Control of the classical and the MBL pathway of complement activation

Molecular Immunology 37 (2000) 803– 811 www.elsevier.com/locate/molimm Control of the classical and the MBL pathway of complement activation Steen Vang Petersen a,*, Steffen Thiel a, Lisbeth Jensen a, Thomas Vorup-Jensen a, Claus Koch b, Jens Christian Jensenius a a Department of Medical Microbiology and Immunology, The Bartholin Building, Uni6ersity of Aarhus, DK-8000 Aarhus, Denmark b Laboratory of Immune De6elopments, Statens Serum Institut, DK-2300 Copenhagen S, Denmark Received 30 October 2000; received in revised form 13 December 2000; accepted 27 December 2000 Abstract The activation of complement via the mannan-binding lectin (MBL) pathway is initiated by the MBL complex consisting of the carbohydrate binding molecule, MBL, two associated serine proteases, MASP-1 and MASP-2, and a third protein, MAp19. In the present report we used an assay of complement activation specifically reflecting the physiological activity of the MBL complex to identify biological and synthetic inhibitors. Inhibitor activity towards the MBL complex was compared to the inhibition of the classical pathway C1 complex and to a complex of MBL and recombinant MASP-2. A number of synthetic inhibitors were found to differ in their activities towards complement activation via the MBL pathway and the classical pathway. C1 inhibitor inhibited both pathways whereas a2-macroglobulin (a2M) inhibited neither. C1 inhibitor and a2M were found to be associated with the MBL complex. Upon incubation at 37°C in physiological buffer, the associated inhibitors as well as MASP-1, MASP-2, and MAp19 dissociated from MBL, whereas only little dissociation of the complex occurred in buffer with high ionic strength (1 M NaCl). The difference in sensitivity to various inhibitors and the influence of high ionic strength on the complexes indicate that the activation and control of the MBL pathway differ from that of the classical pathway. MBL deficiency is linked to various clinical manifestations such as recurrent infections, severe diarrhoea, and recurrent miscarriage. On the other hand, impaired control of complement activation may lead to severe and often chronically disabling diseases. The results in the present report suggests the possibility of specifically inhibiting of the MBL pathway of complement activation. © 2001 Elsevier Science Ltd. All rights reserved. Keywords: MASP; C4 cleavage; Inhibition; Complement activation 1. Introduction The complement system can be activated via three different routes: the alternative pathway, the classical pathway, and the mannan-binding lectin (MBL) pathway. The alternative pathway is controlled by the action of inhibitory proteins present on membranes and in plasma. When C3b is deposited on a non-self surface Abbre6iations: AT III, antithrombin III; a2M, alpha-2-macroglobulin; C1 INH, C1 inhibitor; HSA, human serum albumin; mAb, monoclonal antibody; MASP-1 and MASP-2, mannan-binding lectin associated serine protease-1 and 2; MAp19, MBL-associated protein of 19 kDa; MBL, mannan-binding lectin; SBTI, soybean trypsin inhibitor; TRIFMA, time-resolved immunofluorometric assay. * Corresponding author. Tel.: + 45-89421778; fax: + 45-86196128. E-mail address: svp@microbiology.au.dk (S.V. Petersen). or on immune complexes the tight control of the pathway is intercepted, leading to activation and elimination by phagocytosis or lysis by the terminal complement components (Law and Reid, 1995). The activation of classical pathway characteristically relies on the presence of specific antibodies bound to non-self determinants. The catalytic sub-components, C1r and C1s, are sequentially activated when the C1 complex (C1qC1r2C1s2) binds to immune complexes via the recognition protein, C1q. Activated C1s cleaves complement components C4 and C2 generating the classical C3 convertase, C4bC2b. The activation of C1 is controlled by C1 inhibitor (C1 INH) (Cooper, 1985). C1 INH disassembles the activated C1 complex into C1rC1s(C1 INH)2 complexes, while the recognition protein C1q remains bound to the activator (Ziccardi and Cooper, 1979; Sim et al., 1979a). 