND METHODS heat-inactivated serum, 10%; 0.1 to 0.2 JACi of DSource of bacteria. A smooth-intermediate strain, [1-'4C]glucose and HBSS in a final volume of 3 ml. B. abortus 45/0, was obtained from M. Meyer, Uni- The 25-ml Erlenmeyer reaction flask was sealed with versity of California, Davis, Calif., and the rough a serum stopper containing a plastic well trap. A 0.3strain, B. abortus 45/20, was provided by B. L. Deyoe, ml amount of ethanolamine-methyl cellosolve (1:2, National Animal Disease Center, Ames, Iowa. Cul- vol/vol) was added to the trap, and the reaction flasks tures were maintained and grown as previously de- were incubated with shaking at 370C for 60 min. The scribed as was the history of these established smooth reactions were stopped by injection of 0.1 ml of 5 N HCL The contents of the trap were transferred to and rough laboratory strains (1). Isolation of PMNs. PMNs were purified from he- minivials containing XDC counting solution (4) for parinized human blood by the technique of Boyum (2, determination of radioactivity. Preparation of granule lysates. PMNs were 20). Isolated PMNs were resuspended in Hanks balanced salt solution (HBSS) containing 0.1% gelatin washed once with 0.1 M Krebs-Ringer phosphate and stored at 4°C until used. Leukocytes were also buffer, pH 7.0, and suspended in approximately 5 ml isolated from adult rats and guinea pigs after intraper- of 0.25 M sucrose containing 10 mM trisitoneal injection of sterile casein (10%, wt/vol). The (hydroxymethyl)aminomethane, pH 7.0, and 1 mM PMNs were removed 17 h after injection by lavage ethylenediaminetetraacetic acid. The suspension was with phosphate-buffered saline containing heparin (10 homogenized with a Teflon homogenizer at room temU/ml). The cells were washed once with phosphate- perature until greater than 90% cell lysis had occurred. buffered saline and once with HBSS and then sus- The cell homogenate was centrifuged at 500 x g for 15 pended in HBSS at a known viable cell concentration min to remove intact cells and debris followed by centrifugation of the supernatant at 30,000 x g for 20 determined by trypan blue exclusion. Brucella-PMN association studies. The associ- min. The pellet containing the granules was resusation of Brucella with PMNs was determined with pended in 2 ml of 0.1 Krebs-Ringer phosphate buffer, bacteria labeled during growth with [14C]uracil (10 pH 5.5, and disrupted by homogenization with a Teflon mCi/ml). The radioactive bacteria were harvested by pestle followed by repeated freeze-thawing (6x) in an centrifugation and washed three times with HBSS acetone-dry ice bath (5). MPO activity of granule lysates. The myelobefore use. Incubation mixtures (3 ml) were prepared containing bacteria and PMNs at a 5:1 ratio with continuous gentle agitation with a New Brunswick (G76) gyratory water bath. At 30 min, 1.0-ml samples were removed from the reaction vessel and added to 9 ml of HBSS. After centrifugation at 150 x g, the cell pellet was resuspended in 10 ml of XDC counting solution containing 4% Cab-O-Sil (4), and a 0.5-ml portion of the supernatant was transferred to 10 ml of XDC counting solution. Radioactivity was determined with a Packard 3375 liquid scintillation spectrometer. The data were expressed as the percentage of the total radioactivity found in the supernatant and the pellet. Bactericidal activity of isolated leukocytes. The bactericidal activity of isolated PMNs was assayed with a bacteria-PMN ratio of 5:1. The components of the reaction mixture contained in siliconized 25-ml Erlenmeyer flasks included the following: PMN in HBSS (106 to 107), 1.0 ml; bacteria in HBSS, 1.0 ml; homologous heat-inactivated serum (56°C for 30 min), 0.3 ml; and HBSS (pH 7.2), 0.7 ml. All incubations were at 37°C with shaking, and at timed intervals, 0.1ml portions were removed and diluted into 9.9 ml of deionized water or 9.9 ml of phosphate-buffered saline containing Clostridium perfringens phospholipase c (100 ,g/ml) to effect leukocyte lysis. Dilutions were in 0.5% tryptose-0.5% NaCl broth and surviving bacteria were determined by plating 0.1-ml portions of the appropriate dilution on Trypticase soy agar plates (Baltimore Biological Laboratory). The results were similar by either method of leukocyte lysis. All data were expressed as colony-forming units versus time or as percent survival.
peroxidase (MPO) activity of granule lysates was determined by the method of Tagesson and Stendahl (27). The assay mixture contained guaiacol, 0.02 M; H202, 0.03 M; Krebs-Ringer phosphate buffer (pH 5.5), 0.1 M; and granule lysate in a final volume of 300 tl. The reaction was followed by the change in absorbance at 470 nm. One unit of MPO activity was defined as the amount of granule lysate required to effect a change of 0.001 absorbance units per min at 470 nm. The granule lysate was serially diluted in 0.1 M KrebsRinger phosphate buffer, pH 5.5, to the desired concentration. For studies on the effects of halides, 0.1 M phosphate buffer, pH 6.0, was substituted for the Krebs-Ringer phosphate buffer. All lysates were stored at -70°C until used. Brucellacidal activity of granule lysates. Incubation mixtures were prepared in 25-ml Erlenmeyer flasks and contained 0.1 M Krebs-Ringer phosphate, granule lysate at a final concentration of 10 to 100,ug of protein per ml, and bacteria (3 x 105 to 5 x 105 colony-forming units per ml) in a total volume of 3 ml. In some experiments, catalase (100 1ig) was added to the reaction mixture to ensure that there was little or no endogenous H202. Portions were removed after 30, 60, and 120 min of incubation at 370 C, diluted in 0.1 M phosphate buffer (pH 7.0), and plated in triplicate on Trypticase soy agar plates (Baltimore Biological Laboratory). The results were expressed as colonyforming units per milliliter versus time and as percent bacterial survival. To test oxygen-dependent killing systems in granule lysates, incubation mixtures contained 0.1 M KrebsRinger phosphate (pH 5.5), 3.3 x 10-5 M KI, 10-7 M
VOL. 23, 1979
739
PMNs AND B. ABORTUS
H202, granule lysate at various nonbactericidal concentrations, and bacteria at 3 x 105 to 5 x 105 colonyforming units per ml in a reaction volume of 3.0 ml. The mixtures were incubated with gentle shaking at 370C, and samples were withdrawn and plated as just described. The effect(s) of various halides on the oxygen-dependent killing of B. abortus was determined by substituting 0.1 M phosphate buffer, pH 6.0, for Krebs-Ringer phosphate and using either 10mM NaCl or 10 mM KI in separate reaction mixtures. The surviving bacteria were diluted for plating in 0.1 M phosphate buffer, pH 7.0, instead of the tryptose-NaCl broth. Again, the data were expressed as colony-forming units versus time or percent bacterial survival. RESULTS
u10
70 .
