Immuno-detection of anthrose containing tetrasaccharide in the exosporium of Bacillus anthracis and Bacillus cereus strains

Authors


Gerd Pluschke, Molecular Immunology, Swiss Tropical Institute, CH-4002 Basel, Switzerland. E-mail: gerd.pluschke@unibas.ch

Abstract

Aims: Bacillus anthracis strains of various origins were analysed with the view to describe intrinsic and persistent structural components of the Bacillus collagen-like protein of anthracis glycoprotein associated anthrose containing tetrasaccharide in the exosporium.

Methods and Results:  The tetrasaccharide consists of three rhamnose residues and an unique monosaccharide – anthrose. As anthrose was not found in spores of related strains of bacteria, we envisioned the detection of B. anthracis spores based on antibodies against anthrose-containing polysaccharides. Carbohydrate–protein conjugates containing the synthetic tetrasaccharide, an anthrose–rhamnose disaccharide or anthrose alone were employed to immunize mice. All three formulations were immunogenic and elicited IgG responses with different fine specificities. All sera and monoclonal antibodies derived from tetrasaccharide immunized mice cross-reacted not only with spore lysates of a panel of virulent B. anthracis strains, but also with some of the B. cereus strains tested.

Conclusions:  Our results demonstrate that antibodies to synthetic carbohydrates are useful tools for epitope analyses of complex carbohydrate antigens and for the detection of particular target structures in biological specimens.

Significance and Impact of the Study:  Although not strictly specific for B. anthracis spores, antibodies against the tetrasaccharide may have potential as immuno-capturing components for a highly sensitive spore detection system.

Introduction

Anthrax is an acute zoonotic disease caused by the spore-forming, rod-shaped bacterium Bacillus anthracis that can lie dormant in the soil for decades. Upon ingestion by grazing animals, the spores germinate, changing from the resistant form into the growing and toxin producing vegetative form. After the bacteria disseminate, they typically kill the infected animal and return to the environment converting again to spores. Bacillus anthracis has been described as the ultimate biological weapon because of its virulence and persistence when disseminated as spores (Inglesby et al. 1999; Borio et al. 2001).

Human vaccines for anthrax are available (Turnbull 2000; Pittman et al. 2001), but there has been much controversy over the safety and effectiveness of the current vaccines. Research on a second-generation vaccine in recent years was based on the observation that antibodies to protective antigen (PA) are crucial for protection against exposure to virulent anthrax spores (Brey 2005). As antibodies to PA address the toxaemia component of anthrax disease, it is assumed that an effective anthrax-subunit vaccine should contain multiple antigens. Inclusion of killed B. anthracis spores enhances the protective efficacy of PA-based vaccines in animal models (Brossier et al. 2002). To generate an immunity that protects from infection with B. anthracis, the capsule or somatic antigens in the spore may represent critical vaccine components.

The exosporium, the primary permeability barrier of the spore and the source of spore surface antigens has been the focus of recent investigations. As the outermost surface of the spore, the exosporium is likely to be the most immunologically accessible structure of the spore. The glycoprotein Bacillus collagen-like protein of anthracis (BclA) was the first exosporium protein that was identified and has been shown to be an immunodominant protein suggesting a role for this protein in spore–host interactions (Steichen et al. 2003). BclA has conserved amino- and carboxy-termini and a long, central collagen-like region that is similar to mammalian collagen proteins (Sylvestre et al. 2002, 2003). This polymorphic collagen-like region consists of GXX repeats, including a large proportion of GPT triplets and (GPT)5GDTGTT repeats. The latter 21-amino-acid repeat has been named BclA repeat (Sylvestre et al. 2002) and appears to be an essential feature of the BclA protein important for the structural organization of BclA on the spore surface. The number of GXX repeats in the collagen-like region varies among strains (Steichen et al. 2003; Sylvestre et al. 2003) and is responsible for the different lengths of the exosporium filament found on spores of different B. anthracis strains (Sylvestre et al. 2003). Two O-linked carbohydrates attached to BclA, a 715-Da tetrasaccharide and a 324-Da disaccharide, have been identified (Daubenspeck et al. 2004). Multiple copies of the tetrasaccharide are linked to the collagen-like region of BclA, whereas the disaccharide may be attached outside of this region. Most of the collagen-like region repeating units contain a threonine residue that provides sites for potential glycosylation via a GalNAc linker. The tetrasaccharide is composed of three rhamnose residues and anthrose, an unusual sugar that was not found in spores of strains belonging to the phylogenetically most similar species B. cereus and a B. thuringiensis (Daubenspeck et al. 2004; Tamborrini et al. 2006). Here, we demonstrate cross-reactivity of antisera and monoclonal antibodies raised against the synthetic B. anthracis tetrasaccharide and truncated anthrose containing structures using immunoblots with lysed spores of B. anthracis and B. cereus strains.

