Development of multiplex real-time hybridization probe reverse transcriptase polymerase chain reaction for specific detection and differentiation of Enterovirus 71 and Coxsackievirus A16

Available online at www.sciencedirect.com

Diagnostic Microbiology and Infectious Disease 61 (2008) 294 – 301 www.elsevier.com/locate/diagmicrobio

Virology

Development of multiplex real-time hybridization probe reverse transcriptase polymerase chain reaction for specific detection and differentiation of Enterovirus 71 and Coxsackievirus A16 Eng Lee Tan a,b , Vincent Tak Kwong Chow a , Seng Hock Quak c , Wei Cheng Andrea Yeo c , Chit Laa Poh a,d,⁎ a

Department of Microbiology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore 117597 b School of Chemical and Life Sciences, Singapore Polytechnic, Singapore 139651 c Department of Pediatrics, National University Hospital, Singapore 119074 d Faculty of Life and Social Sciences, Swinburne University of Technology, Hawthorn, Victoria 3122, Australia Received 14 July 2007; accepted 13 February 2008

Abstract Large outbreaks of hand, foot, and mouth disease have been reported in the Asia Pacific region over the last few years and resulted in significant fatalities. The 2 main etiologic agents are Enterovirus 71 (EV71) and Coxsackievirus A16 (CA16). Both viruses are closely related genetically and show similar clinical symptoms. However, EV71 are associated with neurologic complications and can lead to fatalities. In this study, we developed a multiplex real-time hybridization probe reverse transcriptase polymerase chain reaction to detect and differentiate EV71 from CA16 using the LightCycler (Roche Molecular Biochemicals). Specific primers and hybridization probes were designed based on highly conserved VP1 region of EV71 or CA16. Our results showed high specificity and sensitivities in detecting EV71 or CA16 from 67 clinical specimens, and no other enterovirus serotype was detected. Rapid diagnosis to differentiate EV71 from CA16 in outbreak situations will enable pediatricians to identify and manage the patients more effectively. © 2008 Elsevier Inc. All rights reserved. Keywords: Hand, foot, and mouth disease (HFMD); Enterovirus 71; Coxsackievirus A16; Multiplex real-time RT-PCR; Rapid detection and differentiation

1. Introduction The Asia Pacific region had experienced large epidemics of hand, foot, and mouth disease (HFMD) over the last few years, and the 2 main etiologic agents of HFMD were Enterovirus 71 (EV71) and Coxsackievirus A16 (CA16) (Brown et al., 1999; Chang et al., 1999). EV71 and CA16 are closely related genetically, sharing 77% nucleotide and 89% amino acid homologies (Brown et al., 1999). However, infection by EV71 is more often associated with severe neurologic diseases like aseptic meningitis, and brainstem and cerebellar encephalitis, which are not observed in HFMD cases caused by CA16 (AbuBakar et al., 1999; Lum ⁎ Corresponding author. Faculty of Life and Social Sciences, Swinburne University of Technology, Hawthorn, Victoria 3122, Australia. Tel.: +61392148878; fax: +61-392145908. E-mail address: [email protected] (C.L. Poh). 0732-8893/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.diagmicrobio.2008.02.009

et al., 1998). In the major HFMD outbreaks in Southeast Asia, EV71 is the main etiologic agent that caused 41 deaths in Malaysia in 1997 (Lum et al., 1998), 78 deaths in Taiwan in 1998 (Ho et al., 1999), and 4 deaths in Singapore in 2000 (Ahmad, 2000). Because the clinical symptoms of EV71- and CA16associated HFMD are similar, diagnosis depends largely on virus isolation and serotyping (Lim and Benyesh-Melnick, 1960). However, this requires 2 to 3 weeks of growth and neutralization of the viral isolates. However, antigenic typing could be hindered by nonneutralizable viruses because of aggregation, antigenic drifts, or the presence of multiple viruses in the specimen (Melnick, 1996). The development of polymerase chain reaction (PCR) techniques has contributed significantly to laboratory diagnosis of viral infections in terms of sensitivity, specificity, and the rate of detection in comparison with the cell culture method. Using conventional reverse transcriptase PCR

