Role of the complement-lectin pathway in anaphylactoid reaction induced with lipopolysaccharide in mice

Authors


Abstract

We show that Proteus vulgaris O25 (PO25) lipopolysaccharide (LPS) induced an anaphylactoid reaction not only in wild-type and in lipid A non-responding mice but also in recombinase-activating gene-2-deficient (RAG-2–/–) and in mast cell-deficient (W/Wv) animals. Western blot analysis indicated that PO25 LPS bound to Ra-reactive factor (RaRF), the complex of mannan-binding lectins (MBL) and MBL-associated serine proteases. Binding of RaRF to PO25 LPS led to the activation of C4 component without participation of either C1 or Ig, via the lectin pathway. Relative concentration of RaRF and hemolytic activity in mouse serum decreased rapidly during the process of anaphylactoid reaction. A significant drop of MBL-A, but not MBL-C level was observed. Administrationwith antiserum to RaRF prevented animals from death as a consequence of the inhibition of interaction of RaRF with the carbohydrate target and complement activation. These results indicate that complement-lectin pathway activation is responsible for the anaphylactoid reaction induced with LPS in muramyldipeptide-primed mice. RaRF also activated fibrinogen in vitro suggesting the involvement of the coagulation system in the process investigated.

Abbreviations:
KO3:

Klebsiella pneumoniae O3

MDP:

Muramyldipeptide

PO25:

Proteus vulgaris O25

OPS:

O-specific polysaccharide

GlcNAc:

N-Acetyl-D-glucosamine

Hep:

L-Glycero-D-mannoheptose

RaRF:

Ra-reactive factor

MBL:

Mannan-binding lectin

MASP:

MBL-associated serine protease

C4D:

C4-deficient guinea pig serum

TBS-Ca2+:

Tris-buffered saline containing 1 mM CaCl2

GVB:

Gelatin-supplemented veronal buffer

GVB2+:

Gelatin- and divalent cation-supplemented veronal buffer

MGVB2+:

Mannitol-, divalent cation- and gelatin-supplemented veronal buffer

PO25RP:

PO25 LPS-reacting protein

LP:

Lectin pathway

Rag:

Recombinase-activating gene

1 Introduction

Ra-reactive factor (RaRF) is a complex of mannan-binding lectins (MBL-A, MBL-C) and MBL-associated serine proteases (MASP) 16. The lectin component of RaRF, via carbohydrate recognition domain, binds to surface polysaccharides, lipopolysaccharides (LPS) and glycoproteins of numerous bacteria, fungi, protozoa and viruses. MBL ligands, such as mannose, N-acetyl-D-glucosamine (GlcNAc) and fucose are wide-spread in the microbial world 7, 8. One of MBL target structures is Klebsiella pneumoniae O3(KO3) LPS, whose O-polysaccharide region consists of mannose homopolymer. This LPS is also an inducer of the anaphylactoid reaction in muramyldipeptide (MDP)-sensitized mice 9, 10. This phenomenon, thought to be connected with polysaccharide LPS region, leads to the characteristic symptoms (convulsions, spreading and unconsciousness) followed by death within 15–60 min 11. Unlikely to lipid A-dependent endotoxin shock, anaphylactoid reaction can be evoked in LPS-hyporesponsive mice, like C3H/HeJ strain, with defect of the TLR4 receptor12.

Previously we found that LPS of Proteus vulgaris O25 (PO25) shows biological activities similar to that of KO3 13, although it does not contain any mannose residue in O-specific polysaccharide (OPS) (Fig. 1A) or in core region (Fig. 1B) 14, 15. At the nonreducing end of the core, however, GlcNAc and L-glycero-D-mannoheptose (Hep) residues are present 15. Binding of RaRF to bacterial cell envelope components may initiate complement activation via the lectin pathway (LP) 16. We present data indicating that LP activation is involved in the development of LPS-induced anaphylactoid reaction in MDP-sensitized mice.

Figure 1.

Structure of polysaccharide region of PO25 LPS. Repeating unit of OPS (A) and core region (B) are shown. α-Hep: L-glycero-α-D-mannoheptose; α-DDHep: D-glycero-α-D-mannoheptose. Terminal residues of GlcNAc and Hep are indicated with bold italics.

