Open Access

Update on molecular diagnosis of human leptospirosis

Teerasit Techawiwattanaboon
Teerasit Techawiwattanaboon
 and 
Kanitha Patarakul
Kanitha Patarakul
   | Jun 25, 2020
Asian Biomedicine's Cover Image
About this article
Previous Article
Next Article
Cite
Article Category: Review article
Published Online: Jun 25, 2020
Page range: 207 - 216
DOI: https://doi.org/10.1515/abm-2019-0063
Keywords
© 2019 Teerasit Techawiwattanaboon et al., published by Sciendo
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.
Leptospira and leptospirosis

Pathogenic Leptospira spp. are causative agents of human leptospirosis, a neglected zoonosis [1, 2]. The genus Leptospira is composed of pathogenic and saprophytic members which can be further classified into at least 20 species, 25 serogroups, and more than 250 serovars [2, 3]. The bacteria share conserved genetic sequences which can be used for the detection of pathogenic Leptospira [4]. The pathogens infect and persist chronically in the kidney of reservoir hosts before getting excreted into their urine and contaminating surrounding environment [5]. Leptospirosis has a globally widespread distribution but it is mostly endemic in tropical and subtropical regions [2, 6]. The disease typically affects poor living populations in urban slum and agricultural workers in endemic areas [7, 8]. The climate changes with heavy rainfalls and floods, and uncontrolled broaden of slum societies increase the number of leptospirosis cases [9, 10, 11]. Approximately 1 million cases of leptospirosis with 5%–10% fatal rate per year are estimated worldwide [12].

Humans are mainly infected by exposure to urine of infected animals or contaminated environments [1, 2]. The clinical manifestations of leptospirosis range from milder anicteric leptospirosis to severe illness called icteric leptospirosis or Weil's disease. In the mild form, patients present with acute undifferentiated fever, that is mild flu-like illness similar to other tropical infectious diseases such as malaria, dengue fever, enteric fever, typhus, influenza, and hepatitis [1, 13, 14]. Severe cases may present with multi-organ failure, such as jaundice, renal failure, lung hemorrhage, and septic shock [7, 15]. Prompt antibiotic treatment can reduce bacterial burden in patients. Inaccurate or delayed diagnosis leads to serious complications or even fatality. Currently, several diagnostic methods including direct examination with dark-field microscopy, culture, serological tests, and genomic DNA detection by molecular methods have been utilized by diverse laboratories based on their resource settings [3, 15]. This review provides updated information of leptospirosis diagnostic assays, especially rapid identification of pathogenic Leptospira by molecular techniques in the acute phase of infection.

Types of clinical specimens

Anicteric leptospirosis usually presents as a biphasic illness, i.e., the primary septicemic phase during the first week of infection followed by the secondary immune phase [1, 2, 16]. After acquisition through cracked skin or mucous membrane, pathogenic Leptospira disseminate and survive in the bloodstream [17, 18]. Leptospira may be detected in the blood during the leptospiremic phase within a few days post infection. The following immune phase is the concurrent clearance of leptospires from the blood and the appearance of Leptospira-specific antibodies which can be detected by serological assays. In the second phase, leptospires spread to target organs including kidneys from where the spirochete is excreted in urine [19]. Therefore, optimal specimens for genome-based assays are blood and urine in the primary and the secondary phases, respectively [1].

Previous studies suggested that whole blood and serum are more suitable for molecular detection than buffy coat [20, 21, 22]. Leptospires can be engulfed by leukocytes and enriched in a buffy coat [23]. However, the buffy coat is not generally used in routine diagnostic laboratories because of an additional step for isolation and higher risk of contamination with chemical and biological inhibitors than serum [20, 22, 24, 25]. Serum is less contaminated with inhibitors, e.g., heme, anticoagulants, and host cell DNA than whole blood [20, 26, 27, 28]. Nevertheless, whole blood samples containing both free-living and phagocytosed leptospires might be better than serum because bacteria are in total components of blood in contrast to serum from which bacteria could be trapped in blood clot after separation [21, 29, 30]. High-performance DNA extraction methods with additional step of inhibitor removal for whole blood may improve diagnostic sensitivity and accuracy.

