Characterization of physiologic breakdown products of the complement fragment Ba

Molecular Immunology,Vol. 31, No. 4, pp. 307-314,

1994 0 1994 Elsevier Science Ltd Copyright Printed in Great Britain. All rights reserved 0161-5890194 $6.00 + 0.00

0161-5890(93)EOO20-M

CHARACTERIZATION OF THE

OF PHYSIOLOGIC BREAKDOWN COMPLEMENT FRAGMENT Ba

PRODUCTS

MARTIN OPPERMANN* and OTTO G~TZE Department

of Immunology,

University of Gottingen,

D-37075 Giittingen,

Germany

(First received 24 September 1993; accepted in revised form 12 October 1993) Abstract-To better characterize the activation products of factor B which are generated under physiologic conditions Ba was purified directly from human EDTA-plasma by immunoaffinity chromatography using anti-Ba Sepharose. SDS-PAGE analysis revealed the existence of degradation products of the Ba fragment which were truncated at the carboxyterminus. A monoclonal antibody (mAb D22/3) was produced by immunizing mice with a synthetic peptide which corresponds to the Ba carboxyterminus (Glu,,,-Arg,,,). This mAb was found to react with an epitope (Ba neo-epitope), which is newly formed after the generation of Ba from its precursor protein factor B. This neoantigenic determinant is absent both in factor B and the desArg/Lys Ba derivatives. The conversion of Ba by carboxypeptidases in human serum was monitored using an assay which is based on mAb D22/3, revealing a half-life of Ba in serum of 150 min. Furthermore, this assay allowed to quantitate plasma levels of intact and degraded Ba in healthy probands and in patients with chronic renal failure. The processing of the Ba carboxyterminus may be of functional relevance as the biological activity of the Ba fragment which had been shown to suppress human B lymphocyte functions in vitro resides in its carboxyterminal amino acid sequence. Key words: complement

fragment Ba, carboxypeptidase,

peptide-specific antibody.

obtained that Ba is degraded at its carboxyterminus once it is formed by proteolytic activation of B. A monoclonal antibody with specificity for a neoantigenic determinant on the intact carboxyterminus of Ba allowed to directly monitor the conversion of Ba into its lower mol. wt degradation products.

INTRODUCTION

Factor B (B) is a central component of the alternative pathway of complement (G&e, 1988). The reversible Mg’+-dependent interaction of the zymogen B with C3b or C3(H20) enables factor D (D) to cleave a single peptide bond between ArgTj4 and Lys,,, of B, thereby generating the fragments Bb and Ba. Ba is a glycoprotein of approximately 34 kDa which consists of 234 amino acids. Several reports have shown that the Ba fragment suppresses human B-lymphocyte functions in vitro (Zierz and Giitze, 1987; Ambrus et al., 1990). The finding of

MATERIALS

Buffers and protease

AND METHODS

inhibitors

containing VB-Mg’+: 9mM Verona1 buffer 2mM Mg’+. PBS: 9 mM phosphate buffer, pH 7.4, containing 140 mM NaCl. PBS-Tween: PBS containing 0.05% (v/v) Tween 20 (Serva, Heidelberg, Germany); PBS-Tween-EDTA: PBS-Tween containing 20 mM EDTA (Serva, Heidelberg, Germany). Coating buffer: 50 mM carbonate, pH 10.6. Blocking buffer: coating buffer containing 1% (w/v) gelatin (Difco, Hamburg, Germany). Substrate buffer: 100 mM sodium-acetate, 50 mM sodium-phosphate, pH 4.2. The protease inhibitors aprotinin, PMSF and E-64 were purchased from Boehringer Mannheim, Germany, p-APMSF and MERGETPA were from Calbiochem, Frankfurt, Germany.

highly elevated plasma levels of Ba in patients with chronic renal failure has led us to consider that Ba may contribute to the defective immune response in these patients (Oppermann et al., 1991). It was the aim of this study to characterize the physiologic Ba fragments which are generated under in vivo conditions. Evidence was *Author to whom correspondence should be addressed at: Howard Hughes Medical Institute, Duke University Medical Center, P.O. Box 3821, Durham, NC 27710, U.S.A. Abbreviations: B, factor B of the alternative complement pathway; CFA, complete Freund’s adjuvant; CRF, chronic renal failure; D, factor D of the alternative complement pathway; EDTA ethylenediaminetetraacetic acid; ESRD, end-stage renal disease; IFA, incomplete Freund’s adjuvant; KLH, keyhole limpet hemocyanin; mAb, monoclonal antibody; MERGETPA, DL-mercaptomethyl-3-guanidino-ethylthio-propanoic acid; p-APMSF, (p-amidinophenyl)methanesulfonyl fluoride; SMPB, succinimidyl 4(p-maleimidophenyl) butyrate.

