The effect of specific antibody on antibody-independent interactions between E. coli J5 and human complement.

We previously reported that Escherichia coli J5, the galactose epimerase-deficient mutant of E. coli O111:B4, can bind and activate purified human C1. The effects of hyperimmune rabbit anti-J5 IgG or IgM on E. coli J5 interactions with human C have been examined. Specific IgG or IgM increased the binding of 125I-C1 by J5. However, the rate of C1 activation, as determined by SDS-PAGE of eluted 125I-C1s, was decreased if bacteria were preincubated with immune IgG. Complexes formed between J5 preincubated with immune Ig and C1, under conditions in which all of the C1 was allowed to activate, consumed more C4 than J5 alone plus C1. However, the amount of C4 consumed per C1 molecule was identical for all bacteria preparations. Concentrations of specific IgG or IgM that significantly increased C1 binding did not appear to enhance C3b deposition upon incubation of E. coli J5 in NHS. Thus, although specific antibody may enhance C1 binding by E. coli J5, the ability of these additional C1 molecules to alter later events in the C cascade may depend on the control of C1 activation and its subsequent activity when bound to different membrane components.

Antibody-independent interactions between Escherichia coli J5 and human complement components.

With the characterization of an increasing number of molecules that are capable of activating the 1st component of the classical pathway of complement (C), the possibility that some Gram-negative bacteria may activate C1 independent of naturally occurring antibody has been reexamined. We have confirmed a previous report that purified C1 (the activated form of C1) can bind to certain strains of bacteria and it retains its enzymatic activity when thus bound. The availability of purified C1 in its precursor form has allowed us to extend these observations to the native C1 molecule. Using a semirough mutant of Escherichia coli, the galactose epimerase-deficient strain E. coli J5, we have examined the binding and activation of radiolabeled C1. J5 bound radiolabeled C1 in a dose-dependent manner and essentially all of the bound C1 was activated as judged by SDS-PAGE. The bacteria-C1 complex consumed purified C4 and C2 and the consumption of C2 was proportional to the C4 concentration. Subsequent addition of terminal C components C3-9 supplied as serum-EDTA caused a highly significant decrease in bacterial viability. These results demonstrate that C1 may bind to the bacterial membrane in such a manner as to initiate a bactericidal reaction. Therefore, antibody-independent binding and activation of C1 must be considered in the assessment of serum sensitivity of Gram-negative bacteria.

Production and release of platelet-activating factor (PAF); dissociation from degranulation and superoxide production in the human neutrophil.

The generation of human PAF was examined using purified neutrophils from normal and CGD donors. PAF release occurred in response to all of the neutrophil stimuli examined, including PMA and the calcium ionophore A23187, which initiate the relatively selective release of specific granule constituents. PAF release was dissociable from neutrophil degranulation by 1) the lack of correlation between PAF titers and extent of enzyme release for different stimuli, 2) the strict dependence of PAF release on the presence of extracellular Ca++, 3) the different kinetics for release of PAF and granule enzymes, and 4) the ability of neutrophils significantly depleted of azurophil and specific granule constituents to release normal levels of PAF when stimulated with opsonized zymosan. PAF release was also normal, if not elevated, in neutrophils from patients with chronic granulomatous disease, demonstrating that the formation and release of PAF does not depend upon an intact superoxide generating pathway.

Generation and release of platelet-activating factor (PAF) from enriched preparations of rabbit basophils; failure of human basophils to release PAF.

Rabbit mononuclear cells containing up to 20% basophils and uncontaminated by neutrophils release PAF when stimulated with goat antiserum to rabbit IgE. The amount of PAF detected was a function of basophil concentration but decreased on a per cell basis at high basophil or high total cell concentrations. Calcium ionophore A23187, but not protein A, C5a, or FMLP, initiated rabbit basophil degranulation and PAF release. By contrast, extensive studies using a variety of human leukocyte preparations failed to demonstrate the release of significant levels of PAF from human basophils by IgE-dependent or -independent mechanisms. These results suggest that cells other than the peripheral blood basophil (e.g., the neutrophil) may act as the primary site of PAF production in man.

The release of a platelet-activating factor by stimulated rabbit neutrophils.

Normal rabbit peripheral blood neutrophils released a platelet-activating factor upon stimulation by opsonized zymosan. The liberation was Ca++ dependent and the time course of release was closely associated with phagocytosis. The material extracted into chloroform and exhibited an identical mobility by thin layer chromatography to basophil-derived, IgE-stimulated, platelet-activating factor (PAFb). It was similar to PAFb in its effect on platelets in both aggregation and release but was distinguished from ADP, thrombin, arachidonic acid, and thromboxanes. This factor appears to be responsible for some previously reported neutrophil-platelet interactions.

