A C3 convertase assay for nephritic factor functional activity

Journal of Immunological Methods 251 (2001) 45–52 www.elsevier.nl / locate / jim

A C3 convertase assay for nephritic factor functional activity ¨ c, Emiliana Jelezarova a , Markus Schlumberger a , Salima Sadallah b , Peter J. Spath ¨ A. Schifferli b , Hans U. Lutz a , * Jurg a

Institute of Biochemistry, Swiss Federal Institute of Technology, ETH-Zentrum, CH-8092 Zurich, Switzerland b Department of Medicine, University Hospital Basle, Basle, Switzerland c ZLB Central Laboratory Blood Transfusion Service, SRC, (now ZLB Bioplasma AF), Bern, Switzerland Received 27 July 2000; received in revised form 21 November 2000; accepted 11 December 2000

Abstract C3 nephritic factor (C3NeF) is an autoantibody against the C3 convertase which stabilizes this otherwise inherently labile neoenzyme and induces a continuous activation of the alternative pathway with C3 depletion. NeF is found in patients with membranoproliferative glomerulonephritis and / or partial lipodystrpohy. NeF activity is usually detected in plasma by hemolytic tests. In order to obtain reproducible data for the functional activity of purified C3NeF IgG a solid phase assay was developed. C3 convertase was generated on immobilized C3b by incubation with factors B and D in the presence of Ni 21 . Convertase sites were left to decay in the presence of normal IgG or NeF IgG. Residual convertase activity was measured by adding 125 I-C3 and capturing nascent 125 I-C3b on the plate surface via covalently coupled NH 2 -Glu-Tyr dipeptide. In the presence of factor H during C3 convertase decay, a dose dependent stabilizing activity was shown for NeF IgG including NeF IgG purified from urine. A second format of the assay was developed in which C3 convertase was assembled on C3b 2 –IgG complexes in the presence of Mg 21 . Since these complexes are more efficient as convertase precursors the signal was five-fold higher than with C3b. Convertase decay, on the other hand, was not influenced by the nature of the precursor and in both systems the stabilizing activity of NeF IgG was similar.  2001 Elsevier Science B.V. All rights reserved. Keywords: C3 nephritic factor; C3 convertase; Complement; C3b 2 –IgG

1. Introduction Abbreviations: C3b 2 –IgG, covalent complex consisting of two C3b molecules bound to one heavy chain of IgG; C3bBb, alternative pathway C3 convertase; C3bBb(Mg 21 )P, C3 convertase, generated with (Mg 21 ) and properdin; C3NeF, C3 nephritic factor; C4bC2a, classical C3 convertase; EY, dipeptide Glu-Tyr; HC, IgG heavy chain; IgG–C3b 2 Bb, C3 convertase, generated on C3b 2 –IgG; MPGN, membranoproliferative glomerulonephritis; NeF, nephritic factor; SLE, systemic lupus erythematosus *Corresponding author. Tel.: 141-1-632-3009; fax: 141-1-6321269. E-mail address: [email protected] (H.U. Lutz).

Membranoproliferative glomerulonephritis (MPGN) is a renal disorder characterized by dense deposits in the mesangial area and capillary wall of glomeruli. Between 10% and 30% of the individuals presenting with MPGN have in their plasma an autoantibody against the C3 convertase termed nephritic factor, NeF (West, 1994; West and McAdams, 1998). Many other conditions, apart from MPGN, can be associ-

0022-1759 / 01 / $ – see front matter  2001 Elsevier Science B.V. All rights reserved. PII: S0022-1759( 01 )00295-2

