of Complement C5 Convertase Functional Role of the ...

Functional Role of the Noncatalytic Subunit of Complement C5 Convertase Nenoo Rawal and Michael K. Pangburn This information is current as of June 13, 2013.

References Subscriptions

Email Alerts

This article cites 26 articles, 13 of which you can access for free at: http://www.jimmunol.org/content/164/3/1379.full#ref-list-1 Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscriptions Submit copyright permission requests at: http://www.aai.org/ji/copyright.html Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/cgi/alerts/etoc

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 9650 Rockville Pike, Bethesda, MD 20814-3994. Copyright © 2000 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606.

Downloaded from http://www.jimmunol.org/ by guest on June 13, 2013

Permissions

J Immunol 2000; 164:1379-1385; ; http://www.jimmunol.org/content/164/3/1379

Functional Role of the Noncatalytic Subunit of Complement C5 Convertase1 Nenoo Rawal and Michael K. Pangburn2

he C5 convertases are serine proteases that cleave C5, the fifth component of complement, into two biologically important products, C5a and C5b (1). Both fragments play an important role in killing microorganisms. C5a, the smaller fragment, is a potent chemotactic and spasmogenic anaphylatoxin. It mediates inflammatory responses by stimulating neutrophils and phagocytes. C5b, the larger fragment, initiates the formation of the membrane attack complex (C5b-9) which results in the lysis of bacteria and other pathogens. In our previous work, we have shown that the bimolecular C3 convertase (C3b,Bb) of the alternative pathway of complement can cleave C5 without the help of a second C3b molecule (2). Moreover, this C5 convertase in which the noncatalytic subunit was a single C3b exhibited a catalytic rate constant that was similar to the surface-bound C5 convertase formed on zymosan particles (ZymC3b,Bb) containing multiple C3b. These findings were unexpected since the C5 convertase of the alternative pathway of complement was thought to be made up of a C3 convertase and an additional C3b molecule (3– 8). To further understand the functional role of the noncatalytic subunit, we have determined the C5-cleaving properties of the cobra venom factor (CVF)3-dependent C5 convertases. CVF is a glycoprotein found in the venom of snakes (9, 10). The protein has some unique properties that make it a valuable tool for laboratories studying the role of complement in the pathogenesis of diseases. CVF combines with human factor B and the bound

T

factor B is activated by factor D to form a fluid phase C3/C5 convertase (CVF,Bb). This enzyme cleaves C3 and C5, the third and fifth components of the human complement system (11–20). The convertase formed with CVF, unlike the alternative pathway C3b-dependent C5 convertases (C3b,Bb and ZymC3b,Bb), is very stable and resistant to inactivation by the regulatory proteins, factors H and I (14, 16). In the present study, we examined the enzymatic properties of the C5 convertase formed with CVF purified from the venom of two species of cobras: Naja naja and Naja haje. These two species were examined because the enzyme formed with CVF from Naja naja cleaves C5 (12, 17) much more efficiently than that from Naja haje (17, 21). In designing this study, we thought that a comparison of the kinetic parameters of the four C5 convertases (ZymC3b,Bb, C3b,Bb, CVFn,Bb, and CVFh,Bb) having similar catalytic subunits (human Bb) but different noncatalytic subunits would enhance our understanding of the mechanism of action of this enzyme. The results presented in this paper provide insight into the structure/ function of the C5 convertase. The kinetic and binding data demonstrate that it is the noncatalytic subunit of the C5 convertase that modulates the affinity for the substrate C5. The data also show that the noncatalytic subunit of the enzyme does not influence the catalytic efficiency of the enzyme. The catalytic rate constants measured for the four C5 convertases indicate that the rate of C5 cleavage is one of the slowest enzymatic reactions known. This is apparent in the low turnover numbers of these enzymes of approximately one C5 cleaved per 4 min per enzyme at Vmax.

Department of Biochemistry, University of Texas Health Science Center, Tyler, TX 75708

Materials and Methods

Received for publication August 27, 1999. Accepted for publication November 10, 1999. The costs of publication of this article were defrayed in 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 indicate this fact. 1

This research was supported by Grant DK-35081 from the National Institutes of Health and the 1998 award from the Research Council at the University of Texas Health Center at Tyler. 2 Address correspondence and reprint requests to Dr. Michael K. Pangburn, Department of Biochemistry, University of Texas Health Center at Tyler, 11937 U.S. Highway 271, Tyler, TX 75708-3154. E-mail address: [email protected]

Copyright © 2000 by The American Association of Immunologists

Reagents Lyophilized venom from the Thailand cobra Naja naja kaouthia and from the Egyptian cobra Naja haje haje was obtained from Miami Serpentarium 3 Abbreviations used in this paper: CVF, cobra venom factor; C3b and C5b, the proteolytically activated form of C3 and C5, respectively; Ec, chicken erythrocytes; NHS, normal human serum; C3b,Bb, soluble form of the C5 convertase; ZymC3b,Bb, surface-bound C5 convertase; CVFn, cobra venom factor from Naja naja kaouthia; CVFh, cobra venom factor from Naja haje haje; CVFn,Bb, CVFn-dependent C5 convertase; CVFh,Bb, CVFh -dependent C5 convertase; RU, reference unit measurements from BIAcore.

