The Two Major Oligomeric Forms of Human Mannan-Binding Lectin: Chemical Characterization, Carbohydrate-Binding Properties, and Interaction w
http://www.100md.com
免疫学杂志 2005年第5期
Abstract
Mannan-binding lectin (MBL) is an oligomeric C-type lectin assembled from homotrimeric structural units that binds to neutral carbohydrates on microbial surfaces. It forms individual complexes with MBL-associated serine proteases (MASP)-1, -2, -3 and a truncated form of MASP-2 (MAp19) and triggers the lectin pathway of complement through MASP-2 activation. To characterize the oligomerization state of the two major MBL forms present in human serum, both proteins were analyzed by mass spectrometry. Mass values of 228,098 ± 170 Da (MBL-I) and 304,899 ± 229 Da (MBL-II) were determined for the native proteins, whereas reduction of both species yielded a single chain with an average mass of 25,340 ± 18 Da. This demonstrates that MBL-I and -II contain 9 and 12 disulfide-linked chains, respectively, and therefore are trimers and tetramers of the structural unit. As shown by surface plasmon resonance spectroscopy, trimeric and tetrameric MBL bound to immobilized mannose-BSA and N-acetylglucosamine-BSA with comparable KD values (2.2 and 0.55 nM and 1.2 and 0.96 nM, respectively). However, tetrameric MBL exhibited significantly higher maximal binding capacity and lower dissociation rate constants for both carbohydrates. In contrast, no significant difference was detected for binding of the recombinant MASPs or MAp19 to immobilized trimeric or tetrameric MBL. As shown by gel filtration, both MBL species formed 1:2 complexes with MASP-3 or MAp19. These results provide the first precise analysis of the major human MBL oligomers. The oligomerization state of MBL has a direct effect on its carbohydrate-binding properties, but no influence on the interaction with the MASPs.
Introduction
Mannan-binding lectin (MBL)3 is a member of the collectin family, a group of proteins that comprise a collagen-like region and a C-type lectin domain and are involved in innate immunity against infection (2, 3). MBL recognizes patterns of neutral carbohydrates such as mannose and N-acetylglucosamine (GlcNAc) on a wide range of microbial surfaces and is the only collectin able to initiate activation of the lectin pathway of complement (4, 5). Circulating MBL forms complexes with three MBL-associated serine proteases (MASP-1, MASP-2, and MASP-3) (6, 7, 8), its association with MASP-2 yielding a complex that triggers complement activation through cleavage of components C4 and C2 (9, 10, 11). In addition to the MASPs, serum MBL binds to a truncated form of MASP-2 called 19-kDa MBL-associated protein (MAp19) or small MBL-associated protein (12, 13). The biological role of MAp19 as well as the physiological substrates of MASP-1 and MASP-3 remain to be identified.
Human MBL is assembled from a single polypeptide chain, consisting of a short N-terminal cysteine-rich region, a collagenous region comprising 19 repeating Gly-X-Y triplets with an interruption at the 8 triplet, a 34-residue hydrophobic stretch, and a C-terminal C-type lectin domain. Three polypeptide chains associate to form a homotrimeric structural unit comprising a collagen-like triple helix, an -helical coiled-coil called the neck region, and three carbohydrate recognition domains (5). The three cysteines in the N-terminal region form interchain disulfide bonds that mediate formation of oligomers comprising two to six structural units and exhibiting bouquet-like structures when viewed by electron microscopy (14). The collagen-like triple helices associate laterally at their N-terminal ends and then diverge at the level of the interruption in the Gly-X-Y sequence, thus defining a hinge in the molecule.
There are large variations in the MBL serum concentration, arising from polymorphism in the promoter region as well as in the structural moiety of the MBL gene (15), and several clinical studies have provided evidence for a link between a low MBL titer and increased susceptibility to various infectious diseases (16). The structural mutations located in exon 1 of the MBL gene generate amino acid substitutions in the collagen-like region that are believed to impair oligomerization of the protein and result in functional deficiency (15, 17, 18, 19). The reduced ability of these mutants to activate complement has been proposed to result from failure to bind to the MBL-associated serine proteases (20), and from the reduced carbohydrate-binding ability of the lower order MBL oligomers (17, 18, 21).
Purified serum MBL appears to be a mixture of oligomers of different sizes with two major forms, termed MBL-I and MBL-II (8), that can be separated using anion-exchange chromatography (8, 22). The objective of the present study was to precisely characterize the oligomerization state of both forms, their carbohydrate-binding ability, and their interaction properties with the MASPs and MAp19. Mass spectrometry analysis provides unambiguous evidence that MBL-I and -II are trimers and tetramers of the structural unit, respectively. The oligomerization state of MBL has a direct effect on its carbohydrate-binding properties, with the tetrameric form showing increased binding capacity and higher affinity, but has no significant influence on the interaction with the MASPs and MAp19 or on the stoichiometry of the resulting complexes.
Materials and Methods
Materials
N-Acetylglucosamine-BSA (10 average carbohydrate residues per BSA molecule) was purchased from Dextra Laboratories. Mannose (Man)- and galactose-BSA (30 average carbohydrate residues per BSA molecule) were from EY Laboratories.
Proteins
MBL was purified from human serum as described by Thielens et al. (23). Separation of MBL-I and MBL-II was performed by anion-exchange chromatography on a MonoQ HR 5/5 column (Amersham Pharmacia Biotech) as described by Dahl et al. (8) with the following modifications. The column was run at a flow rate of 0.5 ml/min, and MBL fractionation was achieved with a NaCl gradient to 250 mM in 40 min. Fractions of 1 ml were collected and analyzed for the presence of MBL-I and MBL-II by SDS-PAGE and Coomassie Blue staining. Recombinant MASP-1, MASP-2, and MAp19 were expressed using a baculovirus/insect cell system and purified as described previously (23, 24). Recombinant MASP-3 was expressed using the same system and purified as described by Zundel et al. (25) with the following changes. After the first anion-exchange chromatography step, fractions containing MASP-3 were dialyzed against 50 mM NaCl, 1 mM CaCl2, 50 mM triethanolamine-hydrochloride, pH 6.9, and loaded at 1.0 ml/min onto a MonoQ HR 5/5 column (Amersham Biosciences) equilibrated in the same buffer. Elution was conducted with a linear NaCl gradient to 500 mM in 1 h. MASP-3-containing fractions were identified by SDS-PAGE analysis; they were dialyzed against 145 mM NaCl, 1 mM CaCl2, 50 mM triethanolamine hydrochloride, pH 7.4; and further purification was achieved by high pressure gel permeation on a TSK G3000 SWG column (7.5 x 600 nm) (Tosohaas) equilibrated in the same buffer.
The concentrations of purified MBL-I, MBL-II, MASP-3, and MAp19 were determined using molar extinction coefficients at 280 nm of 198,000, 264,000, 160,600, and 29,000 M–1 x cm–1, respectively, which were determined experimentally by amino acid analysis. The molecular masses of MBL-I (228,098 Da), MBL-II (304,899 Da), MASP-3 (87,524 Da), and MAp19 (19,088 Da) were determined by MALDI-TOF mass spectrometry analysis. Due to the low amount of material recovered, estimation of the concentration of recombinant MASP-1 and MASP-2 was based on Coomassie Blue staining after SDS-PAGE analysis using MASP-3 and MAp19 as internal standards and molecular masses of 82,000 and 75,100 Da, respectively (24).
Amino acid and N-terminal sequence analyses
Amino acid analyses were performed on acid hydrolysates of the samples using a Beckman 7300 amino acid analyzer as described previously (26). N-terminal sequence analyses were performed using an Applied Biosystems model 477A protein sequencer as described previously (27).
MALDI-TOF mass spectrometry analysis
Mass spectra were obtained using a Perseptive Biosystems Voyager Elite Xl time of flight mass spectrometer with delayed extraction, operating with a pulsed nitrogen laser at 337 nm. Positive ion mass spectra were acquired using linear, delayed extraction mode with an accelerating potential of 25 kV, an 85% grid potential, a 0.3% guide wire voltage, and a delay time of 600 ns. Samples were mixed with an equal volume of a saturated solution of sinapinic acid prepared in 50% (v/v) acetonitrile, 0.1% trifluoroacetic acid on the stainless steel sample plate and air-dried before analysis. External calibration was performed with BSA using m/z values of 66,431 and 33,216 for the mono- and doubly-charged species, respectively. The accuracy of molecular mass determination was estimated at 0.075%.
Gel permeation chromatography
Complexes between trimeric or tetrameric MBL on one hand and MASP-3 or MAp19 on the other were assembled by incubation of the proteins for 1 h at room temperature and analyzed by gel permeation chromatography on a Superose 6 10/300 GL column (Amersham Biosciences) equilibrated in 145 mM NaCl, 1 mM CaCl2, 50 mM triethanolamine hydrochloride, pH 7.4, and run at 0.5 ml/min. Proteins were detected from their optical density at 280 nm.
Real time surface plasmon resonance spectroscopy and data evaluation
Analyses were performed using a Biacore 3000 instrument (Biacore AB). Carbohydrate-BSA glycoconjugates and MBL were diluted to 10 and 30 μg/ml, respectively, in 10 mM sodium acetate, pH 4.0, and immobilized on the surface of a CM5 sensor chip (Biacore AB) using amine-coupling chemistry as described by Zundel et al. (25). Binding of trimeric and tetrameric MBL to immobilized Man-BSA (2400 resonance units (RU)), GlcNAc-BSA (3000 RU), or Gal-BSA (2300 RU) was measured at a flow rate of 20 μl/min in 145 mM NaCl, 1 mM CaCl2, 50 mM triethanolamine hydrochloride, pH 7.4, containing 0.005% surfactant P20 (Biacore AB). Binding of MASP-1, MASP-2, MASP-3, and MAp19 to immobilized trimeric MBL (9,000 RU) and tetrameric MBL (10,000 RU) was measured under the same conditions. Equivalent volumes of each protein sample were injected over a surface with immobilized BSA to serve as blank sensorgrams for subtraction of the bulk refractive index background. Regeneration of the surfaces was achieved by injection of 10 μl of the running buffer containing 5 mM EDTA instead of 1 mM CaCl2.
Data were analyzed by global fitting to a 1:1 Langmuir binding model of both the association and dissociation phases for several concentrations simultaneously, using the BIAevaluation 3.1 software (Biacore AB). The apparent equilibrium dissociation constants (KD) were calculated from the ratio of the dissociation and association rate constants (koff/kon). Maximal binding capacities (Rmax) were determined using the same model. The 2 value, which is a standard statistical measure of the closeness of the fit (BIAevaluation 3 Software Handbook), was <2 in all cases.
Results
Chemical characterization of MBL-I and MBL-II
Purified human serum MBL appears as a mixture of oligomers of different sizes that can be fractionated using anion-exchange chromatography (8). SDS-PAGE analysis of the two major MBL species isolated from serum, MBL-I and MBL-II, showed that they migrate as diffuse bands with apparent molecular masses of 235 and 310 kDa, respectively, under nonreducing conditions (Fig. 1A). In the case of MBL-II (Fig. 1, lane 2), two additional minor bands were detected, one corresponding to trace amounts of MBL-I and the other one with an apparent molecular mass of 385 kDa, corresponding to the higher order oligomeric species termed MBL-III by Dahl et al. (8). All forms yielded a single chain with an apparent molecular mass of 32 kDa under reducing conditions (Fig. 1B).
