Biochem. J. (1985) 232, 841-850 (Printed in Great Britain) 841 Structural and functional analysis of the complement component Factor H with the use of different and monoclonal antibodies to Factor H Jochem ALSENZ,*II Thomas F. SCHULZ,t John D. LAMBRIS,T Robert B. SIM§ and Manfred P. DIERICHt *Institut fur Medizinische Mikrobiologie, Johannes-Gutenberg-Universitiit, Augustusplatz, D-6500 Mainz, Federal Republic of Germany, tInstitut fur Hygiene, Universitiit Innsbruck, Fritz-Pregel-Strasse 3, A-6020 Innsbruck, Austria, $Department of Immunology, Scripps Clinic and Research Foundation, La Jolla, CA 92037, U.S.A., and §M.R.C. Immunochemistry Unit, Department of Biochemistry, University of Oxford, South Parks Road, Oxford OXI 3QU, U.K.

The action of six different enzymes on the function and structure of Factor H was investigated by use of sodium dodecyl sulphate/polyacrylamide-gel electrophoresis, haemagglutination, two -linked immunosorbent assay systems and an assay for Factor I activity. Six monoclonal antibodies directed against the 38 kDa tryptic fragment ofFactor H [which contains the for C3b (a 180 kDa fragment of the third component ofcomplement) and the cofactor activity] were also used to detect cleavage products derived from the same fragment. , A4 or first cleaved Factor H to 36-38 kDa fragments carrying all six monoclonal anti-(Factor H)-binding sites. In parallel, the interaction of Factor H with surface-bound C3b was lost, whereas the cofactor function was preserved. Further cleavage of the 36-38 kDa fragments into two 13-19 kDa fragments (one carrying the MAH4 and MRC OX 24 epitopes, the other the MAHI, MAH2, MAH3 and MRC OX 23 epitopes) destroyed cofactor activity. Pepsin, bromelain or papain rapidly split offa 13-15 kDa fragment of Factor H carrying the MAH1, MAH2, MAH3 and MRC OX 23 epitopes and destroyed all tested functions of Factor H. Ficin cleaved Factor H into disulphide-linked fragments smaller than 25 kDa, but did not affect the functions of the Factor H molecule. The 38 kDa tryptic fragment of Factor H is the N-terminal end of the Factor H molecule, as determined by N-terminal sequence analysis. A model is presented of the substructure of Factor H.

INTRODUCTION Factor H is a single-chain plasma glycoprotein of about 150-160 kDa containing 9.3% carbohydrate Factor H (formerly fllH) was first described by Nilsson (Sim & DiScipio, 1982). Its frictional ratio of 2.11 & Muller-Eberhard (1965) as a contaminant of C3 and (Sim & DiScipio, 1982), its unusual c.d. spectrum C5 preparations. Its function as a control of the (DiScipio & Hugli, 1982) and its appearance in electron alternative complement pathway was recognized by microscopy (Smith et al., 1983) suggest that Factor H is Whaley & Ruddy (1976a,b). Factor H has several an elongated asymmetric molecule approx. 28 nm x 3 nm regulatory functions in the . Binding (280 A x 30 A), one end of which is larger and more of Factor H to C3b prevents the interaction of C3b with globular than the other. Factor H contains about 33 Factor B (Pangburn & Muller-Eberhard, 1978) and C5 disulphide bridges, which not only stabilize the secondary (Isenmann et al., 1980) and displaces the Bb fragment of conformation of the molecule but are also functionally Factor B from the active C3 convertase and C5 relevant, since their cleavage results in a total loss of all convertase of the alternative pathway (Weiler et al., activities (DiScipio & Hugli, 1982). 1976). Factor H also acts as a cofactor for the serine Tryptic cleavage of Factor H generates 38 kDa and proteinase Factor I, which cleaves C3b to iC3b, thus 142 kDa fragments linked together by disulphide bonds abrogating the action of C3b (Pangburn et al., 1977). (Sim & DiScipio, 1982; Hong et al., 1982; Alsenz et al., Other studies suggest that Factor H releases endogenous 1984). The cleaved Factor H no longer interacts with Factor I from lymphocytes (Lambris et al., 1980) via a surface-bound C3b, whereas the cofactor function for the specific Factor H receptor (Lambris & Ross, 1982; Schulz cleavage of fluid-phase C3b by Factor I remains un- et al., 1984), triggers the proliferation of mouse spleen affected (Sim & DiScipio, 1982; Hong et al., 1982; Alsenz cells (Hammann et al., 1981), initiates the oxidative et al., 1984). We have shown that the C3b-binding site and metabolism in monocytes (Schopfet al., 1982), blocks the the cofactor function of Factor H are localized in the differentiation but not the proliferation ofB-cells (Tsokos 38 kDa tryptic fragment ofFactor H (Alsenz et al., 1984). et al., 1984) and is probably involved in the recognition In the present study Factor H was treated with different ofactivating and non-activating cell surfaces (Horstmann enzymes, and the resulting products were tested for C3b et al., 1984). binding, cofactor activity, and the binding ofsix different

