Activation of Complement Component C5

Although the initiating complex of lectin pathway (called M1 in this study) generates C3/C5 convertases similar to those assembled by the initiating complex (C1) of the classical pathway, activation of complement component C5 via the lectin pathway has not been examined. In the present study kinetic analysis of lectin pathway C3/C5 convertases assembled on two surfaces (zymosan and sheep erythrocytes coated with mannan (EMan)) revealed that the convertases (ZymM1,C4b,C2a and EManM1,C4b,C2a) exhibited a similar but weak affinity for the substrate, C5 indicated by a high Km (2.73-6.88 μm). Very high affinity C5 convertases were generated when the low affinity C3/C5 convertases were allowed to deposit C3b by cleaving native C3. These C3b-containing convertases exhibited Km (0.0086-0.0075 μm) well below the normal concentration of C5 in blood (0.37 μm). Although kinetic parameters, Km and kcat, of the lectin pathway C3/C5 convertases were similar to those reported for classical pathway C3/C5 convertases, studies on the ability of C4b to bind C2 indicated that every C4b deposited on zymosan or EMan was capable of forming a convertase. These findings differ from those reported for the classical pathway C3/C5 convertase, where only one of four C4b molecules deposited formed a convertase. The potential for four times more amplification via the lectin pathway than the classical pathway in the generation of C3/C5 convertases and production of pro-inflammatory products, such as C3a, C4a, and C5a, implies that activation of complement via the lectin pathway might be a more prominent contributor to the pathology of inflammatory reactions.

C3/C5 convertases are enzymes that cleave C3 and C5. Products generated upon the cleavage of C3 and C5 (C3a, C3b, C5a, and C5b) have important biological activities (20). C3b, the larger cleavage product of C3, opsonizes bacteria, whereas C3a, the smaller product, is an anaphylatoxin with important functions such as chemotaxis of mast cells and contraction of smooth muscle cells (21)(22)(23)(24). C5b, the larger cleavage product of C5, initiates formation of the cytolytic complex (C5b-9) that causes lysis of bacteria and pathogens, whereas C5a, the smaller product, is a strong chemotactic and spasmogenic anaphylatoxin that mediates inflammatory responses by stimulating neutrophils and phagocytes to the site of injury or infection (21,25). These products help fight infections, but they also contribute to the pathology of several diseases and conditions such as inflammatory diseases, reperfusion injury, and xenotransplantation rejection (26).
Although activation of the lectin pathway generates C3/C5 convertases similar to those generated via the classical pathway (17), C5 activation via the lectin pathway has not been examined. The classical pathway C3/C5 convertase (C4b,C2a) is bimolecular in nature and is made up of C4b as the noncatalytic subunit and C2a as the catalytic subunit. The specificity for the substrate changes depending on the nature of the noncatalytic subunit of the C3/C5 convertase (27)(28)(29). The heterodimer C4b-C3b has been shown by Takata et al. (30,31) to form a high affinity C5 binding site, whereas kinetic studies from our laboratory have shown C4b,C2a to cleave both C3 and C5. But C4b,C2a has a weak affinity for C5, and the convertase primarily cleaves C3; therefore, the C3/C5 convertase C4b,C2a is called a C3 convertase (31). Deposition of additional C3b molecules by C4b,C2a converts the low affinity C3 convertase to a high affinity C5-binding convertase (C3bC4b,C2a) (31).
Although there are strong similarities between the M1 complex (MBL⅐MASPs) of the lectin pathway and the C1 complex (C1qrs) of the classical pathway, reports from various studies examining different aspects of complement activation have suggested that differences in the efficiency of C3/C5 convertase formation might exist between the pathways. Activation of the classical pathway occurs when the C1q component of the C1 complex binds to antigen-antibody complexes (32). The C1 complex consists of three components, C1q and the serine proteases C1r and C1s, in the ratio 1:2:2. C1r and C1s each form a dimer that interacts to form a C1r 2 C1s 2 tetramer that binds to C1q. Cleavage of C4 and C2 are mediated by C1s in the C1r 2 -C1s 2 tetramer, which gets activated upon the binding of C1 to antigen-antibody complexes (33,34). Similarly, cleavage of C4 and C2 upon activation of the lectin pathway is mediated by MASP-2 in the M1 complex (13,(17)(18)(19)35). Each type of MASP has been reported to form a homodimer (10,36,37), but the stoichiometry of MASPs in association with MBL has yet to be established. Crystallography studies have shown that, even though C1s and MASP-2 have identical substrate specificity, different enzyme-substrate interactions sites are recognized by C1s and MASP-2 for C4 and C2 (38), and kinetic studies have shown that, although C1s and MASP-2 have comparable C2 cleaving activities, MASP-2 is more efficient than C1s in cleaving C4 due to a 26-fold lower K m for the substrate (18).
In the present study, C5 activation via the lectin pathway of complement was examined by analyzing the structural/functional properties of C3/C5 convertases. Purified complement components were employed to assemble lectin pathway C3/C5 convertases on zymosan cells, a natural surface rich in mannose residues. For comparison with the classical pathway C3/C5 convertases that were assembled on sheep erythrocytes (E S ) coated with antibody (EA cells), lectin pathway C3/C5 convertases were assembled on sheep erythrocytes coated with mannan (E Man ). The data presented here show that, although kinetic properties such as K m , V max , and k cat of C3/C5 convertases generated via these two pathways of complement are similar, complement activation via the lectin pathway is at least four times more efficient than the classical pathway because every C4b deposited via the lectin pathway is capable of forming a convertase in contrast to only one of four C4b molecules deposited via the classical pathway.
