Profiling the Enzymatic Properties and Inhibition of Human

Home , Complement factor B, Factor D, Trypsin 1

Profiling the Enzymatic Properties and Inhibition of Human Complement Factor B*□S Received for publication, July 10, 2007, and in revised form, September 25, 2007 Published, JBC Papers in Press, October 5, 2007, DOI 10.1074/jbc.M705646200 Giang Thanh Le, Giovanni Abbenante, and David P. Fairlie1 From the Centre for Drug Design and Development, Institute for Molecular Bioscience, University of Queensland, Brisbane, Queensland 4072, Australia

Human complement factor B is the crucial catalytic compo- three-complement control protein (CCP) domain, a von Will- nent of the C3 convertase enzyme that activates the alternative ebrand factor type A (vWFA)2 module, and a serine protease pathway of complement-mediated immunity. Although a serine domain (Fig. 1). Following binding to C3b, factor B is recog- protease in its own right, factor B circulates in human serum as nized and hydrolyzed between its CCP and vWFA domains by an inactive zymogen and there is a crystal structure only for the factor D, leading to formation of the heterodimer complex inactive state of factor B and various fragments. To provide C3bBb, the C3 convertase of the complement alternative greater insight to the catalytic function and properties of fac- pathway. tor B, we have used short para-nitroanilide derivatives of 4- Recent crystal structures of the serine protease domain (3), Downloaded from to 15-residue peptides as substrates to profile the catalytic the vWFA module (4), a modified Bb (5), and factor B itself (6) properties of factor B. Among factors found to influence cat- have revealed that factor B adopts a chymotrypsin-like fold with alytic activity of factor B was an unusual dependence on pH. several novel structural motifs. The catalytic triad residues Non-physiological alkaline conditions strongly promoted sub- Asp551, His501, and Ser674 were in a typical active conformation

strate cleavage by factor B, consistent with a pH-accessible con- found in serine proteases; however, residues that normally http://www.jbc.org/ formation of the enzyme that may be critical for catalytic func- define the oxyanion hole produced an unusual conformation tion. Small N-terminal extensions to conventional hexapeptide because of the inward orientation of the carbonyl group of para-nitroanilide substrates significantly increased catalytic Arg671 hydrogen-bonded to the NH group of Ser674 to form a activity of factor B, which was more selective for its cleavage site 310 helix. This abnormal conformation is believed to be respon- than trypsin. The new chromogenic assay enabled optimiza- sible for the zymogen property of factor B under physiological tion of catalysis conditions, the profiling of different sub- conditions. The vWFA domain contains an integrin-like at UQ Library on October 9, 2016 strate sequences, and the development of the first reversible MIDAS (metal ion-dependent adhesion site) motif that may and competitive substrate-based inhibitor of factor B. The serve as a binding site for C3b, consistent with the requirement inhibitor was also shown to prevent in vitro formation of C3a for Mg2ϩ to form C3bBb from C3 and factor B. from C3 by factor B, by synthetic and by natural C3 conver- In its native state, factor B has negligible proteolytic activity tase of the alternative complement activation pathway, and to on its native substrate protein C3 at physiological pH. When block formation of membrane attack complex. The availabil- factor B is bound to C3b and processed by factor D (Fig. 2), the ity of a reversible substrate-based inhibitor that could stabi- resulting C3bBb is an efficient and specific serine protease that ␮ Ϫ1 lize the active conformation of factor B, in conjunction with a cleaves C3 (Km 5.9 M, kcat 1.78 s ) (7). However, upon irre- pH-promoted higher processing activity, may offer a new ave- versible dissociation, the catalytic domain Bb retains only nue to obtain crystal structures of factor B and C3 convertase diminished (1%) proteolytic activity (8). Despite having an inac- in an active conformation. tive zymogen-like oxyanion conformation, factor B still exhibits some esterolytic activity against thiobenzyl ester substrates, including dipeptide analogue Cbz-KR-SBzl, for which the high- Ϫ1 Ϫ1 Factor B (B) is a crucial component of the innate immune est catalytic efficiency has been reported (kcat/Km 1370 M s ) system, contributing to the formation of a multicomponent (9). Under the same conditions, the isolated Bb domain dis- complement protease C3 convertase (C3bBb) that plays a vital played slightly improved catalytic efficiency (kcat/Km 2320 Ϫ1 Ϫ1 role in triggering complement-mediated immune responses to M s ) due mainly to a lower Km (0.58 versus 1.19 mM). Nei- infection and injury (1, 2). Factor B circulates in human serum ther a factor B-specific assay nor any other substrates have been in an inactive zymogen conformation (1, 2), itself consisting of a reported to date. Recently, our understanding about the precise role of factor * This work was supported in part by the National Health and Medical B in the alternative pathway has expanded through the avail- Research Council of Australia and by an Australian Research Council fel- ability of gene-depleted mice and monoclonal antibodies (10, lowship (to D. P. F.). The costs of publication of this article were defrayed in 11). There is now growing evidence for an essential role of fac- part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to tor B in activating the alternative complement pathway in mod- indicate this fact. els of numerous inflammatory diseases (10, 11). Studies using □S The on-line version of this article (available at http://www.jbc.org) contains supplemental Figs. S1–S9. 1 Recipient of an Australian Research Council fellowship. To whom correspond- 2 The abbreviations used are: vWFA, von Willebrand factor type A; pNA, para- ence should be addressed. Fax: 61-733462990; E-mail: [email protected]. nitroanilide; PBS, phosphate-buffered saline.

NOVEMBER 30, 2007•VOLUME 282•NUMBER 48 JOURNAL OF BIOLOGICAL CHEMISTRY 34809 Enzymatic Properties of Human Complement Factor B

nephritis (11, 21–25). When factor B was neutralized with a specific monoclonal antibody, mice were significantly protected from complement activation and fetal loss stimulated by antiphospholipid antibody (11, 26). Human complement receptor of the immunoglobulin superfamily (CRIg), a factor B-specific inhibitor, also reversed inflammation and bone loss in two mouse models of arthritis (27). To date, no specific small molecule inhibitors of factor B have been described, and only a few moderately active and non-selective inhibitors of C3/C5 convertase are known, such as FIGURE 1. Schematic representation of factor B showing components ␮ FUT175 (IC50,4 M) (28) and the allosteric inhibitor compsta- CCP1–3 (three-complement control proteins), vWFA (von Willebrand ␮ factor type A), and SP (serine protease). tin (IC50,12 M) (29). The former compound (nafamostat) is in clinical use in Japan to treat acute pancreatitis and glomerulo- nephritis, but it also inhibits a wide range of targets, including many human serine proteases (30, 31). A major problem in developing inhibitors of C3 convertase has been the difficulty in readily measuring C3 convertase activity. Additionally, substrate-based analogues have so far failed to inhibit processing of

C3 by C3 convertase (30–32). The availability of small molecule Downloaded from inhibitors of factor B and C3 convertase could be valuable for probing physiological and patho-physiological roles of factor B in vivo and for evaluating these enzymes as possible therapeutic targets for new complement-based drugs. Here we report the

