Suppression of Blastogenic Transformation of Lymphocytes by Bacteroides fragilis in vitro and in vivo

Zbl. Bakt. 274, 406-416 (1990) © Gustav Fischer Verlag, StuttgartlNew York

Suppression of Blastogenic Transformation of Lymphocytes by Bacteroides fragilis in vitro and in vivo * ARNE C. RODLOFF!, PETRA WIDERA z , STEFAN EHLERS! "*, THOMAS MONTAG z , MARTIN LUCAS!, GERHARD SCHMIDT z , and HELMUT HAHW ! Institute of Medical Microbiology and Immunology, Free University of Berlin, D-1000 Berlin, and 2 Institute for Medical Microbiology, Humboldt University, D-O 1040 Berlin

With 5 Figures· Received March 14, 1990 . Accepted March 20, 1990

Summary Bacteroides species and Enterobacteriaceae are known to cause synergistic infections. However, the mechanisms behind this synergy are not completely understood. Several authors have shown that Bacteroides species may inhibit the phagocytosis of Enterobacteriaceae by polymorphonuclear leukocytes as well as by macrophages. With the present study we have addressed the question of whether Bacteroides fragi/is (BF) is also capable of suppressing specific immune functions. When incubated together with murine lymphocytes, BF significantly inhibited the blastogenic transformation of these cells stimulated by Escherichia coli-lipopolysaccharide (LPS) or concanavalin A. This effect was dose dependent and was not mediated by prostaglandins. Other bacteria such as E. coli or Listeria monocytogenes did not show such an extensive suppression, while Streptococcus pneumoniae was equally active. BF also inhibited the pokeweed mitogen induced blastogenic transformation of human lymphocytes. Moreover, lymphocytes from BF-injected animals obtained 3 to 12 hours after infection proved to be partly refractory for LPSstimulation. Finally, BF injections also affected T-cell dependent immunity as judged from the aggravation of an experimental listeriosis in mice.

Zusammenfassung Eine Reihe von Untersuchungen hat gezeigt, daLl sich Bacteroides species und Enterobacteriaceae bei Mischinfektionen synergistisch verhalten k6nnen. Die Mechanismen dieses Synergismus sind aber bisher nicht vollstandig geklart. Verschiedene Autoren haben berichtet, daLl Bacteroides species die Phagocytose von Enterobacteriaceae durch polymorphker,:. Dedicated to Prof. Dr. Dr. h.c. G. Pulverer on the occasion of his 60th birthday. Present address: Dept. of Medicine, Dartmouth Medical School, Hanover, NH 03756, USA H

Lymphocyte Suppression by Bacteroides fragilis

407

nige neutrophile Granulocyten bzw. durch Macrophagen inhibieren. Die vorliegende Arbeit untersucht die Fragestellung, ob Bacteroides fragi/is (BF) auch Mechanismen der spezifischen Immunitiit beeintriichtigen kann. Murine Lymphocyten, die zusammen mit hitzeinaktiviertem BF inkubiert wurden, zeigten nach Stimulation durch Escherichia coli-Lipopolysaccharid (LPS) oder Concanavalin A eine signifikante Suppression der blastogenen Transformation. Dieser Effekt war dosisabhiingig und wurde nieht durch Prostaglandine vermittelt. Andere Bakterien, wie z. B. E. coli oder Listeria monocytogenes waren weniger suppressiv, wiihrend Streptococcus pneumoniae iihnliche Effekte ausloste. BF inhibierte dariiber hinaus auch die blastogene Transformation von humanen Lymphocyten. Lymphocyten von BF-infizierten Miiusen, die 3 bis 12 Stunden nach Injektion der Erreger gewonnen wurden, erwiesen sich ebenfalls als teilweise refraktiir gegeniiber einer Stimulation mit LPS. Schlie/Slieh konnte im Listeriosemodell der Maus eine Beeintriichtigung auch der T-zell-vermittelten Immunitiit beobachtet werden.