Purified Proteins-The MBL⅐MASPs complex (M1) was purified from human plasma by modifying the method of Tan et al. (40) as described (41). Briefly, M1 was purified by affinity chromatography on a mannose-Sepharose followed by a maltose-Sepharose column. The M1 complex was further purified by ion-exchange chromatography on a Mono Q column using fast protein liquid chromatography as described (41). To deposit C4b on zymosan or E Man , M1 eluted from a maltose-Sepharose or a Mono Q column was used as specified. The protein concentration of M1 was determined spectrophotometrically by assuming 10.0 as the value for E 280 1% . Complement proteins, C2 (43), C3 (44,45), C4 (46), and C6 (47), were purified from normal human plasma as described in the references cited. C5 was purified by modifying the method of Discipio (48) as described (31). Complement components, C1, activated C1s, C4, and C5b,6 were obtained from Complement Technology, Inc. (Tyler, TX).
All complement proteins used in the study were homogenous by polyacrylamide gel electrophoresis. The functional activity of various purified proteins, determined by densitometric analysis of the cleaved protein on SDS-PAGE, was observed to be Ͼ95% as described (49). Protein concentrations of C2, C3, C4, C5, C6, and C5b,6 were determined spectrophotometrically as described (49). All purified proteins were stored at Ϫ76°C. M r values employed in the calculations were 204,000 for C4, 195,000 for C4b, 176,000 for C3b, 190,000 for C5, 179,000 for C5b, 120,000 for C6, 299,000 for C5b,6, 102,000 for C2, and 70,000 for C2a.
Preparation of ZymM1,C4b and ZymM1,C3bC4b Cells-ZymM1 cells were prepared by incubating 25 l of zymosan cells (7.5 ϫ 10 8 cells/ml) with 20 l of M1 (0.03 mg/ml) overnight at 22°C. The cells were washed with GVB ϩϩ to remove unbound M1, suspended in 37.5 l of GVB ϩϩ , and incubated with 12.5 l of C4 (either labeled with 125 I or unlabeled) at 0.92 mg/ml for 20 min at 22°C. The total number of C4b deposited on washed ZymM1 cells was determined by the uptake of radiolabeled C4 as C4b. ZymM1,C4b cells were washed, and the process was repeated until the desired numbers of C4b/Zym were obtained. Zymosan cells that had no M1 were employed as controls for background C4b binding.
ZymM1,C4b cells were converted to ZymM1,C3bC4b cells by incubating 0.15 ϫ 10 8 ZymM1,C4b cells with 1.0 l C2 at 0.4 mg/ml and 10 l of C3 (either labeled with 125 I or unlabeled) at 1.05 mg/ml for 10 min at 22°C. The cells were washed five times with GVB ϩϩ to remove C3 and unbound C2. The total number of radiolabeled C3b molecules bound to ZymM1,C3bC4b was determined by using [ 125 I]C3 (50) as described for alternative pathway C5 convertases (51,52). ZymM1,C3bC4b cells were washed and the procedure repeated until the desired numbers of C3b/Zym were obtained. Zymosan cells that had no M1 were employed as controls for background C3b binding.
Preparation of E Man M1,C4b and E Man M1,C3bC4b Cells-Sheep erythrocytes (E S ) were coated with mannan by the method of Ikeda et al. (53). Briefly, 200 l of E S at 1-1.5 ϫ 10 9 /ml was incubated for 5 min at 22°C with an equal volume of GVB ϩϩ containing 1 mM chromium chloride and 4 g of mannan. Cells were washed three times with GVB ϩϩ to remove unbound mannan and chromium chloride. E Man M1 cells were prepared by incubating 60 l of E Man cells (0.9 ϫ 10 9 /ml) with 30 l of M1 off a Mono Q column at 0.2 mg/ml for 1.5 h at 22°C followed by overnight incubation at 4°C in GVB ϩϩ . The cells were washed with GVB ϩϩ to remove unbound M1. E Man M1,C4b cells were prepared by incubating E Man M1 with 12.5 l of C4 (either labeled with 125 I or unlabeled) at 3.0 mg/ml for 10 min at 37°C. The total number of C4b deposited on E Man M1 cells was determined by the uptake of radiolabeled C4 as C4b. E Man M1,C4b cells were washed, and the process was repeated until the desired numbers of C4b/E Man M1 were obtained. E S cells that had no M1 were employed as controls for background C4b binding. E Man M1,C4b cells were converted to E Man M1,C3bC4b cells by incubating 0.72 ϫ 10 8 /ml E Man M1,C4b cells with 20 l of native C3 (30 g) and 10 l of C2 (1.25 g) for 30 min at 22°C. The cells were washed 5 times with GVB ϩϩ to remove unbound C2 and C3. The number of C3b deposited per E Man was determined by measuring the number of radiolabeled factor B molecules bound as Bb, the activated product of factor B, to E Man M1,C3bC4b cells as described (51). The procedure was repeated until the desired numbers of C3b/ E Man were obtained.
Formation and Assay of C5 Convertases: ZymM1,C4b,C2a, ZymM1,C3bC4b,C2a, E Man M1,C4b,C2a, and E Man M1,C3bC4b, C2a-Because the C5 convertase required less than a minute to form (Fig. 1A), the convertase was assembled in the same reaction mixture in which C5 cleavage assays were done. Enzyme velocities were determined under saturating concentrations of C2 and C6 in 0.5-ml siliconized microcentrifuge tubes. Assay mixtures contained varying concentrations of C5, C2 (0.4 g), C6 (1.0 g), 0.15 mM CaCl 2 , and 0.5 mM MgCl 2 . Reactions were started with the addition of ZymM1,C4b or ZymM1,C3bC4b cells. The concentration of cells was adjusted so as to have 0.15-1.9 ng of bound C2a in a final volume of 25 l of GVB ϩϩ resulting in 0.09 -1.1 nM enzyme in the assays. Enzyme concentration was determined by measuring the number of C2a bound as described under "Experimental Procedures" under "Determination of the C5 Convertase Concentration." After 10-min incubation at 37°C, further cleavage of C5 was prevented by transferring assay tubes to an ice bath and adding 225 l of ice-cold GVBE. Appropriately diluted assay mixtures were titrated for C5b,6 formation by hemolytic assays using E C , and the amount of C5b,6 formed was quantitated from standard curves generated with purified C5b,6 as described (31,51).
