Serotype III group B streptococci (GBS) are a common cause of neonatal sepsis and meningitis. Although deficiency in maternal capsular polysaccharide (CPS)-specific IgG correlates with susceptibility of neonates to the GBS infection, serum deficient in CPS-specific IgG mediates significant opsonophagocytosis. This IgG-independent opsonophagocytosis requires activation of the complement pathway, a process requiring the presence of both Ca2+ and Mg2+, and is significantly reduced by chelating Ca2+ with EGTA. In these studies, we defined a role of L-ficolin/mannose-binding lectin-associated serine protease (MASP) complexes in Ca2+-dependent, Ab-independent opsonophagocytosis of serotype III GBS. Incubation of GBS with affinity-purified L-ficolin/MASP complexes and C1q-depleted serum deficient in CPS-specific Ab supported opsonophagocytic killing, and this killing was inhibited by fluid-phase N-acetylglucosamine, the ligand for L-ficolin. Binding of L-ficolin was proportional to the CPS content of individual strains, and opsonophagocytic killing and C4 activation were inhibited by fluid-phase CPS, suggesting that L-ficolin binds to CPS. Sialic acid is known to inhibit alternative complement pathway activation, and, as expected, the bactericidal index (percentage of bacteria killed) for individual strains was inversely proportional to the sialic acid content of the CPS, and L-ficolin-initiated opsonophagocytic killing was significantly increased by addition of CPS-specific IgG2, which increased activation of the alternative pathway. We conclude that binding of L-ficolin/MASP complexes to the CPS generates C3 convertase C4b2a, which deposits C3b on GBS. C3b deposited by this lectin pathway forms alternative pathway C3 convertase C3bBb whose activity is enhanced by CPS-specific IgG2, leading to increased opsonophagocytic killing by further deposition of C3b on the GBS.
Group B streptococci (GBS3; Streptococcus agalactiae) are the most common cause of neonatal sepsis and meningitis ( 1). GBS are subclassified into nine serotypes according to the immunologic reactivity of the capsular polysaccharide (CPS). Serotype III GBS are particularly important, because this serotype causes a significant percentage of early-onset disease (within the first week of life), most late-onset disease (after the first week of life) in neonates, and the majority of neonatal GBS meningitis ( 1).
Deficiency in maternal type III CPS-specific IgG correlates with susceptibility of neonates to the GBS infection ( 2). Sialic acid residues in CPS render GBS resistant to opsonophagocytosis by inhibiting activation of the alternative complement pathway, thereby decreasing bacterial surface deposition of the opsonic C3b and iC3b fragments that are critical for optimal phagocytosis ( 3). CPS-specific IgG overcomes this resistance, and the amount of CPS-specific IgG correlates with the efficiency of opsonophagocytic killing by polymorphonuclear leukocytes (PMNs) ( 4, 5) and with the production of the complement-derived chemoattractant C5a ( 6) in vitro. Serum deficient in type III CPS-specific IgG still mediates significant opsonophagocytic killing ( 5) and C5a production ( 6) in the presence of both Ca2+ and Mg2+ by a process that involves activation of the complement pathway. This Ab-independent activation of complement by serotype III GBS is reduced by chelating Ca2+ with EGTA, suggesting a critical role of the Ca2+-dependent classical pathway C3 convertase C4b2a. The observation that susceptibility to GBS infection is increased in C4-deficient mice further supports a role of the C4b2a in host defense to this bacteria ( 7). Although C1q, the first component of classical pathway, binds directly to GBS, C1q binding is not sufficient for Ab-independent opsonophagocytic killing ( 8).
