Quantitative detection of Proteus species by real-time polymerase chain reaction using SYBR Green I

Abstract

A SYBR Green real-time polymerase chain reaction (PCR) method for rapid detection of Proteus species was developed and evaluated. Of 322 clinical and food samples tested, 75 samples were positive for Proteus species by using conventional PCR and real-time PCR assays. The results were consistent with standard culture methods and the Vitek auto-microbe system, indicating a 100 % specificity obtained by both PCR assays. For the real-time PCR method, the minimum detectable level was 10 colony forming units (CFU) /ml, which was a 10³ multiple higher than the conventional PCR method. Correlation coefficients of standard curves which were constructed using the threshold cycle (Ct) versus copy numbers of Proteus showed good linearity (R ² = 0.997). In conclusion, several significant advantages such as higher sensitivity and rapidness were observed by using the SYBR Green real-time PCR method for identifying Proteus species.

Findings

Proteus is carried in the intestines of many invertebrate and vertebrate animals, in soil, wastewater, and decomposing organic remains, making up a part of the normal flora of the human gastrointestinal tract (Zych et al. 2007). Following Escherichia coli, Proteus is the second leading cause of Gram-negative bacterial urinary tract infections (Bauernfeind et al. 1987; Liu et al. 1992; Mobley and Belas 1995). Under certain circumstances, Proteus species can also cause other infections, such as bacteremia, respiratory infection, wound and burn infection, enteritis, pneumonia, chronic tympanitis, meningitis, peritonitis, keratitis, mastoiditis, and septicemia (Memmel et al. 2004; Tumialan et al. 2006; Bloch et al. 2011). Because of the increasing clinical relevance of Proteus (O'Hara et al. 2000), the development of efficient detection methods is of great epidemiological importance. Unfortunately, traditional culture methods are still the main detection techniques for Proteus. There have only been a few of reports of PCR-based molecular markers for molecular fingerprinting of Proteus (Serwecińska et al. 1998; Michelim et al. 2008). For real-time PCR, this has been reported for the detection of various bacterial species but with the exception of Proteus species (Alfaresi and Elkosh 2006; Gubala 2006; Fenicia et al. 2007).

To develop a SYBR Green real-time PCR assay for detection of Proteus, a total of 322 clinical and raw meat samples (including 165 urine samples, 139 stool samples, 6 pork, and 12 chicken) obtained in Guangzhou, China, were examined for Proteus species. Totals of 165 urine samples and 139 stool samples originated from 165 patients with urinary tract infections and 139 patients with diarrhoea in the First Affiliated Hospital of Jinan University from 2004 to 2008, respectively. The remaining 18 raw meat samples were purchased from local butcher shops in 2008. Ten reference strains (including Proteus mirabilis, Proteus vulgaris, Providencia stuartii, Providencia alcalifaciens, Yersinia enterocolitica, Listerella monocytogenes, E. coli, Salmonella enteritidis, Vibrio cholerae, and Shigella dysenteriae) were provided by the Guangzhou Centre for Disease Control and Prevention, China, for specificity tests of conventional PCR and real-time PCR assays. The P. mirabilis strain of these reference strains was also used for positive control and sensitivity tests of both PCR assays.

Of 322 samples, 75 Proteus strains were isolated from 75 clinical and food samples by biochemical tests and serological reactions. All the isolates were identified by the Vitek auto-microbe system (BioMérieux, La Balme les Grottes, France). Bacterial strains were grown aerobically in Luria–Bertani (LB) broth, at 37 °C overnight with shaking at 150 rpm, except Y. enterocolitica which was grown at 25 °C. For the raw meat samples, 10 g were transferred aseptically into a sterile filter bag and a 10-ml volume of 0.9 % (w/v) saline was poured into this bag. After the bag were stomached for 30 s using a BagMixer lab blender 400 (Interscience, Saint Nom, France), 1 ml of the homogenized solution was then transferred to be used for DNA extraction. DNA extraction of bacterial cultures and homogenized solution of the 18 raw meat samples was performed with the DNeasy tissue kit (Qiagen, Hilden, Germany) according to the manufacturer′s instructions. Sensitivity studies for the two PCR assays were implemented using DNA of seven 10-fold serial dilutions of Proteus grown overnight in LB broth. Diluted from 10² to 10⁸ times, serial dilutions of the cultures were plated on LB agar to determine the colony forming unit (CFU) counts. This experiment was performed in triplicate to ensure reproducibility.

Based on the report that atpD gene was more conservative than 16 S rRNA in bacterial species (Young and Park 2007), the atpD gene was selected as the PCR target for detection of Proteus species. The similarity of Proteus atpD gene sequences was analyzed by NCBI BLAST server (http://www.ncbi.nlm.nih.gov). High specific primers SL/PF (5′-GTATCATGAACGTTCTGGGTAC-3′) and SL/QPR (5′-TGAAGTGATACGCTCTTGCAG-3′) were designed for both conventional PCR and real-time PCR to differentiate Proteus and non-Proteus species. The amplified fragment was 101 bp in length.