Material and methods

Synthesis of carbohydrates

The tetrasaccharide I was prepared as described before by our group (Werz and Seeberger 2005; Werz et al. 2007). Following our initial report, four other groups achieved the synthesis of this tetrasaccharide or corresponding sequences (Adamo et al. 2005; Saksena et al. 2005, 2006, 2007; Mehta et al. 2006; Crich and Vinogradova 2007; Guo and O’Doherty 2007). The disaccharide II as well as the anthrose III used for further vaccination studies were prepared in an analogous manner by our modular approach (Werz and Seeberger 2005; Werz et al. 2007). The experimental details for the preparation of II and III as well as their analytical data are described in the Supporting Information (Figs S1 and S2). The terminal double bonds of II and III were transformed into thiol functionalities suitable for reaction with a maleimide-functionalized KLH carrier protein. These KLH-glycoconjugates were used for the mouse immunogenicity studies.

Bacterial strains

Bacillus anthracis strains selected from different genotypes (Maho et al. 2006; Pilo et al. 2008) are described in Table 1. Bacillus thuringiensis ATCC 29730; B. cereus ATCC 10876; B. cereus ATCC 10876; B. cereus ATCC 13061; B. cereus ATCC 14579: B. cereus ATCC 33019, were purchased from the American Type Culture Collection (Rockville, MD, USA). Bacillus cereus F4370/75 (Cereus III lineage, ST-27); B. cereus m1545 (Cereus I lineage; ST-5); B. cereus m1564 (Cereus I lineage; ST-6); B. cereus m1293 (Cereus II lineage; ST-45); B. cereus F4810/72 (Cereus II lineage; ST-26) were obtained from F.G. Priest (Priest et al. 2004).

Table 1.   Comparative analysis of the CLR composition of BclA from sequenced PCR products
StrainsCluster (according to Pilo et al. 2008)OriginBclA protein of CLR
GPT repeatsBclA repeats [(GPT)5GDTGTT]GXX repeats (present once)
No. 1
JF3788
A4Human anthrax, Switzerland, from wool factory, 1981546GAT
GLT
GPS
GLG
No. 2
JF3853
A3BCattle, central Switzerland, 1952575GAT
GLT
GPS
GLG
GDT
GTT
No. 3
JF3852
B2Cattle, Bern, Switzerland, 1953211GAT
GLT
GPS
GLG
No. 4
JF3854
B2Cattle, central Switzerland, 1957211GAT
GLT
GPS
GLG
No. 5
JF3783
A4Human anthrax, Switzerland, from wool factory, 1981171GAT
GLT
GPS
GLG
No. 6
JF3784
A4Air filter wool processing factory with outbreak 1981171GAT
GLT
GPS
GLG
No. 7
JF3785
A4Goat hair, wool processing factory with outbreak 1981171GAT
GLT
GPS
GLG
No. 8
JF3851
A4Unknown origin171GAT
GLT
GPS
GLG
No. 9
JF3786
A4Goat hair, wool processing factory with outbreak 1981171GAT
GLT
GPS
GLG
No. 10
JF3787
A4Goat hair, wool processing factory with outbreak 1981171GAT
GLT
GPS
GLG
No. 11
A73
ndUnknown origin171GAT
GLT
GPS
GLG

Production of spores from Bacillus anthracis

Strains of Bacillus anthracis were cultured on tryptone soya agar with 5% sheep blood (Oxoid, Basel, Switzerland) at 37°C for 18 h. Then, the culture plates were kept at room temperature for 4 weeks until the colonies appeared dry. Subsequently, the colonies were suspended in 1 ml PBS buffer (50 mmol l−1 Na2HPO4 - NaH2PO4 pH 8·0, 140 mmol l−1 NaCl) per culture plate, heated at 80°C for 10 min. Spores were then collected by centrifugation at 4000 g for 15 min and suspended in PBS at a titre of 109 spores per ml.