FIGURE 1. SDS-PAGE analysis of purified MBL-I and MBL-II. A, Unreduced MBL-I (lane 1) and MBL-II (lane 2). The molecular masses of unreduced standard proteins (expressed in kilodaltons) are shown on the left side of the gel. B, Reduced MBL-I (lane 1) and MBL-II (lane 2). The molecular masses of reduced standard proteins (expressed in kilodaltons) are shown on the right side of the gel.
With a view to determining precisely the oligomerization state of MBL-I and -II, both species were submitted to MALDI-TOF mass spectrometry analysis, either in their native state or after mild reduction in the presence of 50 mM DTT. Both MBL samples yielded very similar spectra after reduction, with all peaks appearing as doublets (Fig. 2). A major peak (a) with an average mass of 25,449 ± 18 Da was observed in both cases, as well as an additional peak (e) with a mass of 12,726 ± 9 Da consistent with the m/z ratio expected for the corresponding doubly-charged species (Table I). The difference between the highest mass value (25,449) and the calculated mass of the unmodified polypeptide moiety (24,021 Da) yields an average extra mass of 1,428 ± 18 Da, consistent with the presence of 4 hydroxyprolines and 4 glucosylgalactosylhydroxylysines in the collagen-like region (calculated total extra mass, 1,424 Da). It can be assumed consequently that the difference of 330 ± 18 Da observed between peaks a and b of the major doublet corresponds to the lack of glycosylation of 1 hydroxylysine (expected mass difference, 324 Da). Assuming that peaks a and b account for two-thirds and one-third of the major species, respectively, an average mass value of 25,340 ± 18 Da can be assessed for the elementary chain of MBL-I and -II. Minor peaks with average mass values of 23,586 ± 18 Da (peak c) and 11,792 ± 9 Da (peak g), corresponding to mono- and doubly-charged species, respectively, were also detected, suggesting proteolytic cleavage of the major species. N-terminal sequence analysis of both MBL-I and -II yielded the expected sequence Glu-Thr-Val-Thr-Cys-Glu-Asp-Ala-Gln-Lys ..., ruling out the possibility of a degradation at the N-terminal end of the protein. In addition, a single band was detected by SDS-PAGE analysis of reduced MBL samples (Fig. 1B), strongly suggesting that the relative amount of this minor lower mass species might have been overestimated upon mass spectrometry analysis.
FIGURE 2. MALDI-TOF mass spectrometry analysis of reduced MBL-I and MBL-II. MBL-I and -II were incubated for 1 h at 37°C in the presence of 50 mM DTT and submitted to mass spectrometry analysis as described in Materials and Methods. Each spectrum is the result of 100 averaged laser pulses.
Table I. Mass spectrometry analysis of the reduced MBL samples
As shown in Fig. 3, the mass spectrum of native MBL-I yielded three major peaks at 228,098 ± 170 (peak A), 113,929 ± 85 (peak B), and 76,081 ± 57 Da (peak C). The latter two values correspond to the m/z ratios expected for the doubly- and triply-charged ions of the former species (Table II). The MBL-II spectrum yielded two major peaks at 304,899 ± 229 (peak D) and 152,151 ± 114 Da (peak E), and minor peaks at 228,327 ± 171 (peak A), 114,214 ± 86 (peak B), 101,547 ± 76 (peak F) and 76,311 ± 57 Da (peak C). Peaks A and B account for the presence of trace amounts of MBL-I, as expected from SDS-PAGE analysis of MBL-II (Fig. 1A). The four other peaks can be interpreted as deriving from a single species with a molecular mass of 304,899 ± 229 Da, with z values of 1, 2, 3 and 4 (Table II). Peak C could also correspond to the triply-charged ion of the contaminating MBL-I species (expected value, 76,109). Interestingly, no species of higher mass (up to 500,000 Da) was detected in the MBL-II sample. Comparison of the mass values obtained for native MBL-I (228,098 Da) and MBL-II (304,899 Da) with the average mass derived for the elementary chain (25,340 Da) indicates that MBL-I and -II contain 9.00 and 12.03 polypeptide chains, respectively, and are therefore trimers and tetramers of the structural unit.
FIGURE 3. MALDI-TOF mass spectrometry analysis of native MBL-I and MBL-II. Samples were analyzed as described in Materials and Methods. Each spectrum is the result of 100 averaged laser pulses.
Table II. Mass spectrometry analysis of the native MBL samples
Carbohydrate-binding properties of trimeric and tetrameric MBL
The carbohydrate-binding ability of trimeric and tetrameric MBL was investigated by surface plasmon resonance spectroscopy, using glycoconjugates (Man-BSA and GlcNAc-BSA) as the immobilized ligands. As shown in Fig. 4, each MBL form bound to Man and N-acetylglucosamine in the presence of 1 mM CaCl2. Complete elution of the bound protein could be achieved in all cases by a pulse injection of EDTA or of a 0.3 M solution of mannose or GlcNAc. The binding curves exhibited quite different shapes, with a higher binding level and a slower dissociation of tetrameric MBL for both glycoconjugates (Fig. 4). The kinetic parameters of the interactions were determined by recording sensorgrams at varying MBL concentrations and evaluation of the data by global fitting as described in Materials and Methods. The values of maximal binding capacity (Rmax), association (kon) and dissociation (koff) rate constants and the resulting apparent KD values are listed in Table III. Consistent with our preliminary analyses (see Fig. 4), tetrameric MBL exhibited much lower koff values than trimeric MBL for binding to both Man and GlcNAc (13- and 3-fold, respectively). In contrast, trimeric MBL exhibited slightly higher kon values than tetrameric MBL for binding to either carbohydrate. The resulting apparent KD values were in the nanomolar range in all instances, with comparable values for both MBL forms in the case of GlcNAc and a 4-fold higher affinity for tetrameric MBL in the case of mannose (Table III). Tetrameric MBL showed a 4- and 12-fold higher maximal binding capacity (Rmax) for Man and GlcNAc, respectively, compared with the trimeric form (Table III).
FIGURE 4. Analysis by surface plasmon resonance spectroscopy of the interaction of trimeric and tetrameric MBL with Man-BSA (A) and GlcNAc-BSA (B). Man-BSA (2400 RU) and GlcNAc-BSA (3000 RU) were immobilized on the sensor chip as described in Materials and Methods. Sixty microliters of 2 nM trimeric or tetrameric MBL were injected in 145 mM NaCl, 1 mM CaCl2, 50 mM triethanolamine hydrochloride, pH 7.4, containing 0.005% surfactant P20, at a flow rate of 20 μl/min.
Table III. Kinetic and dissociation constants for binding of trimeric and tetrameric MBL to immobilized Man-BSA and GlcNAc-BSA
No binding of trimeric or tetrameric MBL to either glycoconjugate was observed when EDTA was substituted for Ca2+ in the running buffer or when Gal-BSA was immobilized on the sensor chip (data not shown), as expected from the known carbohydrate specificity of the C-type lectin domain of MBL (28).
Interaction properties of trimeric and tetrameric MBL with the MASPs and MAp19
Trimeric and tetrameric MBL were next immobilized on a sensorchip to compare their interaction properties with the recombinant MASPs and MAp19 used as soluble ligands. As shown in Fig. 5, MASP-1 (Fig. 5A), MASP-2 (Fig. 5B), MASP-3 (Fig. 5C), and MAp19 (Fig. 5D) each individually bound to immobilized trimeric or tetrameric MBL in very similar ways, with virtually identical association and dissociation phases in both cases. Likewise, as previously determined in the case of MASP-3 (25), the kinetic rate constants for MASP-1, MASP-2, and MAp19 were of the same order in the case of trimeric and tetrameric MBL, and the resulting KD values exhibited no significant difference (Table IV). No binding of the ligands to either trimeric or tetrameric MBL was observed when EDTA was substituted for CaCl2 in the running buffer.
FIGURE 5. Analysis by surface plasmon resonance spectroscopy of the interaction of the recombinant MASPs and MAp19 with trimeric and terameric MBL. Sixty microliters of 50 nM MASP-1 (A), 50 nM MASP-2 (B), 100 nM MASP-3 (C), and 100 nM MAp19 (D) were injected over 9,000 resonance units (RU) of immobilized trimeric MBL or 10,000 RU of immobilized tetrameric MBL.
Table IV. Kinetic and dissociation constants for binding of recombinant MASP-1, -2, -3 and MAp19 to immobilized trimeric and tetrameric MBL
Stoichiometry of the complexes between trimeric or tetrameric MBL and MASP-3 or MAp19
To further characterize the interaction properties of trimeric and tetrameric MBL, individual MBL-MASP complexes were assembled from the isolated components, and their molecular composition was studied using gel permeation chromatography. Due to the low amount of recombinant MASP-1 and MASP-2 available, experiments were performed using MASP-3 and MAp19. Both proteins have been shown to form dimers in the presence of Ca2+ ions (23, 25). Separate analyses of trimeric MBL and the MASP-3 dimer were first performed, indicating that the isolated proteins eluted at 24.1 and 25.2 min, respectively (Fig. 6A). When trimeric MBL and MASP-3 were mixed in a 1:2 molar ratio and then applied to the column, no residual peak corresponding to either protein was detected, but a new peak eluting ahead of trimeric MBL was observed, consistent with the formation of a 1:2 MBL/MASP-3 complex (Fig. 6A). When trimeric MBL and MASP-3 were mixed in a 1:1 or 1:4 ratio, the amount of complex formed and its elution position did not change, whereas extra peaks corresponding to either excess trimeric MBL or excess MASP-3 were observed. A minor peak eluting at 31 min and showing no interaction with MASP-3 was also detected in the trimeric MBL sample (Fig. 6A). This peak was unlikely to correspond to a MBL form of lower mass (i.e., with only one or two structural units) because no such species was detected to a significant extent by SDS-PAGE analysis of MBL-I (Fig. 1A). Analysis of the interaction between MASP-3 and tetrameric MBL yielded results very similar to those obtained with trimeric MBL, fully consistent with the formation of a 1:2 MBL/MASP-3 complex, even in the presence of a molar excess of MASP-3 (Fig. 6B).
FIGURE 6. Gel filtration analysis of complexes between trimeric or tetrameric MBL and MASP-3. Trimeric (A) or tetrameric (B) MBL was mixed with MASP-3 at molar ratios of 1:1 (70 pmol each), 1:2 (35 pmol of MBL and 70 pmol of MASP-3), and 1:4 (35 pmol of MBL and 140 pmol of MASP-3) in a volume of 100 μl and incubated for 1 h at room temperature. Isolated trimeric MBL, tetrameric MBL (35 pmol each), and MASP-3 (70 pmol), as well as the different mixtures were injected on a Superose 6 column equilibrated in 145 mM NaCl, 1 mM CaCl2, 50 mM triethanolamine hydrochloride, pH 7.4, and run at 0.5 ml/min.
