Rapid activation of the complement system by cuprophane depends on complement component C4

Abstract

Rapid activation of the complement system by cuprophane depends on complement component C4. Hemodialysis with cuprophane dialyzer membranes promotes rapid activation of the complement system, which is thought to be mediated by the alternative pathway. Complete hereditary deficiency of complement C4, a classical pathway component, in two hemodialysis patients provided the opportunity to investigate a possible role of the classical pathway. In two hemodialysis patients with both C4 isotypes, C4A and C4B, and in one patient with C4B deficiency complement activation occurred immediately after the onset of hemodialysis, with peak levels of C3a and terminal complement complex (TCC) after ten to fifteen minutes. In patients with complete C4 deficiency, C3a and TCC remained unchanged for fifteen minutes and increased thereafter, reaching the highest level after thirty minutes. The leukocyte nadir was also delayed from fifteen to thirty minutes. In vitro incubation of normal, C4A- or C4B-deficient serum with cuprophane caused complement activation after fifteen minutes. In contrast, no activation was observed in sera of four C4-deficient patients. The addition of normal serum or purified human C4 restored the capacity for rapid complement activation. In one patient with severe immunoglobulin deficiency, C3a and TCC levels increased only moderately after 25 minutes of cuprophane dialysis. This patient's serum also exhibited delayed complement activation in vitro, which was normalized after pretreatment of cuprophane with immunoglobulins. Preincubation of normal serum with MgEGTA, a blocker of the classical pathway, inhibited rapid complement activation through cuprophane. As basal levels of C4a are markedly increased in hemodialysis patients (3450 ± 850 ng/ml) compared to healthy controls (224 ± 81 ng/ml), no further elevation of C4a was detectable during cuprophane hemodialysis. Incubation of normal serum with cuprophane, however, caused a slight increase in C4a after five minutes. These results indicate that the initial deposition of complement C3b on the cuprophane membrane, necessary for activation of the amplification loop of the alternative pathway, is mediated by the classical pathway C3-convertase C4b2a. We propose an extended concept of complement activation through cuprophane, which is based on four steps: (a) binding of anti-polysaccharide antibodies, (b) classical pathway activation, (c) alternative pathway activation and (d) terminal pathway activation. Rapid activation of the complement system by cuprophane depends on complement component C4. Hemodialysis with cuprophane dialyzer membranes promotes rapid activation of the complement system, which is thought to be mediated by the alternative pathway. Complete hereditary deficiency of complement C4, a classical pathway component, in two hemodialysis patients provided the opportunity to investigate a possible role of the classical pathway. In two hemodialysis patients with both C4 isotypes, C4A and C4B, and in one patient with C4B deficiency complement activation occurred immediately after the onset of hemodialysis, with peak levels of C3a and terminal complement complex (TCC) after ten to fifteen minutes. In patients with complete C4 deficiency, C3a and TCC remained unchanged for fifteen minutes and increased thereafter, reaching the highest level after thirty minutes. The leukocyte nadir was also delayed from fifteen to thirty minutes. In vitro incubation of normal, C4A- or C4B-deficient serum with cuprophane caused complement activation after fifteen minutes. In contrast, no activation was observed in sera of four C4-deficient patients. The addition of normal serum or purified human C4 restored the capacity for rapid complement activation. In one patient with severe immunoglobulin deficiency, C3a and TCC levels increased only moderately after 25 minutes of cuprophane dialysis. This patient's serum also exhibited delayed complement activation in vitro, which was normalized after pretreatment of cuprophane with immunoglobulins. Preincubation of normal serum with MgEGTA, a blocker of the classical pathway, inhibited rapid complement activation through cuprophane. As basal levels of C4a are markedly increased in hemodialysis patients (3450 ± 850 ng/ml) compared to healthy controls (224 ± 81 ng/ml), no further elevation of C4a was detectable during cuprophane hemodialysis. Incubation of normal serum with cuprophane, however, caused a slight increase in C4a after five minutes. These results indicate that the initial deposition of complement C3b on the cuprophane membrane, necessary for activation of the amplification loop of the alternative pathway, is