Thermal injury, intravascular hemolysis, and toxic oxygen products.

Acute thermal injury of rat skin produces an early, acute hemoglobinemia that is associated with the presence in blood of osmotically fragile red cells (RBC) that do not contain on their surfaces measurable amounts of complement components. The hemoglobinemia and the appearance in blood of osmotically fragile RBC appear to be the result of complement activation, which leads to oxygen radical production by neutrophils and damage of RBC. This has been demonstrated in vitro as well as in vivo by the ability of antioxidant interventions or neutrophil or complement depletion procedures to prevent the appearance of osmotically fragile RBC and the release of hemoglobin. These data may be relevant to the complications of hemoglobinemia and hemoglobinuria accompanying thermal injury in humans.

Intravascular activation of complement and acute lung injury. Dependency on neutrophils and toxic oxygen metabolites.

Intravascular activation of the complement system with cobra venom factor results in acute lung injury, which has been quantitated by increases in lung vascular permeability. Cobra venom factor preparations devoid of phospholipase A2 activity retain full lung-damaging capacity. The lung injury is associated with the preceding appearance of chemotactic activity in the serum coincident with the development of a profound neutropenia. The chemotactic activity is immunochemically related to human C5a. Morphologic studies have revealed discontinuities in the endothelial cell lining of lung alveolar capillaries, damage and/or destruction of endothelial cells in these areas, plugging of pulmonary capillaries with neutrophils that are in direct contact with vascular basement membrane, the presence of fibrin in alveolar spaces and in areas adjacent to damaged endothelial cells, and intraalveolar hemorrhage. Lung injury is dramatically attenuated in animals that have been previously neutrophil depleted. Teh intravenous injection of superoxide dismutase or catalase also provides significant protection from the pulmonary damage. Very little protection from the pulmonary damage. Very little protection is afforded by pretreatment of rats with antihistamine. These studies suggest that intravascular activation of the complement system leads to neutrophil aggregation and activation, intrapulmonary capillary sequestration of neutrophils, and vascular injury, which may be related to production of toxic oxygen metabolites by complement-activated neutrophils.

Evidence for role of hydroxyl radical in complement and neutrophil-dependent tissue injury.

Using our recently described model of acute lung injury in rats after systemic activation of complement by cobra venom factor (CVF), we demonstrated that pretreatment of animals with human milk apolactoferrin (in its native or derivatized form), but not iron-saturated lactoferrin, provides significant protection against complement- and neutrophil-mediated lung injury. The synthetic iron chelator deferoxamine mesylate also affords protection from lung injury. The protective effects of apolactoferrin are not related to a blocking of CVF-induced complement activation. We also demonstrated that infusion of ionic iron, especially Fe3+, greatly potentiates lung vascular injury after systemic complement activation. Finally, protection from lung injury occurs in animals pretreated with the potent scavenger of hydroxyl radicals (OH.), dimethyl sulfoxide. Based on transmission electron microscopy, CVF-treated rats show leukoaggregates and endothelial cell destruction in interstitial pulmonary capillaries, along with intraalveolar hemorrhage and fibrin deposition. In animals protected with apolactoferrin, deferoxamine mesylate, or dimethyl sulfoxide, the morphological studies reveal leukoaggregates but no endothelial cell damage, hemorrhage, or fibrin deposition. These data support the concept that tissue injury that is complement and neutrophil dependent may be related to generation of OH. derived from H2O2 after leukocytic activation.

Acute lung injury in rat caused by immunoglobulin A immune complexes.