Abbreviations used: C3b, 180 kDa fragment of the third component of complement; Factor H, a controlling protein of the complement system; MAH, monoclonal antibody to Factor H; Factor I, the C3b inactivator; SDS, sodium dodecyl sulphate; e.l.i.s.a., enzyme-linked immunosorbent assay. 11 To whom correspondence should be addressed. Vol. 232 842 J. Al,enz and others monoclonal antibodies to Factor H (MAH 1, MAH2, Factor H fragments was carried out in 150 mm- MAH3, MAH4, MRC OX 23 and MRC OX 24). In NaCl/lO mM-sodium phosphate buffer, pH 7.3, and addition, we determined the sequence of the N-terminal bound fragments were eluted with 150 mM-NaCl/10 mm- ten amino acids of the 38 kDa tryptic fragment. sodium phosphate/3 M-KSCN buffer, pH 7.3 (MAH1), or 150 mM-NaCl/10 mM-sodium phosphate buffer, MATERIALS AND METHODS pH 4.0 (MAH4). Materials E.l.i.s.a. chymotrypsin A4 The binding of monoclonal antibodies to Factor H, Bromelain (6 units/mg), cleaved Factor H or isolated Factor H fragments was (90 units/mg), ficin (3 units/mg), papain (30 units/mg) assayed as described by Schulz et al. (1984). The binding and pepsin (2500 units/mg) were purchased from of fluid-phase C3b to micro-titre plates coated with Boehringer, Mannheim, Germany. Elastase (type 3; Factor H or fragmented Factor H was detected by 70 units/mg) and trypsin [1-chloro-4-phenyl-3-tosyl- polyclonal goat anti-C3 antibodies and a peroxidase- amidobutan-2-one-(' TPCK'-)treated; 400 units/mg] conjugated anti-(goat IgG) antibody (Alsenzet al., 1984). were from Sigma Chemical Co., St. Louis, MO, U.S.A., There was no difference in the binding capacity of the and Serva, Heidelberg, Germany, respectively. All other micro-titre plates for Factor H and cleaved forms of reagents were of analytical grade and obtained from Factor H, as tested with a polyclonal anti-(Factor H) Merck, Darmstadt, Germany, or Sigma Chemical Co. antibody. The interaction of Factor H or cleaved Factor H with C3b-coated micro-titre plates was described by Complement components Alsenz et al. (1984). Human C3, C3b, Factor H and Factor I were purified from fresh human plasma as previously described (Alsenz Factor H haemagglutination assay et al., 1984; Lambris et al., 1980). 125I-C3b was prepared The binding of serially diluted Factor H or enzyme- by the iodogen method (Pierce Chemical Co., Rockford, treated Factor H with cell-bound C3b was tested in a IL, U.S.A.) (Fraker & Speck, 1978) to a specific haemagglutination assay (Schmitt et al., 1981). radioactivity of 5 x 106-5.5 x 106 c.p.m./,ug of C3b. Assay for the Factor H cofactor activity Enzymic fragmentation of Factor H The cofactor activity of Factor H for the cleavage of Enzymic cleavage of Factor H was carried out in C3b by Factor I was assayed by the method of Alsenz 150 mM-NaCl/10 mM-sodiumphosphatebuffer,atpH 7.3 et al. (1984). (trypsin, ficin and chymotrypsin A4), at pH 8.5 (elastase), at pH 6.2 (bromelain and papain) or at pH 2.0 (pepsin). SDS/polyacrylamide-gel electrophoresis and Western The enzymic reaction with trypsin, chymotrypsin A4 or blotting elastase was stopped with 1 mM-di-isopropyl phosphoro- SDS/5-20% -polyacrylamide-gel electrophoresis fluoridate. The cleavage ofFactor H with ficin, bromelain (Laemmli, 1970), Western blot analysis (Towbin et al., or papain was stopped with the same volume of 1979), estimation of molecular masses with reference 100 mM-iodacetamide (in 0.1 M-Tris/HCl buffer, pH 10) (Bio-Rad Laboratories, Richmond, CA, U.S.A.) and incubated for 15 min at 37 'C. The pH was then and other conditions have been described previously adjusted to pH 7.0 with 0.1 M-Tris/HCl buffer, pH 2.0. (Alsenz et al., 1984). The reaction with pepsin was terminated by adjusting the pH to 8.0 with 0.01 M-NaOH and allowing 10 min to Protein sequencing analysis inactivate the pepsin. Before use, bromelain and papain Edman degradation was performed in a Beckman were activated by incubating 5,l ofeach (5 mg/ml) with model 890D sequencer, with the methods described by 15 1 of 150 mM-NaCl/10 mM-sodiumphosphate/10 mM- Brauer et al. (1975) and Zimmermann & Pisconzo (1977). L-cysteine/HCl/2 mM-EDTA buffer, pH 6.2, for 5 min at 22 'C. RESULTS Antibodies Cleavage of Factor H with elastase and chymotrypsin A4 The affinity-purified monoclonal antibodies against Factor H was treated with elastase at an enzyme/ human Factor H, namely MAHI, MAH2, MAH3, substrate ratio of 1: 10 (w/w) and the fragments were MAH4, MRC OX 23 and MRC OX 24, were prepared analysed by SDS/polyacrylamide-gel electrophoresis. In as previously described (Alsenz et al., 1984; Schulz et al., the non-reduced gel, the 148 kDa band of intact Factor 1984; Sim et al., 1983). H was degraded to a 32 kDa fragment via several Polyclonal goat antibodies to Factor H and C3 were intermediates (128 kDa, 106 kDa and 83 kDa) (Fig. la). prepared as described by Schulz et al. (1984). Peroxidase- Western blottinganalysisdemonstrated thatthe fragments conjugated rabbit anti-(goat y-globulin) antibodies and reacted with each of the six monoclonal antibodies peroxidase-conjugated rabbit antibodies to mouse (results not shown). Under reduced conditions the immunoglobulins were purchased from E-Y Labora- 160 kDa band of the intact Factor H was first cleaved by tories, San Marcos, CA, U.S.A., and Dako, Copenhagen, elastase within 2 min into a 124 kDa and a 36 kDa Denmark. fragment (Fig. lb), the latter carrying all antigenic sites for the binding ofthe six monoclonal antibodies (Fig. lc). Affinity chromatography The 124 kDa fragment was then stepwise degraded to The monoclonal antibodies MAH1 and MAH4 were 92 kDa, 85 kDa and 79 kDa pieces. In parallel with the coupled to CNBr-activated Sepharose 4B (Pharmacia, first cleavage, the ability of Factor H to interact with Freiburg, Germany). Purification of pepsin-generated surface-attached C3b was lost (Fig. 2a). In comparison, 1985 Structural and functional studies of Factor H 843