0022-1767/00/$02.00


The C5 convertase is a serine protease that consists of two subunits: a catalytic subunit which is bound in a Mg2ⴙ-dependent complex to a noncatalytic subunit. To understand the functional role of the noncatalytic subunit, we have determined the C5cleaving properties of the cobra venom factor-dependent C5 convertase (CVF,Bb) made with CVF purified from the venom of Naja naja (CVFn) and Naja haje (CVFh) and compared them to those for two C3b-dependent C5 convertases (ZymC3b,Bb and C3b,Bb). A comparison of the kinetic parameters indicated that although the four C5 convertases (CVFn,Bb, ZymC3b,Bb, CVFh,Bb, and C3b,Bb) had similar catalytic rate constants (kcat ⴝ 0.004 – 0.012 sⴚ1) they differed 700-fold in their affinity for the substrate as indicated by the Km values (CVFn,Bb ⴝ 0.036 ␮M, ZymC3b,Bb ⴝ 1.24 ␮M, CVFh,Bb ⴝ 14.0 ␮M, and C3b,Bb ⴝ 24 ␮M). Analysis of binding interactions between C5 and the noncatalytic subunits (CVFh or C3b, or CVFn) using the BIAcore, revealed dissociation binding constants (Kd) that were similar to the Km values of the respective enzymes. The kinetic and binding data demonstrate that the binding site for C5 resides in the noncatalytic subunit of the enzyme, the affinity for the substrate is solely determined by the noncatalytic subunit and the catalytic efficiency of the enzyme appears not to be influenced by the nature of this subunit. The Journal of Immunology, 2000, 164: 1379 –1385.

1380 (Miami, FL). Chicken erythrocytes (Ec) were isolated from chicken blood purchased from Colorado Serum (Denver, CO). Nonidet P-40, a nonionic detergent, and EDTA were purchased from Sigma (St. Louis, MO). PBS contained 10 mM phosphate and 145 mM NaCl (pH 7.4). Gelatin PBS (GPBS) was PBS containing 0.1% gelatin, whereas GPBSE was GPBS containing 10 mM EDTA. HEPES-buffered saline (HBS) contained 10 mM HEPES (pH 7.4), 150 mM NaCl, 3.4 mM EDTA, and 0.005% P20 surfactant (BIAcore, Uppsala, Sweden).

Purified proteins

Quantitation of reaction products C5b,6 was measured hemolytically as described previously (2). Briefly, tubes containing chicken Ec (1.2 ⫻ 107) and 5 ␮l of pooled normal human serum (NHS) as a source of complement proteins C7–C9 in a final volume of 225 ␮l GPBSE were kept ready on ice. An aliquot (25 ␮l) of the diluted sample from C5 convertase assays or purified C5b,6 was added, and the mixtures were incubated for 10 min at 37°C. The unlysed cells were removed by centrifugation for 1 min at 10,000 ⫻ g. The amount of hemoglobin released was quantitated spectrophotometrically at 414 nm. To determine 100% lysis, Ec were lysed in 2% Nonidet P-40. Control enzyme assays containing C5 and C6, but no C5 convertase, were subtracted as the background. To check whether C5b,6 was formed from the C5 and C6 in the pooled NHS used as a source of C7–C9 during the lysis of Ec, reactions containing C5 convertase but no purified C5 or C6 were used as controls. Controls demonstrated that no significant amount of C5b,6 was formed from NHS.

the kinetic constants of the enzyme. After a 15-min incubation at 37°C, further cleavage of C5 was prevented by transferring the assay tubes to an ice bath and diluting with 100 ␮l ice cold GPBSE. The assay mixture was immediately titrated for C5b,6 formation by hemolytic assays using Ec and quantitated using standard curves generated with purified C5b,6 as described previously (2). Controls demonstrated that the cold temperature and the dilution were sufficient to reduce the cleavage of C5 during subsequent steps to undetectable levels. Enzyme velocities for CVFh,Bb were measured as described above for CVFn,Bb. The reactions were started with the addition of CVFh,Bb. Because of the high Km of CVFh,Bb, concentrations up to 12 mg C5/ml in the assays was required. This resulted in a high C5b,6 background due to C5b-like C5 (27). It was therefore necessary to use a higher concentration of CVFh,Bb (14 nM) so that the background without the enzyme was relatively small. Even at this high level of CVFh,Bb in the assays, the maximum amount of C5 cleaved was ⬍1% of the C5 present. After a 15-min incubation period at 37°C, further cleavage of C5 was prevented by transferring the assay tubes to an ice bath and diluting with 500 ␮l ice cold GPBSE. The reaction velocity data were analyzed according to the MichaelisMenten equation: v ⫽ Vmax [S]/(Km ⫹ [S]). The results were fit to this equation using nonlinear regression analysis, and the kinetic parameters, Km, Vmax, and kcat were determined using Grafit version 3.0 software (Erithacus Software, London, U.K.).

Measurement of binding interactions using BIAcore Binding interactions were measured using a BIAcore biosensor system (BIAcore AB) based on the principle of surface plasmon resonance. C5 was immobilized on a CM5 sensor chip (BIAcore AB) via amine groups using standard N-ethyl-N⬘-(dimethylaminopropyl)carbodiimide/N-hydroxysuccinimide coupling as described by the manufacturer. C5 (0.16 mg/ml) in 3.5 mM phosphate buffer containing 23 mM NaCl (pH 6.3) was incubated with an activated CM5 sensor chip for 1 min to give ⬃400 response units (RU) of C5 immobilized on the chip. After coupling, excess N-hydroxysuccinimide groups were inactivated with 1 M ethanolamine hydrocloride (pH 8.5). The reference flow cell was mock derivatized and blocked with ethanolamine. Immobilization of C5 to the chip was conducted in PBS buffer at 25°C, whereas binding interactions were determined either in PBS or HBS buffer at 37°C and nearly identical results were obtained. Binding interactions were determined by passing samples simultaneously over both the mock-derivatized flow cell and the flow cell with immobilized C5 so as to obtain the RU after subtraction of the background. To minimize mass transfer effects, the ligand level (C5) was kept low and a flow rate of 30 ␮l/min was used. Irrespective of the amount and the type of ligand coupled to the chip, the maximum functional binding capacity of the immobilized ligand as measured by equilibrium binding was found to range between 20 and 40% of the total ligand immobilized on the chip. Because, the functional activity of the C5 used for coupling was determined (2) to be ⬎95%, the unavailable binding sites were probably inactivated by the coupling procedure.