2 Results

2.1 Induction of the anaphylactoid reaction with LPS

As reported previously 10, 13, i.v. administration of PO25 or KO3 LPS leads to the occurrence of anaphylactoid reaction and death of MDP-pretreated mice. In contrast, Escherichia coli O127 LPS, an effective inducer of lipid A-dependent endotoxin shock, evoked relatively slight symptoms of illness (without convulsions, prostration or crawling and lethal effect) in the same experimental conditions. Interestingly, PO25 LPS was active not only in the wild-type (BALB/c) and lipid A non-responding (C3H/HeJ) mice, but also in recombinase-activating gene-2-deficient (RAG-2–/–), and mast cell-deficient (WBB6/F1-W/Wv) animals (Table 1). This demonstrated neither Ig nor mast cells to be necessary for the development of anaphylactoid reaction.

Table 1. Induction of anaphylactoid reaction in different strains of micea)
 LPS
Mouse strainAnaphylactoid reactionP. vulgaris O25K. pneumoniae O3E. coli O127
  1. a) Mice were sensitized with 100 μg of MDP (i.p.) and 4 h later injected with 100 μg of test LPS (i.v.). Incidence, score and mortality were recorded within 1 h after LPS injection.

  2. b) Incidence and mortality are shown as number/total.

  3. c) The scoring of anaphylactoid reaction 10 was as follows: 1, staggering; 2, crawling and prostration; 3, prostration and weak convulsions; 4, prostration and strong convulsions. n.t., not tested.

C3H/HeJIncidenceb)6/66/66/6
 Scorec)441
 Mortalityb)6/66/60/6
BALB/cIncidence6/66/66/6
 Score441
 Mortality6/66/60/6
RAG-2–/–Incidence10/10n.t.n.t.
 Score3 to 4  
 Mortality5/10  
WBB6F1-+/+Incidence4/4n.t.n.t.
 Score4  
 Mortality4/4  
WBB6F1-Incidence4/4n.t.n.t.
W/WvScore4  
 Mortality4/4  

2.2 Binding of PO25 LPS to RaRF

PO25 LPS-reacting protein (PO25RP) fraction was separated from mouse serum by absorption with PO25 bacteria and elution with GlcNAc solution. Its SDS-PAGE/Western blot analysis showed the presence of an approximately 30-kDa protein band corresponding to monomers of MBL (Fig. 2a). The presence of MASP-1 and MASP-2 in PO25RP was detected with the help of specific Ab (Fig. 2b, c). ELISA employing mAb against mouse MBL-A and MBL-C and rabbit Ab against mouse Ig demonstrated both lectin forms (with higher content of MBL-A) but no Ig to be present in the PO25RP fraction (Fig. 3). MBL in PO25RP was further purified by LPS-epoxy-Sepharose column chromatography, which resulted in a single 30-kDa band after immunostaining with anti-PO25RP serum (data not shown). The N-terminal amino acid sequence of this product was SGSQTCEDTLKTCSVIA, which showed 100% agreement with the N-terminal sequence of MBL-A 17. These results indicated that PO25 LPS binds mouse RaRF, i.e. MBL-MASP complex.

Interaction of PO25RP with PO25 LPS as well as with mannan was prevented by GlcNAc, mannose, Hep, fucose, mannoheptulose and sedoheptulosan (Table 2). The short chain-rich fraction of PO25 LPS showed identical inhibitory activity, while the long chain-rich fraction showed slighter inhibitory activity, in comparison to native LPS. Since in the nonreducing end of PO25 LPS core oligosaccharide GlcNAc and Hep residues are present (Fig. 115) we suggest that outer core rather than OPS region binds MBL. The inhibition of PO25RP-mannan interaction (Table 2) and PO25RP-induced hemolysis of PO25 LPS-SRBC (data not shown) by mannose indicated that ficolins did not play a crucial role in the process investigated.

Table 2. Inhibition of PO25RP binding to PO25 LPS or mannan by monosaccharides and LPSa)
Inhibitor Inhibition
 PO25 LPSMannan
  1. a) After incubation of 35 μl of appropriate dilution of mouse serum with the same volume of inhibitor (in triplicate) for 30 min at 37°C, the mixture was transferred to plates coatedwith PO25 LPS (500 ng/well) or mannan (15 μg/well). ELISA was performed using rabbit anti-PO25RP serum as primary Ab.

