Complement kinetics during continuous arteriovenous hemofiltration: studies with a new polysulfone hemofilter

Biocompatibility of the Renaflo polysulfone hemofilter was assessed during treatments with continuous arteriovenous hemofiltration. Pre- versus postfilter levels of C3a, C4a, C5a and leukocyte counts were measured prior to and 15, 60 and 90 min after placement of the Renaflo. Results revealed a small, and previously unrecognized, increase in postfilter C3a (1,224 +/- 56 vs. 1,535 +/- 150 ng/ml, mean +/- SEM, n = 15; p less than 0.01). A comparable increase was also found during treatment with an Amicon polysulfone filter. There were no other significant changes in any parameter studied and no patient exhibited symptoms related to complement activation. We conclude that the Renaflo filter is safe for use during continuous arteriovenous hemofiltration. Although clinically insignificant, there is a detectable activation of C3 by polysulfone membranes.

Limitation of complement activation by perfluorocarbon emulsions: superiority of lecithin-emulsified preparations

A considerable body of evidence suggests that complement and granulocyte activation may be important in the occasional adverse reactions which occur on the administration of perfluorocarbon emulsions to patients. In an attempt to develop an emulsion which was less prone to activate complement than was the first product clinically tested, we devised an in-vitro protocol for the testing of an emulsion's ability to activate complement: emulsions were serially diluted, and were incubated with minimally-heparinized plasmas from each of several donors; the generation of C3a was assessed by commercial immunoassay. In such a system, F-Adamantane emulsions prepared with highly-purified lecithin as the emulsifier were consistently less complement-activating than comparable emulsions prepared with Pluronic F-68 as the emulsifier. Emulsification technique seemed unimportant, and HPLC fractionation of the Pluronics identified no innocuous subfraction. Application of these findings to an animal model is underway, to see if the superiority of lecithin-based emulsions is also evident in vivo.

Neutrophil chemotactic activity produced by normal and activated human bronchoalveolar lavage cells

Activated macrophages can secrete a number of mediators that can attract inflammatory cells and enhance secretion of phlogistic substances from these cells. The ultimate effect of activated bronchoalveolar lavage (BAL) cells may be fibrotic lung injury. Inasmuch as pulmonary sarcoidosis is a disease associated with spontaneous activation of macrophages and lymphocytes among BAL cells, cells obtained from patients with sarcoidosis were compared with normal cells. We report that adherent BAL cells in culture from patients with sarcoidosis (n = 21) release during a resting period in vitro more chemotactic activity for neutrophils (PMNs) than do BAL cells from normal individuals (n = 14). After density fractionation of the respiratory cells by albumin gradient, cells from high-density fractions in the group with sarcoidosis secrete more chemotactic activity for neutrophils than cells from less dense fractions. The PMN chemotactic activity spontaneously released in vitro by BAL cells from patients with sarcoidosis correlates with the percentage of PMNs recovered by BAL. Immunochemical bioassay and high-performance liquid chromatographic (HPLC) analysis of BAL cell supernatants revealed a complex pattern of chemotactic factors to be present. Generally, three peaks of chemotactic activity were noted on HPLC 1-60 separations at greater than 20 kd, 8 to 10 kd, and less than 1 kd apparent molecular weights. Significantly, interleukin-1 was present in these supernatants, whereas complement components and leukotriene B4 were absent. Sarcoid BAL cells, principally alveolar macrophages, are activated in vivo as manifested by spontaneous secretion of chemotactic factors for PMNs in vitro. Interleukin-1 and other less well characterized molecules were detected. The presence of PMNs among the lavage cells of some patients with sarcoidosis appears to be an in vivo biologic correlate of this activation. These data provide additional criteria of BAL cell activation in patients with pulmonary sarcoidosis and provide further evidence concerning factors that attract inflammatory cells into the lung.

