Development of a single-tube loop-mediated isothermal amplification assay for detection of four pathogens of bacterial meningitis

Authors


Correspondence: Kenji Hirayama, Department of Immunogenetics, Institute of Tropical Medicine (NEKKEN), Nagasaki University, 1-12-4 Sakamoto, Nagasaki 852-8523, Japan. Tel.: +81 95 819 7805; fax: +81 95 819 7821; e-mail: hiraken@nagasaki-u.ac.jp

Abstract

Several loop-mediated isothermal amplification (LAMP) assays have been developed to detect common causative pathogens of bacterial meningitis (BM). However, no LAMP assay is reported to detect Streptococcus agalactiae and Streptococcus suis, which are also among common pathogens of BM. Moreover, it is laborious and expensive by performing multiple reactions for each sample to detect bacterial pathogen. Thus, we aimed to design and develop a single-tube LAMP assay capable of detecting multiple bacterial species, based on the nucleotide sequences of the 16S rRNA genes of the bacteria. The nucleotide sequences of the 16S rRNA genes of main pathogens involved in BM were aligned to identify conserved regions, which were further used to design broad range specific LAMP assay primers. We successfully designed a set of broad range specific LAMP assay primers for simultaneous detection of four species including Staphylococcus aureus, Streptococcus pneumoniae, S. suis and S. agalactiae. The broad range LAMP assay was highly specific without cross-reactivity with other bacteria including Haemophilus influenzae, Neisseria meningitidis and Escherichia coli. The sensitivity of our LAMP assay was 100–1000 times higher compared with the conventional PCR assay. The bacterial species could be identified after digestion of the LAMP products with restriction endonuclease DdeI and HaeIII.

Introduction

Rapid diagnosis of bacterial meningitis (BM) is essential as successful disease outcome is dependent on immediate antibiotic therapy (Saez-Llorens & McCracken, 2003; Zimmerli, 2005). However, accurate and rapid identification of BM is challenging for clinicians as its symptom and laboratory test are often similar and overlapping with those of aseptic meningitis. Conventional diagnosis of BM relies on the detection of bacteria in cerebrospinal fluid and/or blood by Gram staining, latex agglutination and culturing. However, Gram staining and latex agglutination tests are low in sensitivity (Kennedy et al., 2007), while culturing takes few days. Furthermore, antimicrobial therapy prior to lumbar puncture often reduces the frequency of positive cultures from the CSF and blood (Pandit et al., 2005).

PCR assays have recently been developed to detect several bacterial pathogens of BM. These assays have been widely used in clinical practice and proved to have both high sensitivity and specificity. However, the PCR method requires expensive instrument, experienced technician and few-hour performance. To overcome the limitations of current PCR, the loop-mediated isothermal amplification (LAMP) assay has been invented as an accurate, rapid and cost-effective method, which amplifies the target nucleic acid under isothermal conditions, usually between 56 and 65 °C (Notomi et al., 2000). The amplified product of LAMP assay can be detected in < 1 h and identified by agarose gel electrophoresis, simple visual inspection, real-time monitoring of turbidity or visual colour change using fluorescent dye. Importantly, the assay can be performed at bedside and in rural areas using only a water bath (Tomita et al., 2008). Several LAMP assays have been developed to detect common causative pathogens of BM such as Streptococcus pneumoniae, Haemophilus influenzae, Neisseria meningitidis, Escherichia coli and Staphylococcus aureus (Seki et al., 2005; Yamazaki et al., 2008; Hanaki et al., 2011; Kim et al. 2011; McKenna et al. 2011). However, no LAMP assay has been reported to detect Streptococcus agalactiae and Streptococcus suis, which are two of the most common pathogens of BM in some countries (Mai et al., 2008; Chiba et al., 2009). Moreover, it is laborious and expensive by performing multiple reactions for each sample to detect bacterial pathogen. Thus, we aimed to design and develop a LAMP assay capable of detecting multiple bacterial species based on the nucleotide sequences of the 16S rRNA genes of the bacteria.

