Background

Bacillus anthracis and Yersinia pestis are the pathogenic agents of anthrax and bubonic plague, respectively. They are zoonotic pathogens causing life-threatening diseases in humans and animals and posing high risks for public health [1, 2]. The Centers for Disease Control and Prevention (CDC) classified these pathogens as bioterrorism agents in category A of the most dangerous group of biological agents [3]. One of the most typical examples of bioterrorism, the “Anthrax letters” case in September 2001 in New York, USA, has affected the population with a huge impact at a psychological and political level [4]. Rapid screening of suspected specimens, therefore, helps to control the spread of epidemics, to contain the causative agents and to ensure the accurate and specific treatment.

Currently, there are a number of diagnostic methods for identification of B. anthracis and Y. pestis based on their phenotypic features, including bacteriological isolation, serology, and immunoassays. These approaches are time-consuming and required the use of skillful personnel and biosafety level 3 (BSL3) laboratory [3]. Moreover, serology and immune assays have limited sensitivity and specificity, particularly in case of immune assay-based rapid tests [5]. In recent years, a number of molecular approaches targeting genomic DNA have been developed for rapid and accurate identification of B. anthracis and Y. pestis, including conventional and simple multiplex polymerase chain reaction (mPCR), real-time PCR, and microarray assay applied for detection/discrimination of the pathogens in humans and environment [6,7,8,9,10,11,12]. Generally, in these assays, the detection has been targeted to virulence gene(s) on a symbiotic plasmid of each pathogen only; thus, identification and discrimination between pathogenic strains and non-pathogenic strains lacking either one or both virulence plasmids were limited. For B. anthracis, accurate diagnosis of pathogenic strains in nature is based on simultaneous detection of specific genes on their chromosomes and two virulent plasmids, pXO1 and pXO2. One of these plasmids may be lost in some strains, leading to a decrease in their pathogenicity [13]. Similarly, Y. pestis contains two virulence plasmids, pPCP1 and pMT1, with their respective genes encoding for virulence factors [14, 15]. In some cases, these plasmids are unstable and easy to be lost due to the environmental conditions or during serial subcultures [14, 16, 17]. Other factors that may inhibit successful PCR or generate false products regarding amplification of B. anthracis and Y. pestis from the environmental samples are the substances usually found in soil and water. Therefore, internal amplification control (IAC) in diagnostic PCR assays is necessary to be designed to detect false negative amplifications [18].

These two contagious pathogens have been detected and persisted in some areas in Vietnam. The sporadic cases of anthrax still occasionally appear in the North of Vietnam [19], and Y. pestis still persists in rodents in the Tay Nguyen area has the potential to cause epidemics [20]. An accurate and rapid identification assay to detect these pathogens is essential for early treatment and prevention of epidemic breaks. In the present study, we describe a rapid hexaplex PCR assay simultaneously targeting six genes developed for accurate identification of pathogenic B. anthracis and Y. pestis with the IAC included in the reaction to verify the positive and to avoid false negative results.

Materials and Methods

Experiments were carried out in the Division of Microbiology, the Institute of Biomedicine and Pharmacy, and the Department of Medical Microbiology, Military Hospital 103, Vietnam Military Medical University (VMMU).

Bacterial Strains and Growth Conditions

The strains of Bacillus anthracis 17JB, Bacillus cereus ATCC 6464, Yersinia enterocolitica ATCC 9610 and Escherichia coli ATCC 25922, are used as reference strains in this study. The screening bacterial strains from suspected various soil and aquatic environments (B1, B3–5, B8–13, B17–22, Y1), patients (B2, B6, B23, Y2–11), and livestock animals (B7, B14–16) in the north of Vietnam were collected by the research teams of Military Institute of Preventive Medicine, Vietnam Military Department of Medicine and VMMU, identified by conventional culture and PCR. Reference and screening bacterial strains were grown on broth of brain heart infusion (BHI) (Difco Laboratories, USA) and 5% (v/v) sheep blood agar plates (BioMérieux, Marcy l’Etoile, France) at 37 °C for 24 h for B. anthracis, and at 28 °C for 48 h for Y. pestis. Other reference strains were routinely cultivated on 5% (v/v) sheep blood agar overnight at 37 °C. Presumptive colonies studied the shape, the structure, and the colony characteristics; besides, Gram staining and biochemical tests of the isolates were carried out according to the manufacturer’s instructions (API50CHB, API20NE, BioMérieux, Marcy l’Etoile, France, Vitek2 Compact).