GlcNAc50 mM9376
Man50 mM8594
Fuc50 mM8962
Hep50 mM9396
Sedoheptulosan50 mM8866
Mannitol50 mM18 0
Gal50 mM6127
PO25 LPS70 μg/ml8874
Long chain-rich fraction of PO25 LPS70 μg/ml62 5
Short chain-richt fraction of PO25 LPS70 μg/ml9770
Figure 2.

 Electrophoretic analysis of PO25RP. PO25 bacteria were treated with normal mouse serum. The bound proteins were eluted with TBS-Ca2+ containing 10% GlcNAc and fractionated by SDS-PAGE under reducing conditions. Immunostaining with anti-PO25RP serum (a), anti-MASP-1 Ig (b) and anti-MASP-2 serum (c).

Figure 3.

 ELISA analysis of PO25RP. PO25RP was serially diluted in PO25 LPS-coated plates. The bound protein was detected with mAb against MBL-A, MBL-C, rabbit anti-PO25RP serum or rabbit anti-mouse Ig.

2.3 Activation of the complement-lectin pathway

LP activity was determined by serial incubation of PO25 LPS-SRBC with fresh mouse serum, C4 and C4-deficient guinea pig serum (C4D). As shown in Fig. 4, sera of BALB/c mice lysed LPS-coated erythrocytes. The hemolytic activity of sera of Ig-deficient RAG-1–/– and RAG-2–/– mice enabled to exclude the participation of natural, PO25 LPS-reacting Ab in the observed reaction. Similarly, hemolysis occurred only in the presence of C4 when PO25 LPS-SRBC were incubated with various amounts of PO25RP (Fig. 5A). The degree of lysis depended also on the C4 amount (data not shown). Pretreatment of PO25RP with homologous and anti-MASP-2, but not with anti-MASP-1 Ab resulted in suppression of its C4-consuming ability (Fig. 5B). Anti-PO25RP inhibited the binding of MBL-A and MBL-C to mannan (Fig. 5C). Additionally, the C4b deposition ability of PO25RP after binding to PO25 LPS or mannan, in conditions excluding participation of C1, 18 was demonstrated (Fig. 5D). These results indicate that PO25RP activates LP in vitro.

I.v. injection of anti-PO25RP serum 60 min before PO25 LPS led to prevention of the anaphylactoid reaction: weaker symptoms of illness and no death were observed (Table 3). Possible participation of LPS-reacting Ab was excluded by absorption of antiserum with heat-killed bacteria. It should be stressed that this serum protected KO3 LPS-treated mice (Table 3). Rabbit anti-MASP-1 as well as goat anti-C3 Ab also revealed protective effects in the MDP/PO25 LPS-treated animals. A similar result was seen when rabbit serum against heat-stable surface antigens of PO25 bacteria was employed. In contrast, heat-inactivated (56°C, 30 min) pre-immune rabbit serum did not protect animals against lethal effects of LPS.

As shown in Fig. 6A, total loss of LP-activating potency was observed in sera collected from PO25 and KO3 LPS-treated mice at the moment of manifestation of first symptoms of anaphylactoid reaction. The concentration of PO25RP also decreased significantly (p<0.05; Fig. 6B). In contrast, drop of the serum C4-activating potency and decline of the PO25RP level were much weaker in sera of mice that received E. coli O127 LPS. In mice which underwent the anaphylactoid reaction, the consumption of PO25RP was reflected by rapid drop of MBL-A levels detected in ELISA (Fig. 7A), while the decrease of MBL-C concentrations was much slighter (Fig. 7B). It suggests mainly MBL-A to be involved in PO25 LPS-induced complement activation and development of shock. The treatment of animals with anti-PO25RP serum caused MBL-A neutralization, which resulted in its low concentration detected in ELISA (Fig. 7A), while the level of MBL-C was not influenced significantly (Fig. 7B). Additionally, anti-PO25RP serum prevented activation of LP in PO25 LPS-treated mice (Fig. 7C). This indicates that the protective action of anti-PO25RP serum is connected with blocking of MBL-A carbohydrate target-binding/complement-activating potency. Not-consumed complement factors may contribute to a higher clearance rate of LPS from circulation. On the basis of presented data we conclude that the LP plays a crucial role in the development of LPS-induced anaphylactoid reaction.

Figure 4.

 C4-dependent lysis of PO25 LPS-SRBC sensitized with sera of various mouse strains. PO25 LPS-SRBC were serially treated with pooled sera of several individuals of each strain and with C4. The cells were washed with GVB2+ after each treatment. After incubation with C4D, hemolysis rate was determined.

