Mannan-binding lectin (MBL) initiates complement activation by binding to arrays of carbohydrates on the surfaces of pathogenic microorganisms and activating MBL-associated serine proteases (MASPs). Separate point mutations to the collagenous domain of human MBL are associated with immunodeficiency, caused by reduced complement activation by the variant MBLs as well as by lower serum MBL concentrations. In the work reported here, we have used the well characterized rat lectin pathway to analyze the molecular and functional defects associated with two of the variant proteins. Mutations Gly25 → Asp and Gly28 → Glu create comparable structural changes in rat MBL but the G28E variant activates complement >10-fold less efficiently than the G25D variant, which in turn has ∼7-fold lower activity than wild-type MBL. Analysis of mutant MBL · MASP-2 complexes assembled from recombinant components shows that reduced complement activation by both mutant MBLs is caused by failure to activate MASP-2 efficiently on binding to a mannan-coated surface. Disruption of MBL-MASP-2 interactions as well as to changes in oligomeric structure and reduced binding to carbohydrate ligands compared with wild-type MBL probably account for the intermediate phenotype of the G25D variant. However, carbohydrate binding and MASP-2 activation are ostensibly completely decoupled in complexes assembled from the G28E mutant, such that the rate of MASP-2 activation is no greater than the basal rate of zymogen MASP-2 autoactivation. Analogous molecular defects in human MBL probably combine to create the mutant phenotypes of immunodeficient individuals.

The lectin pathway of complement provides a primary line of defense against invading pathogens, selectively neutralizing microorganisms through an Ab-independent mechanism (1). The pathogen recognition components of this pathway, mannan-binding lectin (MBL,3 or mannose-binding protein) and serum ficolins, bind to arrays of carbohydrates or N-acetyl groups on the surfaces of bacteria, viruses, and parasites and activate MBL-associated serum proteases (MASPs) to trigger the downstream reaction cascade (2). Complement neutralizes pathogens directly via lysis and host-mediated phagocytosis and also stimulates and directs an effective adaptive immune response.

MBL is a mixture of large, oligomers of a subunit composed of three identical polypeptide chains (3, 4). Human MBL comprises dimers to hexamers of trimeric subunits, whereas rat MBL consists mainly of dimers, trimers, and tetramers of trimers. Oligomers resemble bouquets in which clusters of three globular carbohydrate-recognition domains (CRDs) are joined to rod-like collagenous stems by α-helical coiled coils. Individual stems are connected to one another by disulfide bonds between the N-terminal ends of polypeptides and splay apart at a break in the collagenous domain. MBL circulates in serum bound to three different zymogen MASPs (MASPs-1, -2, and -3) and a small nonenzymatic protein called MAp19 or sMAP (5, 6). When MBL binds to the surface of a microorganism, MASP-2 activates through autolysis and cleaves complement components C2 and C4 to form the C3 convertase (C4b2a), which in turn activates the downstream reaction cascade. The roles of MASP-1, MASP-3, and MAp19 are unknown.

Three separate mutations within the collagen-like domain of human MBL are associated with a common immunodeficiency (7). Individuals, either homozygous or heterozygous for variant alleles, are susceptible to a wide range of bacterial, viral, and parasitic infections, particularly in early childhood before the adaptive immune system is established (8), or when adaptive immunity is compromised, for example during HIV infection or following chemotherapy (9, 10). Originally, immunodeficiency was thought to be caused by low levels of MBL in serum. However, recent work has shown that serum MBL levels are only slightly reduced but that the variant proteins are dysfunctional (11).

The well characterized rat lectin pathway has greatly facilitated analysis of the structural defects associated with both the homozygous and heterozygous variant MBL genotypes (12, 13, 14). Rat MBLs containing mutations equivalent to those in human MBL have low complement-fixing activities that reflect distinct molecular defects. The mutation R23C (R32C in human MBL) causes adventitious disulfide bond formation that hampers formation of the larger MBL oligomers. The reduced activity of this variant is due to the lower complement-fixing activities of the smaller oligomeric forms. The reduced activities of the G25D and G28E mutations (G34D and G37E in human MBL) are more complex. Comparable changes in oligomeric structure and in MASP binding probably both contribute to the functional defects of the mutants. However, the complement-fixing activity of the G28E mutant is much lower than the activity of the G25E variant, so there must be an additional, unknown defect.