Another potential mechanism for C4b2a formation is the lectin pathway, which is initiated by binding of a carbohydrate recognition subcomponent to a carbohydrate structure on the activating entity. Three different carbohydrate recognition subcomponents have been described: mannose-binding lectin (MBL), L-ficolin, and H-ficolin ( 9, 10, 11, 12). All recognition subcomponents consist of homotrimers of a single polypeptide chain with an N-terminal collagen-like domain, a neck region, and a C-terminal carbohydrate-binding domain, and are present as higher-order oligomers of the homotrimeric subunits ( 13). In serum, the oligomers form complexes with MBL-associated serine proteases (MASP), namely MASP-1, MASP-2, and MASP-3, to form a lectin pathway activation complex ( 14, 15, 16, 17, 18, 19) in the presence of Ca2+ ( 20). MASP-2 cleaves C4 and C2 to generate the C4b2a complex ( 17, 21, 22, 23). Thus, the C3 convertase C4b2a complexes of the lectin pathway are generated by a mechanism that is both Ca2+ dependent, and Ab and C1 independent. It seems unlikely that MBL, which has low affinity for GBS ( 24, 25), can initiate the lectin pathway on GBS. Although L-ficolin/MASP complexes activate the lectin pathway after binding to lipoteichoic acid extracted from GBS ( 26), L-ficolin has not been shown to activate the lectin pathway on intact GBS. L-Ficolin is also known to bind to N-acetylglucosamine (GlcNAc), one of the four subunit monosaccharides that comprise serotype III GBS CPS ( 27, 28). In the present study, we therefore determined whether L-ficolin/MASP complexes bind to serotype III CPS and mediate opsonophagocytic killing of type III GBS by initiating the lectin pathway of complement activation.
Materials and Methods
Chemicals and media
Common reagents were obtained from Wako Pure Chemical Industries. Human serum albumin (HSA) was from Sigma-Aldrich. Todd-Hewitt broth (THB) was from BBL Microbiology Systems. PBS, HBSS, Medium 199 Hanks’ salts, and HEPES were from Invitrogen Life Technologies.
Abs and complement
Rabbit antiserum against the type III-specific CPS was provided by Dr. F. Gondaira (Denka Seiken). Anti-human L-ficolin mAb 2F5 was prepared as previously described ( 10). mAbs against human C1q, C4d, Bb, and C3d were purchased from Quidel. mAbs against human IgG, IgG1, IgG2, IgG3, and IgG4, and alkaline phosphatase-conjugated goat anti-mouse Ig were from Zymed Laboratories. Alkaline phosphatase conjugation of Ab was performed using EZ-Link Maleimide-Activated Alkaline Phosphatase kit (Pierce). L-Ficolin/MASP complexes were purified from Cohn fraction III from human plasma by using GlcNAc-agarose (Sigma-Aldrich) column and asialofetuin-Sepharose column made asialofetuin (Sigma-Aldrich) conjugated to CNBr-activated Sepharose 4B (Amersham Biosciences) as previously described ( 13, 20), and their concentrations were determined using Lowry assay kit (Sigma-Aldrich). The affinity-purified L-ficolin/MASP complexes prepared in this manner have previously been shown to be composed of L-ficolin/MASP-1, -2, -3, and MBL-associated protein 19 ( 20). Human complements C1 and C4 were from Sigma-Aldrich.
Bacterial strains and capsule-deficient isogenic mutant
GBS isolates used in this study have been characterized ( 6) and are listed in Table I. Deletion of serotype Ia GBS cpsE, a glucosyl-1-phosphate transferase, results in complete loss of capsule expression ( 29). To examine the contribution of the CPS to opsonization by L-ficolin-initiated complement activation, a capsule-deficient (cps−) isogenic mutant strain of the homologous gene in type III GBS strain 874391 was constructed by deleting cpsIIIE. The 5′ and 3′ flanking genomic DNA were amplified by PCR, using primers CpsE5F (5′-AGCAGAAGCTTATAAGATTATCTGTAG-3′), CpsE5R (5′-AACTACAACTGTTTGAATTCTCGCTATA-3′), CpsE3F (5′-AGGTAGAATTCCTCGGGACAGGAG-3′), and CpsE3R (5′-CAATTCTCTTCGAAATAAAGTTGGC-3′). Amplification products were digested with HindIII, ligated, reamplified with primers CpsE5F and CpsE3R, and cloned into pHY304, generously provided by Dr. C. Rubens (University of Washington, Seattle, WA), to generate pHY304/cpsIIIE−. This shuttle vector was transformed into competent strain 874391 and an isogenic mutant with an in-frame deletion of the complete cpsIIIE coding region, CpsE−, was generated as previously described ( 30). Genomic DNA was prepared, and PCR amplification using cpsIIIE specific primers and Southern blotting with a cpsIIIE probe were performed to confirm the absence of the cpsIIIE gene.
