Detection of B. anthracis Spores and Vegetative Cells with the Same Monoclonal Antibodies

Bacillus anthracis, the causative agent of anthrax disease, could be used as a biothreat reagent. It is vital to develop a rapid, convenient method to detect B. anthracis. In the current study, three high affinity and specificity monoclonal antibodies (mAbs, designated 8G3, 10C6 and 12F6) have been obtained using fully washed B. anthracis spores as an immunogen. These mAbs, confirmed to direct against EA1 protein, can recognize the surface of B. anthracis spores and intact vegetative cells with high affinity and species-specificity. EA1 has been well known as a major S-layer component of B. anthracis vegetative cells, and it also persistently exists in the spore preparations and bind tightly to the spore surfaces even after rigorous washing. Therefore, these mAbs can be used to build a new and rapid immunoassay for detection of both life forms of B. anthracis, either vegetative cells or spores.


Introduction
Bacillus anthracis, the causative agent of anthrax disease, is a Gram-positive spore-forming bacterium. Fully virulent bacilli carry two plasmids, pXO1 and pXO2, which contain genes to produce the lethal factor and edema factor toxins, and a poly-c-Dglutamic acid capsule, respectively [1,2]. In response to nutrient deprivation, B. anthracis will produce spores that can withstand harsh conditions, including temperature, radiation, chemical assault, time and even the vacuum of outer space [3]. These remarkable characteristics allow B. anthracis to be used as a biological threat agent.
Recent studies have indicated that B. anthracis is genetically similar to other members of the Bacillus genus. The 16S rRNA, 23S rRNA and 16S-23S internal transcribed spacer sequences of B. anthracis share a high degree of similarity with those in B. cereus, B. subtilis, B. megaterium, B. mycoides, and B. thuringiensis [4,5]. These sequence similarities make identification of B. anthracis challenging, and substantial effort has been devoted to developing identification methods. The conventional method is the bacteriological assay, which is reliable but time consuming. Other advanced approaches have been proposed for B. anthracis detection, including immunological assays [6,7], PCR-based methods [8,9,10,11,12,13], molecular fingerprinting [14,15,16] and mass spectrometric (MS) analyses [17,18]. The detection targets have mainly focused on the pXO1 and pXO2, gene polymorphisms, specific gene sequences (Ba813, rpoB, bclA) and small acid soluble proteins (SASPs) in the spores. However, these methods fail to eliminate the need for complicated protocols such as cell disruption, nucleic acid extraction and protein purification, and cannot support convenient, rapid and real-time detection of B. anthracis. Therefore, direct detection of B. anthracis is attractive for on-site application. So far, several immunoreagents, directed against the surface of B. anthracis, have been developed for the direct detection of B. anthracis vegetative cells or B. anthracis spores [19,20,21,22,23].
In the current study, attempts were also made to generate mAbs using fully washed B. anthracis spores as an immunogen. Unexpectedly, we identified three high affinity mAbs (8G3, 10C6 and 12F6) capable of direct and species-specific recognition of both B. anthracis spores and vegetative cells. Furthermore, these mAbs were all directed against the EA1 protein, which is well known as a major S-Layer protein in B. anthracis vegetative cells [24,25]. Some recent studies have revealed that EA1 is also retained in the proteomic profiling of rigorous washed spores and even salt/detergent washed exosporium [26,27]. Although Williams and Turnbough suggested EA1 was not a true spore surface protein, they stated this protein persistently existed in the spore surface [28]. Therefore, it is valid to use EA1 as a detection target of B. anthracis spores herein, as EA1 is highly associated with the spore surface. In this study, we conclude that the mAbs we prepared, directed against EA1, can recognize the surface of B. anthracis spores as well as vegetative cells, and we also suggest EA1 protein can serve as a potential marker for the detection of B.
anthracis. This study is significant because until now, there have been no monoclonal antibodies reported for direct and speciesspecific detection of both life forms of B. anthracis.