Introduction Gram-negative obligately anaerobic bacteria constitute a major part of the normal indigenous flora of humans; however, they also have to be considered as important opportunistic pathogens. Bacteroidaceae often participate in polymicrobial infections and they are found especially in association with Enterobacteriaeae. Therefore, efforts were made to study the interaction of aerobes and anaerobes in mixed infections with a number of different animal models. Altemeier (1) showed as early as 1942 that bacterial mixtures isolated from several cases of human peritonitis had higher pathogenic potential in experimental animals than pure cultures of any of the microorganisms involved. A more precise evaluation of the synergistic effects of mixed inocula was given by Hite et al. (8) and later by Socransky and Gibbons (21), who emphazised the role of the Bacteroidaceae in augmenting the virulence of the bacterial associations. Weinstein, Onderdonk and co-workers (15, 25) reported on synergistic effects of Escherichia coli and Bacteroides fragilis studied in a model of peritonitis (intra-abdominal sepsis) in Wistar rats, while Kelly (13) described synergy for the same bacteria in experimental wound infections. Rodloff et al. (18, 19) have demonstrated synergistic lethality in experimental sepsis with E. coli and different Bacteroides species. A number of possible mechanisms have been offered to account for the synergistic effects observed for mixed aerobidanaerobic infections. Mayrand and McBride (14) reported on the production of a growth factor (succinate) by aerobes (Klebsiella) benefitting the anaerobes. Ingham and co-workers (10) demonstrated that Bacteroides species may inhibit phagocytosis and killing of aerobes by polymorphonuclear leukocytes. Tofte et al. (22) as well as Jones and Gemmel (12) suggested that this suppression of phagocytosis is due to competition for serum opsonins. In addition, Radloff et al. (17) have shown that Bacteroides fragilis also impairs the phagocytic capacity of macrophages. Hence, a number of experimental studies have suggested that Bacteroides species are suppressive of non-specific immune functions. However, no information is available on the effects the anaerobe might have on specific immune functions. Therefore, with the present study we addressed the question of whether Bacteroides fragilis affects the blastogenic transformation of lymphocytes. Material and Methods

Experimental animals. Female (CS7Bl/6 x DBN2)F J hybrid mice at the age of approx. 10 weeks were used throughout the experiments. The animals were raised at our own breeding facilities and kept under specific pathogen free conditions.