Formation and assay of lectin pathway C5 convertases assembled on E Man cells (E Man M1,C4b,C2a and E Man M1,C3bC4b,C2a) were done similarly as described above for lectin pathway C5 convertases assembled on zymosan. The concentration of cells was adjusted so as to give 0.08 -2.57 nM enzyme in the assay.
Quantitation of Reaction Products-The sensitivity of chicken erythrocytes (E C ) to human C5b,6 was employed to measure the amount of C5b,6 formed in hemolytic assays as described (31,51). Controls with C5 and C6 but no C2 were subtracted as background. To ensure that C5b,6 was not formed from the C5 and C6 in the pooled normal human serum used as a source of C7-9 during the lysis of E C , reaction mixtures containing C5 convertase but no purified C5 or C6 were used as controls. C5b,6 concentration was quantitated from a standard curve using purified C5b,6 as described previously (51).

Determination of the Number of C4b Molecules Bound per
Cell-C4b was deposited on cells by incubating ZymM1 with radiolabeled C4 at 0.92 mg/ml (specific activity ϭ 0.05 Ci/g).
To determine the total number of C4b deposited on ZymM1, the ZymM1C4b cells were washed and incubated for 10 min at 37°C in 25 l of GVB ϩϩ after which they were diluted with 75 l of GVB ϩϩ . Bound and free radiolabel were separated by layering 75 l of the mixture on 250 l of 20% sucrose in GVB ϩϩ and centrifuging for 1 min at 10,000 ϫ g at 22°C. To measure the number of radiolabeled C4b bound to cells, the tubes were cut and the amount of radioactivity in the pellet was counted (50). Nonspecific background binding of radiolabeled C4 to zymosan cells was subtracted. The process was repeated to obtain the desired number of C4b molecules per cell. Cells bearing different densities of radiolabeled C4b ranging from 60,000 to 576,000 C4b/cell were made. C4b was deposited on E Man M1 cells by incubating E Man M1 with radiolabeled C4 at 3 mg/ml (specific activity ϭ 0.057-0.085 Ci/g) for 10 min at 37°C. The number of C4b deposited per cell was calculated as described above for ZymM1 cells.
Determination of the C5 Convertase Concentration-The number of C5 convertase sites formed on ZymM1,C4b or ZymM1,C3bC4b cells was determined by measuring only those C4b molecules that were capable of forming an enzyme with radiolabel C2 under saturating conditions. ZymM1,C4b or ZymM1,C3bC4b cells were incubated for 10 min at 37°C with varying amounts of [ 125 I]C2. Bound and free radiolabel was separated by layering 75 l of the mixture on 250 l of 20% sucrose in GVB ϩϩ and centrifuging for 1 min at 10,000 ϫ g at 22°C. The concentration of C5 convertase in the experiment was calculated from the amount of C2a uptake, the specific activity of C2a, and the molecular weight of C2a as described (31). Specific C2a uptake was determined after subtracting nonspecific binding and contribution from 125 I-labeled C4b and C3b if present in the experiment. The number of C5 convertase sites formed on E Man M1,C4b or E Man M1,C3bC4b was determined as described for zymosan.
Data Analysis-The reaction velocity data were analyzed according to the Michaelis-Menten equation . The results were fit to this equation using nonlinear regression analysis and kinetic parameters, K m , V max , and k cat , were determined using GraFit version 5.06 software (Erithacus software, Staines, Middlesex, London, UK).
Preparation of Radiolabeled Proteins-C2, C3, C4, and factor B (100 g) were radiolabeled with 125 I for 30 min at 0°C in a glass tube coated with Iodogen (Pierce). Free 125 I was removed by centrifugal desalting (54). Specific activities of radiolabeled C2 and factor B ranged from 0.12 to 8.7 Ci/g. Radiolabeled C3 and C4 were diluted with unlabeled C3 or C4 to give specific activities of 0.01-0.085 Ci/g.

Measurement of the Formation and Decay of C5 Convertase-
The rate of formation of surface-bound C5 convertases was determined by assembling the enzyme under concentrations of C2 (0.33 g) in 20 l of GVB ϩϩ shown to be saturating in Fig.  1C. The reaction was started with the addition of ZymM1,C4b or ZymM1,C3bC4b. Reaction mixtures were incubated at 37°C for different time intervals (0.5, 1, 2, 3, and 4 min), and enzyme formation stopped by addition of EDTA. A mixture of C5 (0.3 g), C6 (1.0 g), and 71.4 mM EDTA in 3.5 l of GVB was added to reaction mixtures containing ZymM1,C3bC4b,C2a while a mixture of C5 (9.75 g), C6 (0.75 g), and 40 mM EDTA in 5.0 l of GVB was added to reaction mixtures containing ZymM1,C4b,C2a. These reactions were incubated for 10 min at 37°C, and the amount of C5b,6 formed was quantitated hemolytically using E C (51).
The half-life of C5 convertases was measured by preparing the enzyme for 2 min with C2 (4.3 g), 0.5 mM MgCl 2 , and 26 l of ZymM1,C4b cells (1.5 ϫ 10 8 /ml) bearing 19,000 C4b/cell in a final volume of 130 l of GVB ϩϩ . At zero time additional formation of enzyme was stopped with EDTA. At various time intervals, aliquots of the enzyme preparation (21 l) were withdrawn and assayed for remaining C5 convertase activity by adding 4.0 l containing C5 (9.0 g) and C6 (1.0 g). The amount of C5b,6 formed after 10 min at 37°C was determined hemolytically using E C . The half-life of surface-bound C5 convertase assembled with C3b-containing complexes (ZymM1,C3bC4b, C2a) was determined under assay conditions similar to those described above for ZymM1,C4b,C2a, except that ZymM1, C3bC4b bearing 64,000 C4b and 433,000 C3b per cell were employed for assembling the enzyme and a lower concentration of C5 (0.24 g) was used.