development of a valuable new rapid chromogenic assay for http://www.jbc.org/ measuring factor B activity in vitro, the discovery of some important and unexpected molecular and catalytic properties of factor B for substrate processing, and the discovery of the first substrate-based inhibitor of factor B that also inhibits C3 convertase. These results provide a valuable beginning to better at UQ Library on October 9, 2016 understanding the activity of factor B and C3 convertase, to a more robust method of screening potential inhibitors, and to the rational development of potent and selective inhibitors of FIGURE 2. Complement activation paths. Activation of the classical/lectin C3 cleavage resulting from complement activation. pathways results in assembly of C4b2a whereas activation of the alternative pathway leads to formation of C3bBb, both of which may be C3 convertases. The cleavage product C3b of either pathway can bind to factor B, generating EXPERIMENTAL PROCEDURES more C3bB and amplifying the alternative pathway of complement Materials—Factor B, factor D, C3, and cobra venom factor activation. were obtained from Calbiochem, or Quidel. Protein molecular weight markers for gels were purchased from Sigma. Fmoc wild type, factor B-deficient, or C4-deficient mice in conjunc- (N-(9-fluorenyl)methoxycarbonyl)-protected amino acids were tion with specific anti-factor B monoclonal antibodies have from Novabiochem. Gel ingredients, casting apparatus, and demonstrated that factor B is critical for allergen-induced resolving chamber (Mini-PROTEAN Tetra CellTM) were from development of airway hyper-responsiveness and inflamma- Bio-Rad. Proteins were visualized using colloidal Coomassie tion (11, 12). Factor B deficiency has been reported to amelio- Brilliant Blue staining solution. All other chemical reagents rate K/BxN serum transfer and collagen-induced models of were analytical grade from Sigma. rheumatoid arthritis (11, 13). As an indispensable component Para-nitroanilide (pNA) Substrate Synthesis—pNA sub- of the alternative pathway, factor B has also been implicated in strates were synthesized according to the general method of inflammatory disorders of renal tissue injury (11, 14–16). A Abbenante et al. (33) and characterized by analytical high per- functionally intact alternative, but not classical, complement formance liquid chromatography and mass and NMR spectros- pathway was demonstrated to be crucial in renal ischemia- copy (see supplemental material). reperfusion injury, whereas both pathways are required for Enzymatic Characterization and Substrate Analysis—Factor development of intestinal ischemia/reperfusion tissue damage B was assayed against short peptide substrates corresponding (11, 17–19). Brain ischemia results in complement activation to the relevant cleavage site in the endogenous substrate C3 but with C3 levels that are significantly higher in infarcted than in with a chromogenic pNA group in the P1Ј-position at the C non-injured brain (20). Both C3 and neuronal ischemia are terminus. Cleavage of pNA from the peptides by factor B pro- attenuated by intravenous immunoglobulin (IgG) that scav- duces a yellow color that enables continuous monitoring of enges complement fragments (20). The alternative pathway absorbance changes at the wavelength 405 nm. The assay was also proved to be important in kidney inflammation such as conducted in a 96-well plate, with a final reaction volume of 150 lupus nephritis and type II membranoproliferative glomerulo- ␮l containing factor B at concentrations of 25, 50, or 75 nM

34810 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282•NUMBER 48•NOVEMBER 30, 2007 Enzymatic Properties of Human Complement Factor B

FIGURE 3. Factor B cleaves C3 at high pH. C3 (10 ␮g) and factor B (4 ␮g) were incubated at 37 °C in 32 ␮l of buffer at the indicated pH (154 mM NaCl in all buffers) for 3 h. a, top region of 7% gel, loading of 2.5 ␮g of C3/well. b, bottom region of 12% gel, loading of 5.0 ␮g of C3/well. Lane 1, molecular mass mark- FIGURE 4. Hydrolysis of decapeptide Ac-ARASHLGLAR-pNA (1) by factor B ers (SigmaMarker; Sigma). Lane 2, C3 showing ␣ (112 kDa) and ␤ (68 kDa) or trypsin. Substrate (1 mM) was incubated with factor B (50 nM)atpH9.5or chains. Lane 3, factor B (top) and C3a (bottom). Lanes 4–7, C3 plus factor B (pH trypsin (50 nM) at pH 8.0, 37 °C for the indicated time periods. The samples 7.4 (PBS), 8.3 (Tris), 9.3 (glycine), 10.2 (glycine) showing pH-dependent forma- were filtered through a size exclusion filter with 10-kDa cut-off and analyzed Downloaded from tion of C3a after3hat37°C.Lane 8, C3 plus C3 convertase (pre-formed from by analytical high performance liquid chromatography. Identities of products cobra venom factor, factor B, and factor D in PBS). Full gels of panels a and were confirmed by liquid chromatography-mass spectrometry. b are available in supplemental Fig. S1, a and b, respectively. using semi-optimized conditions (100 mM glycine-HCl buffer, non-reducing or reducing conditions, respectively. The ability pH 9.5, 154 mM NaCl, 1% Me2SO). Eight different substrate of factor B to process C3 at alkaline pH is independent of the http://www.jbc.org/ ϩ ϩ concentrations, each in duplicate, were used for determining concentrations of Na ,Mg2 , or the metal ion scavenger kinetic constants. After preincubation in separate wells (5 min, EDTA (supplemental Fig. S2, a and b). 37 °C), catalysis was initiated by mixing substrate with enzyme Because these results indicated that hydrolytic activity of buffer solution with automatic shaking for 5 s. The optical den- factor B proportionally increased with pH, we also examined sity was measured at 405 nm every 30–60 s for 30 to 120 min the ability of factor B alone to hydrolyze synthetic peptides at UQ Library on October 9, 2016 (depending on activity) in a Fluostar Optima (BMG Labtech) of varying length, corresponding to the native cleavage reader, and the average change in millioptical density/min was sequence in C3, at pH 7.4 or 9.5. We found that factor B calculated. Kinetic parameters were calculated from weighted hydrolyzed the 15-residue synthetic peptide substrate Ac- 2 nonlinear regression of the initial velocities as a function of the QHARASHLGLAR SNL-NH2, corresponding to the native eight substrate concentrations using GraphPad Prism 4 soft- sequence at the cleavage site in C3, significantly faster at pH 9.5 ware. The parameters kcat, Km, and kcat/Km were calculated than at pH 7.4. The enzyme only recognized the native cleavage ϭ 2 assuming Michaelis-Menten equilibrium kinetics, v Vmax[S]/ site LAR SNL with no hydrolysis detected at the non-native ϩ 2 ([S] Km). Duplicate measurements were taken for each data Ac-QHAR AS site. By contrast, trypsin cleaved the same pep- point, and means Ϯ S.E. are reported. tide at both sites (data not shown). Similar observations were made for the corresponding but shorter 10-residue peptide RESULTS Ac-ARASHLGLAR-pNA (1), in which the prime side tripep- Factor B Cleaves C3 at Alkaline pH—Factor B circulates in tide component was replaced by the chromogenic para-ni- serum intact and is known to be inactive against its native sub- troaniline (Fig. 4). At pH 9.5, factor B only catalyzed hydrol- strate C3 at pH 7.4. We now report that factor B alone does in ysis at the native cleavage site LAR2pNA, with negligible fact cleave C3 fairly rapidly but only at higher pH (pH Ͼ 9) and hydrolysis detected even after 24 h at the non-native with moderate efficiency. For example, Fig. 3 shows a mini-gel Ac-AR2ASH site. Trypsin cleaved the same peptide at both analysis indicating that incubating factor B with C3 in buffers of sites at pH 8 (Fig. 4). varying pH Ͼ 8 at 37 °C for 3 h results in formation of a protein Development of a Chromogenic Assay for Factor B—Toward corresponding to C3a (lane 3), verified by a commercial sample the development of a rapid chromogenic assay for monitoring ([MϩH]ϩ 9084 Da). A band for factor B at 97 kDa remains factor B activity in vitro, a library of systematically shortened intact at all pH values examined (lanes 4–7), but in the presence pNA substrates was prepared to examine effects of peptide of C3 another band at ϳ100 kDa is also produced at pH 9.3 and sequence and length. For comparison, we also included Cbz- 10.2 (lanes 5–7), corresponding to the cleavage product of the ␣ KR-pNA, which has the non-prime site identical to that of the chain of C3. These new protein bands are identical to those best reported thiobenzyl ester substrate (9). Preliminary results obtained (Fig. 3) from cleavage (pH 7.4, 37 °C) of C3 by C3 indicated that octapeptide Ac-ASHLGLAR-pNA (3) was one of convertase (lane 8) generated by mixing cobra venom factor, the best substrates, so we used it to optimize the conditions of factor B, and factor D (32). The C3 samples used (Ͼ98%) (Cal- the chromogenic assay described below. biochem) have an apparent molecular mass of 180 kDa, appear- Effect of pH on Enzymatic Activity—The catalytic activity of ing on SDS-PAGE gels as a single band or two bands under factor B against octapeptide Ac-ASHLGLAR-pNA (3) was

NOVEMBER 30, 2007•VOLUME 282•NUMBER 48 JOURNAL OF BIOLOGICAL CHEMISTRY 34811 Enzymatic Properties of Human Complement Factor B

