Jiro Yasuda, Fifth Biology Section for Microbiology, First Department of Forensic Science, National Research Institute of Police Science, 6-3-1 Kashiwanoha, Kashiwa 277-0882, Japan. E-mail: email@example.com
Aims: To develop a rapid and simple system for detection of Bacillus anthracis using a loop-mediated isothermal amplification (LAMP) method and determine the suitability of LAMP for rapid identification of B. anthracis infection.
Methods and Results: A specific LAMP assay targeting unique gene sequences in the bacterial chromosome and two virulence plasmids, pXO1 and pXO2, was designed. With this assay, it was possible to detect more than 10 fg of bacterial DNA per reaction and obtain results within 30–40 min under isothermal conditions at 63°C. No cross-reactivity was observed among Bacillus cereus group and other Bacillus species. Furthermore, in tests using blood specimens from mice inoculated intranasally with B. anthracis spores, the sensitivity of the LAMP assay following DNA extraction methods using a Qiagen DNeasy kit or boiling protocol was examined. Samples prepared by both methods showed almost equivalent sensitivities in LAMP assay. The detection limit was 3·6 CFU per test.
Conclusions: The LAMP assay is a simple, rapid and sensitive method for detecting B. anthracis.
Significance and Impact of the Study: The LAMP assay combined with boiling extraction could be used as a simple diagnostic method for identification of B. anthracis infection.
Bacillus anthracis is a Gram-positive, aerobic, spore-forming bacterium that causes anthrax mainly in herbivorous animals, but can also cause acute disease in humans. Although infection in humans is rare in most countries, anthrax is still one of the most hazardous diseases in many areas where the soil is contaminated with B. anthracis spores (Dragon and Rennie 1995). Depending on the route of entry of the spores, infection can result in three forms of disease: cutaneous, gastrointestinal and pulmonary. Cutaneous anthrax, which occurs following exposure to diseased animals, is the most common form in humans. The symptoms of cutaneous anthrax can be identified relatively easily, and the prognosis of patients treated with antibiotics is usually good (Dixon et al. 1999). The pulmonary form caused by inhalation of aerosolized B. anthracis spores is very unusual naturally. However, its progress is extremely severe and fulminating, and it has a mortality rate approaching 100% if not treated early with antibiotics (Shafazand et al. 1999). The absence of characteristic symptoms at initial presentation makes it difficult to diagnose the disease in the early phase. To confirm B. anthracis infection as soon as possible, sensitive and reliable methods for detection of the organism in clinical specimens are required. In addition, since the anthrax attacks in the United States in 2001 (Jernigan et al. 2001), increased awareness about biological weapons and the use of anthrax as a bioterrorism agent have led to the necessity for improved and more rapid detection methods.
Identification of B. anthracis can be performed by conventional bacteriological methods, including Gram staining, capsule staining, colony morphology, gamma phage sensitivity tests (Turnbull 1999), and infection experiments with guinea pigs or mice (Welkos et al. 1986). However, these methods are time-consuming, labour-intensive, require large equipment and therefore are not appropriate for rapid detection systems (Begier et al. 2005). The nucleic acid-based amplification technique is the first choice for the rapid detection of B. anthracis. Conventional PCR (Makino et al. 1993; Ramisse et al. 1996) or real-time PCR (Makino et al. 2001; Bell et al. 2002; Oggioni et al. 2002) assays based on toxin-encoding genes on the virulence plasmid pXO1 and capsule biosynthesis protein-encoding genes on the capsule plasmid pXO2 have been developed and are in general use.
Recently, the loop-mediated isothermal amplification (LAMP) method was developed as an isothermal nucleic acid amplification technique (Notomi et al. 2000). This technique is based on the principle of strand displacing DNA synthesis by the Bst DNA polymerase with distinct six primers that recognize eight independent sites. DNA amplification is performed under isothermal conditions (60–65°C), thereby obviating the need for a thermal cycler. Moreover, the LAMP method generates an increase in turbidity in positive samples, allowing detection by visual judgement and by real-time monitoring based on the turbidity of the reaction mixture as well as agarose gel electrophoresis. The advantages of this method are its rapidity, simplicity and high sensitivity for detecting the target genes. LAMP assays have been used for rapid detection of several pathogenic viruses, bacteria and blood protozoa (Iwamoto et al. 2003; Hong et al. 2004; Poon et al. 2006; Kurosaki et al. 2007).