Macrophage stimulation by bacterial lipopolysaccharides. II. Evidence for differentiation signals delivered by lipid A and by a protein rich fraction of lipopolysaccharides.

Stimulation of macrophages to lyse tumor cells is a property common to lipopolysaccharide (LPS) extracted from a variety of smooth and rough bacterial strains by several different preparative procedures. The relationship between macrophage stimulation and the structural characteristics of LPS is defined. In protein-free LPS, lipid A bears the stimulatory signal which results in the differentiation of elicited macrophages into killer cells. The polysaccharide moiety is neither stimulatory itself nor does it block the activity of complete LPS on macrophages. Extraction of LPS by the butanol or Boivin procedures produces preparations in which LPS is complexed through its lipid A moiety to a protein rich component, LAP. Isolated LAP delivers a macrophage differentiation signal which is independent of lipid A. The presence of these two structurally distinct constituents in the cell walls of gram-negative bacteria broadens the biological environments in which they can stimulate macrophages in vivo.

Chemical and biological properties of a protein-rich fraction of bacterial lipopolysaccharides. II. The in vitro rat peritoneal mast cell response.

Lipid A-associated protein (LAP) isolated from preparations of bacterial lipopolysaccharides (LPS, endotoxins) has been demonstrated to initiate the release of amines from rat peritoneal mast cells. The release at low concentrations of LAP requires both cellular energy and calcium, and thus appears to be a true secretory response. At higher concentrations the release is independent of these variables. The time required for maximal response is approximately 10 to 15 min at 37 degrees C. The response appears to be a general property of Escherichia coli LAP preparations since LAP isolated from three serotypes of these organisms all have similar activity. On the basis of heat lability at 100 degrees C, the ability of LAP to initiate mast cell secretion appears to be independent of its ability mitogenically to stimulate murine B lymphocytes.

Isolation of a lipid A bound polypeptide responsible for "LPS-initiated" mitogenesis of C3H/HeJ spleen cells.

The experiments by Sultzer and Nilsson (1), and later by Watson and Riblet (2), established that spleen cells from the C3H/HeJ strain of mouse were refractory to the mitogenic effects of bacterial lipopolysaccharides (LPS). More recently, however, experiments from our laboratory (3) demonstrated that spleen cells from C3H/HeJ mice were in fact responsive to some preparations of LPS but not to others, and that the method of extraction played a critical role in determining activity. In particular, preparations of LPS prepared by extraction with aqueous butanol had potent mitogenic activity. Our data showed that the mitogenic activity of such positive preparations of LPS coisolated with the LPS during gel filtration chromatography and subsequent equilibrium banding on CsCl. In addition, lipid A isolated from positive preparations of LPS was also capable of stimulating C3H/HeJ spleen cells. Taken together, these experiments provided rather convincing data that it was the LPS (in particular the lipid A) itself, or some contaminant very tightly bound to the lipid A, which was responsible for its biological activity. We further demonstrated that treatment of positive preparations of LPS with hot phenol rendered such preparations nonmitogenic for C3H/HeJ spleens, yet activity for other strains was only moderately decreased. These experiments would suggest either that the phenol treatment chemically alters the lipid A region of the LPS molecule or that such treatment removes the putative tightly bound contaminant responsible for C3H/HeJ mitogenesis. In the experiments reported here, we have explored in greater detail the role of lipid A in the stimulation of C3H/HeJ spleen cells. For these experiments we have utilized our earlier observations that the antibiotic polymyxin B forms a highly stable molecular complex with the lipid A region of LPS (4), and that such polymyxin B-LPS complexes are unable to mitogenically stimulate B lymphocytes (5). In addition, we have attempted to distinguish between the two potential modes of action of phenol on LPS, namely, the chemical alteration of the lipid A or the removal of a tightly bound contaminant by phenol treatment. The results of the experiments we report here support the interpretation that mitogenic activity of positive preparations of LPS is associated with a low mol wt phenol soluble polypeptide of approximately 10,000 mol wt. After partial purification, this polypeptide intitiates a significant mitogenic response at concentrations as low as 10 mug/ml. We conclude that the C3H/HeJ strain of mouse is a true nonresponder to the stimulatory effects of the lipid A region of LPS.