Mouse IgG and IgA, with reactivity to dinitrophenol conjugated to carrier protein, have been isolated from myeloma proteins by means of a variety of affinity techniques. The IgA was predominantly in the dimeric form. The in vitro and in vivo biological activities of IgA-containing immune complexes were assessed in the rat. IgA-containing immune complexes were demonstrated, in a dose-dependent manner in vitro, to activate neutrophils and to generate O.-2. In addition, these immune complexes showed evidence of complement activation in vitro, by the use of immunofixation techniques. When IgA was instilled into the airways of rats and antigen was injected intravenously, acute lung injury occurred, as reflected by increases in lung permeability and morphological changes consisting of blebbing of endothelial cells, intra-alveolar hemorrhage, and fibrin deposition. The lung changes were directly proportional to the amount of IgA instilled into the airways and failed to occur if intravenous injection of antigen was omitted. Lung injury did not occur in animals that received an intravenous injection of antigen in the absence of an airway injection of IgA. Lung injury related to IgA-containing immune complexes was complement dependent but neutrophil independent. In companion studies with mouse IgG-containing immune complexes, acute lung injury also occurred and had morphological features similar to those associated with IgA-induced lung injury except that, in the case of IgG immune complex-induced damage, neutrophils were more evident. Acute lung injury induced by IgG-containing immune complexes, whether of mouse or rabbit origin, was complement and neutrophil dependent. The similarities and differences between IgG- and IgA-associated acute immune complex-induced injury of rat lung were reinforced by the use of morphometry techniques. Studies with another monoclonal IgA antibody-containing antigen-binding activity to phosphorylcholine also demonstrated the ability of IgA antibody to cause acute lung injury in the rat. Neither antibody alone nor antigen (phosphorylcholine linked to bovine serum albumin) alone produced evidence of lung injury. These studies indicate for the first time that immune complexes containing IgA have lung-damaging properties and that the pathogenic mechanisms are different from those associated with IgG-associated immune complex-induced acute lung injury.

Systemic complement activation, lung injury, and products of lipid peroxidation.

Previously we have demonstrated that systemic activation of the complement system after intravenous injection of cobra venom factor (CVF) results in acute lung injury as reflected by increases in the vascular permeability of the lung as well as by morphologic evidence of damage to lung vascular endothelial cells. In using the vascular permeability of the lung as the reference, the current studies show a quantitative correlation between lung injury and the appearance in plasma of lipid peroxidation products (conjugated dienes) as well as increased concentrations of lactic dehydrogenase (LDH) and one of its isoenzymes (LDH-4). After injection of CVF, extracts of lungs also showed elevated levels of conjugated dienes, whereas no elevations were found in extracts of liver, kidney, and spleen. There was no evidence in CVF-injected rats of renal or hepatic injury as reflected by the lack of development of proteinuria and the failure to detect increased serum levels of liver-related enzymes. Other peroxidation products identified in plasma of CVF-injected rats involved hydroperoxides and fluorescent compounds with features of Schiff bases. Not surprisingly, malondialdehyde was not found to be a reliable plasma indicator of lipid peroxidation associated with oxygen radical-mediated lung vascular injury. In using a model of oxygen radical-independent lung injury induced by oleic acid, although large amounts of LDH and LDH-4 were found in the plasma, no increases in plasma levels of conjugated dienes were detected. In CVF-injected animals treated with interventions protective against lung injury (neutrophil depletion, catalase, hydroxyl radical scavengers, or iron chelators), there were striking reductions in the plasma levels of conjugated dienes, hydroperoxides, and fluorochromic products. Morphometric analysis of lung sections revealed that the protective interventions did not interfere with the accumulation of neutrophils in lung interstitial capillaries after systemic activation of complement. In vitro studies with phorbol-stimulated neutrophils failed to demonstrate appearance of conjugated dienes, suggesting that the dienes appearing in plasma of CVF-injected animals are not the result of autotoxic changes in neutrophils. The data presented in this paper suggest that acute lung injury mediated by oxygen radicals derived from phagocytic cells can be monitored by the appearance in plasma of products of lipid peroxidation.

A major role for neutrophils in experimental bullous pemphigoid.

Bullous pemphigoid (BP) is an inflammatory subepidermal blistering disease associated with an IgG autoimmune response to the hemidesmosomal protein, BP180. Using a passive transfer mouse model, our group has shown previously that antibodies to the murine BP180 (mBP180) ectodomain are capable of triggering a blistering skin disease that closely mimics human BP. In this study, we investigated the role of neutrophils in the immunopathogenesis of this disease model. BALB/c mice depleted of circulating neutrophils by treatment with neutrophil-specific antibodies were no longer susceptible to the pathogenic effects of anti-mBP180 IgG. IgG and complement were deposited at the dermal-epidermal junction of these animals, but there was no evidence of inflammatory infiltration or blistering. C5-deficient mice, which are resistant to the pathogenic activity of anti-mBP180 IgG, could be made susceptible to this IgG-mediated blistering disease by intradermal administration of a neutrophil chemoattractant, IL-8 or C5a. Intraperitoneal injection of IL-8, which sequesters neutrophils in the peritoneal cavity, interferes with anti-mBP180-induced neutrophilic infiltration of the skin and prevented the development of BP disease in BALB/c mice. These findings provide the first direct evidence that neutrophils recruited to the skin via a C5-dependent pathway play an essential role in subepidermal blister formation in experimental BP, and suggest new directions for disease intervention.