Atomic-force microscopy (AFM) analysis of B. anthracis spores
To guarantee the purity of spores that we prepared, both unwashed and fully washed spores were analysed by AFM. As is shown in Figure 1, a larger amount of free spores appeared. Even the unwashed spores were surprisingly clean, as no intact vegetative cells and little vegetative cell debris were present (Fig. 1A). However, because it is likely that some vegetative cell proteins will bind accidentally to the spore surface, the rigorous washing method, as described above, was still employed. There were few differences between the unwashed spores and fully washed spores, but the latter seemed to have a much smoother surface than the former, indicating that our spore purification protocol had removed some unknown material (Fig. 1B). Therefore, we used these fully washed B. anthracis spores as our immunogen and employed them in the subsequent experiments.
Preparation and screening of mAbs against B. anthracis spores By the fusion protocol [29], approximately 600 hybridoma cultures were screened with the indirect ELISA method. Fifteen hybridomas produced high affinity antibodies and were cloned successfully. Following isotyping, five mAbs were IgMs, while the other 10 were IgGs (data not shown). These mAbs were then examined for their reactivity against a range of Bacillus spores using ELISAs. Three mAbs, designated 8G3 (IgG3), 10C6 (IgG1) and 12F6 (IgG1), specifically recognized B. anthracis spores ( Fig. 2A), and showed no cross-reaction with high concentrations of B. cereus, B. subtilis, B. megaterium, B. mycoides and other Bacillus spores.
Additionally, compared to the other mAbs, these three mAbs all had higher affinity against B. anthracis spores. Even with a range of low concentrations, these mAbs, especially mAb 8G3, always showed strong signals when reacting against B. anthracis spores (Fig. 2B). Therefore, these three mAbs were examined further.

Antigen identification
Exosporium, the outermost structure of spores, was isolated from B. anthracis A16 spores using mild sonication in order to avoid the loss of some exosporium proteins and the damage of the remaining spore protein denaturation. Using electron microscopy, it was observed that the exosporium fragments were successfully extracted (Fig. 3A), and there was no obvious disruption to the spores after sonication (Fig. 3B). The exosporium fragments were boiled in sample buffer for 30-40 min to release more soluble proteins. The soluble exosporium proteins were separated by SDS-PAGE, and the mAbs (8G3, 10C6 and 12F6) recognized a 91-93 kDa protein by Western blotting (Fig. 4A). The band was excised and subjected to DE-MALDI-TOF-MS analysis; the protein was identified as EA1 (Fig. 4B). To determine whether the mAbs were all directed against EA1, the coding region of the gene for EA1 was cloned into the pQE-30 plasmid and expressed in E. coli M15, and the expressed proteins were purified with affinity chromatography. By the method reported previously [30], the affinity constants (K aff ) of the mAbs (8G3, 10C6 and 12F6) for the recombinant EA1 protein were measured to be 1 to 3610 9 M 21 Thus, the results not only established that EA1 was the target protein of the three mAbs, but also indicated that the mAbs had a high affinity for EA1.The reaction curves for the mAbs with 4 mg ml 21 recombinant EA1 are shown in Figure 5.
The mAbs could also react with B. anthracis vegetative cells Since EA1 is a major S-Layer protein in B. anthracis vegetative cells [24,31], the mAbs against EA1 were tested for their reactivity and specificity to B. anthracis vegetative cells. The sandwich ELISA method was employed, rather than indirect ELISA, because the intact vegetative cells were so large that were not adequately coated to the microwells in our indirect ELISA assays (data not shown).
As shown in Figure 6, the mAbs specifically recognized B. anthracis vegetative cells with no significant cross-reaction with other strains at high concentrations (10 7 CFU/ml). Of these mAbs, 12F6 seemed the best candidate, because it reacted strongly with different kinds of B. anthracis vegetative cells. However, the other mAbs, especially 8G3, did not strongly recognize B. anthracis strains lacking the pXO1 or pXO2 plasmids. Therefore, these results indicated that the mAbs had different characteristics, despite having been raised against the same protein.