Elevated plasma C3 anaphylatoxin levels in rheumatoid arthritis patients

Because synovial fluid levels of the C3 anaphylatoxins, C3a and C3a desArg, are more than 7 times higher in patients with rheumatoid arthritis (RA) than in patients with degenerative joint disease or traumatic arthritis, we conducted a prospective study of plasma levels of C3 and C5 anaphylatoxins in 11 RA patients. The mean level of C3 anaphylatoxin in 37 RA plasma specimens was twice as high as that in samples obtained from 9 normal volunteers (272.9 +/- 106.1 ng/ml versus 133.9 +/- 19.4 ng/ml) (P less than 0.05). We found that the joint index and the disease activity index were correlated with the plasma C3 anaphylatoxin level, and that the correlation values were similar to those for serum C-reactive protein level and for the Westergren erythrocyte sedimentation rate. In correlations of the joint index and disease activity index with combinations of laboratory test results, the plasma C3a level, the serum C-reactive protein level, and the erythrocyte sedimentation rate, we found that almost any pair of laboratory tests correlated more strongly than did any one test.

Assessment of blood interactions with the Therapore Apheresis System

Biocompatibility evaluations are a complex issue involving not only classical toxicity and animal testing but also the interactions of biological systems. The evaluations of the Therapore Primary System are reported for three model systems: (1) in vitro human blood, (2) a baboon (Papio anubis) animal model, and (3) normal human volunteer apheresis. Measurements of both complement (C3a) and platelet (BTG and TXB2) activation products during the apheresis procedures are given as a sensitive indicator of biocompatibility. No statistically significant elevations of these activation products were observed. Thus, the Therapore System was found to have negligible perturbations of the interacting biological systems evaluated even with the highly sensitive RIA procedures employed. These observations and methods establish a solid foundation for comprehensive biocompatibility evaluations of future extracorporeal devices.

Activation of the complement system by recombinant tissue plasminogen activator

Recent trials have shown that recombinant tissue plasminogen activator (rt-PA) is an effective thrombolytic agent in patients with acute myocardial infarction. Because rt-PA converts plasminogen to plasmin, which is known to activate complement in vitro, we tested the hypothesis that rt-PA can induce in vivo activation of complement. Studies were performed in 12 patients with acute myocardial infarction. Six control patients had patent coronary arteries and did not receive rt-PA; these patients had normal values of the components of the complement system C4a (409 +/- 111 ng/ml) and C5a (8.8 +/- 1.8 ng/ml) with a slight elevation of C3a (204 +/- 6.6 ng/ml) in samples collected before coronary arteriography (253 +/- 25 minutes after onset of pain). After coronary arteriography, there was a slight decrease in the values of C4a (224 +/- 37 ng/ml), C5a (7.3 +/- 1.3 ng/ml) and C3a (164 +/- 35 ng/ml). The remaining six patients had complete coronary occlusion and received rt-PA (80 to 150 mg intravenously). In this treated group, before coronary arteriography the values of C4a (406 +/- 51.6 ng/ml) and C5a (8.1 +/- 1.9 ng/ml) were normal, and those of C3a were slightly elevated (250 +/- 76 ng/ml). All complement values obtained before rt-PA were similar to those in the untreated group. However, after administration of rt-PA (but before any angiographically detectable reperfusion), there was a striking increase in C4a (2,265 +/- 480 ng/ml; p less than 0.01), C3a (600 +/- 89 ng/ml; p less than 0.05) and C5a (30.0 +/- 4.5 ng/ml; p less than 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)

Pulmonary dysfunction is common during a cytomegalovirus infection after renal transplantation even in asymptomatic patients. Possible relationship with complement activation