Peptide

synthesis

and coupling

to carrier proteins

The C-terminal peptide Ba-20C (amino acid single letter code CETIEGVDAEDGHGPGEQQKR) was synthesized according to the published amino acid sequence of Ba (Mole et al., 1984). The extra cysteine was added to the aminoterminus of the peptide in order to 307

308

M. OPPERMANNand 0. G~TZE

facilitate its conjugation to BSA or KLH. The peptide was synthesized using Fmoc-protected and PyBOP-activated amino acids (Coste et al., 1990) and an automatic peptide-synthesizer (Milligen 9050, Millipore, Eschborn, Germany). After cleavage from the resin and removal of protecting groups the peptide was purified by reversedphase HPLC using a Delta Pak Cl&column (Millipore, Eschborn, Germany) and an elution gradient from 0% to 50% acetonitrile in 0.1% trifluoroacetic acid. The fractions corresponding to the major peak were pooled and evaporated to dryness. Judging from HPLC chromatograms the purity of the peptide was >95%. The correct mass was confirmed by plasma desorption mass spectrometry. The peptide was coupled to the carrier proteins through the amino-terminal cysteine residue using the bifunctional reagent SMPB (Green et al., 1982; Liu et al., 1979). Five mg of BSA (Miles, Kanakakee, IL) in 0.5 ml 50mM phosphate, pH 7.5, was reacted (30min/20”C) with 0.5 mg SMPB (Pierce, Hamburg, Germany) which was dissolved in 25 ~1 dimethyl formamide. After the removal of unreacted SMPB by gel filtration with Sephadex G-25 the fractions containing BSA-SMPB derivatives were pooled and reacted with 5 mg of the peptide (3 hr/20”C). The coupling efficiency was determined by measuring the decrease of free cysteine by Ellman’s method (Ellman, 1959). Unspecific binding sites on the carrier protein were blocked by the incubation with a 50-fold molar excess of cysteine. The Ba-20C/KLH conjugate was generated by using KLH (Sigma, Deisenhofen, Germany) instead of BSA as the carrier protein. The Ba-20C peptide was biotinylated in 0.1 M sodium carbonate buffer, pH 8.0, through the aminoterminal cysteine residue by adding N-iodoacetyl-N-biotinylhexylenediamine (Pierce, Hamburg, Germany) to the peptide at a 5-fold molar excess. After incubation (1 hr/37”C) the remaining free biotin-residues were blocked by the addition of a lOO-fold molar excess of cysteine for an additional 30 min. Monoclonal

antibodies

The peptide-specific mAb D22/3 was obtained following the i.p. immunization of BALB/c-mice with 150 pg Ba-20C/BSA conjugate emulsified in CFA (Difco, Detroit, MI). Booster injections of 75 pg conjugate in IFA (Difco, Detroit, MI) were given at 4, 8 and 12 weeks. At days 3-l before the preparation of splenic lymphocytes mice received 100 ,ug carrier-bound peptide in PBS. The fusion was performed according to standard techniques using the myeloma cell line X63-Ag8.653 (Peters and Baumgarten, 1992). Culture supernatants were screened by a solid-phase ELISA employing the Ba-20C peptide (500 ng/ml) or peptide-KLH conjugates (5 pg/ml) as the antigen in coating buffer adsorbed to polystyrene microtiter plates (Immunoplate IIF, Nunc, Wiesbaden, Germany). Hybridoma supernatants which contained peptide-specific antibodies were detected by rabbit antimouse Ig peroxidase conjugates (Dako, Hamburg, Germany). Two out of 18 peptide-specific mAb were identified which reacted with the Bau fragment, as well.

The mAb D22/3 was further characterized in detail. The antibodies P21/15 (IgG2a/rc) and M20/6 (IgGl/K) had previously been obtained following the immunization of mice with factor B (Baumgarten et al., 1985). The mAb P21/15 and M20/6 react with an epitope which is expressed on all tested Ba-fragments and on the precursor protein factor B, as well. The mAb M20/6 was biotinylated using biotin-E -aminocaproic acid N-hydroxysuccinimide ester (Calbiochem, Frankfurt, Germany) as described (Oppermann et al., 1990). Isolation of in vitro generated

Ba (Ba34)