Genomic DNA Extraction from Bacterial Culture

All bacterial culture samples were heat inactivated at 121 °C for 20 min, and follow-up experiments were carried out in a BSL2 facility [21]. Genomic DNA was extracted using the commercial QIAamp DNA Mini Kit (Qiagen GmbH, Hilden, Germany) according to the manufacturer’s instructions. Extracted DNA samples were prepared at a concentration of 105 copies/μl and stored at 4 °C until use.

Primer Design and Standard PCR Assays

Sequences of specific genes in GenBank for both pathogenic and non-pathogenic B. anthracis and Y. pestis were selected as references for primer design for multiplex PCR. The vrrA (chromosome), pagA (pXO1 virulent plasmid), and capA (pXO2 virulent plasmid) genes, in particular, were selected as markers for B. anthracis. For Y. pestis, the ypo2088 (chromosome), pla (pPCP1 virulent plasmid), and caf1 (pMT1 virulent plasmid) genes, similarly, were selected. All primer pairs for the selected genes were self-designed in silico by using Primer3 tool in the National Center for Biotechnology Information (NCBI, https://www.ncbi.nlm.nih.gov/) and BioEdit software. List of primers designed, target genes, and amplicon sizes are shown in Table 1. Standard PCR and multiplex PCR assays were performed in a Master proS Cycler (Eppendorf, Hamburg, Germany) in the volume of 25 μl containing 1× PCR buffer (2 mM MgCl2), 0.2 mM of each dNTP, 0.2 μM of each primer, 1 U Dream Taq polymerase (Thermo Scientific, Wilmington, DE, USA), and 5 μl genomic DNA (approximately 10 ng). Thermocycling parameters were as follows: an initial denaturation at 94 °C for 4 min, followed by 30 cycles consisting of 94 °C for 1 min, 56 °C for 45 s, and 72 °C for 50 s, and a final extension at 72 °C for 7 min.

Table 1 List of primer pairs developed for internal amplification control and multiplex PCR in this study
Full size table

Optimization of Hexaplex PCR Assay

A hexaplex PCR was carried out by simultaneous addition of six primer pairs for vrrA, pagA, capA, ypo2088, pla, and caf1 genes (Table 1) in the same reaction mixture. Initially, equimolar primer concentration of 0.2 μM each was used in the multiplex PCR and adjustment of the proportion of each primer pair from 0.1 to 0.6 μM in the reactions to obtain optimal amplification of targeted genes. Secondly, a gradient of PCR annealing temperature was performed from 54 to 58 °C to find the optimal temperature allowing simultaneous and specific amplification of six different amplicons. Additionally, to increase the product yield of hexaplex PCR assay, other PCR components were also evaluated including MgCl2 0.2–2.0 mM, dNTPs 0.25–0.8 mM, buffer 1–1.5×, and number of PCR cycles from 30 to 40 cycles. Adjuvants such as BSA (bovine serum albumin), betaine (trimethylglycine) and DMSO (dimethyl sulfoxide) to be considered additional if necessary. After each amplification assay, PCR fragments were analyzed on 1.5% agarose gel electrophoresis in 0.5× TBE buffer at 110 V for 50 min and 0.5 μg/ml of ethidium bromide (Bio Basic, Markham, Ontario, Canada) stained and visualized with a BioDoc-it Imaging System (UVP, Upland, CA, USA).

Sensitivity and Specificity of the Hexaplex PCR Assay

Genomic DNA samples from B. anthracis and Y. pestis were serially diluted from approximately 20 ng/μl to 20 fg/μl with 5 μl from each dilution used for hexaplex PCR to identify the analytical sensitivity, corresponding to a range from 2 × 107 to 20 genome copies/reaction, respectively. Specificity of primer pairs in the developed hexaplex PCR was assessed using genomic DNA samples from B. cereus, Y. enterocolitica, and E. coli. The time stability of developed PCR assay which was stored at − 20 °C was tested monthly in the same standard concentration sample of B. anthracis and Y. pestis.