Mouse immunogenicity studies

Mice carrying human immunoglobulin Cγ1 heavy and Cκ light chain gene segments (Pluschke et al. 1998) were immunized with synthetic saccharide antigens (Fig. 1) conjugated to keyhole-limpet-haemocyanine (KLH) formulated in ImmunEasyTM adjuvant (Qiagen). Starting on day 0, they received three doses of 40 μg conjugate at three-weekly intervals. Blood was collected before each immunization and 2 weeks after the final injection. The production of anti-tetrasaccharide monoclonal antibodies is described in (Tamborrini et al. 2006). Anti-B. thuringiensis antiserum was generated by immunization with strain ATCC 29730 B. thuringiensis spores (Raven Biological Laboratories, INC, Omaha, NE, USA). Mice were immunized subcutaneously at three-weekly intervals with 106B. thuringiensis spores formulated in 0·2 ml PBS without adjuvant. Two weeks after the third injection, heart blood was collected and serum stored at −20°C until use.

Figure 1.

 Structure of the synthetic tetrasaccharide (I) of the major surface glycoprotein of Bacillus anthracis and deletion sequences of it: anthrose–rhamnose (II; disaccharide); anthrose (III; monosaccharide). Attachment via a pentenyl handle to KLH carrier protein afforded the conjugates for immunological studies.

Animals were housed in temperature-controlled rooms (22 ± 3°C). Conventional laboratory feeding and unlimited drinking water were provided to the mice. Approval for animal experimentation has been obtained from the responsible authorities and all experiments have been performed in strict accordance with the Rules and Regulations for the Protection of Animal Rights laid down by the Swiss Bundesamt für Veterinärwesen. All animal manipulations have been performed under controlled laboratory conditions by specifically qualified personnel in full conformity to Swiss and European regulations.

Enzyme-linked immunosorbent assay (ELISA)

ELISA microtitre plates Immunolon 4 (Dynex Technologies Inc., Chantilly, VA, USA) were coated at 4°C overnight with 50 μl of a 10 μg ml−1 solution of saccharide-BSA conjugate in PBS, pH 7·2. Wells were then blocked with 5% milk powder in PBS for 1 h at room temperature followed by three washings with PBS containing 0·05% Tween-20. Plates were then incubated with two-fold serial dilutions of mouse serum or mAbs in PBS containing 0·05% Tween-20 and 0·5% milk powder for 2 h at room temperature. After washing, the plates were incubated with alkaline phosphatase-conjugated goat anti-mouse IgG (γ-chain specific) antibodies (Sigma, St Louis, MO, USA) for 1 h at room temperature and then washed. Phosphatase substrate (1 mg ml−1p-nitrophenyl phosphate (Sigma)) in buffer (0·14% Na2CO3, 0·3% NaHCO3, 0·02% MgCl2, pH 9·6) was added and incubated at room temperature. The optical density (OD) of the reaction product was recorded after appropriate time at 405 nm using a microplate reader (Titertek Multiscan MCC/340; Labsystems, Finland).

PCR and DNA sequencing

PCR was performed using FIREPol® (Solis BioDyne, Tartu, Estonia) Taq polymerase and synthetic oligonucleotides hybridizing to flanking or internal sequences of the bclA gene. A segment including the entire bclA gene was amplified by using the primers 5′-CCGTTAGAATCCATTGCAAGATGATAAGGC-3′ and 5′-CGACCAACCATACTGTGTGCAGCTCTTGGC-3′ (Sylvestre et al. 2003). The sequence encoding the BclA variable collagen-like region containing the GXX-repeats was amplified by using the primers 5′-CCCTAATCTTGTAGGACCTACATTACCACC-3′ and 5′-CCCACCGGAGTTAAATGCATATAGTCCTGC-3′ (Sylvestre et al. 2003). Thirty microlitre reactions were carried out with a GeneAmp® 9700 PCR System (Applied Biosystems). Template DNA was denatured at 94°C for 4 min. Thirty amplification cycles of 30 s of denaturation at 94°C, 30 s of annealing at 58°C and 2 min of extension at 72°C were performed followed by one cycle of 10 min at 72°C. The PCR products were analysed by agarose gel electrophoresis and sequenced by using two primers flanking the bclA gene (5′-CGTGTCATTTTCTTTCGGTTTTGCATCTAC-3′ and 5′-GTGCCTCCTACGGAATGTCATACAAC-3′) by Macrogen (Korea).