E.L. Tan et al. / Diagnostic Microbiology and Infectious Disease 61 (2008) 294–301

(RT-PCR) strategy, specific detection of EV71 or CA16 had been reported (Bendig et al., 2001; Brown et al., 2000; Nix et al., 2006; Singh et al., 2002; Tsao et al., 2002). However, the 2-step RT-PCR process is time consuming, and the risk of cross-contamination is increased by this approach. In addition, detection of the viruses based on conventional RTPCR requires at least 6 to 8 h and could be a limiting factor during outbreak situations. In a recent study, a conventional multiplex RT-PCR in combination with the microarray method was developed to differentiate EV71 from CA16 using specific primers targeting each of the viral RNA. RT-PCR was 1st carried out to amplify EV71 and CA16 viral RNA. The amplicons were then labeled with fluorescent dyes and added to array slides, which were spotted with 60-mer degenerate oligonucleotide probes specific for EV71 or CA16. A diagnostic accuracy of 92% and 95.8% was achieved for specific detection of EV71 and CA16, respectively (Chen et al., 2006). In another study, Tsao et al. (2006) evaluated a microchip method developed by DR. Chip Biotechnology (Miao-Li, Taiwan). The detection was based on the biotinylated PCR products hybridizing to specific probes, and this approach achieved a sensitivity of 82% (Tsao et al., 2006). However, the detection of EV71 from clinical specimens using both approaches is laborious. Besides, the entire diagnostic process using both methods require at least 10 h which is too long in outbreak situations. In recent years, real-time PCR has gained wider acceptance for viral diagnosis in laboratories because of higher sensitivity and specificity, and faster rate of detection, and it provides real-time monitoring of the amplification process through fluorescence emission (Mackay et al., 2002). Previous studies have shown that enteroviruses that cause HFMD could be differentiated from other viruses such as varicella-zoster virus, poliovirus, and Herpes simplex virus using real-time PCR SYBR Green I-based (Read et al., 2001) and TaqMan probe-based assays (Nijhuis et al., 2002; Petitjean et al., 2006). In all the 3 studies, the primers and probes were designed to target at the 5′-untranslated region (5′UTR) of the enterovirus genome, which is highly conserved among enteroviruses, and molecular epidemiologic studies have shown that the diversity in this region did not correlate well with enterovirus serotypes (Oberste et al., 1999). Thus, in these studies, different enterovirus serotypes could not be differentiated effectively from 1 another without DNA sequencing. We have previously developed a real-time hybridization probe RT-PCR to detect EV71 specifically from clinical specimens (Tan et al., 2006). However, with increased concerns over the HFMD infections caused by EV71 and CA16, there is a need for a rapid and highly specific method to distinguish these 2 viruses from other enteroviruses in large outbreak situations. In this study, we developed a 1-step quantitative multiplex real-time hybridization probe-based RT-PCR to detect and differentiate EV71 specifically from CA16 directly from clinical specimens within 1 to 2 h. We