|Strain .||Sialic Acid Content (μg/mg of cells, dry weight) .|
|cps− mutant of 874391||0|
|Strain .||Sialic Acid Content (μg/mg of cells, dry weight) .|
|cps− mutant of 874391||0|
Type III-specific CPS
Late-exponential growth phase strain 874391 grown in THB were treated with 0.5% formalin and centrifuged. Supernatant was filtered and concentrated with a PM30 ultrafilter (Millipore). After NaCl concentration of the material was adjusted to 150 mM, cold ethanol was added to a concentration of 50% to precipitate most proteins and nucleic acids. Cold ethanol was again added to the supernatant to a concentration of 80% to precipitate CPS. This crude CPS material was purified by gel filtration with Sephacryl S-300 HR (Amersham Biosciences) and ion exchange chromatography with DEAE-Sepharose CL-6B (Amersham Biosciences) as previously described ( 31). The purified CPS was dialyzed in water and lyophilized. Type III and group B antigenic activities in fractions were assayed by the method of Ochterlony ( 32) with rabbit antiserum against the type III CPS and by Phadebact Streptococcus Test (Boule Diagnostics), respectively. The sialic acid content of the purified CPS was confirmed by the thiobarbituric acid method ( 33).
Naturally occurring type III CPS-specific IgG2
CPS-specific IgG was purified from an i.v. human IgG preparation (Glovenin-I; Takeda) by affinity chromatography. Purified CPS (4 mg) was coupled with 16 ml of swelled epoxy-activated Sepharose 6B (Amersham Biosciences) according to the manufacturer’s instructions. The IgG was loaded into a CPS-Sepharose 6B column equilibrated with 10 mM Tris-HCl buffer (pH 7.4) containing 150 mM NaCl. After washing the column, CPS-specific IgG was eluted with 100 mM glycine-HCl buffer (pH 2.7) containing 1.0 M NaCl, and each fraction was collected into a tube containing 0.1 vol of 1 M Tris-HCl (pH 9.0). Fractions containing CPS-specific IgG were pooled, concentrated with a YM10 ultrafilter (Millipore), and dialyzed three times with a membrane with a molecular-weight cutoff of 25,000 (Spectrum) against 50 mM Tris-HCl buffer (pH 7.4) containing 1.0 M NaCl to remove sialic acid residues released from CPS-Sepharose 6B. Although the IgG preparation was almost entirely composed of IgG2 (>99%) as measured by ELISA, the preparation was successively passed through each of three affinity columns made monoclonal anti-human IgG1, IgG3, and IgG4 conjugated to HiTrap NHS-activated HP (Amersham Biosciences). The concentration of the IgG2 was determined by absorbance at 280 nm.
Serum from a normal healthy volunteer, with a low level of IgG specific for type III CPS, was chelated with 2.5 mM EDTA. A portion of the serum was selectively depleted of C1q, a highly basic protein ( 34), by weak cation exchange chromatography on a column of CM FF (Amersham Biosciences). Serum was applied to the column equilibrated with 50 mM phosphate buffer containing 82 mM NaCl and 2 mM EDTA (pH 7.4), and eluted with the same buffer. The volume of C1q-depleted serum thus obtained was restored to the starting volume by concentrating with a YM10 ultrafilter. C1q-depleted serum or EDTA-chelated serum (each 10 ml) was absorbed for >6 h at 4°C with formalin-fixed bacterial pellets of the four GBS strains to be examined (each harvested from 200 ml of 0.6-OD600 suspensions) to remove Abs directed against all Ags present on the bacterial surface. The absorption was performed three successive times, and the bacteria were removed by centrifugation following each absorption. Formalin-fixed bacterial pellets were prepared as follows: bacteria were grown at 37°C in THB to late-exponential growth phase, washed with PBS, fixed for 2 h at 37°C in PBS containing 5% formalin, then washed five times with PBS containing 1 mM EDTA, and pelleted. The C1q/Ab-depleted serum or Ab-absorbed serum thus obtained was dialyzed five times with a membrane with a molecular-weight cutoff of 3500 (Spectrum) against HBSS without Mg2+ and Ca2+ to remove Mg2+- and Ca2+-EDTA. The volume of each of C1q/Ab-depleted serum or Ab-absorbed serum thus obtained was restored to 10 ml by concentrating on a YM10 ultrafilter, filter-sterilized, and stored at −80°C. Depletion of C1q and CPS-specific IgG were confirmed by ELISA.
Blood taken from healthy donors was heparinized, and PMNs were isolated by the gradient-centrifugation method with Polymorphprep (Nycomed Pharm) according to the manufacturer’s instructions, washed twice, and resuspended at a density of 4.6 × 106/ml in Medium 199 containing 0.4% HSA and 10 mM HEPES. Dye exclusion indicated that the viability of the PMNs was >95%.
