Development of ELISA based on Bacillus anthracis capsule biosynthesis protein CapA for naturally acquired antibodies against anthrax

Abstract

Anthrax is a zoonotic disease caused by the gram-positive spore-forming bacterium Bacillus anthracis. Detecting naturally acquired antibodies against anthrax sublethal exposure in animals is essential for anthrax surveillance and effective control measures. Serological assays based on protective antigen (PA) of B. anthracis are mainly used for anthrax surveillance and vaccine evaluation. Although the assay is reliable, it is challenging to distinguish the naturally acquired antibodies from vaccine-induced immunity in animals because PA is cross-reactive to both antibodies. Although additional data on the vaccination history of animals could bypass this problem, such data are not readily accessible in many cases. In this study, we established a new enzyme-linked immunosorbent assay (ELISA) specific to antibodies against capsule biosynthesis protein CapA antigen of B. anthracis, which is non-cross-reactive to vaccine-induced antibodies in horses. Using in silico analyses, we screened coding sequences encoded on pXO2 plasmid, which is absent in the veterinary vaccine strain Sterne 34F2 but present in virulent strains of B. anthracis. Among the 8 selected antigen candidates, capsule biosynthesis protein CapA (GBAA_RS28240) and peptide ABC transporter substrate-binding protein (GBAA_RS28340) were detected by antibodies in infected horse sera. Of these, CapA has not yet been identified as immunoreactive in other studies to the best of our knowledge. Considering the protein solubility and specificity of B. anthracis, we prepared the C-terminus region of CapA, named CapA322, and developed CapA322-ELISA based on a horse model. Comparative analysis of the CapA322-ELISA and PAD1-ELISA (ELISA uses domain one of the PA) showed that CapA322-ELISA could detect anti-CapA antibodies in sera from infected horses but was non-reactive to sera from vaccinated horses. The CapA322-ELISA could contribute to the anthrax surveillance in endemic areas, and two immunoreactive proteins identified in this study could be additives to the improvement of current or future vaccine development.

Fig 1. Expression and immunoreactivity of candidate proteins.

(A) Coomassie brilliant blue (CBB) staining and (B–D) Western blotting of the proteins in cell pellet fractions of the control and candidate protein-expressing Escherichia coli strains grown in terrific broth. In Western blotting, the proteins in the cell pellets were probed with different antibodies: (B) anti-glutathione S-transferase (GST) immunoglobulin G; (C) horse hyperimmunized anti-Bacillus anthracis serum (PC1); (D) naive horse serum (NC1). Lane 1, Mw, molecular weight marker (in kDa); lane 2, control strain which is E. coli BL21 harboring empty pGEX-6P-2 without IPTG induction; lane 3, control strain which is E. coli BL21 harboring empty pGEX-6P-2 with 0.2 mM IPTG induction; lane 4, E. coli BTZ001 expressing recombinant hypothetical protein (GST-GBAA_RS28110: 44 kDa); lane 5, E. coli BTZ002 expressing recombinant capsule biosynthesis protein CapA (GST-GBAA_RS28240: 72 kDa); lane 6, E. coli BTZ003 expressing recombinant signal peptidase (GST-GBAA_RS28275: 47 kDa); lane 7, E. coli BTZ004 expressing recombinant peptide ABC substrate-binding protein (GST-GBAA_RS28340: 84 kDa); lane 8, E. coli BTZ005 expressing recombinant metal-dependent hydrolase (GST-GBAA_RS28430: 46 kDa). The arrows indicate target proteins in the expected sizes.

Fig 2. Expression and purification of CapA322.

(A) Genetic map of Bacillus anthracis capsule-coding operon revealing the CapA322, C-terminus region (amino-acid 322–411) of CapA used in antigen preparation. (B) Coomassie brilliant blue staining (CBB) and (C, D, and G) Western blotting of the fractions collected after affinity batch purification. In Western blotting, proteins were probed with different horse sera: (C) horse hyperimmunized anti-B. anthracis serum (PC1); (D) naive horse serum (NC1); (G) vaccinated horse serum (Vac_H3D21). Lane 1, Mw, molecular weight marker (in kDa); lane 2, the supernatant fraction applied to affinity beads (GST-CapA322: 37 kDa); lane 3, beads bound; lane 4, elute after treatment with PreScission Protease (CapA322: 11 kDa). (E and F) CBB and Western blotting of the fractions collected during the cation exchange process. Lane 1, Mw, molecular weight marker (in kDa); lane 2, sample loaded onto the cation exchange chromatography column; lanes 3 and 4 flow-throughs; lane 5, the eluted protein. Host cell-derived contaminants indicated as Escherichia coli’s protein.

Fig 3. Horses anti-PAD1 immunoglobulin G (IgG) responses against subcutaneously injected Bacillus anthracis Sterne 34F2 strain spore vaccine.

(A) anti-PAD1 IgG response of vaccinated horses in serum dilution 1:100. Each serum sample was tested in technical triplicate by PAD1-ELISA. Checkerboard titration between PAD1 and (B) horse hyperimmunized anti-B. anthracis serum (PC1) or (C) naive horse serum (NC1) in PAD1-ELISA. Each dilution of serum sample was tested in technical triplicate. The twofold serum dilution starts with a dilution of 1:100, and PAD1 dilution starts with 1.6 μg/well. From the result, we determined the optimal concentrations of the antigen, antibody, and serum dilutions as follows: antigen, 0.4 μg/well; serum dilution, 1:100; second antibody dilution, 1:15,000.

Fig 4. Comparison of PAD1-ELISA and CapA322-ELISA.

Checkerboard titration between CapA322 and (A) horse hyperimmunized anti-Bacillus anthracis serum (PC1) or (B) naive horse serum (NC1) in CapA322-ELISA. Each dilution of serum sample was tested in technical triplicate by CapA322-ELISA. The twofold serum dilution starts with a dilution of 1:100, and CapA322 dilution starts with 1.6 μg/well. From the result, we determined the optimal concentrations of antigen, antibody, and serum dilutions as follows: antigen, 0.8 μg/well; serum dilution, 1:100; second antibody dilution, 1:15,000. (C) Comparison of PAD1-ELISA and CapA322-ELISA on horse sera. Each serum samples of horse hyperimmunized anti-Bacillus anthracis (PC1), horse naturally infected with B. anthracis (PC2 and PC3), naive horses (NC1 and NC2), and horse vaccinated with Sterne34F2 strain (Vac_H1D21-Vac_H4D21) was analyzed in technical triplicate by PAD1-ELISA and CapA322-ELISA. (D) One-way ANOVA with Tukey’s multiple comparison test on relative OD values of positive (PC; n = 3), negative (NC; n = 2), and vaccinated (Vac; n = 3) groups.