295

also evaluated the efficacies of the real-time PCR assay in detecting EV71 or CA16 directly from clinical specimens such as sera, saliva, urine, stools, throat, and rectal swabs. 2. Materials and methods 2.1. Viral isolates A CA16 strain (CA16-G-10) was kindly provided by Dr. M.A. Pallansch, CDC, Atlanta, GA. Three Japanese strains, namely, 1585-Yamagata-01 (genogroup C2), 75-Yamagata03 (genogroup C4) and 2933-Yamagata-03 (genogroup B5) were kind gifts from Prof. K. Mizuta, Yamagata Prefectural Institute of Public Health, Yamagata, Japan. One Singapore strain, the fatal 5865/SIN/00009 strain (designated as strain 41), was isolated from patients during the outbreak in October 2000 and cultivated in tissue cultures. Other enterovirus isolates analyzed in this study included CA24, Coxsackieviruses B1 (CB1), CB2, CB3, and Echoviruses 6 and 7. 2.2. Sample processing and storage A total of 67 clinical specimens were obtained from 40 pediatric patients who were admitted to the National University Hospital, Singapore, with suspected HFMD. The clinical specimens included 8 stools, 12 rectal swabs, 10 blood serum, 13 throat swabs, 11 saliva, and 13 urine specimens. The saliva and the urine specimens were processed for RNA extraction directly upon receiving them. The throat and rectal swabs were 1st suspended in 1% phosphate-buffered saline (PBS) before being processed for RNA extraction. A 10% stool suspension was made by adding 0.5 g of stool (0.5 mL for fluid stools) to 5 mL of 1% PBS. The suspension was then centrifuged at 12 000 × g for 10 min and filtered. The filtrate was then subsequently processed. Blood samples were allowed to stand in a vertical position for about 15 to 20 min. After centrifuging for 10 min at 12 000 × g, the serum was then aspirated and transferred to a clean 1.5-mL sterile Eppendorf tube. 2.3. RNA extraction Viral RNA extraction was carried out using QIAamp Viral RNA Mini Kit according to the manufacturer's instruction (QIAGEN, Valencia, CA). Briefly, the specimens were lysed with a QIAGEN viral lysis (AVL) buffer, and the RNA released would be bound to the membrane. After washing twice with wash buffers to remove any contaminating proteins and lipids, the RNA was then eluted with an elution buffer (QIAGEN). 2.4. Design of primers and hybridization probes The VP1 nucleotide sequences of EV71 strains and CA16 strains from the GenBank were analyzed. Using BLAST (http://www.ncbi.nlm.nih.gov/BLAST) and DNASTAR program, we defined a highly conserved VP1 region of EV71

296

E.L. Tan et al. / Diagnostic Microbiology and Infectious Disease 61 (2008) 294–301

strains that differed from CA16. A pair of primers (designated as EVFP and EVRP) and a pair of hybridization probes (EV-FL and EV-LC) were designed for specific detection of EV71 as described previously (Tan et al., 2006; Table 1). Another pair of primers (CAFP and CARP) and a pair of hybridization probes (CA-FL and CA-LC) were designed for specific detection of CA16 (Table 1). The EVFL and the CA-FL probes were labeled with LC640 and LC705 fluorescence dyes, which enabled them to be detected at 640 and 705 nm, respectively. 2.5. Multiplex real-time hybridization probe RT-PCR Multiplex real-time hybridization probe RT-PCR was performed to detect and differentiate EV71 and CA16 using the 1-step LightCycler RNA Amplification Hybridization Probe kit. The test kit allows a 1-step RT-PCR in glass capillaries using the LightCycler instrument (Roche Molecular Biochemicals, Mannheim, Germany). The enzyme mix contains a mixture of RT and “Faststart” Taq Polymerase that allows reverse transcription of RNA template and subsequent cDNA amplification. Each reaction contained 300 ng of RNA, 5 mmol/L of MgCl2, 0.5 μmol/L of EVFP, 0.3 μmol/L of EVRP, 0.4 μmol/L of CAFP, 0.4 μmol/L of CARP, 0.2 μmol/L each of EV-FL, EV-LC, CA-FL, and CA-LC, 4.0 μL of RNA hybridization probe amplification reaction solution, and 0.4 μL of enzyme mix, and is made up to 20 μL with water. cDNA was 1st synthesized from the RNA for 10 min at 55 °C and followed by 40 cycles of amplification at 95 °C for 30 s, 58 °C for 15 s, and 72 °C for 12 s. The temperature transition rates for all preceding steps were set at 20 °C/s for rapid thermal ramping. Detection of

Table 1 Nucleotide sequences of the specific primers and hybridization probes designed for the specific amplification of EV71 (EVFP, EVRP, EV-FL, and EV-LC) a and CA16 (CAFP, CARP, CA-FL, and CA-LC) Primer/probes Primers EVFP EVRP CAFP CARP Probes EV-FL c EV-LC c CA-FL d CA-LC d a

Nucleotide sequence (5′–3′)