Formation of the C5 convertases CVF n,Bb and CVF h,Bb

Analysis of binding data

CVFn,Bb was made by incubating 0.025– 0.125 ␮g of CVFn (1.67– 8.4 nM) with excess factor B (2.5 ␮g, 397 nM) and excess factor D (2.5 ␮g, 1042 nM) in a total volume of 100 ␮l GPBS containing 0.5 mM MgCl2. After a 15-min incubation at 37°C, 25 ␮l of 15 mM EDTA in GPBS was added. An aliquot of this enzyme mixture was added to 25-␮l assay mixtures measuring the kinetic properties of CVFn,Bb. Preliminary experiments were done to determine the concentrations of factor B and factor D that would give the maximum amount of enzyme. The C5 convertase, CVFh,Bb was made as described for CVFn,Bb except that higher concentrations of CVFh (1.26 ␮g, 84 nM) and factor B (37.5 ␮g, 5952 nM) were used. A total of 4 ␮l of the enzyme mixture containing CVFh,Bb was added to a final volume of 25-␮l assay mixtures measuring the kinetic properties of CVFh,Bb. CVFh,Bb and CVFn,Bb were made just before use.

Dissociation constants (Kd) were calculated from the analysis of equilibrium binding (RU) measured as a function of the concentration of the analyte. The binding data were analyzed according to a one-site binding equation using nonlinear regression analysis, and the binding constant Kd determined using Grafit version 3.0 software (Erithacus Software). The binding constant was also determined based on the association (kon) and dissociation (koff) rate constants using BIAevaluation version 3.0 software (BIAcore). The constants were calculated based on a one-site binding interaction with mass transfer model, and the overlay plots were globally fitted to simultaneous kon/koff simple Langmuir system (A ⫹ B 7 AB). The dissociation rate constant was calculated from the decay curve (data from approximately 125 to 150 s on the x-axis in Fig. 3A). The association rate constant was calculated from the slope of the association curve (data from approximately 0 to 25 s on the x-axis in Fig. 3A) and the dissociation rate constant. As a result of the low affinity of CVFh for C5, the Kd of this interaction was determined by first measuring the maximum binding of the C5-CM5 sensor chip with the high-affinity CVFn. The degree of saturation was then measured at several CVFh concentrations, and the binding constant was calculated at each concentration using the following equation: Kd ⫽ (Rmax ⫺ RCVF)(CVFh conc)/RCVF, where Rmax was the maximum RU determined with saturating CVFn, RCVF was the RU observed at equilibrium and (CVFh conc) was the concentration of CVFh injected over the chip.

Kinetic measurements of the C5 convertases CVFn,Bb and CVF h,Bb Enzyme velocities were determined under saturating concentrations of C6 in siliconized microfuge tubes. Assay mixtures contained 2.5 ␮g of C6 (833 nM) and varying concentrations of C5. The reaction was started by the addition of CVFn,Bb in a final volume of 25 ␮l of GPBS. Different concentrations of CVFn,Bb in the range of 0.27–1.3 nM were used to measure


Complement proteins used in the present studies were all purified from normal human plasma. C3 (22, 23), factor B (11), and factor D (24) were isolated as described in the references cited. C5 was isolated as described (25) except that ceramic hdroxylapatite (Bio-Rad, Richmond, CA) was used instead of hydroxylapatite. C3b was prepared from C3 by cleavage with factors B and D in the presence of Ni2⫹ at 37°C as described (2). C6 was purified as described by Kolb et al. (26) and purified C5b,6 was obtained from Advance Research Technologies (San Diego, CA). All proteins were homogenous by polyacrylamide gel electrophoresis. Protein concentrations of C3b, C5, C6, factor B, factor D, and C5b,6 were determined spectrophotometrically using 11.0, 11.0, 10.8, 12.7, 11.0, and 10.3, as the 1% value for E280 , respectively. All purified proteins were stored at ⫺76°C. Mr values employed in the calculations were 176,000 for C3b, 190,000 for C5, 179,000 for C5b, 120,000 for C6, 299,000 for C5b,6, 150,000 for CVF, 93,000 for factor B, 63,000 for Bb, and 24,000 for factor D. CVF either from the venom of Naja naja or from the venom of Naja haje was purified on a Hi load Q-Sepharose HP column using the fast protein liquid chromatography (FPLC, Amersham Pharmacia Biotech, U.K.). Fractions containing CVF were pooled, concentrated, and further purified by chromatography over a Bio-Gel (0.5 M) column followed by a Mono Q column using fast protein liquid chromatography. The purity of CVF from both the species was determined by SDS-PAGE, which showed a single 150-kDa band in the absence of reducing agents. Protein concentration of CVF was determined spectrophotometrically using 10.0 as the 1% value for E280 .