0161-5890/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 1 6 1 - 5 8 9 0 ( 0 1 ) 0 0 0 0 4 - 9 804 S.V. Petersen et al. / Molecular Immunology 37 (2000) 803–811 The MBL pathway of complement activation is initiated upon binding of MBL to carbohydrates on microorganisms (Kawasaki et al., 1989; Ihara et al., 1991). Two serine proteases: mannan-binding lectin associated serine protease 1 (Matsushita and Fujita, 1992; Takahashi et al., 1993; Sato et al., 1994) and 2 (Thiel et al., 1997) (MASP-1 and MASP-2) are associated with MBL. Also found in the complex is a protein of 19 kDa referred to as small MBL-associated protein (sMAP) or MBL-associated protein of 19 kDa (MAp19) (Takahashi et al., 1999; Stover et al., 1999). The stoechiometric composition of the complex has not been determined. Upon binding and activation the MBL complex is capable of cleaving complement components C4 and C2 (Ji et al., 1988; Matsushita et al., 2000) generating the C3 convertase, C4bC2b. The physiological relevant control of the MBL pathway of complement activation is still to be determined. Both a2-macroglobulin (a2M) (Terai et al., 1995; Storgaard et al., 1995) and C1 INH (Wong et al., 1999; Matsushita et al., 2000) have been found associated with the MBL complex. In this study we have estimated the inhibitory capacity of various biological and synthetic inhibitors in an assay mimicking the physiological activity of the MBL complex, i.e. estimation of C4b deposition on a relevant carbohydrate surface. The activity of the components towards the classical pathway of complement activation was also determined. One further examined the fate of the complexes upon activation in different buffers. 2. Materials 2.1. Reagents Human serum was prepared from normal human blood after coagulation at room temperature for 2 h. MBL deficient serum was similarly prepared from volunteers of the staff who had below 20 ng MBL pr. ml plasma. Benzamidine (B-6506), pepstatin A (P-4265), soybean trypsin inhibitor (SBTI) (T-9003), phenylmethylsulfonyl fluoride (PMSF) (P-7626), o-phenanthroline (P-9375), a1-antitrypsin (A-9024), antithrombin III (AT III) (A-7388), a2M (M-6159), C1 INH (C-2412), goat anti-C1 INH antiserum (C-8159), and goat antia2M antiserum (M-5649) were obtained from Sigma, St. Louis, MO, USA; a2M was also a gift from Lars Sottrup-Jensen, University of Aarhus (Sottrup-Jensen et al., 1980); basic pancreatic trypsin inhibitor (aprotinin) (596676) from Bayer AG, Leverkusen, Germany. Pefabloc®SC (1429876) and leupeptin (1017101) from Boehringer Mannheim, Mannheim, Germany; Cyklokapron (073494) from Kabi Pharmacia AB, Uppsala, Sweden; human serum albumin (HSA) (440511), normal human IgG (007740), mAb anti-C4 (162-1 and 162-2), and mAb anti-MBL (131-1) were from Statens Serum Institut, Copenhagen, Denmark; rabbit anti-mouse Ig (Z259) and rabbit anti-rat Ig (Z147) from DAKO A/S, Glostrup, Denmark. Polyclonal antibodies were purified from goat anti-C1 INH antiserum and from goat anti-a2M antiserum on protein G beads (HiTrap Protein G, Amersham-Pharmacia, Uppsala, Sweden) as described by the manufacturer. Biotinylation of proteins was done as described by Guesdon et al. (1979). Antibodies were labelled with europium as described by the manufacturer (Wallac Oy, Turku, Finland). Briefly, europium-labelling reagent (1244-301, Wallac Oy) was added to the antibody and incubated overnight at room temperature. Residual active groups on the labelling reagent were blocked by adding 1 M ethanolamine, pH 8.0 and incubating for 1 h at room temperature. After addition of 10 mM Tris – HCl, 140 mM NaCl, 1.5 mM NaN3, pH 7.4 (TBS) with 0.1% (w/v) HSA to a final concentration of 50 mg Ab/ml the labelled antibody was dialysed in TBS to remove excess Eu3 + -labelling reagent. 2.2. MBL and MASP preparations Ten millilitres of a MBL preparation (clinical grade MBL, lot MO-04, Statens Serum Institut, 300 mg MBL and 5 mg HSA/ml saline) containing MASPs and MAp19 was diluted with an equal volume of TBS with 0.05% (v/v) Tween 20, pH 7.4 (TBS/tw), 10 mM CaCl2. The solution was passed through a column containing 50 ml mannose-TSK beads pre-equilibrated in TBS/tw with 5 mM CaCl2 (TBS/tw/Ca). The mannose derivatised TSK HW-75 beads (14985, Merck KgaA, Darmstadt, Germany) were prepared by coupling mannose to divinylsulphone-activated TSK 75 HW beads as described (Fornstedt and Porath, 1975). The column was washed extensively with TBS/tw/Ca, and bound material eluted with TBS/tw with 15 mM EDTA. Fractions containing MBL complex were pooled and diluted in 50 mM CH3COONa, 7.5 mM NaN3, 1 mM EDTA, 0.01% (v/v) polyoxyethylene 10 tridecyl ether (Emulfogen, Sigma, P-2393), pH 4.5 (buffer A) and passed through a Mono S HR 5/5 column (Amersham-Pharmacia) pre-equilibrated in buffer A. After washing with buffer A, bound material was eluted with a stepwise increase of NaCl in buffer A (final concentration 1 M NaCl). Eluted fractions were neutralised with 1 M Tris – HCl, pH 8.5. In order to purify MBL devoid of associated proteins fractions containing MBL with no detectable C4 activating activity, i.e. below 10 − 5 of the activity of the MBL complex and no detectable MASP-1, MASP-2 or MAp19 by Western blotting, were pooled and used as MBL, i.e. defined as MASP-1/MASP-2/MAp19 depleted MBL. The