408

A. C. Rodloff et al.

Bacteria. Bacteroides fragilis ATCC 25285 (BF) was grown in Schaedler-broth (BBL Microbiology Systems, Cockeysville, MD, USA; supplemented with 0.1 mg/l Vitamin K1 ) under anaerobic conditions for 24 h at 37°C. Appropriate bacterial concentrations were established by adding fresh medium and confirmed by plating suitably diluted aliquots. Prior to experimental use, BF was passaged repeatedly through animals to ensure maximal encapsulation. Escherichia coli ATCC 25922 (EC) was included to evaluate an aerobic organism often found together with BF in mixed infections. Listeria monocytogenes EGD (LM) and Streptococcus pneumoniae ATCC 6303 (SP) were employed to evaluate unrelated organisms. All aerobes were either grown in nutrient broth or in brain-heart infusion broth (Oxoid, Basingstoke, Great Britain). When called for, microorganisms were inactivated by incubating for 30 min at 60°C, washed three times and resuspended in cell culture medium. Appropriate concentrations were established by counts with a Petroff-Hausser and Helber chamber (Hausser Scentific, Blue Bell, PA, USA). Murine Spleen Cells. Spleens were obtained from healthy mice and disrupted into single cell suspensions. Red blood cells were lysed, remaining cells washed three times and resuspended at a concentration of 2.5 X 10 6 cells/ml in RPMI 1640 (Biochrom KG, Berlin, FRG; supplemented with ( per liter) 3.7g NaHC0 3 , 100 ml fetal bovine serum, 100000 I.U. penicillin, and 10 mg streptomycin). When called for, RPMI 1640 was further supplemented with 5 mg/l indomethacin (Sigma Chemical Co., St. Louis, MO, USA). Additional experiments were carried out with spleen cells that were obtained from infected mice. These cells were treated as described above. Human Lymphocytes. Mononuclear cells were isolated from the venous blood of eight different healthy adults by gradient centrifugation according to Boyum (4). The cells were washed three times in RPMI 1640 and then resuspended at a concentration of 1 X 106/ml in culture medium (RPMI 1640 supplemented with (per liter) 2 g NaHC0 3 , 4.76 g HEPES, 100 ml fetal bovine serum, 100000 I.U. penicillin). Blastogenic Transformation. Twohundred microliters of the spleen cell suspensions were added to wells of flat bottomed microtitration plates and 0.05 ml heat killed bacteria were added where appropriate. Control cultures were kept without bacterial challenge. The cultures were then incubated for 6h at 37°C and 5% CO 2 before mitogens were added. Mitogens employed were either lipopolysaccharide of E. coli 0111 :B4 (LPS; Sigma; 0.01 mg/well) or concanavalin A (Con A; Sigma; 0.001 mg/well). Additional cultures were left without stimulation. After further incubation for 24 h, 1 !tCi of eH)thymidine (Amersham Laboratories, Amersham, Great Britain) was added to each wei!. The incubation was continued for another 24 hours, cells were then harvested onto filter paper and radioactivity incorporated into the cells was counted with a liquid scintillation spectrometer (Packard Instrument Co., Rockville, MD, USA). All assays were done at least in triplicates. Spleen cells from infected animals were treated similarly exept that the addition of heat killed bacteria was omitted. Lymphocytes from human peripheral blood were also tested in amounts of 0.2 m!. Cell cultures were incubated together with different numbers of heat killed BF (ratio BF/lymphocytes either 500: lor 50: 1) for 72h at 37°C in 5% CO 2 , When called for, pokeweed mitogen (PWM, Serva, Heidelberg, FRG) was added at a final concentration of 2 !tg/m!. Proper controls were kept without BF and/or PWM. 24 h prior to harvesting, 1 !tCi of eH)thymidine (UVVVR, Prague, CSFR) was added to each culture. Cell associated radioactivity was counted employing a liquid scintillation spectrometer (LKB Wallac Oy, Turku, Finland). All assays were done in triplicate and for each blood donor individually. Experimental infections. In order to evaluate the effect of BF on lymphocytes in vivo, live organisms were introduced into experimental animals by intravenous injections (tail vein) of 0.2 ml fresh broth culture containing 1 X 10 8 du. Spleen cells of these animals were harvested at indicated times and were assayed in vitro for their blastogenic response to LPS. Furthermore, the effect of BF on the outcome of an infection with L. monocytogenes was studied. Mice received an i.v. injection of 5 X 103 LM and were subsequently inoculated with 1 X 108 BF at indicated times. Neither of these inocula alone causes prolonged or even

Lymphocyte Suppression by Bacteroides fragilis

409

lethal infections in mice. Groups of at least 4 mice were killed every 24 h and numbers of LM in the spleens were established by appropriate culture of organ homogenates. Results Effect of BF and Other Organisms on Spleen Cells in vitro

The effect of different concentrations of heat-inactivated BF on the mitogen induced blastogenic transformation of lymphocytes was studied. Spleen cells were incubated for 6 h together with BF at a cell/bacteria ratio of approx. 500: 1, 50: 1, or 5: 1. Subsequently, the cells were treated with mitogens or were left untreated. LPS was employed to stimulate B-lymphocytes, while Can A served as a T-lymphocyte mitogen. The eH)thymidine uptake of differently treated cell populations is depicted in Fig. 1. The results indicate that BF by itself had some mitogenic property for splenocytes. However, BF clearly inhibited the LPS-mediated blastogenic transformation in a dosedependent manner. For Con A-stimulated lymphocytes, a bi-phasic response was observed: At lower doses BF was capable of enhancing the effect of Con A, while 10 9 heat-killed BF again caused inhibition of ('H)thymidine uptake.

eon A

LPS

!ZI

SF 5 x 10e9/ml

lID

SF 5 x 10e8/ml

S

SF 5 x 10e7/ml

•

no mitogen

o

20000

40000

60000

control

80000

cpm

Fig. 1. Uptake of (lH)thymidine by mouse spleen cells that were incubated for 6 h with different concentrations of heat-inactivated Bacteroides fragilis (BF) and were then stimulated with LPS or Can A or were kept untreated. Controls were incubated without bacteria.