Figure 5.

 C4 activation by PO25RP and specificity of anti-PO25RP serum. (A) C4-dependent lysis of PO25 LPS-SRBC sensitized with PO25RP. PO25 LPS-SRBC were treated with various amounts of PO25RP and with or without C4. After incubation with C4D, hemolysis rate was determined. (B) Inhibition of C4-dependent lysis of PO25 LPS-SRBC. PO25 LPS-SRBC were incubated with PO25RP, which had been pre-incubated with various dilutions of anti-PO25RP, anti-MASP-2 serum or anti-MASP-1 Ig. After washing, cells were incubated with C4 and precipitated by centrifugation followed by incubation with C4D. (C) Inhibition of MBL-A and MBL-C binding to mannan. PO25RP was pre-incubated with homologous antiserum and then transferred to mannan-coated plates. The amount of MBL-A and MBL-C bound to mannan was determined in ELISA with the help of specific Ab. (D) C4b deposition activity of PO25RP-mannan or PO25RP-PO25 LPS complexes. PO25RP was incubated with mannan-(10 μg/ml) or PO25 LPS-(50 μg/ml) coated ELISA plates. After addition of C4 (5 μg/ml), the deposited C4b was detected with the help of rabbit anti-C4 serum.

Figure 6.

 C4-activating potency and PO25RP concentration in sera of LPS-treated mice. BALB/c mice (five individuals per group) were sensitized with MDP followed by LPS injection and sacrificed at the moment of manifestation of first symptoms of anaphylactoid reaction (mice treated with PO25 and KO3 LPS) or 30 min after the last injection (when O127 LPS or NaCl were used). (A) C4-activating potency of serum was evaluated by measuring hemolysis of PO25 LPS-SRBC after serial treatments of the cells with mouse serum, C4 and C4D. (B) PO25RP concentration was measured in ELISA using mannan-coated plates and rabbit anti-PO25RP serum (mouse sera diluted 1:800). The figure represents one experiment of two with similar results; *p<0.05 vs. control group.

Figure 7.

 MBL-A (A) and MBL-C (B) levels and LP activation (C) in sera of mice treated with rabbit anti-PO25RP serum and/or LPS. BALB/c mice (five individuals per group) sensitized with MDP were injected with PO25 LPS and sacrificed at the moment described in Sect. 4. Mice that received anti-PO25RP serum only, animals treated with antiserum and 1 h later with LPS, as well as controls (injected with NaCl) were sacrificed 1 h after the last injection. MBL-A (sera diluted 1:800) and MBL-C (sera diluted 1:3,200) levels were measured in ELISA using mannan-coated plates with the help of specific mAb. MBL-deficient human serum was employed as C4 source (mouse sera diluted 1:2,000). The figure represents one experiment of two with similar results; *p<0.05 vs. control group.

Table 3. Protection of LPS-treated mice with Ab of various specificitiesa)
LPSSerum or Ab Number of mice
 TotalAnaphylactoid reaction
 IncidenceScoreb)Survivors
  1. a) MDP-treated C3H/HeJ mice were injected with test serum or Ab 1 h before LPS administration. Death was recorded up to 72 h.

  2. b) Numbers indicate the degree of symptoms 10 as follows: 1, staggering; 2, crawling and prostration; 3, prostration and weak convulsions; 4, prostration and strong convulsions.

P. vulgaris O25 664 0
K. pneumoniae O3 444 0
P. vulgaris O25Non-immune 664 0
P. vulgaris O25Anti-PO25RP1060 – 210
K. pneumoniae O3Anti-PO25RP 442 4
P. vulgaris O25Anti-MASP-1 Ab 443 – 4 3
P. vulgaris O25Anti-C3 Ab 442 – 3 4
 Anti-PO25    
P. vulgaris O25bacteria 662 5

2.4 Activation of fibrinogen in vitro

Fibrinogen activation by PO25RP was tested according to the procedure described by Hajela et al. 19. Incubation of this substrate with PO25RP resulted in its cleavage in a dose-dependent manner (Fig. 8). PO25RP (25 μg) caused similar effect to thrombin, a natural activator of fibrinogen, used at dose of 0.1 μg. These data indicate the role of PO25RP not only in the complement-LP, but also in coagulation system activation.

Figure 8.

