Rapid, Sensitive Detection of Bartonella quintana by Loop-Mediated Isothermal Amplification of the groEL Gene

Trench fever, caused by Bartonella quintana, is recognized as a re-emerging and neglected disease. Rapid and sensitive detection approaches are urgently required to monitor and help control B. quintana infections. Here, loop-mediated isothermal amplification (LAMP), which amplifies target DNA at a fixed temperature with high sensitivity, specificity and rapidity, was employed to detect B. quintana. Thirty-six strains, including 10 B. quintana, 13 other Bartonella spp., and 13 other common pathogens, were applied to verify and evaluate the LAMP assay. The specificity of the LAMP assay was 100%, and the limit of detection was 125 fg/reaction. The LAMP assay was compared with qPCR in the examination of 100 rhesus and 20 rhesus-feeder blood samples; the diagnostic accuracy was found to be 100% when LAMP was compared to qPCR, but the LAMP assay was significantly more sensitive (p < 0.05). Thus, LAMP methodology is a useful for diagnosis of trench fever in humans and primates, especially in low-resource settings, because of its rapid, sensitive detection that does not require sophisticated equipment.


Introduction
Bartonella are fastidious Gram-negative bacteria. They are transmitted by arthropods, for example lice and fleas are the vectors of B. quintana and B. henselae, respectively, to humans. They can cause several human diseases [1,2]. All Bartonella species are α-proteobacteria. Among them, B. quintana is the causative agent of trench fever and regarded as a re-emerging pathogen which infects humans and other primates [3]. Clinical manifestations include bacillary angiomatosis, chronic lymphadenopathy, endocarditis and trench fever [4][5][6]. Humans and primates are the major B. quintana reservoir and the human body louse has been considered the principal vector [7][8][9]. However, recently, this bacterium has been detected in specimens collected from cat fleas [10,11] and other arthropod families such as bed bugs [12,13], suggesting that a range of insects may act as vectors in the spread of trench fever. Human migration, habitat destruction, and changes in weather patterns or host dynamics increase the in weather patterns or host dynamics increase the potential threat of sporadic and occasional epidemics of trench fever [14]. Hence, there is a need for rapid and specific methods to identify B. quintana and differentiate it from other Bartonella species to aid both diagnosis and treatment.
Diagnosis of trench fever remains challenging. Conventional methods for the isolation and identification of B. quintana require up to 4 weeks before they can be considered negative [15], which has obvious disadvantages in the clinical setting. Serological tests such as IFA (Indirect Immunofluoreseent, Assay, IFA) or immunoblotting are proven and renowned methods for diagnosis Bartonella infections. However, require paired samples from the acute and recovery phases and may not be useful for the diagnosis of acute disease [16]. PCR amplification of DNA is sensitive and specific, but requires sophisticated apparatus, which may not be available in in resource-poor settings, so this approach is impractical for diagnosis of B. quintana infection in many areas [17].
Loop-mediated isothermal amplification (LAMP) is a nucleic acid amplification method that can amplify up to 10 9 copies of a DNA target in isothermal conditions (60-65 °C) in 1 h, and the results can be observed by a visual assessment of turbidity [18]. This assay has not only been applied to the detection of other bacterial species [19,20], but also to the detection of other Bartonella species other than B. quintana [21]. It is simpler than PCR-based methods and requires less equipment [22]. Here, we developed a LAMP assay targeting the molecular chaperone gene groEL, a member of the heat shock regulon, to detect B. quintana, and evaluated the diagnostic specificity and sensitivity of the assay.

Confirmation and Detection of B. quintana Loop-Mediated Isothermal Amplification (LAMP) Reaction Products
Amplification reactions were performed in the presence or absence of B. quintana genomic DNA to test the B. quintana-LAMP assay. Positive amplification was indicated by a color change from light gray to green, while the negative controls remained light gray ( Figure 1A). After 2.5% agarose gel electrophoresis, positive reactions showed a ladder-liker pattern, but negative controls did not ( Figure 1B).