Position

GAG AGT TCT ATA GGG GAC AGT AGC TGT GCT ATG TGA ATT AGG AA ACC AGG CYG TCA CTT CCC RA b GTG GTG GGC ATG GTA ATA ATA CT

2466–2489 2669–2647 2467–2486 2734–2715

GAT GAC TGC TCA CCT GTG TGT TTT GAC C-FL LC640-GCT GGC AGG GCC TGG GTA AGT GCC-P CTC ACT AGC ATT GGT GTT GGC GGC GGT-FL LC705-GGT TCT ACC TGT AAG GAT GTC AGC GCA C-P

2553–2526 2518–2494

EV71 was observed in the F2 channel of the LightCycler that detects fluorescence emitted at 640 nm. Detection of CA16 was observed in the F3 channel that detects fluorescence emission at 705 nm. During the amplification process, fluorescence emission was monitored by the Ct values. The Ct threshold value represents the cycle number from which significant increase in fluorescence was detected. Melting curve analysis was carried out after the PCR reaction. A rapid thermal ramp to 95 °C was achieved for complete denaturation of the PCR amplicons, followed by hybridization at 60 °C for 10 s. A slow steady transition from 60 to 95 °C was then performed. Subsequently, a graph was plotted using dF/ dT against temperature, which would show a peak corresponding to the Tm of the PCR amplicon. To eliminate any possible crosstalk of the fluorescence between both detection channels, we carried out color compensation test using the LightCycler color compensation kit, according to the manufacturer's instruction (Roche Molecular Biochemicals). 2.6. Quantitation of multiplex real-time hybridization probe RT-PCR Quantitative analysis of the real-time hybridization probe RT-PCR for specific detection of EV71 was determined as previously reported (Tan et al., 2006). Similar approach was carried out to determine the detection limit of CA16 viral RNA. Briefly, conventional RT-PCR was 1st carried out to amplify a 268-bp VP1 fragment using the primers CAFP and CARP. The PCR product was then cloned into the pGEM-T easy vector (Promega, Madison, WI), orientated downstream of the T7 promoter. The plasmid was then linearized by digestion with SacI and purified by phenol– chloroform extraction. RNA standards were synthesized as “runoff” transcripts by using the T7 RNA polymerase in vitro transcription kit according to the manufacturer's protocol (Promega). After determining the concentration of the RNA standards spectrophotometrically at 260 nm, a series of 10-fold viral dilutions ranging from 5 × 109 copies to 5 × 100 copies was prepared and analyzed. The standard curve was calculated and generated by the LightCycler based on the regression analysis, which was calculated based on the regression between the Ct values versus the log10 of the standard viral RNA copies. Based on the standard curve generated, the correlation coefficient (R2) was also calculated to determine the linearity association between the Ct values and the viral RNA standard. All experiments were carried out in 2 independent experiments.

2549–2523 2521–2493

The sequences have been published in Tan et al. (2006). b Y—C, T; R—A, G. c EvVP1-FL was labeled with fluorescein at the 3′ end and EvVP1-LC was labeled with LC Red 640 at the 5′ end and phosphorylated at the 3′ end. d CA-FL was labeled with fluorescein at the 3′ end, and CA-LC was labeled with LC Red 705 at the 5′ end and phosphorylated at the 3′ end.

2.7. Nucleotide sequence analysis The purified PCR amplicons were cycle sequenced using the ABI PRISM Big Dye Terminator cycle sequencing ready reaction kit. DNA sequencing was performed using the ABI PRISM 377 DNA sequencer (Applied Biosystems, Foster City, CA). Homology searches were carried out using the BLAST algorithm (http://www.ncbi.nlm.nih.gov/BLAST).

E.L. Tan et al. / Diagnostic Microbiology and Infectious Disease 61 (2008) 294–301

297

Fig. 1. Multiplex detection of EV71 and CA16 using different detection channels of the LightCycler (Roche Molecular Biochemicals). (A) Specific detection of EV71 was observed in the F2 channel that detects fluorescence emission at 640 nm. (B) A Tm value of 72.8 °C was observed by melting curve analysis. (C) Specific detection of CA16 observed in the F3 channel that detects fluorescence emission at 705 nm. (D) A Tm value of 71.8 °C was observed by melting curve analysis.