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting

Endospore suspensions were solubilized using 2× loading buffer [1·7 ml of 0·5 mol l−1 Tris-HCl (pH 6·8), 2 ml of glycerol, 4·5 ml of 10% sodium dodecyl sulfate, 1 ml of β-mercaptoethanol, 0·8 ml of bromophenol blue (0·3%, w/v)] and heated for 10 min at 95°C before loading onto the 10% SDS-PAGE. As molecular weight marker SeeBlue®Plus (Invitrogen) was used. Separated proteins were electrophoretically transferred to a nitrocellulose filter (Protran Nitrocellulose, BA85; Schleicher & Schuell) by semidry blotting. Blots were blocked with PBS containing 5% milk powder and 0·1% Tween 20 overnight at 4°C. Whole blots or cut strips were incubated with appropriate dilutions of immune serum or mAbs in blocking buffer for 2 h at room temperature. In competition experiments, primary antibodies were pre-incubated for 30 min with synthetic competitors. After several washing steps, blots were incubated with goat anti-mouse IgG horseradish peroxidase conjugated Ig (Bio-Rad Laboratories, CA, USA) or with alkaline phosphatase-conjugated goat anti-mouse (gamma) heavy-chain antibodies (Sigma, St Louis, MO, USA) for 1 h. Blots were finally developed either using the ECL system according to manufacturer’s instructions or with 5-bromo-4-chloro-3-indolylphosphate (Bio-Rad, Reinach, Switzerland) and nitroblue tetrazolium (Bio-Rad) to visualize bands.

Results

Fine specificity of antibodies raised against anthrose-containing synthetic carbohydrates

Synthetic B. anthracis tetrasaccharide, a rhamnose–anthrose disaccharide and anthrose (Fig. 1) were covalently attached to the keyhole-limpet-haemocyanine (KLH) carrier protein by reductive amination to improve their immunogenicity by the recruitment of carrier-specific T cells. Mice were repeatedly immunized with a CpG based adjuvant (ImmunEasyTM, Qiagen) formulation of the conjugates and the fine specificity of the immune sera was analysed by ELISA and immunoblotting.

All anthrose monosaccharide-immunized animals developed high IgG ELISA titres to anthrose. Cross-reactivity with the disaccharide and the tetrasaccharide antigen was comparatively low (Fig. 2a). Sera from anthrose–rhamnose disaccharide immunized mice not only reacted best with the disaccharide immunogen itself, but also showed cross-reactivity with the anthrose monosaccharide and the B. anthracis tetrasaccharide (Fig. 2b). Sera and mAbs from tetrasaccharide immunized mice were highly specific for the tetrasaccharide (Fig. 2c,d respectively). The three tetrasaccharide specific mAbs that were generated previously (Tamborrini et al. 2006) shared the IgG2b : λ isotype and exhibited similar fine specificity in all analyses.

Figure 2.

 Development of mouse IgG responses specific for the anthrose-containing tetrasaccharide, disaccharide and monosaccharide compounds are shown in Fig. 1. Mice were immunized with monosaccharide-KLH (a), disaccharide-KLH (b) or tetrasaccharide-KLH (c) formulations. Shown are ELISA readouts obtained with serial dilution of mouse sera taken pre-immune (inline image, PI) and 2 weeks after the second (inline image, 2.imm) third (inline image, 3.imm) or fourth (inline image, 4.imm) immunization. Response patterns of individual tetrasaccharide specific mAbs (MTA1, inline image; MTA2, inline image; MTA3, inline image) are depicted in (d). In (a) and (b), mean ELISA readouts ± SD of sera from three mice are shown.