FIGURE 6. Hydrolysis of octapeptide substrate (3). Ac-ASHLGLAR-pNA (3) FIGURE 5. Dependence on pH of hydrolysis of Ac-ASHLGLAR-pNA (3) by for different concentrations of factor B enzyme (25, 50, 75 nM) and pNA sub- factor B. Ac-ASHLGLAR-pNA (1 mM) was incubated with 75 nM factor B at pH strate (0.1, 0.2, 0.4, 0.8, 1.2, 1.6, 2.0, 3.0 mM) at pH 9.5, 37 °C. Initial velocity 7.5, 8.5, 9.0, 9.25, 9.5, 9.75, 10.0, 10.25, 10.5, 10.75, 11.0, 11.25, and 11.5. Initial (␮M/s) was calculated as amount of free pNA released from increase in mil- velocity (␮M/s) was calculated as amount of free pNA released from increase lioptical density/s (at 405 nm). in millioptical density/s (at 405 nm). All buffers contained 154 mM NaCl. Downloaded from investigated over a wide pH range, using a variety of buffers (Tris-HCl, pH 7.5–8.50; glycine-HCl, pH 9.00–10.50; piperi- dine-HCl, pH 10.75–11.5). The optimal pH for most rapid processing of this substrate by factor B was ϳ10.5 (Fig. 5). Auto- hydrolysis of this substrate was negligible at or below pH 11.5 in http://www.jbc.org/ the absence of factor B, but the catalytic activity decreased abruptly above pH 10.5, presumably due to denaturation of factor B. The effect of pH on the catalytic activity of factor B was ϩ independent of Na (154 mM NaCl, Fig. 5; 0 mM NaCl, supplemental Fig. S3). at UQ Library on October 9, 2016 Effect of Divalent Cations, Ion Strength, and Other Additives— Effects of diluents were probed at pH 9.5. Although the vWFA domain of factor B contains an integrin-like MIDAS (metal ion- FIGURE 7. Effect of substrate length on factor B catalysis. Substrates are Ac- ARASHLGLAR-pNA (1, 10-mer), Ac-RASHLGLAR-pNA (2, 9-mer), Ac-ASHLGLAR- dependent adhesion site) motif that could bind divalent cations, pNA (3, 8-mer), Ac-SHLGLAR-pNA (4, 7-mer), Ac-HLGLAR-pNA (6-mer), Ac- neither calcium (10 mM), magnesium (10 mM), nor the divalent LGLAR-pNA (5-mer), Ac-GLAR-pNA (4-mer), and Cbz-KR-pNA. Substrates (1 mM) were incubated with 50 nM factor B in glycine-HCl buffer, pH 9.5, at cation-sequestering agent EDTA (10 mM) detectably influ- 37 °C. Change in optical density at 405 nm in the first 5–10 min was used to enced the activity of factor B in vitro. Similarly, neither varia- calculate the initial velocities. tion of ionic strength (NaCl 0–250 mM) nor addition of bovine serum albumin (1%), polyethylene glycol (1%), or glycerol (10%) TABLE 1 had any effect on the substrate-processing ability of factor B. Kinetic parameters for pNA peptide substrates Km and kcat were determined using initial velocities for selected 7–10-residue sub- Under the semi-optimal conditions (pH 9.5, 100 mM glycine- strates at concentrations 0.2, 0.6, 1.0, 2.0, 4.0, 6.0, 8.0, 10.0 mM. M HCl buffer, 154 m NaCl), we found that factor B exhibited Substrate kcat Km kcat/Km Ϫ1 classic Michaelis-Menten kinetics against the pNA substrate 3, s mM Ϫ Ϯ 1 Ϯ a a a with kcat 10.19 0.19 s , Km 2.19 0.08 mM, and kcat/Km Ac-ARASHLGLAR-pNA (1) 12.6 12.4 1016 Ϫ1 Ϫ1 2 a a a 4666 Ϯ 253 M s (Fig. 6). Ac-RASHLGLAR-pNA ( ) 12.8 15.8 810 Ac-ASHLGLAR-pNA (3) 10.19 Ϯ 0.19 2.19 Ϯ 0.08 4666 Ϯ 253 Effects of Substrate Length—A number of pNA peptide sub- Ac-SHLGLAR-pNA (4) 8.38 Ϯ 1.4 1.05 Ϯ 0.14 8315 Ϯ 2458 strates were synthesized, ranging from 4 to 10 amino acids cor- Ac-AHLGLAR-pNA (5) 8.20 Ϯ 0.02 2.84 Ϯ 0.05 2884 Ϯ 54 Ac-KHLGLAR-pNA (6) 8.25 Ϯ 0.22 2.79 Ϯ 0.5 3045 Ϯ 467 responding to the native cleavage sequence of C3. Because Cbz- a Values estimated from a single experiment at substrate concentrations of 0.1, 0.2, KR-SBzl has been reported to be the best thioester substrate for 0.4, 0.8, 1.2, 1.6, 2.0, 3.0 mM. factor B (9), its corresponding pNA analogue, Cbz-KR-pNA, was also prepared for comparison. this substrate was examined using factor B obtained from a Among eight putative substrates examined (Fig. 7), the 8- different supplier (Quidel), comparable catalytic efficiency was Ϯ Ϯ Ϫ1 Ϯ and 7-residue pNA peptides (3 and 4) proved to be the best observed (Km 1.15 0.21 mM, kcat 9.1 0.6s , kcat/Km 8258 Ϫ1 Ϫ1 Ϫ1 substrates, with initial velocity of 0.17–0.19 ␮Ms (50 nM fac- 2070 M s ). tor B, pH 9.5, (substrate) ϭ 1mM). The Cbz-KR-pNA substrate Effect of the 7th Residue in Substrates—The 7th residue serine displayed Ͼ10-fold lower initial velocity under identical condi- in the heptapeptide substrate Ac-SHLGLAR-pNA (4) was var- tions. Km and kcat values for selected substrates are recorded in ied to probe substrate requirements at this position. Large Ϫ1 Table 1. Heptapeptide substrate 4 (Km 8.4 mM, kcat 1s , hydrophobic or negatively charged side chains at this position kcat/Km 8315) was processed most efficiently by factor B. When significantly decreased hydrolysis by factor B (Fig. 8). Of the

34812 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282•NUMBER 48•NOVEMBER 30, 2007 Enzymatic Properties of Human Complement Factor B

FIGURE 8. Initial velocities for factor B processing of heptapeptide-pNA substrates. Substrates varied in the 7th residue: Ac-AHLGLAR-pNA (5, Ala), FIGURE 10. Effect of substrate concentration on inhibition by 7. Substrate Ac-LHLGLAR-pNA (Leu), Ac-FHLGLAR-pNA (Phe), Ac-DHLGLAR-pNA (Asp), Ac- 4 at concentrations of 0.5, 2.5, 5, or 10 mM was incubated with 50 nM factor B in the presence of 100 ␮M inhibitor 7, 100 mM glycine-HCl buffer, pH 9.5, at KHLGLAR-pNA (6, Lys), Ac-Isox-HLGLAR-pNA (isoxazole-5-carboxylic acid, ␭ ϭ Isox), Ac-Dapa-HLGLAR-pNA (L-diaminopropionic acid, Dapa), Ac-SHLGLAR- 37 °C. Increases in absorbance were monitored at 405 nm. pNA (4, Ser). Substrates (1 mM) were incubated with 50 nM factor B in glycine-

HCl buffer, pH 9.5, at 37 °C. Change in optical density at 405 nm in the first Downloaded from 5–10 min was used to calculate initial velocities. http://www.jbc.org/ at UQ Library on October 9, 2016