The role of complement in experimental bullous pemphigoid.

Bullous pemphigoid (BP) is a blistering skin disease associated with an IgG autoimmune response directed against the ectodomain of the hemidesmosomal protein, BP180. An animal model of BP has recently been developed by our laboratory based on the passive transfer of rabbit antimurine BP180 antibodies into neonatal BALB/c mice. The experimental animals develop a blistering disease that reproduces all of the key immunopathological features of BP. In the present study we have investigated the role of complement in the pathogenesis of subepidermal blistering in the mouse model of BP. We demonstrate the following. (a) Rabbit anti-murine-BP180 IgG was effective in inducing cutaneous blisters in a C5-sufficient mouse strain, but failed to induce disease in the syngeneic C5-deficient strain; (b) neonatal BALB/c mice, pretreated with cobra venom factor to deplete complement, became resistant to the pathogenic effects of the anti-BP180 IgG; (c) F(ab')2 fragments generated from the anti-BP180 IgG exhibited no pathogenic activity in the mouse model; and (d) histologic evaluation of the skin of mice described in points b and c above showed minimal or no neutrophilic cell infiltration in the upper dermis. Thus, anti-BP180 antibodies trigger subepidermal blistering in this BP model via complement activation. This experimental model of BP should greatly facilitate future studies on the pathophysiology of autoantibody-mediated diseases of the dermal-epidermal junction.

Requirement and role of C5a in acute lung inflammatory injury in rats.

The complement activation product, C5a, may play a key role in the acute inflammatory response. Polyclonal antibody to rat C5a was used to define the requirements for C5a in neutrophil-dependent inflammatory lung injury after systemic activation of complement by cobra venom factor (CVF) or after intrapulmonary deposition of IgG immune complexes. In the CVF model, intravenous infusion (but not intratracheal instillation) of anti-C5a produced a dose-dependent reduction in lung permeability and in lung content of myeloperoxidase. In C6-deficient rats, CVF infusion caused the same level of lung injury (measured by leak of 125I-albumin) as found in C6-sufficient rats. In the IgG immune complex model of lung injury, anti-C5a administered intratracheally (but not intravenously) reduced in a dose-dependent manner both the increase in lung vascular permeability as well as the buildup of lung myeloperoxidase. Treatment with anti-C5a greatly suppressed upregulation of lung vascular intercellular adhesion molecule-1 (ICAM-1). This was correlated with a substantial drop in levels of TNFalpha in bronchoalveolar fluids. These data demonstrate the requirement for C5a in the two models of injury. In the IgG immune complex model, C5a is required for the full production of TNFalpha and the corresponding upregulation of lung vascular ICAM-1.

Lung injury and complement activation: role of neutrophils and xanthine oxidase.

Evidence is presented that oxygen products generated from xanthine oxidase (XO) may also be involved in the pathogenesis of neutrophil-mediate lung injury following intravascular activation of complement with cobra venom factor (CVF). CVF injection in rats resulted in a rapid increase in plasma of both XO activity (but not xanthine dehydrogenase) and its reaction product, uric acid. These changes were greatly attenuated in allopurinol-treated animals. The appearance of XO activity was paralleled by a rise in plasma of histamine. Prevention of histamine release by pretreatment of rats with cromolyn abolished both the rise in plasma histamine and the increase in XO activity. Since we have previously shown that histamine can enhance XO activity in vitro and in vivo (Am. J. Pathol. 135:203, 1989), these observations suggest that the increase in plasma XO activity following CVF injection is related to the appearance in plasma of histamine. Accordingly, pretreatment of rats with xanthine oxidase inhibitors (allopurinol, lodoxamide) or prevention of histamine release by pretreatment with cromolyn significantly attenuated development of lung injury following injection of CVF. Our data support the concept that oxygen radicals derived from both neutrophils and XO are playing a role in the CVF-induced acute lung injury.

Role of oxygen radicals in experimental allergic uveitis.

The experimental autoimmune disease elicited by a large dose of retinal S antigen in guinea pigs is characterized by massive necrotizing uveitis and retinitis. Treatment of these animals with the antioxidants superoxide dismutase, catalase, and sodium benzoate resulted in marked reduction of uveal inflammation. The attenuated inflammation was characterized by a relatively well-preserved retina and retinal pigment epithelium along with a reduction of subretinal exudate and vitreous inflammation. These findings suggest that reactive oxygen metabolites may play a role in the destruction of ocular tissue and amplification of the inflammatory process in experimental uveitis.