410

A. C. Rodloff et al.

Further experiments were designed to investigate the specificity of the effects demonstrated for BF. For this purpose, other bacteria (EC, LM, SP) were tested under the same experimental conditions, employing a 500: 1 ratio of bacteria to lymphocytes. In these experiments, EC and LM showed a moderate inhibition of the LPS-mediated blastogenic transformation of spleen cells (Fig. 2). However, this suppression was not as extensive as the effect observed for BF. Surprisingly, SP proved to be equally as active as BF. In contrast, the Con A-mediated response was further enhanced by EC and LM while BF and SP were again almost equally suppressive. Since such inhibitory effects might be mediated by the induction of prostaglandins, the experiments were repeated with 5 mg/l indomethacin present in the culture medium. Again, essentially the same results were obtained (data not shown), suggesting that prostaglandins are not involved in the suppression of blastogenic transformation of lymphocytes described here. Effect of BF on Human Lymphocytes

In order to assess whether the effects described for the mouse system might be conferrable upon the human situation, we performed additional experiments employing human peripheral blood lymphocytes. Since human cells are not stimulated by LPS,

liD s. pneumoniae II

Con A (control. 46992 cpm)

• • E:.l

L. monocytogenes E. coli B. fragilis control

LPS (control. 33278 cpm)

tl63%

no mitogen

1605%

534%

o

50

100

150

200

250

cpm in % of control

Fig. 2. Uptake of (3H)thymidine by mouse spleen cells that were incubated for 6 h with different heat-inactivated bacteria and were then stimulated with either LPS or Con A or were kept untreated. Controls were incubated without bacteria.

Lymphocyte Suppression by Bacteroides fragilis

411

PWM

Ea

SF 5 x 10e9/ml

II1l

SF 5 x 10e8/ml

•

control

no mitogen

o

50

100

150

cpm in % of control

Fig. 3. Uptake of (3H)thymidine by human peripheral lymphocytes that were incubated together with different concentrations of heat-inactivated Bacteroides fragilis (BF) and were then stimulated with PWM or were kept untreated. Controls were incubated without bacteria. The values given here are means of the results established for eight different blood donors individually.

PWM was used as a lymphocyte mitogen activating both, T-cells and B-cells. The results of all assays showed a high degree of conformity, therefore, mean values for all eight blood donors are given in Fig.3. The data indicate that BF by itself did not stimulate blastogenic transformation of human lymphocytes. However, BF did suppress PWM-stimulated incorporation of (3H)thymidine (Fig. 3), again in a dose-dependent fashion.

Effect of BF on Spleen Cells in vivo Splenocytes of BF-infected mice were harvested 3,6, 12, and 24h after introduction of the bacteria into the animals and then tested in vitro for their capacity to respond to stimulation with LPS. The results clearly demonstrate that - when compared to normal splenocytes - cells from infected animals were inhibited in their mitogenic response (Fig.4). The degree of inhibition was time-dependent and followed the same pattern already described for other effects of BF (e.g. inhibition of macrophage phagocytosis, LPS-hypersensitivity of experimental animals, interferon induction; 3, 17, 18).

412

A. C. Rodloff et al.

24 h p.i.

12h p.i.

6 h p.i.

3 h p.i.

control

o

20

40

60

80

100

120

cpm in % of control

Fig. 4. Uptake of (3H)thymidine by spleen cells that were obtained at different times post infection (p.i.) from mice injected intravenously with 1 x 108 Bacteroides fragilis. The cells were stimulated in vitro with LPS.