-J
60
50
4
40
60
a
30
0>20 >
° X
D
50
30
JI
so 90 120
I
W
40
J
Brucellacidal activity of guinea pig and human PMNs. The brucellacidal activities of guinea pig peritoneal and human peripheral leukocytes were compared. Guinea pig PMNs had no bactericidal activity against strain 45/0 and exhibited minimal killing of strain 45/20 (Fig. 1). Human peripheral leukocytes exhibited increased bactericidal activity compared with guinea pig PMNs against both strains of B. abortus, although B. abortus 45/0 was more resistant to killing than was B. abortus 45/20
tn
30
W
0 -
> 20 D W
p
p
0
100
30
60
90 (mins)
120
TIME FIG. 2. Bactericidal activity of human peripheral PMNs against B. abortus strains 45/0 and 45/20.
Strain 45/0, no PMNs (0); strain 45/20, no PMNs (0); PMNs + 45/0 (0); PMNs + 45/20 (-). Inset shows bactericidal activity against S. aureus 502A, bacteria, no PMNs (A); PMNs + S. aureus 502 (A).
90 /
to O
VI'/
X/
/
(Fig. 2). In parallel incubation mixtures, human PMNs displayed normal bactericidal activity
/J
against S. aureus 502A, which indicated that the leukocytes were not affected by Ficoll-HyW paque purification. The rqugh strain B. abortus aR (45/20 has a 1.7- to 2-fold greater generation time 60 \ than the smooth parent strain, 45/0 (D. L. and D. C. Robertson, unpublished a: wKreutzer O data). The basis for the observed difference in > growth rates is unknown; however, rough strains 50 50 . of gram-negative bacteria generally have longer D \ generation times than smooth bacilli (C. S. 70
/ -
/
Buller, personal communication).
The association of PMNs with B. abortus 45/ 0 and 45/20 was examined by using radioactive bacilli to determine whether the increased sur. . . vival of strain 45/0 was due to decreased ingesA tion. The smooth-intermediate strain (45/0) 0 60 120 30 90 readily associated with both guinea pig PMNs T I M E (mins) FIG. 1. Bactericidal activity of guinea pig PMNs and human PMNs after 30 min of incubation to against B. abortus strains 45/0 and 45/20. Strain the extent of 74 and 88%, respectively. The as45/0, no PMNs (0); 45/20, no PMNs (0); PMNs + sociation with the rough strain was slightly less, with 53% of the bacteria associating with guinea 45/0 (0); PMNs + 45/20 (U). 40
O
740
INFECT. IMMUN.
KREUTZER, DREYFUS, AND ROBERTSON
pig PMNs and 56% associating with human PMNs during a 30-min incubation. These data show that the increased survival of B. abortus 45/0 was not likely due to decreased ingestion, but to resistance of smooth Brucella to intracellular killing systems. Stimulation of the HMP pathway during phagocytosis of B. abortus 45/0 and B. abortus 45/20. Stimulation of the HMP shunt is a primary metabolic event associated with the bactericidal activity of PMNs (13, 22). Furthermore, measurement of the extent of D-[1-14C]glucose oxidation by PMN provides an indication of the bactericidal potential of the phagocytic cell (6, 23). As shown in Table 1, there was no measurable stimulation of D-[1-'4C]glucose oxidation by either guinea pig or human PMNs incubated with each strain of B. abortus. The lack of HMP pathway stimulation was apparently due to the unique surface properties of the bacilli because there was significant stimulation in PMNs exposed to heat-killed Salmonella typhimurium. One explanation for the observed lack of bactericidal activity toward injected Brucella was that the bacilli inhibited the normal metabolic activity after ingestion, especially the generation of H202. To test this, leukocytes were preincubated with either 45/0 or 45/20 before being exposed to heat-killed S. typhimurium. Preincubation of guinea pig PMN with whole cells and cell envelopes did not prevent [14C]_ glucose oxidation when the guinea pig PMNs were subsequently exposed to heat-killed S. typhimurium (Table 2). TABLE 1. Glucose oxidation by leukocytes in the presence and absence of bacterial stimulus -Fold stimulation of HMP pathway
Source of PMN
Guinea pig Human
rium (heat B. abortus B. abortus 45/0 45/20 killed) 2.9 1.1 1.3 2.5 0.8 1.0
TABLE 2. Glucose oxidation by leukocytes preincubated with B. abortus 45/0 and 45/20 and cell fractions followed by stimulation with S. typhimurium -Fold stimPreincubation condition
Brucellacidal activity of granule lysates. Because there appeared to be ingestion of both smooth and rough strains of B. abortus, but no metabolic stimulation usually associated with degranulation, the brucellacidal activity of granule lysates was examined. Guinea pig granule lysates (10 jig of protein per ml) were not bactericidal to B. abortus 45/0 at pH 5.5 but did kill 50% of the rough strain 45/20. Under the same incubation conditions, human PMN granule lysates (10 ,ug of lysate protein per ml, pH 5.5) were more bactericidal to both strains of Brucella, with 13.5% of strain 45/0 and 71% of strain 45/20 being killed during a 60-min incubation. Boiling the granule extract or addition of catalase abolished the brucellacidal activity of the granule lysates against both strains. These data suggested that an oxygen-dependent killing system was mainly responsible for the limited bactericidal activity of granule lysates. The bactericidal activity of an oxygen-dependent killing system effective against Brucella was assayed by adding H202 and halides (Cland I-) at nontoxic concentrations. In the presence of H202, Cl-, and I- (Fig. 3), both guinea pig and human granule lysates, 1 MPO unit per ml, caused a greater than 99.9% decrease in viability of both strains. Because the I--mediated bactericidal activity of MPO is known to be different from Cl--mediated killing, the effect of each halide on the bactericidal activity of
100
%I
4>
80 r W I-
z w u
60 40
Id
a.