In the work described here, we have used recombinant rat proteins to elucidate and quantify the functional and molecular defects of the variant G25D and G28E MBLs and thereby discover the basis of their different complement-fixing activities. The data reveal that although changes in sugar binding and in MBL.MASP-2 interactions probably both contribute to the reduced activities of the mutants, the more severely impaired phenotype of the G28E mutant is caused by failure of the MBL · MASP-2 complex to autoactivate on attachment to a mannan-coated surface. We propose a model to explain how carbohydrate binding is decoupled from MASP-2 activation in this MBL variant.

Recombinant wild-type and mutant rat MBLs were produced in Chinese hamster ovary cells using serum-free medium and were purified by affinity chromatography using mannose-Sepharose (4). MASP-2K, in which the arginine residue at the cleavage site for zymogen activation (Arg424) is replaced by a lysine residue, was produced and purified as described (15).

To remove endogenous MASPs, guinea-pig serum (5 ml) was incubated with 1 ml of MBL-agarose (16), equilibrated in 50 mM Tris-HCl (pH 7.5), containing 145 mM NaCl, 5 mM MgCl2, and 5 mM CaCl2, with mixing at 4°C. After 16 h, the serum was separated from the MBL agarose by filtration and was loaded sequentially on to two mannose-Sepharose columns (0.5 ml), equilibrated in the same buffer, to remove endogenous MBL. The resulting MBL- and MASP-depleted serum was stored frozen until required.

To measure lectin pathway-dependent hemolysis, aliquots of MBL (0.4 ml containing concentrations ranging from 0.02 to 60 μg/ml) were incubated with mannan-coated sheep erythrocytes (1 × 108 cells in 0.1 ml) for 30 min in gelatin-Veronal buffer at 25°C. Cells were then washed with 0.5 ml of buffer to remove any unbound MBL and were incubated with a fixed amount of MASP-2 (0.65 μg in a total volume of 460 μl). After 10 min at 37°C, guinea pig serum (40 μl), depleted of MASPs and MBL, was added, and cells were incubated for a further 1 h with mixing at 37°C. Finally, gelatin-Veronal buffer (700 μl) was added to each sample, the remaining cells were pelleted by centrifugation, and the amount of cell lysis was measured using absorbance at 541 nm. Data were expressed as a percentage of the absorbance of an equivalent volume of cells totally lysed in water, correcting for lysis observed in the absence of MBL. The data were fitted to a sigmoidal curve using Microcal Origin (Microcal Software). Relative complement-fixing activities were calculated from the concentrations of MBL required for 50% hemolysis compared with wild-type MBL in assays performed using the same batch of mannan-coated erythrocytes.

Microtiter plates (Nalge Nunc) were coated with mannan (1 μg; Sigma-Aldrich) in 15 mM Na2CO3, 35 mM NaHCO3 (pH 9.6), blocked with 0.1% (w/v) human serum albumin in TBS (10 mM Tris-Cl, 140 mM NaCl (pH 7.4)) then washed with TBS containing 0.05% (v/v) Tween 20, and 5 mM CaCl2 (wash buffer). Serial dilutions of rat MBL were prepared in binding buffer (TBS with 10 mM CaCl2, 0.05% (v/v) Triton X-100, 0.1% (w/v) human serum albumin), added to the plates and incubated for 16 h at 4°C. The plate was washed three times with wash buffer to remove any unbound protein and incubated with anti-rat MBL-A antiserum (17). Bound Ab was detected with alkaline phosphatase-conjugated anti-rabbit IgG and the colorimetric substrate p-nitrophenyl phosphate (Sigma-Aldrich).

Rat MBL · MASP-2 complexes were formed by mixing 1 μg of MASP-2K with 1 μg of recombinant MBL in 1 ml of binding buffer and incubating overnight at 4°C. The following day, C4 activation was measured using a modification of the assay developed by Petersen et al. (18). A microtiter plate was coated with mannan, blocked, and washed. Serial dilutions of preformed MBL · MASP complexes were prepared in binding buffer, added to the plate, and incubated for 1 h at room temperature. The plate was washed three times with wash buffer and 0.1 μg of purified human C4 (19) in 4 mM barbital, 145 mM NaCl, 2 mM CaCl2, 1 mM MgCl2 (pH 7.4) was added to each well. After 90 min at 37°C, the plate was washed again and bound C4b detected using alkaline phosphatase-conjugated chicken anti-human C4c (Immunsystem AB) and the colorimetric substrate p-nitrophenyl phosphate.