Leptospira has been reported to shed and appear in urine since the acute phase of infection [31]. The pathogen can persistently colonize in proximal renal tubules over the period of infection, thus it can be found in urine for a longer period than in blood [2]. Previous report demonstrated that the detection of leptospiral DNA in urine and blood specimens by molecular methods had comparable and high sensitivity rate [22]. The urine sample preparation with an additional step of low-speed centrifugation and then washing with sterile phosphate-buffered saline before genome extraction increased yields of positive results [30]. However, urine may possibly be contaminated with other bacteria leading to false-positive detection of the 16S rRNA (rrs) target [22].

In cases of leptospirosis with acute meningitis or encephalitis syndrome, leptospiral DNA might be detected in the cerebrospinal fluid (CSF) [25]. Some patients had higher bacterial loads in CSF than in blood [25, 32].

Diagnostic strategies of leptospirosis

Patients with signs and symptoms compatible with leptospirosis should be confirmed by laboratory tests using microbiological and serological techniques [7, 15]. The specimen collection and assay selection are based on symptomatic period and laboratory performance.

Detection of Leptospira in clinical specimens and culture

A simple and rapid method to determine leptospiral infection is a direct examination of the pathogen in clinical specimens such as blood and urine under dark-field microscopy. Leptospira are visualized as thin, spiral-shaped, motile bacteria with hooked ends. However, this method has low sensitivity and specificity [33, 34] and it requires rarely available dark-field microscope. Culture of Leptospira from clinical samples requires special enriched media, such as Ellinghausen–McCullough–Johnson–Harris (EMJH) and Fletcher's media, and it needs up to 4 weeks for positive detection because the pathogen is fastidious and grows slowly [35]. Long-term incubation increases risk of contamination with other microorganisms. Although culture using modified medium formulations showed better efficiency to eliminate contaminants, it was still time-consuming and unsuitable for acute leptospirosis cases [36, 37]. Additional methods, such as serological tests with specific antibodies or polymerase chain reaction (PCR) with sequencing, are required to confirm the presence of pathogenic Leptospira.

Serological determination for human leptospirosis

Microscopic agglutination test (MAT) has been widely accepted as a standard method for confirmation of leptospirosis [38]. This method detects specific agglutinating antibodies in patient sera against reference serovars of Leptospira [2]. The positive case is characterized by either a fourfold or greater rising in MAT titers between acute and convalescent sera or a higher MAT titer than a cutoff level of single serum depending on the baseline titers among healthy individuals in the area [7]. This method has limitations due to low antibody level during acute phase; therefore, a convalescent serum is usually required to confirm the diagnosis which may not be practically obtained. In addition, it is laborious and the maintenance of viable pathogenic leptospires is required.

The enzyme-linked immunosorbent assay (ELISA) to measure specific IgM may have higher sensitivity than MAT for diagnosis of acute leptospirosis [39, 40, 41, 42, 43, 44, 45, 46, 47]. Whole-cell Leptospira and various recombinant outer membrane proteins have been used as target antigens. The major lipoprotein LipL32 of pathogenic Leptospira exhibited outstanding outcomes [40, 41, 44, 45, 46, 47]. The multi-subunit ELISA using LipL32 and other surface proteins could be alternative to MAT [44, 45]. Moreover, the multi-epitope fusion proteins containing LipL32 and other proteins were developed to improve the efficacy [46, 47]. These chimeric antigens showed a better specificity than individual antigens. However, ELISA is time-consuming and it needs expensive equipment and expertise which may not be appropriate for resource-poor settings.