Hexaplex PCR Assay for the Detection of Soil Pseudo-Sample Contaminated Bacteria

In order to validate the multiplex PCR with the environment sample, we spiked the inactivated B. anthracis and Y. pestis culture samples into the soil. The final concentration of pseudo-samples was from 100 to 107 colony forming units/milliliter (CFU/ml). One milliliter of each sample was added to aliquots of soil (1 g) and incubated the contaminated soil for 24 h [22]. The suspension that was created from the soil by adding 2 ml of distilled water and carefully vortex was used to extract the DNA using the Soil DNA isolation kit (Norgen Biotek Corp., Canada) according to the manufacturer’s instruction.

Internal Amplification Control Designing

A competitive IAC used to verify the diagnosis results was a chimeric plasmid, and its construction is briefly described. A 654 bp fragment of lef gene of B. anthracis was amplified using IC1F/BamHI and IC1R/NdeI primer pairs (Table 1). This PCR product was cloned into a pJET1.2-capA plasmid consisting of the sequence of pJET1.2-blunt vector (Thermo Fisher Scientific Inc., Waltham, USA) and the capA gene of B. anthracis after it was linearized by BamHI and NdeI restriction enzyme. The ligation product was conventionally transformed into E. coli DH5α (Invitrogen Corp., Carlsbad, USA) [18]. The recombinant IAC named pJET1.2-capA-IC1 was sequenced to confirm its identity. Determination of the lowest reproducible IAC concentration during genomic DNA extraction was evaluated from 109 to 100 copies/sample from IAC dilutions.

Evaluation of Developed Hexaplex PCR with IAC

Evaluation of hexaplex PCR assay with pJET1.2-capA-IC1 was performed with a total of 38 samples, including B. anthracis (n = 11), B. cereus (n = 14), Y. pestis (n = 1), Y. enterocolitica (n = 1), and E. coli (n = 11), and compared with conventional culture methods. Additionally, the strains of B. anthacis and Y. pestis specifically determined by the hexaplex PCR assay were exactly confirmed by gel extraction of target bands and direct Sanger sequencing.

All the experiments in the present study were performed in triplicate and controls included.

Results

Selection and Primer Pair Design of Three Target Genes for B. anthracis and Three Target Genes for Y. pestis

In B. anthracis, to distinguish between pathogenic strains and non-pathogenic strains, we selected both the pagA gene (751 bp amplicon) and capA gene (610 bp amplicon) from pXO1 and pXO2, respectively, and a B. anthracis-specific target of the variable repeat region A—the vrrA gene (222 bp amplicon) as a stable chromosomal marker for a B. anthracis detection triplex PCR assay.

Similarly, the pla and caf1 are common target genes which are used to identify Y. pestis in molecular methods. Particularly, the pla gene (amplicon size 473 bp) encodes plasminogen activator located on the virulent plasmid pPCP1. Whereas, the cafF1/R1 primer pairs targeted the caf1 gene (amplicon size 271 bp) located on the virulent pMT1 plasmid which encoded the F1 fraction. Furthermore, the ypo2088 gene (amplicon size 103 bp) is a specific chromosomal marker of Y. pestis. Uniplex PCR amplicons of different genes of B. anthracis as well as Y. pestis yielded the products of expected sizes on agarose gel electrophoresis. The specificity of the primer pairs was checked in pooled two DNA templates of B. anthracis and Y. pestis by tri-primer PCR assays of each agent, to ensure that there was no cross-amplification (Fig. 1, Table 1).

Fig. 1
figure 1

Mono- and tri-primer PCR assay for the detection of B. anthracis and Y. pestis. Lane M, 100 bp DNA ladder (Thermo Scientific); NC, negative control (no DNA template). For the detection of B. anthracis: lane 1, mono-primer PCR assay consisting of pagAF1/R1 (virulent pXO1 plasmid); lane 2, capAF1/R1 (virulent pXO2 plasmid); lane 3, vrrAF1/R1 (chromosome). For the detection of Y. pestis: lane 4, mono-primer PCR assay consisting of plaAF1/R1 (virulent pPCP1 plasmid); lane 5, cafF1/R1 (virulent pMT1 plasmid); lane 6, ypoF1/R1 (chromosome). For the detection of pathogenic Y. pestis and B. anthracis with pooled two DNA templates: lane 7, tri-primer PCR assay including three primer pairs (ypoF1/R1, plaF1/R1, cafF1/R1); lane 8, tri-primer PCR assay including three primer pairs (vrrAF1/R1, pagAF1/R1, capAF1/R1)