Similar experiments were performed using MAp19 (Fig. 7). The Ca2+-dependent MAp19 dimer yielded a peak eluting at 34.8 min, whereas MBL/MAp19 complexes eluted only slightly ahead of trimeric or tetrameric MBL, due to the relatively small size of the MAp19 dimer (38 kDa) compared with either MBL species (Fig. 7). Complex formation was also clearly assessed by the disappearance of the MAp19 peak. When trimeric or tetrameric MBL and MAp19 were mixed in a 1:1 ratio, the peak corresponding to the complex became broader, with a shoulder corresponding to free MBL (Fig. 7). Conversely, a peak corresponding to the MAp19 dimer was observed when a 4-fold molar excess of MAp19 was used.
FIGURE 7. Gel filtration analysis of complexes between trimeric or tetrameric MBL and MAp19. Trimeric (A) or tetrameric (B) MBL was mixed with MAp19 at molar ratios of 1:1 (114 pmol each), 1:2 (57 pmol of MBL and 114 pmol of MAp19), and 1:4 (57 pmol of MBL and 228 pmol of MAp19) in a volume of 100 μl and analyzed as described in the legend to Fig. 6. Isolated MBL (57 pmol) and MAp19 (114 pmol) were also injected as controls.
It was concluded from these results that trimeric and tetrameric MBL each form 1:2 complexes with MASP-3 and MAp19. This stoichiometry is not modified in the presence of excess MBL, MASP-3, or MAp19.
Discussion
Human serum MBL is usually represented as a hexameric molecule, although structures ranging from dimers to hexamers have been observed previously by electron microscopy (14). Various relative amounts of these different oligomeric forms have been reported, with a predominance of hexamers (29), of dimers and trimers (15), or of trimers and tetramers (14). Two major forms of serum MBL, originally called MBL-I and -II, have been described recently, and it was suggested from SDS-PAGE analysis that they correspond to trimers and tetramers of the structural unit, respectively (8, 22). Using MALDI-TOF mass spectrometry analysis of purified native or reduced MBL-I and -II, we provide unambiguous experimental evidence that the MBL-I and -II forms contain 9 and 12 elementary chains and therefore consist of trimers and tetramers of the trimeric structural unit, respectively. It has been proposed that oligomerization involves interchain disulfide bridges between cysteine residues located at positions 5, 12, and 18 of MBL (21). Our results are fully consistent with this hypothesis and show that both assembly of the structural unit and oligomer formation involve disulfide bonds that are sensitive to reduction under nondenaturing conditions and therefore fully accessible to the solvent. The apparent masses determined for native trimeric and tetrameric MBL by SDS-PAGE analysis under nonreducing conditions (235 and 310 kDa, respectively) are close to those determined by mass spectrometry (228,098 ± 170 and 304,899 ± 229 Da). In contrast, the mass of the reduced chain determined by SDS-PAGE analysis (32 kDa) is clearly overestimated compared with the average value determined by mass spectrometry (25,340 ± 18 Da). It may be inferred from this observation that the major serum MBL species with apparent masses of 210 and 300 kDa, previously interpreted as dimers and trimers (15), rather represent trimers and tetramers, respectively. It should also be emphasized that, despite the fact that our MBL-II preparation was contaminated by a higher molecular mass band corresponding to the species termed MBL-III (8) (Fig. 1A), no such form could be detected by mass spectrometry analysis (Fig. 3). As reported in recent papers showing the oligomerization pattern of MBL (17, 18, 19, 30), higher oligomeric forms may exist in serum, but in much lower amounts than those of the trimeric and tetrameric forms. In contrast, the oligomerization pattern of human serum MBL may be more complex if the different common variant alleles of the MBL gene are considered. It has indeed been shown recently that these alleles give rise to significant amounts of lower oligomeric forms with reduced carbohydrate-binding ability (17, 18) that would likely not bind to the affinity column used in our purification protocol.
The mass value determined for the reduced MBL chain by MALDI-TOF mass spectrometry is close to the 25,500-Da value obtained by Larsen et al. (21), using surface-enhanced laser desorption/ionization mass spectrometry. Compared with the unmodified polypeptide, the MBL chain shows a mass increase that accounts for the posttranslational modifications usually observed in collagens, i.e., hydroxylation of the proline and lysine residues in the Y position of the repeating Gly-X-Y sequence and glycosylation of hydroxylysine residues. Previous analyses have indicated the presence of 3.7–4 hydroxyproline and 3.2–4 hydroxylysine residues per MBL chain (31, 32). Our own results are consistent with the presence of 4 hydroxyproline and 4 hydroxylysine residues, among which 3–4 carry the characteristic O-linked glucosylgalactosyl disaccharide observed in other proteins containing collagen-like sequences such as rat MBL (33), human C1q (34), or adiponectin (35). The fact that the mass values for the elementary chain of trimeric and tetrameric MBL are strictly identical indicates that preferential association into trimers or tetramers is not dependent on the extent of modification in the collagen-like region.
Analysis of the carbohydrate-binding properties of trimeric and tetrameric MBL reveals that they bind to Man-BSA with KD values of 2.2 and 0.55 nM and to GlcNAc-BSA with KD values of 1.2 and 0.96 nM, respectively, indicating high affinity in all cases. These values are in the same range as those determined previously for binding of rat MBL to glycosylated BSA (1 nM) (36) or of human MBL to mannan (2.3 nM) (37). Although the apparent KD values are of the same order for trimeric and tetrameric MBL, these two forms show strikingly different kinetic rate constants. Thus, tetrameric MBL exhibits a 13-fold lower dissociation rate constant from Man-BSA than trimeric MBL does, reflecting a much higher stability of the carbohydrate-tetrameric MBL interaction. Although the kon value is 3-fold lower in the case of tetrameric MBL, the maximal binding capacity to Man-BSA is 4-fold higher, and therefore much more mannose/tetrameric MBL complexes are expected to be formed in the presence of equivalent concentrations of both MBL forms (Fig. 4A). Comparable results were obtained using GlcNAc-BSA as a ligand, although in that case the major difference is that tetrameric MBL has a 12-fold higher maximal binding capacity than trimeric MBL. Thus, the oligomerization state of MBL has a direct effect on its carbohydrate-binding properties, with the tetrameric form showing increased binding capacity and higher interaction stability. It was shown previously that monovalent interaction between a single carbohydrate recognition domain of rat MBL-A and mannose is characterized by a very weak affinity (KD in the millimolar range) (38). Interaction of oligomeric MBL with carbohydrate arrays allows simultaneous engagement of several lectin domains, hence strengthening binding through an avidity phenomenon. In this respect, our data clearly indicate that the tetravalent MBL form has more avidity than the trivalent MBL form. Consistent with these observations, it was reported recently that mutated MBL molecules with a dimeric structure bind mannan with a markedly decreased capacity compared with wild-type MBL (21).
MBL-I and MBL-II were found to be preferentially associated in serum with MASP-1 and MAp19 and with MASP-3 and MASP-2, respectively (8). In contrast, using surface plasmon resonance spectroscopy, we have shown that recombinant MASP-3 binds to serum trimeric and tetrameric MBL with comparable binding constants (25). We have now extended this study to recombinant MASP-1, MASP-2, and MAp19, and we show that these proteins each interact individually with trimeric and tetrameric MBL with similar binding constants. Our data therefore indicate that the oligomerization state of MBL has no significant effect on its MASP-binding properties in vitro. In this respect, it was reported recently that purified tetrameric MBL induces C4 cleavage more efficiently than does trimeric MBL when added to MBL-deficient plasma (22), and it was suggested that this arises from a preferential association of serum MASP-2 with higher MBL oligomers. In light of our data, it should be pointed out that this difference may also be accounted for by the differential carbohydrate-binding capacity of trimeric and tetrameric MBL, because the complex between trimeric MBL and MASP-2 would exhibit a reduced capacity to bind to the mannan surface used in these experiments and therefore would trigger MASP-2 activation and C4 cleavage less efficiently.
Gel filtration analysis of the complexes between trimeric or tetrameric MBL and MASP-3 or MAp19 reveals a 1:2 stoichiometry in all cases. Although similar experiments could not be done using MASP-1 and MASP-2, it should be stressed that MASP-1 and MASP-3 share the same N-terminal CUB1-epidermal growth factor (EGF)-CUB2 domains known to mediate binding to human MBL or rat MBL-A (24, 39). In agreement with the observation that these proteases bind trimeric and tetrameric MBL with comparable affinities (Table IV), it can be inferred that MASP-1 associates with either MBL species with the same stoichiometry as determined for MASP-3. This reasoning likely also applies to MASP-2, because this protease comprises the same N-terminal CUB1-EGF modules as MAp19 (12, 13). The present results do not fully agree with those of Chen and Wallis (40) who reported that, whereas rat recombinant MBL-A dimers bind two MASP-1 or MASP-2 CUB1-EGF-CUB2 fragments, trimeric and tetrameric MBL-A forms bind up to four. It should be emphasized, however, that the latter complexes were observed only in the presence of a large molar excess (6- to 12-fold) of the MASP fragments over MBL-A.
The x-ray structure of human MAp19 has been solved recently, and site-directed mutagenesis studies have allowed us to map the MBL-binding site at the distal end of the CUB1 module (41). A three-dimensional model of the interaction between the MAp19 dimer and a triple-helical segment of MBL containing the putative MASP-binding site was proposed. This information was used to construct models of the trimeric and tetrameric MBL molecules and of their complexes with MAp19. As illustrated in Fig. 8B, the complex between the MBL tetramer and the MAp19 dimer is expected to be symmetrical, the latter protein interacting with binding sites located on two opposite collagen triple helices of tetrameric MBL. In contrast, it is clear that such a symmetry cannot be conserved in the complex between trimeric MBL and MAp19, because in this case MAp19 should interact with two contiguous collagen helices (Fig. 8A).
FIGURE 8. Model of the complexes between MAp19 and trimeric (A) or tetrameric MBL (B). The CUB and EGF modules of MAp19 (PDB ID: 1SZB; Ref. 39 ) are blue and green, respectively. The lectin domain of human MBL (positions 88–228) (PDB ID: 1HUP; Ref. 42 ) is red, and the collagen-like segment (positions 45–72) containing the MAp19 binding site as modeled by Gregory et al. (41) is gray. The structure of the MBL collagen-like segment encompassing residues 73–87 is unknown and is shown as gray dots. A bottom view of the complexes is shown.
In summary, our results show that the major MBL forms in human serum are trimers and tetramers of the structural unit. These two forms exhibit similar interaction properties with the MASPs, as shown by their common ability to form 1:2 complexes with MASP-3 and MAp19. In contrast, they markedly differ in their carbohydrate-binding characteristics, with the tetrameric form showing higher binding capacity and increased interaction stability. This differential behavior has potentially important functional implications because a prolonged interaction of the MBL/MASP-2 complex with microbial carbohydrates may be expected to facilitate MASP-2 activation and hence to trigger complement activation through C4 and C2 cleavage with more efficiency.
Disclosures
The authors have no financial conflict of interest.
Footnotes
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 work was supported by the Commissariat à l’Energie Atomique, the Centre National de la Recherche Scientifique, and the Université Joseph Fourier (Grenoble, France). A preliminary report of this study was presented by N.M.T. at the 20th International Complement Workshop in Honolulu, HI, July 13–18, 2004.