The mAbs bound to surfaces of both B. anthracis spores and vegetative cells
Binding of mAbs (8G3, 10C6 and 12F6) to the surface of B. anthracis A16 spores and vegetative cells was investigated with immunofluorescence. Normal mouse IgG was used as a negative control ( Figure S1), and photographs from the 8G3 experiments as examples are shown in Figure 7. In the immunofluorescence assays, the fully washed spores incubated with 8G3 were extensively stained with R-phycoerythrin-conjugated goat anti-mouse IgG (red). The vegetative cells were incubated with the mAb, and were visualized with FITC-conjugated goat anti-mouse IgG (green). The pattern of fluorescence indicated that these mAbs not only recognized the surfaces of spores, but also bound to the intact vegetative cells.

EA1 was persistently present in B. anthracis spore extracts
During preparation, the B. anthracis spores were washed extensively with water and pelleted through 20% and 50% Renografin to remove vegetative cell debris [10]. The spores were monitored and collected at four stages of the standard washing procedure: unwashed spores from the culture plate (Preparation 1), spores washed with MilliQ water for 24 h, three times (Preparation 2), Renografin purified spores (Preparation 3), and spores subjected to three additional washes (Preparation 4). In addition, the supernatants from sedimentation during Preparation 3 were diluted 1:10 with water and centrifuged at 10,0006g for 30 min to collect the debris. Equal volumes of all preparations and the debris were subjected to SDS-PAGE analysis.
As a result, the B. anthracis A16 SDS-PAGE profiles showed no obvious differences between Preparations 1 and 2, except that a negligible amount of protein was removed by the washing steps ( Figure 8). The same results were observed when Preparations 3 and 4 were compared. In contrast, most proteins, including EA1, decreased after sedimentation through Renografin. These removed proteins were present in the ''debris'' of the supernatant, which contained a similar amount of protein to Preparations 2 and 4. These results suggest that the Renografin step washed out some of the EA1 along with many other spore proteins, which were then detected by SDS-PAGE. Despite the presence of EA1 in the debris, EA1 was also persistently present in the B. anthracis A16 spore extracts of each preparation. Furthermore, B. anthracis strains lacking pXO1 or pXO2 produced similar results to those presented above (data not shown). Therefore, EA1 was persistently present in B. anthracis spore extracts.