mediated by the classical pathway C3-convertase C4b2a. We propose an extended concept of complement activation through cuprophane, which is based on four steps: (a) binding of anti-polysaccharide antibodies, (b) classical pathway activation, (c) alternative pathway activation and (d) terminal pathway activation. anaphylatoxins complement component that is a 200 kDa glycoprotein complement C4 isotypes ethylenediaminetetraacetate interleukin-1β meagnesium ethylene glycol tetra-acetic acid terminal complement complex tumor necrosis factor alpha Hemodialysis with cellulose membranes leads to vigorous activation of the complement system with release of the anaphylatoxins C3a and C5a and formation of the terminal complement complex (TCC)1.Chenoweth D.E. Cheung A.K. Henderson L.W. Anaphylatoxin formation during hemodialysis: Effects of different dialyzer membranes.Kidney Int. 1983; 24: 764-769Abstract Full Text PDF PubMed Scopus (266) Google Scholar,2.Deppisch R. Schmitt V. Bommer J. HÄNSCH G.M. Ritz E. Rauterberg E.W. Fluid phase generation of terminal complement complex as a novel index of biocompatibility.Kidney Int. 1990; 37: 696-706Abstract Full Text PDF PubMed Scopus (117) Google Scholar. Complement activation commences immediately after the onset of treatment1.Chenoweth D.E. Cheung A.K. Henderson L.W. Anaphylatoxin formation during hemodialysis: Effects of different dialyzer membranes.Kidney Int. 1983; 24: 764-769Abstract Full Text PDF PubMed Scopus (266) Google Scholar,2.Deppisch R. Schmitt V. Bommer J. HÄNSCH G.M. Ritz E. Rauterberg E.W. Fluid phase generation of terminal complement complex as a novel index of biocompatibility.Kidney Int. 1990; 37: 696-706Abstract Full Text PDF PubMed Scopus (117) Google Scholar. Both acute and chronic sequelae of hemodialysis have been ascribed to complement activation3.Hakim R.M. Clinical implication of hemodialysis membrane biocompatibility.Kidney Int. 1993; 44: 484-494Abstract Full Text PDF PubMed Scopus (284) Google Scholar. Acute effects include dialysis neutropenia and the first-use syndrome with chest and back pain and shortness of breath4.Kaplow L.S. Goffinet J.A. Profound neutropenia during the early phase of hemodialysis.JAMA. 1968; 203: 1135-1137Crossref PubMed Scopus (271) Google Scholar, 5.Craddock P.R. Fehr J. Brigham K.L. Kronenberg R.S. Complement and leukocyte-mediated pulmonary dysfunction in hemodialysis.N Engl J Med. 1977; 296: 769-774Crossref PubMed Scopus (746) Google Scholar, 6.Hakim R.M. Breillatt J. Lazarus J.M. Port F.K. Complement activation and hypersensitivity reactions to dialysis membranes.N Engl J Med. 1984; 311: 878-882Crossref PubMed Scopus (292) Google Scholar. Anaphylatoxins C3a and C5a activate neutrophils and monocytes with the release of proteases, reactive oxygen species and arachidonic acid metabolites, and increase transcription of interleukin (IL)-1β and tumor necrosis factor-alpha (TNF-α)7.Himmelfarb J. Lazarus J.M. Hakim R.M. Reactive oxygen species production by monocytes and polymorphonuclear leukocytes during dialysis.Am J Kidney Dis. 1991; 3: 271-276Abstract Full Text PDF Scopus (173) Google Scholar, 8.Strasser T. Schiffl H. Generation of leukotriene B4 by hemodialyzer membranes: A novel index of biocompatibility.Klin Wochenschr. 1991; 69: 808-812Crossref Scopus (16) Google Scholar, 9.Luger A. Kovarik J. Stummvoll H.K. Urbanska A. Luger T.A. Blood-membrane interaction in hemodialysis leads to increased cytokine production.Kidney Int. 1987; 32: 84-88Abstract Full Text PDF PubMed Scopus (172) Google Scholar. Complement activation may also play a role in increased production of β2-microglobulin and amyloid bone disease10.Jahn B. Betz M. Deppisch R. Jannsen O. HÄNSCH G.M. Ritz E. Stimulation of β2-microglobulin synthesis in lymphocytes after exposure to cuprophan dialyzer membranes.Kidney Int. 1991; 40: 285-290Abstract Full Text PDF PubMed Scopus (43) Google Scholar,11.Van Ypersele De Strihou C. Jadoul M. Malghem J. Maldague B. Jamart J. Effect of dialysis membrane and patient's age on signs of dialysis-related amyloidosis.Kidney Int. 1991; 39: 1012-1019Abstract Full Text PDF PubMed Scopus (311) Google Scholar, and the high incidence of infections and malnutrition12.Hornberger J.C. Chernew M. Petzersen J. Garber A.M. A multivariate analysis of mortality and hospital admissions with high flux dialysis.J Am Soc Nephrol. 