2 Address correspondence and reprint requests to Dr. Nicole Thielens, Laboratoire d’Enzymologie Moléculaire, Institut de Biologie Structurale Jean-Pierre Ebel, 41 rue Jules Horowitz, 38027 Grenoble Cedex 1, France. E-mail address: nicole.thielens{at}ibs.fr
3 Abbreviations used in this paper: MBL, mannan-binding lectin; MASP, MBL-associated serine protease; GlcNAc, N-acetylglucosamine; MAp19, 19-kDa MBL-associated protein; CUB module, module originally found in complement proteins C1r/C1s, Uegf, and bone morphogenetic protein-1; EGF, epidermal growth factor; Man, mannose; RU, resonance unit(s); the nomenclature of protein modules is that defined by Bork and Bairoch (1 ).
Received for publication October 6, 2004. Accepted for publication December 6, 2004.
References
Bork, P., A. Bairoch. 1995. Extracellular protein modules: a proposed nomenclature. Trends Biochem. Sci. 20:(Suppl.):C03. March.
Lu, J., C. Teh, U. Kishore, K. B. Reid. 2002. Collectins and ficolins: sugar pattern recognition molecules of the mammalian innate immune system. Biochim. Biophys. Acta 1572:387.
Holmskov, U., S. Thiel, J. C. Jensenius. 2003. Collectins and ficolins: humoral lectins of the innate immune defense. Annu. Rev. Immunol. 21:547.
Neth, O., D. L. Jack, A. W. Dodds, H. Holzel, N. J. Klein, M. W. Turner. 2000. Mannose-binding lectin binds to a range of clinically relevant microorganisms and promotes complement deposition. Infect. Immun. 68:688.
Turner, M. W.. 1996. Mannose-binding lectin: the pluripotent molecule of the innate immune system. Immunol. Today 17:532.
Matsushita, M., T. Fujita. 1992. Activation of the classical complement pathway by mannose-binding protein in association with a novel C1s-like serine protease. J. Exp. Med. 176:1497.
Thiel, S., T. Vorup-Jensen, C. M. Stover, W. Schwaeble, S. B. Laursen, K. Poulsen, A. C. Willis, P. Eggleton, S. Hansen, U. Holmskov, K. B. M. Reid, J. C. Jensenius. 1997. A second serine protease associated with mannan-binding lectin that activates complement. Nature 386:506.
Dahl, M. R., S. Thiel, M. Matsushita, T. Fujita, A. C. Willis, T. Christensen, T. Vorup-Jensen, J. C. Jensenius. 2001. MASP-3 and its association with distinct complexes of the mannan-binding lectin complement activation pathway. Immunity 15:127.
Vorup-Jensen, T., S. V. Petersen, A. Hansen, K. Poulsen, W. Schwaeble, R. B. Sim, K. B. M. Reid, S. J. Davis, S. Thiel, J. C. Jensenius. 2000. Distinct pathways of mannan-binding lectin (MBL)- and C1-complex autoactivation revealed by reconstitution of MBL with recombinant MBL-associated serine protease-2. J. Immunol. 165:2093.
Rossi, V., S. Cseh, I. Bally, N. M. Thielens, J. C. Jensenius, G. J. Arlaud. 2001. Substrate specificities of recombinant mannan-binding lectin-associated serine proteases-1 and -2. J. Biol. Chem. 276:40880.
Ambrus, G., P. Gal, M. Kojima, K. Szilagyi, J. Balczer, J. Antal, L. Graf, A. Laich, B. E. Moffatt, W. Schwaeble, R. B. Sim, P. Zavodszky. 2003. Natural substrates and inhibitors of mannan-binding lectin-associated serine protease-1 and -2: a study on recombinant catalytic fragments. J. Immunol. 170:1374.
Stover, C. M., S. Thiel, M. Thelen, N. J. Lynch, T. Vorup-Jensen, J. C. Jensenius, W. J. Schwaeble. 1999. Two constituents of the initiation complex of the mannan-binding lectin activation pathway of complement are encoded by a single structural gene. J. Immunol. 162:3481.
Takahashi, M., Y. Endo, T. Fujita, M. Matsushita. 1999. A truncated form of mannose-binding lectin-associated serine protease (MASP)-2 expressed by alternative polyadenylation is a component of the lectin complement pathway. Int. Immunol. 11:859.
Lu, J. H., S Thiel, H. Wiedemann, R. Timpl, K. B. Reid. 1990. Binding of the pentamer/hexamer forms of mannan-binding protein to zymosan activates the proenzyme C1r2C1s2 complex, of the classical pathway of complement, without involvement of C1q. J. Immunol. 144:2287.
Lipscombe, R. J., M Sumiya, J. A. Summerfield, M. W. Turner. 1995. Distinct physicochemical characteristics of human mannose binding protein expressed by individuals of differing genotype. Immunology 85:660.
Turner, M. W.. 1998. Mannose-binding lectin (MBL) in health and disease. Immunobiology 199:327.
Garred, P., F. Larsen, H. O. Madsen, C. Koch. 2003. Mannose-binding lectin deficiency revisited. Mol. Immunol. 40:73.
Roos, A., P. Garred, M. E. Wildenberg, N. J. Lynch, J. R. Munoz, T. C. Zuiverloon, L. H. Bouwman, N. Schlagwein, F. C. Fallaux van den Houten, M. C. Faber-Krol, et al 2004. Antibody-mediated activation of the classical pathway of complement may compensate for mannose-binding lectin deficiency. Eur. J. Immunol. 34:2589.
Terai, I., K. Kobayashi, M. Matsushita, H. Miyakawa, N. Mafune, H. Kikuta. 2003. Relationship between gene polymorphisms of mannose-binding lectin (MBL) and two molecular forms of MBL. Eur. J. Immunol. 33:2755.
Matsushita, M., R. A. Ezekowitz, T. Fujita. 1995. The Gly-54Asp allelic form of human mannose-binding protein (MBP) fails to bind MBP-associated serine protease. Biochem. J. 311:1021.
Larsen, F., H. O. Madsen, R. B. Sim, C. Koch, P. Garred. 2004. Disease-associated mutations in human mannose-binding lectin compromise oligomerization and activity of the final protein. J. Biol. Chem. 279:21302.
Dumestre-Perard, C., D. Ponard, G. J. Arlaud, N. Monnier, R. B. Sim, M. G. Colomb. 2002. Evaluation and clinical interest of mannan binding lectin function in human plasma. Mol. Immunol. 39:465.
Thielens, N. M., S. Cseh, S. Thiel, T. Vorup-Jensen, V. Rossi, J. C. Jensenius, G. J. Arlaud. 2001. Interaction properties of human mannan-binding lectin (MBL)-associated serine proteases-1 and -2, MBL-associated protein 19, and MBL. J. Immunol. 166:5068.
Cseh, S., L. Vera, M. Matsushita, T. Fujita, G. J. Arlaud, N. M. Thielens. 2002. Characterization of the interaction between L-ficolin/p35 and mannan-binding lectin-associated serine proteases-1 and -2. J. Immunol. 169:5735.
Zundel, S., S. Cseh, M. Lacroix, M. R. Dahl, M. Matsushita, J.-P. Andrieu, W. J. Schwaeble, J. C. Jensenius, T. Fujita, G. J. Arlaud, N. M. Thielens. 2004. Characterization of recombinant mannan-binding lectin-associated serine protease (MASP)-3 suggests an activation mechanism different from that of MASP-1 and MASP-2. J. Immunol. 172:4342.
Thielens, N. M., C. A. Aude, M. B. Lacroix, J. Gagnon, G. J. Arlaud. 1990. Ca2+ binding properties and Ca2+-dependent interactions of the isolated NH2-terminal fragments of human complement proteases C1-r and C1-s. J. Biol. Chem. 265:14469.
Rossi, V., C. Gaboriaud, M. Lacroix, J. Ulrich, J. C. Fontecilla-Camps, J. Gagnon, G. J. Arlaud. 1995. Structure of the catalytic region of human complement protease C1s: study by chemical cross-linking and three-dimensional homology modeling. Biochemistry 34:7311.
Drickamer, K.. 1992. Engineering galactose-binding activity into a C-type mannose-binding protein. Nature 360:183.
Wong, N., R. B. Sim. 1997. Comparison of the complement system protein complexes formed by C1q and MBL. Biochem. Soc. Trans. 25:41S.
Ohlenschlaeger, T., P. Garred, H. O. Madsen, S. Jacobsen. 2004. Mannose-binding lectin variant alleles and the risk of arterial thrombosis in systemic lupus erythematosus. N. Engl. J. Med. 351:260.
Ohtani, K., Y. Suzuki, S. Eda, T. Kawai, T. Kase, H. Keshi, Y. Sakai, S. Yamamoto, T. Sakamoto, N. Wakamiya. 1999. High-level and effective production of human mannan-binding lectin (MBL) in Chinese hamster ovary (CHO) cells. J. Immunol. Methods 222:135.
Vorup-Jensen, T., E. S. Sorensen, U. B. Jensen, W. Schwaeble, T. Kawasaki, Y. Ma, K. Uemura, N. Wakamiya, Y. Suzuki, T. G. Jensen, et al 2001. Recombinant expression of human mannan-binding lectin. Int. Immunopharmacol. 1:677.
Colley, K. J., J. U. Baenziger. 1987. Identification of the post-translational modifications of the core-specific lectin: the core-specific lectin contains hydroxyproline, hydroxylysine, and glucosylgalactosylhydroxylysine residues. J. Biol. Chem. 262:10290.
Shinkai, H., K. Yonemasu. 1979. Hydroxylysine-linked glycosides of human complement subcomponent C1q and various collagens. Biochem. J. 177:847.
Wang, Y., A. Xu, C. Knight, L. Y. Xu, G. J. Cooper. 2002. Hydroxylation and glycosylation of the four conserved lysine residues in the collagenous domain of adiponectin: potential role in the modulation of its insulin-sensitizing activity. J. Biol. Chem. 277:19521.
Lee, R. T., Y. Ichikawa, T. Kawasaki, K. Drickamer, Y. C. Lee. 1992. Multivalent ligand binding by serum mannose-binding protein. Arch. Biochem. Biophys. 299:129.
Kawasaki, N., T. Kawasaki, I. Yamashina. 1983. Isolation and characterization of a mannan-binding protein from human serum. J. Biochem. 94:937.
Iobst, S. T., M. R. Mormald, W. I. Weis, R. A. Dwek, K. Drickamer. 1994. Binding of sugar ligands to Ca2+-dependent animal lectins. I. Analysis of mannose binding by site-directed mutagenesis and NMR. J. Biol. Chem. 269:15505.
Wallis, R., R. B. Dodd. 2000. Interaction of mannose-binding protein with associated serine proteases: effects of naturally occurring mutations. J. Biol. Chem. 275:30962.
Chen, C. B., R. Wallis. 2001. Stoichiometry of complexes between mannose-binding protein and its associated serine proteases: defining functional units for complement activation. J. Biol. Chem. 276:25894.
Gregory, L. A., N. M. Thielens, M. Matsushita, R. Sorensen, G. J. Arlaud, J. C. Fontecilla-Camps, C. Gaboriaud. 2004. The x-ray structure of human mannan-binding lectin-associated protein 19 (MAp19) and its interaction site with mannan-binding lectin and L-ficolin. J. Biol. Chem. 279:29391.