The rapid immunoassays have been used as diagnostic tests for leptospirosis. Agglutination and immunochromatography to detect antibodies, mostly IgM, against leptospires are the key principles of most available kits. They are fast, simple, and user-friendly, thus they can be used for point-of-care testing. However, various studies revealed inconsistent sensitivity and specificity [48, 49, 50, 51, 52, 53]. The antibody baseline varies in different endemic areas, thus local validation is required. Recent study evaluated commercially available rapid diagnostic tests to detect IgM against Leptospira at the time of admission showing approximately 90% of both sensitivity and specificity [50, 51, 54]. Primary screening using Dual Path Platform (DPP) or Test-it and further confirmation with SD-IgM might enhance their accuracy. The evaluation of DPP with finger-stick blood presented similar results to venous whole blood and serum samples [51]. Therefore, the rapid diagnostic tests are more practical and less invasive for point-of-care testing. However, low levels of anti-Leptospira antibodies at the early phase of infection limited rapid IgM detection to be used for screening tests.

Detection of leptospiral antigens should be a better rapid detection platform in the acute phase of leptospirosis. Our group is developing a lateral flow assay targeting LipL32 antigen. The sensitivity and specificity are acceptable in a pilot study. The product has been patented and will be validated for the field use.

Molecular diagnosis for human leptospirosis

Currently, the affordable price of equipment and reagents promotes the use of molecular techniques for diagnosis of infectious diseases. Laboratory capacity has been gradually expanded in developing countries. DNA detection technology methods especially PCR-based methods have been applied for rapid detection of many emerging and re-emerging infectious diseases and have the potential to become a point-of-care testing in endemic areas.

In the early leptospiremic phase, serological assays for antibody detection are limited [1, 2, 16]. Although eradication of leptospires by either immune response or antibiotic treatment results in failure of bacterial culture, genetic materials of leptospires may still be maintained at different sites in patients depending on the course of infection [20, 55, 56]. The molecular assays can detect both live and dead leptospires in various clinical specimens from both phases of leptospirosis leading to increased sensitivity of detection. The common targets for leptospiral detection are located on their housekeeping genes [57, 58, 59] and pathogen-specific regions [29, 30, 59, 60, 61, 62]. The amplification and detection systems included conventional PCR, real-time PCR, and real-time reverse transcriptase PCR (real-time RT-PCR) [57, 63, 64, 65, 66]. Molecular diagnosis of human leptospirosis is summarized in Table 1.

Molecular diagnosis of human leptospirosis

Detection methodMolecular targetLimit of detectionaClinical EvaluationReferences
Sample% Sensitivity% Specificity
SYBR greensecY60 GE/PCRWhole blood67.790.0[63, 71, 72, 73]
lfb1103 cells/mLNDNDND
lipL32102 cells/mL, 3 GE/PCRSerum30100
TaqManlipL3240 cells/mL, 1 GE/PCRWhole blood43.0–60.693.0–99.0[21, 60, 74, 75]
Serum29.199.0
Blood culture86100.0
UrineND100.0
rrs40 cells/mL, 1 GE/PCRWhole blood18.4–96.490.0–99.5[20, 74, 75, 76, 77]
Serum50.0–51.099.2
Buffy coat35.799.7
Blood culture69.5–100.095.2–97.0
Urine39.191.5–92.5
rrs/lipL32NDSerum53.999.6[22]
Buffy coat58.899.9
Urine45.099.6
RT-PCRrrs1 cell/mLWhole blood64.0100.0[66]
ddPCRlipL3210 cells/mL, 1 GE/PCRSpiked whole bloodNDNDOur study
RPAlipL32≤2 GE/RxSerum94.797.7[78]
LAMPlipL3210 GE/RxSerum, urine91.67100[79]
rrs10 GE/RxSerum43.683.5

SYBR green, SYBR green real-time PCR; TaqMan RT-PCR, TaqMan real-time PCR; RT-PCR, reverse transcription PCR; ddPCR, droplet digital PCR; RPA, recombinase polymerase amplification; LAMP, loop-mediated isothermal amplification; ND, not determined.

cells/mL, Leptospira cells per milliliter; GE/PCR, genome equivalent per PCR reaction; GE/Rx, genome equivalent per reaction.