STRUCTURE/FUNCTION OF COMPLEMENT C5 CONVERTASE

The Journal of Immunology

1381

Results Rates of formation and decay of CVF,Bb The rate of formation of CVF,Bb was determined by addition of CVF to assays containing excess factor B, factor D, C5, C6, and Mg2⫹. Formation of CVF,Bb was stopped by the addition of EDTA and the amount of C5b,6 produced in the next 15 min was measured. Maximum amount of enzyme formation was achieved in 8 min. Thereafter the level of CVF,Bb remained constant (data not shown). Therefore, all enzymatic assays employed CVF,Bb that was preformed for 15 min. The rate of formation of CVF,Bb was slow when compared with the rate of formation of the C3bdependent C5 convertases. Both C3b,Bb and ZymC3b,Bb have been shown to form in ⬍1 min under saturating concentrations of factor B, factor D, and metal ions (2). Because CVF,Bb has been reported to have a half-life of 7 h (18), the rate of decay of CVF,Bb in a 15-min assay period was considered negligible. Furthermore, kinetic experiments conducted with and without enzyme regeneration during the 15-min assays showed no significant differences.

To measure the kinetic constants of the C5 convertase activity of CVFn,Bb made with CVF from Naja naja, cleavage of C5 was determined as the amount of C5b,6 formed during a 15-min assay period under saturating concentrations of factors B and D and C6. The rate of C5 cleavage was found to be linear for the 15-min assay period even at the lowest concentration of C5 used in this study. Initial velocities were then determined at various C5 concentrations and at a fixed concentration of enzyme determined from the amount of CVF added to the assay mixture. The data obtained were found to fit well to the theoretical curve based on the Michaelis-Menten equation, v ⫽ Vmax [S]/(Km ⫹ [S]), as shown in Fig. 1. The data were also observed to fit well to the linearized form of the Michaelis-Menten equation shown as the Eadie-Hofstee plot (Fig. 1, inset). The kinetic constants calculated from the data in Fig. 1 are summarized in Table I. The Km of CVFn,Bb for C5 was 0.036 ␮M, and the kcat was 0.0071 s⫺1. The low Km value of CVFn,Bb revealed a strong binding interaction of the enzyme complex with the substrate. Due to the low Km, care was taken to ensure that significant substrate consumption did not occur during the assays. In most assays C5 cleavage was ⬍1%. In assays at the lowest substrate concentrations (2–20 nM), cleavage of C5 ranged from 6% to a maximum of 20% at the end of the 15-min assays. C5 consumption was kept low by using low concentrations of enzyme. The Km of CVFn,Bb when compared with the Km reported for the C3b-dependent C5 convertases (2) was found to be 34-fold lower than the surface-bound C5 convertase (ZymC3b,Bb, Km ⫽ 1.24 ␮M) and 700-fold lower than the soluble form of the C5 convertase (C3b,Bb, Km ⫽ 24 ␮M) as shown in Table I. Although the Km values varied widely, the catalytic rate constant of CVFn,Bb was within 3-fold of those of the two C3b-dependent C5 convertases of the human complement system (Table I) (2). Measurement of kinetic parameters of CVF h,Bb Because the convertase formed with CVF from Naja haje has been previously reported to have minimal ability to cleave C5 due to its poor affinity for C5 (17), we wanted to determine the kinetic parameters of the enzyme and compare it with those of the enzyme formed with CVF from Naja naja. CVFh,Bb not only cleaved C5, but did so at almost exactly the same catalytic rate (kcat) as the enzyme formed with CVF from Naja naja (Table I). The velocity vs substrate plot (Fig. 2) showed an excellent fit to the MichaelisMenten equation. Analysis of the Km value revealed that CVFh,Bb had a Km (14.0 ␮M) that was 400-fold greater than that of

CVFn,Bb (Km ⫽ 0.036 ␮M) (Table I). The high Km value of CVFh,Bb for C5 was observed to be similar to the high Km value (24 ␮M) of the soluble form of the C3b-dependent C5 convertase (C3b,Bb) (Table I). Equilibrium binding interactions between CVF

n

and C5

A comparison of the kinetic properties of the CVF-dependent C5 convertases (CVFn,Bb and CVFh,Bb) with those of the C3b-dependent C5 convertases (ZymC3b,Bb and C3b,Bb) indicated that all four C5 convertases had catalytic rate constants that were within 3-fold but differed 700-fold in their Km values. Since the four C5 convertases had a similar catalytic subunit (human Bb), but different noncatalytic subunits, the results indicated that the noncatalytic subunit was responsible for the differences observed in the Km values. To evaluate the contributions of the Bb subunit and the catalytic rate constant to the Km values, we measured binding interactions between the noncatalytic subunit of the enzyme and the substrate C5 using surface plasmon resonance. Fig. 3A shows an overlay plot of the binding response observed between immobilized C5 and various concentrations of CVFn ranging from 2.5 to 160 nM. Fig. 3B shows the binding curve obtained by nonlinear regression analysis of the equilibrium binding data in Fig. 3A. Linear transformation of the data in Fig. 3B is shown as a Scatchard plot in Fig. 3C. The low dissociation constant (Kd ⫽ 0.042 ␮M, Table II) for the interaction between CVFn and C5 agreed very well with the Km value determined for the bimolecular enzyme CVFn,Bb (0.036 ␮M, Table II), suggesting that the observed Km is governed by the interaction of C5 with the noncatalytic subunit of the enzyme.