FIGURE 9. Dose-dependent inhibition of factor B by compound 7. Condi- FIGURE 11. Compound 7 inhibits cleavage of C3 by factor B in a concentration-dependent manner. C3 (10 ␮g) and factor B (4 ␮g) were incubated tions were 100 mM glycine-HCl buffer, pH 9.5, 37 °C, 50 nM factor B, 0.5 mM ϳ ␮ at 37 °C in 32 ␮l of 100 mM glycine-HCl buffer, pH 10.2, with various concen- pNA substrate 4, eleven concentrations (from 1 M to1mM)of7. Positive ␮ control was buffer alone (not shown). trations of inhibitor 7. a, top region of 7% gel, loading of 2.5 g of C3/well. b, bottom region of 12% gel, loading of 5.0 ␮g of C3/well. Lane 1, molecular mass markers (SigmaMarker; Sigma). Lane 2, C3a. Lane 3, C3 showing ␣ (112 kDa) and ␤ (68 kDa) chains. Lane 4, factor B. Lanes 5–9, C3, factor B plus buffer small number of heptapeptide substrates investigated (Fig. 8), (lane 5) or inhibitor (lanes 6–9, 0.01, 0.1, 0.25, 0.5 mM compound 7) showing compound 4 with Ser at position 7 proved to be the best. A serine inhibition of C3 cleavage to C3a after3hat37°C.Full gels of panels a and b are mimetic, isoxazole-5-carbonyl, was poorly tolerated, while small available in supplemental Fig. S4, a and b, respectively. side chains (Ala) or positively charged moieties (Lys, Dapa (diaminopropionic acid)) caused small decreases in initial velocities was also inhibited by aldehyde 7 in a concentration-dependent (Fig. 8). manner (Fig. 11). Because factor B also provides the crucial Substrate-based Inhibitor of Factor B and C3 Convertase—To catalytic unit within C3 convertase of the alternative pathway, convert substrate 4 to a putative inhibitor, the pNA moiety was compound 7 was also examined for inhibition of C3 convertase. replaced by an aldehyde group to furnish compound 7 (Ac- Cleavage of C3 by C3 convertase generated either from a com- SHLGLAR-H). This compound proved to be a moderately bination of cobra venom factor, factor B, and factor D (CVF.Bb; ␮ potent inhibitor of factor B (IC50 19 M, Fig. 9), blocking the Fig. 12 and supplemental Fig. S5, a and b) or from human C3, cleavage of C3 in a concentration-dependent manner under the factor B, and factor D (hC3bBb, supplemental Fig. 6, a and b)in same assay conditions (Fig. 11, supplemental Fig. S4, a and b). the presence of Mg2ϩ was indeed inhibited by aldehyde 7 in a In this inhibitor, the aldehyde functional group is capable of dose-dependent manner. Under the same conditions, general forming a reversible covalent bond with the catalytic Ser side serine protease inhibitors leupeptin or phenylmethylsulfonyl chain, resulting in a tetrahedral configuration that mimics the fluoride at 1-mM concentrations did not inhibit C3 cleavage transition state. (supplemental Fig. S6, a and b). Compound 7 is a reversible and competitive inhibitor of factor B, since increasing concentrations of substrate 4 overcame DISCUSSION the inhibition of factor B caused by 100 ␮M 7 (Fig. 10). Cleavage Factor B is currently considered to be an inactive zymogen of the native substrate C3 to C3a and C3b by factor B at pH 10.2 serine protease involved in the alternative pathway of comple-

NOVEMBER 30, 2007•VOLUME 282•NUMBER 48 JOURNAL OF BIOLOGICAL CHEMISTRY 34813 Enzymatic Properties of Human Complement Factor B

NMR and x-ray crystal structures for an inactive state of factor B, it has been proposed that activation of factor B likely pro- ceeds through multiple conformational changes (1, 2, 6, 36) induced by binding to C3b and then through cleavage by factor D (1, 2, 6, 36). On the basis of our results it is conceivable that, at alkaline pH, factor B changes its protonation state and adopts an active conformation that can more effectively recognize and cleave C3. Other serine proteases are known to be activated under alkaline conditions. For example, NS2B-NS3 serine proteases from flaviviruses (e.g. West Nile, dengue, yellow fever) (37–42) become substantially active in vitro only at alkaline pH (Յ10.5). Those enzymes, like factor B, are thought to interact with other protein domains and to deposit on membrane surfaces under FIGURE 12. Compound 7 inhibits cleavage of C3 by C3 convertase native conditions for proteolytic processing. Human kallikrein (CVF.Bb) in a concentration-dependent manner. a, top region of 7% gel, loading of 2.5 ␮g of C3/well. b, bottom region of 12% gel, loading of 5.0 ␮gof 6 has been reported to exhibit significantly higher catalytic C3/well. C3 convertase was generated by incubating 12 ␮g of cobra venom activity at pH 9.0 than at pH 7.5 (43); the basis for such pH factor, 8 ␮g of factor B, and 0.2 ␮g of factor D in 100 ␮l of PBS, pH 7.4, con- ␮ dependence is not well understood at present. taining 1 mM MgCl2 for1hat37°C,andthen 100 l of PBS, pH 7.4, containing 20 mM EDTA was added. C3 (10 ␮g) and the C3 convertase (6 ␮l) were incu- The catalytic activity of factor B against small pNA substrates Downloaded from bated at 37 °C for 30 min in final 32 ␮l of PBS, pH 7.4, with buffer or various was also found here to exhibit a strong pH dependence. The inhibitor concentrations. Lane 1, molecular mass markers (SigmaMarker; Sigma). Lane 2, C3 showing ␣ (112 kDa) and ␤ (68 kDa) chains. Lane 3, C3a. hydrolysis of pNA substrates by factor B followed classic Lanes 4–8, C3, C3 convertase plus buffer (lane 4) or inhibitor (lanes 5–8, 0.01, Michaelis-Menten kinetics. The magnitude of kinetic parame- 0.1, 0.25, 0.5 mM) showing inhibition of C3a formation after 37 °C for 30 min. ters (Table 1) for our pNA substrates supports the idea that Full gels of panels a and b are available in supplemental Fig. S5, a and b, respectively. factor B adopts an active conformation at higher pH, since all http://www.jbc.org/ Ͼ Ϫ1 substrates are processed quite fast kcat ( 8s ), comparable ment activation (1, 2). Recent studies using knock-out mice and with the catalytic hydrolysis of C3 by C3 convertase (kcat 1.78 Ϫ1 a specific antibody have demonstrated important contributions s ). On the other hand, kcat/Km values for the pNA substrates from factor B to a number of inflammatory disorders (10, 11), are only modest (Table 1), due to a high Km (low mM). This such as airway hyper-responsiveness (12), lupus nephritis (34), suggests a much lower substrate affinity for the pNA substrates ␮ at UQ Library on October 9, 2016 and antiphospholipid syndrome (26, 35), and have provided than for C3 (Km 5.9 M for C3 processing by C3 convertase). insights to the role of factor B in complement activation via the Strikingly, factor B prefers longer substrates than other ser- alternative pathway. The new data from those studies suggest ine proteases. We found that pNA substrates with short peptide that factor B may be a potential therapeutic target in its own sequences (Ͻ7 residues) were very poor substrates for factor B, right. Moreover, selective inhibitors of factor B (none is known) unlike other serine proteases that tend to recognize shorter could potentially be valuable new molecular probes for investi- segments of polypeptide substrates and inhibitor analogues (30, gating the biology of factor B and its homologue C3 convertase, 44, 45). Aminomethyl courmarin substrates of Յ6 residues as well as providing clues to the development of novel thera- have been reported not to be processed effectively by C3 con- peutics that might regulate the alternative pathway of comple- vertase (32). Substrates with 7 or 8 residues, in contrast to ment activation (10, 11). To date, only a couple of weakly potent shorter peptides, were found here to be readily processed by and non-selective inhibitors of C3 convertase are known (28, factor B. Trypsin and thrombin, which recognize shorter sub- 29), but they are not substrate analogues and have been of little strate sequences, still process the heptapeptide AcSHLGLAR- use for understanding C3 cleavage relevant to complement pNA and are inhibited by the corresponding AcSHLGLAR-al- activation. No protease inhibitors for factor B are known, and dehyde (supplemental Fig. S8, a and b). However, trypsin can there is no chromogenic assay available for monitoring enzyme process AcARASHLGLARSNL-NH2, corresponding to the activity for either factor B or C3 convertase in vitro. native sequence at the cleavage site in C3, at both the AR2A Factor B is known to be inactive against its native substrate and AR2S sites, whereas factor B only cleaved at the AR2S C3 at pH 7.4 (1, 2), and this may be the reason why factor B site and only effectively in sequences with 7 or more residues on function has not been heavily studied. Experiments reported the N-terminal side. The crystal structure of factor B in its inac- here have shown that factor B alone can hydrolyze C3 under tive state (3, 6) shows shallow putative binding pockets at S1-S6 mild to strong alkaline conditions. Importantly, the cleavage in the enzyme, consistent with the need for some additional products from factor B-mediated hydrolysis of C3 at pH 10.2 binding residues attached to hexapeptide substrates. However, are identical to those produced by cobra venom factor-contain- the unusual requirement of small P7 (Ser) and P8 (Ala) residues ing C3 convertase at pH 7.4, as demonstrated by gel analysis for enhancement of substrate processing was surprising. Using (Fig. 3). The significance of these new findings of pH-induced NMR spectroscopy, we did not detect any conformational activation of factor B may be important in understanding catal- changes in water going from 6- to 7- residue peptides, which 3 ysis by factor B. The structure of the active conformation of were both random coil structures ( JNH-CH␣ coupling constants factor B remains elusive, as it is stable only in complex with C3b for 6-mer were 6.8, 5.8, 7.5, 5.9, 6.6, 6.8 Hz and for 7-mer were or the C3b surrogate, cobra venom factor. On the basis of recent 6.7, 8.0, 5.9, 7.2, 5.9, 6.1 Hz), so it would appear that the sub-