1993; 3: 1227-1237Google Scholar,13.Hakim R.M. Levin N. Malnutrition in hemodialysis patients.Am J Kidney Dis. 1993; 21: 125-137Abstract Full Text PDF PubMed Scopus (295) Google Scholar. The complement system can be either activated by the classical pathway, the lectin pathway or the alternative pathway. According to current understanding, complement activation by hemodialysis membranes is mediated exclusively by the alternative pathway. Complement C4, a 200 kDa glycoprotein, is a major component of the classical pathway of complement. The two C4 genes on chromosome 6 encode for two isoforms C4A and C4B, which differ in their physicochemical properties14.Law S.K.A. Dodds A.W. Porter R.R. A comparison of the properties of two classes, C4A and C4B, of the human complement component C4.EMBO J. 1984; 3: 1819-1823Crossref PubMed Scopus (224) Google Scholar. Two of our hemodialysis patients suffer from hereditary complete C4 deficiency, an extremely rare condition15.Lhotta K. KÖNIG P. Hintner H. Spielberger M. Dittrich P. Renal disease in a patient with hereditary complete deficiency of the fourth component of complement.Nephron. 1990; 56: 206-211Crossref PubMed Scopus (16) Google Scholar, 16.Hauptmann G. Tappeiner G. Schifferli J.A. Inherited deficiency of the fourth component of complement.Immunodeficiency Rev. 1988; 1: 3-22PubMed Google Scholar, 17.Lhotta K. Thoenes W. Glatzl J. Hintner H. Kronenberg F. Joannidis M. KÖNIG P. Hereditary complete deficiency of the fourth component of complement: Effects on the kidney.Clin Nephrol. 1993; 39: 117-124PubMed Google Scholar. These two patients enabled us to investigate a possible role of the classical pathway for complement activation during hemodialysis with cuprophane membranes in vivo. As one of our hemodialysis patients had severe immunoglobulin deficiency, it was also possible to study the influence of antibodies. In vitro studies of complement activation by cuprophane using sera from two additional patients with complete C4 deficiency17.Lhotta K. Thoenes W. Glatzl J. Hintner H. Kronenberg F. Joannidis M. KÖNIG P. Hereditary complete deficiency of the fourth component of complement: Effects on the kidney.Clin Nephrol. 1993; 39: 117-124PubMed Google Scholar and of individuals with deficiency of C4 isotypes C4A and C4B were performed. Levels of C4a, which is released upon activation of C4, have not been assessed in detail in patients undergoing hemodialysis. We therefore measured C4a during cuprophane hemodialysis and in the in vitro system. Our investigations indicate that immediate and rapid complement activation during hemodialysis depends on the presence of immunoglobulins, C4 and the classical pathway of complement. Patient 1, now a 34-year-old man, developed severe Henoch Schönlein purpura at the age of seventeen. Seven years later hemodialysis had to be started. He lost a renal allograft due to recurrence of disease and chronic rejection15.Lhotta K. KÖNIG P. Hintner H. Spielberger M. Dittrich P. Renal disease in a patient with hereditary complete deficiency of the fourth component of complement.Nephron. 1990; 56: 206-211Crossref PubMed Scopus (16) Google Scholar. Patients 2, 3 and 4 are siblings. They suffer from a lupus-like illness with early onset and mesangial and endocapillary proliferative glomerulonephritis. Patient 2, a 24-year-old woman, recently developed end-stage renal failure. Patient 3, a 15-year-old girl, has stable renal function at a serum creatinine of 200 μmol/liter. Patient 4, an 18-year-old boy, has a well-functioning renal allograft17.Lhotta K. Thoenes W. Glatzl J. Hintner H. Kronenberg F. Joannidis M. KÖNIG P. Hereditary complete deficiency of the fourth component of complement: Effects on the kidney.Clin Nephrol. 1993; 39: 117-124PubMed Google Scholar. All four patients are homozygous for the HLA haplotype HLA A30 B18 C4AQ0 C4BQ0 DR7. Individual 5 is healthy. The other two patients, 6 and 7, suffer from insulin-dependent diabetes mellitus. None of our dialysis patients has C4A deficiency. Patient 8, who is also included in the hemodialysis studies, suffers from reflux nephropathy. Patients 9 and 10 have Wegener's granulomatosis that is in remission. . Patient 11 suffered from chronic lymphatic leukemia with severe secondary immunoglobulin deficiency (IgG 78 mg/dl, IgA < 10 mg/dl, IgM < 7 mg/dl). The cause of renal failure was unknown. Patient 12 with end-stage renal disease of unknown origin and patient 13 with arteriosclerotic nephropathy both had a normal C4A3B1 phenotype, and were included as controls in the hemodialysis experiments. Complement C4 was phenotyped by agarose gel electrophoresis and immunoprecipitation18.Sim E. Cross S.J. Phenotyping of human complement component C4, a class-III HLA antigen.Biochem J. 1986; 239: 763-767Crossref PubMed Scopus (168) Google Scholar. For quantification of C3a and C4a the following radioimmunoassays were used: C3a des Arg 125I RIA RPA 518, and C4a des Arg 125I RIA RPA 519 (Amersham, UK). The terminal complement complex (TCC) was determined by ELISA19.WÜRZNER R. Schulze M. Happe L. Franzke A. Bieber F.A. Oppermann M. GÖTZE O. Inhibition of terminal complement complex formation and cell lysis by monoclonal antibodies.Complement Inflamm. 