Fig. 2. Detection of EV71 from other genogroups. The multiplex real-time hybridization probe RT-PCR was tested against EV71 strains belonging to other genogroups, including 1585-Yamagata-01 (genogroup C2), 75-Yamagata-03 (genogroup C4), and 2933-Yamagata-03 (genogroup B5). Strain 41 was used as a positive control. The negative control represents PCR reaction without any RNA template. No amplification was observed for other enteroviruses, including CA24, CB1, CB2, CB3, Echo 6, and Echo 7.

298

E.L. Tan et al. / Diagnostic Microbiology and Infectious Disease 61 (2008) 294–301

3. Results 3.1. Real-time RT-PCR specificity and sensitivity The detection and differentiation of EV71 from CA16 by multiplex real-time hybridization probe RT-PCR was analyzed using the EV71 strain 41 and the prototype CA16-G-10 stain. Specific amplification of EV71 strain 41 was observed at 640 nm (Fig. 1A). Melting curve analysis of the EV71 PCR amplicons showed a Tm of 72.8 °C (Fig. 1B). On the other hand, specific detection of CA16 was detected at 705 nm (Fig. 1C) and had a melting temperature (Tm) of 71.8 °C (Fig. 1D). Sequence analysis of the PCR amplicons generated from the multiplex real-

time hybridization probe RT-PCR showed high homologies to the EV71 or CA16 VP1 region (99% sequence homology, data not shown). We also tested other EV71 strains such as the 3 EV71 strains belonging to other genogroups, including 1585-Yamagata-01 (genogroup C2), 75-Yamagata-03 (genogroup C4), and 2933-Yamagata-03 (genogroup B5), and other enterovirus serotypes such as CA24, CB1, CB2, CB3, Echo 6, and Echo 7. Amplification of all the EV71 strains tested was observed but not the other enterovirus serotypes (Fig. 2). The analytical sensitivity of the multiplex real-time hybridization probe RT-PCR developed was assessed using the serially diluted EV71 or CA16 RNA in 2 independent experiments. Consistent with our previous study, the

Fig. 3. Quantitative analysis of multiplex real-time hybridization probe RT-PCR for the detection of (A) EV71 and (B) CA16. The standard curve was obtained by plotting the Ct values against starting copy number. The coefficients (R2) and linear equations are as shown. The experiment was carried out in 2 independent experiments.

E.L. Tan et al. / Diagnostic Microbiology and Infectious Disease 61 (2008) 294–301

detection limit for EV71 was 5 viral copies (Tan et al., 2006) (Fig. 3A), and statistical analysis of the standard curve showed high correlation coefficient (R2 = 1.00) between the Ct value and the standard RNA. The same detection limit (5 viral copies) was also observed for quantitative analysis of CA16, and the standard curve showed a high correlation coefficient (R2 = 0.98; Fig 3B). The linearity was consistent for both experiments (P b 0.001). 3.2. Comparison of multiplex real-time hybridization probe RT-PCR with tissue culture method The efficacy of the multiplex real-time RT-PCR in detecting EV71 or CA16 directly from clinical specimens was evaluated. For comparison purpose, the traditional viral tissue culture method was also carried out on the clinical specimens as described previously (Singh et al., 2002). A total of 67 clinical specimens from pediatric patients admitted to the National University Hospital with suspected HFMD were analyzed. Our results showed that the viral tissue culture method failed to detect EV71 in all the 67 clinical specimens. On the other hand, the multiplex realtime hybridization probe-based RT-PCR detected presence of EV71 in 48 clinical specimens and CA16 in 11 clinical specimens. One clinical specimen was found to contain both EV71 and CA16, whereas no amplification of either EV71 or CA16 was observed in the remaining 7 clinical specimens (Table 2). To verify that the positive clinical specimens did contain EV71 or CA16, we carried out a conventional RT-PCR with all the clinical specimens using specific primers targeting at the 5′UTR region (Nijhuis et al., 2002). Amplification of the 5′UTR region was observed for all the clinical specimens that tested positive for EV71 or CA16 by the multiplex real-time RT-PCR, and DNA sequence analysis showed that the PCR amplicons correlated highly with the 5′UTR region of EV71 or CA16, respectively (96–99% sequence homology, data not shown). The conventional RT-PCR was also carried out with the 7 clinical specimens that were tested negative for EV71 and CA16. Sequencing results showed that 4 clinical