Effect of BF on an Ongoing Infection with Listeria monocytogenes

Since - in vitro - BF suppressed both, LPS- and Con A-mediated lymphocyte stimulation, we addressed the question of whether BF also affects T-lymphocyte dependent immunity in vivo. For this purpose, we studied the effect of BF on the well characterized model of murine listeriosis. Mice were intravenously infected with a sublethal dose of LM and received an additional injection with BF either together with the LMapplication or 12, 24, 36, 48, or 72 h later. Spleens were harvested from experimental animals at 24 h intervals and cultured from experimental animals at 24 h intervals and cultured for LM. The bacterial concentrations found in these organs are shown in Fig. 5. It is evident that BF injected 12 to 48 h after the listerial challenge enhanced the amount of vital Listeria organisms in animal spleens. Thus, the mice treated with BF were less capable of controlling the listerial infection. Moreover, all animals that received an increased dose of LM (1 x 104 du, normally sublethal) and were then treated with BF (24 h later), died from that infection 3 to 4 days post infection (p.i., data not shown).

Lymphocyte Suppression by Bacteroides fragilis

413

7

6

log cfu of Listeria! spleen

5

....................r .......··

... 4

3

1

--.-

--

····0···

2

3

4

5

6

days p.i. LM without BF LM + BF together LM + BF t2 h later

--,.--

LM + BF 36 h later

._.0'C-'.

LM + BF 48 h later

........--

LM + BF 72 h later

.~.-

LM + BF 24 h later

Fig. 5. Concentrations of Listeria monocytogenes (LM) in spleens of mice that were obtained at different times post infection (p.i.). The animals were intravenously infected with a sublethal dose of LM (5 x 10 3 du) and received an additional injection of 1 x 108 BF at indicated times. For each point in time, at least 4 spleens were evaluated. The standard deviation was always less than 5%.

Discussion Employing our model of mixed aerobidanaerobic sepsis, we have shown earlier that not only vital bacteria may cause synergistic lethality in mice. Rather, it became clear that even heat killed (metabolically inactive) organisms and/or bacterial components (E. coli-LPS) are capable of exerting synergistic effects in experimental animals (17, 18). Therefore, it seemed likely that alterations in the immune response of the host play a role in bringing about the synergistic effects described. In support of this notion, a number of investigators have demonstrated that anaerobes interfere with the phagocytosis of aerobic bacteria by polymorphonuclear leukocytes in vitro (5, 10, 12,20, 23). In addition, Bacteroides fragilis was shown to be capable of inhibiting the phagocytic

