Complement-mediated tissue injury in humans occurs upon deposition of immune complexes, such as in autoimmune diseases and acute respiratory distress syndrome. Acute lung inflammatory injury in wild-type and C3-/- mice after deposition of IgG immune complexes was of equivalent intensity and was C5a dependent, but injury was greatly attenuated in Hc-/- mice (Hc encodes C5). Injury in lungs of C3-/- mice and C5a levels in bronchoalveolar lavage (BAL) fluids from these mice were greatly reduced in the presence of antithrombin III (ATIII) or hirudin but were not reduced in similarly treated C3+/+ mice. Plasma from C3-/- mice contained threefold higher levels of thrombin activity compared to plasma from C3+/+ mice. There were higher levels of F2 mRNA (encoding prothrombin) as well as prothrombin and thrombin protein in liver of C3-/- mice compared to C3+/+ mice. A potent solid-phase C5 convertase was generated using plasma from either C3+/+ or C3-/- mice. Human C5 incubated with thrombin generated C5a that was biologically active. These data suggest that, in the genetic absence of C3, thrombin substitutes for the C3-dependent C5 convertase. This linkage between the complement and coagulation pathways may represent a new pathway of complement activation.

It is becoming increasingly clear that the autonomic nervous system and the immune system demonstrate cross-talk during inflammation by means of sympathetic and parasympathetic pathways. We investigated whether phagocytes are capable of de novo production of catecholamines, suggesting an autocrine/paracrine self-regulatory mechanism by catecholamines during inflammation, as has been described for lymphocytes. Here we show that exposure of phagocytes to lipopolysaccharide led to a release of catecholamines and an induction of catecholamine-generating and degrading enzymes, indicating the presence of the complete intracellular machinery for the generation, release and inactivation of catecholamines. To assess the importance of these findings in vivo, we chose two models of acute lung injury. Blockade of alpha2-adrenoreceptors or catecholamine-generating enzymes greatly suppressed lung inflammation, whereas the opposite was the case either for an alpha2-adrenoreceptor agonist or for inhibition of catecholamine-degrading enzymes. We were able to exclude T cells or sympathetic nerve endings as sources of the injury-modulating catecholamines. Our studies identify phagocytes as a new source of catecholamines, which enhance the inflammatory response.

In humans with sepsis, the onset of multiorgan failure (MOF), especially involving liver, lungs, and kidneys, is a well known complication that is associated with a high mortality rate. Our previous studies with the cecal ligation/puncture (CLP) model of sepsis in rats have revealed a C5a-induced defect in the respiratory burst of neutrophils. In the current CLP studies, MOF occurred during the first 48 h with development of liver dysfunction and pulmonary dysfunction (falling arterial partial pressure of O2, rising partial pressure of CO2). In this model an early respiratory alkalosis developed, followed by a metabolic acidosis with increased levels of blood lactate. During these events, blood neutrophils lost their chemotactic responsiveness both to C5a and to the bacterial chemotaxin, fMLP. Neutrophil dysfunction was associated with virtually complete loss in binding of C5a, but binding of fMLP remained normal. If CLP animals were treated with anti-C5a, indicators of MOF and lactate acidosis were greatly attenuated. Under the same conditions, C5a binding to blood neutrophils remained intact; in tandem, in vitro chemotactic responses to C5a and fMLP were retained. These data suggest that, in the CLP model of sepsis, treatment with anti-C5a prevents development of MOF and the accompanying onset of blood neutrophil dysfunction. This may explain the protective effects of anti-C5a in the CLP model of sepsis.

This study defines the molecular basis for defects in innate immunity involving neutrophils during cecal ligation/puncture (CLP)-induced sepsis in rats. Blood neutrophils from CLP rats demonstrated defective phagocytosis and defective assembly of NADPH oxidase, the latter being due to the inability of p47phox to translocate from the cytosol to the cell membrane of neutrophils after cell stimulation by phorbol ester (PMA). The appearance of these defects was prevented by in vivo blockade of C5a in CLP rats. In vitro exposure of neutrophils to C5a led to reduced surface expression of C5aR and defective assembly of NADPH oxidase, as defined by failure in phosphorylation of p47phox and its translocation to the cell membrane, together with failure in phosphorylation of p42/p44 mitogen-activated protein kinases. These data identify a molecular basis for defective innate immunity involving neutrophils during sepsis.