MASP-2K and either wild-type or mutant MBL were incubated with mannan-Sepharose (20 μl in a total volume of 130 μl), in 50 mM Tris-HCl (pH 7.5), containing 145 mM NaCl, 5 mM MgCl2, and 5 mM CaCl2, at 37°C with mixing. To ensure that MASP-2 was fully bound to MBL, the concentration of MASP-2 (30 μg/ml, equivalent to 2 × 10−7 M) was higher than the KI values for binding (∼3 × 10−8 M) (14) and a 2-fold excess of MBL was used (60 μg/ml). MBL concentrations were also much greater than the apparent Kd for mannan (≤20 ng/ml), thereby ensuring maximal binding to the mannan-Sepharose. At various time points, aliquots were removed and proteins were pelleted following precipitation using 30% trichloroacetic acid. Proteins were separated by SDS-PAGE and the amount of zymogen activation was quantified by scanning gels using a ChemiGenius (2) (Syngene).

To quantify the amounts of MBL and MASP-2 that were associated with the mannan-Sepharose in the activation assays, protein mixtures prepared as described above were mixed with mannan-Sepharose and were incubated at 4°C for 30 min. The mannan-Sepharose was pelleted by centrifugation, washed twice with buffer containing 1.25 M NaCl, and bound proteins were eluted using 25 mM EDTA. Proteins were separated on a 12% SDS polyacrylamide gel under reducing conditions and were quantified by scanning gels. Data are expressed as the mean ± S.E. from four separate experiments.

Mutations to glycine residues at positions 25 and 28 in rat MBL destabilize the collagen-like domain and the adjacent N-terminal domain, thereby disrupting the oligomeric structure and interactions with MASPs (13, 20). To quantify and thus explain the reduced complement-fixing activities of the G25D and G28E mutants, we analyzed each of the key steps in the activation process, including carbohydrate binding, MASP-2 autoactivation and C4 activation by MASP-2. The results reveal how defects at different levels combine to produce the overall phenotypes of the mutants.

Although the G25D and G28E mutations are distinct from the MASP-binding sites on MBL, they indirectly affect interactions with all three MASPs (14). It was therefore important to establish that the phenotypes of the mutants can be explained by reduced activation of MASP-2 alone. Consequently, we developed a novel hemolytic assay to measure complement activation by recombinant MBL · MASP-2 complexes assembled from wild-type or mutant MBLs, using serum depleted of endogenous MBL and MASPs and mannan-coated erythrocytes as targets. To ensure that complexes were reconstituted entirely from zymogen MASP-2, we used a modified form of the enzyme, called MASP-2K, which is secreted and purified as a zymogen and autoactivates more slowly than native MASP-2 to retain full enzymic activity (15). Addition of increasing amounts of MBL with MASP-2K resulted in increasing lectin pathway-dependent hemolysis of coated erythrocytes (Fig. 1 A). MBL or MASP-2K alone had no detectable hemolytic activity, even at high concentrations, confirming that there was no residual MBL or MASP activity in the depleted serum and demonstrating that MBL · MASP-2 complexes are sufficient to trigger lectin pathway activation in absence of MASP-1 and MASP-3.

FIGURE 1.

Complement activation by recombinant wild-type and mutant MBL · MASP-2K complexes. Complement activation was measured by hemolysis of mannan-coated sheep erythrocytes, using serum depleted of endogenous MBL and MASPs. A, Reconstitution of lectin pathway activation, using recombinant wild-type MBL and MASP-2K. B, Complement activation by wild-type and mutant MBL · MASP-2K complexes.

FIGURE 1.

Complement activation by recombinant wild-type and mutant MBL · MASP-2K complexes. Complement activation was measured by hemolysis of mannan-coated sheep erythrocytes, using serum depleted of endogenous MBL and MASPs. A, Reconstitution of lectin pathway activation, using recombinant wild-type MBL and MASP-2K. B, Complement activation by wild-type and mutant MBL · MASP-2K complexes.