414

A. C. Radloff et al.

activity of macrophages in vitro and in vivo (17). While this suppression of phagocytic systems might serve to explain the enhancement of bacterial concentrations in organs of experimental animals suffering from mixed infections (19), it does not as readily offer an interpretation for the hypersensitivity of BF-infected animals to LPS (18). Therefore, we have addressed the question of whether BF is capable of altering other (specific) immune functions. An early response of the cellular immune system to an invading microorganism in the clonal expansion of specifically reactive lymphocytes. Moreover, in the mouse system, lipopolysaccharides of aerobic Gram-negative bacteria are potent inducers of poly clonal B-lymphocyte proliferation. Although the lipopolysaccharides of Bacteroidaceae are known to have little biological activity, they have been described as potent lymphocyte mitogens (11,24). Nevertheless, heat killed BF induced only a very limited uptake of (3H)thymidine by splenocytes (Fig. 1), especially when compared with other Gram-negative microorganisms such as EC (Fig. 2). At the same time, treatment of lymphocytes with BF resulted in a severe inhibition of their response to concomitantly administered mitogens. This effect was not only operative in vitro but could be substantiated with lymphocytes obtained from BF-infected animals (Fig. 4). Hence, infections with BF induce a broad immunosuppression of the host, affecting the nonspecific phagocytic systems and specific immune functions. Since both, LPS-stimulated B-cell- and Con A-stimulated T-cell-functions are compromised, it came as no surprise that BF is capable of aggravating experimental infections with E. coli (19) as well as with L. monocytogenes, a facultatively intracellular organism (Fig. 5). Finally, we have presented evidence suggesting that these immunosuppressive effects of BF might be of importance in cases of human infections, since BF seems at least to be able to impair the PWM-medited profileration of human peripheral blood lymphocytes (Fig. 3). By contrast, EC and LM caused only a mild suppression of LPS-mediated blastogenic transformation of lymphocytes. These organisms were even capable of enhancing the effect of Con A (Fig. 2). Conversely, S. pneumoniae was similar in inhibitory activity to BF. Since the latter two microorganisms are set apart from EC and LM, especially by the presence of large amounts of polysaccharides on their surfaces, it is tempting to speculate that this capsular material might playa role in mediating the effects described here. This hypothesis needs further evaluation. Since BF inhibits regular functions of different cell types (polymorphonuclear phagocytes, macrophages, lymphocytes) it seems possible that these effects are due to the induction of a mediator with broad suppressive potential. The failure of indomethacin to antagonize suppression suggests that prostaglandins are not responsible for the lymphocyte inhibition observed here. Earlier, we have reported on the induction of an atypical acid-labile interferon-alpha by BF (3). This type of interferon has also been found in immunosuppressed patients suffering from AIDS, lupus erythematodes, etc. (6,9, 16). However, the biological functions of this interferon have not been characterized as yet. Finally, in both, experimental E. coli-infections and murine listeriosis, tumor necrosis factor (TNF) induction has been recognized as a mediator of infectious lethality (2, 7). Therefore, at we are currently studying the effects BF might have on TNF-production. Whether other cytokines might be involved in the immunosuppression induced by BF is presently unclear. Acknowledgments. The expert technical assistance of Mrs. U. Ruschendorf and Mrs. is gratefully acknowledged.