20
elation of way
S. typhimurium (no preincubation) B. abortus 45/0 + S. typhimurium B. abortus 45/0 cell envelopes + S. typhimurium B. abortus 45/20 + S. typhimurium B. abortus 45/20 cell envelopes + S. typhimurium
2.9 2.5 2.0 2.4 4.2
0
15
30 TIME (mins)
FIG. 3. Brucellacidal activity of guinea pig granule lysates, granule lysate + I- + Cl- (0); H202 + I+ Cl- (0); granule lysate + H202 + I- + Cl- (A). The concentration of each reaction component is given in the text.
PMNs AND B. ABORTUS
VOL. 23, 1979
granule lysates was examined. The granule lysate v H202 IE system was extremely bactericidal to both strains; whereas, the granule lysate H202. C1 system was much less bactericidal (Table 3). In summary, the smooth strain of B. abortus 45/0 was more resistant to Clmediated killing than the rough strain 45/20; however, both strains were extremely sensitive to the presumed MPO-E-mediated killing activity. DISCUSSION PMNs constitute an important line of defense against microbial invaders and are likely the first phagocytic cells which interact with Brucella at the foci of infection. The inability of these cells to efficiently kill virulent Brucella at the primary site of infection is a key factor in the dissemination of Brucella to the regional lymph nodes and eventually infection of the reticuloendothelial system. Even though there have been numerous studies on the survival of Brucella within mononuclear cells (3, 7, 12, 17, 18, 24, 26), we chose to look at the interaction of these bacteria with PMNs for the main reason just cited and to ascertain whether the presence of a unique phenol-soluble LPS, the only significant difference noted in comparison of cell envelopes of a smooth-intermediate and rough strain of B. abortus (14, 15), might be a factor in intracellular survival. Guinea pig PMNs had no bactericidal activity against the smooth strain of B. abortus 45/0, and exhibited minimal killing of the rough strain 45/20 (34% loss of viability in 2 h), whereas human PMNs were somewhat more lethal to both strains, but yet far less than would be expected for extracellular parasites. The lower survival of rough strains has also been observed with LPS mutants of enteric bacilli (21) and might be expected with other gram-negative bacilli. A high degree of association of both strains was detected with human and guinea pig PMNs; therefore, it appeared that ingestion was not a function of LPS differences. However, no stimulation of the HMP pathway was observed after ingestion of either strain, whereas, in control TABLE 3. Effect of halides on the brucellacidal activity ofguinea pig granule lysates % Survival of strain:
Incubation mixture
Control NaCl + KI + H202
Granule lysate + H202 + KI Granule lysate + H202 + NaCl
tus 45/0
B. abortus 45/20
100
100
B. abor-
0
0
86
45
741
experiments with other scnmuli, three- to fivefold stimulation was common. There was no evidence of a generalized metabolic inhibition due to ingested Brucella because there was increased glucose oxidation via the HMP pathway upon incubation with S. typhimurium after preincubation with B. abortus. If the HMP pathway disposes of H202 generated after ingestion (6), it appeared either that Brucella inhibited a specific course of events subsequent to ingestion or the bacterial surface never generated the appropriate stimulus by interaction with the plasma membrane of the phagocyte. The in vitro brucellacidal activity of mixed granule lysates of human and guinea pig PMNs was tested as an indicator of survival probability of the bacteria within the hostile environment of the phagolysosome. It was significant that extracts of either guinea pig or human granule lysates had limited brucellacidal activity at pH 5.5, unless supplemented with exogenous H202 and halides. Furthermore, boiling the granule extracts or addition of catalase to decompose endogenous H202 completely abolished the bactericidal activity. It should be noted that the concentrations of H202, NaCl, and KI were not lethal in the absence of granule extract; however, in their presence, the granule extracts were potent bactericidal agents. The resistance of the smooth strain 45/0 to Cl-mediated killing may be significant as shown by the increased killing activity in the presence of I- relative to Cl-. Perhaps in addition to the ability to catabolize erythritol (25), virulent Brucella have developed resistance to an oxygen-killing system, presumably MPO, mediated by Cl-. The intracellular survival of Brucella within phagocytes contributes to the chronic nature of Brucellosis and complicates chemotherapeutic approaches toward the disease. For a parasite to survive intracellularly, the organism must be adaptable to intracellular nutritional conditions and be resistant to or inhibit the killing systems in phagolysosomes (19). Like Brucella, each facultative intracellular parasite so far examined appears to have developed different mechanisms for survival. For example, Mycobacterium lepraemurium survives degranulation and multiplies within phagolysosomes of rat macrophages (11), whereas M. tuberculosis appears to inhibit degranulation (1, 10) by the presence of toxic sulfolipids on the tubercle bacilli cell surface (9). The vole bacillus M. microti appears to increase cyclic adenosine 3',5'-monophosphate levels and inhibit the fusion of lysosomes with the phagocytic vacuoles (16). Results of experiments in progress in our laboratory (L. K. Riley and D. C. Robertson, unpublished data) from electron microscopy show that there is limited degranu-
742
KREUTZER, DREYFUS, AND ROBERTSON
INFECTr. IMMUN.