The pathogenicity of B. anthracis differs among strains, and both plasmids pXO1 and pXO2 are required for virulence of B. anthracis (Uchida et al. 1986; Tournier et al. 2007). pXO1 encodes tripartite protein exotoxins, protective antigen (pag), lethal factor (lef) and oedema factor (cya) (Okinaka et al. 1999), while pXO2 carries three genes (capB, capC and capA) essential for biosynthesis of the poly-γ-d-glutamic acid capsule (Makino et al. 1989). These genes have been used as markers to discriminate B. anthracis from other Bacillus species, especially closely related Bacillus cereus and Bacillus thuringiensis, which constitute the B. cereus group with B. anthracis. Although the genomes of the B. cereus group exhibit a high level of similarity (Radnedge et al. 2003; Okinaka et al. 2006), the unique chromosomal sequences of B. anthracis genes, rpoB (Qi et al. 2001), gyrA (Hurtle et al. 2004) and Ba813 (Ramisse et al. 1999) were also used for discriminating B. anthracis from other bacterial species by real-time PCR analysis. The sap gene, which encodes the S-layer constituent protein (Etienne-Toumelin et al. 1995), was described for identification of B. anthracis and used for the detection of B. anthracis in meat and tissue samples (Cheun et al. 2001).
The present study describes the development of a system for detection of B. anthracis by LAMP based on unique nucleotide sequences on pXO1, pXO2 and the chromosomal DNA.
Materials and methods
Bacterial strains, culture conditions and spore preparation
The bacterial strains used in this study are listed in Table 1. Bacillus anthracis and strains of other Bacillus species were grown on nutrient agar plates at 37°C overnight. Pasteur II derivative strains lacking pXO1 or both plasmids were prepared as described previously (Uchida et al. 1985, 1986). The other B. anthracis strains listed in Table 1 are isolates previously obtained from natural sources (Uchida et al. 1985; Makino and Cheun 2003) except for a vaccine strain 34-F2.
Table 1. Specificity of the loop-mediated isothermal amplification (LAMP) assay for Bacillus anthracis
*ATCC, American Type Culture Collection; NBRC, National Institute of Technology and Evaluation Biological Resource Center, Japan; IAM, Institute of Molecular and Cellular Biosciences, the University of Tokyo, Japan; HSCC, Higeta Shoyu Co. Ltd., Japan.
†Each assay was tested on the turbidimeter LA-200 platform against 1 ng of purified bacterial DNAs listed. Samples were examined in triplicate.
Bacillus anthracis spores were prepared as described previously (Fujinami et al. 2007). Briefly, the Pasteur II strain was grown on modified Schieffer’s sporulation agar for 3 days followed by harvesting in ice-cold water. The spore pellet was washed at least six times with ice-cold water, resuspended in PBS and stored at 4°C. Spore preparations were checked by microscopy to confirm the absence of vegetative cells.
Bacterial DNAs were extracted from cultured colonies and used as templates for PCR or LAMP. Individual bacterial colonies of each strain were harvested and suspended in PBS. The cell pellets were obtained by low speed centrifugation and resuspended with 180 μl of lysis buffer containing 20 mg ml−1 of lysozyme, 20 mmol l−1 Tris–HCl (pH 8·0), 2 mmol l−1 EDTA, 1·2% Triton-X100 and 0·2% SDS. After incubation at 36°C for 30 min, 20 μl of proteinase K (20 mg ml−1; Qiagen, Hilden, Germany) was added and incubated at 56°C overnight, followed by incubation at room temperature for 2 min with 28 units of RNase A (Qiagen). The cell lysates were subjected to phenol/chloroform/isoamyl alcohol (25 : 24 : 1) extraction followed by ethanol precipitation. Isolated DNA was resuspended in 50 μl of sterile water, and the DNA concentration was determined by measuring the optical density at 260 nm (DU-800; Beckman Coulter, Fullerton, CA, USA).