Opsonization of bacteria
Bacteria were cultured overnight in THB, inoculated at a 1/20 dilution into fresh THB, and incubated at 37°C for 90 min. Bacteria were washed twice and resuspended in HBSS without Mg2+ and Ca2+ to an OD600 of 0.60. Each 1 ml of a 0.60-OD600 suspension contains 168 μg of cell dry weight and between 0.4 × 108 and 1.8 × 108 CFU ( 35). Bacteria (0.12 OD600) in HBSS containing 0.4% HSA and 10 mM HEPES were successively incubated with the following components or serum in a 200-μl mixture (in silicone-coated glass tubes): L-ficolin/MASP complexes in the presence of 2 mM CaCl2 (30 min); 10% C1q/Ab-depleted serum in the presence of 2 mM MgCl2 and 2 mM CaCl2 (1 min); and 10% C1q/Ab-depleted serum or 10% Ab-absorbed serum in the presence of 4 mM MgCl2/16 mM EGTA (4 min). In some experiments, as indicated in Results, bacteria were incubated with human C1 and CPS-specific IgG2 in the presence of 2 mM CaCl2 (30 min). Opsonized bacteria were washed three times and resuspended at 0.06 OD600 in HBSS with Mg2+ and Ca2+. The opsonized bacterial suspensions (100%) thus obtained were diluted with HBSS with Mg2+ and Ca2+ to prepare 56.2, 31.6, 17.8, and 10.0% suspensions, and all five were used as standards in the opsonophagocytic killing assay.
Opsonophagocytic killing assay
Opsonophagocytic killing of bacteria was determined by using a modification of the previously developed quantitative assay ( 35). Briefly, 350 μl of a suspension of 4.6 × 106/ml PMNs in Medium 199 containing 0.4% HSA and 10 mM HEPES was added to 50 μl of 100% opsonized GBS, and incubated at 37°C for 60 min with shaking, and then immediately cooled in an ice water bath. The five standards of opsonized GBS suspensions (100, 56.2, 31.6, 17.8, and 10.0%) were treated in the same manner, except that Medium 199 containing 0.4% HSA, 10 mM HEPES, and RBCs of a concentration adjusted to that of PMN preparations was substituted for PMN suspensions. To determine the percent bacterial survival (extra-PMN bacteria) after opsonophagocytic killing, 3.6 ml of THB was added to the mixture and thoroughly vortexed. The PMNs and RBCs were pelleted by centrifugation at 200 × g at 4°C for 5 min, and 2 ml of supernatant was transferred to a cuvette and incubated at 37°C. During the incubation process, growth was monitored by recording OD values of the cuvettes with the spectrophotometer at 640 nm, and all suspensions were then immediately cooled in an ice water bath when the OD value of the 100% standard reached 0.2. The equation for the regression line between the OD value and the log of the percent concentrations of five standard suspensions was obtained by the least-squares method. The log of the percent bacterial survival after opsonophagocytic killing was estimated from the regression line, and results were expressed as bactericidal index (percentage of bacteria killed). The percentage of bacteria surviving in the extra-PMN fluid was approximately the same as the percentage of bacteria surviving in both intra-PMN and extra-PMN locations ( 35).
Measurement of complement component binding to opsonized bacteria
The opsonized bacteria (80 μl of 0.0015 OD600) described above were coated to wells of the ELISA plate (Asahi Techno Glass) overnight at 4°C. Wells were blocked with 300 μl of 0.1% HSA in TBS (10 mM Tris-HCl, 150 mM NaCl (pH 7.4)) containing 5 mM CaCl2 for 1 h at 37°C, and then washed with TBS containing 0.05% Tween 20 and 5 mM CaCl2 (wash buffer). After washing, complement components bound to the bacteria were detected using mAbs specifically directed against complement component fragments, alkaline phosphatase-conjugated goat anti-mouse Ig as the secondary Ab, and the ELISA Amplification System (Invitrogen Life Technologies) according to the manufacturer’s instructions. For negative controls, the primary Ab was omitted. L-Ficolin binding was determined in the same manner except that the bacteria were not incubated with the C1q/Ab-depleted serum before the ELISA determination.