46

E. Jelezarova et al. / Journal of Immunological Methods 251 (2001) 45 – 52

ated with the presence of NeF and in all cases hypocomplementemia is a feature. These include poststreptococcal glomerulonephritis (FremeauxBacchi et al., 1994), systemic lupus erythematosus (SLE) (Walport and Davies, 1996), urticarial vasculitis (Carmichael and Marsden, 1993) and partial lipodystrophy (Wayte et al., 1996). NeF binds and stabilizes the alternative pathway convertase, C3bBb (C3NeF) or the classical pathway convertase, C4bC2a (C4NeF) (Daha et al., 1976; Ohi and Yasugi, 1994). Most C3NeF antibodies are of the IgG class (Daha et al., 1978). NeF functional activity depends on antibody specificity, since the ability to stabilize red cell bound C3bBb was confined to F(ab9) 2 and Fab9 fragments (Scott et al., 1978). It is still unclear which portion of the C3bBb neoenzyme is recognized by C3NeF. C3NeF might bind to Bb or to neoepitopes on C3bBb. Properdin, the only known positive physiological regulator of complement, also stabilizes the alternative pathway convertase by decreasing the rate of its intrinsic decay. Functional studies of C3NeF and properdin have shown that both prolong the half-life of red cell bound C3bBb sites several-fold, but the difference between them appeared to reside in the relative resistance of NeF-stabilized convertase to the action of factor H (Weiler et al., 1976; Whaley and Ruddy, 1976). The nature of such resistance is not fully understood. The mechanism of action of C3NeF is of potential interest not only with regard to clinical practice, but also in understanding the regulation of the complement system in general. More reproducible measurements are required to avoid a possible binding of NeF IgG to red cells (Marin et al., 1991), to clarify the existence of NeF neutralizing antibodies in certain patients (Fremeaux-Bacchi et al., 1994) and to differentiate between NeF and similar antibodies in patients and apparently healthy persons (Gewurz et al., 1983; Spitzer et al., 1990). The major difficulty arising from the inherent instability of C3bBb(Mg 21 ) was overcome by using Ni 21 for C3bBb generation (Fishelson and Muller-Eberhard, 1982) or by generating alternative pathway convertase on C3b 2 –IgG complexes (Jelezarova et al., 2000). Different acceptors were tested for covalent capturing of nascent 125 I-C3b to the plate.

2. Materials and methods

2.1. Materials As whole human IgG we used Sandoglobulin (ZLB Central Laboratory, Blood Transfusion Service, Bern, Switzerland). Sephadex, sodium 125 Iiodide and Protein G Sepharose Fast Flow were obtained from Amersham Pharmacia Biotech AB (Uppsala, Sweden); factor B, factor H and properdin were either from Calbiochem-Novabiochem (San Diego, CA) or from Advanced Research Technologies (San Diego, CA). Factor D was purified from the peritoneal fluid of patients with end stage renal failure (Catana and Schifferli, 1991). Complement C3 was isolated from fresh plasma (Hammer et al., 1981; Lutz et al., 1996) and was deprived of inactivated C3 (iC3) before use (Janatova et al., 1980). Microtitre plates were purchased from Costar (Corning, NY) and modified to contain aldehyde groups (‘Chemobond plates’, manufactured by Dr ¨ Ernst Fischer Laboratories, Dubendorf, Switzerland). Peptide Glu-Tyr was obtained from Bachem (Bubendorf, Switzerland).

2.2. Patients Patient A presented with partial lipodystrophy, meningitis and nephritis (Schifferli and Blanc, 1986). Patient B was a 22-year-old male Caucasian, had proteinuria (nephrotic range 3–5 g / 24 h) but a normal serum albumin level and no edema. Renal biopsy showed dense deposit disease.

2.3. Protein purification C3 was labeled with Na 125 I by the Chloramine T procedure as previously described (Lutz et al., 1993) and had a specific activity of 2.5 to 7.5310 6 cpm / mg. C3b 2 –IgG complexes were generated from C3b and biotinylated IgG and subsequently purified (Jelezarova et al., 2000). The pooled human IgG (Sandoglobulin  ) will be referred to as ‘normal’ IgG. As NeF IgG we used the IgG fraction of serum or urine from patients with MPGN. IgG was purified on Protein G Sepharose according to the manufacturer’s instructions. Briefly,

E. Jelezarova et al. / Journal of Immunological Methods 251 (2001) 45 – 52

serum was diluted 5-fold in 20 mM phosphate buffer, pH 7.0 and loaded onto the column (1.635 cm). After washing, bound IgG was eluted with 0.1 M glycine buffer, pH 2.5. Fractions were immediately neutralized with 1 M Tris–HCl, pH 9, and dialyzed against PBS, pH 7.4. Urine, collected in 0.04% sodium azide, was dialyzed against distilled water to remove urea and salts. Insoluble material was removed by centrifugation for 10 min at 13 0003g. The pH was adjusted to 7.4 and solid ammonium sulfate was slowly added at room temperature to achieve 50% saturation. The solution was further stirred for 1 h at room temperature and the formed precipitate collected by centrifugation for 30 min at 13 0003g. The pellet was dissolved in PBS and dialyzed against 20 mM phosphate buffer, pH 7.0. Purification on Protein G Sepharose was carried out as for serum IgG. All IgG preparations were treated with 1.1 mM diisopropylfluorophosphate and stored in aliquots at 2708C. Protein concentrations were determined spectrophotometrically using A 1% 280, 1 cm values of 9.7 for C3 and C3b, and 14.1 for IgG.