Measurement of kinetic parameters of CVF n,Bb

FIGURE 1. Kinetic analysis of the C5 convertase activity of CVFn,Bb in the presence of varying concentrations of C5. Initial rates of C5 cleavage by the C5 convertase made with CVF from Naja naja were determined by measuring the amount of C5b,6 formed in 15 min at 37°C after demonstrating that the reaction was linear over this interval. Reactions contained saturating levels of C6 (2.5 ␮g, 833 nM) and the indicated concentration of C5. Reactions were initiated by adding 5 ␮l of CVFn,Bb to a final volume of 25 ␮l GPBS which resulted in 0.27 nM C5 convertase in the assay. After 15 min at 37°C, the reactions were analyzed for C5b,6 formation after appropriate dilution with cold GPBSE. An aliquot of 25 ␮l was analyzed for C5b,6 formation by adding 225 ␮l of GPBSE containing Ec (1.2 ⫻ 107) and 5 ␮l of pooled NHS as a source of C7–C9 to the reaction mixture and incubating it for 10 min at 37°C. Percent hemolysis was determined from the release of hemoglobin at 414 nm. CVFn, Bb was made at 37°C for 15 min before use as described in Materials and Methods. The data were fitted by nonlinear regression according to the Michaelis-Menten equation to determine the Km and Vmax values using Grafit version 3.0 Erithacus software. Inset, Analysis of the C5 convertase initial velocity data using EadieHofstee function.

1382

STRUCTURE/FUNCTION OF COMPLEMENT C5 CONVERTASE Table I. Kinetic parameters of C5 convertasesa

Enzyme

Km (␮M)

kcat (s⫺1)

kcat/Km (s⫺1M⫺1)

Turnover Number (min⫺1)

CVFn,Bb CVFh,Bb C3b,Bb ZymC3b,Bb

0.036 ⫾ 0.013 14 ⫾ 6.6 24 ⫾ 6.6 1.24 ⫾ 0.27

0.0071 ⫾ 0.0026 0.0121 ⫾ 0.0061 0.0110 ⫾ 0.0038 0.0038 ⫾ 0.0013

197,000 ⫾ 67,200 560 ⫾ 360 470 ⫾ 170 3,080 ⫾ 790

0.43 ⫾ 0.16 0.73 ⫾ 0.37 0.68 ⫾ 0.22 0.23 ⫾ 0.08

a Kinetic parameters of CVFn,Bb and CVFh,Bb were determined by nonlinear regression as shown in Figs. 1 and 2, respectively. Kinetic parameters of C3b,Bb and ZymC3b,Bb have been reported from Rawal and Pangburn (2). Values are mean ⫾ SD (n ⫽ 3 or 4).

Equilibrium binding interactions between CVF n

h


The Kd of CVFn for C5 was also derived from the association and dissociation rate constants by globally fitting the data in Fig. 3A to a one-binding site model as described in Materials and Methods. CVFn was found to bind to C5 with a rate constant (kon) of 2.0 ⫻ 106 M⫺1s⫺1, whereas it dissociated with a rate constant (koff) of 0.085 s⫺1. The Kd (0.043 ␮M) obtained from these rate constants agreed well with the Kd (0.042 ␮M, Table II) obtained from the equilibrium binding data in Fig. 3B. The half-life of the CVFn,C5 complex was calculated to be 9.4 s. When the immobilized ligand was CVFn instead of C5, a similar Kd was found (Table II). Immobilized CVF was converted to CVFn,Bb by injection of saturating levels of factors B and D in the presence of magnesium. Formation of CVFn,Bb was followed by measuring the binding of Bb to CVFn as indicated by an increase in plasmon resonance. A Kd determined using various concentrations of C5 indicated that the presence of the catalytic Bb subunit had no effect on the affinity of CVFn for C5 (Table II). and C5

h

In contrast to CVF , the affinity between CVF and C5 was found to be extremely weak (Fig. 3B). A Kd value of 16 ␮M (Table II) was determined from the degree of saturation at equilibrium binding, indicating a very weak interaction between CVFh and C5 when compared with the 0.042 ␮M Kd of CVFn. These measurements of affinity were in excellent agreement with the Km value of the enzyme CVFh,Bb (Km ⫽ 14 ␮M).

FIGURE 2. Kinetic analysis of the C5 convertase activity of CVFh,Bb in the presence of varying concentrations of C5. Initial rates of C5 cleavage by the C5 convertase made with CVF from Naja haje were determined by measuring the amount of C5b,6 formed in 15 min at 37°C. Assay mixtures contained saturating levels of C6 (2.5 ␮g, 833 nM) and the indicated concentration of C5. Reactions were initiated by adding 4 ␮l of CVFh,Bb to a final volume of 25 ␮l GPBS which resulted in 13.5 nM C5 convertase in the assay. After 15 min at 37°C, the reactions were stopped by the addition of cold GPBSE. Appropriately diluted aliquots were quantitated for C5b,6 formation, and the data were analyzed as described in Fig. 1.

FIGURE 3. Binding of CVF from Naja naja and from Naja haje to immobilized C5 using the BIAcore. A, An overlay plot of response curves obtained from the BIAcore instrument when various concentrations of CVFn from Naja naja (2.5–160 nM) prepared in PBS buffer containing 0.1% BSA were passed over a CM5 sensor chip with 70 RU of functional C5 immobilized via amine coupling. To minimize mass transfer limitations, the ligand level was kept low and a flow rate of 30 ␮l/min was used. Binding of CVF to immobilized C5 was determined at 37°C for 2 min. B, The analysis of the binding data in A is presented showing the change in RU at equilibrium due to CVF binding to the C5-derivatized chip as a function of CVF concentration. B also shows data from BIAcore-binding assays using CVFh. The data were analyzed by nonlinear regression using Grafit version 3.0 Erithacus software and employing a one-binding site equation. C, Scatchard plot of the binding data of C5 and CVFn in B.