The complement activation product, C5a, is a powerful phlogistic factor. Using antibodies to detect human or rat C5a, incubation at pH 7.4 of human blood neutrophils or rat alveolar macrophages (AMs) with C5 in the presence of phorbol 12-myristate 13-acetate (PMA) led to generation of C5a. Rat AMs activated with lipopolysaccharide also generated C5a from C5. With activated neutrophils, extensive cleavage of C5 occurred, whereas activated macrophages had much more selective proteolytic activity for C5. Peripheral blood human or rat mononuclear cells and rat alveolar epithelial cells when stimulated with phorbol ester all failed to demonstrate an ability to cleave C5, suggesting a specificity of C5 cleavage by phagocytic cells. With rat AMs, C5a generation was time-dependent and was blocked if AMs were pretreated with inhibitors of transcription or protein synthesis (actinomycin D or cycloheximide). Similar treatment of activated human polymorphonuclear leukocytes only partially reduced C5a generation after addition of C5. C5a generated by activated AMs was biologically (chemotactically) active. This generation was sensitive to serine protease inhibitors but not to other classes of inhibitors. These data indicate that phagocytic cells, especially lung macrophages, can generate C5a from C5. In the context of the lung, this may represent an important C5a-generating pathway that is independent of the plasma complement system. The complement system generating the complement activation products, C3a, C5a, and C5b-9 and the cellular defense system involving macrophages and neutrophils are known to form the first line of defense (innate immunity) against microorganisms and other tissue-damaging factors.

Innate immune functions are known to be compromised during sepsis, often with lethal consequences. There is also evidence in rats that sepsis is associated with excessive complement activation and generation of the potent anaphylatoxin C5a. In the presence of a cyclic peptide antagonist (C5aRa) to the C5a receptor (C5aR), the binding of murine 125I-C5a to murine neutrophils was reduced, the in vitro chemotactic responses of mouse neutrophils to mouse C5a were markedly diminished, the acquired defect in hydrogen peroxide (H2O2) production of C5a-exposed neutrophils was reversed, and the lung permeability index (extravascular leakage of albumin) in mice after intrapulmonary deposition of IgG immune complexes was markedly diminished. Mice that developed sepsis after cecal ligation/puncture (CLP) and were treated with C5aRa had greatly improved survival rates. These data suggest that C5aRa interferes with neutrophil responses to C5a, preventing C5a-induced compromise of innate immunity during sepsis, with greatly improved survival rates after CLP.

Although acute lung injury (ALI) is an important problem in humans, its pathogenesis is poorly understood. Airway instillation of bacterial LPS, a known complement activator, represents a frequently used model of ALI. In the present study, pathways in the immunopathogenesis of ALI were evaluated. ALI was induced in wild-type, C3−/−, and C5−/− mice by airway deposition of LPS. To assess the relevant inflammatory mediators, bronchoalveolar lavage fluids were evaluated by ELISA analyses and various neutralizing Abs and receptor antagonists were administered in vivo. LPS-induced ALI was neutrophil-dependent, but it was not associated with generation of C5a in the lung and was independent of C3, C5, or C5a. Instead, LPS injury was associated with robust generation of macrophage migration inhibitory factor (MIF), leukotriene B4 (LTB4), and high mobility group box 1 protein (HMGB1) and required engagement of receptors for both MIF and LTB4. Neutralization of MIF or blockade of the MIF receptor and/or LTB4 receptor resulted in protection from LPS-induced ALI. These findings indicate that the MIF and LTB4 mediator pathways are involved in the immunopathogenesis of LPS-induced experimental ALI. Most strikingly, complement activation does not contribute to the development of ALI in the LPS model.

We evaluated antibodies to different peptide regions of rat C5a in the sepsis model of cecal ligation and puncture (CLP) for their protective effects in rats. Rabbit polyclonal antibodies were developed to the following peptide regions of rat C5a: amino-terminal region (A), residues 1-16; middle region (M), residues 17-36; and the carboxyl-terminal region (C), residues 58-77. With rat neutrophils, the chemotactic activity of rat C5a was significantly inhibited by antibodies with the following rank order: anti-C > anti-M >> anti-A. In vivo, antibodies to the M and C (but not A) regions of C5a were protective in experimental sepsis, as determined by survival over a 10-day period, in a dose-dependent manner. The relative protective efficacies of anti-C5a preparations (in descending order of efficacy) were anti-C > anti-M >> anti-A. In CLP rats, a delay in infusion of antibodies, which were injected at 6 or 12 h after CLP, still resulted in significant improvement in survival rates. These in vivo and in vitro data suggest that there are optimal targets on C5a for blockade during sepsis and that delayed infusion of anti-C5a antibody until after onset of clinical evidence of sepsis still provides protective effects.