J. Orso

Lymphocyte Suppression by Bacteroides fragilis

415

References 1. Altemeier, W. A.: The pathogenicity of the bacteria of appendicitis peritonitis. Surgery 11 (1942) 374-384 2. Beutler, B., I. W. Milsark, and A. C. Cerami: Passive immunization against cachectinl tumor necrosis factor protects mice from lethal effect of endotoxin. Science 229 (1985) 869-871 3. Blanchard, D. K., A. C. Rodloff, W. E. Steward II, H. Hahn, and H. Friedman: Induction of an atypical interferon by Bacteroides fragilis and Escherichia coli in experimental infctions and in leukocyte cultures. Infection 14 (1986) 289-293 4. Boyum, A.: Isolation of mononuclear cells and granulocytes from human blood. J. Clin. Lab. Invest. 21 (1968) 77-89 5. Connolly, J. c., C. McLean, and S. Tabaqchali: The effect of capsular polysaccharide and lipopolysaccharide of Bacteroides fragilis on polymorph function and serum killing. J. Med. Microbiol. 17 (1984) 259-271 6. DeStefano, E., R. M. Friedman, A. E. Friedman-Kien, J. j. Goedert, D. Hendriksen, O. T. Preble, J. A. Sonnabend, and J. Vilcek: Acid-labile human leukocyte interferon in homosexual men with Kaposi's sarcoma and lymphadenopathy. J. Infect. Dis. 146 (1982) 451-455 7. Havell, E. A.: Production of tumor necrosis factor during murine listeriosis. J. Immunol. 139 (1987) 4225-4231 8. Hite, K. E., M. Locke, and H. C. Hesseltine: Synergism in experimental infections with nonsporulating anaerobic bacteria. J. Infect. Dis. 84 (1949) 1-9 9. Hocks, J. j., H. M. Montsopoulos, S. A. Geis, N. I. Stahl, j. L. Decker, and A. L. Notkins: Immune interferon in the circulation of patients with autoimmune disease. N. Engl. J. Med. 301 (1979) 5-8 10. Ingham, H. R., P. R. Sissons, D. Thargonnet, J. B. Selkon, and A. A. Codd: Inhibition of phagocytosis in vitro by obligate anaerobes. Lancet II (1977) 1252-1254 11. joiner, K. A., K. P. W. J. McAdam, and D. L. Kasper: Lipopolysaccharides from Bacteroides fragilis are mitogenic for spleen cells from endotoxin responder and nonresponder mice. Infect. Immun. 36 (1982) 1139-1145 12. jones, G. R. and C. G. Gemmel: Impairment by Bacteroides species of opsonisation and phagocytosis of enterobacteria. J. Med. Microbiol. 15 (1982) 351-361 13. Kelly, M. J.: The quantitative and histological demonstration of pathogenic synergy between Escherichia coli and Bacteroides fragilis in guinea pig wounds. J. Med. Microbiol. 11 (1978) 513-523 14. Mayrand, D. and B. C. McBride: Ecological relationship of bacteria involved in a simple, mixed anaerobic infection. Infect. Immun. 27 (1980) 44-50 15. Onderdonk, A. B., J. G. Bartlett, T. Louie, N. Sullivan-Seigler, and S. L. Gorbach: Microbial synergy in experimental intra-abdominal abscess. Infect. Immun. 13 (1976) 22-26 16. Preble, O. T., R. J. Black, and R. M. Friedman: Systemic lupus erythematosus: presence in human serum of an unusual acid-labile leukocyte interferon. Science 216 (1982) 429-431 17. Rodloff, A. c., J. Becker, D. K. Blanchard, T. W. Klein, H. Hahn, and H. Friedman: Inhibition of macrophage phagocytosis by Bacteroides fragilis in vivo and in vitro. Infect. Immun. 52 (1986) 488-492 18. Rodloff, A. c., P. Giidke, F. Lux, and H. Hahn: Experimentelle Pathogenitiit von Bacteroides fragilis. Fortschr. Antimikrob. Antineoplast. Chemother. 4 (1985) 901-910 19. Rodloff, A. C. and H. Hahn: Synergistic lethality in experimental infections with Escherichia coli and Bacteroides fragilis. Zbl. Bakt. Hyg. A 258 (1984) 112-119 20. Rodloff, A. c., F. Hillinger, H. Friedman, and H. Hahn: Effects of anti-Bacteroidesantibodies on Escherichia coli and different Bacteroides species in vitro and vivo. Zbl. Bakt. Hyg. A 262 (1986) 483-491 28 Zbl. Bakt. 274/3

416

A. C. Rodloff et al.

21. Socransky, S. S. and R. j. Gibbons: Required role of Bacteroides melaninogenicus in mixed anaerobic infections. ]. Infect. Dis. 115 (1965) 247-253 22. Tofte, R. W., P. K. Peterson, D. Schmeling, j. Bracke, Y. Kim, and P. G. Quie: Opsonization of four Bacteroides species: Role of the classical complement pathway and immunoglobulin. Infect. Immun. 27 (1980) 784-792 23. Wade, B. H., D. L. Kasper, and G. L. Mandell: Interactions of Bacteroides fragilis and phagocytes: studies with whole organisms, purified capsular polysaccharide and clindamycin-treated bacteria. J. Antimicrob. Chemother. 17, Suppl. C (1983) 51-62 24. Wannemuehler, M. j., S. M. Michalek, E. jirillo, S. I. Williamson, M. Hirsawa, and j. R. McGhee: LPS regulation of the immune response: Bacteroides endotoxin induces mitogenic, polyclonal and antibody responses in classical LPS responsive but not C3H1 He] mice. J. Immunol. 133 (1984) 299-305 25. Weinstein, W. M., A. B. Onderdonk, j. G. Bartlett, and S. L. Gorbach: Experimental intra-abdominal abscesses in rats: development of an experimental model. Infect. Immun. 10 (1974) 1250-1255 Dir. u. Prof. Dr. Arne C. Rodloff, Paul-Ehrlich-Institut, Paul-Ehrlich-Str. 51-59, D-6070 Langen 1