2.4. Protein coupling to Chemobond  plates The coupling procedure includes incubation of the protein in the presence of NaBH 3 CN and blocking of the remaining aldehydes with NaBH 4 (Lutz et al., 1990). In the modified version, the primary protein, C3b, was first coupled at 20 mg / ml for 6 h at room temperature. In a second coupling step, pooled human IgG (0.5 mg / ml) or 10 mM peptide Glu-Tyr (EY) were incubated overnight at RT. The plates were then blocked with NaBH 4 as usual and stored at 48C. For C3b 2 –IgG plates, neutro-avidin was coupled at 20 mg / ml as the primary protein, followed by peptide Glu-Tyr in the second coupling step. Prior to use, biotinylated C3b 2 –IgG was adsorbed on the plate at 5 mg / well (9.65 nM) in VBS, pH 7.0, overnight at RT.

2.5. Convertase activity assay Free binding sites on the plates were blocked either with 20 mg / ml of human IgG (C3b regular

47

coupling) or with 0.5% BSA (C3b / IgG and C3b / EY plates) in VBS, pH 7.2. C3 convertase generation was carried out in VBS, containing 0.1% gelatin, pH 7.2 (GVBS), with 5 mM MgCl 2 or 1 mM NiCl 2 for 15 min at 378C in a final volume of 50 ml / well (Jelezarova et al., 2000). Unless otherwise stated, factor B was used at 50 mg / ml, factor D at 2 mg / ml and properdin, where indicated, was added at final concentrations ranging from 1 to 40 mg / ml. The plates were washed three times with GVBS (150 ml / well). Decay occurred in GVBS in the presence of the indicated proteins and the plates were washed once with GVBS. The enzyme activity of the remaining convertase sites was determined with 1 mg / well of 125 I-labeled C3 for 15 min at room temperature unless otherwise indicated. The plates were washed three times with GVBS containing 0.05% Tween (GVBS-T) and bound radioactivity was determined. Control wells were treated in exactly the same way, except that convertase generation was omitted. Deposition of 125 I-labeled C3 in those wells was subtracted from the corresponding test data. Each experimental point was based on determinations in quadruplicate. In the IgG concentration dependence experiment (Fig. 4), 1 mg / ml of properdin was used during 125 I-C3 cleavage in order to enhance the signal.

3. Results

3.1. C3 convertase assay on immobilized C3 b Alternative pathway C3 convertase was assembled by addition of factors B and D to C3b that was covalently bound to the surface of the plate. After washing, enzyme activity was determined by its capacity to cleave 125 I-C3 followed by the covalent binding of nascent 125 I-C3b to target molecules on the plate. It was therefore essential to provide sufficient acceptor sites on the plate surface. Human IgG (Gadd and Reid, 1981) and tyrosine (Sahu and Pangburn, 1995) are among the best acceptors of nascent C3b. Three different possibilities for capturing nascent 125 I-C3b on the plate were tested. Noncovalently adsorbed IgG gave satis-

48

E. Jelezarova et al. / Journal of Immunological Methods 251 (2001) 45 – 52

sessed and showed that at C3 concentrations up to 1 mg / ml virtually all C3b was bound to the plate (not shown). C3 convertase activity depended on the properdin concentration at a fixed factor B and factor D input and reached approximately 6 ng / well of 125 I-C3b at 40 mg / ml of properdin (not shown). For the subsequent experiments, C3 convertase activity was determined on C3b / EY plates.