Conventional PCR

Conventional PCR was initially developed for the detection of leptospires in various clinical samples such as blood, CSF, and urine [64, 65, 67, 68, 69]. A previous study reported excellent specificity (100%) and fair sensitivity (62%) of PCR in serum samples from acute phase of leptospirosis [70]. In this study, PCR was positive in three of the four (75%) patients presenting no detectable antibodies in MAT, IgM ELISA, or slide agglutination test [70]. Moreover, the combination of the serological methods and PCR could enhance the sensitivity (93.1%–96.5%) in the first phase of leptospirosis [70].

Real-time PCR

The following generation of PCR technique, real-time PCR, has been applied to gain detection capability [57, 63]. Real-time PCR has faster turnaround time and more robust sensitivity and fidelity than conventional PCR [58, 61]. The single-tube assay with simultaneous detection prevents false-positive detection caused by amplicon contamination. Unlike conventional PCR, real-time PCR is a one-step assay with no requirement of downstream process such as gel electrophoresis and ethidium bromide staining, thus it reduces the risk of exposure to chemical hazards. In addition, this platform could be further applied to quantitate leptospiral burden in patient bodies, which may be useful to monitor response to antibiotic treatment, vaccine efficacy, and correlation of bacterial burden with disease severity [20, 80].

SYBR green real-time PCR

The SYBR green real-time PCRs targeting different genes, such as lipL32, lfb1, gyrB, and secY, have been used for rapid diagnosis of leptospirosis [58, 59, 60, 63, 72] and demonstrated comparable limit of detection of approximately 103 leptospiral cells/mL in the samples (1 cell or 5 genome equivalents per PCR reaction). The excellent performance was achieved by the detection of lipL32 which exhibited similar sensitivity and specificity to TaqMan real-time PCR targeting the same region [72]. The lipL32-specific SYBR green detection could detect 3 and 10 leptospiral genomes in spiked serum and urine, respectively [57]. The information suggests that TaqMan real-time PCR may not be required if the appropriately optimized condition for SYBR green dye is used.

Although SYBR green real-time PCR needs extra time for melting curve analysis, the cost of reagents is currently affordable and may compensate for the drawback. The melting peak is unique for each amplicon; therefore, it can be used to discriminate genome identity. Simultaneous pathogen detection and species classification is the advantage of SYBR green exceeding molecular probe real-time PCR. The combination of SYBR green-based detection and high-resolution melting (HRM) analysis becomes a novel and rapid diagnostic assay for leptospiral genotyping based on melting profiles [81, 82, 83]. This technique consists of amplification and detection of target genes, e.g., rrs, lfb1, and secY, followed by amplicon melting temperature determination. The recent report revealed the excellent diagnosis accuracy (100%) of SYBR green real-time HRM to identify leptospiral DNA from clinical specimens comparing with nested PCR [81]. This assay could differentiate six genospecies including four known Leptospira spp. (L. interrogans, L. borgpetersenii, L. kirschneri, and L. noguchii) and two unclassified clinical isolates.

TaqMan real-time PCR

TaqMan probe real-time PCRs targeting unique genes of pathogenic serovars, such as lipL32 gene, or genus-specific genes, such as 16S rRNA gene, have been widely employed in several laboratories [20, 21, 62, 76, 80]. The detection of pathogenic Leptospira DNA using the lipL32-specific probe and primers showed satisfactory performance [21, 30, 74, 75]. A previous study revealed that lipL32-specific TaqMan PCR had 86% sensitivity and 99% specificity to detect leptospiral DNA in whole blood during the first week of infection [21]. The high-performance (90.3% positive detection value, 95% confidence interval [CI]) was observed in specimens collected during 3–4 days of illness [21]. The limit of detection was estimated to be three genome copies per reaction or 103 cells/mL in spiked whole blood, which is almost equivalent to the lowest theoretical possibility. The efficiency of lipL32 detection was improved by optimization of probe and primer sequences and their final concentrations [29, 72, 75]. Therefore, local validation in each laboratory should be performed to determine the optimal condition.