Discussion
The primary goal of this study was to generate mAbs with high affinity and specificity that could be applied to rapid detection of B. anthracis spores. The mAbs were produced against formaldehyde-inactivated B. anthracis A16 spores and reacted with a range of live Bacillus spores, including B. anthracis. Most of the mAbs we produced were highly specific for B. anthracis spores. For each screening of the hybridoma cultures, spores from B. cereus and B. thuringiensis, the two closest relatives of B. anthracis, were used as negative controls. To identify mAbs with high affinity and specificity, hybridomas were selected if the mAbs reacted strongly with B. anthracis but did not recognize either negative antigen. As a result, the three mAbs (8G3, 10C6 and 12F6) we prepared have no cross reaction with many B. thuringienesis subspecies and B. cereus isolates (Table S1).
The three mAbs recognized not only the surface of B. anthracis spores but could also detect intact B. anthracis vegetative cells (Fig. 7). Furthermore, these mAbs were capable of reaction with live B. anthracis as well as dead B. anthracis (inactivated by 1.5% formaldehyde), which is critical for the detection of biological warfare agents in unknown ''white powders'', since it has been suggested that Bacillus inactivation would affect antibody detection assays [32]. Although these three mAbs were directed toward the same target protein, EA1, they had different characteristics. The mAb 12F6 was superior at reacting with different kinds of B. anthracis vegetative cells, while 8G3 had a higher affinity for B. anthracis spores and the target protein EA1 (Fig. 2, Fig. 5 and Fig. 6). Besides this, in the epitope mapping, the epitopes of mAb 8G3 and 10C6 were concluded to be located from the amino acid 275 to 435 on the EA1 protein, and the epitope of mAb 12F6 was exactly located from the amino acid 465 to 554 (data not shown). We suggested that the different positions of the mAb epitopes caused the mAbs to exhibit different behavior the detection of B. anthracis.
As to whether EA1, a major S-layer component of B. anthracis vegetative cells, is also a spore protein, much research has indicated that this protein is retained in the proteomic profiling of spores and salt/detergent washed exosporium [26,27]. However, Williams and Turnbough stated that this protein was merely a persistent contaminant in spore preparations [28]. Whichever is correct, it does not matter for the detection of B. anthracis spores, because this protein does persistently exist in each of the spore preparations (Fig. 8), and the B. anthracis spores, even after full washing, can be detected with our anti-EA1 mAbs ( Fig. 2 and Fig. 7).
Although EA1 could be partially washed out during the rigorous washing step in our study, the other proteins detected by SDS-PAGE also apparently decreased and were present as debris in the supernatant of centrifuged purified samples (Fig. 8). The supernatant debris contained significant amounts of true spore proteins with similar protein profiles to fully washed spores. Therefore, rigorous washing methods, such as Renografin purification, can cause serious loss of spore surface associated proteins and are not suggested for proteomic analysis [26,27]. This suggests that EA1 is certainly at least a highly spore-associated protein: it might be a true spore protein, or may anchor onto particular spore surface components.
In conclusion, this study reports three mAbs (8G3, 10C6 and 12F6) that can bind to B. anthracis spores and intact vegetative cells with high species-specificity and affinity. It also indicates that EA1, the target protein of our mAbs, could serve as a potential detection

Bacterial strains
The bacterial strains used in this study are listed in Table 1. All assays involving live B. anthracis spores and vegetative cells were carried out in a P3 biosafe laboratory. The experimenters were equipped with masks, gloves and exposure suits.

Preparation of spores
Spores were prepared by growing bacteria at 37uC on modified Difco sporulation medium (DSM) [33], containing 6 g tryptone, 3 g yeast extract, 10 g NaCl, 1 g KCl, 0.25 g Mg 2 SO 4 ?7H 2 O, 0.23 g Ca(NO 3 ) 2 , 0.197 g MnCl 2 ?4H 2 O, 0.0002 g FeSO 4 , and 15 g agar per liter. When more than 95% free spores appeared, the spores were collected with cold, sterile, ultrapure water, centrifuged and washed extensively three times, before sedimentation through 20% and 50% Renografin (Bracco Sine, China) [10] and another three washes. The washed spores were then resuspended in sterile saline and stored at 4uC. The spores were quantified by serially diluting the spore stock, and plating 100 ml aliquots on Luria broth (LB) plates, in triplicate. The number of spores (CFU/ml) was counted following overnight culture at 37uC.

Atomic-force microscopy (AFM) analysis of spores
An aqueous spore suspension of 10 ml, containing 10 8 -10 9 spores, was spread on a 1 cm61 cm silicon wafer and air-dried. The sample was not placed in the AFM chamber for imaging until the spores settled on the substrate. Picoscan TM 2500 AFM (Molecular Imaging, US) and commercial single MAClever type II cantilevers (Molecular Imaging, US) were used to obtain images in tapping mode.

Exosporium extraction
Approximately 0.5 g of purified B. anthracis spores (wet weight) were collected by centrifugation at 9,0006g for 5 min at 4uC and resuspended in 15 ml cold sterile TE buffer (50 mmol l 21 Tris-HCl, 0.5 mmol l 21 EDTA, pH 7.2), containing 1 ml ml 21 protease inhibitor cocktail (Sigma, US). All subsequent centrifugations were performed at 4uC. To extract larger exosporium fragments, the spore suspension was divided into three partitions and each was treated using a Vibra-Cell Ultrasonic Processor (Sonics & Materials, USA) with 120 pulse cycles on ice (5 s pulse -9 s interval -5 s pulse). The three partitions were sonicated, pooled and centrifuged for 45 min at 1,2006g. Then the pellet was washed five times with cold TE buffer and centrifuged as above. All the supernatant samples were saved and centrifuged at 184,0006g for 1 h. The sediments, which were composed of exosporium fragments, were then resuspended in 500 ml sterile PBS for further study.