P. D., J. A. Armstrong, C. A. Brown, and P. lation after ingestion of B. abortus. Although 11. Hart, Draper. 1972. Ultrastructural study of the behavior of the mechanism limiting degranulation is unmacrophages toward parasitic mycobacteria. Infect. Imknown, lack of fusion of either type of granule mun. 5:803-807. with vacuoles containing bacilli would constitute 12. Holland, J. J., and M. J. Pickett. 1958. A cellular basis in experimental Brucella infection. J. Exp. Med. 108: an important survival mechanism. Thus, Bru343-360. cella appear to survive within phagocytic cells 13. Karnovsky, M. L. 1962. Metabolic basis of phagocytic by a mechanism similar to M. tuberculosis and activity. Physiol. Rev. 42:143. M. microti. It will be most interesting to deter- 14. Kreutzer, D. L, and D. C. Robertson. 1978. Surface macromolecules and virulence in intracellular parasitmine whether Brucella actively inhibit the deism: comparison of cell envelope components of smooth granulation process or the surface of these bacilli and rough strains of Brucella abortus. Infect. Immun. is such that the proper stimulus is never gener23:819-828. ated during interaction with the plasma mem- 15. Kreutzer, D. L., C. S. Buller, and D. C. Robertson. 1978. Chemical characterization and biological properbrane of the leukocyte. ties of lipopolysaccharides isolated from smooth and
ACKNOWLEDGMENTS This research was supported by the University of Kansas General Research Fund and by Public Health Service training grant GM-703 from the National Institute of General Medical Sciences. LITERATURE CITED 1. Armstrong, J. A., and P. D'Arcy Hart. 1975. Phagosome-lysosome interactions in cultured macrophages infected with virulent tubercle bacilli. J. Exp. Med. 142: 1-16. 2. Boyum, A. 1968. Isolation of mononuclear cells and granulocytes from human blood. Scand. J. Clin. Lab. Invest. 21:77-89. 3. Braun, W., A. Pomales-Lebron, and W. R. Stinebring. 1958. Interactions between mononuclear phagocytes and Brucella abortus strains of different virulence. Proc. Soc. Exp. Biol. Med. 97:393-397. 4. Bruno, G. A., and J. E. Christian. 1961. Determination of carbon-14 in aqueous bicarbonate solutions by liquid scintillation counting techniques: application to biological fluids. Anal. Chem. 33:1216-1218. 5. Cohn, Z. A., and J. G. Hirsch. 1960. The isolation and properties of the specific cytoplasmic granules of rabbit PMN leukocytes. J. Clin. Med. 112:983-1004. 6. DeChatelet, L. R. 1975. Oxidative bactericidal mechanisms of polymorphonuclear leukocytes. J. Infect. Dis. 131:295-303. 7. Freeman, B. A., and L. Vana. 1958. Host-parasite relationships in Brucellosis. I. Infection of normal guinea pig macrophages in tissue culture. J. Infect. Dis. 102: 258-267. 8. Goren, M. B. 1977. Phagocyte lysosomes: interactions with infectious agents, phagosomes, and experimental perturbations in function. Annu. Rev. Microbiol. 3: 507-533. 9. Goren, M. B., P. D. Hart, M. R. Young, and J. A. Armstrong. 1976. Prevention of phagosome-lysosome fusion in cultured macrophages by sulfatides of Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. U.S.A. 73:2510-2514. 10. Hart, P. D., and J. A. Armstrong. 1974. Strain virulence and the lysosomal response in macrophages infected with Mycobacterium tuberculosis. Infect. Immun. 10: 742-746.
16.
17.
18.
19. 20.
21.
22.
23.
24.
25.
26.
27.
rough strains of Brucella abortus. Infect. Immun. 23: 811-818. Lowrie, D. B., P. S. Jackett, and N. A. Ratcliffe. 1975. Mycobacterium microti may protect itself from intracellular destruction by releasing cAMP into phagosomes. Nature (London) 254:600-602. McGhee, J. R., and B. A. Freeman. 1970. Osmotically sensitive Brucella in infected normal and immune macrophages. Infect. Immun. 1:146-150. Macrae, R. A., and H. Smith. 1964. The chemical basis of the virulence of Brucella abortus. VI. Studies on immunity and intracellular growth. Br. J. Exp. Pathol. 45:595-603. Moulder, J. W. 1974. Intracellular parasitism: life in an extreme environment. J. Infect. Dis. 130:300-306. Renshaw, H., W. C. Davis, H. H. Fudenberg, and G. Padgett. 1974. Leukocyte dysfunction in the bovine homologue of the Chediak-Higashi syndrome of humans. Infect. Immun. 10:928-937. Rest, R. F., M. H. Cooney, and J. K. Spitznagel. 1977. Susceptibility of lipopolysaccharide mutants to the bactericidal action of human neutrophil lysosomal fractions. Infect. Immun. 16:145-151. Sbarra, A. J., and M. L. Karnovsky. 1959. The biochemical basis of phagocytosis. I. Metabolic changes during the ingestion of particles by polymorphonuclear leukocytes. J. Biol. Chem. 234:1355-1362. Sbara, A., B. Paul, R. Strauss, and M. Mitchell. 1972. Biochemical aspects of phagocytic cells as related to bactericidal function. RES J. Reticuloendothiel. Soc. 11:492-502. Smith, H., and R. B. Fitzgeorge. 1964. The chemical basis of the virulence of Brucella abortus. V. The basis of intracellular survival and growth in bovine phagocytes. Br. J. Exp. Pathol. 45:174-186. Sperry, J. F., and D. C. Robertson. 1975. Erythritol catabolism by Brucella abortus. J. Bacteriol. 121: 619-630. Stinebring, W. R., and R. Kessel. 1959. Continuous growth of Brucella abortus in mononuclear phagocytes of rats and guinea pigs. Proc. Soc. Exp. Biol. Med. 101: 412-415. Taggesson, C., and 0. Stendahl. 1973. Influence of the cell surface lipopolysaccharide structure of Salmonella typhimurium on resistance to intracellular bactericidal systems. Acta Pathol. Microbiol. Scand. Sect. B 81: 473-480.
Vol. 23, No. 3
0019-9567/79/03-0737/06$02.00/0
Interaction of Polymorphonuclear Leukocytes with Smooth and Rough Strains of Brucella abortus DONALD L. KREUTZERt LAWRENCE A. DREYFUS, AND DONALD C. ROBERTSON* Department of Microbiology, University of Kansas, Lawrence, Kansas 66045 Received for publication 8 December 1978
The bactericidal activity of guinea pig and human polymorphonuclear leukocytes (PMNs) against a smooth-intermediate strain (45/0) and a rough strain (45/20) of Brucella abortus has been examined. After incubation for 120 min, guinea pig PMNs incubated with either the smooth strain 45/0 or the rough strain 45/20 exhibited no bactericidal activity against the former and caused only a 34% decrease in viability of the latter. Human PMNs were more bactericidal than guinea pig PMNs to both strains; however, the killing of strain 45/20 by human PMNs was less than that observed in control experiments with S. aureus strain 502A. Both strains of B. abortus readily associated with guinea pig and human PMNs, and the bacteria were apparently ingested without stimulation of the hexose monophosphate pathway. Lysates (10 pg/ml, pH 5.5), prepared from guinea pig or human granules, were not particularly toxic to either strain unless supplemented with H202 and a halide (I- or Cl-). An oxygen-dependent killing system appeared to be lethal against both strains of B. abortus, with I- being more active than Cl- in the presence of H202 and granule lysate. The data suggest that degranulation after ingestion of Brucella by phagocytes does not occur due to the lack of a proper stimulus or possibly the baccilli actively inhibit the degranulation process thereby protecting the microbe from killing systems normally effective against extracellular parasites.