1991; 8: 328-340PubMed Google Scholar. In patients with the complete C4 deficiency and in patient 11 with immunoglobulin deficiency C3, C4 and factor B were measured by standard nephelometry. Factor D and factor H were determined by ELISA20.Oppermann M. Baumgarten H. Brandt E. Gottsleben W. Kurts C. GÖTZE O. Quantitation of components of the alternative pathway of complement (APC) by enzyme-linked immunosorbent assays.J Immunol Methods. 1990; 133: 181-190Crossref PubMed Scopus (40) Google Scholar. Routine hemolytic assays were used for determination of CH50 and alternative pathway AP-CH50. Hemodialysis was performed with a standard cuprophane membrane (Alwall GFE 15; GAMBRO, Hechingen, Germany). All patients received a standard heparin regimen consisting of a 2500 IU bolus and 1000 IU per hour continuously during treatment. Blood flow was kept at 250 ml/min. For determination of complement activation products ethylenediaminetetraacetate (EDTA) blood samples were taken before the onset of treatment and from the venous outlet of the dialyzer at 5, 10, 15, 30 and 60 minutes after blood had entered the dialyzer, and was handled as described (immediate cooling to 4°C and centrifugation at 2000 g for 15 min, storage at -70°C)21.Cheung A.K. Complement activation as index of hemodialysis membrane biocompatibility: The choice of methods and assays.Nephrol Dial Transplant. 1994; 9: 96-103PubMed Google Scholar. In the C4-deficient patients and the immunoglobulin-deficient patient additional samples were drawn at 20 and 25 minutes. Blood for leukocyte counts was taken at 0 and 15 minutes, when the leukocyte nadir is usually observed, and after 30 and 45 minutes. For determination of C4a cuprophane hemodialysis was performed in four patients having both C4 isotypes. Two of them suffered from arteriosclerotic nephropathy, one from diabetic nephropathy and one from analgesic nephropathy. Samples were taken before and 1, 3, 5 and 10 minutes after the start of hemodialysis. All in vitro experiments were carried out in glass tubes at 37°C. In addition to the serum of the above-mentioned patients, sera from four healthy individuals I to IV having both C4 isotypes were used. All blood samples were immediately centrifuged at 4°C and 2000 g and sera stored at -70°C. Hollow fibers of the same dialyzer as used for hemodialysis studies were cut in 1-mm-long pieces. After taking a serum sample at 0 minutes, 20 mg of these pieces were incubated with 1 ml of serum. This amount ensured that enough serum could be retrieved from the glass tube for testing. After fifteen and then thirty minutes of incubation, 300 μl of serum were taken and immediately stored at -70°C. Normal serum without the cuprophane pieces was used as a negative control. The same experiments were made with C4-deficient sera reconstituted with human C4. Fifty microliters of purified C4 (Sigma Chemical Co., St. Louis, MO, USA) were added to 950 μl of serum from C4-deficient patients, resulting in a C4 concentration of 50 μg/ml. In addition, 800 μl of serum from C4-deficient patients were mixed with 200 μl of serum from a healthy individual with C4A3B1 phenotype, yielding a C4 concentration of 48 μg/ml. C3a and TCC were determined by the tests described above. To test the influence of immunoglobulins, cuprophane fibers were preincubated for ten minutes with a polyvalent immunoglobulin preparation (Pentaglobin; Biotest, Dreieich, Germany), diluted 1:5 in phosphate buffered saline (PBS). This solution contained IgG 760 mg/dl, IgA 120 mg/dl and IgM 120 mg/dl. The fibers were then centrifuged at 3000 g for five minutes, and all excess antibodies removed and washed three times for thirty minutes each in PBS. The pretreated cuprophane fibers were then incubated with the serum of the immunoglobulin-deficient patient 11 as described. To measure the effect of magnesium ethylene glycol tetra-acetic acid (MgEGTA), an inhibitor of classical pathway activation, 8.5 ml blood of the four healthy controls were drawn directly into a syringe containing 0.75 ml of 100 mM EGTA and 0.75 ml of 50 mM MgCl2. These samples were then handled and incubated with cuprophane as above. The same incubation procedure was performed for the determination of C4a, using sera from the four healthy controls. Samples were taken at 0, 5 and 10 minutes. C4a levels in EDTA plasma were measured in the four healthy controls, six hemodialysis patients and seven patients with chronic renal failure. Levels of C3 and Factor H were normal. Factor B was elevated in patients 1, 2 and 3. Factor D was increased in hemodialysis patients 1 and 2. AP-CH50 was normal or elevated. These results show that all C4-deficient patients had normal function