Table 2 Detection of EV71 or CA16 in clinical specimens a by cell culture method in comparison with multiplex real-time hybridization probe RT-PCR Enterovirus serotypes

Tissue Culture method Multiplex real-time hybridization probe RT-PCR b

EV71

CA16

Other serotypes

0 49 c

0 11

0 7

a Clinical specimens isolated from pediatric patients included stools, blood, urine, rectal, and throat swabs. b Multiplex real-time hybridization probe RT-PCR with primers and probes targeting at the VP1 region of EV71 or CA16. c Included 1 clinical specimen that tested positive for both EV71 and CA16.

299

specimens contained Echo 7, whereas 2 clinical specimens were found to contain CB2. No amplification was observed for 1 clinical specimen.

4. Discussion In 2006, EV71 reemerged as the main causative agent in 2 HFMD outbreaks in Sarawak, Malaysia, as well as in Singapore. In Sarawak, more than 8000 cases were reported. There were 13 fatalities, and EV71 was isolated from more than half of the infected patients (Ministry of Health, Malaysia). In Singapore, more than 3000 HFMD cases were reported, and 75% of the pediatric patients were infected with EV71 (Ministry of Health, Singapore). At present, there is no antiviral or vaccine available, and most cases of HFMD are managed with symptomatic treatment. It has been reported by clinicians that most deaths from HFMD caused by EV71 occurred within 24 h when children were admitted with complications associated with HFMD (Ho et al., 1999; Wang et al., 2006). Thus, rapid and specific diagnosis of EV71 from clinical specimens will enable clinicians to identify those who require closer monitoring or even hospitalization before deterioration of clinical conditions. In this study, we developed a multiplex real-time hybridization probe-based RT-PCR that detects EV71 or CA16 directly from clinical specimens. The primers and hybridization probes were designed based on the VP1 region of either the EV71 or the CA16 genome because VP1 possesses a high degree of antigenic and genetic diversity that can be used to distinguish enterovirus serotypes (Oberste et al., 1999). Several studies have indicated that this region correlated more accurately with enterovirus serotypes when compared with the 5′UTR (Brown et al., 2000; Oberste et al., 1999). No amplification was observed for the other enterovirus serotypes, and the EV71 and CA16 amplicons generated were verified by DNA sequence analysis. Thus, our multiplex real-time RT-PCR assay was shown to exhibit 100% specificity in detecting EV71 or CA16. The procedure for the multiplex hybridization probe-based RT-PCR required only 1 to 2 h upon receiving the clinical specimens, and no post-PCR handling such as agarose gel electrophoresis was required. This approach for detection of EV71 and CA16 is more rapid than the tissue culture method, which requires 1 to 2 weeks, and the conventional seminested RTPCR methods, which requires 1 to 2 days (Brown et al., 2000; Singh et al., 2002). The real-time PCR assay was also shown to be more sensitive in detecting EV71 and CA16 in our study because quantitative analysis showed that the multiplex real-time hybridization probe RT-PCR was able to detect as low as 5 EV71 or CA16 viral copies. The ability of the multiplex real-time hybridization probe RT-PCR to detect EV71 strains from other genogroups was shown in this study. Phylogenetic analysis has shown that strain 41 were classified under genogroup B4 (Cardosa et al., 2003; McMinn et al., 2001). Other than strain 41, the

