OUP user menu

Following the course of human leptospirosis: evidence of a critical threshold for the vital prognosis using a quantitative PCR assay

Johann Truccolo, Ophélie Serais, Fabrice Merien, Philippe Perolat
DOI: http://dx.doi.org/10.1111/j.1574-6968.2001.tb10904.x 317-321 First published online: 1 November 2001

Abstract

In order to follow the course of acute human leptospirosis, an ELISA microtiter plate hybridization method was developed for the quantitative determination of Leptospira spp. in biological samples after PCR. The biotin-labelled amplified product (331 bp from the rrs gene) was hybridized with a complementary capture probe covalently linked onto aminated polystyrene wells, and detected using a colorimetric reaction. The mean detection limit was 50 copies per 10 μl. In a prospective study of human leptospirosis cases, we obtained evidence that a density of 104 leptospires per ml of blood is a critical threshold for the vital prognosis of the patients. The practicability of the method makes it suitable for use in tropical areas for multicentric studies. Such studies could lead to a better knowledge of the natural history of the human disease. The method is also suitable for experimental evaluation of improved antibiotic treatments for leptospirosis.

Keywords
  • Leptospirosis
  • Leptospira interrogans
  • Quantitative polymerase chain reaction
  • Pathogenesis

1 Introduction

Leptospirosis, which is caused by pathogenic species of the genus Leptospira, is a worldwide zooanthroponosis affecting wild and domestic animals and humans. Because of the broad range of clinical symptoms, the diagnosis of human leptospirosis remains difficult [1]. Laboratory confirmation is, therefore, essential for effective therapy and for epidemiological surveys. However, standard methods (i.e., microscopic agglutination test and culture in EMJH medium) are time-consuming and lack sensitivity [1]. A combined polymerase chain reaction (PCR)–hybridization test has previously been proposed for qualitative detection of a specific target DNA sequence (16S RNA gene) of Leptospira[2]. This did not, however, allow the evaluation of bacterial density [3].

Quantification of a pathogen contributes to the knowledge of the natural history of the disease and may provide useful information for the monitoring of patients. Various colorimetric quantitative PCR techniques in microtiter plates have been described for many micro-organisms [46]. In leptospirosis, the kinetics of infection remain poorly investigated, and little is known about the pathogenesis of the human disease. As the majority of the human cases occur in tropical areas, a thorough study of the natural history of the disease requires tools to be available that are cost-effective as well as reliable.

In the present work, a rapid and simple microtiter plate hybridization assay was developed, for the quantification of pathogenic leptospires after PCR amplification with biotinylated primers. The capture probe was covalently linked onto aminated wells of microtiter plates, using carbodiimide as a coupling agent. This method was validated in a prospective study of blood and urine samples from confirmed cases of human leptospirosis.

2 Materials and methods

2.1 Preparation of DNA for standard curve

For the standard curve, various numbers of leptospires (serovar Icterohaemorrhagiae strain Verdun), ranging in an geometric ratio of 10 from 20 to 2×107, were heated (in 100 μl sterile distilled water) for 10 min at 96°C, and a 10-μl aliquot (corresponding to 2 to 2×106 leptospires) was used for PCR.

2.2 Treatment of clinical specimens and DNA amplification

A total of 27 biological samples (blood, urine) submitted to our laboratory were obtained from 12 patients with biologically confirmed leptospirosis. The microscopic agglutination test was performed according to standard procedures [1]. Serologically confirmed cases were those showing seroconversion (negative first serum sample and a titer ≥1:100 in the second sample) or a rise in titer of at least two dilutions between the acute and the convalescent phases. When available, clinical specimens were cultured in EMJH medium for primary isolation [1]. For the quantification of DNA in biological samples, the DNA from Leptospira spp. was purified according to Boom et al. [7]. 10 μl of eluted DNA in TE buffer were used for subsequent amplifications. A 331-bp fragment and its complementary probe (CD probe; 289 bp) of the 16S rRNA gene of Leptospira species were amplified as previously described [2,3]. 30 μl of each amplified PCR product was purified with a QIAEX II extraction kit (Qiagen, Hilden, Germany), eluted in 30 μl of TE buffer, and checked by agarose gel electrophoresis.

2.3 Coupling of the capture CD probe to CovaLink NH microtiter plates

Coupling was performed as previously described [8]. The phosphorylated CD probe was denatured by heating at 95°C for 10 min, and then quickly chilled on ice to prevent secondary structure formation. For each microwell, a mixture (100 μl/well) containing 500 ng of denatured CD probe was diluted in a cold solution of 1-methylimidazole, 14.3 mM, pH 7.0 (Sigma Chemical Co., St. Louis, MO, USA) containing 167 mM 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide (Sigma). The plates were incubated overnight at 50°C and washed three times with a preheated (50°C) solution of 0.4 M NaOH containing 0.25% SDS. They were then incubated for 5 min with this solution at room temperature, and washed again three times with the same heated buffer. Finally, the microwells were washed three times at room temperature with TE buffer (10 mM Tris–HCl, 1 mM EDTA, pH 8.0), then filled with TE buffer, and plates were stored at 4°C.

