A.M.J.J. Verweij-van Vught

Transferrins and heme-compounds as iron sources for pathogenic bacteria.

The low concentration of free iron in body fluids creates bacteriostatic conditions for many microorganisms and is therefore an important defense factor of the body against invading bacteria. Pathogenic bacteria have developed several mechanisms for acquiring iron from the host. Siderophore-mediated iron uptake involves the synthesis of low molecular weight iron chelators called siderophores which compete with the host iron-binding glycoproteins lactoferrin (LF) and transferrin (TF) for iron. Other ways to induce iron uptake, without the mediation of siderophores, are the possession of outer membrane protein receptors that actually recognize the complex of TF or LF with iron, resulting in the internalization of this metal, and the use of heme-compounds released into the circulation after lysis of erythrocytes. In this review, the nonsiderophore-mediated iron-uptake systems used by certain pathogenic bacteria are emphasized. The possible contribution of these iron-uptake systems to the virulence of pathogens is also discussed.

An enzyme-linked immunosorbent assay (ELISA) for the measurement of antibodies to different parts of the gram-negative lipopolysaccharide core region

Bovine serum albumin was complexed with the core antigens of either Escherichia coli J5 LPS, Salmonella minnesota R595 LPS or E. coli lipid A. These core-BSA complexes were used for solid-phase coating in ELISAs for anti-core antibodies. Antibodies, binding to various parts of the core region were easily quantified in a single experimental set-up, which was hitherto not possible. The ELISA has only 3 incubation steps and is not costly as only moderate amounts of the core antigens (i.e., 1 microgram per test) were needed for coating. The sensitivity proved to be excellent and the complexes were biologically fully active (compared to native, smooth LPS), which make them suitable for the screening (after fusion) of monoclonal anti-core antibodies. Another possible application is the large-scale screening of blood-bank sera in order to find samples with a high anti-core antibody content.

Outer membrane proteins of Bacteroides fragilis and Bacteroides vulgatus in relation to iron uptake and virulence

A virulent strain B. fragilis BE1 and an avirulent strain B. vulgatus BE20 were grown in a culture medium with and without the addition of a synthetic chelator (Bipyridyl) to induce iron limitation. Cells grew more slowly under iron stress, although the growth rate of the B. vulgatus strain was more affected under these conditions than the strain of B. fragilis. The outer membrane protein profile of these strains was studied in relation to the iron concentration in the growth medium by means of SDS-polyacrylamide gel electrophoresis. Four proteins, with the apparent molecular weights of 89, 49, 44 and 23.5 kDa, were consistently present in the outer membrane of B. fragilis BE1 grown under iron restricted conditions. In B. vulgatus BE20 cells a 44 and a 23.5 kDa protein were absent and only the expression of an 89 kDa protein was clearly seen under these conditions. The iron regulated proteins, particularly the 44 kDa protein, could be involved to an iron uptake mechanism in B. fragilis. So the presence of these proteins might play an important role in the virulence of this anaerobic bacterium.

Characterization of fimbriae from Bacteroides fragilis

Fimbriae derived from Bacteroides fragilis strain BE1 (BE1 fimbriae) appeared to be composed of subunits with a molecular weight of 40,000. Under the electron microscope the fimbriae could be visualized as straight filaments with a diameter of 4 nm. It appeared that production of the BE1 fimbriae is repressed under conditions of iron limitation, and at a growth temperature of 20 degrees C. Antibodies raised against the 40,000 dalton polypeptide, purified by means of preparative SDS-polyacrylamide gelelectrophoresis, recognized the native fimbriae, as was shown by immunogold labelling of intact bacterial cells, and by immunoprecipitation. Immunoblot experiments showed that other strains of B fragilis tested produced polypeptides, ranging in molecular weight from 40,000 to 42,000, that are antigenically related to the BE1 fimbrial subunit. No haemagglutination activity could be associated with the BE1 fimbriae.

Pathogenic synergy: mixed intra-abdominal infections

In this article we review our researches into the pathogenesis of mixed infections. These may conveniently be divided into in vitro and in vivo studies.In vitro we confirmed that interference with the killing of aerobes by polymorphonuclear leucocytes (PMN’s) is a property of theBacteroides strains tested and appears to depend on competition for opsonins i.e. complement factors. Further studies are in progress to define which complement factors and which bacterial structures are involved. The influence ofB. fragilis on chemotaxis has also been studied. Our preliminary data suggest thatB. fragilis is itself poorly chemotactic and reduces the chemoattractivity ofProteus mirabilis. This observation is surprising when we consider that abscess formation is the hall-mark ofB. fragilis infections and needs clarification.In vivo we have developed a skin infection model in mice which is economical and gives reproducible and quantitative results. In this model we have demonstrated pathogenic synergy betweenEscherichia coli andB. fragilis. Further studies are planned to assess the role of complement and bacterial factors in this in vivo synergy.