300

E.L. Tan et al. / Diagnostic Microbiology and Infectious Disease 61 (2008) 294–301

multiplex real-time hybridization probe RT-PCR developed in this study was also able to detect EV71 strains belonging to genogroups C2, C4, and B5. 4.1. Detection of EV71 or CA16 directly from clinical specimens We compared the efficacy of the multiplex real-time hybridization probe RT-PCR and the tissue culture method in detecting EV71 or CA16 directly from clinical specimens. In the present study, the clinical specimens such as throat swabs, urine, and saliva were collected by noninvasive means. Only those patients who required blood taken as part of their clinical investigations had their blood collected for EV71 detection. This is important because it will not cause discomfort to the pediatric patients because clinical specimens other than blood could be collected easily especially in large outbreak situations. We showed that the tissue culture method failed to detect EV71 in all 67 clinical specimens examined. This supported previous observations by Singh et al. (2002) in the low sensitivity in detecting EV71 directly from clinical specimens using the cell culture method. This is due to the fact that clinical specimens such as the cerebrospinal fluid (CSF) and swabs were often characterized by low viral load and may be diagnosed as false negative for EV71 by the tissue culture method (Singh et al., 2002). Other possible reasons could be due to antigenic drifts or the presence of multiple viruses in the specimens, which could hinder accurate diagnosis of the virus (Melnick, 1996). In this study, the multiplex real-time hybridization probe RT-PCR achieved high sensitivities in detecting EV71 or CA16 directly from clinical specimens. Sequence analysis of the 5′UTR PCR amplicons generated from conventional RTPCR confirmed the true positive identification of EV71 or CA16 from the clinical specimens by the multiplex real-time hybridization probe RT-PCR. No EV71 or CA16 was detected in 1 clinical specimen by the multiplex real-time hybridization probe RT-PCR. Because no amplification was observed for this clinical specimen when conventional RTPCR was carried out with specific primers targeting the 5′ UTR region, it is very likely that this clinical specimen contained other pathogens. In conclusion, the multiplex real-time hybridization probe-based RT-PCR developed in this study achieved high specificity and sensitivity in detecting EV71 or CA16 directly from clinical specimens. Because the clinical symptoms of HFMD caused by EV71 and CA16 are indistinguishable from each other, rapid diagnosis to differentiate EV71 from CA16 in outbreak situations will enable pediatricians to identify and manage the patients more effectively. Acknowledgments We thank Dr. M.A. Pallansch, CDC, Atlanta, GA, for providing the CA16 strain. We would also like to thank Prof.