Facultative intracellular parasites (e.g., Mycobacterium tuberculosis, Brucella, and Listeria monocytogenes) survive conditions within phagocytic cells where most pathogens are killed. Their survival is likely influenced by the surface properties of bacilli, but it is not clear as to which surface components can be considered virulence factors except for the sulfolipids of the tubercle bacilli which inhibit degranulation after ingestion by monocytes (8, 9). Virulence factors of intracellular parasites play at least two possible roles: (i) provide protection from the hostile environment within the phagolysosome and (ii) inhibit or not stimulate the sequence of events after ingestion which would normally be bactericidal. It is difficult to compare virulence factors important in the pathogenesis of Mycobacteria with Brucella, the latter being a gram-negative rod with very different surface properties. Therefore, our approach has been to compare the surface macromolecules of two closely related strains of Brucella abortus that differ significantly in virulence and infectivity. The rough strain, B. abortus 45/20, was derived from the t Present address: Department of Pathology, University of Connecticut Health Center, Farmington, CT 06032.
smooth-intermediate strain, 45/0 (14). The chemical composition of cell walls of both strains has been examined with respect to total lipids, proteins, peptidoglycan content, and composition, as well as lipopolysaccharides (LPS) (14, 15). The only significant difference was in the LPS composition; that is, when the two strains were extracted by the Westphal procedure, LPS was found in the phenolic and aqueous phases of B. abortus strain 45/0, but only in the aqueous phase of B. abortus strain 45/20. The phenolsoluble LPS was the only fraction toxic to mice and contained low levels of heptose and dideoxyaldoses, which may suggest a unique core region and 0-specific side chain (15). It is equivocal whether the phenolic-phase LPS can be considered a virulence factor; however, it is the only surface difference observed in comparison studies of the two strains. We chose to examine the interaction of smooth and rough strains of B. abortus with polymorphonuclear leukocytes (PMNs) for two reasons: (i) the PMN is likely the first phagocyte which the bacteria encounter, especially if infection occurs through a cut or the conjunctival route, and (ii) it was of interest to ascertain whether the surface differences detected in B. 737
738
KREUTZER, DREYFUS, AND ROBERTSON
INFECT. IMMUN.
Assay of HMP activity. For the assay of hexose abortus strain 45/0 and 45/20 were related to enhanced intracellular survival of one strain over monophosphate (HMP) activity, the reaction mixture contained a portion of the PMN suspension, 3 x 107; the other. bacteria, 3 X 109 colony-forming units; homologous MATERLALS AND METHODS heat-inactivated serum, 10%; 0.1 to 0.2 JACi of DSource of bacteria. A smooth-intermediate strain, [1-'4C]glucose and HBSS in a final volume of 3 ml. B. abortus 45/0, was obtained from M. Meyer, Uni- The 25-ml Erlenmeyer reaction flask was sealed with versity of California, Davis, Calif., and the rough a serum stopper containing a plastic well trap. A 0.3strain, B. abortus 45/20, was provided by B. L. Deyoe, ml amount of ethanolamine-methyl cellosolve (1:2, National Animal Disease Center, Ames, Iowa. Cul- vol/vol) was added to the trap, and the reaction flasks tures were maintained and grown as previously de- were incubated with shaking at 370C for 60 min. The scribed as was the history of these established smooth reactions were stopped by injection of 0.1 ml of 5 N HCL The contents of the trap were transferred to and rough laboratory strains (1). Isolation of PMNs. PMNs were purified from he- minivials containing XDC counting solution (4) for parinized human blood by the technique of Boyum (2, determination of radioactivity. Preparation of granule lysates. PMNs were 20). Isolated PMNs were resuspended in Hanks balanced salt solution (HBSS) containing 0.1% gelatin washed once with 0.1 M Krebs-Ringer phosphate and stored at 4°C until used. Leukocytes were also buffer, pH 7.0, and suspended in approximately 5 ml isolated from adult rats and guinea pigs after intraper- of 0.25 M sucrose containing 10 mM trisitoneal injection of sterile casein (10%, wt/vol). The (hydroxymethyl)aminomethane, pH 7.0, and 1 mM PMNs were removed 17 h after injection by lavage ethylenediaminetetraacetic acid. The suspension was with phosphate-buffered saline containing heparin (10 homogenized with a Teflon homogenizer at room temU/ml). The cells were washed once with phosphate- perature until greater than 90% cell lysis had occurred. buffered saline and once with HBSS and then sus- The cell homogenate was centrifuged at 500 x g for 15 pended in HBSS at a known viable cell concentration min to remove intact cells and debris followed by centrifugation of the supernatant at 30,000 x g for 20 determined by trypan blue exclusion. Brucella-PMN association studies. The associ- min. The pellet containing the granules was resusation of Brucella with PMNs was determined with pended in 2 ml of 0.1 Krebs-Ringer phosphate buffer, bacteria labeled during growth with [14C]uracil (10 pH 5.5, and disrupted by homogenization with a Teflon mCi/ml). The radioactive bacteria were harvested by pestle followed by repeated freeze-thawing (6x) in an centrifugation and washed three times with HBSS acetone-dry ice bath (5). MPO activity of granule lysates. The myelobefore use. Incubation mixtures (3 ml) were prepared containing bacteria and PMNs at a 5:1 ratio with continuous gentle agitation with a New Brunswick (G76) gyratory water bath. At 30 min, 1.0-ml samples were removed from the reaction vessel and added to 9 ml of HBSS. After centrifugation at 150 x g, the cell pellet was resuspended in 10 ml of XDC counting solution containing 4% Cab-O-Sil (4), and a 0.5-ml portion of the supernatant was transferred to 10 ml of XDC counting solution. Radioactivity was determined with a Packard 3375 liquid scintillation spectrometer. The data were expressed as the percentage of the total radioactivity found in the supernatant and the pellet. Bactericidal activity of isolated leukocytes. The bactericidal activity of isolated PMNs was assayed with a bacteria-PMN ratio of 5:1. The components of the reaction mixture contained in siliconized 25-ml Erlenmeyer flasks included the following: PMN in HBSS (106 to 107), 1.0 ml; bacteria in HBSS, 1.0 ml; homologous heat-inactivated serum (56°C for 30 min), 0.3 ml; and HBSS (pH 7.2), 0.7 ml. All incubations were at 37°C with shaking, and at timed intervals, 0.1ml portions were removed and diluted into 9.9 ml of deionized water or 9.9 ml of phosphate-buffered saline containing Clostridium perfringens phospholipase c (100 ,g/ml) to effect leukocyte lysis. Dilutions were in 0.5% tryptose-0.5% NaCl broth and surviving bacteria were determined by plating 0.1-ml portions of the appropriate dilution on Trypticase soy agar plates (Baltimore Biological Laboratory). The results were similar by either method of leukocyte lysis. All data were expressed as colony-forming units versus time or as percent survival.