The Journal of Immunology

1383

Table II. Comparison of the kinetic constant Km with the binding constant Kda

Enzyme n

CVF ,Bb CVFh,Bb C3b,Bb

Substrate

Km (␮M)

C5 C5 C5

0.036 ⫾ 0.013 14 ⫾ 6.6 24 ⫾ 6.6

Analyte n

CVF CVFh C3b C5 C5

Immobilized Ligand

Kd (␮M)

C5 C5 C5 CVFn CVFn,Bb

0.042 ⫾ 0.004 16 ⫾ 2 37 ⫾ 5 0.105 ⫾ 0.03 0.092 ⫾ 0.01

a Km of C3b,Bb has been reported (2). Kd was determined from BIAcore data by nonlinear regression. Values are mean ⫾ SD (n ⫽ 3 or 4). Binding constant of C5 for CVFn, CVFh and C3b was derived from Figs. 3B and 4, respectively. Binding constant of C5 for CVFn was determined with a CM5 chip containing 850 RU of CVFn immobilized on the chip, and binding interactions were determined under conditions described in Fig 3. Binding constant of C5 for CVFnBb was obtained by converting the CVFn chip to CVFnBb by passing excess factor B and factor D in the presence of 1.25 mM MgCl2. Binding interactions between C5 and CVFnBb were determined under conditions described in Fig. 3.

Equilibrium binding interactions between C3b and C5

Comparison of C5 convertase activities at the normal plasma concentration of C5 A comparison of the activities of the CVF- and the C3b-dependent C5 convertases is shown as a plot of enzyme velocity vs substrate concentration in Fig. 5. The data presented show that at normal physiologic concentrations of C5 in plasma (0.37 ␮M, vertical dashed line) the four C5 convertases will cleave C5 at different

FIGURE 4. Binding of soluble monomeric C3b to immobilized C5 using BIAcore. Binding of C3b to C5 was measured as a change in RU at equilibrium. A CM5 sensor chip was derivatized with 750 RU of C5 via amine coupling, and C3b binding was determined by the degree of saturation of C5 in PBS buffer at 37°C at a flow rate of 5 ␮l/min. Inset, An overlay plot of response curves obtained when 1.0 and 2.5 ␮M C3b prepared in HBS buffer was passed over a CM5 sensor chip containing 7500 RU of C5 immobilized.

Discussion The present study provides insight into the structure/function of the C5 convertase of the alternative pathway of complement. Using surface plasmon resonance and conventional kinetic analysis, we have demonstrated that the noncatalytic subunit of the C5 convertase determines the affinity for the substrate C5. The functional role of the noncatalytic subunit of the enzyme was revealed by comparing the enzymatic and substrate-binding properties of the CVF-dependent C5 convertases (CVFn,Bb and CVFh,Bb) and the C3b-dependent C5 convertases (ZymC3b,Bb and C3b,Bb). The four C5 convertases examined had a common catalytic subunit (human Bb) but differed in their noncatalytic subunit. Although the catalytic rate constants (kcat) of the C5 convertases were within 3-fold of each other, the Km varied 700-fold (Table I). Binding

FIGURE 5. Comparison of the activities of the different C5 convertases at the normal plasma concentration of C5. Initial rates of C5 cleavage by the different C5 convertases (CVFn,Bb, ZymC3b,Bb, CVFh,Bb, and C3b, Bb) were determined by measuring the amount of C5b,6 formed at 37°C. The y-axis represent the velocities that have been normalized by dividing by Vmax and multiplying by 100. - - -, Normal plasma concentration of C5 (0.37 ␮M). Data for CVFn,Bb (E) and CVFh,Bb (F) were from Figs. 1 and 2, respectively. Data for ZymC3b,Bb (Œ) and C3b,Bb (‚) were from Rawal and Pangburn (2).


The affinity between monomeric C3b and C5 was also examined by equilibrium binding using the BIAcore. The dissociation constant obtained from this analysis (Fig. 4) indicated a weak (Kd ⫽ 37 ␮M) interaction between C3b and C5. This affinity derived from equilibrium binding was very similar to the Km of the enzyme C3b,Bb (Km ⫽ 24 ␮M). Very fast on and off rates of C3b for immobilized C5 (inset in Fig. 4) precluded measurement of rate constants and therefore a rate-derived Kd could not be determined. Attempts to measure the on and off rates at C3b concentrations that were 100-fold below the Kd value obtained from equilibrium binding (37 ␮M) indicated that the on and off rates were faster than the maximum rates detectable by the BIAcore. These findings suggest that the rate of association of C3b and C5 is considerably ⬎5 ⫻ 106 M⫺1s⫺1 and that the rate of dissociation is faster than 10⫺1s⫺1.

rates. The Km of CVFn,Bb (0.036 ␮M, Table I) is 10-fold below the normal plasma concentration of C5, suggesting that CVFn,Bb will function close to Vmax, whereas that of ZymC3b,Bb (1.24 ␮M, Table I) is 3-fold above, indicating that the natural surface-bound C5 convertase will operate at ⬃20% of Vmax. In contrast, the rate of C5 cleavage by the C5 convertases, CVFh,Bb and C3b,Bb, will be extremely slow when compared with that of CVFn,Bb and ZymC3b,Bb.

1384

strate complex with C5, the remaining 10% of the enzyme will be cleaving C3 at a rate many times faster than the rate of C5 cleavage. In contrast, the CVF from Naja haje will consume C3 long before it consumes C5. Its rate of C5 cleavage, and therefore the rate of C5a release, will be ⬃20-fold slower than that of CVFn. These findings suggest the following results might be expected in vivo. First, for rapid and complete inactivation of the complement system in laboratory animals, CVF from Naja naja is the better choice because C5 levels will be reduced rapidly to very low levels. CVFh will cleave C3 much faster, but C5 levels will decrease slowly and the steady-state level of C5 will be high even in the presence of enzyme due to the poor affinity for C5. Second, these observations suggest that CVF from Naja naja will be more toxic than that from Naja haje due to rapid release of the more potent anaphylatoxin C5a. In agreement with this conclusion, it has been reported that the lethal dose of whole venom from Naja naja in a mouse is 3– 6 ␮g (s.c.) whereas seven times more of the Naja haje venom is required (30). In conclusion, if the goal is maximum suppression of complement function in vivo, then the better choice will be CVF from Naja naja provided that the high rate of C5a release and of C5b–C9 formation is tolerated.