Production of monoclonal antibodies and polyclonal antibodies
Preparations containing 10 6 spores of B. anthracis strain A16 and inactivated by 1.5% formaldehyde were injected subcutaneously  into six-week-old BALB/c mice. The immunization was repeated three times at two-week intervals before boosting by intraperitoneal injection. The spleen cells were removed 3 d later and fused with SP2/0 myeloma cells, according to the procedures of Kohler and Milstein (1975) [29]. The hybridomas were cloned by limit dilution, screened using ELISA (described below) and then injected intraperitoneally into BALB/c mice. The mAbs were isotyped using an isotyping kit (Sigma, US), according to the manufacturer's directions, and were purified by caprylic acidammonium sulfate precipitation of ascites [34].
For the polyclonal antibodies against anthracis vegetative cells, the rabbit was injected subcutaneously with 10 7 cells (inactivated by 1.5% formaldehyde) for three times at two week before boosting by double injection dose. The polyclonal antibodies were purified from the antiserum by caprylic acid-ammonium sulfate precipitation of ascites.

ELISA
For indirect ELISA detection, 96 well microtiter plates were coated with 100 ml per well of carbonate-bicarbonate (CB) buffer (pH 9.6) containing either 10 7 -10 8 /ml spores or recombinant EA1 (2, 4, 8, 16 mg/ml) and incubated overnight at 4uC. The wells were blocked with 200 ml of blocking buffer (5% skim milk in phosphate buffered saline (PBS) for 2 h at 37uC. Hybridoma culture supernatants or a concentration series of purified mAbs were added in 100 ml aliquots to individual wells and incubated for 30 min at 37uC. Horseradish peroxidase (HRP)-conjugated goat anti-mouse antibodies were added at a dilution of 1/900 and incubated for 20 min.
For sandwich ELISA, the microtiter plates were coated with purified mAbs (10 mg/ml) and blocked as described above. The plates were incubated with a range of Bacillus vegetative cells at 37uC for 1 h and then reacted with 10 mg/ml polyclonal antibodies against anthrax at 37uC for 40 min. HRP-conjugated goat anti-rabbit antibodies were added to the plate at a dilution of 1/3000 (Boster, China) and reacted for 30 min at 37uC.
For both indirect ELISA and sandwich ELISA, five washes with PBS with 0.05% Tween 20 (PBST) were carried out between each step. Normal mouse IgG was used as a negative control, and all antibody dilutions were prepared in PBS, containing 1% skim milk. A tetramethylbenzidine substrate (0.1 mg/ml 100 ml per well) was added for approximately 10-15 min at 37uC to start the reaction, and 50 ml of 2 M H 2 SO 4 was added to stop the reaction. Then the absorbance was measured at 450 nm, and each assay was performed in quadruplicate.

Analysis by SDS-PAGE and immunoblotting
Samples were mixed with an equal volume of sample buffer (50 mM Tris-HCl at pH 6.8, 0.5 mol l 21 b-mercaptoethanol,  20% v/v glycerol, 10% w/v SDS, 0.2% w/v bromophenol blue) and boiled for 10-40 min. The soluble proteins were analyzed by SDS-PAGE using a Bio-Rad gel apparatus, according to the manufacturer's instructions. For immunological detection, the proteins were transferred from the gels onto polyvinylidene difluoride (PVDF) membranes (Millipore, US). The membranes were blocked in Tris-buffered saline (TBS) containing 5% skim milk at 4uC overnight, and were washed three times for 5 min with Trisbuffered saline Tween-20 (TBST, with 0.05% Tween 20). Then the membranes were incubated in 5 mg ml 21 mAbs, which were diluted in TBS containing 1% skim milk, for 1.5 h at room temperature, followed by three additional 10 min washes. The membranes were incubated for 1 h with a 1/100 dilution of HRP-conjugated goat anti-mouse IgG. Four thorough 10 min washes were completed before the substrate buffer (6 mg 3,39-Diaminobenzidine (DAB) and 10 ml of 37% H 2 O 2 in 10 ml TBS) was added and allowed to react for 3-5 min. Images were acquired using a Canon digital camera and analyzed using Adobe Photoshop software.