Effect ofBacteroides fragilis grown in the presence of clindamycin, metronidazole and fusidic acid on opsonization and killing ofEscherichia coli

Bactericidal action of human polymorphonuclear leucocytes onEscherichia coliin the presence ofBacteroides fragilisgrown in subinhibitory concentrations of clindamycin, metronidazole and fusidic acid was studied.Bacteroides fragilisgrown in the absence of drugs significantly inhibited the killing ofEscherichia coli.Bacteroides fragilis grown in the presence of the drugs had a reduced inhibitory effect on the killing ofEscherichia colibut this reduction was only significant forBacteroides fragilisgrown in 1/2 MIC of clindamycin. The phagocytosis ofBacteroides fragilisgrown with and without clindamycin, as measured by killing, was the same. Complement consumption ofBacteroides fragilisgrown with and without clindamycin did not differ. Clindamycin-treatedBacteroides fragilisfixed C3 to a significantly lower degree than did untreated bacteria. The chemiluminescence ofEscherichia coliopsonized with serum preincubated with clindamycin-treatedBacteroides fragiliswas significantly higher than with serum preincubated with untreated bacteria. These results suggested that in killing experiments of mixedEscherichia coliandBacteroides fragilis,the mechanism underlying the reduced inhibitory capacity of clindamycin-exposedBacteroides fragilisis related to greater availability of C3 in serum for opsonization ofEscherichia coli.

Prevention of lethal endotoxemia in actinomycin D-sensitized mice by incubation of Salmonella minnesota R595 lipopolysaccharide with monoclonal antibodies to R595

Murine monoclonal antibodies reacting with lipopolysaccharide (LPS) of Salmonella minnesota strain R595 (Re chemotype) were prepared, and tested for their ability to protect actinomycin D-sensitized mice against lethal endotoxemia. Protection was found with some antibodies up to a 90-fold increase in LD50, whereas others exhibited no protection. The various protective antibodies did not all bind to the same epitope. The same applied for non-protective clones. Protective and non-protective clones could not be discriminated by ELISA. One protective monoclonal antibody (clone 20) was specific for ketodeoxyoctonate, a structural element common to various LPS. These findings show that the involvement of lipid A in the binding site of monoclonal antibodies is no prerequisite for protection.

Influence of rosmarinic acid on opsonization and intracellular killing ofEschericha coli andStaphylococcus aureus by porcine and human polymorphonuclear leucocytes

The influence of rosmarinic acid on the function of porcine and human polymorphonuclear leucocytes was tested. Rosmarinic acid inhibited the chemiluminescence of PMNL, induced by preopsonized Zymosan or phorbol myristate acetate. The killing ofEscherichia coli was inhibited by rosmarinic acid at a concentration of 2 mM, but not that ofStaphylococcus aureus. The inhibition of the killing was due to an impaired opsonization, caused by an adverse influence of rosmarinic acid on complement activation. Direct effects of rosmarinic acid on the killing mechanisms of PMNL were not observed.

Murine ascitic fluids contain varying amounts of an inhibitor that interferes with complement-mediated effector functions of monoclonal antibodies

The ability of murine monoclonal antibodies (mAbs), directed to the inner core of Gram-negative bacterial lipopolysaccharide (LPS, endotoxin), to enhance complement-mediated killing of bacteria, was investigated. The mAbs were tested as present in ascitic fluid. It was found that ascites contains an factor that inhibited the activity of complement. This effect was evident in assays for complement-mediated lysis of antibody-coated Gram-negative bacteria (bacterial killing) or of opsonised red blood cells. Moreover, the amount of inhibitor was found to vary from one ascites to another and spanned a 60-fold range. Thus, in vitro or in vivo experiments where complement is known to play a determining role may yield incorrect results when ascites is used as a source of antibody; the use of ascites prepared from irrelevant antibody as a negative control does not eliminate this problem.

Further characterization of monoclonal antibodies to lipopolysaccharide of Salmonella Minnesota strain R595.

We have shown here that despite the use of monoclonal antibodies with well-defined epitope-specificities, and despite testing them in the most simple animal model available (i.e., mixing of homologous LPS with Mab prior to injection), we are not yet able to explain why some of the antibodies were effective and others not. For some of the clones (e.g., clone 20), an even better definition of binding sites is currently taking place in an attempt to obtain this understanding. We also do not yet understand why clone 20 was not effective in the mucin model, while using much lower amounts of injected antibody, and much higher challenge doses, this Mab was effective against E. coli in the gentamicin-treated mouse model. Very clear is, however, that in order to be protective in the latter model, Mabs are not required to be specific for lipid A. In the future it will be essential to develop procedures which measure specific interaction between smooth LPS/bacteria and antibodies to the LPS core region. In addition, it will be of great help when the chemical structure of non-substituted, rough-form LPS, as occurring in smooth LPS preparations, would be defined. This applies also to O-substituted core molecules.