concentration of MBL was evaluated by a sandwich time-resolved immunofluorometric assay (Christiansen et al., 1999). S.V. Petersen et al. / Molecular Immunology 37 (2000) 803–811 3. Methods 3.1. Inhibition of C1 and MBL complex mediated C4b deposition Microtiter wells (FluoroNunc, Nunc, Kamstrup, Denmark) were coated overnight at room temperature with 1 mg of mannan (prepared according to Nakajima and Ballou, 1974)or 1 mg of normal human IgG in 100 ml of 15 mM Na2CO3, 35 mM NaHCO3, 1.5 mM NaN3, pH 9.6 (coating buffer). The wells were emptied and blocked at room temperature for 1 h with 200 ml of TBS with 1 mg HSA/ml. The mannan-coated wells were incubated with 100 ml MBL-sufficient human serum diluted 200-fold in 20 mM Tris – HCl, 10 mM CaCl2, 1 M NaCl, 0.05% (v/v) Triton X-100, 0.1% (w/v) HSA, pH 7.4 (MBL binding buffer). The IgG coated wells were incubated with 100 ml of the same serum diluted 300-fold in 4 mM barbital, 145 mM NaCl, 2 mM CaCl2, 1 mM MgCl2, 3.8 mM NaN3, pH 7.5 (BBS2 + ) with 0.05% (v/v) Tween 20. Dilutions were selected to yield the same amount of C4b deposited on the activating surfaces. After incubation overnight at 4°C in a humid chamber the wells were washed three times in TBS/tw/Ca. Dilution series of synthetical and biological inhibitors were made in BBS2 + . Three fold dilution series were made of a1-antitrypsin (1 mg/ml; 19 mM), AT III (10 U/ml; 0.49 mM), C1 INH (0.4 mg/ml; 3.6 mM) and a2M (1 mg/ml; 1.4 mM). Ten fold dilution series were made of Pefabloc®SC (100 mM), benzamidine (100 mM), SBTI (2 mg/ml; 0.1 mM), PMSF (10 mM), cyklokapron (100 mM), pepstatin A (1 mM), leupeptin (10 mM), o-phenanthroline (10 mM), and aprotinin (430 mM). The serial dilutions of the reagents were further diluted 10-fold in BBS2 + containing 5.6 mg C4/ml (29.4 mM) (prepared according to Dodds, 1993) before adding to mannan or IgG coated wells. All samples were in duplicate. The plates were incubated at 37°C for 90 min in a humid chamber. Following three washes in TBS/tw/Ca, deposited C4b was detected by a mixture of two biotinylated monoclonal anti-C4 antibodies (mAb 162-1 and 162-2), each at 0.5 mg/ml TBS/ tw/Ca. The buffer for IgG plates contained additionally 0.86 M NaCl (1 M NaCl total) to prevent the binding of biotinylated mAb via surface bound C1q. After 1 h incubation and wash in TBS/tw/Ca, bound antibody was detected by incubation for 1 h with 10 ng of Eu3 + labelled streptavidin (1244-360, Wallac Oy) in 100 ml TBS/tw with 25 mM EDTA, followed by wash and addition of enhancement solution (Wallac Oy). The fluorescent chelate was estimated by time-resolved immunofluorometry using a DELFIA® fluorometer (Wallac Oy). Similar experiments were carried out with MBL and recombinant MASP-2 (rMASP-2): MBL and purified rMASP-2 (Vorup-Jensen et al., 2000) were incubated 805 for 2 h at 4°C in MBL binding buffer and transferred to mannan coated wells for incubation overnight at 4°C. Control wells received MBL binding buffer without MBL/rMASP-2. The effect of inhibitors on C4 activation activity was thereafter assessed as described above. 3.2. Detection of inhibitors associated with the MBL complex Microtiter wells (FlouroNunc), coated with mannan or IgG and blocked with HSA, as described above, were incubated with 100 ml of MBL sufficient serum or MBL deficient serum diluted in BBS2 + . As control, the MBL-sufficient serum was diluted in BBS2 + with 100 mM mannose. The MBL deficient serum was also analysed after the addition of purified MBL complex to a final concentration of 3 mg MBL/ml serum. For the detection of inhibitors in the presence of 1 M NaCl samples were also diluted in MBL binding buffer. All samples were in duplicate. Wells were left over night at 4°C. Following this incubation, some of the plates were transferred to 37°C in a humid chamber for 2 h to allow for activation while other plates were kept at 4°C. The wells were then washed and incubated with 100 ml biotinylated goat anti-C1 INH (2.7 mg/ml) or goat anti-a2M (1.6 mg/ml) in TBS/tw/Ca. After incubation and washing in TBS/tw/Ca, bound antibody was detected with Eu-labelled streptavidin as described above. Detection of inhibitors was also done in capture antibody-coated wells. Microtiter wells (FlouroNunc) were coated overnight at room temperature with 100 ng of mAb 131-1 anti-MBL in 100 ml PBS. After blocking for 15 min in TBS/tw, wells received dilution series of MBL-sufficient or -deficient serum diluted in cold BBS2 + . Following incubation overnight at 4°C wells were washed with TBS/tw/Ca and biotinylated antibodies were added: 100 ng of mAb 131-1 anti-MBL, goat anti-C1 INH, or goat anti-a2M diluted in cold BBS2 + . After incubation and washing in TBS/tw/Ca, bound antibody was detected as described above. 