RESULTS
Formation and Decay of C5 Convertase-The rate of formation of lectin pathway C3/C5 convertases depends on the activation of C4 and C2 by MASP-2 and this may differ from C1s activation of C4 and C2 via the classical pathway of complement. Therefore, in the present study the rate of formation of lectin pathway C3/C5 convertases was determined in assays using M1 complex under saturating concentrations of C2 (129 nM) as shown in Fig. 1C, 0.15 mM CaCl 2 and 0.5 mM MgCl 2 . As seen in Fig. 1A, maximum levels of cell-bound lectin pathway C3/C5 convertases (ZymM1,C4b,C2a and ZymM1,C3bC4b,C2a) were achieved in less than a minute under these conditions. The rate of formation of C3/C5 convertases via the lectin pathway of complement is similar to that reported for the classical pathway C3/C5 convertases (EAC1,C4b,C2a and EAC1,C3bC4b,C2a) (31), indicating that under the conditions employed activation of C2 by MASP-2 in the M1 complex is similar to that observed with C1s in the C1 complex.
Because activation of the lectin pathway of complement generates C3/C5 convertases similar to that generated by the classical pathway, differences in the rate of decay of the C3/C5 convertases of the two pathways were not expected. Nevertheless, for comparison with the classical pathway C3/C5 convertases, the rate of decay of the lectin pathway C3/C5 convertases was determined. As seen in Fig. 1B the surface-bound C3/C5 convertases generated via the lectin pathway had a very short half-life as indicated by a t1 ⁄ 2 of ϳ3 min at 37°C (ZymM1,C4b,C2a ϭ 3.4 min and ZymM1,C3bC4b,C2a ϭ 3.2 min) similar to that reported for C3/C5 convertases (EAC1,C4b,C2a ϭ 2.0 min and EAC1,C3bC4b,C2a ϭ 2.5 min) generated via the classical pathway (31).
Constant C5 Convertase Levels during Assays-During the 10-min assay period employed for determining initial velocities of C5 cleavage by lectin pathway C3/C5 convertases, the enzyme would decay due to its short half-life of ϳ3 min (Fig.  1B). It was therefore important to maintain a constant level of enzyme during the assay. Preliminary experiments were done to determine the C2 concentration required for maintaining constant amounts of lectin pathway C3/C5 convertases during a 10 min assay. As seen in Fig. 1C, with the C2 amounts used in the assay, a constant amount of the product C5b,6 was generated throughout the assay. These results demonstrate that a constant amount of enzyme was maintained during a 10-min assay even though the lectin pathway C3/C5 convertase was decaying and reforming. These C2 concentrations were used in all assays unless otherwise stated.
Determination of Number of Cell Bound C4b and the Concentration of C5 Convertase Formed on Zymosan Cells-The number of lectin pathway-generated C3/C5 convertases was determined from the number of cell bound C4b that could form a C5 convertase measured as the number of C2a-binding sites. First, the number of C4b deposited per cell was determined from the uptake of radiolabeled C4 by ZymM1 cells. The number of C5 convertase sites that were generated on the ZymM1,C4b cells FIGURE 1. Characterization of the C5 convertase assay. A, rate of formation of lectin pathway C5 convertases. C5 convertases were assembled in reaction mixtures containing saturating concentrations of C2 in GVB ϩϩ . The reaction was started with the addition of ZymM1,C4b or ZymM1,C3bC4b cells, and the reaction mixture was incubated for different time intervals at 37°C as indicated in the figure to allow formation of enzyme. EDTA was added at the indicated time interval to stop additional formation of enzyme. The amount of enzyme formed was determined by adding a mixture of C5 and C6 to the reaction mixture, and the amount of C5b,6 formed in the subsequent 10-min incubation at 37°C was quantitated hemolytically using E C as described (51). E, ZymM1,C4b,C2a; ‚, ZymM1,C3bC4b,C2a. B, rate of decay of lectin pathway C5 convertases. C5 convertases were prepared for 2 min at 37°C. After adding GVBE to prevent additional formation of enzyme, the enzyme was allowed to decay at 37°C. Aliquots were removed at the indicated times and assayed for remaining C5 convertase activity by incubating with C5 and C6. C5b,6 formed was quantitated using lysis of E C and EDTA containing pooled normal human serum as a source of C7-C9. E, ZymM1,C4b,C2a; ‚, ZymM1,C3bC4b,C2a. The data obtained were fit to a single exponential decay equation using GraFit version 5.06 Erithacus software. C, demonstration that a constant amount of C5 convertase activity was maintained during the assays by using saturating concentrations of C2, 0.5 mM MgCl 2 (E, ZymM1,C4b,C2a; ‚, ZymM1,C3bC4b,C2a). Reaction mixtures of the convertases were incubated for the indicated length of time. C5 and C6 were then added, and C5b,6 formation was allowed to continue for the next 5 min. The data obtained were fit by linear regression using GraFit version 5.06 Erithacus software.
was then determined as the number of C2a binding sites. As seen in Fig. 2A, the binding of [ 125 I]C2a to C4b-coated ZymM1 cells bearing 78,000 C4b/cell was saturable. From the maximum amount of radiolabeled C2a bound to ZymM1 cells, the number of C5 convertase sites generated was calculated. Increasing the density of C4b on the surface of zymosan by ϳ10-fold (60,000 C4b to 576,000 C4b per zymosan) did not change the ratio of C2a per C4b as shown in Fig. 2B. The average ratio of C2a per C4b was found to be 0.95 Ϯ 0.19. These results suggest that every C4b molecule deposited via the lectin pathway is capable of forming a C5 convertase. These findings are in contrast to those reported for the classical pathway C3/C5 convertases in which the ratio of C2a per C4b was found to be 0.28 Ϯ 0.10 indicating that only one out of every four C4b molecules deposited via the classical pathway was capable of forming a C3/C5 convertase (Fig. 2B in Ref. 31).