(a) (b) kDa kDa kDa 200 148 ) 128 P IWu - 124 116 106 > 92 83 > 92 W 85 4 78 66

45 _M 40 _am_ -436 ....-% -4 30 32w. 31 - - A -_428 25 - _ (E) (E) 21 4_. - 15 - 14 -416 -4 13 1 5 15 45 120 1 5 15 45 120 Time of incubation of Factor H with elastase (min)

(c) kDa

o 124 < 92 85 78

W 36 Wm

--416 -413 A~~ A A.A1 6

44~AILW MAH 1 MAH 2 MAH 3 MAH 4 MRC MRC Poly- OX 23 0X 24 clonal anti- (Factor H)

Fig. 1. Cleavage of Factor H with elastase Factor H was treated with elastase (E) (enzyme/substrate ratio 1: 10) for different periods of time. The samples were analysed by SDS/5-20%-polyacrylamide-gel electrophoresis (30,ug of Factor H/track) under non-reducing (a) and reducing (b) conditions. In addition, a mixture (1 + 1) of min and 45 min elastase-cleaved Factor H was analysed by the immunoblotting technique on SDS/5-20% -polyacrylamide gels under reducing conditions (c).

the binding of fluid-phase C3b to surface-attached enzyme/substrate ratio of 1: 100 (w/w) was similar to that elastase-cleaved Factor H (Fig. 2b) as well as the cofactor of elastase. were not decreased to the same extent activity (Fig. 2c) of Factor H bromelain and papain as was the binding of Factor H to surface-bound C3b. Cleavage by pepsin, The 36 kDa fragment was further cleaved (probably After enzymic digestion of Factor H with pepsin at an via a 30 kDa fragment) into a 13 kDa fragment carrying enzyme/substrate ratio of 1:1000 (w/w) followed by the MAH1, MAH2, MAH3 and MRC OX 23 epitopes SDS/polyacrylamide-gel electrophoresis under non- and a 16 kDa fragment carrying the MAH4 and MRC reducing conditions, Factor H was degraded very fast OX 24 epitopes. Concomitant with the disappearance of into several main bands (134 kDa, 73 kDa, 68 kDa, the 36 kDa fragment, the binding of fluid-phase C3b to 62 kDa, 53 kDa and 15 kDa) and numerous minor bands surface-bound and cleaved Factor H disappeared, and (Fig. 3a). Immunoblotting showed that MAH1, MAH2, the cofactor activity was decreased (Figs. 2b and 2c, MAH3 and MRC OX 23 reacted with a 15 kDa samples 5 and 6). The action of chymotrypsin A4 at an fragment, whereas MAH4 and MRC OX 24 bound to the