C4 cleavage assay
L-Ficolin-initiated lectin pathway activation was quantified using a modification of the previously described C4 cleavage assay ( 36). Briefly, the wells of the ELISA plate were coated with 2.56 μg of CPS in 80 μl of PBS. After overnight incubation at 37°C, wells were blocked with 0.1% HSA in TBS, and then washed with TBS containing 0.05% Tween 20 and 5 mM CaCl2 (wash buffer). L-Ficolin/MASP complexes (80 ng) in 80 μl of 20 mM Tris-HCl, 1 M NaCl, 10 mM CaCl2, 0.05% Triton X-100, and 0.1% HSA (pH 7.4) were added to the wells and incubated overnight at 4°C. The wells were washed thoroughly with wash buffer, and then 80 ng of human C4 in 80 μl of veronal buffer containing 2 mM CaCl2 and 1 mM MgCl2 was added to each well. After 90-min incubation at 37°C, wells were thoroughly washed again, and L-ficolin and C4d binding were detected using mAbs specifically directed against L-ficolin and C4d, alkaline phosphatase-conjugated goat anti-mouse-Ig Ab, and the ELISA Amplification System. For negative controls, the primary Ab was omitted.
Determination of conditions for Ab- and classical pathway-independent opsonophagocytic killing of type III GBS
Serum depleted of C1q and of Abs specific to GBS was used to define the role of L-ficolin/MASP complexes in Ab-independent opsonophagocytic killing of serotype III GBS. C1q was selectively removed from human serum by weak cation exchange chromatography, and then Abs specific to GBS were selectively removed by absorption with GBS. C1q remaining in the serum was 3.7 ± 0.12% (mean ± SD) of that in the untreated serum, and CPS-specific IgG was undetectable (<0.2 μg/ml), as measured by ELISA. The bactericidal index (percentage of bacteria killed) for serotype III strain 874391 after incubation in C1q/Ab-depleted serum was <5% when the bacteria were incubated with the C1q/Ab-depleted serum in the presence of Ca2+ and Mg2+ for 1 min to permit formation of a C4b2a complex (Fig. 1,A). The bactericidal index remained quite low when the bacteria were subsequently incubated for 4 min in the presence of Mg2+/EGTA to allow alternative pathway activation, but rose significantly after 8 min (Fig. 1 A).
It is known that complement factor D is absorbed to the cation exchange resin ( 34). To confirm that the C1q/Ab-depleted serum contains factor D and other complement components to support L-ficolin-initiated opsonophagocytic killing, we tested whether opsonic activity of the serum could be increased by adding human C1 and CPS-specific IgG2. As shown in Fig. 1, the bactericidal index of C1q/Ab-depleted serum (A) and the binding of factor Bb and C3d to the bacteria (B) were significantly increased by adding human C1 and CPS-specific IgG2. These data indicate that the C1q/Ab-depleted serum removes classical pathway contribution to opsonization under these conditions, and that these conditions allow activation of the alternative pathway during the subsequent 4-min incubation in the presence of Mg2+/EGTA. These experimental conditions were therefore used to examine the contribution of the lectin pathway to initiating opsonization.
Binding of L-ficolin/MASP complexes to type III GBS and activation of the lectin pathway
Using the conditions described above, killing by PMNs was observed when strain 874391 was preincubated with L-ficolin/MASP complexes and successively incubated with C1q/Ab-depleted serum (Fig. 2,A). The opsonophagocytic killing increased with increasing concentration of L-ficolin/MASP complexes and was completely inhibited when the ligand for L-ficolin, GlcNAc, was incubated together with the GBS and L-ficolin/MASP complexes (Fig. 2,A). Moreover, concentration-dependent binding of L-ficolin and complement components C3d, Bb, and C4d to strain 874391 was observed (Fig. 2 B). These observations indicate that L-ficolin binds to serotype III GBS, and that MASP activates complement to increase opsonophagocytic killing by PMNs.