 Fibrinogen cleavage by PO25RP. Lane a, fibrinogen (60 μg); lane b, fibrinogen treated with PO25RP (25 μg); lane c, fibrinogen treated with PO25RP (2.5 μg); lane d, PO25RP (25 μg); lane e, fibrinogen treated with thrombin (0.1 μg). Products were separated in SDS-PAGE (reducing conditions) and stained with Coomassie Brilliant Blue.

3 Discussion

The binding of MBL-MASP complex to microbial surface structures rich in mannose, Hep, GlcNAc, and fucose (among them LPS), leads to the activation of C4 and C3 components via the LP and finally to killing of the bacteria 1, 16, 20, 21. Certain LPS, such as those of PO25 and KO3, induce severe anaphylactoid symptoms in mice 10, 13. This reaction has been hypothesized to be connected with LP activation 22. Shibazaki et al. demonstrated that administration of KO3 LPS causes rapid accumulation and degradation of platelets in liver and lung. Some factors released, including serotonin, were supposed to be responsible for shock induction 10. Luskin and Luskin 23 reported that mastocyte mediators might contribute to the IgE-independent anaphylaxis. Contrarily, we excluded the significant role of these cells in LPS-evoked anaphylactoid reaction, since W/Wv mice showed the characteristic symptoms and died within 15 min after PO25 LPS injection (Table 1). The anaphylactoid reaction is also Ig-independent because it can be induced in RAG-2–/– mice, lacking ability to produce Ig (Table 1). However, their sensitivity to the LPS was lower than in those of wild type. This may reflect the enhancement effect of Ig on LP activity due to H factor neutralization as was reported by Suankratay et al. 24. In contrast, Ig-deficient animals are more sensitive to the lipid A-dependent endotoxin shock, due to impaired LPS clearance, than the wild-type mice 25.

Recently it was demonstrated that LPS being inducers of the anaphylactoid shock bind MBL in vitro22. Our data are in agreement with this observation. On the basis of the analysis of SDS-PAGE/Western blot patterns, the sequence of N-terminal protein fragment, and ELISA employing mAb, we found that the PO25RP fraction contains MBL and MASP while the presence of Ig was excluded (Fig. 2 and 3). Moreover, it activates C4 without participation of C1 (Fig. 5D). The lysis of PO25 LPS-SRBC occurs even when PO25RP is replaced by Ig-deficient mice serum (Fig. 4). Therefore, it is evident that PO25 LPS binds MBL-MASP complex and activates the LP.

The hemolytic activity against PO25 LPS-coated SRBC as well as PO25RP concentration in serum decrease rapidly during the anaphylactoid reaction (Fig. 6A, B), suggesting the involvement of LP in this process. The drop of PO25RP level is connected with MBL-A but not MBL-C consumption (Fig. 7), although both lectin forms are able to activate complement 26. Recently, the contribution ofMBL-A to the pathogenesis of septic shock was reported by Takahashi et al. 27, who observed a higher survival rate of MBL-A-deficient mice undergoing acute septic peritonitis. The role of MBL-A in myocardial ischemia-reperfusion injury was shown by Jordan et al. 28. Pretreatment with anti-MBL-A mAb decreased the infarct size, neutrophil accumulation and C3 deposition, and attenuated the expression of pro-inflammatory genes after ischemia-reperfusion.

Previously, Kawabata et al. 11 showed that the inhibition of C5 with K-76 COOH or its consumption with cobra venom factor protected mice from death induced by Prevotellaintermedia and Salmonella typhimurium LPS, respectively. We were able to save animals from death by blocking the complement activation at earlier stages with Ab against PO25RP and C3 component (Table 3).

The most important fact, indicating the crucial role of MBL-MASP complex and complement-LP in the anaphylactoid reaction, is the protective activity of anti-PO25RP (connected with blocking of the carbohydrate recognition domain in MBL molecules and inhibition of LP activation) and anti-MASP-1 sera (Fig. 7A, Table 3). Rossi et al. 29 reported that recombinant MASP-1 possesses rather weak C4- and C2-cleaving potency; however, protease isolated from plasma was demonstrated to lyse C3 and C2 30. In our experiment, anti-MASP-1 Ig did not inhibit PO25RP-induced hemolysis of PO25 LPS-SRBC (Fig. 5B).