Vol. 232 844 J. Alsenz and others

{a) (b) _

._

tn > l0 t 1500 0

0 a u 'J Q) a) _ VZp o 1000 UZ-°+ 6 0 0 I 1- c:n u 4 C- z 500 _ _ - 2

on

0 0 1 2 3 4 5 6 7 1 2 3 4 5 6 7

(c) kDa

115

< 46

a o c d 2 3 4 5 6 Fig. 2. Functional activities of elastase-cleaved Factor H

Factor H was incubated with elastase (enzyme/substrate ratio 1:10) for 0 (sample 1), 1 (sample 2), 5 (sample 3), 15 (sample 4), 45 (sample 5), and 120 min (sample 6) at 37 'C. The samples were tested by incubation with C3b-coated sheep erythrocytes in a haemagglutination test (a) or attached to micro-titre plates and tested in an e.l.i.s.a. for the binding ofC3b (b). In the cofactor assay (c) the samples (1-6) were incubated with 1211-C3b and Factor I for2 hat 37 'C, subjected to SDS/9-20% -polyacrylamide-gel electrophoresis under reducing conditions and the gels exposed to a Kodak X-ray film for 24 h. 125I-C3b incubated with buffer (a), factor I (b), Factor H (c) or di-isopropyl phosphorofluoridate-pretreated elastase (d) served as controls.

134 kDa fragment and to a variety of fragments with 62 kDa and 15 kDa fragments and traces of a 73 kDa molecular masses between 134 kDa and 35 kDa (results fragment. The eluates of MAHI-Sepharose 4B and not shown). Under reducing conditions, none of the MAH4-Sepharose 4B consisted of intact Factor H and fragments had a molecular mass of more than 59 kDa a 15 kDa fragment and ofintact Factor H and a 134 kDa (Fig. 3b). The antibodies MAHI, MAH2, MAH3 and fragment respectively. Both eluates reacted with all six MRC OX 23 bound to one or two smaller fragments monoclonal antibodies, whereas the breakthrough of the (14-16 kDa), whereas MAH4 and MRC OX 24 reacted MAHI-Sepharose 4B and MAH4-Sepharose 4B bound only with higher-molecular-mass fragments (largest only MAH4 and MRC OX 24 and MAHI, MAH2, fragment: 48 kDa) (Fig. 3c). MAH3 and MRC OX 23 respectively (results not shown). Functional assays showed that treatment of Factor H A comparison of the breakthroughs of the two columns with pepsin caused a complete loss of all activities on SDS/polyacrylamide-gel electrophoresis under re- (C3b-binding, cofactor activity) within 10 min (Figs. ducing conditions revealed that both contained the 4a-4c). In addition, Factor H was treated with pepsin for 59 kDa, 32 kDa and 26 kDa fragments, indicating that 3 min at 37 °C, and the resulting fragments were purified none of these fragnents carried epitopes responsible for by affinity chromatography on MAH I-Sepharose 4B and MAH 1 and MAH4 binding. MAH4-Sepharose 4B. SDS/polyacrylamide-gel electro- The C3b-binding and cofactor activites were only phoresis under non-reducing conditions revealed that the found in the eluates (results not shown). This was breakthrough of the MAHI-Sepharose 4B contained presumably due to the presence of uncleaved Factor H 134 kDa, 73 kDa and 62 kDa fragments, whereas the in the eluates. breakthrough of the MAH4-Sepharose 4B contained Treatment of Factor H with bromelain or papain 1985 Structural and functional studies of Factor H 845

Markers (a) (kDa) 200 W kDa

_ _ _ 4_ ,-1344< 92 66 W *i -o 62 45 W "53

31bo

21 > 15 14 W 1 3 10 30 60 120 (b) 200 -

92 - 66 0 59 45 W_- '46

3 1 W 32 '22 r15 I11 3 10 .30 60 -I20 nepsin .1,,irKPerSnr Lenre -.me of incubation of Factor H with (min) 3 Da |kDaI

6 J mw?*_ t

4 <46 45 ?-

31 FY ; "Oft, -4 32 4 21m ^ X^ 22 14 0- _ 1 5 'O JW AAP 10 db j j

......