The complement components B, C3 and D were purified from human serum according to standard protocols (Gotze and Miiller-Eberhard, 1971; Tack and Prahl, 1976; Lesavre et al., 1979). The Ba fragment was prepared by the cleavage (30 min/37”C) of B (20 mg) by D (40 pg) in the presence of C3 (20 mg) in 5 ml VBMg2+. A 34 kDa Ba fragment was obtained following anion-exchange chromatography on QAE-Sepharose fast flow (Pharmacia, Freiburg, Germany) at pH 7.8 and gel filtration on Sephadex G-75 superfine (Pharmacia, Freiburg, Germany). The Ba3, fragment was incubated at 600 pg/ml with porcine carboxypeptidase B (Boehringer, Mannheim, Germany) (final concentration 60 pg/ml) for 2 hr at 37°C to convert Ba into its desArg/Lys form. Isolation of Ba from zymosan activated human serum and from

uremic plasma

A 30 kDa degradation product of Ba (Ba,,) was purified from human serum which had been activated with zymosan in vitro. After the incubation of human serum with zymosan (Serva, Heidelberg, Germany) (5 mg/ml; 48 hr/37”C) factor B and other serum proteins were precipitated by 12% PEG 6000 (Perrin et al., 1975). To prevent the further degradation of Ba the following chromatographic steps were done in the presence of 1 mM PMSF and 5 mM EDTA at 4°C. The sample was dialyzed against 50 mM Tris-HCl, pH 7.9, containing 0.3 M NaCl and was applied to an immunoaffinity column (anti-Ba IgG M20/6-Sepharose). After washing with the same buffer the column was eluted with 50 mM citrate, pH 4.2. Fractions containing Ba were identified by the ELISA which detects all Ba forms and were dialyzed against PBS. Physiologic breakdown products of the Ba fragment were isolated by PEG precipitation of 50 ml EDTA plasma from a single uremic blood donor and immunoaffinity chromatography using an anti-Ba immunoaffinity column as described for the isolation of Ba from zymosan activated serum. It had previously been shown that Ba plasma levels are highly elevated in patients with end-stage renal disease as compared to healthy blood donors (Oppermann et al., 1991). Trace amounts of factor B which were still detectable after immunoaffinity chromatography, were removed by anion-exchange chromatography on Mono Q (Pharmacia, Freiburg, Germany) at pH 7.8. The Ba containing

309

Degradation of complement fragment Ba fractions were identified by ELISA, pooled and dialyzed against PBS. Endog~~cos~dase F digestion The 34 kDa and the 30 kDa Ba fragments were deglycosylated by endoglycosidase F (Boehringer, Mannheim, Ge~any) in the presence of 0.1% SDS and 1% 2-mercaptoethanol for 18 hr at 37°C (Elder and Alexander, 1982). CNBr cleavage The 34 kDa and the 30 kDa Ba fragments were treated with a lOOO-fold molar excess of CNBr (Serva, Heidelberg, Germany) in 70% formic acid. The tube was flushed with nitrogen and incubated in the dark for 18 hr. The reaction was quenched by diluting formic acid with a ten-fold excess of Hz0 and lyophilization using a Speed-Vat centrifuge (Bachofer, Reutlingen, Germany). SDS-PAGE

and immunoblotting

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis was performed according to Laemmli (Laemmli, 1970). Proteins were stained with Coomassie Brilliant Blue R250. Molecular weights were determined by comparison with a commercially available calibration kit (Pharmacia, Freiburg, Germany). Immunoblotting was performed as described (Hempelmann and Oppermann, 1992).

Plasma samples for the quantitation of Ba Human EDTA plasma samples were taken from healthy donors and from patients with chronic renal failure (CRF) or with end-stage renal disease undergoing inte~ittent hemodialysis (ESRD). Clinically relevant and laboratory parameters of these probands have been described in detail ~Oppe~ann et al., 1991). RESULTS Isolation and characterization of Ba from plasma and from zymosan activated serum To characterize the Ba fragments which are generated under in viva conditions Ba was isolated from uremic plasma by immunoabsorption with M20/6 IgG anti-Basepharose. On SDS-PAGE proteins with apparent molecular weights of 33.7 kDa and 30 kDa were identified (Fig. 1, lane 1). Both polypeptides reacted with the anti-Ba mAb M20/6 (not shown) and were heterogeneous, as they consisted of at least two separate bands. The 33.7 kDa protein migrated slightly faster than Ba which had been obtained by the factor D induced MW