Binding of the L-ficolin/MASP complexes to type III CPS and activation of the lectin pathway
Binding of L-ficolin/MASP complexes to four GBS strains with varying amounts of capsule and to a cps− isogenic mutant of strain 874391 that completely lacks capsule demonstrated that binding of L-ficolin/MASP to GBS correlates with capsular content (Fig. 3,B). L-Ficolin also bound to the cps− isogenic mutant (Fig. 3,B), indicating that there is another ligand or ligands for L-ficolin on the GBS surface in addition to capsule. However, the bactericidal index for each strain was inversely proportional to the sialic acid content of the CPS (Fig. 3,A), presumably because the alternative pathway is critical for opsonophagocytosis and is inhibited to a degree that is directly proportional to sialic acid content. Of interest, the bactericidal index was increased by preincubation of the strains with L-ficolin/MASP complexes in a manner that was roughly proportional to CPS content (as indexed by sialic acid content) and, therefore, the amount of L-ficolin/MASP binding of each strain. These data indicate that L-ficolin/MASP complexes can bind to CPS and activate complement. These conclusions were strengthened by additional experiments demonstrating that opsonophagocytic killing (Fig. 4,A), and L-ficolin binding and C4 activation (B), were inhibited by preincubating excess fluid-phase CPS together with the bacteria and the L-ficolin/MASP complexes. Finally, L-ficolin binding and C4 activation on CPS-coated wells were also observed, and both were inhibited by preincubating the L-ficolin/MASP complexes with fluid-phase CPS (Fig. 5).
Effect of CPS-specific IgG2 on L-ficolin-initiated opsonophagocytic killing
Previous studies demonstrated that CPS-specific IgG overcomes the inhibition of the alternative pathway of complement exerted by sialic acid in type III CPS ( 3, 4, 5, 6). Therefore, we tested whether CPS-specific IgG2 increases L-ficolin-initiated opsonophagocytic killing of highly encapsulated serotype III GBS. L-Ficolin-initiated opsonophagocytic killing of strain 874391 was significantly increased when bacteria were preincubated with CPS-specific IgG2 and L-ficolin/MASP complexes before adding C1q/Ab-depleted serum containing Ca2+ and Mg2+, followed by Ab-depleted serum containing Mg2+/EGTA (Fig. 6,A, F,IgG/CaMg/Mg). A slightly larger increase in L-ficolin-initiated opsonophagocytic killing was observed when bacteria were incubated with CPS-specific IgG2 after L-ficolin/MASP binding and incubation with C1q/Ab-depleted serum containing Ca2+ and Mg2+, but before addition of Ab-depleted serum containing Mg2+/EGTA (F/CaMg/IgG/Mg). No increase in opsonophagocytic killing was observed when bacteria were incubated with the CPS-specific IgG2 after complement had been activated (F/CaMg/Mg/IgG), suggesting that the increased killing of the bacteria in the presence of IgG2 is not due to FcR-mediated binding of the GBS to the PMNs. The negative control, in which bacteria were not preincubated with L-ficolin/MASP complexes, exhibited <10% bactericidal index (IgG/CaMg/Mg). These data suggest that CPS-specific IgG2 binds and neutralizes the sialic acid residues, thereby allowing the alternative pathway C3 convertase C3bBb to amplify the C3b that had been deposited on the bacteria by activation of the L-ficolin/MASP complexes. Interestingly, CPS-specific IgG2 was more effective under these experimental conditions at enhancing L-ficolin-initiated opsonophagocytic killing than in Ab-initiated classical pathway activation-induced killing (Fig. 6 B). The enhancing effect of CPS-specific IgG2 on L-ficolin-initiated opsonophagocytic killing was also observed using the more heavily encapsulated strain i05 (data not shown).
Opsonization of GBS with the complement fragments C3b and iC3b promotes efficient phagocytosis and killing of the bacteria. Although CPS-specific IgG appears to increase C3b deposition and opsonophagocytic killing by activation of the alternative pathway, serum deficient in type III CPS-specific IgG still mediates significant opsonophagocytic killing ( 5) in the presence of both Ca2+ and Mg2+, suggesting a critical role of C3 convertase C4b2a of the classical and/or lectin pathways of complement. The L-ficolin-mediated lectin pathway is a potential mechanism for initiating opsonization of serotype III GBS, because L-ficolin has been shown to bind to GBS lipoteichoic acid ( 26), and because the ligand for L-ficolin, GlcNAc, is one of four subunit carbohydrates comprising the serotype III GBS CPS ( 27, 28).