1 2 MAH 1 MAH 2 MAH 3 MAH 4 MRC MRC Poly- OX 23 OX 24 cional anti- (Factor H) Fig. 3. Breakdown of Factor H by pepsin Factor H was incubated with pepsin for 1-120 min (enzyme/substrate ratio 1:1000). The resulting fragments were analysed by SDS/5-20% -polyacrylamide-gel electrophoresis (30 #sg of Factor H/track) under non-reducing (a) and reducing (b) conditions. Under reducing conditions the immunoblotting technique was used to determine the binding of the different monoclonal antibodies to the fragrnents generated by pepsin treatment of 10 min (c).

(enzyme/substrate ratio 1:10, w/w) gave results similar the molecular-mass range of 45-130 kDa. By immuno- to those of the cleavage with pepsin. blotting (non-reducing conditions) all these cleavage products were shown to react with the six monoclonal Cleavage of Factor H with ficin antibodies. SDS/polyacrylamide-gel electrophoretic After Factor H was incubated with ficin at an analysis under reducing conditions demonstrated that enzyme/substrate ratio of 1:50 (w/w) for 120 min, Factor H was cleaved by ficin within 3 min into a variety SDS/polyacrylamide-gel electrophoresis under non- offragments, none ofwhich had a molecular mass greater reducing conditions showed hardly any degradation than 25 kDa (Fig. Sb). Immunoblotting analysis under fragments (Fig. 5). Some very faint bands were seen on reducing conditions showed two types of fragments: one the original gel after treatment for more than 120 min in carrying the MAHI, MAH2, MAH3 and MRC OX 23 Vol. 232 846 J. Alsenz and others

(a) (D)

(A 1000 1- 10 i_

u cn GD I 0 u LL 0 Lw -o a 500 kI I- 5 C) -0 ; ._- -c c C) CO co --m

0 o 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9

(c) kDa M- .- ON" 1415

'O 68 .__ <~~~~~- 46

a b c d 3 4 5 Fig. 4. Effect of pepsin on the functional activities of Factor H Factor H was treated with pepsin for 0 (sample 1), 1 (sample 2), 3 (sample 3), 10 (sample 4), 30 (sample 5), 60 (sample 6) and 120 min (sample 7) at 37 'C. The fragments were fixed to micro-titre plates and tested for C3b binding in an e.l.i.s.a. (a) or incubated with sheep erythrocytes (haemagglutination assay) (b). Inactivated pepsin incubated with Factor H (sample 8) or buffer (sample 9) for 120 min at 37 'C served as controls. In the cofactor assay (c) the samples 3, 4, and 5 were incubated with 1251-C3b and Factor I for 2 h at 37 'C and analysed by SDS/9-20% -polyacrylamide-gel electrophoresis. The controls were '251-C3b incubated with buffer (a), Factor I and Factor H (b), Factor I (c) or sample no. 8 (d). epitopes (15 and 13 kDa), and the other the binding sites DISCUSSION for MAH4 and MRC OX 24 (16 kDa) (Fig. Sc). All antibodies bound to a 21 kDa fragment. For functional Treatment of Factor H with different enzymes activities, after 120 min of ficin treatment there was only produced a large variety of fragments. By using six a small reduction in the binding of fluid-phase C3b to monoclonal antibodies, some of the fragments were cleaved Factor H (Fig. 6a). In addition, the binding identified as parts of the 38 kDa tryptic fragment of of cleaved Factor H to C3b-coated sheep erythrocytes Factor H that carried the C3b-binding site and cofactor (Fig. 6b) and C3b-coated micro-titre plates (results not activity. In addition, the consequences of different shown), as well as the cofactor activity, were not affected enzymic cleavages on the functions of Factor H were (Fig. 6c). tested. The six monoclonal antibodies to Factor H, namely N-Terminal amino acid analysis of the tryptic 38 kDa MAHI, MAH2, MAH3 and MAH4 (Alsenz et al., 1984; fragment of Factor H Schulz et al., 1984) and MRC OX 23 and MRC OX 24 The first ten residues of the N-terminal sequence of the (Sim et al., 1983) used here were directed against the 38 kDa tryptic fragment of Factor H, which carry the 38 kDa tryptic fragment of Factor H. Since that part of C3b-binding site and the cofactor activity (Alsenz et al., the molecule carries the C3b-binding site and the cofactor 1984), were obtained by automated Edman degradation activity (Alsenz et al., 1984), we tested smaller fragments and shown to be identical with the N-terminal amino acid to determine if they retained their functional activity. sequence of Factor H published by Sim & DiScipio Thus Factor H was treated with-different enzymes, and (1982), namely: the resulting fragments were characterized by SDS/ 1 5 10 polyacrylamide-gel electrophoresis, by the immuno- Glu-Asp-Glu-Asn-Glu-Leu-Pro-Pro-Arg-Arg blotting technique and by functional assays. 1985 Structural and functional studies of Factor H 847