Adhesion molecules in experimental phacoanaphylactic endophthalmitis.

Intraocular accumulation of inflammatory neutrophils is an important feature of experimental phacoanaphylactic endophthalmitis (EPE). Increasing evidence suggests that localization of neutrophils to the site of inflammation requires the participation of neutrophil and endothelial adhesion molecules. These studies were undertaken to determine if blocking of adhesion molecules on neutrophils (CD18) or endothelium (ELAM-1) could attenuate EPE in Lewis rats. Treatment of experimental animals with anti-CD18 or anti-ELAM-1 significantly suppressed intraocular neutrophil accumulation, retinal hemorrhage, and vasculitis, and attenuated retinal edema formation by 48% and 70%, respectively. These observations demonstrate that antibodies directed against adhesion molecules on the neutrophil (CD18) or the vascular endothelial cell (ELAM-1) exhibit potent anti-inflammatory effects, resulting in a striking amelioration of injury in EPE in rats.

Requirement for C5a in lung vascular injury following thermal trauma to rat skin.

Previous studies in rats have shown that deep second degree dermal burns, involving 28-30% of total body surface area, result in systemic complement activation, appearance in plasma of chemotactic activity, sequestration of blood neutrophils in lung capillaries, and development of neutrophil-dependent dermal vascular and lung vascular injury. Although blockade of complement activation or depletion of complement before skin burns has resulted in significant attenuation of tissue injury both locally and distally (in lung), a role for C5a in these events is unclear. In the following studies, we demonstrate the presence of C5a and neutrophil chemotactic activity in serum and in lung homogenates after thermal injury. C5a has also been found in bronchoalveolar lavage fluids of thermally injured animals. Treatment of animals with a polyclonal neutralizing rabbit antibody to rat C5a was lung protective. The protective effects of the antibody (anti-C5a) were associated with diminished vascular permeability changes, as well as reduced tissue build-up of myeloperoxidase. Anti-C5a also prevented up-regulation of lung vascular ICAM-1 (intercellular adhesion molecule-1) in skin-burned rats. These observations indicate that C5a is essential for development of neutrophil accumulation and vascular permeability increases in distant (lung) organs after thermal trauma to skin. The protective effects of anti-C5a in lung, appear to be related to prevention of up-regulation of vascular ICAM-1. Accordingly, C5a may represent a target for clinical approaches in the treatment of organ injury following thermal trauma.

Complement and experimental respiratory failure.

Activation of the complement system within the lung can lead to acute pulmonary damage and dysfunction. Based on a variety of experimental models it is now apparent that lung injury is related to complement-induced generation of oxygen derived free radicals from neutrophils and from macrophages. In addition to the oxygen radicals, it is also possible that the conversion of hydrogen peroxide by myeloperoxidase to hypochlorous acid also contributes to the injury. Exposure of the pulmonary microvasculature to oxygen radicals generated from complement-activated neutrophils causes focal damage and necrosis of endothelial cells. IgG immune complex-induced injury of lung is also complement and neutrophil dependent and oxygen radical mediated. In contrast, lung injury produced by IgA immune complexes is neutrophil independent, complement dependent and oxygen radical mediated. There is now increasing evidence that oxygen radicals are not only directly tissue-toxic but also able to potentiate the activity of leukocytic proteases. In all of these models the lung can be protected from injury by pretreatment of the animals with either scavengers of hydroxyl radical or with agents that prevent its formation (e.g. catalase, iron chelators). Data from these models may have direct clinical relevance to conditions such as adult respiratory distress syndrome where lung injury is probably oxygen radical mediated.

Systemic effects of prosthetic vascular graft implantation.

The effect of prosthetic vascular graft implantation on systemic complement activation, as well as leukocyte and platelet counts, was studied in rabbits. Animals underwent aortic clamping and declamping alone (group I) or with implantation of an externally supported PTFE graft (group II) or preclotted knitted Dacron graft (group III). Fifteen minutes after declamping, mild activation of complement reflected as a fall in CH50 to 72% of preoperative baseline values was observed in group I animals. Comparatively greater activation was seen in group II (43% baseline) and group III (28% baseline) animals. The differences between groups I and II (p less than 0.05) and I and III (p less than 0.01) were significant. The drop in platelet count was greater in group II (66% baseline) and group III (53% baseline) animals compared with group I (86% baseline). Neutrophil counts decreased in group I animals (75% baseline), but, again, this fall was more marked in group II (35% baseline) and group III (43% baseline) animals. The difference between groups I and II was significant (p less than 0.05). Similarly, monocyte count decreased in group I animals (80% baseline), which was also more marked in group II (32% baseline) and group III (31% baseline) animals. The differences between group I and groups II and III were significant (p less than 0.05 for each). These trends persisted throughout the first postoperative hour. These results suggest that the blood-surface interactions produced systemic effects in vivo.