Determination of Number of Cell Bound C4b and the Concentration of C5 Convertase Formed on E Man Cells-Because the ratio of C2a per C4b for lectin pathway C3/C5 convertases assembled on zymosan was found to be 4-fold greater than reported for the classical pathway C3/C5 convertases that were assembled on sheep erythrocytes coated with antibody (EA), we wanted to determine if the surface on which lectin pathway C3/C5 convertases were assembled had any influence on the ratio of C2a per C4b. Therefore lectin pathway C3/C5 convertases were assembled on sheep erythrocytes coated with mannan (E Man ) and the number of C2a binding sites determined as described for ZymM1 cells. As seen in Fig. 3A Table  1). The result indicates that the lectin pathway C5 convertase has a weak affinity for C5 when compared with the normal concentration of C5 in blood (0.37 M). The k cat of the enzyme was calculated from the V max obtained from the velocity versus substrate curve (Fig. 4A) using the equation, k cat ϭ V max /[enzyme concentration], and the data obtained are shown in Table  1. The enzyme concentration was determined by measuring the maximum [ 125 I]C2a bound as shown in Fig. 2A. The average k cat was 0.011 Ϯ 0.003 s Ϫ1 (Table 1). Increasing the density of C4b on the cell surface by ϳ16-fold (21,000 to 328,000 C4b/ Zym) did not influence the K m (Fig. 4B) or the k cat (Fig. 4C) of C3/C5 convertases of the lectin pathway.
Lectin pathway C3/C5 convertases were also assembled on sheep erythrocytes that were coated with mannan (E Man ). Determination of kinetic constants of the C5 convertase (E Man M1,C4b,C2a) assembled on E Man indicated properties similar to that obtained with the convertase assembled on zymosan. As seen in Fig. 5 and Table 1, the average K m of E Man M1,C4b,C2a for C5 was 6.88 Ϯ 4.82 M at C4b/E Man while the average catalytic rate of the enzyme was 0.0082 Ϯ 0.0054 s Ϫ1 . These kinetic parameters were ϳ2-fold of that observed FIGURE 2. Determination of number of C5 convertase sites on ZymM1,C4b cells. A, the number of bound C4b that were capable of forming a convertase was measured by quantitating the amount of radiolabeled [ 125 I]C2a (3.2 ϫ 10 6 cpm/g) bound to ZymM1,C4b. The binding assays were performed using excess C2 and 0.5 mM MgCl 2 . Although measurements with cells bearing many different densities of C4b/zymosan were made, data only for enzyme assembled with 78,000 C4b/zymosan are shown. After 10 min at 37°C, the cells were washed by centrifugation through 20% sucrose. The amount of C2a bound was calculated from the radioactivity in the pellets and the specific activity of the C2a subunit. The data obtained were fit to one-site binding equation using GraFit version 5.06 Erithacus software. B, comparison of the number of C5 convertase sites versus the number of C4b/cell. The number of C4b bound to ZymM1 cells was determined by quantitating the amount of radiolabeled 125 I-labeled C4b bound. The amount of cell-bound C4b capable of forming a convertase was then determined by quantitating the amount of 125 I-labeled C2a bound to ZymM1,C4b cells as described in 2A. The amount of C2a bound was calculated from the radioactivity in the pellets and the specific activity of the C2a subunit after subtracting the contribution from 125 I-labeled C4b, and both were corrected for nonspecific binding. The data obtained were fit by linear regression using GraFit version 5.06 Erithacus software. when the lectin pathway enzyme was assembled on zymosan (Table 1). These results suggest that the type of surface employed for assembling the lectin pathway C3/C5 convertase does not influence the K m or the k cat of the lectin pathway C5 convertase. Likewise, increasing the density of C4b on E Man was observed to have no influence on the K m or k cat of the enzyme (data not shown).

Kinetic Parameters of High Affinity Surface-bound C3/C5 Convertase (ZymM1,C3bC4b,C2a and E Man M1,C3bC4b,C2a)-
The C4b-dependent lectin pathway C3/C5 convertases exhibited a weak affinity for the substrate C5. Deposition of additional C3b molecules have been shown to convert the low affinity C5 convertase of the classical pathway (C4b,C2a) to a high affinity C5 convertase (31). In the present study the effect of C3b molecules on the affinity for C5 and k cat of the lectin pathway enzyme were examined. Kinetic parameters of lectin pathway C3/C5 convertase assembled with C3b-containing complexes (C3b-C4b complexes) as the noncatalytic subunit were determined. Because the MASPs in the M1 complex do not cleave C3, ZymM1,C3bC4b cells were prepared by allowing the lectin pathway C3 convertase, ZymM1,C4b,C2a to cleave native C3 and deposit C3b on and around themselves. After C3b deposition, cells were washed to remove excess C3. ZymM1,C3bC4b cells were incubated for 15 min at 37°C to ensure decay of any remaining catalytic subunit, C2a, on the convertase. Initial velocity data obtained for C5 cleavage by C5 convertases assembled with saturating levels of C2 and ZymM1,C3bC4b cells that had 55,000 C4b and 180,000 C3b per cell, performed as described above, indicated a K m in the nanomolar range for C5 (7.7 nM) as shown in Fig. 6A. This was observed for all C5 convertases (K m range was 6.7-12.1 nM; average K m was 8.6 nM, Table 1) assembled at different levels of C4b/zymosan ranging from 39,300 to 78,000 C4b/Zym.