(a) (b) (Fig. 2a), whereas it could still react with fluid-phase C3b kDa (Fig. 2b) and had cofactor activity (Fig. 2c). Prolonged 200 digestion resulted in the cleavage of the smaller fragment Factor H into a 13 kDa fragment, which reacted with the MAH1, MAH2, MAH3 and MRC OX 23 antibodies, and a 116 16 kDa/ 19 kDa fragment, which reacted with the MAH4 92 and MRC OX 24 antibodies. In parallel, Factor H lost 66 the functional activites that were tested. The second type of enzyme (pepsin, papain and 45 bromelain) rapidly split off a 13-15 kDa fragment from the Factor H molecule that reacted with the MAH1, 31 MAH2, MAH3 and MRC OX 23 antibodies (Fig. 3). This action destroyed all functional activities that were tested (Fig. 4). In addition, affinity-purified pepsin-induced 21 fragments that carry the binding sites for either the antibodies MAH1, MAH2, MAH3 and MRC OX 23 or the antibodies MAH4 and MRC OX 24 were functionally inactive.

3 10 30 60 120 3 10 30 60 120 The third type of enzyme (ficin) cleaved Factor H into Time of incubation of Factor H with ficin (min) several fragments held together by disulphide bonds (Fig. 5). Under non-reducing conditions all the monoclonal antibodies bound to the same fragment (results not (c) shown), whereas under reducing conditions the binding sites for MAHi, MAH2, MAH3 and MRC OX 23 and for MAH4 and MRC OX 24 were found on different kDa fragments. Despite the fact that the molecule was cleaved by ficin at numerous sites, the functional activities of Factor H were not significantly affected (Fig. 6). The 38 kDa tryptic fragment of Factor H containing the binding sites for all six monoclonal antibodies, the C3b-binding site and the cofactor activity was shown here to be the N-terminal end of the Factor H molecule. 21 I- Since Smith et al. (1983) reported that C3b binds to the more-globular segment in the Factor H molecule, 16 - awn 1 5 10 4.1 '' the globular portion must be the N-terminal end of the 4Xs.:::... molecule, including the 38 kDa tryptic fragment. 1 3 01 V Pepsin split off from the Factor H molecule a 15 kDa fragment carrying the MAH 1, MAH2, MAH3 and MRC OX 23 epitopes. Since the pepsin-generated 48 kDa A I i A I .- fragment (reduced conditions) carried the MAH4 and MAH I MAH 2 MAH 3 MAH 4 MRC MRC MRC OX 24 epitopes and was larger than the 38 kDa OX 23 OX 24 tryptic fragment, it must be the connecting part between Fig. 5. Treatment of Factor H with ficin the trypsin-induced 38 kDa and 142 kDa fragments. Factor H was incubated with ficin for 3-120 min (enzyme/ Therefore the 48 kDa fragment consists ofthe C-terminal substrate ratio 1: 50). The resulting fragments were end of the 38 kDa tryptic fragment and the N-terminal analysed by SDS/5-20% -polyacrylamide-gel electro- end of the 142 kDa tryptic fragment (Fig. 7). Therefore phoresis under non-reducing (a) and reducing (b) the 15 kDa fragment must be split offfrom the N-terminal conditions. The binding of the monoclonal antibodies that end of the Factor H molecule (Fig. 7). In addition, the -react with Factor H to the fragments generated by ficin pepsin-generated 59 kDa, 32 kDa and 26 kDa fragments treatment (45 min, 37 °C, enzyme/substrate ratio 1:10) (reducing conditions), which were shown by affinity was determined by the Western blotting technique after chromatography to carry no monoclonal-antibody- SDS/ 15% -polyacrylamide-gel electrophoresis under binding sites, must be derived from the 142 kDa tryptic reducing conditions (c). fragment. Ficin-cleaved Factor H contained under reducing conditions 13 kDa and 16 kDa fragments carrying the MAHI, MAH2, MAH3 and MRC OX 23 epitopes and a 16 kDa fragment bearing the binding sites for MAH4 Three types of enzymes were found that were different and MRC OX 24. Under non-reducing conditions these in their action on Factor H, as follows. epitopes were found on one fragment. Therefore the The first type (elastase/chymotrypsin A4) cleaved 38 kDa tryptic fragment must contain at least one Factor H similarly to trypsin. Initially, Factor H was disulphide bond. cleaved into a larger 124 kDa/ 125 kDa fragment and a Previously we showed that the monoclonal antibodies smaller disulphide-linked 36 kDa/37 kDa fragment, the MAHI, MAH2, MAH4 and MRC OX 24 blocked the latter carrying the binding sites that reacted with the six binding of Factor H to surface-bound C3b (Alsenz et al., monoclonal antibodies. Concomitant with that cleavage, 1984; Sim et at., 1983). In addition, the binding sites for Factor H no longer interacted with surface-bound C3b MAH 1, MAH2 and MAH4 are not available in Vol. 232 848 J. Alsenz and others