peroxidase (MPO) activity of granule lysates was determined by the method of Tagesson and Stendahl (27). The assay mixture contained guaiacol, 0.02 M; H202, 0.03 M; Krebs-Ringer phosphate buffer (pH 5.5), 0.1 M; and granule lysate in a final volume of 300 tl. The reaction was followed by the change in absorbance at 470 nm. One unit of MPO activity was defined as the amount of granule lysate required to effect a change of 0.001 absorbance units per min at 470 nm. The granule lysate was serially diluted in 0.1 M KrebsRinger phosphate buffer, pH 5.5, to the desired concentration. For studies on the effects of halides, 0.1 M phosphate buffer, pH 6.0, was substituted for the Krebs-Ringer phosphate buffer. All lysates were stored at -70°C until used. Brucellacidal activity of granule lysates. Incubation mixtures were prepared in 25-ml Erlenmeyer flasks and contained 0.1 M Krebs-Ringer phosphate, granule lysate at a final concentration of 10 to 100,ug of protein per ml, and bacteria (3 x 105 to 5 x 105 colony-forming units per ml) in a total volume of 3 ml. In some experiments, catalase (100 1ig) was added to the reaction mixture to ensure that there was little or no endogenous H202. Portions were removed after 30, 60, and 120 min of incubation at 370 C, diluted in 0.1 M phosphate buffer (pH 7.0), and plated in triplicate on Trypticase soy agar plates (Baltimore Biological Laboratory). The results were expressed as colonyforming units per milliliter versus time and as percent bacterial survival. To test oxygen-dependent killing systems in granule lysates, incubation mixtures contained 0.1 M KrebsRinger phosphate (pH 5.5), 3.3 x 10-5 M KI, 10-7 M
VOL. 23, 1979
739
PMNs AND B. ABORTUS
H202, granule lysate at various nonbactericidal concentrations, and bacteria at 3 x 105 to 5 x 105 colonyforming units per ml in a reaction volume of 3.0 ml. The mixtures were incubated with gentle shaking at 370C, and samples were withdrawn and plated as just described. The effect(s) of various halides on the oxygen-dependent killing of B. abortus was determined by substituting 0.1 M phosphate buffer, pH 6.0, for Krebs-Ringer phosphate and using either 10mM NaCl or 10 mM KI in separate reaction mixtures. The surviving bacteria were diluted for plating in 0.1 M phosphate buffer, pH 7.0, instead of the tryptose-NaCl broth. Again, the data were expressed as colony-forming units versus time or percent bacterial survival. RESULTS
u10
70 .
-J
60
50
4
40
60
a
30
0>20 >
° X
D
50
30
JI
so 90 120
I
W
40
J
Brucellacidal activity of guinea pig and human PMNs. The brucellacidal activities of guinea pig peritoneal and human peripheral leukocytes were compared. Guinea pig PMNs had no bactericidal activity against strain 45/0 and exhibited minimal killing of strain 45/20 (Fig. 1). Human peripheral leukocytes exhibited increased bactericidal activity compared with guinea pig PMNs against both strains of B. abortus, although B. abortus 45/0 was more resistant to killing than was B. abortus 45/20
tn
30
W
0 -
> 20 D W
p
p
0
100
30
60
90 (mins)
120
TIME FIG. 2. Bactericidal activity of human peripheral PMNs against B. abortus strains 45/0 and 45/20.
Strain 45/0, no PMNs (0); strain 45/20, no PMNs (0); PMNs + 45/0 (0); PMNs + 45/20 (-). Inset shows bactericidal activity against S. aureus 502A, bacteria, no PMNs (A); PMNs + S. aureus 502 (A).
90 /
to O
VI'/
X/
/
(Fig. 2). In parallel incubation mixtures, human PMNs displayed normal bactericidal activity
/J
against S. aureus 502A, which indicated that the leukocytes were not affected by Ficoll-HyW paque purification. The rqugh strain B. abortus aR (45/20 has a 1.7- to 2-fold greater generation time 60 \ than the smooth parent strain, 45/0 (D. L. and D. C. Robertson, unpublished a: wKreutzer O data). The basis for the observed difference in > growth rates is unknown; however, rough strains 50 50 . of gram-negative bacteria generally have longer D \ generation times than smooth bacilli (C. S. 70
/ -
/
Buller, personal communication).
The association of PMNs with B. abortus 45/ 0 and 45/20 was examined by using radioactive bacilli to determine whether the increased sur. . . vival of strain 45/0 was due to decreased ingesA tion. The smooth-intermediate strain (45/0) 0 60 120 30 90 readily associated with both guinea pig PMNs T I M E (mins) FIG. 1. Bactericidal activity of guinea pig PMNs and human PMNs after 30 min of incubation to against B. abortus strains 45/0 and 45/20. Strain the extent of 74 and 88%, respectively. The as45/0, no PMNs (0); 45/20, no PMNs (0); PMNs + sociation with the rough strain was slightly less, with 53% of the bacteria associating with guinea 45/0 (0); PMNs + 45/20 (U). 40
O
740
INFECT. IMMUN.