34814 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282•NUMBER 48•NOVEMBER 30, 2007 Enzymatic Properties of Human Complement Factor B

strate extensions cause higher receptor affinity (note lower Km the inhibitor is competitive and reversible and thus directed at values) because of their side chains rather than a conforma- the catalytic substrate-binding active site of both enzymes. tional change to the substrate. These results for this new assay, developed for factor B but also The substrate-processing results may reflect that enzymatic validated here for native and synthetic forms of C3 convertase processing of C3 is only slightly dependent on affinity or selec- from the alternative pathway, suggest that potent substrate- tivity for substrate residues immediately surrounding the cleav- based inhibitors of these serine proteases may now be achieva- age site. Indeed, the sequence of C3 at the cleavage site is not ble. Development of such inhibitors into even more potent, greatly different from that of substrates for other serine pro- selective, and more drug-like compounds may facilitate ratio- teases. Instead, the selectivity for processing of C3 during com- nal intervention in the alternative pathway of complement acti- plement activation is probably heavily influenced by the multi- vation and the blockade of membrane attack complex ple domains of C3 convertase, evolved for selective recognition formation. of this particular substrate under different complement activation conditions produced by different environmental stimuli. It REFERENCES has been estimated that ϳ40% of proteases are multidomain 1. Xu, Y. Y., Narayana, S. V. L., and Volanakis, J. E. (2001) Immunol. Rev. 180, enzymes (46), and many of those appear to use the appended 123–135 2. Sim, R. B., and Tsiftsoglou, S. A. (2004) Biochem. Soc. Trans. 32, 21–27 domains for selective substrate recognition. 3. Jing, H., Xu, Y. Y., Carson, M., Moore, D., Macon, K. J., Volanakis, J. E., and It has been proposed that C3 binds C3 convertase through Narayana, S. V. L. (2000) EMBO J. 19, 164–173 several docking sites on the C3b component of C3bBb complex, 4. Bhattacharya, A. A., Lupher, M. L., Staunton, D. E., and Liddington, R. C.

which in turn interacts with Bb to induce the active conforma- (2004) Structure 12, 371–378 Downloaded from tion (5). In a simple model of binding between C3 and Bb, the 5. Ponnuraj, K., Xu, Y. Y., Macon, K., Moore, D., Volanakis, J. E., and Naray- 14, positively charged loop 2 of Bb could possibly interact with the ana, S. V. L. (2004) Mol. Cell 17–28 6. Milder, F. J., Gomes, L., Schouten, A., Janssen, B. J. C., Huizinga, E. G., negative charges on the anchor region and/or MG8 of C3 (47). Romijn, R. A., Hemrika, W., Roos, A., Daha, M. R., and Gros, P. (2007) Nat. The proposed interactions may confer the relatively high Struct. Mol. Biol. 14, 224–228

␮ http://www.jbc.org/ implied affinity of C3 (Km 5.9 M) for the C3bBb complex, and 7. Pangburn, M. K., and Mullereberhard, H. J. (1986) Biochem. J. 235, thus only C3 convertase (but not Bb) can bind to C3 efficiently 723–730 for hydrolysis to occur. Consistent with this hypothesis, Bb 8. Fishelson, Z., and Muller-Eberhard, H. J. (1984) J. Immunol. 132, 2ϩ 2ϩ 1425–1429 requires the presence of divalent cations (Mg or Ni )to 9. Kam, C. M., McRae, B. J., Harper, J. W., Niemann, M. A., Volanakis, J. E., cleave C3, although with only 1% activity of factor B (8). and Powers, J. C. (1987) J. Biol. Chem. 262, 3444–3451 Until now, rational design of substrate-based C3 convertase 10. Mollnes, T. E., and Kirschfink, M. (2006) Mol. Immunol. 43, 107–121 at UQ Library on October 9, 2016 inhibitors such as ␣-ketoheterocycles has reportedly failed (32), 11. Thurman, J. M., and Holers, V. M. (2006) J. Immunol. 176, 1305–1310 possibly because of the limitation of low affinity of the quite 12. Taube, C., Thurman, J. M., Takeda, K., Joetham, A., Miyahara, N., Carroll, short peptide sequences examined to date. Despite the low M. C., Dakhama, A., Giclas, P. C., Holers, V. M., and Gelfand, E. W. (2006) Proc. Natl. Acad. Sci. U. S. A. 103, 8084–8089 affinity of substrate 4 for factor B, replacing the pNA moiety 13. Ji, H., Ohmura, K., Mahmood, U., Lee, D. M., Hofhuis, F. M. A., with an aldehyde C-terminal cap afforded a moderately Boackle, S. A., Takahashi, K., Holers, V. M., Walport, M., Gerard, C., ␮ active factor B inhibitor 7 (IC50 19 M). This compound also Ezekowitz, A., Carroll, M. C., Brenner, M., Weissleder, R., Verbeek, inhibits the cleavage of C3 at high pH by factor B or at pH 7.4 J. S., Duchatelle, V., Degott, C., Benoist, C., and Mathis, D. (2002) by C3 convertase (generated from cobra venom factor, factor Immunity 16, 157–168 B, and factor D or from human C3, factor B, and factor D) in 14. Bao, L. H., Haas, M., Boackle, S. A., Kraus, D. M., Cunningham, P. N., Park, P., Alexander, J. J., Anderson, R. K., Culhane, K., Holers, V. M., and Quigg, a concentration-dependent manner (Figs. 11 and 12, and R. J. (2002) J. Immunol. 168, 3601–3607 supplemental Figs. S4-S6). Furthermore, aldehyde 7 is able 15. Watanabe, H., Garnier, G., Circolo, A., Wetsel, R. A., Ruiz, P., Holers, to inhibit downstream formation of membrane attack com- V. M., Boackle, S. A., Colten, H. R., and Gilkeson, G. S. (2000) J. Immunol. plex to a comparable extent via the alternative pathway (sup- 164, 786–794 plemental Fig. S7). 16. Wang, Y., Hu, Q. L., Madri, J. A., Rollins, S. A., Chodera, A., and Matis, L. A. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 8563–8568 Access to the active conformation of factor B could be quite 17. Thurman, J. M., Ljubanovic, D., Edelstein, C. L., Gilkeson, G. S., and Hol- valuable for better understanding of the molecular mechanism ers, V. M. (2003) J. Immunol. 170, 1517–1523 by which C3 convertase carries out its catalytic function and for 18. Park, P., Haas, M., Cunningham, P. N., Bao, L. H., Alexander, J. J., and devising more effective therapeutic strategies to intervene in Quigg, R. J. (2002) Am. J. Physiol. 282, F352–F357 prolonged complement activation. The novel results presented 19. Zhou, W. D., Farrar, C. A., Abe, K., Pratt, J. R., Marsh, J. E., Wang, Y., Stahl, above suggest that factor B may adopt an active enzyme con- G. L., and Sacks, S. H. (2000) J. Clin. Investig. 105, 1363–1371 20. Arumugam, T. V., Tang, S. C., Lathia, J. D., Cheng, A., Mughal, M. R., formation at alkaline pH and, in complex with a reversible sub- Chigurupati, S., Magnus, T., Chan, S. L., Jo, D. G., Ouyang, X., Fairlie, D. P., strate-based aldehyde inhibitor, could potentially present the Granger, D. N., Vortmeyer, A., Basta, M., and Mattson, M. P. (2007) Proc. first opportunity to obtain crystal structures for factor B and C3 Natl. Acad. Sci. U. S. A., 104, 14104–14109 convertase in their catalytically active forms. The chromogenic 21. Appel, G. B., Cook, H. T., Hageman, G., Jennette, J. C., Kashgarian, M., assay reported here is quite robust and allows for rapid screen- Kirschfink, M., Lambris, J. D., Lanning, L., Lutz, H. U., Meri, S., Rose, N. R., ing of compounds as putative inhibitors of factor B and C3 Salant, D. J., Sethi, S., Smith, R. J. H., Smoyer, W., Tully, H. F., Tully, S. P., Walker, P., Welsh, M., Wurzner, R., and Zipfel, P. F. (2005) J. Am. Soc. convertase. Compound 7, designed and evaluated using this Nephrol. 16, 1392–1403 assay, inhibits cleavage of C3 by both factor B and C3 conver- 22. Pickering, M. C., Cook, H. T., Warren, J., Bygrave, A. E., Moss, J., Walport, tase in a concentration-dependent manner. This indicates that M. J., and Botto, M. (2002) Nat. Genet. 31, 424–428