The prevalence of leptospirosis cases caused by intermediate Leptospira spp. has been gradually reported [84]. The lipL32-based molecular diagnosis might give false-negative results; therefore, the detection of 16S rRNA gene, which is generally found in all species, should be used in these cases. TaqMan real-time PCR using rrs-specific probe and primers with comparable efficacy to lipL32 detection were reported [20, 22, 57, 74, 76, 77]. The rrs detection was promising to elicit consistent results with MAT [74]. The direct comparison studies showed that rrs detection provided higher sensitivity but lower specificity than lipL32 detection of pathogenic Leptospira in clinical specimens [74, 75]. The rrs platform detected a few intermediate and nonpathogenic species resulting in its less specificity. The duplicate copies of rrs gene on single leptospiral chromosome may be the reason for better sensitivity compared with single copy per genome of lipL32 gene. The design of a new rrs-specific probe without the recognition site at the conserved region in nonpathogenic serovars improved the test sensitivity and specificity [62].

Additionally, the combination of lipL32 and rrs detection (rrs/lipL32 real-time PCR) [22] showed comparable sensitivity and specificity to single 16S rRNA detection in blood and buffy coat samples, but it had significantly greater specificity than 16S rRNA detection in urine [22, 77]. Pre-incubation of blood specimens in routine hemoculture media containing tryptic hydrolysate of casein and soy peptone with sodium polyanethol sulfonate before DNA isolation improved sensitivity in a pilot study but showed poor sensitivity in a subsequent prospective study suggesting that directly obtained venous blood samples are more preferable for qPCR [77].

Reverse transcription PCR

The transcript of constitutively expressed 16S rRNA gene is more abundant than its corresponding genomic DNA in viable bacteria; therefore, detection of the gene transcript should be more sensitive. The 16S rRNA gene transcripts of Leptospira were measured by RT-PCR or cDNA-based detection [66]. The SYBR green real-time RT-PCR using rrs-specific primers for clinical blood samples provided 64% sensitivity and 100% specificity [66], which was better than a parallel DNA detection assay. Using the same primers for spiked whole blood, detection of RNA targets yielded 100-fold higher sensitivity than the DNA-based qPCR. Another study indicated that RNA-based detection by TaqMan real-time RT-PCR showed similar results to the SYBR green platform [55]. The real-time RT-PCR produced 5.6 cycle threshold of positive detection earlier than real-time PCR when the same DNA samples were analyzed in parallel indicating enhanced detection by RT-PCR [55, 66].

Droplet digital PCR

Droplet digital PCR (ddPCR) is a novel nucleic acid quantification method that can be subsequently applied to TaqMan probe, SYBR green, or reverse transcription ddPCR [85, 86, 87, 88]. The reaction mixture is generated into water–oil emulsion nanosize particles that separate a single DNA template molecule into individual droplets. The positive detection refers to the fluorescent signal in each droplet representing single target copy per droplet. The direct quantification is estimated by the total count of positive droplets; therefore, the DNA standard curve is not required. Several publications reported the advantages of ddPCR in terms of sensitivity, accuracy, and reproducibility than real-time PCR to detect the causative agents of infectious diseases in clinical samples with very low level of targeted genomes [86, 88, 89, 90, 91, 92, 93]. However, current ddPCR systems compared with qPCR are very expensive, have low throughput, and have longer turnaround time to be used for routine diagnosis.