Sheriff, S., C. Y. Chang, R. A. Ezekowitz. 1994. Human mannose-binding protein carbohydrate recognition domain trimerizes through a triple -helical coiled-coil. Nat. Struct. Biol. 1:789.(Florence Teillet, Bernard)
Mannan-binding lectin (MBL) is an oligomeric C-type lectin assembled from homotrimeric structural units that binds to neutral carbohydrates on microbial surfaces. It forms individual complexes with MBL-associated serine proteases (MASP)-1, -2, -3 and a truncated form of MASP-2 (MAp19) and triggers the lectin pathway of complement through MASP-2 activation. To characterize the oligomerization state of the two major MBL forms present in human serum, both proteins were analyzed by mass spectrometry. Mass values of 228,098 ± 170 Da (MBL-I) and 304,899 ± 229 Da (MBL-II) were determined for the native proteins, whereas reduction of both species yielded a single chain with an average mass of 25,340 ± 18 Da. This demonstrates that MBL-I and -II contain 9 and 12 disulfide-linked chains, respectively, and therefore are trimers and tetramers of the structural unit. As shown by surface plasmon resonance spectroscopy, trimeric and tetrameric MBL bound to immobilized mannose-BSA and N-acetylglucosamine-BSA with comparable KD values (2.2 and 0.55 nM and 1.2 and 0.96 nM, respectively). However, tetrameric MBL exhibited significantly higher maximal binding capacity and lower dissociation rate constants for both carbohydrates. In contrast, no significant difference was detected for binding of the recombinant MASPs or MAp19 to immobilized trimeric or tetrameric MBL. As shown by gel filtration, both MBL species formed 1:2 complexes with MASP-3 or MAp19. These results provide the first precise analysis of the major human MBL oligomers. The oligomerization state of MBL has a direct effect on its carbohydrate-binding properties, but no influence on the interaction with the MASPs.
Introduction
Mannan-binding lectin (MBL)3 is a member of the collectin family, a group of proteins that comprise a collagen-like region and a C-type lectin domain and are involved in innate immunity against infection (2, 3). MBL recognizes patterns of neutral carbohydrates such as mannose and N-acetylglucosamine (GlcNAc) on a wide range of microbial surfaces and is the only collectin able to initiate activation of the lectin pathway of complement (4, 5). Circulating MBL forms complexes with three MBL-associated serine proteases (MASP-1, MASP-2, and MASP-3) (6, 7, 8), its association with MASP-2 yielding a complex that triggers complement activation through cleavage of components C4 and C2 (9, 10, 11). In addition to the MASPs, serum MBL binds to a truncated form of MASP-2 called 19-kDa MBL-associated protein (MAp19) or small MBL-associated protein (12, 13). The biological role of MAp19 as well as the physiological substrates of MASP-1 and MASP-3 remain to be identified.
Human MBL is assembled from a single polypeptide chain, consisting of a short N-terminal cysteine-rich region, a collagenous region comprising 19 repeating Gly-X-Y triplets with an interruption at the 8 triplet, a 34-residue hydrophobic stretch, and a C-terminal C-type lectin domain. Three polypeptide chains associate to form a homotrimeric structural unit comprising a collagen-like triple helix, an -helical coiled-coil called the neck region, and three carbohydrate recognition domains (5). The three cysteines in the N-terminal region form interchain disulfide bonds that mediate formation of oligomers comprising two to six structural units and exhibiting bouquet-like structures when viewed by electron microscopy (14). The collagen-like triple helices associate laterally at their N-terminal ends and then diverge at the level of the interruption in the Gly-X-Y sequence, thus defining a hinge in the molecule.
There are large variations in the MBL serum concentration, arising from polymorphism in the promoter region as well as in the structural moiety of the MBL gene (15), and several clinical studies have provided evidence for a link between a low MBL titer and increased susceptibility to various infectious diseases (16). The structural mutations located in exon 1 of the MBL gene generate amino acid substitutions in the collagen-like region that are believed to impair oligomerization of the protein and result in functional deficiency (15, 17, 18, 19). The reduced ability of these mutants to activate complement has been proposed to result from failure to bind to the MBL-associated serine proteases (20), and from the reduced carbohydrate-binding ability of the lower order MBL oligomers (17, 18, 21).
Purified serum MBL appears to be a mixture of oligomers of different sizes with two major forms, termed MBL-I and MBL-II (8), that can be separated using anion-exchange chromatography (8, 22). The objective of the present study was to precisely characterize the oligomerization state of both forms, their carbohydrate-binding ability, and their interaction properties with the MASPs and MAp19. Mass spectrometry analysis provides unambiguous evidence that MBL-I and -II are trimers and tetramers of the structural unit, respectively. The oligomerization state of MBL has a direct effect on its carbohydrate-binding properties, with the tetrameric form showing increased binding capacity and higher affinity, but has no significant influence on the interaction with the MASPs and MAp19 or on the stoichiometry of the resulting complexes.
Materials and Methods
Materials
N-Acetylglucosamine-BSA (10 average carbohydrate residues per BSA molecule) was purchased from Dextra Laboratories. Mannose (Man)- and galactose-BSA (30 average carbohydrate residues per BSA molecule) were from EY Laboratories.
Proteins
MBL was purified from human serum as described by Thielens et al. (23). Separation of MBL-I and MBL-II was performed by anion-exchange chromatography on a MonoQ HR 5/5 column (Amersham Pharmacia Biotech) as described by Dahl et al. (8) with the following modifications. The column was run at a flow rate of 0.5 ml/min, and MBL fractionation was achieved with a NaCl gradient to 250 mM in 40 min. Fractions of 1 ml were collected and analyzed for the presence of MBL-I and MBL-II by SDS-PAGE and Coomassie Blue staining. Recombinant MASP-1, MASP-2, and MAp19 were expressed using a baculovirus/insect cell system and purified as described previously (23, 24). Recombinant MASP-3 was expressed using the same system and purified as described by Zundel et al. (25) with the following changes. After the first anion-exchange chromatography step, fractions containing MASP-3 were dialyzed against 50 mM NaCl, 1 mM CaCl2, 50 mM triethanolamine-hydrochloride, pH 6.9, and loaded at 1.0 ml/min onto a MonoQ HR 5/5 column (Amersham Biosciences) equilibrated in the same buffer. Elution was conducted with a linear NaCl gradient to 500 mM in 1 h. MASP-3-containing fractions were identified by SDS-PAGE analysis; they were dialyzed against 145 mM NaCl, 1 mM CaCl2, 50 mM triethanolamine hydrochloride, pH 7.4; and further purification was achieved by high pressure gel permeation on a TSK G3000 SWG column (7.5 x 600 nm) (Tosohaas) equilibrated in the same buffer.
The concentrations of purified MBL-I, MBL-II, MASP-3, and MAp19 were determined using molar extinction coefficients at 280 nm of 198,000, 264,000, 160,600, and 29,000 M–1 x cm–1, respectively, which were determined experimentally by amino acid analysis. The molecular masses of MBL-I (228,098 Da), MBL-II (304,899 Da), MASP-3 (87,524 Da), and MAp19 (19,088 Da) were determined by MALDI-TOF mass spectrometry analysis. Due to the low amount of material recovered, estimation of the concentration of recombinant MASP-1 and MASP-2 was based on Coomassie Blue staining after SDS-PAGE analysis using MASP-3 and MAp19 as internal standards and molecular masses of 82,000 and 75,100 Da, respectively (24).
Amino acid and N-terminal sequence analyses
Amino acid analyses were performed on acid hydrolysates of the samples using a Beckman 7300 amino acid analyzer as described previously (26). N-terminal sequence analyses were performed using an Applied Biosystems model 477A protein sequencer as described previously (27).
MALDI-TOF mass spectrometry analysis
Mass spectra were obtained using a Perseptive Biosystems Voyager Elite Xl time of flight mass spectrometer with delayed extraction, operating with a pulsed nitrogen laser at 337 nm. Positive ion mass spectra were acquired using linear, delayed extraction mode with an accelerating potential of 25 kV, an 85% grid potential, a 0.3% guide wire voltage, and a delay time of 600 ns. Samples were mixed with an equal volume of a saturated solution of sinapinic acid prepared in 50% (v/v) acetonitrile, 0.1% trifluoroacetic acid on the stainless steel sample plate and air-dried before analysis. External calibration was performed with BSA using m/z values of 66,431 and 33,216 for the mono- and doubly-charged species, respectively. The accuracy of molecular mass determination was estimated at 0.075%.
Gel permeation chromatography
Complexes between trimeric or tetrameric MBL on one hand and MASP-3 or MAp19 on the other were assembled by incubation of the proteins for 1 h at room temperature and analyzed by gel permeation chromatography on a Superose 6 10/300 GL column (Amersham Biosciences) equilibrated in 145 mM NaCl, 1 mM CaCl2, 50 mM triethanolamine hydrochloride, pH 7.4, and run at 0.5 ml/min. Proteins were detected from their optical density at 280 nm.
Real time surface plasmon resonance spectroscopy and data evaluation
Analyses were performed using a Biacore 3000 instrument (Biacore AB). Carbohydrate-BSA glycoconjugates and MBL were diluted to 10 and 30 μg/ml, respectively, in 10 mM sodium acetate, pH 4.0, and immobilized on the surface of a CM5 sensor chip (Biacore AB) using amine-coupling chemistry as described by Zundel et al. (25). Binding of trimeric and tetrameric MBL to immobilized Man-BSA (2400 resonance units (RU)), GlcNAc-BSA (3000 RU), or Gal-BSA (2300 RU) was measured at a flow rate of 20 μl/min in 145 mM NaCl, 1 mM CaCl2, 50 mM triethanolamine hydrochloride, pH 7.4, containing 0.005% surfactant P20 (Biacore AB). Binding of MASP-1, MASP-2, MASP-3, and MAp19 to immobilized trimeric MBL (9,000 RU) and tetrameric MBL (10,000 RU) was measured under the same conditions. Equivalent volumes of each protein sample were injected over a surface with immobilized BSA to serve as blank sensorgrams for subtraction of the bulk refractive index background. Regeneration of the surfaces was achieved by injection of 10 μl of the running buffer containing 5 mM EDTA instead of 1 mM CaCl2.
Data were analyzed by global fitting to a 1:1 Langmuir binding model of both the association and dissociation phases for several concentrations simultaneously, using the BIAevaluation 3.1 software (Biacore AB). The apparent equilibrium dissociation constants (KD) were calculated from the ratio of the dissociation and association rate constants (koff/kon). Maximal binding capacities (Rmax) were determined using the same model. The 2 value, which is a standard statistical measure of the closeness of the fit (BIAevaluation 3 Software Handbook), was <2 in all cases.
Results
Chemical characterization of MBL-I and MBL-II
Purified human serum MBL appears as a mixture of oligomers of different sizes that can be fractionated using anion-exchange chromatography (8). SDS-PAGE analysis of the two major MBL species isolated from serum, MBL-I and MBL-II, showed that they migrate as diffuse bands with apparent molecular masses of 235 and 310 kDa, respectively, under nonreducing conditions (Fig. 1A). In the case of MBL-II (Fig. 1, lane 2), two additional minor bands were detected, one corresponding to trace amounts of MBL-I and the other one with an apparent molecular mass of 385 kDa, corresponding to the higher order oligomeric species termed MBL-III by Dahl et al. (8). All forms yielded a single chain with an apparent molecular mass of 32 kDa under reducing conditions (Fig. 1B).