The Optimal Temperature for the B. quintana LAMP Assay
The optimal temperature for the B. quintana LAMP reaction was determined using the reference strain B. quintana Toulouse as a positive control with 0.5 pg genomic DNA per reaction. The LAMP reactions were carried out at 60-67 °C and monitored by real-time turbidity measurement.

The Optimal Temperature for the B. quintana LAMP Assay
The optimal temperature for the B. quintana LAMP reaction was determined using the reference strain B. quintana Toulouse as a positive control with 0.5 pg genomic DNA per reaction. The LAMP reactions were carried out at 60-67 • C and monitored by real-time turbidity measurement. Figure 2 shows typical kinetics, and Figure 3 shows agarose gel electrophoresis of the reaction products. A temperature of 63 • C was chosen as optimal for the LAMP reaction and used for the remainder of this study.
shows typical kinetics, and Figure 3 shows agarose gel electrophoresis of the reaction products. A temperature of 63 °C was chosen as optimal for the LAMP reaction and used for the remainder of this study.   shows typical kinetics, and Figure 3 shows agarose gel electrophoresis of the reaction products. A temperature of 63 °C was chosen as optimal for the LAMP reaction and used for the remainder of this study.

Specificity of the B. quintana LAMP Assay
The assay specificity was determined with DNA templates from 23 members of the genus Bartonella (B. quintana 10 strains, other Bartonella spp. 13 strains) and 13 other common pathogenic bacteria. Reactions containing B. quintana genomic DNA produced a color change within 1 h, and a specific ladder of multiple bands observed after gel electrophoresis. No color change or bands on agarose gels were observed for the 23 non-B. quintana templates ( Figure 4). Thus, the LAMP assay was highly specific for screening B. quintana.

Specificity of the B. quintana LAMP Assay
The assay specificity was determined with DNA templates from 23 members of the genus Bartonella (B. quintana 10 strains, other Bartonella spp. 13 strains) and 13 other common pathogenic bacteria. Reactions containing B. quintana genomic DNA produced a color change within 1 h, and a specific ladder of multiple bands observed after gel electrophoresis. No color change or bands on agarose gels were observed for the 23 non-B. quintana templates ( Figure 4). Thus, the LAMP assay was highly specific for screening B. quintana.

Sensitivity of the B. quintana LAMP Assay
The LAMP assay was nearly fourfold more sensitive than qPCR for the detection of B. quintana groEL ( Table 1), with detection limits of 125 and 500 fg DNA per reaction, respectively ( Figure 5). B. quintana LAMP amplifications were monitored by real-time turbidity ( Figure 5A) and agarose gel electrophoresis ( Figure 5B,C). The LAMP reaction required incubation periods of 12, 14, 16, 17 and 18 min at genomic DNA levels of 5 ng, 50 pg, 500 fg, 250 fg and 125 fg per reaction, respectively ( Figure 5A).

Sensitivity of the B. quintana LAMP Assay
The LAMP assay was nearly fourfold more sensitive than qPCR for the detection of B. quintana groEL (Table 1), with detection limits of 125 and 500 fg DNA per reaction, respectively ( Figure 5). B. quintana LAMP amplifications were monitored by real-time turbidity ( Figure 5A) and agarose gel electrophoresis ( Figure 5B,C). The LAMP reaction required incubation periods of 12, 14, 16, 17 and 18 min at genomic DNA levels of 5 ng, 50 pg, 500 fg, 250 fg and 125 fg per reaction, respectively ( Figure 5A).

Evaluation of Practical Application of the B. quintana LAMP Assay
When testing 100 rhesus blood samples, we detected 22 positive results for B. quintana by the LAMP assay and eight by qPCR. The sensitivity of the LAMP assay was significantly higher than that of the qPCR (p < 0.05) ( Table 2). Similarly, we tested 20 rhesus-feeder blood samples using these assays; the total positive rates were 20.0% for the LAMP assay (4/20) and 5.0% for the real-time PCR (1/20; p < 0.05) ( Table 2).