Full size image

Optimization of Hexaplex PCR Assay

The optimization of a hexaplex PCR assay for high amplification efficiency of the specific PCR products has been conducted as follows. Firstly, the concentration of individual primer pairs was evaluated, and the optimized concentrations of three primer pairs for B. anthracis and three primer pairs for Y. pestis were 0.6 and 0.1 μM, respectively. Secondly, an optimal annealing temperature of 56 °C was found. Next, MgCl2 was obtained for 0.6 mM as the best concentration. After the concentration of MgCl2 was determined as of 0.6 mM, the dNTP concentration from 0.25 to 0.8 mM was tested and 0.25 μM was found the most optimal concentration. Finally, other components found for the hexaplex PCR were 2.0 U of Dream Taq polymerase, 1.2× buffer, and 32 amplification cycles. The optimal mPCR assay revealed the presence of six amplicons clearly separated on agarose gel by the electrophoresis (Fig. 2).

Fig. 2
figure 2

Optimal hexaplex PCR assay with pooled two DNA templates of B. anthracis and Y. pestis. Lane M, 100 bp DNA ladder (Thermo Scientific); lane 1, hexaplex PCR assay; lane 2, negative control

Full size image

Evaluation of the Detection Capability for Multiplex PCR Assay

The limit of detection of hexaplex PCR by using serious dilutions of bacterial genomic DNA as templates, for the individual detection of B. anthracis and Y. pestis, was 0.5 pg/reaction, equivalent to approximately 100 bacterial cells. The sensitivity of hexaplex PCR with spiking soil study for B. anthracis and Y. pestis was about 104 CFU/ml, and there were no notable differences in sensitivity for B. anthracis and for Y. pestis observed. However, some weak unspecific products appeared in the no-spiked and spiked soil samples (Fig. 3). The results proofed the overabundance of other microorganisms on soil samples in the study. Sequencing of six specific bands extracted from the gel confirmed the correspondent designed chromosomal and virulent targets of B. anthracis and Y. pestis (data not shown).

Fig. 3
figure 3

Multiplex PCR products from soil samples inoculated with B. anthracis and Y. pestis. Lane M, 100 bp DNA ladder (Thermo Scientific); lanes 1, 4, and 7, bacterial culture of B. anthracis, Y. pestis, and B. anthracis and Y. pestis; lanes 2, 5, and 8, soil not inoculated with B. anthracis, Y. pestis, and B. anthracis and Y. pestis; lanes 3, 6, and 9, soil inoculated with 106 CFU/ml of B. anthracis, Y. pestis, and B. anthracis and Y. pestis

Full size image

Optimization and Evaluation of the Multiplex PCR Assay Including Internal Amplification Control

Plasmid pJET1.2-capA-IC1 was designed to contain the primer sequence of capAF1/R1 of B. anthracis which was amplified simultaneously in the hexaplex PCR, for IAC product with a size of 1000 bp different from the six diagnosis amplicons of the hexaplex PCR in terms of size. The IAC concentration of 5 × 104 copies per sample was determined, and the components of the hexaplex PCR assay including IAC were adjusted to 0.64 μM concentration of capAF1/R1 primer set (Fig. 4). The IAC was added to the samples to control the DNA extraction procedure and verify PCR negative cases.

Fig. 4
figure 4

Evaluation of the detectability of the optimal multiplex PCR including IAC from genomic DNA from screening isolates. M, 100 bp DNA ladder (Thermo Scientific); NC, negative control (B. cereus + E. coli); PC, positive control (B. anthracis and/or Y. pestis DNA); lane 1, Y. pestis Y1 and B. anthracis B2; lane 2, B. anthracis B4; lane 3, Y. pestis Y1; lane 4, B. anthracis B3; lane 5, B. anthracis B5; lane 6, B. cereus B9; lane 7, Y. enterocolitica

Full size image

The developed hexaplex PCR assay was validated on 38 bacterial samples including the reference strains and screening isolates. Sensitivity and specificity were calculated for each target gene individually for the hexaplex PCR results of B. anthracis and Y. pestis. Out of 38 genomic DNA samples, 24 samples (B. anthracis 17JB, B1-B23) showed a PCR product of the vrrA gene amplification (222 bp) for the “B. cereus” group, and four samples (B. anthracis 17JB, B2, B3, B23) showed two PCR products of both 751 and 610 bp of the pagA gene and capA gene, respectively, which are specific to pathogenic B. anthracis. One sample (B5) showed a single product of 751 bp of the pagA gene present in the plasmid pXO1 in B. anthracis strain only and no amplification for the capA gene due to the absence of plasmid pXO2, indicating this is of a not fully virulent strain. One sample showed three bands of 103, 275, and 438 bp, which are specific to pathogenic Y. pestis (Y1). The rest samples did not show any band except IAC product in all examined samples (illustrated in Fig. 4). The hexaplex PCR assay had, thus, 100% specificity for Y. pestis.