Bacillus anthracis-specific primers used for LAMP were designed based on the published sequences of pag (GenBank accession number AF306782), cap (accession no. M24150) and sap (accession no. Z36946) genes. Potential target regions were selected from the sequences aligned with other strains of B. anthracis, and LAMP primers were designed with the LAMP primer design support software program PrimerExplorer ver. 3 (Net Laboratory, Tokyo, Japan; http://primerexplorer.jp/e/). Six primers were used for each LAMP assay, including two outer primers (F3 and B3), two loop primers (LF and LB), a forward inner primer (FIP) and a reverse inner primer (BIP) (Nagamine et al. 2002). FIP consisted of F1c complementary to the F1 sequence, a TTTT spacer and the F2 sequence. BIP consisted of B1c complementary to the B1 sequence, a TTTT spacer and B2 sequence. The sequences of the primers are listed in Table 2.
Table 2. Oligonucleotide sequences of loop-mediated isothermal amplification (LAMP) primers
PCR primers for detection of the pag, cap and sap genes of B. anthracis were used as reported previously (Makino et al. 1989; Etienne-Toumelin et al. 1995; Price et al. 1999). Aliquots of template DNA and 5 pmol of each primer were added to a PCR mixture containing 0·26 U of KOD Dash DNA polymerase (Toyobo, Tokyo, Japan), 0·2 mmol l−1 dNTP and 1× reaction buffer. The final volume was adjusted to 25 μl with distilled water, and the reaction mixture was then amplified by PCR according to the profile: 1× 95°C for 5 min; 30× (95°C for 30 s, followed by 60°C for 30 s, 72°C for 30 s); 1× 72°C for 5 min using a thermal cycler (Mastercycler ep; Eppendorf, Hamburg, Germany). Amplified PCR products were electrophoresed in 2% agarose gels and stained with ethidium bromide for visualization on an imaging analyser (LAS-3000; FujiFilm, Tokyo, Japan). The samples showing DNA fragments of the correct sizes were considered positive.
LAMP reaction was performed with a Loopamp DNA amplification kit (Eiken Chemical, Tokyo, Japan) in accordance with the manufacturer’s protocol. Reaction mixtures containing 40 pmol of each primer FIP and BIP, 5 pmol of each of the outer primers F3 and B3, 20 pmol of each of the primers LF and LB, 12·5 μl of 2× reaction mix, 1·0 μl of Bst DNA polymerase and 5·0 μl of DNA sample in a final volume of 25 μl were incubated at 63°C for 60 min. A portion of each amplified product was analysed on a 3% agarose gel, followed by staining with ethidium bromide. For real-time monitoring of amplification by LAMP, the reaction mixtures were incubated and observed by spectrophotometric analysis using a real-time turbidimeter (LA-200; Teramecs, Kyoto, Japan). The time of positivity observed through real-time LAMP assay was determined as the time at which the turbidity reached the threshold value fixed at 0·1, which was double the average turbidity value of negative controls in several replicates.
Blood sample preparation
Specific pathogen-free female BALB/c mice were obtained from Sankyo Labo Service (Tokyo, Japan) and were 6-week old at the time of the experiments. Aliquots of 10 μl of PBS containing 2·2 × 107 CFU of B. anthracis Pasteur II (pXO1+/pXO2+) spores were administered into the nasal cavities of the mice. The inoculum dose was confirmed by viable count assay after plating on nutrient agar plates following infection. Three days after inoculation, mice were anaesthetized with diethyl ether, and approx. 300 μl of blood was collected by cardiac puncture. Viable counts in blood were determined by tenfold serial dilution in sterile PBS and plating on nutrient agar plates.
For DNA extraction, aliquots of 10 μl of heparinized blood were serially diluted in 50 μl of uninfected mouse blood, and samples of 20 μl of the diluted samples were treated using two commonly used extraction methods, using DNeasy Blood and Tissue kit (Qiagen), boiling method. DNA extraction using the Qiagen kit was carried out according to the manufacturer’s instructions for Gram-positive bacteria with minor modifications as follows. The samples (20 μl) were added to 180 μl of lysis buffer containing 20 mmol l−1 Tris–HCl (pH 8·0), 2 mmol l−1 EDTA, 1·2% Triton-X100 and 50 μg of lysostaphin (Sigma, St Louis, MO, USA), and incubated at 30°C for 30 min. This sample mixture was combined with 25 μl of proteinase K and 200 μl of AL buffer, followed by incubation at 55°C for 30 min. After incubation, 200 μl of ethanol was added, mixed by vortexing, loaded onto a spin column and washed according to the manufacturer’s instructions. DNA was eluted in 50 μl of AE buffer through the column.