As shown in this study, L-ficolin binds both to purified, immobilized CPS and to intact GBS, and the binding to GBS is directly proportional to the CPS content of the GBS, when sialic acid content of the bacteria is used as the index of the amount of CPS on the bacterial surface. L-Ficolin also binds to the cps− isogenic mutant, indicating that there is another ligand or ligands on the bacterial surface in addition to capsule. Activation of L-ficolin/MASP complexes generates a C3 convertase C4b2a that produces C3b fragments, which, in turn, combine with factor B to generate the alternative pathway C3 convertase C3bBb and amplifies C3b deposition on the bacteria. Indeed, it has been demonstrated that the lectin pathway may contribute substantially to the formation of the alternative pathway C3 convertase ( 37, 38). We used serum depleted of C1q and Ab to study the contribution of L-ficolin/MASP complexes to opsonophagocytic killing without classical pathway involvement, and then permitted subsequent alternative pathway activation in the presence of Mg2+/EGTA. Under these conditions, opsonophagocytic killing of GBS strain 874391 by PMNs was proportional to the concentration of L-ficolin/MASP complexes in the reaction mixture, and significant killing was observed at concentrations of L-ficolin similar to that found in the cord blood of human neonates ( 39). We propose that MASP-2 is responsible for the complement activation and C3b deposition necessary for the opsonophagocytic killing initiated by L-ficolin/MASP complexes, because MASP-2 is a component of the L-ficolin/MASP complexes ( 20) and has previously been shown to activate complement by cleaving C4 and C2 ( 17, 21, 22, 23). Experiments with further purified or recombinant proteins will be necessary to define the more precise mechanisms involved.
L-Ficolin binding to different strains of GBS was directly proportional to the sialic acid content of each strain, whereas opsonophagocytic killing was inversely proportional to sialic acid content. This observation is consistent with the previous one that encapsulated strains of type III GBS do not activate the alternative pathway in the absence of CPS-specific IgG ( 5, 6). Failure to activate the alternative pathway appears to reflect the inhibitory effect of sialic acid, which is present as a terminal side-chain residue of the type III CPS ( 27, 28). Cell surfaces having sialic acid increase the affinity of factor H for bound C3b, resulting in cleavage of C3b to iC3b by factor I and interruption of the C3b amplification loop ( 40, 41). This mechanism has been demonstrated on type III GBS by experiments in which removal of sialic acid from the capsule by neuraminidase treatment ( 3) or by transposon-insertional mutagenesis ( 42, 43) changed the bacteria from a nonactivating to an activating surface for the alternative pathway of complement. The results of the experiments reported in this paper suggest that L-ficolin/MASP complexes binding to GBS leads to C3b deposition on the bacteria, but that sialic acid residues inhibit activation of the C3b amplification loop and optimal opsonophagocytic killing. Consistent with this hypothesis, addition of CPS-specific IgG2 before alternative pathway activation resulted in increased L-ficolin-initiated opsonophagocytic killing. The inability of CPS-specific IgG2 to enhance L-ficolin-initiated opsonophagocytic killing when CPS-specific IgG2 were added after complement activation is consistent with Ab causing increased phagocytic killing through enhanced complement activation and not by interacting with PMN FcRs ( 5). Taken together, these observations are consistent with a model in which L-ficolin/MASP complexes initiate C3b deposition, and CPS-specific IgG2 increases C3b deposition by overcoming sialic acid-mediated inhibition of the alternative pathway C3 convertase.
Although maternal CPS-specific Ab protects colonized neonates from invasive GBS disease, most infants that are heavily colonized with GBS at birth do not develop invasive disease, and many of these uninfected infants lack protective levels of CPS-specific Ab. The reason why colonized infants lacking CPS-specific Ab do not get infected is not clear, but may be due to maternal Abs against GBS protein Ags ( 44), or to intrinsic differences in virulence between GBS clones, as we and others have proposed ( 30, 45, 46, 47, 48). These studies have focused only on serotype III GBS, which are associated with a large percentage of invasive neonatal disease, but the presence of GlcNac in the CPS of all serotype of GBS ( 27, 31, 49, 50, 51, 52) excepting VI (53) and VIII ( 54) raises the possibility that the L-ficolin/MASP complexes can initiate C3b deposition on these serotypes. The results of these studies suggest that the L-ficolin-mediated lectin pathway of complement activation is critical in preventing neonatal infections with GBS in both the presence and absence of CPS-specific Ab.
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.
This study was supported by The Promotion Corporation for Private School of Japan and by the Thrasher Research Fund.
Abbreviations used in this paper: GBS, group B streptococcus; CPS, capsular polysaccharide; PMN, polymorphonuclear leukocyte; MBL, mannose-binding lectin; MASP, MBL-associated serine protease; GlcNAc, N-acetylglucosamine; HSA, human serum albumin; THB, Todd-Hewitt broth.