Recently, Hajela et al. 19 suggested that MASP-1 is important in defense against spreading of infection due to fibrinogen and plasma transglutaminase (factor XIII) cleavage.Fibrinogen is a dimer of subunits built up of three chains: α, β and γ of molecular masses of 67, 56 and 48 kDa, respectively. Thrombin, being its natural activator, liberates fibrinopeptides A and B from α and β chains, forming fibrin monomers. Thrombin also activates factor XIII, which is responsible for fibrin web stabilization being a result of cross-linking of γ and α chains. Our investigation showed a cleavage of fibrinogen by PO25RP (Fig. 8). Therefore, neutralization of MASP-1 with specific Ab might interrupt the intravascular coagulation processes, which could be speculated to participate in the development of anaphylactoid shock. It is also imaginable that binding of the Ab to any element of RaRF complex disturbs its activation simply by conformation change or steric effect. These data also suggest that clinical use of Ab to MBL and MASP for treating certain causes of LPS hypersensitivity shock might be possible.

We suppose that the PO25 LPS core oligosaccharide rather than OPS is the MBL target (Table 2). In contrast to the long chain-rich LPS fraction, the short chain-rich one inhibited the binding of PO25RP to mannan and PO25 LPS as efficiently as a not-separated product. We have previously demonstrated that at the nonreducing end of the PO25 LPS core, GlcNAc and Hep residues are present 15. As reported by Childs et al., MBL-A preferentially binds to oligosaccharides with terminal GlcNAc 31. Both mentioned sugars, even in monosacharidic form, also weakened the interaction between LPS or mannan and PO25RP. A similar phenomenon was reported in the case of S. typhimurium Ra LPS 1. We found, however, despite the high efficiency of MBL binding, its shock-inducing activity to be low (data not shown), as that of rough mutant of KO3 bacteria 10. This may be explained by a higher clearance rate of Ra LPS from blood circulation in comparison to smooth forms (32 and our unpublished data), probably due to positive charge and/or larger micelle size. Devyatyarova-Johnson et al. suggested the significance of LPS molecule arrangement for MBL attachment. They demonstrated that MBL binds more efficiently to shorter polysaccharide chains of endotoxin 33. One could speculate that to induce anaphylactoid reaction PO25 LPS molecules must be sufficiently short to bind MBL via outer core, and simultaneously, sufficiently long for low clearance rate.

4 Materials and methods

4.1 Animals

LPS-hyporesponsive C3H/HeJ mice were purchased from Charles River Wiga (Germany). BALB/c mice were obtained from the Institute of Microbiology and Immunology, University of Łódź, Poland. RAG-2–/– mice of BALB/c origin were provided by the Central Institute for Experimental Animals, Kawasaki, Japan. Mast cell-deficient WBB6F1-W/Wv mice and their wild-type littermate controls were obtained from Japan SLC Inc., Hamamatsu, Japan. New Zealand white rabbits came from the Nofer Institute of Occupational Medicine, Łódź.

4.2 Reagents

C4D was obtained from animals with genetic deficiency of C4 that were raised at the Kitasato University, Sagamihara, Japan. Human C4 and thrombin were purchased from Sigma, and fibrinogen fromCalbiochem. Peroxidase-labeled streptavidin came from Vector Laboratories. Gelatin was purchased from Merck (Germany). Tris-buffered saline containing 1 mM CaCl2 (TBS-Ca2+) contained 10 mM Tris-HCl, 120 mM NaCl and 1 mM CaCl2 (pH 8.0); gelatin-supplemented veronal-buffered saline (GVB) contained 0.1% gelatin (pH 7.4); gelatin- and divalent cation-supplemented veronal-buffered saline (GVB2+) contained 0.3 mM CaCl2 and 2 mM MgCl2. Low-ionic-strength GVB2+ (μ=0.07) contained 2.5% mannitol (mannitol-, divalent cation- and gelatin-supplemented veronal buffer; MGVB2+). Imidazole buffer contained 40 mM imidazole, 1.25 M NaCl, 50 mM CaCl2 and 5% BSA (pH 7.8). MBL-binding buffer contained 20 mM Tris-HCl, 10 mM CaCl2, 1 M NaCl, 0.05% Triton X-100 and 0.5% BSA (pH 7.4). Hepes buffer contained 20 mM Hepes, 140 mM NaCl and 5 mM CaCl2 (pH 8.0).