Evaluation of Practical Application of the B. quintana LAMP Assay
When testing 100 rhesus blood samples, we detected 22 positive results for B. quintana by the LAMP assay and eight by qPCR. The sensitivity of the LAMP assay was significantly higher than that of the qPCR (p < 0.05) ( Table 2). Similarly, we tested 20 rhesus-feeder blood samples using these assays; the total positive rates were 20.0% for the LAMP assay (4/20) and 5.0% for the real-time PCR (1/20; p < 0.05) ( Table 2).

Discussion
A novel LAMP assay to detect B. quintana by targeting the groEL gene was developed in this study. The limit of detection of the LAMP assay was 125 fg DNA/reaction, which was fourfold more sensitive than qPCR. No non-specific amplification or cross-reaction was observed when testing a panel of closely related bacteria (other Bartonella species) and common pathogens. In practical tests of rhesus blood samples and rhesus-feeder blood samples, the sensitivity of the LAMP assay was significantly higher (p < 0.05) than that of qPCR.
Trench fever caused by B. quintana is considered an important re-emerging infectious disease [23]. Once infected, patients suffer headache, recurrent fever, and pretibial pain. Due to persistent bacteremia [24], several major epidemics of trench fever occurred among soldiers in Europe during World Wars I and II. More recently, the disease has occurred occasionally in urban areas in Europe and the USA, mainly among the homeless, drug-addicts, and HIV-positive patients [4]. Thus, sensitive and accurate detection methods are required to monitor and study B. quintana.
Bacterial culture is the preferred method for identification of Bartonella infections, but is slow, labor-intensive and often poorly effective. PCR-based methods are a sensitive and selective approach to the detection of Bartonella and have been used to target genes including gltA, rpoB, the 16S-23S rRNA ITS, and ftsZ [25,26]. However, these techniques require a high-precision thermocycler, limiting their use in areas with basic clinical facilities, for example in rural endemic and impoverished areas. As an alternative, the LAMP method is rapid and simple to perform, requiring only a water bath or heating block for amplification. The test reaction can be performed within 18 min in isothermal conditions (in the present case 63 • C), enabling rapid molecular diagnosis [22,27]. Recent studies have shown that LAMP has high sensitivity and specificity compared to conventional, nested or qPCR for the detection of several intracellular bacteria, including Coxiella burnetii and Orientiatsu tsugamushi [28,29] and in distinguishing intraerythrocytic protozoan parasites, such as Plasmodium spp. and Babesia spp. [30,31]. The present study is the first report of a LAMP assay to identify B. quintana, and can be used in the field, clinic and veterinary laboratories in surveillance and diagnosis of trench fever.

Ethics
The protocol of this study was reviewed and approved by the Ethics Committee of the China Institute for Communicable Disease Control and Prevention, based on the medical research regulations of the Ministry of Health (Approval No. ICDC-2016003). All experimental procedures conformed to institutional guidelines for the care and use of laboratory animals as described by the China CDC, and to the National Institutes of Health Guide for Care and Use of Laboratory Animals (Publication No 85-23, revised 1985). Animal blood sampling was conducted with the consent of the animals' owners.

Bacterial Strains and Culture Conditions
Thirty-six bacterial strains were used in this study, including 10 B. quintana, 13 other Bartonella species, and 13 non-Bartonella species (Table 3). All Bartonella strains were cultured on Tryptic Soy Agar (TSA) containing 10% sheep red blood cells using standard methods [32]. Other strains were cultured on Columbia agar containing 5% sheep blood at 37 • C for 24-48 h in an atmosphere containing 5% CO 2 . B. quintana str. Toulouse was used as the positive control to determine the optimal conditions for the LAMP assay and to establish baseline sensitivity.

Genomic DNA Extraction
Bacterial genomic DNA was extracted from pure cultures using QIAamp DNA minikits (Qiagen, Hilden, Germany). The non-anticoagulated blood of rhesus macaque (2 mL) was collected, and the serum was separated by centrifugation at 3000× g for 10 min. Blood clots were used to extract DNA using the QIAamp DNA Mini Kit. All of the DNA samples from rhesus blood were tested using the LAMP assay, using 1 µL of the extracted DNA as the template. To verify the LAMP assay, a qPCR assay based on the groEL gene was used. The two methods were compared using the χ 2 test and Fisher's exact test. p < 0.05 was considered significant.