3.3. Detection of MASP-1 and MASP-2 /MAp19 associated with mannan-bound MBL This assay was essentially carried out as described above (Section 3.2) with the following modifications: After incubation with serum at 37°C, wells were washed and added either mouse anti-MASP-1, antibody (mAb 9.D8, S.V. Petersen et al., to be published) at 1 mg/ml TBS/tw/Ca, or polyclonal rat anti-MASP-2 (produced by immunising rats with recombinant MASP-2), diluted 500-fold in TBS/tw/Ca with 0.1% (w/v) heat aggregated IgG (63°C, 30 min) and 0.1% (w/v) heat treated HSA (75°C, 30 min). Following incubation for 2 h at room temperature wells were washed and incubated with 806 S.V. Petersen et al. / Molecular Immunology 37 (2000) 803–811 either rabbit anti-mouse Ig (DAKO A/S) or rabbit anti-rat Ig (DAKO A/S), both europium-labelled as described above. After incubation at room temperature for 90 min, wells were washed and developed as described above. 4. Results 4.2. Inhibition of MBL-rMASP-2 complexes A complex of MBL-rMASP-2 bound to a mannan surface was capable of activating C4 and hence deposit C4b (Vorup-Jensen et al., 2000). The effect of the selected natural inhibitors on MBL-rMASP-2 resembled the effect on the natural MBL complex (Fig. 2c). Fifty percent inhibition of the C4b depositing capacity was obtained by 23 nM of C1 INH. As is the case for the natural MBL complex the presence of aprotinin 4.1. Inhibition of the classical and the MBL pathway of complement acti6ation The functional activity of the C1 and the MBL complex was assessed through the C4 activation activity by the amount of C4b deposited on a surface coated with IgG or mannan, respectively. The serum dilution in the two assays were selected to obtain a similar amount of deposited C4b on the activating surfaces. The capacity of the two pathways to deposit C4b was examined in the presence of a variety of synthetic and biological inhibitors. C4b deposition by the classical pathway was reduced dose dependently by leupeptin and benzamidine. Pefabloc®SC and PMSF shows a tendency to inhibition, whereas none of the other synthetic inhibitors affected the deposition of C4b (Fig. 1a). A 50% reduction of C4 activation activity by leupeptin and benzamidine was obtained with 39 mM and 7 mM, respectively. The activity of the MBL complex was also inhibited by leupeptin (5 mM) and benzamidine (1 mM) but in addition by SBTI (3 mM), Pefabloc®SC (257 mM), and to a lesser extent by PMSF and pepstatin A (Fig. 1b) (the concentration required for 50% inhibition given in parenthesis). Five biological proteinase inhibitors were tested for inhibition of the two pathways: a2M, aprotinin, and the three serpins, C1 INH, a1-antitrypsin, and antithrombin III (AT III). C1 INH inhibited both pathways; a 50% reduction of C4b deposition activity by the classical and the MBL pathway was obtained with 14 and 6 nM, respectively (Fig. 2a and b). Aprotinin proved to be selective in inhibiting the MBL pathway but not the classical pathway. Fifty percent inhibition required 1 mM of aprotinin. A preparation of a2M was initially found to inhibit both complement activating pathways, but further purification by gel permeation chromatography showed that this inhibitory effect could not be assigned to a2M (Fig. 2a and b). This result was confirmed using purified a2M (SottrupJensen et al., 1980). Neither a1-antitrypsin nor AT III affected the C4b depositing activity of the complexes (Fig. 2a and b), although an inhibitory capacity of AT III could be induced in both pathways by the addition of a physiological relevant concentration of heparin (0.25 U/ml) (not shown). Fig. 1. The effect of synthetic protease inhibitors on C4b deposition. The activity of the C1 complex (a) and the MBL complex (b) was measured by incubating diluted NHS in wells coated with IgG (a) or with mannan (b). Following incubation overnight at 4°C and wash, the proenzymes of both complexes were allowed to activate at 37°C in the presence of a constant amount of C4 mixed with inhibitors at various concentrations. Deposited C4b was detected with Eu3 + -labeled mAb anti-human C4. The C4b depositing activity in A and B was expressed as percentage by defining 100% as the mean of the counts produced in all wells with the highest dilution (1 nM) of the inhibitors. The experiment was repeated three times with minor modifications and yielding similar results. S.V. Petersen et al. / Molecular Immunology 37 (2000) 803–811 807 also reduced C4b deposition. A 50% reduction was obtained with 342 nM aprotinin (Fig. 2c). a2M and a2-antitrypsin showed no inhibitory effect on this complex. The addition of AT III enhanced C4b deposition (not shown). 