Close modal

The complement activation assays verified that activities of the mutant MBLs are substantially reduced compared with wild-type MBL (Fig. 1 B). The activity of the G25D mutant was ∼7-fold lower than wild-type MBL, whereas the activity of the G28E mutant was reduced by ∼50-fold. Indeed, the relative activities of the mutant MBL · MASP-2K complexes in the absence of MASP-1 and MASP-3 were almost the same as the relative activities measured in serum containing all three endogenous MASPs (13) implying that faulty MASP-2 activation accounts for lower complement-fixing activities of the mutants.

Reduced activation of MBL · MASP-2 complexes, regardless of the underlying mechanism, would lead to a corresponding decrease in activation of downstream components and a reduction in complement activation overall. Nevertheless, it was still necessary to show that activation of MASP-2 is sufficiently impaired to fully explain the lower complement-fixing activities of the mutant MBLs to exclude the possibility of additional, more complex processes contributing to the mutant phenotypes. We therefore measured C4 activation by reconstituted MBL · MASP-2K complexes (Fig. 2). As expected, higher concentrations of mutant MBL ·MASP complexes were required to activate C4, demonstrating that both mutants fail to activate MASP-2 as effectively as wild-type MBL. Moreover, the relative activities of the G25D mutant and the G28E mutant are ∼11- and ∼40-fold lower than wild-type MBL, which are comparable to their relative complement-fixing activities (Table I). Thus, the functional defects can be explained by failure of the MBL · MASP-2 complexes to activate C4 and are not augmented by additional indirect affects on the complement cascade.

FIGURE 2.

C4 activation by wild-type and mutant MBL · MASP-2K complexes. In the absence of MASP-2K, neither wild-type nor mutant MBLs activated C4 significantly. C4b deposition was inhibited at high concentrations of wild-type MBL · MASP-2K complexes.

FIGURE 2.

C4 activation by wild-type and mutant MBL · MASP-2K complexes. In the absence of MASP-2K, neither wild-type nor mutant MBLs activated C4 significantly. C4b deposition was inhibited at high concentrations of wild-type MBL · MASP-2K complexes.

Close modal
Table I.

Properties of mutant MBLs expressed relative to wild-type MBL

MBLComplement-Fixing Activity of MBL · MASP-2K ComplexC4-ActivationMannan BindingKI for MBL-MASP-2 InteractionaAutoactivation of MBL · MASP-2 Complex
Wild type 
 G25D 0.14 ± 0.04 0.085 ± 0.05 0.26 ± 0.01 0.77 ± 0.08 0.10 ± 0.02 
 G28E 0.020 ± 0.005 0.024 ± 0.012 0.29 ± 0.01 0.40 ± 0.1 0.035 ± 0.010 
MBLComplement-Fixing Activity of MBL · MASP-2K ComplexC4-ActivationMannan BindingKI for MBL-MASP-2 InteractionaAutoactivation of MBL · MASP-2 Complex
Wild type 
 G25D 0.14 ± 0.04 0.085 ± 0.05 0.26 ± 0.01 0.77 ± 0.08 0.10 ± 0.02 
 G28E 0.020 ± 0.005 0.024 ± 0.012 0.29 ± 0.01 0.40 ± 0.1 0.035 ± 0.010 
a

Data from Ref.14 .

Recent work on human MBL has shown that the differences in oligomeric structures of the mutants compared with wild-type MBL lead to reduced affinities for target carbohydrate ligands (21). Both mutations cause comparable changes to the oligomeric structure of rat MBL, resulting in a lower proportion of trimers and tetramers of the trimeric subunit and more dimers and monomers of trimers (13). Similar changes are observed in human MBL. Because the smaller oligomers have fewer CRDs, they bind to arrays of carbohydrate ligands with lower affinities. To quantify these differences, we measured MBL binding to mannan-coated plates, using a simple Ab-based assay. The apparent KD for binding of wild-type MBL to mannan was ∼6 ng of MBL/ml, corresponding to 8 × 10−11 M of MBL subunits. Interestingly, the maximum amount of MBL bound to the mannan surface was comparable for wild-type and mutant MBLs, revealing that the mutations do not significantly affect the extent of binding. However, both mutants bound to the mannan-coated surfaces with ∼4-fold lower affinities than wild-type MBL (Fig. 3). These differences in sugar binding properties could account for part of the reduction in the overall complement-fixing activities of the mutant MBLs.

FIGURE 3.

Binding of wild-type and mutant MBLs to mannan. The apparent KD was the concentration of MBL required for half-maximal binding to mannan.