of the alternative pathway of complement. No C4 could be detected. The classical pathway CH50 was absent or very low. Hemodialysis patient 11 with immunoglobulin deficiency had normal C3, C4 Factor B and H levels, but factor D was increased; CH50 and AP-CH50 were normal. Detailed concentrations are shown in Table 1.Table 1Complement parameters in patients with complete C4 deficiency and immunoglobulin deficiency After 15 minutes of hemodialysis with a cuprophane dialyzer the total leukocyte counts fell to 24% and 21%, respectively, of the predialysis count in the patients 12 and 13 with normal C4 phenotype, and to 22% in the C4B-deficient patient 8. After 30 minutes the leukocyte counts had again increased. In patients 1 and 2 with complete C4 deficiency leukocytes were unchanged at 15 minutes, fell to 19% and 29% after 30 minutes of dialysis and increased again after 45 minutes Figure 1. In the patients with normal C4 phenotype and the one with C4B deficiency C3a levels were elevated after five minutes, reached their maximum at 10 minutes and declined slowly thereafter. The C4-deficient patients showed a different pattern. C3a was unchanged and very low for up to 15 minutes and reached its peak value at 30 minutes Figure 2a. In the patients with normal C4 phenotype or C4B deficiency the terminal complement complex increased from 5 to 10 minutes onwards, reaching a plateau at 10 to 15 minutes. In C4-deficient patients the TCC remained unchanged until 15 to 20 minutes, increasing thereafter to plateau levels at 30 minutes Figure 2b. These results show that complement activation in vivo and the consecutive leukocyte drop are delayed in patients with complete C4 deficiency. In the patient with immunoglobulin deficiency C3a and TCC remained unchanged for 25 minutes and increased only moderately thereafter Figure 2. Due to chronic lymphatic leukemia the patient had a total leukocyte count of 70 G/liter. The number of neutrophils was normal (4 G/liter) and remained unchanged for 45 minutes Figure 1. These results indicate a profound defect of complement activation by cuprophane in the immunoglobulin-deficient patient. The C4a plasma levels of hemodialysis patients are described below. In vitro experiments were performed to confirm the dialysis data and to investigate the influence of reconstitution of C4-deficient sera with C4. Detailed concentrations are listed in Tables 2 and 3. As serum capable of complement activation had to be used in these experiments, levels of C3a at 0 minutes were already elevated. Therefore, a significant complement activation in the negative control without cuprophane stimulation was also observed. Nevertheless, after an incubation time of 15 minutes C3a and TCC were highly increased in serum from probands with a normal C4 phenotype and C4A or C4B deficiency Table 2. C4A-deficient patients had a somewhat lower increase in C3a compared to C4B-deficient probands (P < 0.05, unpaired t-test). In C4-deficient serum only a marginal increase in C3a and TCC concentrations had occurred after 15 minutes, which was comparable to the negative control Table 3. Subsequently, we investigated whether the addition of purified human C4 or small amounts of serum from a non-C4-deficient individual was able to reconstitute the ability for rapid complement activation. Both the addition of C4 and serum, resulting in a C4 concentration of 50 and 48 μg/ml, respectively (normal range 200 to 500 μg/ml), caused a rise in C3a and TCC comparable to normal controls Figure 3 and Table 3. These experiments confirm the in vivo results and clearly establish that C4 deficiency is the crucial factor for delayed complement activation. The serum of the immunoglobulin-deficient patient showed an identical pattern to C4-deficient serum with only a marginal elevation of C3a and TCC after fifteen minutes. Preincubation of cuprophane fibers with a polyvalent immunoglobulin preparation normalized complement activation Figure 3 and Table 2.Table 2Results of C3a and terminal complement complex (TCC) after in vitro stimulation of normal, C4A-deficient, C4B-deficient and immunoglobulin-deficient serum with cuprophaneTable 3C3a and terminal complement complex TCC after incubation of completely C4-deficient sera with cuprophane: Effect of addition of normal serum and purified human C4Table 3C3a and terminal complement complex TCC after incubation of completely C4-deficient sera with cuprophane: Effect of addition of normal serum and purified human C4Figure 3Increase of (A) anaphylatoxin C3a and (B) terminal complement complex (TCC) from 0 to 15 minutes after incubation of serum with cuprophane. Bars show mean value and SD for each patient group. Note that the increase in C4-deficient serum (N = 4) and immunoglobulin-deficient