(a: z .t -:t. 1000 E I 0- u 0 -E L- I CU r-I G w c,;- c- LL. LA -0 0 IC) m 0 t C- -0 - m 5 500 L) -o 0 a 4--l . -17 2 c L) m LL 4-- LU 0 cn C: -o C: co 0 0 34 56 7 1 2 3 4 5 6 n7 12

kDa

^_ i _ 9 115 _~~~~~~~~~~~~~. 75 68 :,Arm. map " _ _ _ R < 46 40w _0 upx- .

a b c d 1 2 3 4 5 6

Fig. 6. Influence of ficin on the functional acdvities of Factor H

Factor H was treated with ficin (enzyme/substrate ratio 1:50) for 0 (sample 1), 5 (sample 2), 10 (sample 3), 30 (sample 4), 60 (sample 5) and 120 min (sample 6) at 37 'C. The binding of fluid-phase C3b to cleaved Factor H was tested in an e.l.i.s.a. (a), and a haemagglutination assay with sheep erythrocytes (b) was used to detect the binding of cleaved Factor H to surface-bound C3b (b). Buffer served as a control (sample 7). The cofactor activity (c) of Factor H was tested by incubating the samples 1-6 with 25I-C3b and Factor I. 1251-C3b treated with buffer (a), Factor H (b), Factor I (c), and iodoacetamide-pretreated ficin (d) served as controls.

preformed C3b-Factor H complexes (Schulz et al., 1984). appear to modulate the C3b-binding site (Fig. 8, form B). These antibodies therefore bind to the C3b-binding site, This form (B) of Factor H has lost the ability to interact or close to it. In contrast, MAH3 and MRC OX 23 have with surface-attached C3b, whereas the cofactor activity no effect on C3b-Factor H interactions, and the binding remains unaffected or is even increased (Hong et al., 1982; site of MAH3 in preformed C3b-Factor H complexes is Alsenz et al., 1984; Sim et al., 1983). Since the Factor H still detectable. In this study, all activities of Factor H molecule was identical in both assays, the model suggests were destroyed when the binding sites for MAHI and an additional difference between the conformation of MAH2 were separated from the MAH4 and MRC OX 24 surface-bound and fluid-phase C3b. Since MAH 1, epitopes. Therefore it is likely that the C3b-binding site MAH2 and MAH4 inhibit both Factor H binding to C3b within the 38 kDa tryptic fragment of Factor H is and cofactor activity, and because separation of the two localized between the determinants where these two groups of determinants destroys both activities, this groups of monoclonal antibodies react, or is formed by suggests a single binding site for both fluid-phase and two structures of the molecule that are separated by that surface-bound C3b in Factor H. However, although it is cleavage (Fig. 7). As the model in Fig. 8 proposes, the likely that only one binding site exists, these data do not cleavage ofFactor H with trypsin, elastase or chymotryp- totally exclude the possibility that two close but distinct sinogen A4, or the binding of the monoclonal antibody binding sites exist within the 38 kDa tryptic fragment of MRC OX 24 to Factor H (without cleavage of H), both Factor H. 1985 Structural and functional studies of Factor H 849

28 nm MRC -i OX 23 OX 24

C-Terminal 3 nm

MAH 2 C3b-binding site

r-S-S-1 .. S-S.- Trypsin 38 kDa 142 kDa (93 kDa + 52 kDa) I

Pepsin l5kDa a 48 kDa 59 kDa 32 kDa 26 kDa 1 I

r-S-S-1 1--S-S7 ,--S-S -i rs A-SS,.IVS,S* [-S-S-iI ISSI paA I I - - I_. - I I I~~ I Ficin j13 kDa ,16 kDa! I I I I I a ~~~ ~~A a a a~~~~~~~~~~~~~~a I Fig. 7. Model of the substructure of Factor H The size and shape of this model were selected from the data of Sim & DiScipio (1982), DiScipio & Hugli (1982) and Smith et al. (1983). The binding sites of the monoclonal antibodies to Factor H were assigned with respect to the data in the Discussion section. The cleavage sites (broken vertical lines) of the different enzymes, the orientation of the resulting Factor H fragments with regard to the N- and C-terminal ends and the possible positions of disulphide bonds are schematically shown below the model.