FIGURE 3.

Binding of wild-type and mutant MBLs to mannan. The apparent KD was the concentration of MBL required for half-maximal binding to mannan.

Close modal

Defective MASP-2 binding might also be a factor. However, the affinity for MASP-2 of the G28E mutant is reduced by only 2.5-fold relative to wild-type MBL, whereas there is only a marginal loss of affinity for the G25D mutant (14) (Table I). Overall therefore, differences in carbohydrate binding and in interactions with MASP-2 could account for 5- to 10-fold decreases in the complement-fixing activities of the mutant MBLs and can therefore largely explain the defective phenotype of the G25D mutant. However, they cannot account for the 50-fold loss in activity of the G28E mutant.

When MBL binds to a carbohydrate surface, it initiates activation by increasing the rate of MASP-2 autocatalysis (15). Having examined differences in sugar binding and considered the effects of MBL-MASP-2 interactions, we next sought to quantify autoactivation itself. To compare activation of MBL · MASP-2 complexes directly, we measured MBL-dependent MASP-2 autolysis, using mannan-Sepharose as a carbohydrate ligand (Fig. 4,A). High protein concentrations were used to minimize differences in the amounts of wild-type and mutant MBL · MASP-2K complexes bound to the affinity matrix. Initially all of the MASP-2K in complexes was in the zymogen form. Upon incubation with mannan-Sepharose, MASP propeptides were progressively cleaved (Fig. 4,B). The half-time for activation of complexes assembled from wild-type MBL was ∼18 min (Fig. 5). By contrast, complexes formed from mutant MBLs autoactivated more slowly, with half-times of ∼200 and 530 min for the G25D and G28E mutants (Figs. 5 and 6), reductions of ∼11- and ∼30-fold, respectively. Thus, MASP-2 activation by both mutant MBLs is clearly defective but the G28E mutant is more severely affected than the G25D mutant.

FIGURE 4.

Autoactivation of MBL · MASP-2K complexes on mannan-Sepharose analyzed by autolysis of MASP-2K. A, Schematic representation of the domain organization of MASP-2. CUB, C1r/C1s/Uegf/bone morphogenetic protein; EGF, epidermal growth factor. Balls on sticks show the positions of the three N-linked glycosylation sites. The arrow indicates the site that is cleaved during autoactivation of MASP-2. A solid line shows a disulfide bond line that connects the N-terminal domains to the serine protease domain in activated MASP-2. B, Proteins were separated on a 12% SDS-polyacrylamide gel under reducing conditions and were detected by staining with Coomassie Blue. The migration positions of MBL, zymogen MASP-2K and the N- and C-terminal fragments of activated MASP-2K are indicated on the right. Migration positions of molecular mass markers are shown on the left.

FIGURE 4.

Autoactivation of MBL · MASP-2K complexes on mannan-Sepharose analyzed by autolysis of MASP-2K. A, Schematic representation of the domain organization of MASP-2. CUB, C1r/C1s/Uegf/bone morphogenetic protein; EGF, epidermal growth factor. Balls on sticks show the positions of the three N-linked glycosylation sites. The arrow indicates the site that is cleaved during autoactivation of MASP-2. A solid line shows a disulfide bond line that connects the N-terminal domains to the serine protease domain in activated MASP-2. B, Proteins were separated on a 12% SDS-polyacrylamide gel under reducing conditions and were detected by staining with Coomassie Blue. The migration positions of MBL, zymogen MASP-2K and the N- and C-terminal fragments of activated MASP-2K are indicated on the right. Migration positions of molecular mass markers are shown on the left.

Close modal
FIGURE 5.

Time-dependent autoactivation of wild-type and mutant MBL · MASP-2K complexes. A, The target ligand was mannan-Sepharose. B, To confirm that mutant MBL · MASP-2K complexes were bound to the mannan-Sepharose, eluted proteins were separated on a 12% SDS polyacrylamide gel and were stained using Coomassie Blue. The amounts of MBL and MASP-2, expressed relative to wild-type MBL and MASP-2 complexes, were 0.86 ± 0.15 and 0.92 ± 0.25 for the G25E MBL · MASP-2K complex and 0.77 ± 0.06 and 0.73 ± 0.08 for the G28E complex. MASP-2K alone did not bind to mannan-Sepharose (data not shown), so it is pulled down only through its association with MBL.