serum is similar to the increase in the negative control without cuprophane. Addition of normal serum or purified C4 to C4-deficient serum restores rapid complement activation. Similarly, preincubation of cuprophane fibers with immunoglobulins normalizes activation in the immunoglobulin-deficient serum. The difference in C3a between C4A-deficient (N = 3) and C4B-deficient (N = 3) serum is significant (P < 0.05, unpaired t-test).View Large Image Figure ViewerDownload (PPT) MgEGTA-treated serum had a markedly reduced preactivation of complement as shown by the low basal C3a levels. MgEGTA also blocked the early increase of C3a and TCC after incubation with cuprophane Figure 4. These results further establish a role of the classical pathway in rapid complement activation by cuprophane in vitro. In four healthy controls C4a concentrations were 224 ± 81 ng/ml. Levels were 10-fold to 20-fold increased in hemodialysis patients (3450 ± 850 ng/ml) compared to controls. Intermediate C4a values were found in six patients with various degrees of renal failure Figure 5. These results show that C4a accumulates with progressive renal failure. Because of these extremely high predialysis levels, no clear changes of C4a were detectable during hemodialysis with a cuprophan membrane. In vitro incubation of four normal sera with cuprophane clearly led to an increase of C4a. Basal serum values were 603 ± 23 ng/ml. At five minutes C4 had increased by 198 ± 30 ng/ml and after 10 minutes by 256 ± 111 ng/ml. This shows that cuprophane causes moderate activation of C4, most of which occurs within the first five minutes of incubation Figure 6.Figure 6C4a serum levels upon complement activation by cuprophane in vitro. Increments of C4a after five and ten minutes of incubation are shown. Symbols are: (•) healthy individuals (N = 4, mean ±SD); (○) negative control without cuprophane. Cuprophane causes a prompt and slight increase in C4a serum levels.View Large Image Figure ViewerDownload (PPT) The current concept of complement activation by cellulosic hemodialysis membranes is based on an exclusive role of the alternative complement pathway [reviewed in22.Johnson R.J. Complement activation during extracorporeal therapy: Biochemistry, cell biology and clinical relevance.Nephrol Dial Transplant. 1994; 9: 36-45Google Scholar. According to this concept, C3(H2O)Bb, which is constantly generated in small amounts in plasma, activates C3 to C3b, leading to exposure of the internal thioester bond of C3. C3b then binds to the cellulosic membrane. Binding of Factor B and its cleavage by Factor D leads to C3bBb, the alternative pathway C3-convertase. The convertase cleaves many more C3 molecules into C3a, which is released into the fluid phase, and C3b, which binds to the membrane surface (amplification loop). In addition, the alternative pathway C5-convertase C3bC3bBb is formed and cleaves C5 into C5a and C5b, thus initiating activation of the terminal complement pathway to form the terminal complement complex (TCC)22.Johnson R.J. Complement activation during extracorporeal therapy: Biochemistry, cell biology and clinical relevance.Nephrol Dial Transplant. 1994; 9: 36-45Google Scholar. The observation, however, that C3b does not spontaneously bind to cellulosic membranes, is inconsistent with this concept and indicates that other serum factors are required for binding of C3b on the surface23.Cheung A.K. Parker C.J. Wilcox L. Janatova J. Activation of the alternative pathway of complement by cellulosic hemodialysis membranes.Kidney Int. 1989; 36: 257-265Abstract Full Text PDF PubMed Scopus (65) Google Scholar. It has been suggested that this additional component is factor B, which binds primarily to the membrane and forms a nidus for C3b24.Cheung A.K. Parker C.J. Janatova J. Brynda E. Modulation of complement activation on hemodialysis membranes by immobilized heparin.J Am Soc Nephrol. 1992; 2: 1328-1337PubMed Google Scholar. Based on our observations we propose that initial deposition of C3b on cuprophane is mediated by the classical pathway C3-convertase C4b2a. Our results of complement activation by cuprophane membranes in patients with hereditary complete C4 deficiency clearly show that the classical pathway C3-convertase C4b2a plays a crucial role in the immediate complement activation after onset of dialysis. Without a functional classical pathway C3-convertase, for example, in a patient with complete C4 deficiency, complement activation is delayed for about fifteen minutes. The extent of complement activation, however, is not influenced by a defective classical pathway. The results obtained from our C4-deficient patients during hemodialysis with a cuprophane dialyzer were confirmed by in vitro studies. Complement activation by