WI 67

-

ELISA procedures An ELISA that allows to specifically quantitate the 43 34 kDa form of the Ba-fragment (Ba,,) was established. The Ba,,-neoepitope specific mAb D22/3 (40 pg/well) was adsorbed into wells of microtiter plates (Immunoplate IJF; Nunc, Wiesbaden, Germany) in coating buffer for 16 hr at 4°C. After blocking with 1% gelatin 100 ~1 samples were applied per well. Purified Ba,, or plasma 30 samples diluted in PBS-Tween-EDTA were allowed to bind for 2 hr and were detected by adding, in sequence, the biotinylated Ba-specific mAb M20/6 (2 pug/ml) and a 500-fold dilution of streptavidin-peroxidase (Amersham, Braunschweig, Germany) for 1 hr. Following extensive washing with PBS-Tween color was developed by the 20.1 addition of 2 mM 2,2-azino-di-(3-ethyl-benzthiazoline sulfonate) (Boehringer Mannheim, Germany) in substrate buffer in the presence of 2.5 mM H,Oz. Absorbances were recorded with a microplate photometer (Dynatech MR 600, Dynatech, Denkendorf, 14.4 Germany) at 410 nm (reference wavelength 490 nm). An ELISA that detects both the 34 kDa and the 30 kDa Ba Fig. 1. SDS-PAGE analysis of Ba (2 gg protein per lane) under reducing conditions. A 10% po~yacrylamide gel was used. Lane fragments has been described in detail (Oppe~ann I: Ba isolated by ~mmunoabsorption from human plasma. et aZ., 1990). This assay which is based on two mAb Lane 2: Ba, which was obtained by the factor D induced (P21/15 and M20/6) detects all forms of Ba. Factor B cleavage from factor B. Lane 3: Ba, which had been treated which is recognized by these antibodies, as well, is with carboxypeptidase B. Lane 4: CNBr cleavage product of removed from the sample by anti-B/Bb mAb coupled to Ba which had been purified from human plasma. Lane 5: CNBr paramagnetic particles. cleavage product of Ba,,.

310

M. OFPERMANNand 0. G~TZE

MW

123412

1234

WI 93 6'7

A

B

C

Fig, 2. (A) SDS-PAGE analysis of Ba (5 pg protein per lane) under reducing conditions. A 10% polyacrylamide gel was used. Lane 1: Ba,, which was obtained by the factor D induced cleavage from factor B. Lane 2: Ba3, which was isolated from zymosan activated serum. Lane 3: Ba,, after deglycosylation with endoglycosidase F. Lane 4: Ba,, after deglycosylation with endoglycosidase F. (B) Immunoblot showing the specificity of the mAb D22/3. Proteins (same order as in A; 250 ng per lane) were detected by the mAb D22/3 and visualized by a rabbit anti-mouse Ig peroxidase conjugate and 4-chloro-l-naphthol as peroxidase substrate. (C) Immunoblot showing the specificity of the mAb P21/15. Lane 1: 250ng Ba,,. Lane 2: 250ng Ba,,. activation of factor B in vitro (Ba,,) and had the same molecular weight as carboxypeptidase B treated Ba,, (Fig. 1, lanes 2 and 3). Whereas the 30 kDa Ba fragments were only minor constituents in plasma, proteins with identical molecular weights were purified with high yield from serum which had been activated in vitro with zymosan for prolonged periods of time. To determine whether differences in the asparagine-linked oligosaccharide moieties accounted for the different molecular weights of the 34 kDa and the 30 kDa forms of Ba both fragments were deglycosylated with endoglycosidase F. The molecular weights of the deglycosylated proteins were approximately 4 kDa lower than those of the respective glycosylated Ba forms (Fig. 2A). The peptide-specific mAb D22/3 reacted with Ba,, in its glycosylated as well as in its deglycosylated form (Fig. 2B). However, it did not recognize Ba, which had been purified from zymosan activated serum. The mAb P21/15 reacted with an epitope which is expressed on both Ba forms (Fig. 2C). The 234 amino acid sequence of Ba includes a single CNBr cleavage site at methionine 198. The CNBr cleavage of a sample containing the heterogeneous Ba forms which had been isolated from uremic plasma resulted in a single 29 kDa protein as analyzed by SDS-PACE. This protein was apparently very similar to the CNBr cleavage product of in vitro generated Baj4 (Fig. 1, lanes 4 and 5). This result indicates that the 30 kDa fragments of Ba proteolytic

in plasma are generated by the cleavage of sites carboxyterminal from methionine 198. Specljicify of antipeptide antibodies One hybridoma clone was identified which produced a mono~lonal antibody (D22/3) with specificity for the synthetic Ba-20C peptide and the physiologic 34 kDa Ba fragment. The mAb D22/3 was typed as IgGZb/K and further characterized by a competitive ELISA. The binding of biotin-labelled Ba-20C peptide to the mAb D22/3 in the presence of several competitors was assessed by ELISA (Fig. 3). Besides the Ba-20C-peptide only Ba,, and serum that had been treated with zymosan (30min/37”C), but not Ba, nor factor B in EDTAplasma prevented the binding of biotinylated Ba-20C to the mAb. The Ba-20C/KLH conjugate was the most efficient competitor in this assay on a molar basis probably because it contained several Ba-20C peptides crosslinked to the carrier protein. It was concluded that the mAb D22/3 recognizes a neoantigenic determinant that is present on the 34 kDa Ba fragment but absent from Ba,, and native factor B in plasma. Degradation of Bas4 in zymosan-activated and by car~oxy~e~~~~se B

human serum

The incubation of human serum with zymosan in vitro rapidly leads to proteoiytic cleavage of factor I3 and the generation of Ba which can be detected by ELISA