In 24 patients with a cadaveric renal allograft, serial measurements after transplantation were made of the diffusing capacity for carbon monoxide (DLCO) together with serial measurements of C3d, the stable conversion product of the complement factor C3, and determinations of the anaphylatoxin C3a. Twelve patients were studied during an active cytomegalovirus (CMV) infection, and 12 patients were studied during allograft rejection or during a stable phase after renal transplantation (control subjects). No patients had pulmonary symptoms nor abnormal chest radiographs or arterial blood gas determinations. During an active CMV infection, DLCO was significantly reduced compared with the measurements made during allograft rejection or during a stable phase after renal transplantation. This was true both with (p less than 0.01) and without (p less than 0.01) correction for the hemoglobin concentration. Serum C3d levels were increased in 8 of the 12 patients with a CMV infection, but not in any of the patients in the control group. In 8 patients with a CMV infection, measurements were made of the anaphylatoxin C3a, and were found to be significantly higher than the levels in the control population (p less than 0.01). We conclude that our data are consistent with pulmonary dysfunction in every patient with an active CMV infection. The concomitant findings of complement activation and formation of anaphylatoxins suggest a causal relationship of the complement activation and a decreased DLCO, although further studies are warranted to determine the exact role of complement in the pulmonary events during an active CMV infection after renal transplantation.(ABSTRACT TRUNCATED AT 250 WORDS)

Solution conformation of carboxy-terminal fragments of the third component of human complement C3: proton nuclear magnetic resonance study of C3a, des-Arg-C3a, and C3a Arg69

A proton nuclear magnetic resonance (NMR) study is reported of the solution conformation of human C3a, that is released on activation of C3, the third component of complement. The intact C3a was used along with des-Arg-C3a, which is formed on cleavage of Arg-77 at the C terminal of C3a, and C3a Arg69, which is a 69-residue fragment produced on tryptic digestion of C3. A method of carboxypeptidase digestion/difference spectroscopy (Endo & Arata (1985) Biochemistry 24, 1561-1568) was extensively used for the spectral assignments of Ile-43, Ile-60, Leu-63, Tyr-15, and Tyr-59. On the basis of the results of nuclear Overhauser effect (NOE) measurements, we discuss the solution conformation of the C3a molecule. It has been concluded that removal of Arg-77, which is essential for expression of the biological activity of C3a, does not induce any significant change in the solution conformation of the C3a molecule. The C3a molecule is known to consist of a core region that comprises segment Tyr-15-Tyr-59. We conclude that in solution the C terminal segment sticks out of the core and takes on a helix-like conformation. Possible roles of the core region and the N terminal segment in maintaining the conformation of the C terminal segment are briefly discussed.

Complement and dengue haemorrhagic fever/shock syndrome

The complement system is activated in DHF/DSS. The peak of activation and the presence of C3a and C5a anaphylatoxins coincided with the onset of shock and leakage. The levels of C3a correlated well with disease severity. This indicated an important role of the complement system in the pathogenesis of shock. Circulating immune complexes as assayed by two standard techniques were not detected in the majority of patients, and if detected were found in small amount. The role of circulating immune complexes in the activation of complement in DHF/DSS needs to be reinvestigated, and other possible mechanisms leading to complement activation should be sought.

Complement and anaphylatoxin responses to cross-clamping of the aorta. Studies during general anaesthesia with or without extradural blockade

Complement activation and anaphylatoxin formation were studied in 28 patients undergoing aortic reconstructive surgery under general anaesthesia. In 12 of the patients an extradural catheter was placed in the L1-2 space and 2% mepivacaine was injected until blockade reached the T4 level. The remaining 16 patients were operated upon under general anaesthesia alone. Complement activation was not obvious until aortic cross-clamping. The release of anaphylatoxin was less pronounced during cross-clamping in the extradural group.

The ischaemic leg as a source of complement activation

Complement activation before and after surgical resection was studied in 40 patients with ischaemic legs (17 with acute and 23 with chronic ischaemia). Plasma anaphylatoxin concentrations (C3a and C5a) and anaphylatoxin inhibitor (AI) activity were studied pre-operatively, and 1 h, 24 h and 1 week postoperatively. Increased plasma anaphylatoxin concentrations were found pre-operatively. Anaphylatoxin levels in plasma had returned to normal one week postoperatively. Patients with acute ischaemia had higher plasma C3a and C5a pre-operatively and during the first 24 h postoperatively than patients with chronic ischaemia. Anaphylatoxins increase vascular permeability, smooth muscle contraction and histamine release from mast cells. Their enhanced activity might be one of the explanations for respiratory, hepatic and renal insufficiency before surgical resection in patients with extensive ischaemia.