For the conventional PCR, DNA was amplified on P×2 Thermal Cycler (Thermo Electron, Madison, USA) with 20-μl reaction mixtures, in which the PerfectShot Taq (loading dye Mix) (TaKaRa, Dalian, China) was used as per the manufacturer’s recommendations. Then, 0.2 μM of primers of SL/PF and SL/QPR were added to each reaction mixture, and 2 μl of extracted DNA were added as the template in the corresponding PCR mixtures. The amplification conditions were as follows: 94 °C for 5 min, 35 amplification cycles were performed at 94 °C for 30 s, 58 °C for 30 s and 72 °C for 30 s, and a final extension which was 7 min at 72 °C.

The real-time PCR and data analysis were performed using 96-well microwell plates and an ABI Fast 7500 instrument (Applied Biosystems, Foster, USA). The 20-μl total reaction volume of each microwell was composed as follows: 10 μl of SYBR^® Premix Ex Taq™ (TaKaRa), 0.4 μl of Rox Reference Dye II, 0.2 μM of primers of SL/PF and SL/QPR, 2 μl of template DNA, and RNase-free distilled water up to 20 μl. The amplification program included an initial denaturation step at 95 °C for 10 s, followed by 40 cycles of denaturation at 95 °C for 3 s, and annealing and extension at 60 °C for 25 s. Fluorescence signals were collected after each cycle. For melting curve analysis, the following step of amplification, the samples were denatured at 95 °C for 15 s, cooled at 65 °C for 60 s, and then heated slowly to 95 °C with a temperature transition rate 0.1 °C/s and a continuous monitoring of fluorescence at the same time.

High specificity was acquired when real-time PCR assay was subjected to 10 reference strains, with no false positive amplification observed. The melt curves revealed peaks at a melting temperature of 80.3 °C corresponding to the melting temperature of the specific amplified products. Simultaneously, the fragments of expected length of 101 bp were shown using agarose gel electrophoresis of the specific amplified products. According to the CFU counts for bacterial cultures diluted serially from 10² to 10⁸ times, the concentrations of the 7 dilutions were obtained. They were 1.08 × 10⁶, 1.08 × 10⁵, 1.08 × 10⁴, 1.08 × 10³, 1.1 × 10², 10, and 1 CFU/ml, respectively. The detection limit of real-time PCR assay was found to be 10 CFU/ml and conventional PCR assay was 1.08 × 10⁴ CFU/ml, indicating that real-time PCR was 10³-fold more sensitive than the conventional PCR (Fig. 1).

Fig. 1

The determination of the sensitivity of SYBR Green real-time PCR. Ten-fold serial dilutions P. mirabilis cells were used as the source of DNA template. Cell concentrations ranged from 1.08 × 10⁶ to 1 CFU/ml, where each curve with increasing Ct values represents amplification with a ten-fold dilution as a template. The representative amplification curves were for the products amplified with the primer pairs for atpD. a Amplification curves of real-time PCR; b dissociation curves of real-time PCR; 1 1.08 × 10⁶ CFU/ml; 2 1.08 × 10⁵ CFU/ml; 3 1.08 × 10⁴ CFU/ml; 4 1.08 × 10³ CFU/ml; 5 1.1 × 10² CFU/ml; 6 10 CFU/ml; 7 1 CFU/ml; 8 negative control

Standard curve was constructed based on the mean threshold cycle (Ct) and 6 serial dilutions using in real-time PCR ranging from 10¹ to 1.08 × 10⁶ CFU/ml, which resulted in a linear relationship between Ct and log CFU (Fig. 2). The equation of the final standard curve was Y = −4.1222X + 40.828. Meanwhile, the correlation coefficients of standard curve achieved 0.997, indicating that this curve can potentially be used for the purpose of relative quantitation or for estimation of unknown quantities of bacterial cells which were added to the assays.

Fig. 2

Standard curves showing the linear relationship between the log₁₀ values derived from 10-fold serial dilutions of P. mirabilis cell numbers versus cycle threshold values (from Fig. 1). Linear correlation coefficient was R ² = 0.997 (y = −4.1222x + 40.828)

This established real-time PCR assay was compared with the conventional PCR by detection of Proteus species from 322 clinical and food samples. Seventy-five samples (including 38 urine samples, 29 stool samples, 3 pork, and 5 chicken) were positive for Proteus species by using both PCR methods. The results were consistent with standard culture methods and the Vitek auto-microbe system, indicating a 100 % specificity obtained by both assays.

In summary, this study showed that the real-time PCR using SYBR Green I, an accurate, rapid, uncomplicated, specific and sensitive method, can be applied to identify Proteus species from clinical and food samples. Both the sensitivity and specificity of this method were excellent. A complete analysis of up to 96 samples only took about 50 min. Also, the method developed for SYBR green real-time PCR can be easily applicable to conventional PC,R though a lower detection limit was shown. The development of SYBR Green real-time PCR assay for detection of Proteus species will facilitate rapid diagnosis, and benefit epidemiological studies of the human carriage, environmental contamination, and/or soil distribution of Proteus species.