NOVEMBER 30, 2007•VOLUME 282•NUMBER 48 JOURNAL OF BIOLOGICAL CHEMISTRY 34815 Enzymatic Properties of Human Complement Factor B

23. Hogasen, K., Jansen, J. H., Mollnes, T. E., Hovdenes, J., and Harboe, M. Rheumatol. 3, 140–147 (1995) J. Clin. Investig. 95, 1054–1061 36. Hinshelwood, J., and Perkins, S. J. (2000) J. Mol. Biol. 301, 1267–1285 24. Meri, S., Koistinen, V., Miettinen, A., Tornroth, T., and Seppala, I. J. T. 37. Peyrefitte, C. N., Pastorino, B. A. M., Grandadam, M., Rolland, D., Tolou, (1992) J. Exp. Med. 175, 939–950 H. J., and Bessaud, M. (2007) Virus Genes 34, 185–191 25. Lenderink, A. M., Liegel, K., Ljubanovic, D., Coleman, K. E., Gilkeson, 38. Li, J., Lim, S. P., Beer, D., Patel, V., Wen, D. Y., Tumanut, C., Tully, D. C., G. S., Holers, V. M., and Thurman, J. M. (2007) Am. J. Physiol. 293, Williams, J. A., Jiricek, J., Priestle, J. P., Harris, J. L., and Vasudevan, S. G. F555–F564 (2005) J. Biol. Chem. 280, 28766–28774 26. Thurman, J. M., Kraus, D. M., Girardi, G., Kang, H. J., Hourcade, D., Royer, 39. Nall, T. A., Chappell, K. J., Stoermer, M. J., Fang, N. X., Tyndall, J. D. A., P. A., Gilkeson, G., Salmon, J. E., and Holers, V. M. (2004) Mol. Immunol. Young, P. R., and Fairlie, D. P. (2004) J. Biol. Chem. 279, 48535–48542 41, 318–318 40. Bessaud, M., Pastorino, B. A. M., Peyrefitte, C. N., Rolland, D., Grand- 27. Katschke, K. J., Helmy, K. Y., Steffek, M., Xi, H. K., Yin, J. P., Lee, W. P., adam, M., and Tolou, H. J. (2006) Virus Res. 120, 79–90 Gribling, P., Barck, K. H., Carano, R. A. D., Taylor, R. E., Rangell, L., Diehl, 41. Bessaud, M., Grard, G., Peyrefitte, C. N., Pastorino, B., Rolland, D., Char- L., Hass, P. E., Wiesmann, C., and Campagne, M. V. (2007) J. Exp. Med. rel, R. N., de Lamballerie, X., and Tolou, H. J. (2005) Virus Res. 107, 57–62 204, 1319–1325 42. Leung, D., Schroder, K., White, H., Fang, N. X., Stoermer, M. J., Ab- 28. Inagi, R., Miyata, T., Maeda, K., Sugiyama, S., Miyama, A., and Nakashima, benante, G., Martin, J. L., Young, P. R., and Fairlie, D. P. (2001) J. Biol. I. (1991) Immunol. Lett. 27, 49–52 Chem. 276, 45762–45771 29. Sahu, A., Kay, B. K., and Lambris, J. D. (1996) J. Immunol. 157, 884–891 30. Leung, D., Abbenante, G., and Fairlie, D. P. (2000) J. Med. Chem. 43, 43. Angelo, P. F., Lima, A. R., Alves, F. M., Blaber, S. I., Scarisbrick, I. A., Blaber, 305–341 M., Juliano, L., and Juliano, M. A. (2006) J. Biol. Chem. 281, 3116–3126 31. Abbenante, G., and Fairlie, D. P. (2005) Med. Chem. 1, 71–104 44. Fairlie, D. P., Tyndall, J. D. A., Reid, R. C., Wong, A. K., Abbenante, G., 32. Furlong, S. T., Dutta, A. S., Coath, M. M., Gormley, J. J., Hubbs, S. J., Lloyd, Scanlon, M. J., March, D. R., Bergman, D. A., Chai, C. L. L., and Burkett, B. A. (2000) J. Med. Chem. 43, 1271–1281 D., Mauger, R. C., Strimpler, A. M., Sylvester, M. A., Scott, C. W., and Downloaded from Edwards, P. D. (2000) Immunopharmacology 48, 199–212 45. Tyndall, J. D. A., Nall, T., and Fairlie, D. P. (2005) Chem. Rev. 105, 973–999 33. Abbenante, G., Leung, D., Bond, T., and Fairlie, D. P. (2000) Lett. Pept. Sci. 46. Puente, X. S., Sanchez, L. M., Overall, C., and Lopez-Otinn, C. (2003) 7, 347–351 Nature Rev. Genet. 4, 544–558 34. Elliott, M. K., Jarmi, T., Ruiz, P., Xu, Y. Y., Holers, V. M., and Gilkeson, 47. Milder, F. J., Raaijmakers, H. C. A., Vandeputte, M., Schouten, A., Huiz- G. S. (2004) Kidney Int. 65, 129–138 inga, E. G., Romijn, R. A., Hemrika, W., Roos, A., Daha, M. R., and Gros, P.

35. Salmon, J. E., Girardi, G., and Lockshin, M. D. (2007) Nat. Clin. Pract. (2006) Structure 14, 1587–1597 http://www.jbc.org/ at UQ Library on October 9, 2016

34816 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282•NUMBER 48•NOVEMBER 30, 2007 Le et al Supplementary Material S1

Figure S1a: pH Dependence of C3 cleavage by Factor B (7% gel analysis). C3 (10 µg) and factor B (4 µg) were incubated at 37°C in 32 µL buffer at the indicated pH for 3h. Loading of 2.5 µg C3/well. Lane 1: molecular mass markers (SigmaMarker, Sigma). Lane 2: C3 showing alpha (112kDa) and beta (68 kDa) chains. Lane 3: factor B. Lanes 4-7: C3 plus Factor B (pH 7.4, 8.3, 9.3, 10.2) showing pH-dependent formation of C3a after 3h at 37°C. Lane 8: C3 plus C3 convertase (pre-formed from cobra venom factor, factor B and factor D). All buffers contain 154 mM NaCl.

Figure S1b: pH Dependence of C3 cleavage by Factor B (12% gel analysis). C3 (10 µg) and factor B (4 µg) were incubated at 37°C in 32 µL buffer at the indicated pH for 3h. Loading of 5.0 µg C3/well. Lane 1: molecular mass markers (SigmaMarker, Sigma). Lane 2: C3 showing alpha (112kDa) and beta (68 kDa) chains. Lane 3: C3a. Lanes 4-7: C3 plus Factor B (pH 7.4, 8.3, 9.3, 10.2) showing pH- dependent formation of C3a after 3h at 37°C. Lane 8: C3 plus C3 convertase (pre-formed from cobra venom factor, factor B and factor D). All buffers contain 154 mM NaCl.