Boiling extraction was carried out as described previously with a few modifications (Makino et al. 1993). Blood samples (20 μl) were diluted in 180 μl of distilled water and incubated at 95°C for 15 min. After centrifugation at 10 000 g for 10 min at 4°C, the supernatants were used for the following assays. Each DNA sample was sterilized by syringing through 0·22-μm filter (Millipore, Billerica, MA, USA). Infected animals were housed in cages within laminar flow safety enclosures in a BSL-3 facility. All animal experiments were performed in accordance with the guidelines for animal experimentation of the National Research Institute of Police Science.
Specificity and sensitivity of the LAMP assay for Bacillus anthracis
For specific amplification of the chromosomal DNA and the plasmids pXO1 and pXO2, which define the virulence of B. anthracis, a set of six primers were designed for each target (Table 2). The target genes for detection of chromosomal DNA, pXO1 and pXO2 were the sap, pag and cap genes respectively.
First, to examine whether the three newly designed primer sets can correctly amplify each target gene, we demonstrated the LAMP assay using these primers for B. anthracis Pasteur II strains harbouring both the virulence plasmids (pXO1+/pXO2+) and its derivatives lacking only the capsule plasmid (pXO1−/pXO2+), or both plasmids (pXO1−/pXO2−) (Fig. 1, upper panel). A successful LAMP reaction produced a characteristic ladder-like pattern on agarose gel electrophoresis, because the LAMP products had a structure consisting of several inverted repeats (Notomi et al. 2000). As shown in Fig. 1, LAMP products were detected according to the presence of each of the virulence plasmids using the primers for pag or cap gene. As expected, the primer set for sap amplified the DNA from all genotypes of the Pasteur II strain regardless of the presence or absence of virulence plasmids. No LAMP products were detected in reactions with the DNA from B. cereus, a species closely related to B. anthracis. The results of LAMP were completely consistent with those of conventional PCR methods (Fig. 1, lower panel).
To determine the sensitivity of LAMP, tenfold serial dilutions of bacterial DNA extracted from Pasteur II strain harbouring both plasmids, i.e., from 1 ng to 10 ag of DNA, were subjected to LAMP or PCR assay in a reaction volume of 25 μl. Amplification by LAMP with each primer set could detect more than 10 fg of bacterial DNA within 60 min (Fig. 2, upper panel), while amplification by PCR using primer sets specific for pag (amplicon size: 596 bp) and cap (amplicon size: 572 bp) genes was successful for more than 100 fg of bacterial DNA, and with primers for sap (amplicon size: 639 bp) for 1 pg of DNA (Fig. 2, lower panel). These observations indicated that the LAMP method was at least 10- or 100-fold more sensitive than the conventional PCR method.
Real-time monitoring of LAMP reaction
The same serial dilutions of the bacterial DNA were also applied to the real-time LAMP assay (Fig. 3). The reaction was performed at 63°C, and amplification was monitored for 60 min by measuring the turbidity caused by the reaction by-product, magnesium pyrophosphate, in the reaction tube (Mori et al. 2001). The threshold value was fixed at a turbidity of 0·1, and samples with turbidity above this threshold value were considered positive. The real-time LAMP assay with each of the three primer sets showed the limit of detection for the assay to be 10 fg DNA per test for the three different targets. The results indicated that real-time monitoring was as sensitive as agarose gel electrophoresis. In addition, the times of positivity observed through real-time monitoring were found to be <30 min for pag and cap genes (Fig. 3a,b), and <42 min for sap (Fig. 3c) at DNA concentrations of more than 10 fg.
The real-time assay was tested for specificity and cross-reactivity with a panel of 29 organisms (Table 1), including nine B. anthracis strains and 20 strains of 14 other Bacillus species. All nine B. anthracis strains were positive on testing with sap-specific primers. In addition, these strains were also tested for the virulence plasmid markers, pag and/or cap, and the results corresponded to the plasmid profile previously defined by nested PCR (Cheun et al. 2001). All of the other 20 Bacillus species, including the closely related B.cereus group, B. cereus and B. thuringiensis, gave negative results with all three primer sets.