FIGURE 5.

Time-dependent autoactivation of wild-type and mutant MBL · MASP-2K complexes. A, The target ligand was mannan-Sepharose. B, To confirm that mutant MBL · MASP-2K complexes were bound to the mannan-Sepharose, eluted proteins were separated on a 12% SDS polyacrylamide gel and were stained using Coomassie Blue. The amounts of MBL and MASP-2, expressed relative to wild-type MBL and MASP-2 complexes, were 0.86 ± 0.15 and 0.92 ± 0.25 for the G25E MBL · MASP-2K complex and 0.77 ± 0.06 and 0.73 ± 0.08 for the G28E complex. MASP-2K alone did not bind to mannan-Sepharose (data not shown), so it is pulled down only through its association with MBL.

Close modal
FIGURE 6.

Comparison of autoactivation of zymogen MASP-2 with MBL · MASP-2 complexes assembled from the G28E variant MBL.

FIGURE 6.

Comparison of autoactivation of zymogen MASP-2 with MBL · MASP-2 complexes assembled from the G28E variant MBL.

Close modal

To quantify the relative amounts of MBL and MASP-2K bound to the mannan, the mannan-Sepharose was pelleted and bound proteins were analyzed. The amounts of mutant MBL · MASP-2 complexes bound to the affinity matrix were only marginally less than amounts of wild-type complexes (Fig. 5 B), thus reduced MASP-2 autoactivation could not be caused simply by a smaller proportion of mutant than wild-type complexes binding to mannan. Nevertheless, interactions between mutant complexes and mannan would be more transient than those between wild-type MBL and mannan, thereby reducing the time available for autoactivation to occur. Thus, defective interactions between MBL and MASP-2 and between MBL and mannan could still account for reduced MASP-2K activation by the G25D mutant and in part for the reduced MASP-2K activation by the G28E mutant. However, the phenotype of the G28E mutant can only be fully explained if MASP-2 activation by the MBL is itself impaired.

To reveal the full extent of the defects in the mutant MBLs, we compared activation of MBL · MASP-2 complexes with autoactivation of zymogen MASP-2K alone. In the absence of MBL, zymogen MASP-2K autoactivated at a low but significant rate (Fig. 6). The half-time, measured in two separate experiments, was 540 ± 15 min, comparable to the half-time for activation of MASP-2 by the G28E mutant (530 ± 20 min). We can therefore conclude that wild-type MBL increases the basal autoactivation rate of MASP-2K by ∼30-fold when complexes bind to mannan-Sepharose. However, the G25D mutant increases the activation rate by only 3-fold, while the G28E mutant fails to enhance the basal rate of autoactivation significantly, even though both MBLs bind to MASP-2 and interact with the carbohydrate ligand. Thus, the link between carbohydrate binding and MASP activation is partially decoupled in G25D mutant and wholly decoupled in the G28E mutant. These molecular defects would account for the low complement-fixing activities of the mutant MBL · MASP-2 complexes and probably contribute to the functional defects of the equivalent human variants.

In the work described here, we have shown how multiple factors combine to reduce activation of lectin pathway by mutant MBLs. Both mutations destabilize the collagenous domain of MBL to comparable extents (14), disrupting assembly of MBL oligomers during biosynthesis. Consequently, fewer of the larger MBL oligomers and more of the smaller oligomers are produced. The reduced activities of the MBLs can be partly explained by weaker, more transient binding of the smaller oligomers for surface carbohydrates. Similar changes have recently been described in preparations of variant human MBLs (21). In vivo, the effects of these changes are likely to depend on the nature and density of the ligands on the surface of pathogens. Although the smaller MBLs oligomers probably still bind to microorganisms with high-density arrays of exposed mannose-like epitopes because sufficient interactions between CRDs and carbohydrate epitopes would permit stable binding, pathogens with fewer exposed ligands may evade MBL altogether, thereby increasing the risk of serious infection.

Weaker or more transient interactions with MASP-2 probably also contribute to the overall reduction in complement-fixing activities of the mutants, albeit to a relatively minor extent. Based on estimates of serum MBL concentrations in humans (∼1 μg/ml) and the KD values for the MBL-MASP interactions (∼3 nM, equivalent to 0.5 and 1 μg/ml for MBL trimers and tetramers (22)) the lower binding and affinities of mutant MBLs for MASP-2 would cause small reductions in the amounts of MBL · MASP-2 complexes circulating in the serum. However, the problem would be further exacerbated, in individuals with low serum MBL concentrations, a predicament frequently associated with the MBL structural mutations, because an even smaller proportion of MBL and MASP molecules would interact with each other under these conditions.