C3a(C3adesArg) induces production and release of interleukin 1 by cultured human monocytes

Purified human C3a(C3adesArg) induced dose-dependent generation of intracellular IL 1 activity and release of IL 1 in cultures of human mononuclear adherent cells in serum-free conditions. Concentrations of C3a(C3adesArg) of 10(-8) M and 6 hr of culture were sufficient to induce production of cell-associated IL 1, as detected in monocyte lysates. Ten- to 100-fold higher concentrations of C3a(C3adesArg) and 24 hr of culture were required for induction of IL 1 release. Release of IL 1 induced by suboptimal amounts of C3a(C3adesArg) was greatly enhanced by the addition of indomethacin to the culture medium. Contamination with C5a of the C3a(C3adesArg) preparation did not account for C3a(C3adesArg)-induced IL 1 production. Induction of IL 1 activity by C3a(C3adesArg) was not due to contaminating LPS, as indicated by the following observations: the amount of contaminating LPS in C3a(C3adesArg) was below that which could induce IL 1 release from human monocytes in serum-free conditions; induction of IL 1 by C3a(C3adesArg) was not suppressed by polymyxin B; kinetics of IL 1 production and release in the presence of C3a(C3adesArg) differed from those observed in the presence of LPS; and sialated gangliosides, which inhibit IL 1 release induced by LPS, had no effect on the induction of IL 1 by C3a(C3adesArg). The C3a(C3adesArg) preparation used in this study mostly contained the desArg derivative, suggesting that, in contrast with the requirement for an intact C-terminal arginyl residue for the spasmogenic activity of C3a, both C3a and its C3adesArg derivative may interact with receptors on human monocytes. By inducing IL 1 production and release, C3a(C3adesArg) may contribute to the generation of the inflammatory process and the regulation of the immune response.

Anaphylotoxin levels following thermal injury

Following thermal injury, anaphylotoxins (C3a and C5a) are generated by complement activity. We measured C3a and C5a levels serially in 18 seriously burned patients (mean burn size = 37.9 per cent body surface area) over the initial 3 weeks following injury. Values for C3a were significantly increased when compared with normal controls, with maximum levels during the second week, while C5a levels were slightly elevated initially. These levels did not correlate with the extent of tissue injury, the development of septic complications or the clinical outcome. However, it appears likely that C3a and C5a may be responsible for some of the pathophysiological alterations observed following thermal injury.

Activation of human serum complement with allergens. I. Generation of C3a, C4a, and C5a and induction of human neutrophil aggregation

To understand the relevance of allergens in generation of C3a, C4a, and C5a in normal human serum, we studied extracts of several allergens (house dust, house dust mite, Aspergillus fumigatus, and perennial ryegrass). Known complement activators zymosan and endotoxin were used as controls. Generation of C3a, C4a, and C5a (determined by radioimmunoassays) occurred with extracts of house dust and Aspergillus more than with house dust mite and ryegrass. Anaphylatoxins were produced both in dose and time-dependent fashions. Serum activated with house dust and Aspergillus extracts induced neutrophil aggregation when this serum was added to neutrophil suspensions. Less aggregation occurred when house dust mite and perennial ryegrass extract-activated sera were added to human neutrophils. We propose that some allergens may induce biologic responses by activation of sera and generation of anaphylatoxins as well as by IgE-mediated responses.