Evaluation of the assay for blood samples from mice infected with Bacillus anthracis
As the LAMP methods were shown to be specific and sensitive for the detection of B. anthracis, the usefulness of the assay for diagnosis was examined. For this purpose, the LAMP assay was evaluated using blood samples from mice infected intranasally with B. anthracis Pasteur II (pXO1+/pXO2+) spores. Three days after infection, blood samples containing B. anthracis at 2·3 × 107 CFU ml−1 were collected from the mouse, and then the blood samples containing bacilli were serially diluted fivefold with blood from uninfected mice, yielding samples containing 2·3 × 107 to 2·9 × 102 CFU ml−1. DNAs were extracted from these diluted blood samples prepared by either boiling or using DNeasy kit, and then used for LAMP or PCR analyses.
First, the inhibitory effects of blood components for LAMP amplification were examined. As described in Materials and Methods, the blood from uninfected mice was diluted tenfold with distilled water and then boiled for 15 min. After centrifugation at 10 000 g for 10 min at 4°C, aliquots of 5 μl of the supernatants were applied to LAMP reactions. As shown in Table 3, addition of mouse blood to the LAMP reaction did not reduce the detection limit in any LAMP assay, although a slight delay in the time for positive reaction was observed in each LAMP assay for three targets. These observations indicated that the reaction sensitivity of LAMP assay was not significantly affected by components from boiled blood, at least up to 5 μl (corresponding to approx. 0·5 μl of whole blood). It was also confirmed that PCR was unaffected by the addition of 5 μl of the same blood samples (data not shown). Therefore, LAMP was compared with PCR using 5 μl of boiled blood samples.
Table 3. The inhibitory effect of boiled blood on loop-mediated isothermal amplification (LAMP)
*Samples were examined in duplicate and considered positive when all tube turbidities reached the threshold value. The average time for a positive reaction are indicated.
†LAMP reaction was carried out with (+) or without (−) 5 μl of blood sample treated, as described in Materials and Methods.
Tt, threshold time; ND, not detected.
Table 4 shows the sensitivity of LAMP or PCR for blood samples containing B. anthracis. The real-time LAMP assay was able to detect all three target genes for the blood samples containing B. anthracis at 7·2 × 103 CFU ml−1 or more. The detection limits of LAMP assay were almost the same in both samples prepared by boiling and with the DNeasy kit. The time required for detection at the detection limit was also almost equivalent for both samples prepared by boiling and with the DNeasy kit. On the other hand, PCR showed different sensitivities to the samples from boiling method or with the DNeasy kit. For the samples prepared with the DNeasy kit, the sensitivities of PCR for all three genes were equivalent to those of LAMP. However, for boiled samples, PCR was less sensitive than LAMP for detection of cap and sap genes, although PCR could detect the pag gene with the same sensitivity as LAMP. These results indicated that LAMP is a more efficient and suitable nucleic acid amplification method than PCR for simply prepared blood samples.
Table 4. Limits of loop-mediated isothermal amplification (LAMP) and PCR for detection of Bacillus anthracis in blood
CFU per test
Avg. Tt (min)†
CFU per test
*Samples were examined in duplicate and considered positive when all tube turbidities reached the threshold value.
†Average threshold time with duplicate results.
‡Applied volumes of the samples extracted using Qiagen DNeasy kit method to the reaction were 5 μl for LAMP and 1 μl for PCR.
7·2 × 103
7·2 × 103
7·2 × 103
1·8 × 105
7·2 × 103
3·6 × 104
7·2 × 103
3·6 × 104
7·2 × 103
3·6 × 104
7·2 × 103
3·6 × 104
Here, a LAMP assay was developed as a rapid and simple system for detection of B. anthracis, and this assay was further evaluated as a method for clinical diagnosis.