The lectin pathway convertase, E Man M1,C3bC4b,C2a, assembled on E Man exhibited kinetic properties similar to those observed for the enzyme assembled on the natural activating surface of zymosan ( Table 1). As seen in Fig. 6B, the velocity versus substrate plots for E Man M1,C3bC4b,C2a showed a very good fit to the theoretical curve based on the Michaelis-Menten equation as well as to the linearized form of the equation (Eadie-Hofstee plot) shown as an inset in Fig. 6B. The average K m of E Man M1,C3bC4b,C2a for C5 (7.5 Ϯ 7.9 nM, Table 1) was  Fig. 2. The rate of C5b,6 formation was measured by hemolytic assays as described under "Experimental Procedures." Data were fit by nonlinear regression according to the Michaelis-Menten equation to determine K m and V max values using GraFit version 5.06 Erithacus software. Inset, analysis of C5 convertase initial velocity data using the Eadie-Hofstee equation. B, effect of C4b density on K m of C3/C5 convertase. The figure compares the K m of C3/C5 convertases ZymM1,C4b,C2a (E) and ZymM1,C3bC4b,C2a (‚) assembled with varying amounts of C4b/cell. The number of C4b/cell was measured using 125 I-labeled C4b as described in Fig. 2B. Initial rates of C5 cleavage were measured as described in Fig. 1. The data obtained were fit by linear regression using GraFit version 5.06 Erithacus software. C, comparison of the k cat of C5 convertases ZymM1,C4b,C2a (E) and ZymM1,C3bC4b,C2a (‚) formed with cells bearing different amounts of C4b/cell as indicated. The k cat for each C5 convertase was calculated from the V max obtained from individual velocity versus substrate concentration plots and the enzyme concentration determined from the number of 125 I-labeled C2a bound/cell under saturating concentrations of C2. The data obtained with the two surface-bound C5 convertases, ZymM1,C4b,C2a and ZymM1,C3bC4b,C2a, were considered as one set of data and fit by linear regression using GraFit version 5.06 Erithacus software. similar to that observed when the enzyme was assembled on zymosan ( Table 1). The very low K m of the lectin pathway C3bcontaining convertases assembled on E Man or zymosan (average K m ϭ 7.5 and 8.6 nM, respectively, Table 1) indicates a high affinity interaction with C5 that is stronger by about three orders of magnitude when compared with convertases assembled with C4b alone. The average catalytic rate of E Man M1,C3bC4b,C2a (0.026 Ϯ 0.012 s Ϫ1 , Table 1) was within 2-fold of that observed when the enzyme was assembled on zymosan (0.013 Ϯ 0.003 s Ϫ1 , Table 1). These results suggest that the type of surface employed for assembling the lectin pathway C3/C5 convertase does not influence the K m or the k cat of the lectin pathway C5 convertase.
Effect of C3b Density on Kinetic Parameters of Surface-bound C3/C5 Convertase-Because deposition of C3b molecules on the cell surface was required to generate high affinity C5 convertases, the results suggested that C5 convertases with varying affinities for C5 exist. And because the two convertases differ a thousand fold in their affinity for C5, the Eadie-Hofstee plot would have a curvature to it. To investigate this possibility, kinetic properties of convertases assembled with varying ratios of C3b:C4b were examined. Initial velocities for C5 cleavage were measured in the presence of saturating concentrations of C2, excess C6, and varying concentrations of C5 ranging from 0.0013 M to 13.7 M. As seen in Fig. 7, the data obtained with ZymM1,C3bC4b cells that had a C3b:C4b ratio of 1.3:1 (i.e. 220,000 C3b/cell and 146,000 C4b/cell) did not fit well to the theoretical curve based on the Michaelis-Menten equation, which assumes a homogenous enzyme (Fig. 7A). The Eadie-Hofstee plot had a strong curvature (Fig. 7B) suggesting the presence of C5 convertases with different affinities for C5. The initial and final slopes of the Eadie-Hofstee plot suggested two C5 convertase species with a K m of 8.8 nM for the high affinity and 0.84 M for the low affinity enzyme, in agreement with the other kinetic data presented here. The biphasic nature of the Eadie-Hofstee plot was not observed in experiments that employed cells with a C3b:C4b ratio greater than 4:1. Considered together, these results suggest that under the con-ditions employed deposition of C3b molecules greater than four times the number of C4b molecules results in the conversion of most low affinity C3/C5 convertases to high affinity C5 convertases.
Comparison of C5 Convertase Activities at the Normal Plasma Concentration of C5-A comparison of the normal plasma concentration of C5 with the activity of various forms of C5 convertase assembled with cell-bound C4b alone or with C3b-containing complexes is shown in Fig. 8. Because the K m of  A, kinetic analysis of C3/C5 convertase, ZymM1,C3bC4b,C2a assembled on zymosan. High affinity C5 convertases were generated by allowing preformed C3/C5 convertases (ZymM1,C4b,C2a) to deposit C3b on and around themselves by providing native C3 during preparation. C3b was deposited by incubating 1.5 ϫ 10 7 ZymM1,C4b cells (bearing 55,000 C4b/cell) with C2 (0.4 g) and C3 (10 g) in a final volume of 16 l of GVB ϩϩ for 5 min at 22°C. The cells were washed with GVB ϩϩ . ZymM1,C4b cells now bearing 180,000 C3b/cell were employed for assembling C5 convertases. The number of C3b per zymosan was determined by measuring the number of radiolabeled factor B molecules bound to ZymM1,C3bC4b cells as described (51). C5 cleavage was measured by incubating reaction mixtures containing saturating concentrations of C2 (0.4 g), C6 (1.0 g), and 0.5 mM MgCl 2 and the indicated concentrations of C5 in a final volume of 25 l of GVB ϩϩ . Reactions were initiated by the addition of ZymM1,C3bC4b cells bearing 0.23 nM C5 convertase (measured as in Fig. 2). Data were fit by nonlinear regression according to the Michaelis-Menten equation to determine K m and V max values using GraFit version 5.06 Erithacus software. Inset, analysis of C5 convertase initial velocity data using the Eadie-Hofstee function. B, kinetic analysis of surface-bound C4b-dependent C3/C5 convertase, E Man M1,C3bC4b assembled with 3,100 C4b/cell and 124,000 C3b/ cell. Initial velocities for C5 cleavage were measured similarly as described above for ZymM1,C3bC4b,C2a in Fig. 6A. E Man M1,C3bC4b cells bearing 0.16 nM C5 convertase were used to initiate C5b,6 formation.
surface-bound C3/C5 convertases assembled on zymosan or E Man with C4b as the noncatalytic subunit is well above the normal physiologic concentrations of C5 in plasma (0.37 M, vertical dashed line in Fig. 8) these C5 convertases will be relatively inefficient in binding C5 and therefore inefficient in cleaving C5. However, these C3/C5 convertases primarily cleave C3 and the resulting deposition of C3b lowers their K m for C5 below the normal plasma concentration of C5 in blood. Thus the C3b-containing enzyme complexes, ZymM1,C3bC4b and E Man M1,C3bC4b,C2a, will be occupied by C5 most of the time, and the convertase will cleave C5 at a catalytic rate close to V max (Fig. 8).