cuprophane hollow fibers was delayed in the sera of four patients with complete C4 deficiency. Factors other than C4 may certainly have an impact on complement activation. All patients had normal concentrations of C3 and of the alternative pathway components, Factors B, D and H. Most importantly, they had an unimpaired alternative pathway hemolytic activity. The C4-deficient patients did not have signs of active immune complex disease and circulating immune complexes were undetectable. C3a levels comparable to controls exclude a preactivation of the complement system in these patients. The most convincing evidence, however, that C4 deficiency was indeed the cause of delayed complement activation is the fact that the addition of purified human C4 to C4-deficient sera restored the capability for immediate complement activation. The C4 concentrations of 50 μg/ml or 48 μg/ml in reconstituted sera are well below the normal C4 range of 200 to 500 μg/ml. Obviously, small amounts of C4 are therefore sufficient to trigger complement activation. In another report of a patient with hereditary complete C4 deficiency, delayed complement activation by zymosan, which, like cuprophane, is considered to be an alternative pathway activator, has been described. The addition of C4 to the serum of this patient also corrected complement activation25.Jackson C.G. Ochs H.D. Wedgwood R.J. Immune response of a patient with deficiency of the fourth component of complement and systemic lupus erythematosus.N Engl J Med. 1979; 300: 1124-1129Crossref PubMed Scopus (94) Google Scholar. C4 also has a role in alternative pathway activation. Nascent C3b binds with high affinity to C4b and is thereby protected from cleavage by factors H and I26.Meri S. Pangburn M.K. A mechanism of activation of the alternative complement pathway by the classical pathway: protection of C3b from inactivation by covalent attachment to C4b*.Eur J Immunol. 1990; 20: 2555-2561Crossref PubMed Scopus (44) Google Scholar. The absence of this protective effect may also contribute to the delayed alternative pathway activation in patients with complete C4 deficiency. To further substantiate a role of C4, we measured C4a, which is released upon activation of C4 by the C1 complex, in our hemodialysis patients. Surprisingly, we found C4a concentrations ten to twenty times higher than those in healthy controls. Our data of hemodialysis patients and those with various degrees of renal insufficiency show that C4a accumulates in chronic renal failure. Anaphylatoxins C5a27.Oppermann M. GÖTZE O. Plasma clearance of the human C5a anaphylatoxin by binding to leucocyte C5a receptors.Immunology. 1994; 82: 516-521PubMed Google Scholar and most likely also C3a are cleared by binding to their receptors on leukocytes and other cells. On the contrary, C4a, a 9 kDa peptide, seems to be cleared by glomerular filtration and tubular metabolization28.Gorski J.P. Quantitation of human complement fragment C4ai in physiological fluids by competitive inhibition radioimmune assay.J Immunol Methods. 1981; 47: 61-73Crossref PubMed Scopus (36) Google Scholar. Because the anaphylatoxin C3a is 100 times more potent than C4a, it is unlikely that this increase of C4a is of any clinical significance for dialysis patients. Unfortunately, the extremely high levels preclude the detection of minor C4a increments during hemodialysis. Our in vitro data, however, clearly show that C4a increases and hence C4 is rapidly activated upon incubation of serum with cuprophane. The two isotypes of human complement C4 differ in their reactivity with acceptor surfaces14.Law S.K.A. Dodds A.W. Porter R.R. A comparison of the properties of two classes, C4A and C4B, of the human complement component C4.EMBO J. 1984; 3: 1819-1823Crossref PubMed Scopus (224) Google Scholar. Both isotypes react with hydroxyl and amino groups. However, whereas C4A preferentially binds to amino groups after activation via its internal thioester, C4B reacts with hydroxyl groups29.Dodds A.W. Ren X.D. Willis A.C. Law S.K.A. The reaction mechanism of the internal thioester in the human complement component C4.Nature. 