_ 7~~~~~~~~4G -~ C

|+Pepsin

Factor H [7'I 1 - LFI - I r-I7 I A

I+ Trypsin

rS-S% rS-S- LA'- B

Surface bound C3b Fluid-phase C3b Fig. 8. Model of the interaction of both native and cleaved Factor H with surface-bound and fluid-phase C3b Interactions of native Factor H (form A), trypsin-cleaved Factor H (form B) and pepsin-cleaved Factor H (form C) with surface-bound and fluid-phase C3b are shown. Vol. 232 850 J. Alsenz and others

This work was supported by grants from the Deutsche Nilsson, U. R. & Muller-Eberhard, H. J. (1965) J. Exp. Med. Forschungsgemeinschaft (SFB 107 and A5) and from the 122, 277-298 Stiftung Volkswagenwerk. J. D. L. is the recipient of an Pangburn, M. K. & Miller-Eberhard, H. J. (1978) Proc. Natl. E.M.B.O. grant (ALTF 121-1982). Acad. Sci. U.S.A. 75, 2416-2420 Pangburn, M. C., Schreiber, R. D. & Muller-Eberhard, H. J. REFERENCES (1977) J. Exp. Med. 146, 257-270 Schmitt, M., Mussel, H. H. & Dierich, M. P. (1981) J. Alzenz, J., Lambris, J. D., Schultz, T. F. & Dierich, M. P. Immunol. Methods 47, 75-85 (1984) Biochem. J. 224, 389-398 Schopf, R. E., Hammann, K. P., Scheiner, O., Lemmel, E.-M. Brauer, A. W., Margolies, M. N. & Haber, E. (1975) & Dierich, M. P. (1982) Immunobiology 46, 307-312 Biochemistry 14, 3029-3035 Schulz, T. F., Scheiner, O., Alsenz, J., Lambris, J. D. & Dierich, DiScipio, R. G. & Hugli, T. E. (1982) Biochim. Biophys. Acta M. P. (1984) J. Immunol. 132, 392-398 709, 58-64 Sim, E., Palmer, M. S., Pulklavec, M. & Sim, R. B. (1983) Fraker, P. J. & Speck, J. C. (1978) Biochem. Biophys. Res. Biosci. Rep. 3, 1119-1131 Commun. 80, 849-857 Sim, R. B. & DiScipio, R. G. (1982) Biochem. J. 205, 285-293 Hammann, K. P., Raile, A., Schmitt, M., Mussel, H. H., Peters, Smith, C. A., Pangburn, M. K., Vogel, C.-W. & Muller- H. & Dierich, M. P. (1981) Immunobiology 160, 289-301 Eberhard, H. J. (1983) Immunobiology 164, 298 Horstmann, R. D., Pangburn, M. K. & Miiller-Eberhard, H. J Towbin, H., Staehelin, T. & Gordon, J. (1979) Proc. Natl. Acad. (1984) Fed. Proc. Fed. Am. Soc. Exp. Biol. 43, 1765 Sci. U.S.A. 76, 4350-4363 Isenmann, D. E., Podack, E. R. & Cooper, N. R. (1980) J. Tsokos, G. C., Inghirami, G., Balow, J. F., Tsoukas, C. D. & Immunol. 124, 326 Lambris, J. D. (1984) Fed. Proc. Fed. Am. Soc. Exp. Biol. Hong, K., Kinoshita, T., Dohi, Y. & Inoue, K. (1982) J. 43, 1763 Immunol. 129, 647-652 Weiler, J. M., Daha, M. R., Austen, K. F. & Fearon, D. T. Laemmli, U. K. (1970) Nature (London) 227, 680-685 (1976) Proc. Natl. Acad. Sci. U.S.A. 73, 3268-3272 Lambris, J. D. & Ross, G. D. (1982) J. Exp. Med. 155, Whaley, K. & Ruddy, S. (1976a) Science 193, 1011-1012 1400-1411 Whaley, K. & Ruddy, S. (1976b) J. Exp. Med. 144, 1147-1163 Lambris, J. D., Dobson, N. J. & Ross, G. D. (1980) J. Exp. Zimmermann, C. L. & Pisconzo, J. J. (1977) Methods Enzymol. Med. 152, 1625-1644 47, 45-51

Received 26 February 1985/22 July 1985; accepted 27 August 1985

1985