Degradation

of complement

fragment

311

Ba

1.0

E

C 2 *

0.0

’

0.6 0.4 0.2 0 2000

1000

500

250

125

1:2

1:4

1:e

1:16

1:32

63 1:64

31

16

nM

1:126 1:256 Dilution

Fig. 3. Specificity of the anti-Ba mAb D22/3 as demonstrated by a competitive ELISA. The binding of biotinylated Ba-20C peptide to the mAb D22/3 was assayed in the presence of various competitors: (A) EDTA plasma, (0) zymosan activated serum, (0) Ba,,, (0) Ba,,, (V) Ba-20C peptide, (+) Ba-20C/KLH. Streptavidin peroxidase and 2,2’-azino-di-(3-ethyl benzthiazolinesulfonate)/H,O, were used for the detection of bound Ba-20C biotin.

(Fig. 4). When the ELISA is applied which detects all Ba fragments maximum values were reached after 1 hr. Whereas the total Ba content in the sample remained constant, declining Ba,, concentrations indicated that Ba was further processed (t,,z = 150 min) yielding degradation products which are not recognized by the mAb D22/3. To characterize those enzymes which degrade the Ba, fragment zymosan activated serum was incubated in the presence of various protease inhibitors. As shown in Table 1 only EDTA and DL-mercaptomethyl3-guanidino-ethylthio-propanoic acid (MERGETPA), both potent inhibitors of the serum carboxypeptidase (Plummer and Ryan, 198 l), efficiently antagonized the degradation of Ba,,. The sulfhydryl compound MERGETPA is an activator of the alternative complement pathway (von Zabern et al., 1987) and thereby

120

,

I

generates even higher levels of Ba,, in zymosan activated serum. To test the hypothesis that a carboxypeptidase participates in the degradation of Ba, purified Ba3, was incubated with porcine carboxypeptidase B (Fig. 5). As determined by ELISA the antigenic epitope on Ba,, which is recognized by the mAb D22/3 is rapidly (t,,z = 5 min) destroyed by carboxypeptidase B. Quantitation of total Ba and of Ba,, in EDTA patients with renal failure

The plasma levels of total Ba and of Ba,, in patients with chronic renal failure or with end-stage renal disease were quantitated by specific ELISA procedures (Table 2). The total Ba levels in these patients had already been communicated in a previous report (Oppermann et al., 1991). The Ba,, concentrations in the plasma

Table

1. The effect of protease inhibitors of Ba,, in zymosan activated

Protease

I

1I 20 II

00

0

12

3

Time

4

5

6

7

0

Ihours]

Fig. 4. Degradation of Ba by serum proteases. Serum was activated by the addition of zymosan (10 mg/ml) and was incubated at 37°C. Total Ba levels (0) and Ba,, (0) in samples which had been taken after the indicated time intervals were determined by ELISA.

plasma of

inhibitor

None Aprotinin (2 pg/ml) p-APMSF (100 PM) E-64 (10 p M) EDTA (20 mM) MERGETPA (1 mM)

Ba,, (pg/ml) 10.8 15.6 21.6 22.1 43.3 62.7

on the degradation serum” Degradation

(%)b

68 53 35 (-::, (-88)

“Normal human serum was activated with zymosan (10 mg/ml) for 30min at 37°C. After removal of zymosan from the sample by centrifugation protease inhibitors were added to individual aliquots. Samples were incubated for additional 7.5 hr and Ba,, was determined by ELISA. bThe degradation of Ba,, in complement activated serum is expressed as [(c, - cZ)/c,] * 100 where c, is the concentration of Bau before the addition of protease inhibitors, i.e. 33.3 pg/ml, and c2 the concentration of Ba,, after 7.5 hr incubation with the protease inhibitors.

M. OPPERMANNand 0. G~TZE

312

400

300 = E . Z -

200

f d 100

0 0

1

2.5

10

5

Time

20

30

60

[mid

Fig. 5. Degradation of Ba,, by carboxypeptidase B. Purified Ba31 (600 pgg/ml) was incubated in the presence of 2 mM Ca2+/Mg2+ with porcine carboxypeptidase B (1: 10 w/w) at 37°C. Proteolytic cleavage was stopped by the addition of 20mM EDTA at the indicated time intervals and Ba,, was determined by ELISA.

of patients with CRF or with ESRD were significantly (p < 0.001) elevated as compared to healthy controls. Although Ba,, levels and total Ba plasma concentrations in the patients under study were positively correlated (Y = 0.80, p < 0.01 as calculated by linear regression analysis) the Ba,, fragment contributed to the total Ba plasma content to different degrees. Whereas in controls mean Ba,, plasma concentrations equalled 77% of total Ba values, in CRF only 36% and in ESRD patients 30% of total Ba consisted of the 34 kDa Ba fragment.