Le et al Supplementary Material S2

Figure S2a: Metal Ion Independence of C3 cleavage by Factor B (7% gel analysis). C3 (10 µg) and factor B (4 µg) were incubated at 37°C in 32 µL of various buffer (all at pH 10.3) for 3h. Loading of 2.5 µg C3/well. Lane 1: molecular mass markers (SigmaMarker, Sigma). Lane 2: factor B. Lane 3: C3 showing alpha (112kDa) and beta (68 kDa) chains. Lanes 4-7: C3 (2.5 µg) and factor B (1 µg). Lane 4: Glycine buffer pH 10.3 plus 154mM NaCl, lane 5: Glycine buffer pH 10.3, no added NaCl, lane 6: Glycine buffer pH 10.3 plus 5mM MgCl2 and no added NaCl, lane 7: Glycine buffer pH 10.3 plus 5mM EDTA and no added NaCl.

Figure S2b: Metal Ion Independence of C3 cleavage by Factor B (12% gel analysis). C3 (10 µg) and factor B (4 µg) were incubated at 37°C in 32 µL of various buffer (all at pH 10.3) for 3h. Loading of 5.0 µg C3/well. Lane 1: molecular mass markers (SigmaMarker, Sigma). Lane 2: factor B. Lane 3: C3 showing alpha (112kDa) and beta (68 kDa) chains. Lanes 4-7: C3 (5.0 µg) and factor B (2 µg). Lane 4: Glycine buffer pH 10.3 plus 154mM NaCl, lane 5: Glycine buffer pH 10.3, and no added NaCl, lane 6: Glycine buffer pH 10.3 plus 2.5mM MgCl2 and no added NaCl, lane 7: Glycine buffer pH 10.3 plus 2.5mM EDTA and no added NaCl.

Le et al Supplementary Material S3

0.35

0.30 Factor B (75nM)

0.25 Buffer only

0.20

V 0.15 (µM/s)

0.10

0.05

0.00

7.5 8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 -0.05 pH

Figure S3: Initial velocity for hydrolysis of Ac-ASHLGLAR-pNA (3) by factor B at varying pH without Na+. Ac-ASHLGLAR-pNA (1mM) was incubated with 75nM factor B at pH 7.5, 8.5, 9.0, 9.25, 9.5, 9.75, 10.0, 10.25, 10.5, 10.75, 11.0, 11.25 and 11.5. Initial velocity (mM/s) was calculated as amount of free pNA released from increase in millioptical density/second (at 405 nm). All buffers contained no added NaCl. Le et al Supplementary Material S4

Figure S4a: Compound 7 inhibits cleavage of C3 by factor B in a concentration dependent manner (7% gel analysis). C3 (10 µg) and factor B (4 µg) were incubated at 37°C in 32 µL, 100mM glycine-HCl buffer, pH 10.2 with various concentrations of inhibitor 7. Loading of 2.5 µg C3/well. Lane 1: molecular mass markers (SigmaMarker, Sigma). Lane 2: C3a. Lane 3: C3 showing alpha (112kDa) and beta (68 kDa) chains. Lane 4: factor B. Lanes 5-9: C3, Factor B plus buffer (lane 5) or inhibitor (lanes 6-9: 0.01, 0.1, 0.25, 0.5 mM compound 7) showing inhibition of C3 cleavage to C3a after 3h at 37°C. All buffers contain 154 mM NaCl.

Figure S4b: Compound 7 inhibits cleavage of C3 by factor B in a concentration dependent manner (12% gel analysis). C3 (10 µg) and factor B (4 µg) were incubated at 37°C in 32 µL, 100mM glycine-HCl buffer, pH 10.2 with various concentrations of inhibitor 7. Loading of 5.0 µg C3/well. Lane 1: molecular mass markers (SigmaMarker, Sigma). Lane 2: C3a. Lane 3: C3 showing alpha (112kDa) and beta (68 kDa) chains. Lane 4: factor B. Lanes 5-9: C3, Factor B plus buffer (lane 5) or inhibitor (lanes 6-9: 0.01, 0.1, 0.25, 0.5 mM compound 7) showing inhibition of C3 cleavage to C3a after 3h at 37°C. All buffers contain 154 mM NaCl. Le et al Supplementary Material S5

Figure S5a: Compound 7 inhibits cleavage of C3 by C3 convertase (CVF.Bb) in a concentration dependent manner (7% gel analysis). Loading of 2.5 µg C3/well. C3 convertase was generated by incubating 12µg of cobra venom factor, 8µg of factor B and 0.2 µg of factor D in 100 µL PBS (pH 7.4) containing 1mM MgCl2 for 1 h at 37°C, then was added 100µL PBS (pH 7.4) containing 20mM EDTA. C3 (10 µg) and the C3 convertase (6 µl) were incubated at 37°C for 30 min. in final 32 µL PBS (pH 7.4) with buffer or various inhibitors concentrations. Lane 1: molecular mass markers (SigmaMarker, Sigma). Lane 2:.C3 showing alpha (112kDa) and beta (68 kDa) chains Lane 3: C3a. Lanes 4-8: C3, C3 convertase plus buffer (lane 4) or inhibitor (lanes 5-8: 0.01, 0.1, 0.25, 0.5 mM) showing inhibition of C3a formation after 37°C for 30 min.

Figure S5b: Compound 7 inhibits cleavage of C3 by C3 convertase (CVF.Bb) in a concentration dependent manner (12% gel analysis). Loading of 5.0 µg C3/well. C3 convertase was generated by incubating 12µg of cobra venom factor, 8µg of factor B and 0.2 µg of factor D in 100 µL PBS (pH 7.4) containing 1mM MgCl2 for 1 h at 37°C, then was added 100µL PBS (pH 7.4) containing 20mM EDTA. C3 (10 µg) and the C3 convertase (6 µl) were incubated at 37°C for 30 min. in final 32 µL PBS (pH 7.4) with buffer or various inhibitors concentrations. Lane 1: molecular mass markers (SigmaMarker, Sigma). Lane 2: C3 showing alpha (112kDa) and beta (68 kDa) chains. Lane 3: C3a. Lanes 4-8: C3, C3 convertase plus buffer (lane 4) or inhibitor (lanes 5-8: 0.01, 0.1, 0.25, 0.5 mM) showing inhibition of C3a formation after 37°C for 30 min. Le et al Supplementary Material S6

Figure S6a: Compound 7 (vs. leupeptin and PMSF negative controls) inhibits cleavage of C3 by native C3 convertase (human C3bBb) in a concentration dependent manner (7% gel analysis). C3 (10 µg), factor B (0.5 µg) and factor D (0.05 µg) were incubated at 37°C in 32 µL, PBS containing 1mM MgCl2, pH 7.4 with various concentrations of inhibitor 7, leupeptin or PMSF. Loading of 2.5 µg C3/well. Lane 1: molecular mass markers (SigmaMarker, Sigma). Lane 2: factor B. Lane 3: C3 showing alpha (112kDa) and beta (68 kDa) chains. Lane 4: C3a, Lanes 5-10: C3, factor B and factor D buffer (lane 5) or inhibitor (lanes 6-8: 1.0, 0.5, 0.25 mM compound 7, lane 9: 1.0mM of leupeptin, lane 10: 1,0 mM of PMSF) showing inhibition of C3 cleavage to C3a after 30 minutes at 37°C. All buffers contain 154 mM NaCl and 1mM MgCl2.

Figure S6b: Compound 7 (vs. leupeptin and PMSF negative controls) inhibits cleavage of C3 by native C3 convertase (human C3bBb) in a concentration dependent manner (12% gel analysis). C3 (10 µg), factor B (0.5 µg) and factor D (0.05 µg) were incubated at 37°C in 32 µL, PBS containing 1mM MgCl2, pH 7.4 with various concentrations of inhibitor 7, leupeptin or PMSF. Loading of 5 µg C3/well. Lane 1: molecular mass markers (SigmaMarker, Sigma). Lane 2: factor B. Lane 3: C3 showing alpha (112kDa) and beta (68 kDa) chains. Lane 4: C3a, Lanes 5-10: C3, factor B and factor D buffer (lane 5) or inhibitor (lanes 6-8: 1.0, 0.5, 0.25 mM compound 7, lane 9: 1.0mM of leupeptin, lane 10: 1,0 mM of PMSF) showing inhibition of C3 cleavage to C3a after 30 minutes at 37°C. All buffers contain 154 mM NaCl and 1mM MgCl2.