3.2. C3 convertase decay

Fig. 1. Effect of C3b acceptor on the detection of captured 125 I-C3b. Regular C3b plates containing covalently bound C3b were blocked with human IgG. C3b / IgG plates and C3b / EY plates had covalently bound C3b and IgG or C3b and EY and were blocked with BSA. C3 convertase was generated from 20 mg / ml of factor B, 2 mg / ml of factor D and 5 mM MgCl 2 in the absence (hatched bars) or in the presence of 5 mg / ml of properdin (filled bars). Convertase activity was determined with 0.4 mg / ml of 125 I-C3. Data is expressed as mean6S.D. from quadruplicates.

factory results with significant binding of nascent 125 I-C3b (Fig. 1: regular C3b plate). The addition of properdin provided only a small increase in 125 I-C3b fixation. Apparently, the high concentration of adsorbed IgG not only trapped nascent 125 I-C3b, but also had a protective effect. The standard deviations were rather high, possibly because some of the IgG was lost during the washing steps before measuring the remaining radioactivity. The capturing of nascent 125 I-C3b was substantially improved in the case of the immobilized peptide (Fig. 1, C3b / EY). The difference between values with and without properdin during convertase generation was bigger and the quadruplicates showed little variation, indicating that the assay was very stable. The plates with covalently bound IgG had a low capturing capacity (Fig. 1, C3b / IgG). The C3b-capturing capacity of the C3b / EY plate was linear up to 1 mg / ml of 125 I-C3 with a tendency to saturate around 7–8 ng / well of 125 I-C3b at 50 to 300 mg / ml (not shown). The occurrence of unbound 125 I-a9-C3b in the fluid phase was separately as-

C3 convertase was generated with Mg 21 or Ni 21 and the effects of properdin and / or factor H were studied. Convertase activity generated with Mg 21 in the absence of properdin was hardly detectable and that generated with Ni 21 showed moderate initial activity (Fig. 2). The addition of 1 mg / ml of properdin during convertase generation significantly increased the number of stable convertase sites and slowed their decay as expected. C3 convertases assembled with Ni 21 were more stable than those assembled with Mg 21 . Factor H produced a rapid decay of all types of convertase and the differences between Mg 21 and Ni 21 became less obvious, although the estimated half-life of C3bBb(Ni 21 )P was approximately 10 min as compared to 6 min for C3bBb(Mg 21 )P. For further experiments the convertase was therefore generated with Ni 21 .

3.3. Effect of NeF and normal IgG on the kinetics of C3 convertase decay in the presence of factor H In the C3bBb(Ni 21 ) system residual C3 convertase activity was studied after decay in the presence of 0.5 mg / ml of factor H and 0.2 mg / ml of normal or NeF IgG. The kinetics of C3 convertase decay in the presence of normal IgG and factor H was virtually indistinguishable from that of factor H alone (Fig. 3). Under these conditions, the estimated half-life of the C3 convertase was approximately 14 min. In the presence of NeF IgG and factor H the remaining convertase sites were significantly stabilized, since after 1 h of incubation 45% of the initial enzyme activity was measured compared to 14% for normal IgG.

E. Jelezarova et al. / Journal of Immunological Methods 251 (2001) 45 – 52

49

Fig. 2. C3 convertase decay as a function of the bivalent cation and the addition of properdin. C3 convertase was generated on a C3b / EY plate in the absence (circles) or in the presence (squares) of 1 mg / ml of properdin. Convertase sites were left to decay either in GVBS (empty symbols) or in the presence of 4 mg / ml of factor H (filled symbols). A mean6S.D. is shown.

3.4. Functional activity of various NeF IgG Our next goal was to test NeF IgG from different patients as well as to compare the properties of NeF IgG isolated from the serum and urine of the same patient. We determined C3 convertase activity remaining after 1 h of decay as a function of IgG concentration. Normal IgG showed no stabilizing effect in the presence of 2 mg / ml of factor H (Fig. 4). NeF IgG purified from the sera of two patients with high NeF titres in the hemolytic assay produced a concentration dependent increase in residual convertase activity. NeF IgG purified from the urine of patient B showed functional activity, which comprised 73–78% of that of serum IgG. The extent of

convertase stabilization correlated with the NeF IgG binding to C3bBb in the ELISA (not shown).