Acknowledgments We thank Nicole S. Narlo, Kerry L. Wadey-Pangburn, and Cherri Bell for excellent technical assistance.

References 1. Rother, K. 1998. Complement in inflammation. In The Complement System, 2nd Ed. K. Rother, G. O. Till, and G. M. Ha¨nsch, eds. Springer, New York, p. 462. 2. Rawal, N., and M. K. Pangburn. 1998. C5 convertase of the alternative pathway of complement: kinetic analysis of the free and surface-bound forms of the enzyme. J. Biol. Chem. 273:16828. 3. Daha, M. R., D. T. Fearon, and K. F. Austen. 1976. C3 requirements for formation of alternative pathway C5 convertase. J. Immunol. 117:630. 4. Medicus, R. G., R. D. Schreiber, O. Go¨tze, and H. J. Mu¨ller-Eberhard. 1976. A molecular concept of the properdin pathway. Proc. Natl. Acad. Sci. USA 73:612. 5. Medicus, R. G., O. Go¨tze, and H. J. Mu¨ller-Eberhard. 1976. Alternative pathway of complement: recruitment of precursor properdin by the labile C3/C5 convertase and the potentiation of the pathway. J. Exp. Med. 144:1076. 6. Vogt, W., G. Schmidt, B. Von Buttlar, and L. Dieminger. 1978. A new function of the activated third component of complement: binding to C5, an essential step for C5 activation. Immunology 34:29. 7. Isenman, D. E., E. R. Podack, and N. R. Cooper. 1980. The interaction of C5 with C3b in free solution: a sufficient condition for cleavage by fluid phase C3/C5 convertase. J. Immunol. 124:326. 8. Kinoshita, T., Y. Takata, H. Kozono, J. Takeda, K. Hong, and K. Inoue. 1988. C5 convertase of the alternative complement pathway: covalent linkage between two C3b molecules within the trimolecular complex enzyme. J. Immunol. 141:3895. 9. Mu¨ller-Eberhard, H. J., and K. E. Fjellstro¨m. 1971. Isolation of the anticomplementary protein from cobra venom and its mode of action on C3. J. Immunol. 107:1666. 10. Vogel, C. W., R. Bredehorst, D. C. Fritzinger, T. Grunwald, P. Ziegelmu¨ller, and M. A. Kock. 1996. Structure and function of cobra venom factor, the complement-activating protein in cobra venom. Adv. Exp. Med. Biol. 391:97. 11. Go¨tze, O., and H. J. Mu¨ller-Eberhard. 1971. The C3-activator system: an alternative pathway of complement activation. J. Exp. Med. 134:90s. 12. Cooper, N. R. 1973. Formation and function of a complex of the C3 proactivator with a protein from cobra venom. J. Exp. Med. 137:451. 13. Vogt, W., L. Dieminger, R. Lynen, and G. Schmidt. 1974. Alternative pathway for the activation of complement in human serum: formation and composition of the complex with cobra venom factor that cleaves the third component of complement. Hoppe-Seyler’s Z. Physiol. Chem. 355:171. 14. Lachmann, P. J., and L. Halbwachs. 1975. The influence of C3b inactivator (KAF) concentration of the ability of serum to support complement activation. Clin. Exp. Immunol. 21:109. 15. Miyama, A., T. Kato, I. Minoda, T. Ueda, and S. Kashiba. 1976. Activation of terminal components of human complement by a trypsin-activated complex of human factor B and cobra venom factor. Jpn. J. Microbiol. 20:507. 16. Nagaki, K., K. Iida, M. Okubo, and S. Inai. 1978. Reaction mechanims of B1H globulin. Int. Arch. Allergy Appl. Immunol. 57:221. 17. von Zabern, I., B. Hinsch, H. Przyklenk, G. Schmidt, and W. Vogt. 1980. Comparison of Naja n. naja and Naja h. haje cobra-venom factors: correlation between binding affinity for the fifth component of complement and mediation of its cleavage. Immunobiology 157:499.