Materials and methods

Design of LAMP assay primers for seven common bacteria in BM

The broad range LAMP primers were designed to be specific for eubacterial 16S rRNA-specific gene. This gene was chosen because of its highly conserved regions among species and has been widely used as a target for broad range PCR method (Gray et al., 1984; Lane et al., 1985). The partial nucleotide sequences of the 16S rRNA genes of S. aureus (GenBank FJ907240.1), S. pneumoniae (Z22807), S. suis (Z22776.1), S. agalactiae (Z22808), N. meningitidis (Z22806), H. influenzae (Z22809.1) and E. coli (AY513502.1) were retrieved from the GenBank database and were aligned to identify potential target regions using multialin software (Corpet, 1988). Several conserved regions were chosen for designing of LAMP primer set using the LAMP primer design software Primer Explorer version 4 (Eiken Chemical Co., Ltd, Tokyo, Japan). A set of four primers including two outer primers (forward primer F3 and backward primer B3) and two inner primers [forward inner primer (FIP) and backward inner primer (BIP)] that identified six distinct regions on the potential target sequence was designed. This study was approved by the institutional ethical review committees of the Institute of Tropical Medicine, Nagasaki University.

Bacterial strains and samples preparation

Serotypes 3 and 10 of S. pneumoniae were isolated from upper respiratory tract in Vietnamese patients. Two strains (8-01 and 8-02) of S. suis serotype 2, E. coli, S. aureus and S. agalactiae were also isolated from Vietnamese patients. In addition, H. influenzae and N. meningitidis were isolated from Japanese patients. The S. pneumoniae was cultured on rabbit blood Muller Hinton agar, while other bacteria were grown on rabbit blood brain heart infusion agar. Grown bacteria were harvested and suspended in normal saline. The cells were pelleted, suspended in TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0) and serially diluted with TE buffer ranging from 108 down to 100 colonies of bacteria mL−1 (CFU mL−1). DNA was released from the bacteria by boiling for 20 min followed by centrifugation at 10 000 g for 10 min. The supernatant was used as the DNA template.

Optimization of LAMP reaction

The LAMP reaction was carried out in a 25-μL reaction mixture with a Loopamp DNA amplification kit (Eiken Chemical Co., Ltd) as described in our previous work (Kubo et al., 2010). The reaction mixture contained 40 pmol (1 μL) each of FIP and BIP, 5 pmol (1 μL) each of F3 and B3 and 20 pmol (1 μL) each of Loop F and Loop B. LAMP reaction was performed at several different temperatures ranging from 55 to 68 °C in 90 min using LA-320C Loopamp real-time turbidimeter (Teramecs, Japan). The best condition for LAMP procedure was at 63 °C and in 60 min. Therefore, all of mixtures were incubated at 63 °C for 90 min, followed by heating at 80 °C for 5 min to inactivate the reaction. Two microlitre of the extracted DNA was used as the template in each reaction mixture. A negative control (a reaction mixture with distilled water instead of DNA template) and a positive control (a confirmed positive sample) were included in each run. Precautions were taken to prevent cross-contaminations.

Analysis of LAMP product

The LAMP product was analysed by three methods including a real-time turbidimeter, agarose gel analysis and naked eye visualization. The LA-320C Loopamp real-time turbidimeter (Teramecs) was used to monitor the LAMP reaction based on the turbidity of magnesium pyrophosphate at 405 nm, a byproduct of the reaction. The turbidity threshold value for a positive sample was fixed at 0.1, and samples above this threshold value were considered as positive. After amplification, 2 μL of the LAMP product was further separated by 2% agarose gel electrophoresis, which was stained with ethidium bromide and visualized under UV light. In addition, 1 μL of SYBR Green I (Invitrogen) was added to the remained LAMP product, a change from orange to fluorescent green colour was considered as positive. To further distinguish bacterial species, 2 μL of the LAMP product was digested with 10 U of DdeI or HaeIII at 37 °C for 90 min. The digested LAMP product was analysed by 2% agarose gel electrophoresis as described above.

Conventional PCR assay

A conventional PCR was also carried out with the universal primer set targeting 16S rRNA genes to compare the sensitivity of the LAMP assay. The paired primers were 5′-CCAGCAGCCGCGGTAATACG-3′ and 5′-ATCGG(C/T)TACCTTGTTACGACTTC-3′ (Lu et al., 2000). Twenty-five microlitre of PCR assay contained 2 μL of DNA template, 1 μL of each primer, 2 mM MgCl2, 0.2 mM dNTPs, 2.5 μL of 10 × buffer and 1.25 U Taq HS DNA polymerase (Takara Bio, Shiga, Japan). The reactions were amplified as follows: initial activation of one cycle at temperature 94 °C for 10 min and then followed by 35 cycles at 94 °C for 30 s, 55 °C for 50 s and 72 °C for 2 min. The final extension step was carried out at 72 °C for 10 min. Amplified products were then detected by ethidium bromide staining after 2% agarose gel electrophoresis.