4.3. Association of inhibitors with the MBL complex Fig. 2. The effect of biological protease inhibitors on C4b deposition. Conditions in (a) and (b) as described in Fig. 1. In (c), a complex of MBL and rMASP-2 was formed by incubating MBL in mannan coated wells followed by wash and addition of rMASP-2. The complex was allowed to form overnight at 4°C. Following wash, the complex was allowed to activate in the presence of a constant amount of C4 mixed with inhibitors at various concentrations. Deposited C4b was detected and depicted as described in Fig. 1. The experiment was repeated three times with minor modifications and yielding similar results. After incubation of MBL-sufficient serum at 4°C in mannan coated wells both C1 INH and a2M were found attached to the surface (Fig. 3a and c). This attachment depended on the binding of the MBL complex, as the inhibitors were not detected after incubation with MBL-sufficient serum in the presence of 100 mM mannose (inhibits the binding of MBL) or in MBL-deficient serum. Binding of C1 INH and a2M was recorded when purified MBL complex was added to the MBL deficient serum. A marked reduction in the levels of associated a2M and C1 INH was observed if the wells were subsequently incubated at 37°C (Fig. 3b and d). The reduction of the amount of inhibitor on the surface was not a result of MBL dissociating from the surface as estimated by binding of anti-MBL antibody (not shown). The dissociation of inhibitors was influenced by the ionic strength as dissociation was reduced significantly in buffer containing 1 M NaCl (Fig. 3b and d). To determine whether the inhibitors were associated with the MBL complex in fluid phase a sandwich TRIFMA was performed. When MBL was captured on plates coated with a monoclonal anti-MBL antibody we could detect both C1 INH and a2M with a secondary antibody (data not shown). In parallel, the association of MASP-1, MASP-2, and MAp19 with mannan-bound MBL, was examined. Both MASP-1 and MASP-2/ MAp191 were detected in wells after incubation of serum at 4°C in mannan-coated wells at physiologic ionic strength (BBS2 + ) or buffer containing high salt (MBL binding buffer) (Fig. 4a and b). A marked reduction in the amount of MBL associated MASP-1 (Fig. 4a) as well as MASP-2/MAp19 (Fig. 4b) was evident after activation was allowed by incubation at physiological ionic strength at 37°C. If the incubation at 37°C was carried out in buffer containing 1 M NaCl, MBL-associated MASP-1 and MASP-2/MAp19 were found to dissociate to a far lesser degree than at physiological ionic strength (Fig. 4a and b). The reduction in MBL-associated proteins were not due to MBL dissociating from the surface as estimated by an antiMBL antibody (not shown). The binding and dissociation of MASPs and MAp19 thus parallels that of C1 INH and a2M. 1 MASP-2/MAp19 does not indicate a complex. The antibody used detects both MASP-2 and MAp19. 808 S.V. Petersen et al. / Molecular Immunology 37 (2000) 803–811 Fig. 3. Inhibitors associated with the MBL complex. Wells coated with mannan were incubated with dilutions of MBL sufficient serum ( ), MBL sufficient serum with 100 mM mannose added ( ), MBL deficient serum ( ), or MBL deficient serum with added MBL (). Buffer were at physiological ionic strength (— ) or at 1M NaCl (· · ·). After overnight incubation at 4°C wells were either left at 4°C (a, c) or transferred to 37°C for 2 h (b, d). Following wash, wells were incubated with biotinylated goat anti-C1 INH or goat anti-a2M. Bound antibody was detected with Eu3 + -labeled streptavidin. The experiment was repeated three times with similar results. S.V. Petersen et al. / Molecular Immunology 37 (2000) 803–811 5. Discussion Upon binding of C1 or MBL complexes onto an activating surface, the proenzymic entity of the complexes is activated and mediates the cleavage of C4 into C4a and C4b, some of the latter being covalently deposited onto the surface. This phenomenon has been used to investigate the effect of different enzyme inhibitors towards the activity of C1 and MBL complexes. Among the synthetic inhibitors only leupeptin and benzamidine inhibited the activity of C1 while the activity of the MBL complex was inhibited by several inhibitors. The observed difference was unexpected since a priory one might predict similar inhibitor susceptibility of enzymes with the same substrate specificity. Murine MASP was shown to be much less sensitive than C1s to substrate alterations (mutations in murine Fig. 4. Detection of MASP-1, MASP-2, and MAp19 associated with MBL. The experiment was carried out as described in Fig. 3, only the detecting antibodies were mAb anti-MASP-1 (a) or rat anti-MASP-2 antiserum (detecting both MASP-2 and MAp19) (b). Bound antibodies were detected with Eu3 + labelled rabbit anti-mouse Ig or rabbit anti-rat Ig. The figure represents one experiment of three with similar results. 