DISCUSSION This study shows that under physiologic conditions, the human Ba fragment, once generated by the factor D-mediated proteolytic activation of factor B, is further processed by serum proteases yielding lower m.w. cleavage products. The physiologic Ba fragments were structurally characterized by SDS-PAGE and immunoblotting and were quantitated by ELISA procedures which allowed to differentiate between the native and the degraded Ba forms. Several proteins with distinct molecular weights were identified when Ba which had been directly isolated from uremic plasma by immunoabsorption on anti-Ba sephaTable 2. Plasma levels of Ba,, and of total Ba in healthy probands and in patients with chronic renal failure (CRF) or with end-stage renal disease (ESRD)

f%, (a/ml) Ba (a/ml)

Controls (N = 55)

CRF (N = 41)

(N =61)

ESRD

0.78 f 0.20 1.01 * 0.30

1.73 + 1.10 4.84 f 3.58

4.83 f 1.55 16.1 k 6.10

rose was analysed by SDS-PAGE. The major constituents appeared to be the desArg/Lys fragments of Ba which are generated by serum carboxypeptidases. The carboxyterminal sequence of Ba (PGEQQKR) reveals two potential cleavage sites for serum carboxypeptidase. Our results are in accord with a previous study by Davrinche et al. (1984) who showed by isoelectrofocusing that during the prolonged in zjitro activation of human serum with inulin charge modifications occur in the Ba fragment which are compatible with the action of serum carboxypeptidases. In order to describe the degradation of Ba in quantitative terms monoclonal antibodies with specificities for the Ba carboxyterminus were generated by the immunization of mice with a synthetic peptide. This technique had successfully been applied in the past to the production of monoclonal antibodies with specificities for neoantigenic determinants on other complement activation products such as human C3a(desArg) (Burger et ul., 1988) or porcine CSa/CSa(desArg) (Hopken et ul., 1992). The monoclonal antibody D22/3 recognizes an epitope on the Ba fragment which is expressed only after the proteolytic cleavage/activation of factor B. This antibody provided the basis for a specific ELISA which allows the direct quantitation of the 34 kDa Ba fragment in human plasma. More importantly, the mAb D22/3 recognized an epitope on the Ba carboxyterminus which is destroyed by carboxypeptidase treatment. The carboxyterminal arginine and/or lysine residues form essential constituents of the epitope which is recognized by this antibody. The ELISA which is based on the mAb D22/3 in combination with an earlier described assay which recognizes all Ba forms in plasma (Oppermann et al., 1990) allowed to monitor the generation of Ba in zymosan activated serum and its subsequent degradation by serum carboxypeptidases. Thus, for the first time the processing of a complement activation product by carboxypeptidases could be directly followed by a quantitative assay. Whereas the purified 34 kDa Ba fragment was rapidly (t,..?= 5 min) degraded by porcine carboxypeptidase B, Ba,, in zymosan activated serum was only slowly converted to its desArg/Lys form. The t,,? of Ba,, in human serum was determined to be approximately 2.5 hr. The rather slow degradation of Ba,, in serum is in contrast to the rapid degradation and inactivation of the anaphylatoxins C3a and C5a by serum carboxypeptidase (Bokisch and Muller-Eberhard, 1970). This may explain why arginated synthetic Ba peptides remain intact for a period sufficient to raise a specific immune response in mice. The predominant role of carboxypeptidases in the degradation of Ba in serum was shown in this study by experiments in which the processing of Ba was completely abrogated by the addition of the specific carboxypeptidase inhibitor DLmercaptomethyl-3-guanidino-ethylthio-propanoic acid (MERGETPA) to zymosan activated serum whereas other protease inhibitors showed no effect. By applying an assay which detects all Ba forms we have previously noticed highly elevated Ba plasma levels in patients with chronic renal failure and with end-stage