Results and discussion

Design of broad range LAMP assay primers

We aimed to develop a LAMP assay capable of detecting many bacterial species (multispecies LAMP assay) for diagnosis of BM. The bacterial species used in this study were S. aureus, S. pneumoniae, S. suis, S. agalactiae, N. meningitidis, H. influenzae and E. coli. The nucleotide sequences of the 16S rRNA genes of these bacteria were retrieved and aligned to design broad range specific LAMP assay primers using explorer version 4 (Eiken Chemical Co., Ltd). We could not design any broad range specific LAMP assay primers for all the seven bacteria due to high level of variation in the target 16S rRNA gene among species (Fig. 1a). Next, we repeatedly aligned the target gene and designed broad range specific LAMP assay primers each time removing each species. However, no broad range specific LAMP assay primers were found for the detection of any set of more than four bacterial species. Finally, we successfully designed a set of broad range specific LAMP assay primers for the detection of four species including S. aureus, S. pneumoniae, S. suis and S. agalactiae (Fig. 1b). The name, positions and nucleotide sequences of all four primers are shown in Fig. 1b and Table 1. The DNA sequence alignment of 16S rRNA gene of these four species indicated a low variation among these species.

Figure 1.

Alignment of nucleotide sequences of the 16S rRNA genes of bacteria. The target gene of seven common bacteria of BM including Neisseria meningitidis, Haemophilus influenzae, Streptococcus pneumoniae, Streptococcus agalactiae, Escherichia coli, Staphylococcus aureus and Streptococcus suis (a) and four bacteria including S. pneumoniae, S. agalactiae, S. aureus and S. suis (b) were aligned to identify the highly conserved regions, which were used for LAMP primers design. Consensus shown similar nucleotides in red colour was used to design a universal set of LAMP primers for simultaneous detection of multiple bacteria.

Table 1. LAMP primer sequences for simultaneous detection of Streptococcus pneumoniae, Streptococcus agalactiae, Staphylococcus aureus and Streptococcus suis
PrimersSequence 5′–3′
FIPCGCTTTCG(C/A)(A/G)C(A/C)TCAGCGTCATGGAGGAACACC(A/G)GTGGC
BIPCACGC(C/T)GTAAACGATGAGTGCTAGGCGGAGTGCTTAATGC
F3CATGTGTAGCGGTGAAATGC
B3TCAACCTTGCGGTCGTACT

Sensitivity and specificity of LAMP assay

The sensitivities of the broad range LAMP assay were performed by running 10-fold serial dilutions of target bacteria (from 107 to 100 CFU mL−1). The detection limit was 100 CFU mL−1 of S. pneumoniae by both real-time turbidimeter and electrophoresis of LAMP products, and 10 000 CFU mL−1 by conventional PCR method (Fig. 2). Similarly, the broad range LAMP assay detected S. suis, S. agalactiae and S. aureus at 100 CFU mL−1, while conventional PCR assay only detected these bacteria with more than 104 CFU mL−1 (Table 2). The results of all the positive samples detected by the LAMP assay were achieved within 60 min.

Figure 2.

Sensitivity of broad range LAMP assay for the detection of Streptococcus pneumoniae. LAMP reactions detected by real-time turbidity (a) and electrophoresis of LAMP products and conventional PCR products (b). The assay was performed in 10-fold serial dilutions (from 107 to 100 CFU mL−1). Streptococcus pneumoniae with more than 100 CFU mL−1 was detected by both real-time turbidimeter and electrophoresis. The conventional PCR only detected S. pneumoniae with more than 104 CFU mL−1.

Table 2. Sensitivities of LAMP and conventional PCR assays
BacteriaThe limit of detection (CFU mL−1)
LAMP assayPCR assay
S. pneumoniae 102104
S. suis 102104
S. agalactiae 102105
S. aureus 102104

The specificity of the LAMP assay was evaluated by cross-reactivity test using DNA extracted from N. meningitidis, H. influenzae and E. coli. There were ladder-like products amplified from S. pneumoniae, S. suis, S. agalactiae and S. aureus but not from N. meningitidis, H. influenzae and E. coli cultures (Fig. 3), suggesting that the broad LAMP assay was specific for S. pneumoniae, S. suis, S. agalactiae and S. aureus.

Figure 3.

