Search

In the absence of antibiotic therapy, viable bacteria can persist within intra-abdominal abscesses in mice for at least 10 weeks. The mechanisms contributing to this survival are unknown, but abscess-derived neutrophils have impaired abilities to kill, in vitro, organisms engulfed in vivo. In order to determine whether subpopulations of abscess neutrophils might be discernible on the basis of phenotypic or functional criteria, cells from murine intra-abdominal abscesses were examined for phagocytic activity, CR3 expression, and H2O2 production in response to soluble and particulate stimuli. With respect to phagocytosis of Proteus mirabilis, abscess cells were no less efficient than peritoneal exudate neutrophils; no significant subpopulation of cells was incapable of phagocytosis in the presence of normal mouse serum. Using flow cytometry to examine abscess neutrophils for CR3 expression, we found that no subpopulations of cells were observed with unstimulated cells or with cells incubated with either phorbol 12-myristate 13-acetate or bacteria and serum. Intracellular H2O2 levels were measured by using the probe 2',7'-dichlorofluorescin diacetate. In general, incubation with phorbol 12-myristate 13-acetate resulted in similar increases in H2O2 production in all cells of the population. However, stimulation with bacteria and serum revealed a variable but consistent, poorly responsive subpopulation of neutrophils in abscess cell populations. Cell-sorting experiments showed that cells from the poorly responsive section of the FACS profile contained significantly higher numbers of abscess-derived bacteria, suggesting the presence of a subpopulation of viable abscess neutrophils harboring persisting viable bacteria.

The microbicidal activity of an acetate extract of human polymorphonuclear leukocyte granules was tested against Bacteroides fragilis, Bacteroides vulgatus, Bacteroides distasonis and Bacteroides thetaiotaomicron. All strains tested were killed by the extract, and there were no significant differences between the different Bacteroides species.

Intraabdominal abscesses were induced in mice by intraperitoneal inoculation of Bacteroides fragilis and Escherichia coli plus bran as the abscess-potentiating agent. Six- or seven-day-old abscesses were mechanically disaggregated in buffer, and the cells obtained were fractionated on discontinuous Percoll density gradients. Neutrophil populations of different density, each approximately 90% pure, were isolated. When the abscess-derived neutrophils were subsequently incubated with normal serum in vitro under aerobic conditions, the viability of the gram-negative bacteria that had been phagocytosed within the abscess did not change significantly. This anergy to intracellular bacteria (on subsequent incubation in vitro under optimal conditions for phagocytic killing) was also found for neutrophils that had been obtained from abscesses induced by a mixture that included Proteus mirabilis plus B. fragilis and from those induced by E. coli plus P. mirabilis. While unable to significantly kill intracellular organisms that had been phagocytosed in vivo, the abscess-derived neutrophils could engulf and kill organisms to which they were exposed in vitro. Neutrophils from abscesses induced by P. mirabilis only plus bran killed that organism introduced in vitro significantly more effectively than the organisms that had been engulfed in vivo. In contrast, neutrophils from abscesses induced by the gram-positive organism Staphylococcus aureus plus bran were able to kill their intracellular organisms on subsequent incubation in vitro as effectively as they could kill added S. aureus. Neutrophils isolated from the peripheral blood and from induced peritoneal exudates of abscess-bearing mice were able to phagocytose and kill organisms in vitro with greater efficiency than abscess-derived neutrophils. The mechanism whereby neutrophils from abscesses induced by the gram-positive organism S. aureus can kill the organisms phagocytosed in vivo on subsequent in vitro incubation, in contrast to the relative anergy to their intracellular organisms displayed by neutrophils derived from abscesses induced by combinations of gram-negative bacteria, is not known.

Athymic mice formed significantly smaller abscesses than euthymic mice in response to the intraperitoneal inoculation of an abscess-inducing mixture of Escherichia coli, Bacteroides fragilis, and autoclaved colonic contents, an abscess-potentiating agent. Adoptive transfer of nonimmune, Thy-1-positive spleen cells to athymic mice restored their ability to make abscesses of sizes similar to those in controls, indicating that T lymphocytes contribute to abscess formation in normal mice.

The role of the Fc and third component of complement (C3) receptors on mouse neutrophils in the control of killing of Proteus mirabilis, opsonized in normal mouse serum (NMS) or heated immune mouse serum (HIMS), was studied. The events following incubation of neutrophils with P. mirabilis and the events associated with bacterial killing were assayed. The respiratory burst was quantified by chemiluminescence (CL). Levels of leukocyte-associated bacteria were determined after a 20-min ingestion period as a measure of phagocytosis. Bacterial killing was measured while ingestion was allowed to continue or as a discrete process when extracellular, noningested bacteria had been removed and neutrophils with intracellular bacteria were incubated in the presence of serum. Modification of these responses in the presence of three monoclonal antibodies (MAb), NIMP-R10 and M1/70, which bind to different epitopes of the mouse C3 receptor, and 2.4G2, which binds to the mouse Fc receptor, was investigated. MAb to the C3, but not to the Fc, receptors reduced CL, ingestion, and intracellular killing of NMS-opsonized P. mirabilis. MAb to the Fc receptor diminished CL to and reduced the rate of ingestion of HIMS-opsonized bacteria. The two MAb to the C3 receptor each produced a similar inhibition of ingestion and intracellular killing of HIMS-opsonized bacteria, but they only partially blocked CL. A range of MAb preparations reactive with other murine antigens did not inhibit these events, either with NMS- or HIMS-opsonized P. mirabilis. The results suggest that C3 receptors on mouse neutrophils played a predominant role in regulation of the killing of P. mirabilis. Similar results were found for Staphylococcus aureus. C3 receptors were necessary for maximal expression of all functions culminating in bacterial kill. That MAb to the C3 receptor inhibited phagocytosis of HIMS-opsonized bacteria in similar fashion to the effect of MAb to the Fc receptor and in contrast to the lack of effect of control MAb may reflect steric hindrance of the Fc receptor by MAb binding to the C3 receptor, or it may reflect that the receptors are linked in murine neutrophils as they are in human neutrophils.

A report that Bacteroides fragilis inhibited the killing of facultative anaerobes by human leukocytes in vitro prompted us to look for such inhibition in the peritoneal cavities of mice. No evidence that this phenomenon was operative could be found.

Granule proteins from human neutrophils were prepared by extraction with acetate, and their antibacterial activity against Bacteroides fragilis was determined. Activity was highly dependent on pH; greatest killing occurred at the most acid pH tested (pH 5.0). Optimum activity was observed at physiological ionic strength and low bacterial numbers. Killing was inhibited by incubation temperatures of less than 37 degrees C. Eight times more extract was required to kill 50% of stationary-phase bacteria, compared with those growing in logarithmic phase. The antibacterial effect of granule extract was destroyed by boiling, but some activity was retained after heating to 56 degrees C and 80 degrees C. Granule extract activity was tested under conditions in which oxygen-dependent antibacterial systems were inhibited. The rate and extent of killing was not affected by anaerobiosis, sodium azide, or cysteine hydrochloride. These results suggest that the activity of granule extract is independent of oxidative antibacterial systems, and therefore, under conditions that occur in anaerobic infections, potent leukocyte granule-associated mechanisms exist for the destruction of B. fragilis.