Degradation

of complement

renal disease undergoing intermittent hemodialysis (Oppermann er al., 1991). This finding was attributed to a combined effect of the direct renal retention of Ba, which is filtered at the glomerulus and subsequently reabsorbed and catabolized by the proximal tubular epithelium (Maack et al., 1979), and by the enhanced activation of the alternative pathway of complement in these patients due to an impaired elimination of the protease factor D (Volanakis et al., 1985). The additional quantitation of the short-lived Ba,, fragment in the patients’ samples allowed to estimate the relative impact of these different mechanisms. If one assumes that Ba,, plasma levels are predominantly controlled by serum carboxypeptidases which convert this fragment into its desArg/Lys forms the five-fold elevation of the Ba3, fragment in ESRD patients as compared to healthy controls reflects the increased generation of Ba during enhanced complement turnover. However, as the total Ba plasma levels in these patients are elevated 16-fold as compared to controls the impaired renal elimination of Ba appears to be an even more important parameter which determines Ba levels in renal failure. Immunoabsorption of uremic plasma on anti-Ba Sepharose revealed the presence of additional 30 kDa degradation products of Ba besides the 34 kDa proteins. These lower molecular weight fragments were also obtained with high yield from serum after a prolonged incubation (48 hr) with zymosan. The finding in guineapigs of two Ba forms which differ in their molecular weight by 4 kDa has been ascribed to the presence of different numbers of oligosaccharide side-chains on the precursor protein factor B (Matsushita et al., 1989; Mizuochi et al., 1990). The possibility that the 34 kDa and the 30 kDa human Ba fragments represent differently glycosylated forms of the same protein was excluded by the demonstration that treatment of these proteins with endoglycosidase F reduced both their molecular weights by 4 kDa. Chemical cleavage with CNBr of a sample which contained heterogeneous Ba fragments from uremic plasma resulted in a single protein with the same molecular weight as a CNBr cleavage product of Ba,,. As limited N-terminal amino acid sequencing of the 30 kDa Ba degradation form revealed an intact Ba aminoterminus and as the 234 amino acid Ba fragment contains a single CNBr cleavage site at methionine 198 this result indicates that the Ba fragment undergoes more extensive proteolytic cleavage at its carboxyterminus besides the degradation by carboxypeptidases. Neither have the proteases which convert Ba into its 30 kDa cleavage products been identified nor have cleavage sites at the Ba carboxyterminus been characterized in more detail, so far. It is therefore not yet clear whether the additional cleavage sites on the Ba fragment which have been described in this study correlate with specific structural features of the Ba fragment. The Ba fragment is composed of internal repeating units which are known as short consensus repeats @CR) or complement control protein (CCP) modules (Baron et al., 1991) and a short, carboxyterminal sequence which shows no homology to other proteins (Morley and

313

fragment Ba

Campbell, 1984). The carboxyterminally truncated Ba fragment identified in human plasma corresponds to those parts of the Ba sequence which are made up of three of these compact protein modules. The finding of carboxyterminal processing of the Ba fragment under in vivo conditions may have functional relevance. The human Ba fragment has been shown to inhibit the proliferation and differentiation of preactivated human B lymphocytes (Zierz and Gotze, 1987; Ambrus et al., 1990). Furthermore, Ambrus et al. (1990) showed that a decamer which corresponds to the 10 carboxyterminal amino acids of Ba also inhibited the proliferation of activated B lymphocytes although at higher concentrations. These results imply that functionally relevant epitopes are located at the carboxyterminus of the Ba fragment. Although possible biological functions of the desArg/Lys derivatives or the 30 kDa degradation products of Ba have yet to be defined these findings imply that processing of the Ba carboxyterminus may represent a control mechanism for this complement activation product which could be of similar importance as the well characterized inactivation of anaphylatoxic peptides C3a or C5a by carboxypeptidases. Acknowledgements-This work was supported by grants DFG, Go 410/4-l and SFB 236, B6. The authors thank MS G. Sonntag for excellent technical assistance. We are grateful to Dr B. Schmidt for providing the synthetic peptide. We would also like to thank Dr B. Zimmermann for performing peptide mass determinations.

REFERENCES Ambrus J. L., Peters M. G., Fauci A. S. and Brown E. J. (1990) The Ba fragment of complement factor B inhibits human B lymphocyte proliferation. J. Zmmun. 144, 154991553. Baron M., Norman D. G. and Campbell I. D. (1991) Protein modules. TIBS 16, 13-17. Baumgarten H., Meyer-Kryst J. and Gotze 0. (1985) Enzymelinked immunosorbent assays for human factor B and its fragments Ba and Bb. Complement 2, 7. Bokisch V. A. and Miiller-Eberhard H. J. (1970) Anaphylatoxin inactivator of human plasma: its isolation and characterization as a carboxypeptidase. J. din. Invest. 49,

2421-2436. Burger R., Zilow G., Bader A., Friedlein A. and Naser W. (1988) The C terminus of the anaphylatoxin C3a generated upon complement activation represents a neoantigenic determinant with diagnostic potential. J. Immun. 141, 553-558. Coste J., Le-Nguyen D. and Castro B. (1990) Py-BOPR: A new peptide coupling reagent devoid of toxic by-product. Tetrahedron Lett. 31, 205-208. Davrinche C., Charlionet R., Rivat C., Helal A. N. and Lefranc G. (1984) Inulin-induced, activation of factor B in whole serum: description of structural modifications in the Ba fragment. Eur. J. Immun. 14, 957-961. Elder J. H. and Alexander S. (1982) Endo-fi-N-acetylglucosaminidase F: Endoglycosidase from flavobacterium meningosepticum that cleaves both high mannose and complex glycoproteins. Proc. natn. Acad. Sci. U.S.A. 79, 4540-4544.