We have shown that the characteristic phenotype of the G28E variant MBL is caused by failure to trigger MASP-2 activation directly when it binds to a carbohydrate surface. Thus, the rate of MASP-2 activation in mutant complexes is comparable to the rate of autoactivation of unassociated zymogen MASP-2. Interestingly, however, MBL · MASP-2K complexes assembled from the G28E mutant still had some residual complement-fixing activity, albeit >100-fold lower than native MBL · MASP-2 complexes. Because the mutant MBL cannot initiate activation of the MASP directly, activation must occur via an alternative mechanism. The most likely explanation is that the low level of complement activation arises through the capacity of the MASP-2 to autoactivate, which is observed even in the absence of MBL. Because the mutant MBL still binds to carbohydrate surfaces, it would tether MASP-2 molecules and could therefore provide a focus for C4 deposition and subsequent complement activation, following spontaneous MASP-2 activation. Thus, some complement activation would still occur, even though it is not triggered through the normal activation mechanism.

It is worth considering that the experiments reported here were conducted using a mutant of MASP-2, which autoactivates more slowly than the native MASP. This strategy was necessary because wild-type MASP-2 partially autoactivates during biosynthesis, yielding a mixture of zymogen and activated enzyme that makes detailed kinetic studies unfeasible. Autoactivation of the native MBL · MASP complex is likely to be even faster than activation of MBL · MASP-2K. Consequently, activation of the G28E mutant MBL · MASP-2 complex might be even more impaired relative to activation by the native complex.

MBL · MASP activation is normally triggered when the CRDs of MBL interact with the surface of a microorganism (15). Changes in the structure of MBL initiate activation by increasing the rate of MASP-2 autocatalysis. Each MASP dimer bridges two MBL subunits by binding close to the break in the collagenous domain (sometimes called the hinge), which causes the stems to splay apart (23) (Fig. 7). When the CRDs bind to the surface of a pathogen, movement at the hinge probably allows the rod-like stems to move relative to one another. Changes to the intersecting angle between subunits transform the associated MASP, thereby initiating activation. Gly28 is immediately adjacent to the hinge toward the N-terminal end of MBL. By disrupting this portion of the collagenous domain, the Gly28 → Glu mutation probably prevents the correct motion of the collagenous stems upon binding to a sugar surface, thereby decoupling carbohydrate binding from MASP-2 activation.

FIGURE 7.

Model for defective MASP-2 activation by the G28E variant MBL. A and B, Activation of wild-type MBL · MASP-2 complex. Binding to a carbohydrate surface, imparts changes in the relative positions of MBL subunits via movement at the hinge region. C and D, Extra flexibility near the hinge region in the G28E mutant allows binding to the carbohydrate surface, without the accompanying conformational changes required to induce MASP-2 activation.

FIGURE 7.

Model for defective MASP-2 activation by the G28E variant MBL. A and B, Activation of wild-type MBL · MASP-2 complex. Binding to a carbohydrate surface, imparts changes in the relative positions of MBL subunits via movement at the hinge region. C and D, Extra flexibility near the hinge region in the G28E mutant allows binding to the carbohydrate surface, without the accompanying conformational changes required to induce MASP-2 activation.

Close modal

We thank Katrine Abelskov for critical reading of the manuscript.

The authors have no financial conflict of interest.

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 Wellcome Trust Programme Grant 060574.

3

Abbreviations used in this paper: MBL, mannan-binding lectin; MASP, MBL-associated serine protease; CRD, carbohydrate-recognition domain.