FIGURE 1. SDS-PAGE analysis of purified MBL-I and MBL-II. A, Unreduced MBL-I (lane 1) and MBL-II (lane 2). The molecular masses of unreduced standard proteins (expressed in kilodaltons) are shown on the left side of the gel. B, Reduced MBL-I (lane 1) and MBL-II (lane 2). The molecular masses of reduced standard proteins (expressed in kilodaltons) are shown on the right side of the gel.
With a view to determining precisely the oligomerization state of MBL-I and -II, both species were submitted to MALDI-TOF mass spectrometry analysis, either in their native state or after mild reduction in the presence of 50 mM DTT. Both MBL samples yielded very similar spectra after reduction, with all peaks appearing as doublets (Fig. 2). A major peak (a) with an average mass of 25,449 ± 18 Da was observed in both cases, as well as an additional peak (e) with a mass of 12,726 ± 9 Da consistent with the m/z ratio expected for the corresponding doubly-charged species (Table I). The difference between the highest mass value (25,449) and the calculated mass of the unmodified polypeptide moiety (24,021 Da) yields an average extra mass of 1,428 ± 18 Da, consistent with the presence of 4 hydroxyprolines and 4 glucosylgalactosylhydroxylysines in the collagen-like region (calculated total extra mass, 1,424 Da). It can be assumed consequently that the difference of 330 ± 18 Da observed between peaks a and b of the major doublet corresponds to the lack of glycosylation of 1 hydroxylysine (expected mass difference, 324 Da). Assuming that peaks a and b account for two-thirds and one-third of the major species, respectively, an average mass value of 25,340 ± 18 Da can be assessed for the elementary chain of MBL-I and -II. Minor peaks with average mass values of 23,586 ± 18 Da (peak c) and 11,792 ± 9 Da (peak g), corresponding to mono- and doubly-charged species, respectively, were also detected, suggesting proteolytic cleavage of the major species. N-terminal sequence analysis of both MBL-I and -II yielded the expected sequence Glu-Thr-Val-Thr-Cys-Glu-Asp-Ala-Gln-Lys ..., ruling out the possibility of a degradation at the N-terminal end of the protein. In addition, a single band was detected by SDS-PAGE analysis of reduced MBL samples (Fig. 1B), strongly suggesting that the relative amount of this minor lower mass species might have been overestimated upon mass spectrometry analysis.
FIGURE 2. MALDI-TOF mass spectrometry analysis of reduced MBL-I and MBL-II. MBL-I and -II were incubated for 1 h at 37°C in the presence of 50 mM DTT and submitted to mass spectrometry analysis as described in Materials and Methods. Each spectrum is the result of 100 averaged laser pulses.
Table I. Mass spectrometry analysis of the reduced MBL samples
As shown in Fig. 3, the mass spectrum of native MBL-I yielded three major peaks at 228,098 ± 170 (peak A), 113,929 ± 85 (peak B), and 76,081 ± 57 Da (peak C). The latter two values correspond to the m/z ratios expected for the doubly- and triply-charged ions of the former species (Table II). The MBL-II spectrum yielded two major peaks at 304,899 ± 229 (peak D) and 152,151 ± 114 Da (peak E), and minor peaks at 228,327 ± 171 (peak A), 114,214 ± 86 (peak B), 101,547 ± 76 (peak F) and 76,311 ± 57 Da (peak C). Peaks A and B account for the presence of trace amounts of MBL-I, as expected from SDS-PAGE analysis of MBL-II (Fig. 1A). The four other peaks can be interpreted as deriving from a single species with a molecular mass of 304,899 ± 229 Da, with z values of 1, 2, 3 and 4 (Table II). Peak C could also correspond to the triply-charged ion of the contaminating MBL-I species (expected value, 76,109). Interestingly, no species of higher mass (up to 500,000 Da) was detected in the MBL-II sample. Comparison of the mass values obtained for native MBL-I (228,098 Da) and MBL-II (304,899 Da) with the average mass derived for the elementary chain (25,340 Da) indicates that MBL-I and -II contain 9.00 and 12.03 polypeptide chains, respectively, and are therefore trimers and tetramers of the structural unit.
FIGURE 3. MALDI-TOF mass spectrometry analysis of native MBL-I and MBL-II. Samples were analyzed as described in Materials and Methods. Each spectrum is the result of 100 averaged laser pulses.
Table II. Mass spectrometry analysis of the native MBL samples
Carbohydrate-binding properties of trimeric and tetrameric MBL
The carbohydrate-binding ability of trimeric and tetrameric MBL was investigated by surface plasmon resonance spectroscopy, using glycoconjugates (Man-BSA and GlcNAc-BSA) as the immobilized ligands. As shown in Fig. 4, each MBL form bound to Man and N-acetylglucosamine in the presence of 1 mM CaCl2. Complete elution of the bound protein could be achieved in all cases by a pulse injection of EDTA or of a 0.3 M solution of mannose or GlcNAc. The binding curves exhibited quite different shapes, with a higher binding level and a slower dissociation of tetrameric MBL for both glycoconjugates (Fig. 4). The kinetic parameters of the interactions were determined by recording sensorgrams at varying MBL concentrations and evaluation of the data by global fitting as described in Materials and Methods. The values of maximal binding capacity (Rmax), association (kon) and dissociation (koff) rate constants and the resulting apparent KD values are listed in Table III. Consistent with our preliminary analyses (see Fig. 4), tetrameric MBL exhibited much lower koff values than trimeric MBL for binding to both Man and GlcNAc (13- and 3-fold, respectively). In contrast, trimeric MBL exhibited slightly higher kon values than tetrameric MBL for binding to either carbohydrate. The resulting apparent KD values were in the nanomolar range in all instances, with comparable values for both MBL forms in the case of GlcNAc and a 4-fold higher affinity for tetrameric MBL in the case of mannose (Table III). Tetrameric MBL showed a 4- and 12-fold higher maximal binding capacity (Rmax) for Man and GlcNAc, respectively, compared with the trimeric form (Table III).
FIGURE 4. Analysis by surface plasmon resonance spectroscopy of the interaction of trimeric and tetrameric MBL with Man-BSA (A) and GlcNAc-BSA (B). Man-BSA (2400 RU) and GlcNAc-BSA (3000 RU) were immobilized on the sensor chip as described in Materials and Methods. Sixty microliters of 2 nM trimeric or tetrameric MBL were injected in 145 mM NaCl, 1 mM CaCl2, 50 mM triethanolamine hydrochloride, pH 7.4, containing 0.005% surfactant P20, at a flow rate of 20 μl/min.
Table III. Kinetic and dissociation constants for binding of trimeric and tetrameric MBL to immobilized Man-BSA and GlcNAc-BSA
No binding of trimeric or tetrameric MBL to either glycoconjugate was observed when EDTA was substituted for Ca2+ in the running buffer or when Gal-BSA was immobilized on the sensor chip (data not shown), as expected from the known carbohydrate specificity of the C-type lectin domain of MBL (28).
Interaction properties of trimeric and tetrameric MBL with the MASPs and MAp19
Trimeric and tetrameric MBL were next immobilized on a sensorchip to compare their interaction properties with the recombinant MASPs and MAp19 used as soluble ligands. As shown in Fig. 5, MASP-1 (Fig. 5A), MASP-2 (Fig. 5B), MASP-3 (Fig. 5C), and MAp19 (Fig. 5D) each individually bound to immobilized trimeric or tetrameric MBL in very similar ways, with virtually identical association and dissociation phases in both cases. Likewise, as previously determined in the case of MASP-3 (25), the kinetic rate constants for MASP-1, MASP-2, and MAp19 were of the same order in the case of trimeric and tetrameric MBL, and the resulting KD values exhibited no significant difference (Table IV). No binding of the ligands to either trimeric or tetrameric MBL was observed when EDTA was substituted for CaCl2 in the running buffer.
FIGURE 5. Analysis by surface plasmon resonance spectroscopy of the interaction of the recombinant MASPs and MAp19 with trimeric and terameric MBL. Sixty microliters of 50 nM MASP-1 (A), 50 nM MASP-2 (B), 100 nM MASP-3 (C), and 100 nM MAp19 (D) were injected over 9,000 resonance units (RU) of immobilized trimeric MBL or 10,000 RU of immobilized tetrameric MBL.
Table IV. Kinetic and dissociation constants for binding of recombinant MASP-1, -2, -3 and MAp19 to immobilized trimeric and tetrameric MBL
Stoichiometry of the complexes between trimeric or tetrameric MBL and MASP-3 or MAp19
To further characterize the interaction properties of trimeric and tetrameric MBL, individual MBL-MASP complexes were assembled from the isolated components, and their molecular composition was studied using gel permeation chromatography. Due to the low amount of recombinant MASP-1 and MASP-2 available, experiments were performed using MASP-3 and MAp19. Both proteins have been shown to form dimers in the presence of Ca2+ ions (23, 25). Separate analyses of trimeric MBL and the MASP-3 dimer were first performed, indicating that the isolated proteins eluted at 24.1 and 25.2 min, respectively (Fig. 6A). When trimeric MBL and MASP-3 were mixed in a 1:2 molar ratio and then applied to the column, no residual peak corresponding to either protein was detected, but a new peak eluting ahead of trimeric MBL was observed, consistent with the formation of a 1:2 MBL/MASP-3 complex (Fig. 6A). When trimeric MBL and MASP-3 were mixed in a 1:1 or 1:4 ratio, the amount of complex formed and its elution position did not change, whereas extra peaks corresponding to either excess trimeric MBL or excess MASP-3 were observed. A minor peak eluting at 31 min and showing no interaction with MASP-3 was also detected in the trimeric MBL sample (Fig. 6A). This peak was unlikely to correspond to a MBL form of lower mass (i.e., with only one or two structural units) because no such species was detected to a significant extent by SDS-PAGE analysis of MBL-I (Fig. 1A). Analysis of the interaction between MASP-3 and tetrameric MBL yielded results very similar to those obtained with trimeric MBL, fully consistent with the formation of a 1:2 MBL/MASP-3 complex, even in the presence of a molar excess of MASP-3 (Fig. 6B).
FIGURE 6. Gel filtration analysis of complexes between trimeric or tetrameric MBL and MASP-3. Trimeric (A) or tetrameric (B) MBL was mixed with MASP-3 at molar ratios of 1:1 (70 pmol each), 1:2 (35 pmol of MBL and 70 pmol of MASP-3), and 1:4 (35 pmol of MBL and 140 pmol of MASP-3) in a volume of 100 μl and incubated for 1 h at room temperature. Isolated trimeric MBL, tetrameric MBL (35 pmol each), and MASP-3 (70 pmol), as well as the different mixtures were injected on a Superose 6 column equilibrated in 145 mM NaCl, 1 mM CaCl2, 50 mM triethanolamine hydrochloride, pH 7.4, and run at 0.5 ml/min.
Similar experiments were performed using MAp19 (Fig. 7). The Ca2+-dependent MAp19 dimer yielded a peak eluting at 34.8 min, whereas MBL/MAp19 complexes eluted only slightly ahead of trimeric or tetrameric MBL, due to the relatively small size of the MAp19 dimer (38 kDa) compared with either MBL species (Fig. 7). Complex formation was also clearly assessed by the disappearance of the MAp19 peak. When trimeric or tetrameric MBL and MAp19 were mixed in a 1:1 ratio, the peak corresponding to the complex became broader, with a shoulder corresponding to free MBL (Fig. 7). Conversely, a peak corresponding to the MAp19 dimer was observed when a 4-fold molar excess of MAp19 was used.