Cross-reactivity of sera raised against the synthetic carbohydrate structures with the tetrasaccharide expressed by B. anthracis on the BclA protein was investigated by immunoblotting after separation of B. anthracis endospore lysate proteins by SDS-PAGE. With the endospore lysate of strain 1 (Table 1), all sera stained a double band with an apparent molecular weight of around ∼250 kDa, corresponding to the BclA glycoprotein (Fig. 3). In competition experiments, the binding of the anti-tetrasaccharide mAbs (Fig. 3c) and of anti-tetrasaccharide serum IgG (data not shown) to the spore antigen of B. anthracis strain 1 was inhibited only by the synthetic tetrasaccharide in a concentration dependent manner. Partial structures, such as trirhamnose, rhamnose-anthrose disaccharide, monorhamnose (data not shown), anthrose and a variant of anthrose (data not shown) missing the 3-hydroxy-3-methylbutamido side chain (Fig. 3d) failed to compete the strong binding to the native tetrasaccharide antigen. Antigen binding of anti-disaccharide antiserum (Fig. 3b) was blocked by tetrasaccharide and disaccharide, but not by anthrose. All competitors containing anthrose blocked antigen binding of anti-anthrose immune serum, whereas the side chain variant of anthrose did not inhibit binding (Fig. 3c).

Figure 3.

 Competition Western blot experiments with a set of overlapping synthetic sugars were used for epitope mapping. Bacillus anthracis endospore suspensions were separated by SDS-10% PAGE under reducing conditions and blotted onto a nitrocellulose membrane. Anti-monosaccharide mouse sera (a), anti-disaccharide mouse sera (b) and mAb MTA1 (c) were pre-incubated with synthetic competitors and afterwards added to cut strips. After incubation with an alkaline phosphatase-conjugate, blots were developed with 5-bromo-4-chloro-3-indolylphosphate and nitroblue tetrazolium to visualize bands. (d) Structures of different competitors used. Anthrose* = anthrose without the 3-hydroxy-3-methylbutamido chain; k = no competitor; a = 100 μg ml−1 competitor; b = 1 μg ml−1 competitor. Immune sera were used at a dilution of 1 : 2000 and mAb MTA1 at a concentration of 1 μg ml−1.

Diversity of post-translational modification of the BclA protein in different B. anthracis strains

Immunoblotting experiments with spore lysates of a set of 11 B. anthracis isolates (Table 1) yielded strain-specific multiple band staining patterns characteristic for the BclA protein (Sylvestre et al. 2003), when an anti-B. thuringiensis antiserum was used as primary antibody (Fig. 4e). Although the BclA proteins have a calculated molecular-weight of only 20–39 kDa, both high molecular weight bands (>150 kDa) and low-molecular weight bands (∼50 kDa) appear, because of glycosylation of the protein (Sylvestre et al. 2003). Patterns obtained with the tetrasaccharide-specific mAbs (Fig. 4b) and immune sera (data not shown) included high-molecular-weight (>150 kDa) bands for all strains tested. In the case of strains 1 and 2 as well as the ‘Sterne’ (34F2) strain (data not shown), staining of the high-molecular weight band was very intense and no other bands became visible. Intense staining of low-molecular weight bands was observed with strains 3 and 4 (∼55 kDa) and 5 to 8 (∼50 kDa), but not with strains 9 to 11, indicating a possible difference between strains in the levels of glycosylation. With disaccharide (Fig. 4c) and anthrose (Fig. 4d), specific immune sera comparable staining patterns were observed. However, with strains 3 to 11 staining of the high-molecular-mass bands in relation to the low-molecular mass bands was increased, indicating that accessibility of the different epitopes differs for the high- and low-molecular weight bands.

Figure 4.

 (a) Differences in the size of the bclA gene in 11 B. anthracis strains (lanes 1 to 11; see also Table 1). The sequences encoding the collagen-like regions were amplified by PCR and analysed by 2% agarose gel electrophoresis. Molecular size markers are indicated in base pairs (bp). (b–e) Western blot analysis with spore lysates of the 11 B. anthracis strains. Endospore lysates were separated by SDS-10% PAGE and blotted onto a nitrocellulose membrane. Blots were incubated with anti-tetrasaccharide mAb MTA1 (0·01 μg/ml) (b), with disaccharide- (c), anthrose- (d) and B. thuringiensis-antisera (e). All immune sera were used at a dilution of 1 : 1000 and blots were developed using the ECL system. The sizes of the molecular weight markers are given in kDa. (f) EZBlueTM (Sigma) protein staining of the 11 B. anthracis separated endospore lysates on a SDS-10% gel.