1996; 379: 177-179Crossref PubMed Scopus (161) Google Scholar. As cellulosic membranes like cuprophane contain abundant hydroxyl groups, one would expect C4B to be predominantly involved in complement activation by such membranes. We therefore included a dialysis patient with a complete deficiency of the isotype C4B in our in vivo study. The time course of complement activation in this patient was identical to that of the patient having both isotypes. In the in vitro studies C4B-deficient serum was somewhat more efficient than C4A-deficient serum. We assume that C4A may be more effective in mediating rapid complement activation by cuprophane. Cleavage of C4 and C2 to form the classical pathway C3-convertase C4b2a is brought about either by the classical pathway or lectin pathway of complement30.Law S.K.A. Reid K.B.M. Complement. (2nd ed). Edited by MALE D. Oxford University Press, Oxford1995Google Scholar. Activation of the classical pathway requires binding of C1q to the cuprophane membrane. The classical pathway is usually activated by binding of C1q through its globular non-collagenous heads to IgG or IgM. Anti-dextran antibodies of IgG isotype, which are present in normal serum, have been shown to bind to cuprophane, and their titer correlates with complement activation31.Maillet F. Kazatchkine M.D. Specific antibodies enhance alternative complement pathway activation by cuprophane.Nephrol Dial Transplant. 1991; 6: 193-197Crossref PubMed Scopus (18) Google Scholar. Classical pathway activation may be started by these anti-dextran antibodies or other natural polyreactive anti-polysaccharide antibodies of IgG or IgM isotypes. Our studies of a patient with severe secondary immunoglobulin deficiency point to a role of immunoglobulin in the initial phase of complement activation by cuprophane. Although classical and alternative pathway hemolytic activities were normal, this patient exhibited delayed and only weak complement activation during cuprophane hemodialysis. Lymphocytes do not express anaphylatoxin receptors30.Law S.K.A. Reid K.B.M. Complement. (2nd ed). Edited by MALE D. Oxford University Press, Oxford1995Google Scholar. An influence, however, of the extremely high leukocyte count on the plasma levels of C3a and TCC cannot be excluded. Using the patient's serum, complement activation in vitro was also severely impaired, but could be normalized by preincubation of cuprophane with immunoglobulins. Anti-polysaccharide antibodies present in the immunoglobulin preparation had probably bound to the cuprophane surface and activated the otherwise intact classical pathway of the patient. In addition, the somewhat greater efficacy of C4A indicates that activated C4b binds to amino groups of antibodies on the membrane rather than to hydroxyl groups of the membrane itself14.Law S.K.A. Dodds A.W. Porter R.R. A comparison of the properties of two classes, C4A and C4B, of the human complement component C4.EMBO J. 1984; 3: 1819-1823Crossref PubMed Scopus (224) Google Scholar. Clearly the role, isotype and specificity of immunoglobulins involved in complement activation by cuprophane requires further investigation. To further establish the involvement of the classical pathway in complement activation by cuprophane, in vitro experiments using sera incubated with MgEGTA, an inhibitor of the classical pathway, were performed. In these experiments it could clearly be demonstrated that blocking the classical pathway with MgEGTA results in delayed complement activation identical to C4 deficiency or immunoglobulin deficiency. In conclusion, our data imply that the rapid activation of complement by cuprophane depends on the classical pathway. Based on our results we propose the following extended concept of complement activation by cellulosic hemodialysis membranes, which is based on four steps: (a) binding of anti-polysaccharide antibodies, (b) classical pathway activation, (c) consecutive alternative pathway activation, and (d) terminal pathway activation Figure 7. Immediately after blood enters the dialyzer anti-polysaccharide antibodies bind to the membrane. Then C1q reacts with these immunoglobulins and C4 is cleaved into C4a and C4b by C1s in the C1 complex. C4b probably binds covalently to amino groups of the antibodies. After binding and cleavage of C2 the classical pathway C3-convertase C4b2a is formed. The classical pathway C3-convertase activates C3 by proteolytic cleavage into C3a and C3b, which is deposited on the membrane. In addition, C4b protects C3b from cleavage by Factors H and I. Binding and activation of Factor B creates the alternative pathway C3-convertase C3bBb, which serves as the nidus for explosive complement activation by the amplification loop of the alternative pathway. Cleavage of C5 and terminal pathway activation is mediated both by the classical pathway C5-convertase C4b2a3b and the alternative pathway C5-convertase C3bBb3b. In patients with complete C4 deficiency without a functioning classical pathway, the initial C3b deposition depends on spontaneous binding of C3b created by C3(H2O)Bb in the fluid phase. Because this is a rather inefficient and slow process, complement activation by cuprophane is delayed in these patients. We are indebted to Friedrich Neumair and Gertraud Beck, who helped in the hemodialysis studies of C4-deficient patients, and to Anna Schlögl and Andrea Eder for technical assistance.