Le et al Supplementary Material S7

Assay procedure for the determination of IC50 for aldehyde inhibitor AcSHLGLAR-H Vs alternative pathway

The assay kit is available from EURO-DIAGNOSTICA (Medeon, SE-205 12 Malmo, Sweden. Catalogue #: COMPL 300). The protocol is modified from the manufacture’s instruction to allow IC50 determination. Principle of the assay: The assay based on the principle of the haemolytic assay for the complement cascade. Briefly, wells of the microtitre strips are coated with specific activators of each classical, alternative or lectin pathways. Human serum are first diluted with buffers containing specific blockers of the other pathways, ensuring that only the desired pathway is activated in each well. After activation, the resulting MAC is detected with alkaline phosphatase labeled antibodies specific to the neoantigen expressed during MAC formation. The amount of the antibodies are quantified based on extent of substrate cleavage by the phosphatase.

Assay protocol: Stock solutions of inhibitors were prepared in PBS (pH 7.4) in a three-fold serial dilutions starting from the highest concentration of 36.45mM.

Dilution of serum (Sigma, Cat. # H4522): Alternative pathway: 264µL of the supplied alternative pathway diluent was mixed with 30µL of human serum, and 6µL of the inhibitor stock solutions or PBS (positive control). The solutions were incubated at room temperature for 30 minutes.

MAC formation To each well specific for the alternative pathway was added 100µL of the above- diluted serum in duplicate. The wells were incubated at 37°C for 1hr with lid. After incubation the wells were emptied and washed three times with 300 µL of the supplied washing buffer using multi channel pipette. After the last wash, the wells were emptied by tapping the strips on absorbent tissues.

MAC quantification: 100µL of supplied conjugate solution was added to the washed wells and allowed to incubate at room temperature for 30 minutes. The wells were washed as described above. To each well was then added 100µL of substrate solution and absorbance of the wells were monitored at 405nm on a microplate reader at 25°C. The increase of absorbance after 30-minute incubation was calculated and taken as proportional to the quantity of the antibodies.The final concentrations of the inhibitors were: 729µM, 243µM, 81µM, 9µM, 3µM, 1µM, and 0µM. IC50 for AcSHLGLAR-H Vs alternative pathway is 15±5µM

Le et al Supplementary Material S8

IC50 of compound 7 vs. alternative pathway

IC50 = 15±5µM 70

50 Enzyme activity (%) 30

-10 -6 -5 -4 -3 -2 Log[Ac-ASHLGLAR-H]

Figure S7. Inhibition of MAC formation by compound 7 via the alternative pathway.

Le et al Supplementary Material S9

Assay procedure for the determination of IC50 for aldehyde inhibitor AcSHLGLAR-H Vs Thrombin and Trypsin

Buffer For Both Assays : 10mM Hepes, 10mM Tris, 150mM NaCl, 0.1%PEG, pH=7.4

Substrate : The substrate used was in both cases was AcSHLGLAR-pNA. A 100mM solution of the substrate in DMSO was prepared and 0.75 µL of this solution was added to each well. Final concentration in each well is 0.5mM or 500µM.

Enzyme concentration for Trypsin: Commercially supplied trypsin (1 mM) was diluted 100X in buffer (to a concentration of 10uM) and 2µL of this solution was used for each well. Final concentration in each well is 133nM.

Enzyme concentration for Thrombin: Commercially supplied thrombin (100U/ml) was diluted 2X in buffer (50U/mL) and 10µL of this solution was used for each well. Final amount of enzyme in each well is 0.5U.

Inhibitor AcSHLGLAR-H: For trypsin a 2X serial dilution of inhibitor was used starting from 2mM solution of the aldehyde in buffer. The highest concentration of inhibitor used in the assay was 26.6µM. Eleven different descending concentrations of inhibitor were used. For thrombin, a 2X serial dilution of inhibitor was used starting from 50mM solution of the aldehyde in buffer. The highest concentration of inhibitor in the assay was 666µM. Eleven different descending concentrations of inhibitor were used.

Total volume of each well is 150µL and reaction was monitored at 405nm at 37°C.

The assays were performed in duplicate (thrombin) and triplicate (trypsin) – see Fig. 8 below. IC50 for AcSHLGLAR-H Vs Thrombin is 24±1 µM IC50 for AcSHLGLAR-H Vs Trypsin is 716±35 nM Le et al Supplementary Material S10

Figure S8a : Inhibition of trypsin mediated cleavage of AcSHLGLAR-pNA by compound 7.

Figure S8b : Inhibition of thrombin mediated cleavage of AcSHLGLAR-pNA by compound 7. Le et al Supplementary Material S11

NMR Spectroscopy. Samples for NMR analysis were prepared by dissolving the 1 compounds (1-3mg) in 0.5 mL H2O/D2O (9/1) and 1D H-NMR spectra were recorded on a Bruker Avance DRX-600 spectrometer. Spectra were processed using Topspin (Bruker, Germany) software.

Fig. S9a. 1D 1H-NMR spectra of Ac-ARASHLGLAR-pNA (1)

Fig. S9b. 1D 1H-NMR spectra of Ac-RASHLGLAR-pNA (2)

Le et al Supplementary Material S12

Fig. S9c. 1D 1H-NMR spectra of Ac-ASHLGLAR-pNA (3)

Fig. S9d. 1D 1H-NMR spectra of Ac-SHLGLAR-pNA (4)

Fig. S9e. 1D 1H-NMR spectra of Ac-HLGLAR-pNA Le et al Supplementary Material S13

Fig. S9f. 1D 1H-NMR spectra of Ac-LGLAR-pNA

Fig. S9g. 1D 1H-NMR spectra of Ac-GLAR-pNA

1 Fig. S9h. 1D H-NMR spectra of Cbz-KR-pNA in H2O/D2O at 25°C. Le et al Supplementary Material S14

In water, Cbz-KR-pNA exists as a mixture of two conformers in the ratio of ~2.5/1. In DMSO, at 25°C the ratio change to about 8.7/1, and at 45°C only one conformer was observed.

1 Fig. S9i. 1D H-NMR spectra of Cbz-KR-pNA in DMSO d6 at 25°C

1 Fig. S9j. 1D H-NMR spectra of Cbz-KR-pNA in DMSO d6 at 45°C

Le et al Supplementary Material S15

Fig. S9k. 1D 1H-NMR spectra of Ac-AHLGLAR-pNA (5)

Fig. S9m. 1D 1H-NMR spectra of Ac-LHLGLAR-pNA

Fig. S9n. 1D 1H-NMR spectra of Ac-FHLGLAR-pNA Le et al Supplementary Material S16

Fig. S9o. 1D 1H-NMR spectra of Ac-DHLGLAR-pNA

Fig. S9p. 1D 1H-NMR spectra of Ac-KHLGLAR-pNA (6)

Fig. S9q. 1D 1H-NMR spectra of Isox-HLGLAR-pNA Le et al Supplementary Material S17

Fig. S9r. 1D 1H-NMR spectra of Ac-Dapa-HLGLAR-pNA

Fig. S9s. 1D 1H-NMR spectra of Ac-SHLGLAR-H (7) Profiling the Enzymatic Properties and Inhibition of Human Complement Factor B Giang Thanh Le, Giovanni Abbenante and David P. Fairlie J. Biol. Chem. 2007, 282:34809-34816. doi: 10.1074/jbc.M705646200 originally published online October 5, 2007

Access the most updated version of this article at doi: 10.1074/jbc.M705646200

Alerts: • When this article is cited • When a correction for this article is posted

Click here to choose from all of JBC's e-mail alerts Downloaded from

Supplemental material: http://www.jbc.org/content/suppl/2007/10/05/M705646200.DC1.html

http://www.jbc.org/ This article cites 47 references, 18 of which can be accessed free at http://www.jbc.org/content/282/48/34809.full.html#ref-list-1

at UQ Library on October 9, 2016