3.5. C3 convertase generated on C3 b2 –IgG complexes C3 convertase was assembled on C3b 2 -IgG complexes in the presence of Mg 21 . The enzyme generated without properdin still had detectable activity after a 10-min decay period (Fig. 5). The value obtained for initial activity was twice as high as that with C3bBb(Ni 21 ) (Fig. 2). The half-life, however, was shorter, which was consistent with the characteristics of the Mg 21 system. In the presence of 1 mg / ml of properdin, initial enzyme activity was five

50

E. Jelezarova et al. / Journal of Immunological Methods 251 (2001) 45 – 52

Fig. 3. Kinetics of C3 convertase decay in the presence of factor H and NeF IgG. C3b / EY plates were prepared as described. C3 convertase was generated in the presence of 1 mM NiCl 2 . Convertase sites were incubated with 0.2 mg / ml of NeF IgG (squares), normal IgG (circles) or GVBS (triangles) in the presence of 0.5 mg / ml of factor H. The residual C3 convertase activity is expressed as a percentage of the initial value (0.2160.08 ng C3b / well). The graph presents results from two independent experiments and for each a mean6S.D. is shown.

Fig. 4. Functional activity of various NeF IgG preparations. C3 convertase was generated on a C3b / EY plate with 1 mM NiCl 2 . Convertase sites were left to decay for 1 h in the presence of 2 mg / ml of factor H and increasing concentrations of normal IgG or NeF IgG. The assignment of the symbols is as follows: d, normal IgG; m, NeF IgG from the serum of patient A; j, NeF IgG from the serum of patient B; and h, NeF IgG from the urine of patient B. Results are shown as mean6S.D. Initial convertase activity was 0.50 ng C3b / well.

times higher than the activity of the convertase generated with C3b (Figs. 5 and 2). At the same time, decay was neither facilitated nor delayed as compared to C3bBb(Mg 21 ). Activity dropped to 50% by 60 min of decay in GVBS on both C3b (Fig. 2) and C3b 2 –IgG (Fig. 5). In the presence of 4 mg / ml of factor H the half-lives of the C3 convertases were shortened as expected and were similar. C3 convertase generated on C3b 2 –IgG complexes with Mg 21 showed a similar pattern of decay with a half-life of approximately 10 min in the presence of IgG and factor H or factor H alone (Fig. 6). When NeF IgG was added together with factor H, C3 convertase sites were stabilized and after 1 h of decay, enzyme activity was still 30% of the initial value. The faster decay of the convertase generated with C3b 2 –IgG is compatible with the use of Mg 21 . The difference between NeF and normal IgG was similar after 1 h of decay in the C3b and the C3b 2 –IgG system. Thus, the assay performed with C3b 2 –IgG complexes as convertase precursors confirmed the essential findings from the C3b system

Fig. 5. C3 convertase generated on C3b 2 –IgG complexes. C3 convertase was generated on a C3b 2 -IgG / EY plate with 5 mM MgCl 2 in the absence (circles) or in the presence (squares) of 1 mg / ml of properdin. Convertase sites were left to decay either in GVBS (empty symbols) or in the presence of 4 mg / ml of factor H (filled symbols). Data is expressed as mean6S.D.

E. Jelezarova et al. / Journal of Immunological Methods 251 (2001) 45 – 52

Fig. 6. Decay of C3 convertase assembled on C3b 2 –IgG complexes. C3b 2 –IgG / EY plates were prepared as described. C3 convertase was generated in the presence of 5 mM MgCl 2 . Convertase sites were incubated with 0.4 mg / ml of NeF IgG (squares), normal IgG (circles) or GVBS (triangles) in the presence of 0.5 mg / ml of factor H. The residual C3 convertase activity is expressed as a percentage of the initial value (1.32460.046 ng C3b / well). The graph shows results from two independent experiments.

and can be used as a more elaborate tool for studying convertase decay under physiological conditions.