4.3 Bacterial strains and LPS

Proteus vulgaris PO25 (strain PrK 48/57) came from the the Czech National Collection of Type Cultures. K. pneumoniae strain K55:O3 was kindly provided by Prof. S. Kałuźewski (National Institute of Hygiene, Warsaw, Poland). LPS were prepared by the phenol-water method 34 and purified according to the described technique 35. To prepare the ester-bound fatty acid-free LPS (alkali-treated), endotoxin was treated with 0.25 N sodium methanolate as described 36. This product was used for the coating of SRBC. To obtain long and short polysaccharide chain-rich fractions, LPS was fractionated by gel-permeation chromatography using Sephadex G200 (Pharmacia, Sweden) 37. The effectiveness of separation was checked by SDS-PAGE. E. coli O127 LPS was purchased from Sigma.

4.4 Isolation of PO25 LPS-reacting protein from mouse serum and Western blot analysis

The PO25RP was isolated according to the described method 16. Briefly, mouse serum was incubated overnight with PO25 bacterial cells. After centrifugation, the cells were washed with TBS-Ca2+, and the bound proteins were eluted from the bacterial surface with 10% GlcNAc solution in TBS-Ca2+. Bacterial cells were removed by centrifugation. Supernatant fraction was termed PO25RP. SDS-PAGE was performed according to the procedure described by Laemmli 38. Molecular mass standards came from Sigma. For the Western blot analysis, samples were transferred to polyvinylidene difluoride membranes (Bio-Rad). Rabbit anti-PO25RP, rat anti-MASP-2 serum or rabbit anti-MASP-1 Ig were used as primary Ab. Peroxidase-conjugated goat anti-rabbit Ig or rabbit anti-rat Ig (DAKO, Denmark) were employed to detect bound proteins. For detection of MASP-1 and MASP-2, the membrane was incubated with enhanced chemiluminescence substrate (Santa Cruz Biotechnology), and then exposed to an X-ray film. In the case of MBL detection, 4-chloro-1-naphtol (Sigma) was employed as a substrate for peroxidase.

For determination of the reactivity of PO25RP with PO25 LPS, PO25RP was charged to LPS-epoxy-Sepharose column (the resin was purchased from Pharmacia). PO25 LPS-Sepharose 6B was prepared according to the described procedure 9. After washing of the column with TBS-Ca2+, the bound materials were eluted with 10% GlcNAc-containing TBS-Ca2+. Fractions giving maximal optical density (OD) at 280 nm were concentrated by 30-kDa Vivaspin 2 (Sartorius, GB) and analyzed by SDS-PAGE/Western blot.

4.5 Sequence analysis

The N-terminal amino acid sequence was determined by a gas-phase sequencer (Model 491, Perkin-Elmer Applied Biosystems) and Microgradient Delivery System Model 140C equipped with a Programmable Absorbance Detector Model 785A (both from Perkin-Elmer Applied Biosystems).

4.6 Antisera and antibodies

To obtain anti-PO25RP serum, rabbits were immunized with 600 μg of PO25RP together with CFA followed by three immunizations with 1,200 μg of PO25RP together with IFA (adjuvants came from The Binding Site, UK). This serum, called anti-PO25RP, and non-immune rabbit serum were used in experiments after heat inactivation (56°C, 30 min) and pre-absorption with PO25 bacteria and/or SRBC. To obtain anti-PO25 serum, rabbits were immunized with heat-killed bacterial cells 39. Anti-MASP-1 Ab (Ig fraction) was prepared from serum of rabbits immunized with peptide (Gly73 to Glu86) of MASP-1 40. Goat anti-human C4 and anti-human C3 Ab were purchased from Sigma. RAG-1–/– serum was prepared from rag1-knockout mice of C57BL/6 origin. mAb against mouse MBL-A (2B4) and MBL-C (14D12) and rat anti-MASP-2 serum were kindly provided by Prof. S. Thiel, University of Aarhus, Denmark.

4.7 Complement activation and its inhibition

The LP of complement activation was determined using hemolytic assay or ELISA-based C4b deposition test. To prepare PO25 LPS-SRBC, 1 ml of SRBC suspension (109/ml) was incubated for 30 min at 37°C with 50 μg of alkali-treated PO25. The cells were washed with cold PBS and resuspended to the concentration 108 cells/ml in MGVB2+.