FIGURE 7. Gel filtration analysis of complexes between trimeric or tetrameric MBL and MAp19. Trimeric (A) or tetrameric (B) MBL was mixed with MAp19 at molar ratios of 1:1 (114 pmol each), 1:2 (57 pmol of MBL and 114 pmol of MAp19), and 1:4 (57 pmol of MBL and 228 pmol of MAp19) in a volume of 100 μl and analyzed as described in the legend to Fig. 6. Isolated MBL (57 pmol) and MAp19 (114 pmol) were also injected as controls.
It was concluded from these results that trimeric and tetrameric MBL each form 1:2 complexes with MASP-3 and MAp19. This stoichiometry is not modified in the presence of excess MBL, MASP-3, or MAp19.
Discussion
Human serum MBL is usually represented as a hexameric molecule, although structures ranging from dimers to hexamers have been observed previously by electron microscopy (14). Various relative amounts of these different oligomeric forms have been reported, with a predominance of hexamers (29), of dimers and trimers (15), or of trimers and tetramers (14). Two major forms of serum MBL, originally called MBL-I and -II, have been described recently, and it was suggested from SDS-PAGE analysis that they correspond to trimers and tetramers of the structural unit, respectively (8, 22). Using MALDI-TOF mass spectrometry analysis of purified native or reduced MBL-I and -II, we provide unambiguous experimental evidence that the MBL-I and -II forms contain 9 and 12 elementary chains and therefore consist of trimers and tetramers of the trimeric structural unit, respectively. It has been proposed that oligomerization involves interchain disulfide bridges between cysteine residues located at positions 5, 12, and 18 of MBL (21). Our results are fully consistent with this hypothesis and show that both assembly of the structural unit and oligomer formation involve disulfide bonds that are sensitive to reduction under nondenaturing conditions and therefore fully accessible to the solvent. The apparent masses determined for native trimeric and tetrameric MBL by SDS-PAGE analysis under nonreducing conditions (235 and 310 kDa, respectively) are close to those determined by mass spectrometry (228,098 ± 170 and 304,899 ± 229 Da). In contrast, the mass of the reduced chain determined by SDS-PAGE analysis (32 kDa) is clearly overestimated compared with the average value determined by mass spectrometry (25,340 ± 18 Da). It may be inferred from this observation that the major serum MBL species with apparent masses of 210 and 300 kDa, previously interpreted as dimers and trimers (15), rather represent trimers and tetramers, respectively. It should also be emphasized that, despite the fact that our MBL-II preparation was contaminated by a higher molecular mass band corresponding to the species termed MBL-III (8) (Fig. 1A), no such form could be detected by mass spectrometry analysis (Fig. 3). As reported in recent papers showing the oligomerization pattern of MBL (17, 18, 19, 30), higher oligomeric forms may exist in serum, but in much lower amounts than those of the trimeric and tetrameric forms. In contrast, the oligomerization pattern of human serum MBL may be more complex if the different common variant alleles of the MBL gene are considered. It has indeed been shown recently that these alleles give rise to significant amounts of lower oligomeric forms with reduced carbohydrate-binding ability (17, 18) that would likely not bind to the affinity column used in our purification protocol.
The mass value determined for the reduced MBL chain by MALDI-TOF mass spectrometry is close to the 25,500-Da value obtained by Larsen et al. (21), using surface-enhanced laser desorption/ionization mass spectrometry. Compared with the unmodified polypeptide, the MBL chain shows a mass increase that accounts for the posttranslational modifications usually observed in collagens, i.e., hydroxylation of the proline and lysine residues in the Y position of the repeating Gly-X-Y sequence and glycosylation of hydroxylysine residues. Previous analyses have indicated the presence of 3.7–4 hydroxyproline and 3.2–4 hydroxylysine residues per MBL chain (31, 32). Our own results are consistent with the presence of 4 hydroxyproline and 4 hydroxylysine residues, among which 3–4 carry the characteristic O-linked glucosylgalactosyl disaccharide observed in other proteins containing collagen-like sequences such as rat MBL (33), human C1q (34), or adiponectin (35). The fact that the mass values for the elementary chain of trimeric and tetrameric MBL are strictly identical indicates that preferential association into trimers or tetramers is not dependent on the extent of modification in the collagen-like region.
Analysis of the carbohydrate-binding properties of trimeric and tetrameric MBL reveals that they bind to Man-BSA with KD values of 2.2 and 0.55 nM and to GlcNAc-BSA with KD values of 1.2 and 0.96 nM, respectively, indicating high affinity in all cases. These values are in the same range as those determined previously for binding of rat MBL to glycosylated BSA (1 nM) (36) or of human MBL to mannan (2.3 nM) (37). Although the apparent KD values are of the same order for trimeric and tetrameric MBL, these two forms show strikingly different kinetic rate constants. Thus, tetrameric MBL exhibits a 13-fold lower dissociation rate constant from Man-BSA than trimeric MBL does, reflecting a much higher stability of the carbohydrate-tetrameric MBL interaction. Although the kon value is 3-fold lower in the case of tetrameric MBL, the maximal binding capacity to Man-BSA is 4-fold higher, and therefore much more mannose/tetrameric MBL complexes are expected to be formed in the presence of equivalent concentrations of both MBL forms (Fig. 4A). Comparable results were obtained using GlcNAc-BSA as a ligand, although in that case the major difference is that tetrameric MBL has a 12-fold higher maximal binding capacity than trimeric MBL. Thus, the oligomerization state of MBL has a direct effect on its carbohydrate-binding properties, with the tetrameric form showing increased binding capacity and higher interaction stability. It was shown previously that monovalent interaction between a single carbohydrate recognition domain of rat MBL-A and mannose is characterized by a very weak affinity (KD in the millimolar range) (38). Interaction of oligomeric MBL with carbohydrate arrays allows simultaneous engagement of several lectin domains, hence strengthening binding through an avidity phenomenon. In this respect, our data clearly indicate that the tetravalent MBL form has more avidity than the trivalent MBL form. Consistent with these observations, it was reported recently that mutated MBL molecules with a dimeric structure bind mannan with a markedly decreased capacity compared with wild-type MBL (21).
MBL-I and MBL-II were found to be preferentially associated in serum with MASP-1 and MAp19 and with MASP-3 and MASP-2, respectively (8). In contrast, using surface plasmon resonance spectroscopy, we have shown that recombinant MASP-3 binds to serum trimeric and tetrameric MBL with comparable binding constants (25). We have now extended this study to recombinant MASP-1, MASP-2, and MAp19, and we show that these proteins each interact individually with trimeric and tetrameric MBL with similar binding constants. Our data therefore indicate that the oligomerization state of MBL has no significant effect on its MASP-binding properties in vitro. In this respect, it was reported recently that purified tetrameric MBL induces C4 cleavage more efficiently than does trimeric MBL when added to MBL-deficient plasma (22), and it was suggested that this arises from a preferential association of serum MASP-2 with higher MBL oligomers. In light of our data, it should be pointed out that this difference may also be accounted for by the differential carbohydrate-binding capacity of trimeric and tetrameric MBL, because the complex between trimeric MBL and MASP-2 would exhibit a reduced capacity to bind to the mannan surface used in these experiments and therefore would trigger MASP-2 activation and C4 cleavage less efficiently.
Gel filtration analysis of the complexes between trimeric or tetrameric MBL and MASP-3 or MAp19 reveals a 1:2 stoichiometry in all cases. Although similar experiments could not be done using MASP-1 and MASP-2, it should be stressed that MASP-1 and MASP-3 share the same N-terminal CUB1-epidermal growth factor (EGF)-CUB2 domains known to mediate binding to human MBL or rat MBL-A (24, 39). In agreement with the observation that these proteases bind trimeric and tetrameric MBL with comparable affinities (Table IV), it can be inferred that MASP-1 associates with either MBL species with the same stoichiometry as determined for MASP-3. This reasoning likely also applies to MASP-2, because this protease comprises the same N-terminal CUB1-EGF modules as MAp19 (12, 13). The present results do not fully agree with those of Chen and Wallis (40) who reported that, whereas rat recombinant MBL-A dimers bind two MASP-1 or MASP-2 CUB1-EGF-CUB2 fragments, trimeric and tetrameric MBL-A forms bind up to four. It should be emphasized, however, that the latter complexes were observed only in the presence of a large molar excess (6- to 12-fold) of the MASP fragments over MBL-A.
The x-ray structure of human MAp19 has been solved recently, and site-directed mutagenesis studies have allowed us to map the MBL-binding site at the distal end of the CUB1 module (41). A three-dimensional model of the interaction between the MAp19 dimer and a triple-helical segment of MBL containing the putative MASP-binding site was proposed. This information was used to construct models of the trimeric and tetrameric MBL molecules and of their complexes with MAp19. As illustrated in Fig. 8B, the complex between the MBL tetramer and the MAp19 dimer is expected to be symmetrical, the latter protein interacting with binding sites located on two opposite collagen triple helices of tetrameric MBL. In contrast, it is clear that such a symmetry cannot be conserved in the complex between trimeric MBL and MAp19, because in this case MAp19 should interact with two contiguous collagen helices (Fig. 8A).
FIGURE 8. Model of the complexes between MAp19 and trimeric (A) or tetrameric MBL (B). The CUB and EGF modules of MAp19 (PDB ID: 1SZB; Ref. 39 ) are blue and green, respectively. The lectin domain of human MBL (positions 88–228) (PDB ID: 1HUP; Ref. 42 ) is red, and the collagen-like segment (positions 45–72) containing the MAp19 binding site as modeled by Gregory et al. (41) is gray. The structure of the MBL collagen-like segment encompassing residues 73–87 is unknown and is shown as gray dots. A bottom view of the complexes is shown.
In summary, our results show that the major MBL forms in human serum are trimers and tetramers of the structural unit. These two forms exhibit similar interaction properties with the MASPs, as shown by their common ability to form 1:2 complexes with MASP-3 and MAp19. In contrast, they markedly differ in their carbohydrate-binding characteristics, with the tetrameric form showing higher binding capacity and increased interaction stability. This differential behavior has potentially important functional implications because a prolonged interaction of the MBL/MASP-2 complex with microbial carbohydrates may be expected to facilitate MASP-2 activation and hence to trigger complement activation through C4 and C2 cleavage with more efficiency.
Disclosures
The authors have no financial conflict of interest.
Footnotes
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 work was supported by the Commissariat à l’Energie Atomique, the Centre National de la Recherche Scientifique, and the Université Joseph Fourier (Grenoble, France). A preliminary report of this study was presented by N.M.T. at the 20th International Complement Workshop in Honolulu, HI, July 13–18, 2004.