4. Discussion Up to now there has been no commercially available assay for NeF functional activity. All available assays for the detection of NeF feature high variability in the results as they rely on assembly of the convertase on the red cell surface. Our goal was to develop a reliable assay which would permit a study of the functional properties of purified NeF IgG and compare them with its antigenic characteristics as determined using the available ELISA assay (Seino et al., 1993). NeF functional assays have been carried out earlier in the fluid phase and enzymatic activity subsequently assayed by determining the remaining C3 convertase sites (Daha et al., 1984). In the assay presented here, C3 convertase was assembled on immobilized precursors and its activity was directly measured on the plate. Enzymatic activity was

51

confined to the solid phase by capturing all nascent C3b on the free tyrosine OH-group of the co-immobilized peptide (Fig. 1). This approach significantly improved 125 I-C3b binding to the plate as compared with immobilized or adsorbed IgG (Fig. 1). A major problem when studying complement convertases is their intrinsic instability, resulting in low residual activity. Ni 21 is known to prolong the half-life of the C3 convertase about 6-fold (Fishelson and Muller-Eberhard, 1982) and it can be used to generate a C3 convertase stable enough to allow satisfactory measurements. On the other hand, C3b 2 – IgG complexes are far more efficient precursors for C3 convertase generation even in the presence of Mg 21 (Jelezarova et al., 2000). Therefore we analyzed C3 convertase generation on two different precursors, C3b and C3b 2 –IgG, and compared their properties. The initial activity of IgG– C3b 2 Bb(Mg 21 ) was twice as high as that of C3bBb(Ni 21 ) (Figs. 2 and 5). Properdin, added during convertase generation, had a much stronger effect on the IgG–C3b 2 Bb system probably due to bivalent binding to C3b 2 –IgG (Jelezarova et al., 2000). The kinetics of C3 convertase decay was not influenced by the nature of the precursor-C3b or C3b 2 –IgG. The decay pattern of C3bBb in Mg 21 was, however, clearly distinct from that in Ni 21 (Fig. 2), and IgG–C3b 2 Bb(Mg 21 ) decayed with similar kinetics to C3bBb(Mg 21 ) (Figs. 5 and 2). Both assay systems can be used to study the functional activity of NeF IgG preparations (Figs. 3 and 6). While the generation of C3bBb(Ni 21 ) is straightforward, the presence of Ni 21 may add artificial stabilization during subsequent decay. Convertase assembly on C3b 2 –IgG requires their purification, but permits the analysis of NeF activity under strictly physiological conditions. The presence of factor H allowed normal and NeF IgG to be distinguished after only 10 min of decay (Fig. 4). In the absence of factor H, differentiation between NeF and normal IgG was possible at incubation times longer than 2 h (not shown). The NeF IgG effect was highly reproducible and dose-dependent (Fig. 4). Furthermore, we could show for the first time that IgG isolated from the patient’s urine retained the convertase stabilizing properties of serum NeF IgG. Evi-

52

E. Jelezarova et al. / Journal of Immunological Methods 251 (2001) 45 – 52

dently, this novel assay is well suited to determine NeF functional activity in IgG preparations.

Acknowledgements This work was supported by a Tandem grant 3238-050664 of The Swiss National Science Foundation.

References Carmichael, A.J., Marsden, J.R., 1993. Urticarial vasculitis: a presentation of C3 nephritic factor [letter]. Br. J. Dermatol. 128, 589. Catana, E., Schifferli, J.A., 1991. Purification of human complement factor D from the peritoneal fluid of patients on chronic ambulatory peritoneal dialysis. J. Immunol. Methods 138, 265–271. Daha, M.R., Austen, K.F., Fearon, D.T., 1978. Heterogeneity, polypeptide chain composition and antigenic reactivity of C3 nephritic factor. J. Immunol. 120, 1389–1394. Daha, M.R., Deelder, A.M., Van Es, L.A., 1984. Stabilization of the amplification convertase of complement by monoclonal antibodies directed against human factor B. J. Immunol. 132, 2538–2542. Daha, M.R., Fearon, D.T., Austen, K.F., 1976. C3 nephritic factor (C3NeF): stabilization of fluid phase and cell-bound alternative pathway convertase. J. Immunol. 116, 1–7. Fishelson, Z., Muller-Eberhard, H.J., 1982. C3 convertase of human complement: enhanced formation and stability of the enzyme generated with nickel instead of magnesium. J. Immunol. 129, 2603–2607. Fremeaux-Bacchi, V., Weiss, L., Demouchy, C., May, A., Palomera, S., Kazatchkine, M.D., 1994. Hypocomplementaemia of poststreptococcal acute glomerulonephritis is associated with C3 nephritic factor (C3NeF) IgG autoantibody activity [see comments]. Nephrol. Dial. Transplant 9, 1747–1750. Gadd, K.J., Reid, K.B., 1981. The binding of complement component C3 to antibody-antigen aggregates after activation of the alternative pathway in human serum. Biochem. J. 195, 471–480. Gewurz, A.T., Imherr, S.M., Strauss, S., Gewurz, H., Mold, C., 1983. C3 nephritic factor and hypocomplementaemia in a clinically healthy individual. Clin. Exp. Immunol. 54, 253– 258. Hammer, C.H., Wirtz, G.H., Renfer, L., Gresham, H.D., Tack, B.F., 1981. Large scale isolation of functionally active components of the human complement system. J. Biol. Chem. 256, 3995–4006. Janatova, J., Lorenz, P.E., Schechter, A.N., Prahl, J.W., Tack, B.F., 1980. Third component of human complement: appearance of a sulfhydryl group following chemical or enzymatic inactivation. Biochemistry 19, 4471–4478.