1
Turner, M. W..
1996
. Mannose-binding lectin: the pluripotent molecule of the innate immune system.
Immunol. Today
17
:
532
.-540.
2
Fujita, T., M. Matsushita, Y. Endo.
2004
. The lectin-complement pathway—its role in innate immunity and evolution.
Immunol. Rev.
198
:
185
.-202.
3
Wallis, R., K. Drickamer.
1997
. Asymmetry adjacent to the collagen-like domain in rat liver mannose-binding protein.
Biochem. J.
325
:
391
.-400.
4
Wallis, R., K. Drickamer.
1999
. Molecular determinants of oligomer formation and complement fixation in mannose-binding proteins.
J. Biol. Chem.
274
:
3580
.-3589.
5
Schwaeble, W., M. Dahl, S. Thiel, C. Stover, J. Jensenius.
2002
. The mannan-binding lectin-associated serine proteases (MASPs) and MAp19: four components of the lectin pathway activation complex encoded by two genes.
Immunobiology
205
:
455
.-466.
6
Wallis, R..
2000
. C-Type lectins and collectins. S. Y. C. Wong, and G. Arsequell, eds.
Immunobiology of Carbohydrates
Wiley-VCH, Weinheim.
7
Sumiya, M., M. Super, P. Tabona, R. J. Levinsky, A. Takayuki, M. W. Turner, J. A. Summerfield.
1991
. Molecular basis of opsonic defect in immunodeficient children.
Lancet
337
:
1569
.-1570.
8
Lipscombe, R. J., Y. L. Lau, R. J. Levinsky, M. Sumiya, J. A. Summerfield, M. W. Turner.
1992
. Identical point mutation leading to low levels of mannose binding protein and poor C3b mediated opsonization in Chinese and Caucasian populations.
Immunol. Lett.
32
:
253
.-258.
9
Garred, P., H. O. Madsen, U. Balslev, B. Hofmann, C. Pedersen, J. Gerstoft, A. Svejgaard.
1997
. Susceptibility to HIV infection and progression of AIDS in relation to varient alleles of mannose-binding lectin.
Lancet
349
:
236
.-240.
10
Neth, O., I. Hann, M. W. Turner, N. J. Klein.
2001
. Deficiency of mannose-binding lectin and burden of infection in children with malignancy: a prospective study.
Lancet
358
:
614
.-618.
11
Garred, P., F. Larsen, H. O. Madsen, C. Koch.
2003
. Mannose-binding lectin deficiency revisited.
Mol. Immunol.
40
:
73
.-84.
12
Wallis, R..
2002
. Dominant effects of mutations in the collagenous domain of mannose-binding protein.
J. Immunol.
168
:
4553
.-4558.
13
Wallis, R., J. Y. Cheng.
1999
. Molecular defects in variant forms of mannose-binding protein associated with immunodeficiency.
J. Immunol.
163
:
4953
.-4959.
14
Wallis, R., J. M. Shaw, J. Uitdehaag, C. B. Chen, D. Torgersen, K. Drickamer.
2004
. Localization of the serine protease-binding sites in the collagen-like domain of mannose-binding protein: indirect effects of naturally occurring mutations on protease binding and activation.
J. Biol. Chem.
279
:
14065
.-14073.
15
Chen, C. B., R. Wallis.
2004
. Two mechanisms for mannose-binding protein modulation of the activity of its associated serine proteases.
J. Biol. Chem.
279
:
26058
.-26065.
16
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
.-25902.
17
Wagner, S., N. J. Lynch, W. Walter, W. J. Schwaeble, M. Loos.
2003
. Differential expression of the murine mannose-binding lectins A and C in lymphoid and nonlymphoid organs and tissues.
J. Immunol.
170
:
1462
.-1465.
18
Petersen, S. V., S. Thiel, L. Jensen, R. Steffensen, J. C. Jensenius.
2001
. An assay for the mannan-binding lectin pathway of complement activation.
J. Immunol. Methods
257
:
107
.-116.
19
Dodds, A. W..
1993
. Small-scale preparation of complement components C3 and C4.
Methods Enzymol.
223
:
46
.-61.
20
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
.-30969.
21
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
.-21311.
22
Teillet, F., B. Dublet, J. P. Andrieu, C. Gaboriaud, G. J. Arlaud, N. M. Thielens.
2005
. The two major oligomeric forms of human mannan-binding lectin: chemical characterization, carbohydrate-binding properties, and interaction with MBL-associated serine proteases.
J. Immunol.
174
:
2870
.-2877.
23
Feinberg, H., J. C. Uitdehaag, J. M. Davies, R. Wallis, K. Drickamer, W. I. Weis.
2003
. Crystal structure of the CUB1-EGF-CUB2 region of mannose-binding protein associated serine protease-2.
EMBO J.
22
:
2348
.-2359.