2 Address correspondence and reprint requests to Dr. Nicole Thielens, Laboratoire d’Enzymologie Moléculaire, Institut de Biologie Structurale Jean-Pierre Ebel, 41 rue Jules Horowitz, 38027 Grenoble Cedex 1, France. E-mail address: nicole.thielens{at}ibs.fr
3 Abbreviations used in this paper: MBL, mannan-binding lectin; MASP, MBL-associated serine protease; GlcNAc, N-acetylglucosamine; MAp19, 19-kDa MBL-associated protein; CUB module, module originally found in complement proteins C1r/C1s, Uegf, and bone morphogenetic protein-1; EGF, epidermal growth factor; Man, mannose; RU, resonance unit(s); the nomenclature of protein modules is that defined by Bork and Bairoch (1 ).
Received for publication October 6, 2004. Accepted for publication December 6, 2004.
References
Bork, P., A. Bairoch. 1995. Extracellular protein modules: a proposed nomenclature. Trends Biochem. Sci. 20:(Suppl.):C03. March.
Lu, J., C. Teh, U. Kishore, K. B. Reid. 2002. Collectins and ficolins: sugar pattern recognition molecules of the mammalian innate immune system. Biochim. Biophys. Acta 1572:387.
Holmskov, U., S. Thiel, J. C. Jensenius. 2003. Collectins and ficolins: humoral lectins of the innate immune defense. Annu. Rev. Immunol. 21:547.
Neth, O., D. L. Jack, A. W. Dodds, H. Holzel, N. J. Klein, M. W. Turner. 2000. Mannose-binding lectin binds to a range of clinically relevant microorganisms and promotes complement deposition. Infect. Immun. 68:688.
Turner, M. W.. 1996. Mannose-binding lectin: the pluripotent molecule of the innate immune system. Immunol. Today 17:532.
Matsushita, M., T. Fujita. 1992. Activation of the classical complement pathway by mannose-binding protein in association with a novel C1s-like serine protease. J. Exp. Med. 176:1497.
Thiel, S., T. Vorup-Jensen, C. M. Stover, W. Schwaeble, S. B. Laursen, K. Poulsen, A. C. Willis, P. Eggleton, S. Hansen, U. Holmskov, K. B. M. Reid, J. C. Jensenius. 1997. A second serine protease associated with mannan-binding lectin that activates complement. Nature 386:506.
Dahl, M. R., S. Thiel, M. Matsushita, T. Fujita, A. C. Willis, T. Christensen, T. Vorup-Jensen, J. C. Jensenius. 2001. MASP-3 and its association with distinct complexes of the mannan-binding lectin complement activation pathway. Immunity 15:127.
Vorup-Jensen, T., S. V. Petersen, A. Hansen, K. Poulsen, W. Schwaeble, R. B. Sim, K. B. M. Reid, S. J. Davis, S. Thiel, J. C. Jensenius. 2000. Distinct pathways of mannan-binding lectin (MBL)- and C1-complex autoactivation revealed by reconstitution of MBL with recombinant MBL-associated serine protease-2. J. Immunol. 165:2093.
Rossi, V., S. Cseh, I. Bally, N. M. Thielens, J. C. Jensenius, G. J. Arlaud. 2001. Substrate specificities of recombinant mannan-binding lectin-associated serine proteases-1 and -2. J. Biol. Chem. 276:40880.
Ambrus, G., P. Gal, M. Kojima, K. Szilagyi, J. Balczer, J. Antal, L. Graf, A. Laich, B. E. Moffatt, W. Schwaeble, R. B. Sim, P. Zavodszky. 2003. Natural substrates and inhibitors of mannan-binding lectin-associated serine protease-1 and -2: a study on recombinant catalytic fragments. J. Immunol. 170:1374.
Stover, C. M., S. Thiel, M. Thelen, N. J. Lynch, T. Vorup-Jensen, J. C. Jensenius, W. J. Schwaeble. 1999. Two constituents of the initiation complex of the mannan-binding lectin activation pathway of complement are encoded by a single structural gene. J. Immunol. 162:3481.
Takahashi, M., Y. Endo, T. Fujita, M. Matsushita. 1999. A truncated form of mannose-binding lectin-associated serine protease (MASP)-2 expressed by alternative polyadenylation is a component of the lectin complement pathway. Int. Immunol. 11:859.
Lu, J. H., S Thiel, H. Wiedemann, R. Timpl, K. B. Reid. 1990. Binding of the pentamer/hexamer forms of mannan-binding protein to zymosan activates the proenzyme C1r2C1s2 complex, of the classical pathway of complement, without involvement of C1q. J. Immunol. 144:2287.
Lipscombe, R. J., M Sumiya, J. A. Summerfield, M. W. Turner. 1995. Distinct physicochemical characteristics of human mannose binding protein expressed by individuals of differing genotype. Immunology 85:660.
Turner, M. W.. 1998. Mannose-binding lectin (MBL) in health and disease. Immunobiology 199:327.
Garred, P., F. Larsen, H. O. Madsen, C. Koch. 2003. Mannose-binding lectin deficiency revisited. Mol. Immunol. 40:73.
Roos, A., P. Garred, M. E. Wildenberg, N. J. Lynch, J. R. Munoz, T. C. Zuiverloon, L. H. Bouwman, N. Schlagwein, F. C. Fallaux van den Houten, M. C. Faber-Krol, et al 2004. Antibody-mediated activation of the classical pathway of complement may compensate for mannose-binding lectin deficiency. Eur. J. Immunol. 34:2589.
Terai, I., K. Kobayashi, M. Matsushita, H. Miyakawa, N. Mafune, H. Kikuta. 2003. Relationship between gene polymorphisms of mannose-binding lectin (MBL) and two molecular forms of MBL. Eur. J. Immunol. 33:2755.
Matsushita, M., R. A. Ezekowitz, T. Fujita. 1995. The Gly-54Asp allelic form of human mannose-binding protein (MBP) fails to bind MBP-associated serine protease. Biochem. J. 311:1021.
Larsen, F., H. O. Madsen, R. B. Sim, C. Koch, P. Garred. 2004. Disease-associated mutations in human mannose-binding lectin compromise oligomerization and activity of the final protein. J. Biol. Chem. 279:21302.
Dumestre-Perard, C., D. Ponard, G. J. Arlaud, N. Monnier, R. B. Sim, M. G. Colomb. 2002. Evaluation and clinical interest of mannan binding lectin function in human plasma. Mol. Immunol. 39:465.
Thielens, N. M., S. Cseh, S. Thiel, T. Vorup-Jensen, V. Rossi, J. C. Jensenius, G. J. Arlaud. 2001. Interaction properties of human mannan-binding lectin (MBL)-associated serine proteases-1 and -2, MBL-associated protein 19, and MBL. J. Immunol. 166:5068.
Cseh, S., L. Vera, M. Matsushita, T. Fujita, G. J. Arlaud, N. M. Thielens. 2002. Characterization of the interaction between L-ficolin/p35 and mannan-binding lectin-associated serine proteases-1 and -2. J. Immunol. 169:5735.
Zundel, S., S. Cseh, M. Lacroix, M. R. Dahl, M. Matsushita, J.-P. Andrieu, W. J. Schwaeble, J. C. Jensenius, T. Fujita, G. J. Arlaud, N. M. Thielens. 2004. Characterization of recombinant mannan-binding lectin-associated serine protease (MASP)-3 suggests an activation mechanism different from that of MASP-1 and MASP-2. J. Immunol. 172:4342.
Thielens, N. M., C. A. Aude, M. B. Lacroix, J. Gagnon, G. J. Arlaud. 1990. Ca2+ binding properties and Ca2+-dependent interactions of the isolated NH2-terminal fragments of human complement proteases C1-r and C1-s. J. Biol. Chem. 265:14469.
Rossi, V., C. Gaboriaud, M. Lacroix, J. Ulrich, J. C. Fontecilla-Camps, J. Gagnon, G. J. Arlaud. 1995. Structure of the catalytic region of human complement protease C1s: study by chemical cross-linking and three-dimensional homology modeling. Biochemistry 34:7311.
Drickamer, K.. 1992. Engineering galactose-binding activity into a C-type mannose-binding protein. Nature 360:183.
Wong, N., R. B. Sim. 1997. Comparison of the complement system protein complexes formed by C1q and MBL. Biochem. Soc. Trans. 25:41S.
Ohlenschlaeger, T., P. Garred, H. O. Madsen, S. Jacobsen. 2004. Mannose-binding lectin variant alleles and the risk of arterial thrombosis in systemic lupus erythematosus. N. Engl. J. Med. 351:260.
Ohtani, K., Y. Suzuki, S. Eda, T. Kawai, T. Kase, H. Keshi, Y. Sakai, S. Yamamoto, T. Sakamoto, N. Wakamiya. 1999. High-level and effective production of human mannan-binding lectin (MBL) in Chinese hamster ovary (CHO) cells. J. Immunol. Methods 222:135.
Vorup-Jensen, T., E. S. Sorensen, U. B. Jensen, W. Schwaeble, T. Kawasaki, Y. Ma, K. Uemura, N. Wakamiya, Y. Suzuki, T. G. Jensen, et al 2001. Recombinant expression of human mannan-binding lectin. Int. Immunopharmacol. 1:677.
Colley, K. J., J. U. Baenziger. 1987. Identification of the post-translational modifications of the core-specific lectin: the core-specific lectin contains hydroxyproline, hydroxylysine, and glucosylgalactosylhydroxylysine residues. J. Biol. Chem. 262:10290.
Shinkai, H., K. Yonemasu. 1979. Hydroxylysine-linked glycosides of human complement subcomponent C1q and various collagens. Biochem. J. 177:847.
Wang, Y., A. Xu, C. Knight, L. Y. Xu, G. J. Cooper. 2002. Hydroxylation and glycosylation of the four conserved lysine residues in the collagenous domain of adiponectin: potential role in the modulation of its insulin-sensitizing activity. J. Biol. Chem. 277:19521.
Lee, R. T., Y. Ichikawa, T. Kawasaki, K. Drickamer, Y. C. Lee. 1992. Multivalent ligand binding by serum mannose-binding protein. Arch. Biochem. Biophys. 299:129.
Kawasaki, N., T. Kawasaki, I. Yamashina. 1983. Isolation and characterization of a mannan-binding protein from human serum. J. Biochem. 94:937.
Iobst, S. T., M. R. Mormald, W. I. Weis, R. A. Dwek, K. Drickamer. 1994. Binding of sugar ligands to Ca2+-dependent animal lectins. I. Analysis of mannose binding by site-directed mutagenesis and NMR. J. Biol. Chem. 269:15505.
Wallis, R., R. B. Dodd. 2000. Interaction of mannose-binding protein with associated serine proteases: effects of naturally occurring mutations. J. Biol. Chem. 275:30962.
Chen, C. B., R. Wallis. 2001. Stoichiometry of complexes between mannose-binding protein and its associated serine proteases: defining functional units for complement activation. J. Biol. Chem. 276:25894.
Gregory, L. A., N. M. Thielens, M. Matsushita, R. Sorensen, G. J. Arlaud, J. C. Fontecilla-Camps, C. Gaboriaud. 2004. The x-ray structure of human mannan-binding lectin-associated protein 19 (MAp19) and its interaction site with mannan-binding lectin and L-ficolin. J. Biol. Chem. 279:29391.
Sheriff, S., C. Y. Chang, R. A. Ezekowitz. 1994. Human mannose-binding protein carbohydrate recognition domain trimerizes through a triple -helical coiled-coil. Nat. Struct. Biol. 1:789.(Florence Teillet, Bernard)