K. Mizuta, Yamagata Prefectural Institute of Public Health, Yamagata, Japan, for providing EV71 strains 1585-Yamagata-01, 75-Yamagata-03, and 2933-Yamagata-03 for this study. This research was supported by an Academic Research Fund R-182-000-076-112 from the National University of Singapore. References AbuBakar S, Chee HY, Al-Kobaisi MF, Xiaoshan J, Chua KB, Lam SK (1999) Identification of enterovirus 71 isolates from an outbreak of hand, foot and mouth disease (HFMD) with fatal cases of encephalomyelitis in Malaysia. Vir Res 61:1–9. Ahmad K (2000) Hand, foot and mouth disease outbreak reported in Singapore. Lancet 356:1338. Bendig JWA, O'Brien PS, Muir P (2001) Serotype-specific detection of Coxsackievirus A16 in clinical specimens by reverse transcriptionnested PCR. J Clin Micro 39:3690–3692. Brown BA, Oberste MS, Alexander JP, Kennett ML, Pallansch MA (1999) Molecular epidemiology and evolution of enterovirus 71 strains isolated from 1970 to 1998. J Virol 74:9969–9975. Brown BA, Kilpatrick DR, Oberste MS, Pallansch MA (2000) Serotypespecific identification of enterovirus 71 by PCR. J Clin Virol 16: 107–112. Cardosa MJ, Perera D, Brown BA, Cheon D, Chan HM, Chan KP, Cho H, McMinn P (2003) Molecular epidemiology of human enterovirus 71 strains and recent outbreaks in the Asia-Pacific region: comparative analysis of the VP1 and VP4 genes. Emerg Infect Dis 9:461–468. Chang LY, Lin TY, Huang YC, Tsao KC, Shih SR, Kuo ML, Ning HC, Chung PW, Kang CM (1999) Comparison of enterovirus 71 and coxsackievirus A16 clinical illness during the Taiwan enterovirus epidemic, 1998. Pediatr Infect Dis J 18:1092–1096. Chen TC, Chen GW, Hsiung CA, Yang JY, Shih SR, Lai YK, Juang JL (2006) Combining multiplex reverse transcription-PCR and a diagnostic microarray to detect and differentiate enterovirus 71 and coxsackievirus A16. J Clin Micro 44:2212–2219. Ho M, Chen ER, Hsu KH, Twu SJ, Chen KT, Tsai SF, Wang JR, Shih SR (1999) An epidemic of enterovirus 71 infection in Taiwan. N Engl J Med 341:929–935. Lim KA, Benyesh-Melnick M (1960) Typing of virus by combination of antiserum pools. Application to typing of enteroviruses. J Immunol 84: 309–317. Lum LC, Wong KT, Lam SK, Chua KB, Goh AY (1998) Neurogenic pulmonary oedema and enterovirus 71 encephalomyelitis. Lancet 352: 1391. Mackay IM, Arden KE, Nitsche A (2002) Real-time PCR in virology. Nucleic Acids Res 30:1292–1305. McMinn P, Lindsay K, Perera D, Chan HM, Chan KP, Cardosa MJ (2001) Phylogenetic analysis of enterovirus 71 strains isolated during linked epidemics in Malaysia, Singapore and Western Australia. J Virol 75: 7732–7738. Melnick JL (1996) Enteroviruses: polioviruses, coxsackieviruses, echoviruses, and newer enteroviruses. In: Fields virology, 3rd ed. Fields BN, Knipe DM, Howley PM, Eds. Philadelphia: Lippincott-Raven Publishers, pp. 655–712. Nijhuis M, van Maarseveen N, Schuurman R, Verkuijlen S, de Vos M, van Loon Hendriksen AMK (2002) Rapid and sensitive routine detection of all members of the genus enterovirus in different clinical specimens by real-time PCR. J Clin Microbiol 40:3666–3670. Nix WA, Oberste MS, Pallansch MA (2006) Sensitive seminested PCR amplification of VP1 sequences for direct identification of all enterovirus serotypes from original clinical specimens. J Clin Microbiol 44: 2698–2704. Oberste MS, Mather K, Kilpatrick DR, Pallansch MA (1999) Molecular evolution of the human enterovirus: correlation of serotype with VP1

E.L. Tan et al. / Diagnostic Microbiology and Infectious Disease 61 (2008) 294–301 sequence and application to picornavirus classification. J Virol 73: 1941–1948. Petitjean J, Vabret A, Dina J, Gouarin S, Freymuth F (2006) Development and evaluation of a real-time RT-PCR assay on the LightCycler for the rapid detection of enterovirus in cerebrospinal fluid specimens. J Clin Virol 35:278–284. Read SJ, Mitchell JL, Fink CG (2001) LightCycler multiplex PCR for the laboratory diagnosis of common viral infections of the central nervous system. J Clin Microbiol 39:3056–3059. Singh S, Chow VTK, Phoon MC, Chan KP, Poh CL (2002) Direct detection of enterovirus 71 (EV71) in clinical specimens from a hand, foot and mouth disease outbreak in Singapore by reverse transcription-PCR with universal enterovirus and EV71-specific primers. J Clin Microbiol 40: 2823–2827.

301

Tan EL, Chow VTK, Kumarasinghe G, Lin RTP, Mackay IM, Sloots TP, Poh CL (2006) Specific detection of Enterovirus 71 directly from clinical specimens using real-time RT-PCR hybridization probe assay. Mol Cell Probes 20:135–140. Tsao KC, Chang PY, Ning HC, Sun CF, Lin TY, Chang LY, Huang YC, Shih SR (2002) Use of molecular assay in diagnosis of hand, foot and mouth disease caused by enterovirus 71 or coxsackievirus A16. J Virol Methods 102:9–14. Tsao LY, Lin CY, Yu YY, Wang BT (2006) Microchip, reverse transcriptionpolymerase chain reaction and culture methods to detect enterovirus infection in pediatric patients. Pediat Int 48:5–10. Wang JN, Yao CT, Yeh CN, Huang CC, Wang SM, Liu CC, Wu JM (2006) Critical management in patients with severe enterovirus 71 infection. Pediatr Int 48:250–256.