2.4 Hybridization assay on CovaLink NH microtiter plates

Drawing of the standard curve and the quantification of PCR products in biological samples were performed simultaneously. 10 μl of QIAEX-purified biotinylated PCR product were denatured in 55 μl of 5×SSC by boiling (10 min), and chilled immediately on ice (10 min). After adding 30 μl of cold hybridization solution (5×SSC, 3% bovine serum albumin (BSA)) to the heat-denatured PCR product, this solution was transferred to the emptied wells. All samples were tested in triplicate. Plates were covered with a plastic seal (plate sealers 3095, Costar, Cambridge, MA, USA) and incubated for 2 h at 48°C with gentle stirring. After the hybridization reaction, plates were washed twice for 5 min at room temperature with 2×SSC–0.1% SDS, followed by one 10-min wash with 0.2×SSC–0.1% SDS at 48°C. Two final washing steps were carried out with 250 μl of phosphate-buffered saline (PBS)–Tween (1×PBS, 0.05% Tween 20, pH 7.4) at room temperature.

2.5 Colorimetric detection

Plates were incubated for 60 min at room temperature with alkaline phosphatase–extravidin conjugate (Sigma), diluted 1/1500 in PBS–Tween buffer containing 1% BSA (200 μl per well). They were then washed five times with PBS–Tween, and finally 200 μl of 0.2 M Tris buffer (pH 8.0), containing 1 mg ml−1para-nitrophenyl phosphate (Sigma), was added. The enzymatic reaction was carried out from 30 min to 1 h at 37°C in darkness. The absorbance (OD: optical density) was measured at 405 nm using a microplate reader (Titertek Multiscan Plus). The PCR fragment concentrations in the samples were determined by reading the individual absorbances at 405 nm on the curve obtained with the calibrating DNA. The detection level was defined as the lowest concentration exceeding the zero dose precision and was estimated using the equation: cut-off=mean absorbance of negative sample+3×S.D.

3 Results

3.1 Standard curve

Serial dilutions corresponding to 2 to 2×106 leptospires were amplified with 30 cycles and quantification was performed. Absorbance values were plotted against the number of Leptospira cells, resulting in a typical sigmoid dose–response curve (Fig. 1). A linear exponential relationship was demonstrated within the range of 102–105. The negative absorbance cut-off, defined as the mean absorbance obtained from three negative samples (TE buffer) plus 3 S.D., was used to determine the detection limit of the assay. In the following experiments, an absorbance of 0.06 was calculated as the cut-off value, corresponding to an initial number of 50 bacteria. The detection limit was 20-fold greater than that of gel electrophoresis, with which 103Leptospira cells were visually detected after amplification for 30 cycles [2]. The reproducibility of the method was determined in the exponential part of the standard curve by testing independently amplified samples (n=3) containing the same number of leptospires (2×102–2×105), and adding each corresponding product to a different well in the same microplate (Fig. 1).

Figure 1

Amplification of Leptospira spp. DNA by PCR (30 cycles) and detection of amplified DNA by colorimetric microtiter plate hybridization. Data are presented with standard deviations. The dotted line represents the cut-off value.

3.2 Detection of Leptospira spp. in human samples

The results of our prospective study are summarized in Table 1. Ages of the 12 patients ranged from 17 to 73 years (mean 36); nine were male and three were female. Twenty-seven samples, including 23 blood specimens and four urines, were collected for testing. Leptospirosis was confirmed both serologically (rise in titer) and by positive culture in six cases (patients A, E, F, H, I and J). The six isolated strains were identified as belonging to serogroup Icterohaemorrhagiae. Among the other six patients, two showed seroconversion (patients C and L), and one (patient G) had a significant rise in titer. In the three remaining cases (patients B, D and K), a significant follow-up was not possible, owing to the availability of a single serum (case B; high titer of 400) or to a too short period between the two serum samples (cases D and K).

View this table:
Table 1

Quantitative PCR, serologic and culture results for 12 leptospirosis patients

PatientAge (years)/sexDay of samplingaBiological samplesCulture in EMJH mediumSerologyPCR estimation of number of leptospires ml−1
A38/FD7blood+ (sg Ictero)1002×103
urinesnsc3.6×103
D10bloodnsc3 2000
B57/MD9bloodnsc4000
urinesnsc3.9×103
C73/MD1blood500
urinesnsc2.3×103
D9bloodnsc2000
Db45/MD3blood1006.1×103
D4bloodnsc1005.7×105
E24/MD4blood+ (sg Ictero)2001.5×103
D21bloodnsc3 2000
Fb29/MD3blood+ (sg Ictero)1005.2×104
D7bloodnsc12 8001.0×102
G35/MD5blood4002.4×102
urinesnsc9.6×101
D9bloodnsc3 2008.0×101
Hb39/MD17blood+ (sg Ictero)1001.5×106
D18bloodnsc12 8007.1×104
I18/MD7blood+ (sg Ictero)1002.0×102
D15bloodnsc3 2007.9×101
J30/FD14blood+ (sg Ictero)4001.1×102
D15bloodnsc12 8008.6×101
K29/FD2blood4002.7×103
D3bloodnsc8005.4×102
L17/MD2blood03.0×102
D6bloodnsc8005.6×101
nsc: not subjected to culture. sg Ictero: isolated strain typed as from serogroup Icterohaemorrhagiae.
  • aD0: estimated day of onset of illness.