809 C4 and C3 near the cleavage site) (Ogata et al., 1995). It has also been shown that the specificity of serine proteases is not exclusively dictated by the S1 amino acid2 (the amino acid of the protease, which interact with the N-terminal amino acid of the substrate peptide bond to be cleaved) (Hedstrom et al., 1992). Thus, although MASP-1, MASP-2, C1r, and C1s are all trypsin-like serine proteases, the differences between MASP specificity and C1s specificity (Ogata et al., 1995) and the difference of interaction with synthetic inhibitors reported here may indicate differences in substrate interaction requirements in the vicinity of the active site. Indeed, the structure of the C1s serine protease domain (Gabriaud et al., 2000) indicates that the high substrate specificity of C1s results from restrictive access to substrate binding sub-sites. These restrictive elements may be modified in the homologous MASPs. The interaction of five different biological enzyme inhibitors with C1 and MBL complexes was examined. C1 INH is recognised as the principal serum inhibitor of the C1 complex (Sim et al., 1979b). The activity of the C1 INH towards C1 is evident in the C4b deposition assay. A total of 11 nM C1 INH were found to reduce the activity 2-fold (Fig. 2a) which is comparable to the physiologic concentration of C1 INH in serum diluted 300-fold (1.9 mM; 0.2 mg/ml in serum). In the C4b deposition assay no interaction was found between a2M and the C1 complex. This is in agreement with previous reports where no interaction between activated C1r or C1s and a2M could be detected (Ziccardi and Cooper, 1976; Sim et al., 1979b). The esterolytic activity of C1s can be inhibited by AT III in the presence of heparin which enhances the activity of AT III, but not in the absence of heparin (Ogston et al., 1976). Also no interaction between activated C1s and a1-antitrypsin was detected (Ogston et al., 1976). In concordance, no inhibition of C1 activity was observed by AT III or a1-antitrypsin (Fig. 2a) but in the presence of heparin, an increased inhibitory capacity of AT III was observed. C1 INH and a2M have previously been observed to be associated with the MBL complex (Terai et al., 1995; Storgaard et al., 1995). A functional interaction between C1 INH and the MBL complex has recently been published (Matsushita et al., 2000). This finding was also demonstrated in the experimental system (Fig. 2b). The association between a2M and the MBL complex reported previously and here is not reflected in a functional interaction in our C4b deposition assay. Terai et al. (1995), show that a2M can inhibit the lysis of mannan-coated sheep erythrocytes sensitised with MBL/MASP. The discrepancy between these results is 2 Nomenclature for interaction sites as described by Schechter and Berger (1967). 810 S.V. Petersen et al. / Molecular Immunology 37 (2000) 803–811 not fully understood. Pre-incubating mannan-bound MBL complex with a2M (as done by Terai et al. 1995) did not result in the reduction of C4b depositing capacity in the assay. These results may indicate that the functional interaction between the MBL complex and a2M reported by Terai et al. (1995) is due to inhibition of an enzymatic activity other than the C4 cleaving step of the complement activation cascade, i.e. the activity of MASP-2. Neither AT III nor a1-antitrypsin affected the activity, but the inhibitory capacity of AT III was expressed in the presence of heparin, as observed for the C1 complex. Aprotinin was found to inhibit the C4 cleaving activity of the MBL pathway but not the classical pathway. Aprotinin belongs to the Kunitz family of protease inhibitors and is present in tissue and blood where the primary target of inhibition is kallikrein, although other serine proteases can be inhibited (Fritz and Wunderer, 1983). The finding that the classical pathway is not affected by aprotinin is in agreement with the findings of Arlaud and Thielens (1993) who showed that C1s can not be inhibited by BPTI (bovine pancreatic trypsin inhibitor or aprotinin). This supports the suggestion that the active site of proteases of the MBL complex is less restricted that the case for C1s. In order to determine whether these inhibitors interacted with MASP-2 we studied a complex of MBL and rMASP-2 capable of activating C4 (Vorup-Jensen et al., 2000). As indicated by the results in Fig. 2c, C1 INH and aprotinin interact with MASP-2, thereby inhibiting C4b deposition by the complex. This finding is consistent with MASP-2 being the C4 cleaving component of the MBL complex. The presence of a2M or a1-antitrypsin did not inhibit