314

M. OPPERMANNand 0. G~TZE

Ellman G. L. (1959) Tissue sulfhydryl groups. Arch. Biochem. Biophys. 82, 70-77. Giitze 0. (1988) The alternative pathway of activation. In The Complement System (Edited by Rother K. and Till G. 0.) pp. 1544168. Springer Verlag, Berlin, Heidelberg. G&e 0. and Mtiller-Eberhard H. J. (1971) The C3-activator system: an alternative pathway of complement activation. J. exp. Med. 134, 9Os-108s. Green N., Alexander H., Olsen A., Alexander T., Shinnick T. M., Sutcliff J. G. and Lerner R. A. (1982) Immunogenic structure of the influenza virus hemagglutinin. Cell 28, 477-487. Hempelmann E. and Oppermann M. (1992) Protein blotting, Immunoblotting. In Monoclonal Antibodies (Edited by Peters J. H. and Baumgarten H.), pp. 448-453. Springer Verlag Berlin, Heidelberg. Hiipken U., Striiber A., Oppermann M., Mohr M., Miicke K. H., Burchardi H., Gijtze 0. (1992) Production and characterization of peptide-specific monoclonal antibodies that recognize a neoepitope on hog CSa. In Host Defense Dysf&rction in Trauma, Shock and Sepsis (Edited by Faist E., Meakins J. and Schildberg F. W.), pp. 411.-417. Springer Verlag, Berlin, Heidelberg. Laemmli U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685. Lesavre P. H., Hugli T. E., Esser A. F. and Miiller-Eberhard H. J. (1979) The alternative pathway C3jC5 convertase: chemical basis of factor B activation, J. Immun. 123, 529-534. Liu F. T., Zinnecker M., Hamaoka T. and Katz D. H. (1979) New procedures for preparation and isolation of conjugates of proteins and a synthetic copolymer of D-amino acids and immunochemical characterization of such conjugates. Biochemistry 18, 690-693. Maack T., Johnson V., Kau S. T, Figueiredo J. and Sigulem D. (1979) Renal filtration, transport and metabolism of low-molecular weight proteins: A review. Kidney Znt. 16, 251-270. Matsushita M. and Okada H. (1989) Two forms of guinea pig factor B of the alternative complement pathway with different molecular weights. Molec. Immun. 26, 6699676.

Mizuochi T., Hamako J., Titani K., Matsushita M. and Okada H. (1990) Structures of the asparagine-linked oligosaccharides of guinea-pig factor B of the alternative complement pathway. Biochem. J. 272, 533-535. Mole J. E., Anderson J. K., Davison E. A. and Woods D. E. (1984) Complete primary structure for the zymogen of human complement factor B. J. biol. Chem. 259,3407-3412. Morley B. J. and Campbell R. D. (1984) Internal homologies of the Ba fragment from human complement component factor B, a class III MHC antigen. EMBO J. 3, 1533157. Oppermann M., Baumgarten H., Brandt E., Gottsleben W., Kurts C. and Giitze 0. (1990) Quantitation of components of the alternative pathway of complement (APC) by enzymelinked immunosorbent assays. J. Immun. Meth. 133, 181-190. Oppermann M., Kurts C., Zierz R., Quentin E., Weber M. H. and Giitze 0. (1991) Elevated plasma levels of the immunosuppressive complement fragment Ba in renal failure. Kidney Int. 40, 939-947. Perrin L. H., Lambert P. H. and Miescher P. A. (1975) Complement breakdown products in plasma from patients with systemic lupus erythematosus and patients with membranoproliferative or other glomerulonephritis. J. Clin. Invest. 56, 165-l 76. Peters J. H. and Baumgarten H. (eds) (1992) Monoclonal Antibodies. Springer Verlag, Berlin, Heidelberg. Plummer T. H. and Ryan T. J. (1981) A potent mercapto bi-product analogue inhibitor for human carboxypeptidase N. Biochem. Biophys. Res. Commun. 98, 448-454. Tack B. F. and Prahl J. W (1976) Third component of human complement: purification from plasma and physicochemical characterization. Biochemistry 15, 45 13-4521. Volanakis J. E., Barnum S. R., Giddens M. and Galla J. H. (1985) Renal filtration and catabolism of complement protein D. N. Engl. J. Med. 312, 395-399. von Zabern I. and Nolte R. (1987) Activation of the alternative pathway of human complement by sullhydryl compounds of analytical and therapeutic use. Int. Archs. Allergy Appl. Immun. 84, 178-184. Zierz R. and Giitze 0. (1987) Regulation of human B lymphocyte functions by the complement fragment Ba. Complement 4, 242-243.