  • bDead patient.

Three patients (D, F and H) were hospitalized in an intensive care unit and died of acute leptospirosis 2–4 days after admission. Two of these (F and H) gave a positive culture in EMJH medium (serogroup Icterohaemorrhagiae). Interestingly, these three patients showed the highest PCR-estimated bacteremia with at least 104 leptospires ml−1 of blood (maximum 1.5×106 ml−1 of blood, patient H). Antibiotic treatment was started on admission to hospital. For patient D, this antibiotic therapy had no effect on the level of PCR-estimated bacteremia, as an increase of two log occurred. Estimated bacteremia was significantly reduced in patients F and H, but the patients nevertheless developed hemorrhagic complications and died.

A weak PCR-estimated bacteremia (less than 104 leptospires ml−1 of blood) was observed in the other nine patients (A, B, C, E, G, I, J, K and L) (maximum 2.7×103 ml−1 of blood, patient K), all of whom recovered. When available, urines were always weakly positive (average of 2.5×103 leptospires ml−1; patients A, B, C and G). In three of these patients (A, B, C), the level of leptospiral DNA in urine was higher than that in the blood. This supports the potential value of urine for early PCR detection of leptospires in humans [9].

4 Discussion

Quantification of viral and bacterial agents in biological samples using PCR amplification has value for clinical and research applications. These applications include the monitoring of treatment, pathogenesis studies, and standardization of assays between laboratories. Microtiter plate-based hybridization offers the most convenient method for detecting and quantifying PCR-amplified fragments. However, variations in the passive binding capacity of polystyrene surfaces limit the use of these commercially available supports in hybridization assays. The recently developed real-time PCR system has great potential. However, the spread of this technique is restricted by its very high cost. This is specially the case in tropical areas where human leptospirosis is endemic.

In the present study, we used direct immobilization onto CovaLink NH plates with carbodiimide condensation to covalently bind a capture probe, complementary to the diagnostic amplicon, onto polystyrene (secondary amino group). This step offers two main advantages compared with passive adsorption: the attachment is quite stable, even after 3 weeks of storage at 4°C, and the denaturation step can be performed directly in the microwells. Satisfactory reproducibility, sensitivity and specificity are obtained, and this colorimetric microtiter plate hybridization assay can be used for quantifying leptospiral DNA using standard laboratory ELISA equipment.

Following patients in this study allowed us to estimate the bacterial density in the blood during the course of the disease. In most cases the estimated density was around 103–104 leptospires ml−1 (average on the first sample in the period D1–D4: 1.0×104 leptospires ml−1) (Table 1). Septicemia in leptospirosis can be considered as moderate [1]. The clinical evolution of our cases indicates that a density of 104 leptospires ml−1 of blood can be considered as a critical threshold for the vital prognosis of the patients. Indeed, the three patients with estimated bacteremia to this level all died. Evidence for the concept of a critical threshold relating to lethality was previously obtained using a mouse model of leptospiral infection [10]. In one case (patient F), a decrease in the estimated number of circulating leptospires under the threshold of 104 leptospires ml−1 was observed, but this patient did not recover. This was consistent with the course of experimental disease in laboratory rodents. They may die even though leptospires have been cleared from the bloodstream, if lesions in target organs (kidneys, liver) can be demonstrated [1,11].

In conclusion, the method described here can be applied to the rapid detection and quantification of leptospires in biological samples. The instrumentation developed for ELISA techniques can be used, allowing a large number of samples to be analyzed safely and conveniently. This quantitative PCR assay has two main potential applications. The first application is in studies of pathogenesis. Most current knowledge about the molecular determinants of pathogens comes from in vitro studies. However, bacterial growth in vivo changes during the course of infection, according to the target tissues and the host defences. In leptospirosis, little is known about the growth rates during infection in target organs in animal models [11]. Our data demonstrating the possibility of a critical threshold of PCR-estimated bacteremia in human disease are of significant interest, and would justify a multicentric prospective study. The other potential application is for the improvement of drug therapy. Leptospires are sensitive to a wide range of antibiotics, and the acquisition of new resistances has not been reported. However, standard antibiotic regimens using β-lactamins are not always effective in treating leptospirosis, as a persistent presence of leptospires has been observed in human patients [3,9]. Usual treatments may be unable to remove leptospires from the sites where they are protected from the immune response (meninges, eyes). Evaluation of various antibiotic regimens using animal models would help to define more efficient therapeutic strategies.

Acknowledgements

We thank Françoise Charavay for technical assistance. This work was supported by the Institut Pasteur, Paris, France (International Network of Institutes Pasteur). Warm thanks are due to Dr. Rod Chappel (National Serology Reference Laboratory, Melbourne, Australia) for editorial advice.

References

  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
View Abstract