MS protein analysis
The protein band identified on the SDS-PAGE gel was excised, washed, and subjected to tryptic digestion [26]. The peptides were collected and analyzed by delayed extraction-matrix assisted laser desorption ionization-time of flight mass spectrometry (DE-MALDI-TOF-MS). The MS spectra were obtained using an Axima-CFR Plus Mass Spectrometer (Shimadzu Biotech, Japan). All MS analyses were performed at the Institute of Biophysics, Chinese Academy of Sciences, China.

Gene cloning and expression
Sequence data for the B. anthracis genome were obtained from the National Center for Biotechnology Information (NCBI) web site at http://www.ncbi.nlm.nih.gov. Based on the sequence information, DNA primers were designed to amplify the entire eag open reading frame of the B. anthracis A16 strain using polymerase chain reaction (PCR). The chromosomal DNA of A16 was used as the template, and the primer sequences were 59-TATTGGATCCATGGCAAA GAC-TAACT-39 and 59-CTATAGAGCTCGTATAGATTTGGGTT A -39. The PCR conditions were as follows: 30 cycles of 94uC for 5 min, 94uC for 45 s, 40uC for 45 s, and 72uC for 2 min for 30 cycles; and lastly 72uC for 5 min. The PCR products were digested with BamHI and SacI and were inserted into plasmid pQE-30 (Qiagen, Germany). The recombinant plasmids were transformed into E. coli M15 to express the cloned gene in vitro. Firstly, the E. coli M15, carrying recombinant plasmids, were incubated at 37uC at 160 rpm to an average OD 600 of 0.5 followed by addition of IPTG to 1 mM final concentration, and were incubated for another 4 h at 30uC. Secondly, the cells were harvested by centrifugation at 3000 g for 5 min, and were sonicated by a Vibra-Cell Ultrasonic Processor (Sonics & Materials, USA) as described above. Lastly, after centrifugation at 12,000 g for 20 min, the supernatant containing recombinant EA1 protein with His tag was purified using Ni-nitrilotriacetic acid agarose (Qiagen, Germany) affinity chromatography, according to the manufacturer's instructions.

Determination of the mAbs affinities
As the antibody affinities reported previously [30]

Indirect immunofluorescence
For the indirect immunofluorescence assays, 200 ml aliquots of spore or vegetative cell suspension (10 9 CFU ml 21 ) were pelleted by centrifugation at 10,0006g for 1 min at room temperature. The spores and vegetative cells were resuspended by adding 500 ml blocking buffer and incubated for 2 h at 37uC with gentle shaking (70-100 rpm). Aliquots of mAbs (500 ml; 1 mg ml 21 ) were added and the mixture was incubated for 1 h at 37uC with shaking. R-phycoerythrin-conjugated goat anti-mouse IgG (Sigma, US) or FITC-conjugated goat anti-mouse IgG (Boster, China) was used as the secondary antibody (100 ml, 1/50 dilution), and was added to the spores or vegetative cell solution respectively. The mixture was incubated for 40 min at room temperature in the dark. Each step was followed by three 10 min washes in PBST, followed by centrifugation at 10,0006g for 1 min to remove the supernatant. The antibodies were collected in PBS containing 1% skim milk, and images were acquired using a Leica confocal microscope. Figure S1 Confocal microscopy images of negative control antibody binding to B. anthracis. The secondary antibodies were