Specificity of broad range LAMP assay. Two microlitre of the DNA template (105 CFU mL−1) extracted from Staphylococcus aureus (lane 1), Streptococcus pneumoniae serotype 3 (lane 2), S. pneumoniae serotype 10 (lane 3), Streptococcus suis serotype 2-8-01 (lane 4), S. suis serotype 2-8-02 (lane 5), Streptococcus agalactiae (lane 6), Haemophilus influenzae (lane 7), Escherichia coli (lane 8) and Neisseria meningitidis (lane 9). Lane P, Positive control; lane N, Negative control; and lane M, DNA size markers.

Visual detection of the LAMP products

LAMP products were further explored by visual inspection based on the intercalation of fluorescent dye SYBR Green I into amplified DNA. As shown in Fig. 4, the product of positive reaction became visible under ultraviolet lamp and was green colour under naked eye, while the negative product was not seen under ultraviolet lamp and remained orange colour under day light.

Figure 4.

Visual detection of LAMP products. Representative visual inspection of Streptococcus pneumoniae by fluorescence under ultraviolet lamp (left Streptococcus) and day light (right). (N), negative control without DNA template; (P), positive reaction in the presence of S. pneumoniae DNA template.

Identification of bacterial species

To identify bacterial species, the LAMP product was digested with specific restriction enzyme and analysed by gel electrophoresis. After digested with DdeI, all LAMP products were digested into several fragments. Staphylococcus aureus gave five bands at 55, 150, 197, 230 and 263 bp (Fig. 5a, lane 1). Streptococcus pneumoniae produced three bands at 55, 150 and 200 bp (Fig. 5a, lane 3). Streptococcus agalactiae (lane 2) and S. suis (lane 4) gave similar pattern. Thus, the LAMP products of S. agalactiae and S. suis were further digested with HaeIII. The result showed that S. agalactiae was digested into four bands at 70, 216, 254 và 292 bp (Fig. 5b, lane 6), while S. suis was not digested by HaeIII (Fig. 5b, lane 5).

Figure 5.

The digested pattern of the LAMP products with restriction enzymes using 2% agarose electrophoresis. (a) DdeI digestion patterns of the LAMP products from Staphylococcus aureus (lane 1), Streptococcus agalactiae (lane 2), Streptococcus pneumoniae (lane 3) and Streptococcus suis (lane 4). (b) HaeIII digestion patterns of the of the LAMP products from S. suis (lane 5) and S. agalactiae (lane 6). Lane M is 100-bp ladder size markers.

To our knowledge, this is the first study that developed a broad range LAMP assay for simultaneous detection of more than four different bacterial species. The sensitivity of our LAMP assay was 100–1000 times higher compared with the conventional PCR assay. The bacterial species could be distinguished among S. pneumoniae, S. suis, S. agalactiae and S. aureus based on the digested pattern of the LAMP products with restriction enzymes of DdeI and HaeIII. In addition, our method has several advantages over the current diagnostic methods. Firstly, the method is rapid (c. 1 h) as compared with the real-time PCR method which requires 6 h to run (Nadkarni et al., 2002). Secondly, the LAMP method does not require expensive fluorimeter and fluorogenic primers and probes. Thirdly, the assay is simple and does not require highly experienced technician. More importantly, the assay can be performed in a water bath at bedside or in rural areas. These advantages suggested that our broad range LAMP assay would improve the early diagnosis and treatment of BM, helping to reduce morbidity and mortality. Furthermore, the assay could detect bacterial species, helping to select an appropriate antibiotic therapy.

One limitation of our LAMP assay was that only four species could be detected. A single-tube LAMP assay for the detection of more than four species is under development using a mixture current broad range LAMP primers and specific LAMP primers of other bacteria species. Additional clinical studies are also required to validate this new assay.

Conclusions

Four common pathogen of BM including S. pneumoniae, S. suis, S. agalactiae and S. aureus could be simultaneously detected using a broad range LAMP assay in single tube in < 1 h. The assay is highly sensitive, rapid and simple and can be performed at bedside in healthcare facilities.

Acknowledgements

We thank Dr Toru Kubo, from Department of Virology, Institute of Tropical Medicine, Nagasaki University, Nagasaki, Japan, for his technical advice. The authors declare no competing interests of the manuscript due to commercial or other affiliations. This study was supported in part by Japan Initiative for Global Research Network on Infectious Diseases (J-GRID) for K.H.

Authors' contributions

N.T.H and L.T.T.H. contributed equally to this work.

Ancillary