Activation of complement by hydroxyl radical in thermal injury.

Complement activation resulting from local burn injury of skin and other soft tissues can be linked to systemic complications, such as intravascular hemolysis, neutrophil activation, and acute lung injury. This study was designed to clarify the relationship between cutaneous thermal injury, oxygen radical formation, and complement activation in vivo. A model for "selective" venous sampling from the area of a partial-thickness cutaneous burn over 25% to 30% of the total body surface in the rat was developed. Interventions involving oxygen radical scavengers, antioxidant enzymes, xanthine oxidase inhibitors, an iron chelator, complement depletion, and neutrophil depletion were used to probe the nature of the oxygen products involved in complement activation. Plasma from the area of burn was examined for total hemolytic complement activity, content of C5a-related chemotactic peptide, and relationship of oxygen products to appearance of this peptide. Xanthine oxidase inhibitors, hydroxyl radical scavengers, and complement depletion diminished the generation of C5a activity at the burn site, whereas neutrophil depletion was without effect. These data suggest that C5a activity may be related to oxygen products from xanthine oxidase. The catalase sensitivity and iron dependency of C5a generation suggest that hydroxyl radical may be related to complement activation and C5a appearance. This is the first report to directly link oxygen radical generation and complement activation in vivo.

Xanthine oxidase: its role in the no-reflow phenomenon.

This study was designed to probe the hypothesis that oxygen-derived free radicals are involved in initiation of the no-reflow phenomenon. We developed a reproducible model of no reflow in the rat hind limb. Laser Doppler studies confirmed that the hind limbs perfused well after 2 or 4 hours of ischemia, but perfusion ceased in the first 10 minutes after 6 hours of ischemia. Venous blood samples and biopsy specimens of skin and muscle were taken after 2 and 4 hours of ischemia to study tissue injury. Blood samples were evaluated for xanthine oxidase (XO), xanthine dehydrogenase, and creatine phosphokinase (CPK) activities. Conjugated dienes and iodine 125-labeled albumin extravasation were quantified in tissue samples. Groups of animals were treated with inhibitors of XO (allopurinol), antioxidant enzymes (superoxide dismutase plus catalase), and free radical scavengers (dimethyl sulfoxide and dimethyl thiourea) to assess the roles of free radicals in ischemia-reperfusion injury in the hind limbs. After 4 hours of ischemia followed by reperfusion, plasma XO activity rose threefold over preischemia levels (p less than 0.05). Xanthine dehydrogenase activity did not change; conjugated diene levels in muscle rose twofold; CPK levels rose sixfold, and 125I albumin extravasation rose twofold (p less than 0.05). Pretreatment with the XO inhibitor allopurinol reduced XO activity to negligible levels and significantly attenuated conjugated diene levels, CPK levels, and albumin extravasation. Albumin extravasation was also significantly attenuated by pretreating animals with superoxide dismutase together with catalase, dimethyl thiourea, and dimethyl sulfoxide. In all animals pretreated with allopurinol or superoxide dismutase and catalase, reperfusion persisted after 6 hours of ischemia. These data suggest that, in ischemia followed by reperfusion, tissue injury is related to oxygen products derived from XO activity.

Acute skin injury releases neutrophil chemoattractants.

BACKGROUND: Progressive or ongoing skin necrosis after traumatic injury is well known. Experimental evidence has associated these events with neutrophil activation and secondary oxidant injury. To determine the mechanism by which neutrophils migrate to a site of injury, cytokine release from injured skin was measured. METHODS: Twenty-five skin biopsy specimens of acute partial thickness skin injuries were compared with uninjured skin of the same patient. Conditioned medium from explanted skin was assayed for tumor necrosis factor (TNF), interleukin-6 (IL-6), and IL-8. RESULTS: Acute skin injury resulted in a significant release of IL-8 but not IL-6 or TNF. In eight patients gradient cytokine release was found; IL-8 levels for partial thickness burn were 26.4 ng/ml, for unburned skin adjacent to the burn were 2.1 ng/ml, and for distal normal skin were 0.2 ng/ml. CONCLUSIONS: IL-8 is released from acutely injured skin; IL-6 and TNF are not. This selective release suggests a mechanism whereby neutrophils are recruited into injured tissue. These neutrophils might then induce further injury, increasing the extent of posttraumatic tissue loss.