To test whether immunoblot patterns correlated with the size and structure of the tetrasaccharide-carrying collagen-like region of the BclA protein, the bclA genes of the 11 B. anthracis strains were amplified by PCR and sequenced. While the amino- and carboxy-terminal sequence stretches were conserved, a wide variation in size of the central collagen-like regions from 69 to 219 amino acids was observed (Table 1), as reflected by size differences of the PCR products (Fig 4a). The collagen-like regions varied in the numbers of GXX units and the number of BclA repeats, defined by (Sylvestre et al. 2003) as the 21-amino-acid sequence (GPT)5GDTGTT. In all eleven B. anthracis strains, the same sets of four nucleotide sequences coded for the GPT repeats and the BclA repeats respectively (data not shown).

Strains 1 and 2 associated with strong staining of the high-molecular-weight (>150 kDa) bands had much larger collagen-like regions than the other strains (Table 1). The collagen-like regions of strains 3 and 4 that exhibited an additional ∼55 kDa molecular weight band in immunoblotting were slightly larger than those of strains 5 to 8 that showed a slightly smaller (∼50 kDa) band. The bclA genes of strains 9 to 11 that exhibited no low-molecular weight bands were identical to those of strains 5 to 8. As indicated by gene exchange experiments (Sylvestre et al. 2003), variations in the BclA protein banding patterns thus can be caused by post-translational modifications that depend more on the genetic background of the strains than on the sequences of the bclA genes.

Cross-reactivity of antisera with endospores of Bacillus cereus

In immunoblotting experiments with spore lysates of nine B. cereus strains covering the Cereus I, II and III lineages from the MLST B. cereus clade 1 (Priest et al. 2004), tetrasaccharide specific mAbs (Fig. 5) as well as tetra-, di- and monosaccharide specific immune sera (data not shown) exhibited reactivity with three strains. Staining of a high-molecular-weight band with an anti-B. thuringiensis spore antiserum demonstrated the expression of the immunodominant BclA protein in all B. cereus strains.

Figure 5.

 Western blot analysis of the reactivity of anti-tetrasaccharide mAb MTA1 (a) and anti-B. thuringiensis antiserum (b) with spore lysates. Total spore-lysates of B. anthracis strain 1 (lane 1), B. cereus F4370/75 (lane 2), B. cereus m1545 (lane 3), B. cereus m1564 (lane 4), B. cereus m1293 (lane 5), B. cereus F4810/72 (lane 6), B. cereus ATCC 10876 (lane 7), B. cereus ATCC 13061 (lane 8), B. cereus ATCC 14579 (lane 9), B. cereus ATCC 33019 (lane 10) and B. thuringiensis ATCC 29730 (lane 11) were separated by SDS-10% PAGE and blotted onto a nitrocellulose membrane. Anti-B. thuringiensis antiserum was used at a dilution of 1 : 1000 and the mAb at a concentration of 0·01 μg/ml. (c) EZBlueTM (Sigma) protein staining of the 11 Bacillus species separated on a SDS-10% gel. The sizes of the molecular mass markers are given in kDa.

Discussion

Specific detection and phenotypic differentiation of B. anthracis is challenging (Williams et al. 2003), because of its genetic similarity to other bacteria of the B. cereus group (Helgason et al. 2000). Extensive genomic studies on strains of B. anthracis, B. cereus and B. thuringiensis have suggested that B. anthracis and B. thuringiensis are subspecies of the species B. cereus. The main difference between these subspecies is the presence of plasmids coding for insecticidal toxins in B. thuringiensis and the presence of the capsule plasmid pXO2 and the toxin plasmid pXO1 in B. anthracis (Edwards et al. 2005). DNA homology studies of different B. anthracis strains revealed evanescent small amounts of genomic variation. This significant homogeneity of B. anthracis may be the result of the organism surviving the majority of its life as a spore, where it is not exposed to DNA-altering events (Henderson et al. 1995; Helgason et al. 2000; Schupp et al. 2000). The BclA protein is an immunodominant spore antigen and it has been shown that the BclA protein itself and not its carbohydrate constituent directs the dominant immune response (Steichen et al. 2003). Identification of a BclA protein-associated tetrasaccharide with a previously undescribed terminal sugar anthrose in the animal vaccine strain ‘Sterne’ of B. anthracis (Daubenspeck et al. 2004) had raised hope to have identified a B. anthracis specific antigen on the surface of the spore. It was proposed to use this antigen to develop a specific oligosaccharide-based detection system for B. anthracis spores, as anthrose was not found in spores of the B. cereus T strain and a B. thuringiensis ssp. kurstaki strain (Daubenspeck et al. 2004). The structural analysis of the tetrasaccharide was confirmed by our finding that monoclonal antibodies raised against the synthetic tetrasaccharide are cross-reactive with intact B. anthracis endospores of B. anthracis avirulent strain ‘Sterne’ and highly virulent strain Ames in immunofluorescence assay (Tamborrini et al. 2006). Here, we have tested these monoclonal antibodies and antisera raised against the tetrasaccharide, a rhamnose–anthrose disaccharide and anthrose for cross-reactivity with immunoblotted spore lysates of panels of B. anthracis and B. cereus strains. In immunoblotting experiments, the mAbs and immune sera cross-reacted not only with all B. anthracis isolates, but also with some of the B. cereus strains tested. The observed cross-reactivity demonstrates a broader presence of anthrose in the B. cereus group, supporting the phenotypic argument that B. cereus, B. thuringiensis and B. anthracis could be considered as a single bacterial species (Helgason et al. 2000). Recent genetic evidence demonstrating the presence of the anthrose biosynthetic operon in some B. cereus strains supports our findings (Dong et al. 2008). The collagen like region of the B. cereus BclA proteins contains GXX units harbouring threonine residues representing sites for attachment of O-linked oligosaccharides (Castanha et al. 2006).