the activity of MBL/rMASP-2. Unexpectedly, AT III enhanced the activity. Such enhancement by AT III was not seen by serum derived MBL complex (Fig. 2b) nor with purified MBL complex (not shown). C1 INH and a2M were found associated with both fluid phase and mannan-bound MBL at 4°C. Associated C1 INH and a2M dissociated from the mannan-bound complex upon incubation at 37°C. One examined if the dissociation of C1 INH and a2M was reflected by a decrease in MBL-associated proteins. Indeed MASP-1 and MASP-2/MAp19 were found to dissociate from the MBL complex upon incubation at 37°C at physiological ionic strength. The dissociation of the MBL complex at 37°C at physiological ionic strength may be a result of activation and active displacement of proteases by associated inhibitors. This would be in analogy with the activated C1 complex where C1r and C1s is actively displaced by C1 INH. If the MBL complex was incubated at 37°C in buffer containing high salt (1 M NaCl) both C1 INH and a2M remained associated with the complex. Also, only small amounts of MASP-1 and MASP-2/MAp19 dissociated from the complex under these conditions. This is in concordance with the observation that only high ionic strength and EDTA in combination can disrupt the MBL complex in fluid phase (Thiel et al., 2000). The presence of 1 M NaCl could tentatively inhibit the activation of the MBL complex thus maintaining the quaternary structure of the complex associated to the surface. These findings is in contrast to the situation of the C1 complex, which is separated into sub-components by incubation in buffer with high ionic strength (Colten et al., 1968). Also, the binding of C1q to immunoglobulin will be disrupted at high ionic strength (Burton et al., 1980). The results presented indicate that the activation and control of the MBL complex differ significantly from that of the C1 complex of the classical pathway. The apparent difference in interaction with synthetic inhibitors suggest the possibility of controlling the MBL pathway without affecting the classical pathway. This may provide a new therapeutic modality in cases where excess activation of the MBL pathway results in clinical manifestations. Acknowledgements We thank Annette G. Hansen for excellent technical assistance. The investigation was supported by grants from The Danish Medical Research foundation and from The Novo Foundation. References Arlaud, G.J., Thielens, N.M., 1993. Human complement serine proteases C1r and C1s and their proenzymes. Methods Enzymol. 223, 61 – 82. Burton, D.R., Boyd, J., Brampton, A.D., Easterbrook-Smith, S.B., Emanuel, E.J., Novotny, J., Rademacher, T.W., van Schravendijk, M.R., Sternberg, M.J.E., Dwek, R.A., 1980. The C1q receptor site on immunoglobulin G. Nature 288, 338– 344. Christiansen, O.B., Kilpatrick, D.C., Souter, V., Varming, K., Thiel, S., Jensenius, J.C., 1999. Mannan-binding lectin deficiency is associated with unexplained recurrent miscarriage. Scand. J. Immunol. 49, 193– 196. Colten, H.R., Borsos, T., Rapp, H.J., 1968. Reversible loss of activity of the first component of complement (C%1) as a function of ionic strength. J. Immunol. 100, 799– 807. Cooper, N.R., 1985. The classical complement pathway: activation and regulation of the first complement component. Adv. Immunol. 37, 151– 216. Dodds, A.W., 1993. Small-scale preparation of complement components C3 and C4. Methods Enzymol. 223, 46 – 61. Fornstedt, N., Porath, J., 1975. Characterization studies on a new lectin found in seeds of Vicia ervilia. FEBS Lett. 57, 187– 191. Fritz, H., Wunderer, G., 1983. Biochemistry and applications of aprotinin, the kallikrein inhibitor from bovine organs. Arzneimittelforschung 33, 479– 494. Gabriaud, C., Rossi, V., Bally, I., Arlaud, G.J., Fontecilla-Camps, J.C., 2000. Crystal structure of the catalytic domain of human S.V. Petersen et al. / Molecular Immunology 37 (2000) 803–811 complement C1s: a serine protease with a handle. EMBO J. 19, 1755– 1765. Guesdon, J.L., Ternynck, T., Avrameas, S., 1979. The use of avidinbiotin interaction in immunoenzymatic techniques. J. Histochem. Cytochem. 27, 1131– 1139. Hedstrom, L., Szilagyi, L., Rutter, W.J., 1992. Converting trypsin to chymotrypsin: the role of surface loops. Science 255, 1249– 1253. Ihara, S., Takahashi, A., Hatsuse, H., Sumitomo, K., Doi, K., Kawakami, M., 1991. Major component of Ra-reactive factor, a complement-activating bactericidal protein, in mouse serum. J. Immunol. 146, 1874– 1879. Ji, Y.H., Matsushita, M., Okada, H., Fujita, T., Kawakami, M., 1988. The C4 and C2 but not C1 components of complement are responsible for the complement activation triggered by the Ra-reactive factor. J. Immunol. 