Partial liquid ventilation protects lung during resuscitation from shock.

Preliminary animal experience with partial liquid ventilation (PLV) suggests that this therapy may diminish neutrophil invasion and capillary leak during acute lung injury. We sought to confirm these findings in a model of shock-induced lung injury. Sixty anesthetized rats were studied. After hemorrhage to an arterial pressure of 25 mmHg for 45 min, animals were resuscitated with blood and saline and treated with gas ventilation alone or with 5 ml/kg of intratracheally administered perflubron. Myeloperoxidase activity was used to measure lung neutrophil content. A permeability index (the bronchoalveolar-to-blood ratio of 125I-labeled albumin activity) quantified alveolar leak. Injury caused an increase in myeloperoxidase that was reversed by PLV (injury = 0.837 +/- 0.452, PLV = 0.257 +/- 0.165; P < 0.01). Capillary permeability also increased with hemorrhage, with a strong trend toward improvement in the PLV group (permeability indexes: injury = 0.094 +/- 0.102, PLV = 0.045 +/- 0.045; 95% confidence interval for injury--PLV: -0.024, 0.1219). We conclude that PLV is associated with a decrease in pulmonary neutrophil accumulation and a trend toward decreased capillary leak after hemorrhagic shock.

Systemic and lung physiological changes in rats after intravascular activation of complement.

Systemic complement activation has been noted in a variety of shock states, and there is growing evidence that, in addition to being proinflammatory effectors, products of complement activation contribute directly to generalized manifestations of shock, such as hypotension and acidosis. To study the effects of complement activation, we examined responses in rats to systemic activation of complement with cobra venom factor (CVF), including blood pressure, metabolic acidosis, changes in vascular permeability, and lung function. High doses of CVF produced circulatory collapse (mean arterial pressure = 110 +/- 16 and 35 +/- 9 mmHg in control and with CVF, respectively, P < 0.05), metabolic acidosis (HCO concentration = 27.8 +/- 1.7 and 9.6 +/- 3.4 meq/l in control and with CVF, respectively, P < 0.05), extravasation of albumin into the lung and gut, and modest arterial hypoxemia (PO2 = 486 +/- 51 and 201 +/- 36 Torr in control and during 100% O2 breathing, respectively, P < 0.05). Prior depletion of complement protected against these abnormalities. Other interventions, including neutrophil depletion and cyclooxygenase inhibition, prevented lung injury but had much less effect on systemic hemodynamics or gut permeability, suggesting that complement activation products induce injury by neutrophil- and cyclooxygenase-dependent pathways in the lung but not in the gut. These studies underscore the significant systemic abnormalities developing after systemic activation of complement.

Detrimental effects of complement activation in hemorrhagic shock.

The complement system has been implicated in early inflammatory events and a variety of shock states. In rats, we measured complement activation after hemorrhage and examined the hemodynamic and metabolic effects of complement depletion before injury and worsening of complement activation after hemorrhage and resuscitation [with a carboxypeptidase N inhibitor (CPNI), which blocks the clearance of C5a]. Rats were bled to a mean arterial pressure of 30 mmHg for 50 min and were then resuscitated for 2 h. Shock resulted in significant evidence of complement consumption, with serum hemolytic activity being reduced by 33% (P < 0.05). Complement depletion before injury did not affect hemorrhage volume (complement depleted = 28 +/- 1 ml/kg, complement intact = 29 +/- 1 ml/kg, P = 0.74) but improved postresuscitation mean arterial pressure by 37 mmHg (P < 0.05) and serum bicarbonate levels (complement depleted = 22 +/- 3 meq/ml, complement intact = 13 +/- 8 meq/ml, P < 0.05). Pretreatment with CPNI was lethal in 80% of treated animals vs. the untreated hemorrhaged group in which no deaths occurred (P < 0.05). In this model of hemorrhagic shock, complement activation appeared to contribute to progressive hypotension and metabolic acidosis seen after resuscitation. The lethality of CPNI during acute blood loss suggests that the anaphylatoxins are important in the pathophysiological events involved in hemorrhagic shock.