Discussion

Rapid, reliable, and simultaneous identification of CDC category A bioterrorism agent plays a crucial role in disease prevention. In this study, we have developed and evaluated a novel hexaplex PCR assay for simultaneous detection of pathogenic B. anthracis and Y. pestis as well as discrimination with non-pathogenic strains in a single PCR tube. A multiplex PCR assay comprising a total of six primer pairs including vrrAF1/R1, pagAF1/R1, capAF1/R1, ypoF1/R1, plaF1/R1, and cafF1/R1 for B. anthracis and Y. pestis detection, respectively, was developed. Especially, the recombinant IAC, namely pJET1.2-capA-IC1, was designed with primer capAF1/R1 and was successfully co-amplified in the hexaplex PCR assay.

In B. anthracis, the pagA gene which encodes the protective antigen belonging to the virulent plasmid pXO1 is recommended by the World Health Organization (WHO) [19]. Besides, the second virulent plasmid of B. anthracis, pXO2, contains the capA, capB, and capC genes encoding poly-d-glutamic capsule, in which the capA gene was widely used for diagnosis of B. anthracis [7, 23, 24]. In Y. pestis, both the pla and caf1 genes were commonly used as the targets because of high estimated copy numbers (3 and 4 in a cell), respectively, which improved the sensitivity of PCR [12, 25, 26]. Additionally, we have used the ypo2088 gene encoding for a putative methyltransferase to distinguish Y. pestis from other Enterobacteriaceae [15, 27].

In the developed hexaplex PCR assay, the recombinant IAC was also designed to control both DNA extraction and PCR processing, allowing the evaluation of the true negative to false negative results. In our study, we used no additional adjuvant supplementary materials, which were recommended by various authors to improve amplification efficiency and specificity of the multiplex PCR [28, 29]. Our results revealed that the simultaneous hexaplex PCR assay with IAC was effectively optimized in this study.

However, the study has some limitations that will be the focus of our future work. Firstly, due to the lack of a clinical sample for both B. anthracis and Y. pestis, we could not evaluate the hexaplex PCR assay for diagnostics in clinical and environmental samples. Instead, we had to validate the mPCR assay in soil pseudo-samples contaminated with the cultured bacteria. Secondly, for the 6 non-pathogenic B. anthracis isolates (B1, B4, B6–8, B10), which do not harbor both the virulent plasmids, pXO1 and pXO2, which habored the vrrA gene as this gene is present in all species of “B. cereus” group including B. anthraics, B. cereus, and B. thuringensis [5], these isolates, therefore, were indistinguished from B. cereus used as reference strains. Thus, the use of both plasmids pXO1 and pXO2 in combination to accurately identify pathogenic B. anthracis was recommended [13, 30, 31]. Here, the three pathogenic B. anthracis were explored by the hexaplex PCR assay which was to the best of our knowledge the first report for cases of anthrax during two past decades in Vietnam. Altogether, the above multipex PCR assay has been proven the ability of accurate and simultaneous identification of pathogenic B. anthracis and Y. pestis as well as discrimination with non-pathogenic strains in nature.

Conclusions

The hexaplex PCR assay was developed for both pathogenic B. anthracis and Y. pestis to allow rapid and accurate detection of six target genes simultaneously including two genes on chromosomes and four genes on virulence plasmids of both bacteria, help distinguish strains that cause disease with non-pathogenic strains of B. anthracis as well as Y. pestis. Furthermore, we successfully developed the recombinant IAC for concurrent PCR amplification, making it as reliable proof for avoiding false negative results. The hexaplex PCR can, therefore, be a useful tool in screening for both pathogenic and non-pathogenic B. anthracis and Y. pestis, necessary in use for biodefense purposes and disease prevention.