KREUTZER, DREYFUS, AND ROBERTSON
pig PMNs and 56% associating with human PMNs during a 30-min incubation. These data show that the increased survival of B. abortus 45/0 was not likely due to decreased ingestion, but to resistance of smooth Brucella to intracellular killing systems. Stimulation of the HMP pathway during phagocytosis of B. abortus 45/0 and B. abortus 45/20. Stimulation of the HMP shunt is a primary metabolic event associated with the bactericidal activity of PMNs (13, 22). Furthermore, measurement of the extent of D-[1-14C]glucose oxidation by PMN provides an indication of the bactericidal potential of the phagocytic cell (6, 23). As shown in Table 1, there was no measurable stimulation of D-[1-'4C]glucose oxidation by either guinea pig or human PMNs incubated with each strain of B. abortus. The lack of HMP pathway stimulation was apparently due to the unique surface properties of the bacilli because there was significant stimulation in PMNs exposed to heat-killed Salmonella typhimurium. One explanation for the observed lack of bactericidal activity toward injected Brucella was that the bacilli inhibited the normal metabolic activity after ingestion, especially the generation of H202. To test this, leukocytes were preincubated with either 45/0 or 45/20 before being exposed to heat-killed S. typhimurium. Preincubation of guinea pig PMN with whole cells and cell envelopes did not prevent [14C]_ glucose oxidation when the guinea pig PMNs were subsequently exposed to heat-killed S. typhimurium (Table 2). TABLE 1. Glucose oxidation by leukocytes in the presence and absence of bacterial stimulus -Fold stimulation of HMP pathway
Source of PMN
Guinea pig Human
rium (heat B. abortus B. abortus 45/0 45/20 killed) 2.9 1.1 1.3 2.5 0.8 1.0
TABLE 2. Glucose oxidation by leukocytes preincubated with B. abortus 45/0 and 45/20 and cell fractions followed by stimulation with S. typhimurium -Fold stimPreincubation condition
Brucellacidal activity of granule lysates. Because there appeared to be ingestion of both smooth and rough strains of B. abortus, but no metabolic stimulation usually associated with degranulation, the brucellacidal activity of granule lysates was examined. Guinea pig granule lysates (10 jig of protein per ml) were not bactericidal to B. abortus 45/0 at pH 5.5 but did kill 50% of the rough strain 45/20. Under the same incubation conditions, human PMN granule lysates (10 ,ug of lysate protein per ml, pH 5.5) were more bactericidal to both strains of Brucella, with 13.5% of strain 45/0 and 71% of strain 45/20 being killed during a 60-min incubation. Boiling the granule extract or addition of catalase abolished the brucellacidal activity of the granule lysates against both strains. These data suggested that an oxygen-dependent killing system was mainly responsible for the limited bactericidal activity of granule lysates. The bactericidal activity of an oxygen-dependent killing system effective against Brucella was assayed by adding H202 and halides (Cland I-) at nontoxic concentrations. In the presence of H202, Cl-, and I- (Fig. 3), both guinea pig and human granule lysates, 1 MPO unit per ml, caused a greater than 99.9% decrease in viability of both strains. Because the I--mediated bactericidal activity of MPO is known to be different from Cl--mediated killing, the effect of each halide on the bactericidal activity of
100
%I
4>
80 r W I-
z w u
60 40
Id
a.
20
elation of way
S. typhimurium (no preincubation) B. abortus 45/0 + S. typhimurium B. abortus 45/0 cell envelopes + S. typhimurium B. abortus 45/20 + S. typhimurium B. abortus 45/20 cell envelopes + S. typhimurium
2.9 2.5 2.0 2.4 4.2
0
15
30 TIME (mins)
FIG. 3. Brucellacidal activity of guinea pig granule lysates, granule lysate + I- + Cl- (0); H202 + I+ Cl- (0); granule lysate + H202 + I- + Cl- (A). The concentration of each reaction component is given in the text.
PMNs AND B. ABORTUS
VOL. 23, 1979
granule lysates was examined. The granule lysate v H202 IE system was extremely bactericidal to both strains; whereas, the granule lysate H202. C1 system was much less bactericidal (Table 3). In summary, the smooth strain of B. abortus 45/0 was more resistant to Clmediated killing than the rough strain 45/20; however, both strains were extremely sensitive to the presumed MPO-E-mediated killing activity. DISCUSSION PMNs constitute an important line of defense against microbial invaders and are likely the first phagocytic cells which interact with Brucella at the foci of infection. The inability of these cells to efficiently kill virulent Brucella at the primary site of infection is a key factor in the dissemination of Brucella to the regional lymph nodes and eventually infection of the reticuloendothelial system. Even though there have been numerous studies on the survival of Brucella within mononuclear cells (3, 7, 12, 17, 18, 24, 26), we chose to look at the interaction of these bacteria with PMNs for the main reason just cited and to ascertain whether the presence of a unique phenol-soluble LPS, the only significant difference noted in comparison of cell envelopes of a smooth-intermediate and rough strain of B. abortus (14, 15), might be a factor in intracellular survival. Guinea pig PMNs had no bactericidal activity against the smooth strain of B. abortus 45/0, and exhibited minimal killing of the rough strain 45/20 (34% loss of viability in 2 h), whereas human PMNs were somewhat more lethal to both strains, but yet far less than would be expected for extracellular parasites. The lower survival of rough strains has also been observed with LPS mutants of enteric bacilli (21) and might be expected with other gram-negative bacilli. A high degree of association of both strains was detected with human and guinea pig PMNs; therefore, it appeared that ingestion was not a function of LPS differences. However, no stimulation of the HMP pathway was observed after ingestion of either strain, whereas, in control TABLE 3. Effect of halides on the brucellacidal activity ofguinea pig granule lysates % Survival of strain:
Incubation mixture
Control NaCl + KI + H202
Granule lysate + H202 + KI Granule lysate + H202 + NaCl
tus 45/0
B. abortus 45/20
100
100
B. abor-
0
0
86
45
741
experiments with other scnmuli, three- to fivefold stimulation was common. There was no evidence of a generalized metabolic inhibition due to ingested Brucella because there was increased glucose oxidation via the HMP pathway upon incubation with S. typhimurium after preincubation with B. abortus. If the HMP pathway disposes of H202 generated after ingestion (6), it appeared either that Brucella inhibited a specific course of events subsequent to ingestion or the bacterial surface never generated the appropriate stimulus by interaction with the plasma membrane of the phagocyte. The in vitro brucellacidal activity of mixed granule lysates of human and guinea pig PMNs was tested as an indicator of survival probability of the bacteria within the hostile environment of the phagolysosome. It was significant that extracts of either guinea pig or human granule lysates had limited brucellacidal activity at pH 5.5, unless supplemented with exogenous H202 and halides. Furthermore, boiling the granule extracts or addition of catalase to decompose endogenous H202 completely abolished the bactericidal activity. It should be noted that the concentrations of H202, NaCl, and KI were not lethal in the absence of granule extract; however, in their presence, the granule extracts were potent bactericidal agents. The resistance of the smooth strain 45/0 to Cl-mediated killing may be significant as shown by the increased killing activity in the presence of I- relative to Cl-. Perhaps in addition to the ability to catabolize erythritol (25), virulent Brucella have developed resistance to an oxygen-killing system, presumably MPO, mediated by Cl-. The intracellular survival of Brucella within phagocytes contributes to the chronic nature of Brucellosis and complicates chemotherapeutic approaches toward the disease. For a parasite to survive intracellularly, the organism must be adaptable to intracellular nutritional conditions and be resistant to or inhibit the killing systems in phagolysosomes (19). Like Brucella, each facultative intracellular parasite so far examined appears to have developed different mechanisms for survival. For example, Mycobacterium lepraemurium survives degranulation and multiplies within phagolysosomes of rat macrophages (11), whereas M. tuberculosis appears to inhibit degranulation (1, 10) by the presence of toxic sulfolipids on the tubercle bacilli cell surface (9). The vole bacillus M. microti appears to increase cyclic adenosine 3',5'-monophosphate levels and inhibit the fusion of lysosomes with the phagocytic vacuoles (16). Results of experiments in progress in our laboratory (L. K. Riley and D. C. Robertson, unpublished data) from electron microscopy show that there is limited degranu-
742
KREUTZER, DREYFUS, AND ROBERTSON
INFECTr. IMMUN.