constants, Kd, measured between C5 and the noncatalytic subunits, CVFn, CVFh, or C3b (0.042, 16, and 37 ␮M, respectively) were found to be similar to the Km values of the respective enzyme as shown in Table II (0.036, 14, and 24 ␮M, respectively). Considered together, these results indicate that although the Km of the enzyme is regulated by the interaction of C5 with the noncatalytic subunit (Kd), the catalytic efficiency of the C5 convertase appears not to be dependent on this subunit of the enzyme. The catalytic subunit Bb was observed to have no effect on the binding affinity of C5 for the noncatalytic subunit CVFn (Table II). This finding suggests a very weak or nonexistent interaction between Bb and C5. This was unexpected because the catalytic subunit of an enzyme has to bind to the substrate to cleave it. One possible explanation of this could be that C5 binds to the noncatalytic subunit of the enzyme to form a pseudo-enzyme-substrate complex which then must wait for catalysis. With C5 waiting near the active site, the slow catalytic rate (1– 4 min/cleavage) must be because of one or more of the following: 1) Bb is rarely in an active conformation, 2) C5 is seldom in a cleavable conformation, 3) both of these proteins are only occasionally active, or 4) they are rarely in the right orientation to each other for catalysis to occur. Because all of these C5 convertases also cleave C3, and because they cleave it ⬃100 times faster than C5, we can conclude that the active site of Bb is proteolytically active more often than C5 is cleaved. This would suggest that other mechanisms are operative here. Either the C5 alone is the rate-limiting entity, orientation is critical, or catalysis must wait until the instant when both C5 and Bb are in position and active. Measurement of the kinetic constants of the CVF-dependent C5 convertases revealed that CVFh,Bb, which has been reported both to cleave and not to cleave C5 (12, 17), not only cleaved C5, but did so at a catalytic rate similar to that of CVFn,Bb and the other C5 convertases (Table I). However, CVFh,Bb differed from CVFn,Bb in having a 400-fold weaker affinity for C5 as indicated by the Km and Kd values shown in Table II. These results suggest that at physiological concentrations of C5 in human plasma (0.37 ␮M), CVFn,Bb will function nearly at Vmax whereas CVFh,Bb will cleave C5 ⬃20-times slower than CVFn,Bb (Fig. 5). Because CVFh,Bb will cleave very little C5 per minute in plasma, the amount of C5a generated when CVF from Naja haje is injected in animals may be insufficient to cause any noticeable physiological damage. This interpretation is supported by the work of Flick et al. (28) who have shown that CVF from Naja haje does not provoke lung injury. On the other hand, the widely used Naja naja CVF has been associated with significant physiological damage (29). Studies have shown that infusion of CVF from Naja naja in sheep causes pulmonary microvascular lung injury (28), and the injury has been attributed to the generation of C5a. The CVF-dependent enzyme CVFn,Bb has been well characterized as a C3 convertase by Vogel and Mu¨ller-Eberhard (18). Now that we have determined its properties as a C5 convertase (Table II), it is interesting to compare both the C3- and C5-cleaving properties of this enzyme. Based on the turnover numbers of the two activities of the enzyme (28 C3/min vs 0.43 C5/min/enzyme), CVFn,Bb will cleave nearly 66 C3 for every one C5 cleaved at Vmax. However, in plasma at physiological concentrations of C3 and C5, this will not occur. Because the Km for C5 (0.036 ␮M) is 10-fold below the plasma concentration of C5 (0.37 ␮M) and the Km reported for C3 (11.6 ␮M) (18) is 2-fold above the plasma concentration of C3 (6.0 ␮M), C5 will compete effectively with C3. Calculations taking into account the Km and Vmax of CVFn,Bb indicate that in plasma this enzyme will cleave ⬃7, instead of 66, C3 molecules for each C5 cleaved. This calculation shows that even though ⬃90% of the convertase will be in an enzyme-sub-

STRUCTURE/FUNCTION OF COMPLEMENT C5 CONVERTASE

The Journal of Immunology 18. Vogel, C. W., and H. J. Mu¨ller-Eberhard. 1982. The cobra venom factor-dependent C3 convertase of human complement: a kinetic and thermodynamic analysis of a protease acting on its natural high molecular weight substrate. J. Biol. Chem. 257:8292. 19. Smith, C. A., C. W. Vogel, and H. J. Mu¨ller-Eberhard. 1982. Ultrastructure of cobra venom factor-dependent C3/C5 convertase and its zymogen, factor B of human complement. J. Biol. Chem. 257:9879. 20. Discipio, R. G., C. A. Smith, H. J. Mu¨ller-Eberhard, and T. E. Hugli. 1983. The activation of human complement component C5 by a fluid phase C5 convertase. J. Biol. Chem. 258:10629. 21. Bauman, N. 1978. Lack of C5 convertase-generating activity in Naja haje cobra factor. J. Immunol. 120:1763. 22. Hammer, C. H., G. H. Wirtz, L. Renfer, H. D. Gresham, and B. F. Tack. 1981. Large scale isolation of functionally active components of the human complement system. J. Biol. Chem. 256:3995. 23. Pangburn, M. K. 1987. A fluorimetric assay for native C3: the hemolytically active form of the third component of human complement. J. Immunol. Methods 102:7. 24. Lesavre, P. H., T. E. Hugli, A. F. Esser, and H. J. Mu¨ller-Eberhard. 1979. The

1385

25.

26. 27.

28.

29.

30.

alternative pathway C3/C5 convertase: chemical basis of factor B activation. J. Immunol. 123:529. Tack, B. F., and J. W. Prahl. 1976. Third component of human complement: purification from plasma and physicochemical characteriztion. Biochemistry 15: 4513. Kolb, W. P., L. M. Kolb, and J. R. Savar. 1982. Biochemical characterization of the sixth component (C6) of human complement. Biochemistry 21:294. Dessauer, A., U. Rother, and K. Rother. 1984. Freeze-thaw activation of the complement attack phase. I. Separation of two steps in the formation of the active C56 complex. Acta Pathol. Microbiol. Immunol. Scand. C92 (Suppl. 284):75. Flick, M. R., J. K. Horn, J. M. Hoeffel, and I. M. Goldstein. 1986. Reduction of total hemolytic complement activity with Naja haje cobra venom factor does not prevent endotoxin-induced lung injury in sheep. Am. Rev. Respir. Dis. 133:62. Van den Berg, C. W., P. C. Aerts, and H. Van Dijk. 1991. In vivo anticomplementary activities of the cobra venom factors from Naja naja and Naja haje. J. Immunol. Methods 136:287. Minton, S. A., Jr., and M. R. Minton. 1980. Venomous Reptiles. Charles Scribner’s Sons, New York, p. 246.