DISCUSSION
In the present study, C5 activation via the lectin pathway was examined by analyzing the structural/functional properties of the lectin pathway C3/C5 convertases and comparing to those previously reported for the classical pathway (31). Kinetic analysis of lectin pathway C3/C5 convertases assembled on zymosan particles revealed a major difference in the number of C4b molecules capable of forming a convertase. In contrast to our published studies on the classical pathway C3/C5 convertases where only one out of four C4b molecules deposited on sheep erythrocytes coated with antibody (EA) formed a convertase (31), in the present study every C4b deposited via the lectin pathway formed a convertase. This was evident from the ratio of C2a/C4b, which was 0.95 Ϯ 0.19 (one C2a per one C4 molecule) at all levels of C4b/zymosan (Fig. 2B). In addition, lectin pathway C3/C5 convertases that were assembled on sheep erythrocytes coated with mannan (E Man M1) also exhibited a C2a/C4b ratio of 0.93 Ϯ 0.21 (one C2a per one C4b molecule, Fig. 3B) similar to that obtained when the enzyme was assembled on zymosan (Fig. 2B). The results indicate that the surface on which lectin pathway C3/C5 convertases are assembled has no influence on the number of C4b molecules capable of forming a convertase and suggest that the lectin pathway has the potential of generating four times more C3/C5 convertases than the classical pathway.
Several factors may contribute to the efficient formation of C3/C5 convertases via the lectin pathway. Kinetic studies by Wallis et al. have shown that the affinity of MASP-2 for C2 in complex with C4b (K m 0.72 M) is 7-fold lower than for C2 alone (K m ϭ 5.1 M), and based on molecular interactions between MASP-2 and C4b, the study reported a slow rate of dissociation of C4b from MASP-2 (t1 ⁄ 2 ϭ 11.5 s) compared with the half-life of the reactive thioester in C4b (Յ0.7 s) (55). Assuming that human C4 activation is comparable to that of rat C4, the study proposed that the relatively slow dissociation of C4b from MASP-2 might help co-localize C4b attachment near the MBL⅐MASP-2 complex on the cell surface. This in turn would facilitate the successive cleavage by the same MASP-2 of C2 bound to the C4b molecule, and hence every C4b deposited via the lectin pathway will be capable of forming a C3/C5 convertase.
Efficient formation of C3/C5 convertases via the lectin pathway may also depend on the stoichiometry of the MBL⅐MASP-2 complex, but the low concentration of MASPs present in normal human serum (56)   . Effect of C3b density on kinetic parameters of surface-bound C3/C5 convertase. A, high affinity C5 convertases (ZymM1,C3bC4b,C2a) were generated by allowing preformed C3/C5 convertases (ZymM1,C4b,C2a) to deposit a limited number of C3b on and around themselves by providing limited amount of native C3 during preparation. C3b was deposited by incubating 0.2 ϫ 10 8 ZymM1,C4b bearing 146,000 C4b/cell with C2 (0.33 g) and C3 (11 g) in a final volume of 16 l of GVB ϩϩ for 10 min at 22°C resulting in the formation of ZymM1,C3bC4b. These ZymM1,C3bC4b cells now bearing 220,000 C3b and 146,000 C4b per cell were employed for assembling C5 convertases. C5 cleavage was measured by incubating reaction mixtures containing saturating concentrations of C2 (0.33 g), C6 (1.0 g), and 0.5 mM MgCl 2 and the indicated concentrations of C5 in a final volume of 25 l of GVB ϩϩ . Data were fit by nonlinear regression according to the Michaelis-Menten equation using GraFit version 5.06 Erithacus software. B, graphical analysis of C5 convertase initial velocity data using the Eadie-Hofstee function.  between MBL and MASP-2 (10). Stoichiometry studies on rat MBL⅐MASP complexes by Chen et al. have shown that rat MBL-A dimer binds to a single MASP-2 homodimer, whereas trimers and tetramers of rat MBL-A were shown to bind two MASP-2 homodimers when a large molar excess of MASP fragments over MBL/A was used (36). Although MBL⅐MASP-2 complexes have been shown to form both 1:1 and 1:2 complexes, the 1:1 complex has been reported to be more stable than the 1:2 complexes (9,10,36).
Structural organization studies by Arlaud's group on human MASPs have shown MASP-1, MASP-2, and MAP19 each associate to form homodimers (37) similar to that reported for rat MASP-1 and MASP-2 (36). Based on the crystal structure of the CUB1-EGF-CUB2 region of human MASP-2, Feinberg et al. proposed a model in which a MASP-2 dimer was shown to easily dock inside a MBL dimer with the serine protease domain of MASP-2 near the flexible ␣-helical neck joint (15). Kinetic studies have shown that although C1s and MASP-2 have comparable C2-cleaving activities, MASP-2 is more efficient than C1s in cleaving C4 due to a 26-fold lower K m for the substrate (K m ϭ 0.074 M and 1.92 M, respectively) (18). Thus even though the stoichiometry of the MBL⅐MASP-2 complex of the lectin pathway might be similar to that reported for the C1 complex of the classical pathway in which one C1q hexamer binds to one C1s dimer (34,59), the reports suggest that the lectin pathway is more efficient than the classical pathway in cleaving C4.