Complement activation by hydroxyethylmethacrylate-ethylmethacrylate copolymers

Certain biomaterials, including nylon oxygenator and cellulosic dialysis membranes, are potent activators of human complement. In this study, the effect of polymers containing 2-hydroxyethyl methacrylate (HEMA) on the human complement system was investigated. Copolymers of HEMA with ethylmethacrylate (EMA) varying from 100 to 40% HEMA in the monomer made by radiation initiation were used to coat glass discs. These were equilibrated with human plasma in vitro and the degree of complement activation was quantitated by C3a radio immunoassay. Significant activation was caused by copolymers made from monomers containing 60% or greater HEMA. A direct relationship between the amount of activation and the percentage of HEMA was found. The degree of activation by poly HEMA, when corrected for surface area, was quite similar to that observed for dialysis and oxygenator membranes. Similar observations were made when solid casts of crosslinked HEMA/N-vinylpyrrolidone (NVP) copolymer gels were tested, but the magnitude of activation was much greater. The results are significant because complement activation may play an important role in the response to foreign surfaces, in both extravascular and intravascular settings. A new concept of molecular biocompatibility is proposed in which surfaces eliciting molecular transformations in any of the biological defense systems are deemed nonbiocompatible. By this criterion, the hydrogel poly-HEMA, which has so frequently been thought of as biocompatible, is considered to be a molecularly non-biocompatible material.

Comparison of complement activation by silicone intraocular lenses and polymethylmethacrylate intraocular lenses with polypropylene loops

Silicone intraocular lenses are undergoing clinical investigation for use in the United States. We compared the ability of silicone intraocular lenses to activate the complement system in human sera with that of polymethylmethacrylate intraocular lenses with polypropylene loops using a radioimmunoassay that measures levels of activated complement fragments. Sera incubated with polymethylmethacrylate lenses with polypropylene loops had higher levels of C3a and C5a, but not C4a, than control sera incubated without intraocular lenses. On the other hand, there were no differences in levels of C3a, C4a, and C5a between sera incubated with silicone lenses and control sera. These results suggest that polymethylmethacrylate lenses with polypropylene loops activate the alternative pathway of complement, while silicone lenses do not.

The role of prostaglandins in C3a-mediated suppression of human in vitro polyclonal antibody responses

Suppression of polyclonal antibody responses in human peripheral blood mononuclear cell cultures by human C3a appears to involve the release of endogenous prostaglandins from monocytes. C3a was found, under the experimental conditions employed, to activate the cyclooxygenase pathway of arachidonic acid metabolism with the release of large amounts of the prostaglandin E2 species. Suppression of the protein A-induced polyclonal antibody response by C3a is abrogated by the prostaglandin synthesis inhibitor indomethacin. In addition, physiologic amounts of exogenous PGE2 were able to inhibit polyclonal antibody secretion in a manner similar to the suppression observed when C3a was added to culture. These results suggest that C3a-induced release of prostaglandins could be a major element in immunosuppression induced by C3a.

Deposits of immunoglobulins, complement, and immune complexes in inflamed human gingiva

Gingival biopsy specimens from 20 patients with moderate to advanced periodontitis were obtained from inflamed sites with pockets of 5 mm or more. Sections were studied by an immunofluorescence technique, using polyclonal rabbit or goat anti-IgG, anti-IgM, anti-C1q, anti-C3a, and anti-C3c and mouse monoclonal anti-C9. Prewashed ethanol-fixed and nonwashed ethanol-fixed or frozen specimens showed many plasma cells staining for IgG or C3a, suggesting the possible occurrence of a receptor for C3a in plasma cells. Plasma cells containing IgM were also seen. Deposits of IgG and IgM with C1q, C3a, and C3c, suggesting immune complexes, were demonstrated by a double staining technique, combining fluorescein (FITC) or rhodamine (TRITC)-labeled anti-immunoglobulins with TRITC- or FITC-conjugated antibody to C3a, C3c, and C1q. The complexes were located mainly within or around vessel walls. Deposits of C3a and C1q were found in vessel walls, in the basement membrane zone of oral gingival epithelium, or diffusely distributed in the tissues. Deposits of C3c were found to a lesser extent and only in vessel walls. Mouse monoclonal anti-C9, visualized with FITC-labeled rabbit anti-mouse and swine anti-rabbit antiserum, showed granular deposits of C9, mainly in the basement membrane zone of oral gingival epithelium. The study indicates the involvement of immune complex vasculitis in inflammatory periodontal lesions. Also, our observations of the occurrence of deposits of complement factors support the hypothesis that complement factors play an important role in the immunopathology of the periodontal lesion.