Jelezarova, E., Vogt, A., Lutz, H.U., 2000. Interaction of C3b2– IgG complexes with complement proteins properdin, factor B and factor H: implications for amplification. Biochem. J. 349, 217–223. Lutz, H.U., Nater, M., Stammler, P., 1993. Naturally occurring anti-band 3 antibodies have a unique affinity for C3. Immunology 80, 191–196. Lutz, H.U., Stammler, P., Fischer, E.A., 1990. Covalent binding of detergent-solubilized membrane glycoproteins to ‘Chemobond’ plates for ELISA. J. Immunol. Methods 129, 211–220. ¨ Lutz, H.U., Stammler, P., Jelezarova, E., Nater, M., Spath, P.J., 1996. High doses of immunoglobulin G attenuate immune aggregate-mediated complement activation by enhancing physiologic cleavage of C3b in C3bn–IgG complexes. Blood 88, 184–193. Marin, M.A., Fontan, G., Lopez-Trascasa, M., 1991. Interaction of C3 nephritic factor (NEF) with erythrocyte membranes complement-independent binding to sheep and patients’ erythrocytes. Mol. Immunol. 28, 133–140. Ohi, H., Yasugi, T., 1994. Occurrence of C3 nephritic factor and C4 nephritic factor in membranoproliferative glomerulonephritis (MPGN). Clin. Exp. Immunol. 95, 316–321. Sahu, A., Pangburn, M.K., 1995. Tyrosine is a potential site for covalent attachment of activated complement component C3. Mol. Immunol. 32, 711–716. Schifferli, J.A., Blanc, E., 1986. Partial lipodystrophy, meningococcal meningitis and nephritis. Dermatologica 173, 9–12. Scott, D.M., Amos, N., Sissons, J.G., Lachmann, P.J., Peters, D.K., 1978. The immunogloblin nature of nephritic factor (NeF). Clin. Exp. Immunol. 32, 12–24. Seino, J., v.d. Wall Bake, W.L., van Es, L.A., Daha, M.R., 1993. A novel ELISA assay for the detection of C3 nephritic factor. J. Immunol. Methods 159, 221–227. Spitzer, R.E., Stitzel, A.E., Tsokos, G.C., 1990. Evidence that production of autoantibody to the alternative pathway C3 convertase is a normal physiologic event. J. Pediatr. 116, S103–S108. Walport, M.J., Davies, K.A., 1996. Complement and immune complexes. Res. Immunol. 147, 103–109. Wayte, J., Bird, G., Wilkinson, J.D., 1996. The clinical significance of partial lipoatrophy and C3 hypocomplementaemia: a report of two cases. Clin. Exp. Dermatol. 21, 131–134. Weiler, J.M., Daha, M.R., Austen, K.F., Fearon, D.T., 1976. Control of the amplification convertase of complement by the plasma protein beta 1 H. Proc. Natl. Acad. Sci. USA 73, 3268–3272. West, C.D., 1994. Nephritic factors predispose to chronic glomerulonephritis [see comments]. Am. J. Kidney Dis. 24, 956–963. West, C.D., McAdams, A.J., 1998. Membranoproliferative glomerulonephritis type III: association of glomerular deposits with circulating nephritic factor-stabilized convertase. Am. J. Kidney Dis. 32, 56–63. Whaley, K., Ruddy, S., 1976. Modulation of the alternative complement pathways by beta 1 H globulin. J. Exp. Med. 144, 1147–1163.