141, 4271– 4275. Kawasaki, N., Kawasaki, T., Yamashina, I., 1989. A serum lectin (mannan-binding protein) has complement-dependent bactericidal activity. J. Biochem. Tokyo 106, 483– 489. Law, S.K.A., Reid, K.B.M., 1995. Activation and control of the complement system. In: Male, D. (Ed.), In fucus: Complement. Oxford University Press, Oxford, pp. 11 – 28. Matsushita, M., Fujita, T., 1992. Activation of the classical complement pathway by mannose-binding protein in association with a novel C1s-like serine protease. J. Exp. Med. 176, 1497– 1502. Matsushita, M., Thiel, S., Jensenius, J.C., Terai, I., Fujita, T., 2000. Proteolytic activities of two types of mannose-binding lectin associated serine protease. J. Immunol. 165, 2637– 2642. Nakajima, T., Ballou, C.E., 1974. Characterization of the carbohydrate fragments obtained from Saccharomyces cerevisiae mannan by alkaline degradation. J. Biol. Chem. 249, 7679– 7684. Ogata, R.T., Low, P.J., Kawakami, M., 1995. Substrate specificities of the protease of mouse serum Ra-reactive factor. J. Immunol. 154, 2351– 2357. Ogston, D., Murray, J., Crawford, G.P.M., 1976. Inhibition of the activated C1s subunit of the first component of complement by antithrombin III in the presence of heparin. Thromb. Res. 9, 217– 222. Sato, T., Endo, Y., Matsushita, M., Fujita, T., 1994. Molecular characterization of a novel serine protease involved in activation of the complement system by mannose-binding protein. Int. Immunol. 6, 665– 669. Schechter, I., Berger, A., 1967. On the size of the active site in proteases. I. Papain. Biochem. Biophys. Res. Commun. 27, 157– 162. Sim, R.B., Arlaud, G.J., Colomb, M.G., 1979a. C1 inhibitor-dependent dissociation of human complement component C1 bound to immune complexes. Biochem. J. 179, 449– 457. Sim, R.B., Reboul, A., Arlaud, G.J., Villiers, C.L., Colomb, M.G., 1979b. Interaction of 125I-labelled complement subcomponents C1r and C1s with protease inhibitors in plasma. FEBS Lett. 97, 111– 115. . 811 Sottrup-Jensen, L., Petersen, T.E., Magnusson, S., 1980. A thiol-ester in alpha 2-macroglobulin cleaved during proteinase complex formation. FEBS Lett. 121, 275– 279. Storgaard, P., Holm Nielsen, E., Skriver, E., Andersen, O., Svehag, S.E., 1995. Mannan-binding protein forms complexes with alpha2-macroglobulin. A protein model for the interaction. Scand. J. Immunol. 42, 373– 380. Stover, C.M., Thiel, S., Thelen, M., Lynch, N.J., Vorup Jensen, T., Jensenius, J.C., Schwaeble, W.J., 1999. Two constituents of the initiation complex of the mannan-binding lectin activation pathway of complement are encoded by a single structural gene. J. Immunol. 162, 3481– 3490. Takahashi, A., Takayama, Y., Hatsuse, H., Kawakami, M., 1993. Presence of a serine protease in the complement-activating component of the complement-dependent bactericidal factor, RaRF, in mouse serum. Biochem. Biophys. Res. Commun. 190, 681– 687. Takahashi, M., Endo, Y., Fujita, T., Matsushita, M., 1999. A truncated form of mannose-binding lectin-associated serine protease (MASP)-2 expressed by alternative polyadenylation is a component of the lectin complement pathway. Int. Immunol. 11, 859– 863. Terai, I., Kobayashi, K., Matsushita, M., Fujita, T., Matsuno, K., 1995. alpha 2-Macroglobulin binds to and inhibits mannose-binding protein-associated serine protease. Int. Immunol. 7, 1579– 1584. Thiel, S., Vorup Jensen, T., Stover, C.M., Schwaeble, W., Laursen, S.B., Poulsen, K., Willis, A.C., Eggleton, P., Hansen, S., Holmskov, U., Reid, K.B., Jensenius, J.C., 1997. A second serine protease associated with mannan-binding lectin that activates complement. Nature 386, 506– 510. Thiel, S., Petersen, S.V., Vorup-Jensen, T., Matsushita, M., Fujita, T., Stover, C.M., Schwaeble, W.J., Jensenius, J.C., 2000. Interaction of C1q and mannan-binding lectin (MBL) with C1r, C1s, MBL-associated serine proteases 1 and 2, and the MBL-associated protein MAp19. J. Immunol. 165, 878– 887. Vorup-Jensen, T., Petersen, S.V., Hansen, A.G., Poulsen, K., Schwaeble, W., Sim, R.B., Reid, K.B.M., Davis, S.J., Thiel, S., Jensenius, J.C., 2000. Distinct pathways of mannan-binding lectin (MBL)- and C1-complex autoactivation revealed by reconstitution of MBL with recombinant MBL-associated serine protease-2. J. Immunol. 165, 2093– 2100. Wong, N.K.H., Kojima, M., Dobó, J., Ambrus, G., Sim, R.B., 1999. Activities of the MBL-associated serine proteases (MASPs) and their regulation by natural inhibitors. Mol. Immunol. 36, 853– 861. Ziccardi, R.J., Cooper, N.R., 1976. Physicochemical and functional characterization of the C1r subunit of the first complement component. J. Immunol. 116, 496– 503. Ziccardi, R.J., Cooper, N.R., 1979. Active disassembly of the first complement component, C1, by C1 inactivator. J. Immunol. 123, 788– 792.