P. D., J. A. Armstrong, C. A. Brown, and P. lation after ingestion of B. abortus. Although 11. Hart, Draper. 1972. Ultrastructural study of the behavior of the mechanism limiting degranulation is unmacrophages toward parasitic mycobacteria. Infect. Imknown, lack of fusion of either type of granule mun. 5:803-807. with vacuoles containing bacilli would constitute 12. Holland, J. J., and M. J. Pickett. 1958. A cellular basis in experimental Brucella infection. J. Exp. Med. 108: an important survival mechanism. Thus, Bru343-360. cella appear to survive within phagocytic cells 13. Karnovsky, M. L. 1962. Metabolic basis of phagocytic by a mechanism similar to M. tuberculosis and activity. Physiol. Rev. 42:143. M. microti. It will be most interesting to deter- 14. Kreutzer, D. L, and D. C. Robertson. 1978. Surface macromolecules and virulence in intracellular parasitmine whether Brucella actively inhibit the deism: comparison of cell envelope components of smooth granulation process or the surface of these bacilli and rough strains of Brucella abortus. Infect. Immun. is such that the proper stimulus is never gener23:819-828. ated during interaction with the plasma mem- 15. Kreutzer, D. L., C. S. Buller, and D. C. Robertson. 1978. Chemical characterization and biological properbrane of the leukocyte. ties of lipopolysaccharides isolated from smooth and
ACKNOWLEDGMENTS This research was supported by the University of Kansas General Research Fund and by Public Health Service training grant GM-703 from the National Institute of General Medical Sciences. LITERATURE CITED 1. Armstrong, J. A., and P. D'Arcy Hart. 1975. Phagosome-lysosome interactions in cultured macrophages infected with virulent tubercle bacilli. J. Exp. Med. 142: 1-16. 2. Boyum, A. 1968. Isolation of mononuclear cells and granulocytes from human blood. Scand. J. Clin. Lab. Invest. 21:77-89. 3. Braun, W., A. Pomales-Lebron, and W. R. Stinebring. 1958. Interactions between mononuclear phagocytes and Brucella abortus strains of different virulence. Proc. Soc. Exp. Biol. Med. 97:393-397. 4. Bruno, G. A., and J. E. Christian. 1961. Determination of carbon-14 in aqueous bicarbonate solutions by liquid scintillation counting techniques: application to biological fluids. Anal. Chem. 33:1216-1218. 5. Cohn, Z. A., and J. G. Hirsch. 1960. The isolation and properties of the specific cytoplasmic granules of rabbit PMN leukocytes. J. Clin. Med. 112:983-1004. 6. DeChatelet, L. R. 1975. Oxidative bactericidal mechanisms of polymorphonuclear leukocytes. J. Infect. Dis. 131:295-303. 7. Freeman, B. A., and L. Vana. 1958. Host-parasite relationships in Brucellosis. I. Infection of normal guinea pig macrophages in tissue culture. J. Infect. Dis. 102: 258-267. 8. Goren, M. B. 1977. Phagocyte lysosomes: interactions with infectious agents, phagosomes, and experimental perturbations in function. Annu. Rev. Microbiol. 3: 507-533. 9. Goren, M. B., P. D. Hart, M. R. Young, and J. A. Armstrong. 1976. Prevention of phagosome-lysosome fusion in cultured macrophages by sulfatides of Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. U.S.A. 73:2510-2514. 10. Hart, P. D., and J. A. Armstrong. 1974. Strain virulence and the lysosomal response in macrophages infected with Mycobacterium tuberculosis. Infect. Immun. 10: 742-746.
16.
17.
18.
19. 20.
21.
22.
23.
24.
25.
26.
27.
rough strains of Brucella abortus. Infect. Immun. 23: 811-818. Lowrie, D. B., P. S. Jackett, and N. A. Ratcliffe. 1975. Mycobacterium microti may protect itself from intracellular destruction by releasing cAMP into phagosomes. Nature (London) 254:600-602. McGhee, J. R., and B. A. Freeman. 1970. Osmotically sensitive Brucella in infected normal and immune macrophages. Infect. Immun. 1:146-150. Macrae, R. A., and H. Smith. 1964. The chemical basis of the virulence of Brucella abortus. VI. Studies on immunity and intracellular growth. Br. J. Exp. Pathol. 45:595-603. Moulder, J. W. 1974. Intracellular parasitism: life in an extreme environment. J. Infect. Dis. 130:300-306. Renshaw, H., W. C. Davis, H. H. Fudenberg, and G. Padgett. 1974. Leukocyte dysfunction in the bovine homologue of the Chediak-Higashi syndrome of humans. Infect. Immun. 10:928-937. Rest, R. F., M. H. Cooney, and J. K. Spitznagel. 1977. Susceptibility of lipopolysaccharide mutants to the bactericidal action of human neutrophil lysosomal fractions. Infect. Immun. 16:145-151. Sbarra, A. J., and M. L. Karnovsky. 1959. The biochemical basis of phagocytosis. I. Metabolic changes during the ingestion of particles by polymorphonuclear leukocytes. J. Biol. Chem. 234:1355-1362. Sbara, A., B. Paul, R. Strauss, and M. Mitchell. 1972. Biochemical aspects of phagocytic cells as related to bactericidal function. RES J. Reticuloendothiel. Soc. 11:492-502. Smith, H., and R. B. Fitzgeorge. 1964. The chemical basis of the virulence of Brucella abortus. V. The basis of intracellular survival and growth in bovine phagocytes. Br. J. Exp. Pathol. 45:174-186. Sperry, J. F., and D. C. Robertson. 1975. Erythritol catabolism by Brucella abortus. J. Bacteriol. 121: 619-630. Stinebring, W. R., and R. Kessel. 1959. Continuous growth of Brucella abortus in mononuclear phagocytes of rats and guinea pigs. Proc. Soc. Exp. Biol. Med. 101: 412-415. Taggesson, C., and 0. Stendahl. 1973. Influence of the cell surface lipopolysaccharide structure of Salmonella typhimurium on resistance to intracellular bactericidal systems. Acta Pathol. Microbiol. Scand. Sect. B 81: 473-480.