Design of LAMP Primers
Using the groEL sequence from B. quintana str. Toulouse (GenBank accession number: BX897700.1) and Primer Explorer v.4 software (http://primerexplorer.jp/), >100 LAMP primer sets were designed. A set of six specific primers was selected for LAMP to target eight distinct regions in groEL. Table 4 and Figure 6 show details of primer design, sequences, locations and target sequences.   Mixtures lacking DNA template were used as negative controls. After amplification, the positive LAMP products could be observed directly by the color change of the FD. Moreover, LAMP products were examined by 2.5% agarose gel electrophoresis. To determine the optimum reaction temperature, the LAMP amplification was carried out for 60 min at constant temperatures from 60 to 67 °C in 1 °C intervals, with final incubation for 5 min at 85 °C to terminate the reaction.

Specificity and Sensitivity of the B. quintana LAMP Assay
To assess specificity, LAMP reactions were carried out as described above using DNA templates from 23 Bartonella species and 13 non-Bartonella species (Table 3). Each sample was examined twice, independently. To determine the analytical sensitivity, genomic DNA from B. quintana str. Toulouse was serially diluted (to 5 ng, 50 pg, 500 fg, 250 fg, 125 fg, 62.5 fg, 31.25 fg, and 15.625 fg, respectively). The limits of detection of the LAMP and qPCR assays were ascertained from the lowest amount of template DNA that could be detected. Table 4 lists the primers and probe for the qPCR assay. qPCR amplification was performed in a 20-µL reaction containing 0.3 mM each primer, 0.2 mM probe, 10 µL PromegaGoTaq ® qPCR Master Mix, and 1 µL of DNA template. The PCR in an ABI PRISM system (Applied Biosystems, Carlsbad, CA, USA) involved predenaturation at 95 °C for 2 min, then 40 cycles of 95 °C for 3 s and extension at 60 °C for 30 s. Samples were tested in duplicate. After amplification, the positive LAMP products could be observed directly by the color change of the FD. Moreover, LAMP products were examined by 2.5% agarose gel electrophoresis. To determine the optimum reaction temperature, the LAMP amplification was carried out for 60 min at constant temperatures from 60 to 67 • C in 1 • C intervals, with final incubation for 5 min at 85 • C to terminate the reaction.

Specificity and Sensitivity of the B. quintana LAMP Assay
To assess specificity, LAMP reactions were carried out as described above using DNA templates from 23 Bartonella species and 13 non-Bartonella species (Table 3). Each sample was examined twice, independently. To determine the analytical sensitivity, genomic DNA from B. quintana str. Toulouse was serially diluted (to 5 ng, 50 pg, 500 fg, 250 fg, 125 fg, 62.5 fg, 31.25 fg, and 15.625 fg, respectively). The limits of detection of the LAMP and qPCR assays were ascertained from the lowest amount of template DNA that could be detected. Table 4 lists the primers and probe for the qPCR assay. qPCR amplification was performed in a 20-µL reaction containing 0.3 mM each primer, 0.2 mM probe, 10 µL PromegaGoTaq ® qPCR Master Mix, and 1 µL of DNA template. The PCR in an ABI PRISM system (Applied Biosystems, Carlsbad, CA, USA) involved predenaturation at 95 • C for 2 min, then 40 cycles of 95 • C for 3 s and extension at 60 • C for 30 s. Samples were tested in duplicate.

Practical Application of the B. quintana LAMP
The utility of the LAMP assay was tested by examining the blood samples of 100 rhesus macaque monkeys and 20 rhesus-feeders collected from the rhesus breeding base, Yibing City, Sichuan Province, China. qPCR and LAMP results were compared using the χ 2 test and Fisher's exact test.

Conclusions
In this study, a novel loop-mediated isothermal amplification (LAMP), which amplifies target DNA at a fixed temperature with high sensitivity, specificity and rapidity, was employed to detect B. quintana. According to the study, LAMP methodology is simpler than PCR-based methods and requires less equipment. It is a useful for diagnosis of trench fever in humans and primates, especially in low-resource settings, because of its rapid, sensitive detection that does not require sophisticated equipment.