Using primers for the sap gene encoded on the chromosome, we could discriminate B. anthracis strains from strains of closely related Bacillus species, including B. cereus and B. thuringiensis (Table 1). LAMP with the primers for pag and cap genes could also precisely detect the presence of the virulence plasmids pXO1 and pXO2 respectively. The combination of tripartite primer sets is a more useful application not only for rapid and reliable identification of B. anthracis but also for determination of its virulence. Molecular typing techniques, including multilocus enzyme electrophoresis (MEE) and multilocus sequence typing (MLST), contributed to the suggestion that B. anthracis is a monophyletic clone derived from the B. cereus sensu lato group (Helgason et al. 2000; Priest et al. 2004). Recently, it has been reported that non-B. anthracis isolates identified as B. cereus (G9241) by genetic analysis of 16S ribosomal RNA possessed a circular plasmid with 99·6% similarity to the B. anthracis pXO1 plasmid (Hoffmaster et al. 2004), and the B. cereus isolates associated with fatal pneumonias were closely related to B. anthracis and harboured B. anthracis virulence genes (Hoffmaster et al. 2006). These closely related isolates are more likely to give discrepant results when assayed by standard microbiological and PCR methods based on the B. anthracis-specific gene sequences (Marston et al. 2006). Although our LAMP assay may be also not able to completely exclude the possibilities of false positive in these closely related Bacillus spp. isolates, the use of multiple targets could decrease the rate of the false positive results and would be applicable to discriminate these Bacillus isolates from B. anthracis strains.
The LAMP assay targeting the chromosomal DNA and virulence plasmids, pXO1 and pXO2, was 10- or 100-fold more sensitive than conventional PCR-based assay (Fig. 2). The real-time LAMP assay also showed high sensitivities similar to the LAMP assay with gel electrophoresis, and the detection limit was 10 fg of DNA per reaction for all targets (Fig. 3). In addition, the assay time needed to detect 1 ng of bacterial DNA was only 15 min (Fig. 3). The detection limit of the real-time PCR system developed previously to detect the same three genes as targeted in this study was 100 pg of bacterial DNA per reaction, and it took about 1 h to obtain a positive result from 1 ng of DNA (Ryu et al. 2003). Thus, the LAMP assay described here is a very rapid and sensitive tool for identification of B. anthracis. More sensitive real-time PCR assays have been reported recently (Hurtle et al. 2004; Christensen et al. 2006). Nevertheless, the LAMP assay still has advantages with regard to rapidity and simplicity.
Inhalational anthrax is known to be associated with systemic bacteraemia and/or sepsis in human and animal cases. Sensitive detection of bacteria in blood is quite important for early diagnosis and treatment of anthrax. To evaluate the usefulness of LAMP for clinical specimens, such as blood, blood samples were prepared from mice inoculated intranasally with a lethal dose of B. anthracis spores, and the sensitivity of the LAMP assay were then examined using these samples. For DNA extraction from blood samples, two methods were examined, i.e. boiling and using Qiagen DNeasy kit, because these are simple, easy and rapid. The time needed to prepare a DNA sample from a whole blood sample was approx. 30 min for the boiling method and 1·5 h for the DNeasy kit method. It has been reported that blood components interfere with PCR (Al-Soud et al. 2000; Al-Soud and Radstrom 2001). Therefore, it was examined whether the blood components have any inhibitory effect on LAMP reaction, as DNA samples extracted by the boiling method should carry blood components into the reaction. It was confirmed that that volumes of up to 5·0 μl of blood samples prepared by the boiling method did not have significant inhibitory effect on LAMP reaction (Table 3). When LAMP and PCR were compared using the samples prepared using the DNeasy kit method, the detection limits of LAMP were indistinguishable from those of PCR for all three target genes (Table 4). However, when samples prepared by the boiling method were used, LAMP was much more sensitive than PCR in the detection of cap and sap genes. Moreover, LAMPs using the boiled samples were slightly more sensitive than those using the samples prepared using the DNeasy kit method for the detection of target genes. These results indicated that the LAMP assay combined with boiling extraction is a simple, rapid and sensitive method and therefore would be an efficient on-site diagnostic tool for direct detection of B. anthracis in whole blood without requiring culture growth or precise DNA extraction.
LAMP has advantages over other nucleic acid amplification techniques in its simple operation and rapid detection. The only equipment required for the LAMP reaction is a water bath or heat block that furnishes a constant temperature of 63°C. The assay does not require the use of sophisticated equipment or highly skilled personnel. This makes it potentially useful for on-site use in clinical diagnosis and field surveillance or in cases of deliberate release during bioterrorism attacks.
This work was supported by a grant from the Japan Science and Technology Agency (JST).