Four different BclA-related multiple-band patterns were observed. The multiplicity of bands is also seen when B. anthracis lysates are analysed with anti-BclA protein antibodies (Sylvestre et al. 2003). We investigated whether the size and structure of the bclA genes and the corresponding proteins explains the observed differences in the anti-carbohydrate immunoblot patterns. Sequence differences were restricted to the polymorphic collagen-like regions of the BclA protein and correlations between the size of the collagen-like regions and the relative apparent molecular weights on SDS-PAGE gels were observed. However, for the B. anthracis strains 5 to 11, two different anti-tetrasaccharide blotting patterns were observed, while their bclA genes are identical. It has been demonstrated by gene exchange experiments (Sylvestre et al. 2003) that genetic background dependent post-translational modifications of the BclA protein can cause such variations in the banding pattern. It has been suggested that post-translational modifications, probably implicated in the stability or glycosylation of BclA are responsible for the multiple-band Western blotting patterns (Sylvestre et al. 2003). Our results with strains 5 to 11 emphasize the role of bclA gene-independent genetic factors in this phenomenon. In contrast to the mono- and di-saccharide specific sera, the tetrasaccharide specific antibodies, which seem to bind to a complex epitope comprising all four sugar residues of the tetrasaccharide, showed preferential binding to the bands with low apparent molecular weight. This observation may indicate that the tetrasaccharide as a whole is better accessible in the low-molecular weight BclA species. While anti-anthrose and anti-disaccharide antibodies are likely to bind, even if only the terminal residues are easily accessible, the anti-tetrasaccharide antibodies are likely to be much more sensitive to shielding of the deeper parts of the tetrasaccharide. Alternatively, glycosyl chains associated with high- and low-molecular weight species may slightly differ in their structure. Changes in the tri-rhamnose core structure, such as lack or substitution of one of the rhamnose residues are likely to affect the tetrasaccharide-specific antibodies, but not those raised against anthrose or anthrose-rhamnose.

Even though the antibodies generated against anthrose-containing structures are not strictly specific for B. anthracis, they detect spore lysates of B. anthracis from a broad phylogenetic spectrum, thus representing a valuable immuno-capturing component for a highly sensitive spore detection system. Such sensitive immunoassays would be plagued by false-positive results, but would afford a quick exclusion of negative samples from further analysis. Subsequent confirmatory expensive and time-consuming diagnostic tests with high specificity, such as real-time PCR or standard culture methods that focus on the vegetative form, are therefore minimized. Moreover, it would be important to detect spores of B. cereus, as it is an opportunistic human pathogen and might be used as basis for engineered biological weapons.

Acknowledgements

We thank the Swiss Tropical Institute, the ETH Zürich, the Alexander von Humboldt Foundation (AvH) and the Deutsche Forschungsgemeinschaft (Feodor Lynen and Emmy Noether Fellowships to D.B.W.) for financial support, F.G. Priest for B. cereus strains and M. Holzer for excellent technical assistance.

Ancillary