In theory, ddPCR might be useful for diagnosis of leptospirosis in the early phase because of low-abundance Leptospira in the blood. Recently, we performed preliminary experiments to detect and enumerate the number of leptospires in whole blood specimens. We used ddPCR with lipL32-specific probe and primers to detect leptospiral DNA in Leptospira-spiked blood. This method was able to detect one leptospiral cell in 5 μL of DNA template per reaction which was not detected by TaqMan real-time PCR using the same probe and primer set (unpublished data). The limit of detection was 10 cells/mL. Therefore, our results supported the greater efficiency of ddPCR. Moreover, we also used this technique to quantitate leptospiral burden in blood and kidneys of vaccinated hamsters. The absolute quantification of leptospiral DNA by ddPCR helped us to obtain more reliable results to determine the vaccine efficacy (unpublished data). However, there is insufficient data to support utilizing ddPCR in diagnosis of leptospirosis. More studies using enough clinical samples are required to determine its performance compared with other techniques.

Isothermal amplification

The leptospirosis cases commonly present at hospitals in the rural areas. The rapid detection with less expensive and simpler methods such as loop-mediated isothermal amplification (LAMP) and isothermal recombinase polymerase amplification (RPA) assay have been determined for diagnosis of leptospirosis [78, 79, 94, 95, 96, 97]. Similar targets to PCR methods, lipL32 and rrs, are also used for LAMP and RPA. The RPA procedure contains the process of DNA synthesis, DNA recombination, and DNA repair. The commercial RPA presented sensitivity of 94% and specificity of 97.7% comparing with culture method and the limit of detection of approximately two genome copies per reaction [78]. The LAMP using 16S rRNA specific primers detected as lower as two leptospiral cells per reaction in urine [96]. In a clinical evaluation of the 16S rRNA LAMP, the sensitivity of 43.6% and specificity of 83.5% were estimated in blood specimens [97]. The sensitivity of 91.67% and specificity of 100% for Leptospira detection in serum and urine were obtained from lipL32-specific LAMP [79]. The nanoparticle-based lateral flow dipstick assay combined with multiplex LAMP had three targets: the LAMP target amplicon, the LAMP internal control amplicon, and a chromatography control, and it showed acceptable performance [95] with 100% specificity and a limit of detection of 0.395 genomic equivalent/mL; therefore, it may serve as a point-of-care test for diagnosis of leptospirosis.

Conclusions

Early diagnosis of leptospirosis especially in severe cases is crucial for prompt and appropriate treatment to minimize morbidity and mortality. Nucleic acid detection tests are recommended for detection of Leptospira in the blood samples during acute leptospiremic phase. So far, there are insufficient data of direct comparison demonstrating which nucleic acid detection test is the most accurate. In practice, we suggest that PCR-based detection either conventional or real-time PCR, depending on availability in the setting, is a method of choice to detect leptospiral DNA in the blood samples collected in the acute phase of leptospirosis. Their accuracy may depend on several factors including timing of sample collection, prevalence in the region, blood sample type (whole blood, plasma, or serum), primers or target genes, and detection method of the amplified products. However, the interpretation of the results must consider the strength of suspicion of leptospirosis cases and the prevalence of leptospirosis in the setting. In case of strong suspicion, diagnosis of leptospirosis cannot be excluded despite of negative PCR results. In the immune phase of disease, nucleic acid detection in urine samples may be performed in addition to serological tests, such as MAT or IgM ELISA to increase the sensitivity of diagnosis. Nevertheless, these tests will be less helpful in the early phase of infection. In the future, high-throughput nucleic acid detection or syndromic panel-based testing using multiplex real-time PCR or microarray, and next-generation sequencing that can be used for simultaneous diagnosis of multiple tropical diseases including leptospirosis, may be commercially available and more widely used if its cost-effectiveness has been determined and justified.

eISSN:
1875-855X
Language:
English
Publication timeframe:
6 times per year
Journal Subjects:
Medicine, Assistive Professions, Nursing, Basic Medical Science, other, Clinical Medicine
Journal RSS Feed