To test the C4-activating potency of PO25RP, serially diluted samples (50 μl) were incubated with 100 μl of PO25 LPS-SRBC for 30 min at 30°C. After washing with GVB2+, the cell pellet was resuspended in 150 μl of MGVB2+ and incubated with various amounts of C4 at the same temperature for 20 min. Next, cells were washed with GVB2+ and suspended in 200 μl of MGVB2+. Then 100 μl of C4D (diluted 1:70) was added and the incubation was prolonged for another 30 min at 37°C. Finally, 1 ml of cold GVB containing 10 mM EDTA was added tostop the reaction. The rate of hemolysis was evaluated by measuring the OD of the supernatant at 414 nm. To test the C4-activating potency of mouse serum, samples were diluted in a volume of 100 μl and the incubation time was prolonged to 60 min and 90 min for PO25 LPS-SRBC and C4D, respectively.

To test the inhibitory activity of antiserum, 25 μl of PO25RP (600 ng) was incubated with the same volume of anti-PO25RP and anti-MASP-2 serum or anti-MASP-1 Ig for 30 min at 30°C before incubation with PO25 LPS-SRBC, C4 and C4D. The hemolysis rate was determined as described above. Alternatively, 25 μl of PO25RP was incubated with the same volume of anti-PO25RP or buffer for 2 hat 37°C and then mixture was transferred to mannan-coated microtiter plates. Detection of bound MBL-A and MBL-C was performed with the help of specific Ab.

C4b deposition property via LP was determined in ELISA according to the method described by Petersen et al., in conditions excluding participation of C1 18. Briefly, PO25RP or mouse serum diluted in MBL-binding buffer was incubated in mannan-or PO25 LPS-coated ELISA plates, which was followed by incubation with 5 μg/ml of C4 or MBL-deficient human serum (1:2,500).The bound C4b was detected with rabbit anti-C4 and peroxidase-labeled goat anti-rabbit Ig.

4.8 Fibrinogen cleavage

Fibrinogen (60 μg) was incubated with PO25RP or thrombin in a final volume of 40 μl of Hepes buffer for 2 h at 37°C 19. Digestion products were analyzed by SDS-PAGE.

4.9 Determination of PO25RP concentration in sera

BALB/c mice (8 weeks old, five per group) were pretreated i.p. with 100 μg of MDP. After 4 h, animals received 100 μg of KO3 or PO25 LPS i.v. Blood was taken at the moment of appearance of first symptoms of anaphylactoid reaction (usually after 5–7 min). The control and the E. coli O127 LPS-treated mice were sacrificed 30 min after i.v. injection of NaCl or LPS, respectively. The PO25RP concentration in serum was determined in ELISA. The MaxiSorp U96 plates (Nunc, Denmark) were coated with PO25 LPS or mannan (Sigma). Test sera or PO25RP diluted in imidazole buffer or in MBL-binding buffer were added and incubated for 90 min at 37°C. Then, rabbit anti-PO25RP serum and peroxidase-conjugated goat anti-rabbit Ig Ab (DAKO) or biotinylated mAb against mouse MBL-A or mouse MBL-C and peroxidase-labeled streptavidin were added. To investigate the inhibiting ability of various reagents equal volumes of test inhibitor and fresh mouse serum were incubated for 30 min at 37°C and then transferred to the PO25 LPS- or mannan-coated microtiter plates.

4.10 Induction of anaphylactoid reaction in MDP-primed mice and protection by Ab

Mice received an i.v. injection of 100 μg of LPS 4 h after i.p. injection of 100 μg of MDP. Incidence, scoring 10 of the anaphylactoid reaction and death were recorded within 1 h. For investigation of protective activity, mice were injected i.v. with heat-inactivated (56°C, 30 min) antisera or Ig fractions of immune sera 60 min before LPS injection. The survival was assessed for 72 h. Controls received heat-inactivated non-immune rabbit serum.

4.11 Statistical analysis

Statistical analysis was performed with the help of the Kruskal-Wallis test.

Acknowledgements

This work was supported by grants 6P04A04816 (Polish State Committee for Scientific Research, KBN, Poland); 11670270 (Ministry of Education, Science and Culture of Japan) and IIC-4 (Ministry of Health and Welfare, Japan).

Footnotes

  1. 1

    WILEY-VCH

  2. 2

    WILEY-VCH

  3. 3

    WILEY-VCH

  4. 4

    WILEY-VCH

  5. 5

    WILEY-VCH

  6. 6

    WILEY-VCH

  7. 7

    WILEY-VCH

  8. 8

    WILEY-VCH

Ancillary