Structural studies by Harmat et al. have reported human MASP-2 to possess the most flexible CCP2/SP junction among the related proteins of the C1r/C1s/MASPs enzyme family (38). Lu et al. have reported that the two joints in MBL, one at the interruption in the collagen region with Gly-X-Y repeats, which results in the kink, and the second between the collagen region with Gly-X-Y repeats and the ␣-helical neck regions, are flexible and make the individual MBL monomer stalks spread out from each other (15,60). In contrast, the kink in the collagen region of C1q has been reported to have limited flexibility (61). These observations suggest that the limited flexibility of C1q may restrict C1s from reaching out to the C4b molecules that are further away from the initiating C1qrs complex. Although these C4b molecules that are further away from the initiating C1qrs complex may bind a C2, they are less likely to get converted to C4bC2a complexes and as a result only one out of four C4b molecules deposited via the classical pathway is capable of forming a convertase (31). Studies of Ziccardi et al. have estimated that only thirty-five C4 molecules and four C2 molecules are activated for one activating C1 complex in serum (62). This rate of turnover (seven C4 molecules for one C2 molecule) is apparently sufficient to trigger activation of the classical pathway.
In the present study analysis of the enzymatic properties of lectin pathway C3/C5 convertases, assembled on zymosan cell surface or on sheep erythrocytes coated with mannan, show that kinetic parameters such as K m and substrate specificity (Table 1) are similar to those reported for the classical pathway C3/C5 convertases (31). This was evident from the kinetic properties of the C4b-dependent C3/C5 convertase assembled on zymosan cells (ZymM1,C4b,C2a), which exhibited a weak affinity for C5 indicated by a high K m (2.73 M) ( Table 1). Increasing the density of C4b on the cell surface by ϳ16-fold for zymosan (21,000 to 328,000 C4b/cell) did not generate high affinity C5 convertase, because similar K m values were obtained for C3/C5 convertases assembled on zymosan with different densities of C4b molecules (Fig. 4B). Kinetic properties of the lectin pathway C4b-dependent C3/C5 convertase assembled on the surface of E Man cells (E Man M1,C4b,C2a, K m ϭ 6.88 M) (Fig.  5, Table 1) was observed to be similar to those assembled on zymosan cells (ZymM1,C4b,C2a) suggesting that the type of surface employed to assemble the convertases plays no role. The findings on the C4b-dependent lectin pathway C5 convertases assembled on zymosan or E Man cells are similar to those reported for the C4b-dependent the classical pathway C5 convertase (EAC1,C4b,C2a) assembled on EA cells (Table 1) (31). The results also show that, because the C3/C5 convertases assembled via the lectin pathway of complement have a weak affinity for the substrate C5, they will primarily cleave C3, depositing C3b on and around the convertase.
Deposition of C3b was not observed in our study when ZymM1,C4b or E Man M1,C4b cells were incubated with C3. This may be attributed to the weak activity of MASP-1 reported for C3 cleavage (13). Deposition of C3b was achieved only when preformed C3/C5 convertases (ZymM1,C4b,C2a and E Man M1,C4b,C2a) assembled on zymosan and E Man cells were allowed to cleave native C3. C3b deposition on and around the low affinity C3/C5 convertases (ZymM1,C4b,C2a and E Man M1,C4b,C2a) generated C3b-containing complexes (ZymM1,C3bC4b and E Man M1,C3bC4b). Analysis of the C5 cleaving properties of the C3b-containing C4b-dependent C3/C5 convertase (ZymM1,C3bC4b,C2a and E Man M1,C3bC4b,C2a) showed a ϳ1000-fold greater affinity for C5 than the convertases lacking C3b as indicated by a 1000-fold decrease in the K m (Figs. 4C, 6A, and 6B and Table  1). The average K m (8.6 to 7.5 nM) was well below the normal plasma concentration of C5 in blood (0.37 M). These findings are similar to those reported for the C3b-containing C4b-dependent classical pathway high affinity C5 convertase (EAC1,C3bC4b,C2a) assembled on EA cells (Table 1) (31). The results show that the lectin pathway C5 convertase will be 90% or more saturated with C5 under normal physiological conditions and will cleave C5 at velocities close to the maximum velocity (V max ) (Fig. 8). The low affinity C3/C5 convertase of the lectin pathway is converted to high affinity C5 convertase by the addition of C3b, and the enzyme now primarily activates C5 instead of C3 producing the cytolytic C5b-9 complex.
Although a 1000-fold difference in the affinity for the substrate C5 (K m ) of the lectin pathway C4b-dependent C3/C5 convertase and C3b-containing C4b-dependent high affinity C5 convertase was observed, the k cat of the two convertases (0.0082 s Ϫ1 to 0.011 s Ϫ1 for the C4b-dependent C5 convertases and 0.013 s Ϫ1 to 0.026 s Ϫ1 for the C3b-containing C4bdependent C5 convertases, Table 1) were within 2-fold of each other. The average k cat was 0.015 Ϯ 0.008 s Ϫ1 , indicating that the lectin pathway C5 convertase has a turnover number of 0.9 C5 molecules cleaved per min per enzyme. The average k cat (0.015 Ϯ 0.008 s Ϫ1 ) and turnover number (0.9 C5 molecule cleaved/min) of the lectin pathway C5 convertases are within 2-fold of the average k cat (0.024 Ϯ 0.010 s Ϫ1 ) and turnover number (turnover number ϭ 1.44/min) reported for the classical pathway C5 convertases) (see "Discussion" in Ref. 31). Although the catalytic rate of the lectin pathway C3/C5 convertases does not contribute to the efficiency of the pathway every C4b deposited via the lectin pathway is capable of forming a convertase and thus the lectin pathway will generate four times more cytolytic C5b-9 complex and chemotactic fragment C5a, than the classical pathway. Because the three naturally occurring structural variants of MBL (MBL/B, -C, and -D) have been reported to activate complement poorly when compared with wild-type MBL/A (58,63,64), the results imply that activation of the lectin pathway and hence production of pro-inflammatory products C3a, C4a, and C5a that are required to mount an effective immune response to infections would be weak in individuals who are carriers for MBL alleles thus making them more susceptible to recurrent infections than individuals with wild-type MBL.