Rapid diagnosis of enterovirus infections by real-time PCR on the LightCycler using the TaqMan format PDF

Preview

Summary

Diagnostic Microbiology and Infectious Disease 42 (2002) 99 –105 www.elsevier.com/locate/diagmicrobio Virology Rapid diagnosis of enterovirus infections by real-time PCR on the LightCycler using the TaqMan format Thomas Watkins-Riedela,*, Markus Woegerbauerb, David Hollemanna, Peter Hufnaglc a The D...

Description

Diagnostic Microbiology and Infectious Disease 42 (2002) 99 –105

www.elsevier.com/locate/diagmicrobio

Virology

Rapid diagnosis of enterovirus infections by real-time PCR on the LightCycler using the TaqMan format Thomas Watkins-Riedela,*, Markus Woegerbauerb, David Hollemanna, Peter Hufnaglc a

The Division of Clinical Virology, Institute of Virology and Infectious Diseases, University Hospital Vienna, A-1090 Vienna, Austria b Department of Internal Medicine, University Hospital Vienna, A-1090 Vienna, Austria c Roche Diagnostics GmbH, A-1211 Vienna, Austria Received 17 July 2001; accepted 27 October 2001

Abstract A 2 h single-tube, reverse-transcription(RT)-PCR/hybridization assay using the TaqMan format for rapid diagnostic screening of enterovirus (EV) infections was optimized for the real-time LightCycler (LC) technology. For low EV load clinical samples an additional 30 min reamplification step using a novel primer-mix/probe design resulted in a 100% concordance with AMPLICOR EV PCR Test and in-house RT-PCR. Combined with maximum specificity, the sensitivity of LC-PCR was 10- to 100-fold higher compared to AMPLICOR EV Test. © 2002 Elsevier Science Inc. All rights reserved.

1. Introduction Enteroviruses (EVs) are responsible for more than 10 million symptomatic infections in the United States annually (Moore, 1982; Rotbart et al., 1994; Strikas, 1986). More than 80% of all cases of aseptic meningitis are caused by EVs worldwide (Young et al., 2000; Rotbart, 1990; Rotbart et al., 1990; Rotbart et al., 1994), posing a threat especially to children (Berlin et al., 1993; Rantakallio et al., 1986) and neonates (Modlin, 1997). EVs, like rhinoviruses a genus within the family of Picornaviridae, are single (⫹)-stranded RNA viruses comprising 66 distinct serotypes (Romero, 1999), which can be subdivided into polioviruses, group A and B coxsackieviruses, echoviruses and non-assigned enterovirus 68 through 71. EVs in clinical specimens were usually detected by tissue culture isolation from cerebrospinal fluid (CSF), throat swab or stool, thus being limited by a slow turnaround time (Romero, 1999; Rotbart et al., 1994) and relative insensitivity (Chonmaitree et al., 1988; Muir et al., 1999; Muir et al., 1993; Romero, 1999; Rotbart, 1997; Rotbart et al., 1994). Reverse-transcription (RT)-PCR based assays have been

described as a more rapid and sensitive approach to detect EV RNA in muscle biopsies, CSF, throat swabs, serum or stool samples (Pozo et al., 1998; Romero et al., 1996; Rotbart, 1997; Schlessinger et al., 1994). A commercially available PCR assay, i.e., AMPLICOR Enterovirus (EV) Test (Roche Molecular Systems, Branchburg, N.J.), which combined sufficient sensitivity and maximum specificity (Lina et al., 1996; Muir et al., 1999; Romero, 1999; Rotbart, 1997; Rotbart et al., 1994) was widely used until marketing of this test was abandoned in 1999. Thus, we developed an alternate 2 h RT-PCR assay for rapid and reliable detection of EV RNA in clinical specimens gaining advantage of the new real-time LightCycler (LC) PCR technology (Wittwer et al., 1997) using the TaqMan format (Nitsche et al., 1999). The LC instrument (Roche) allows on-line fluorometric detection of the target sequence, which is amplified in a glass capillary. Due to the high surface to volume ratio a rapid temperature transmission is achieved allowing ultra-short cycling times. Forty PCR amplification cycles usually are completed in just 30 to 35 min (Wittwer et al., 1997).

2. Materials and methods * Corresponding author. Tel.: ⫹1-43-1-40400-5138 (or 5149, 5150); fax: ⫹1-43-1-40400-5135. E-mail addresss: [email protected] ( T. WatkinsRiedel).

For early detection of EV RNA in clinical specimens, we established in the present study a qualitative 2 h, real-time LightCycler PCR assay (LC-PCR) including a 0.5 h RNA

0732-8893/02/$ – see front matter © 2002 Elsevier Science Inc. All rights reserved. PII: S 0 7 3 2 - 8 8 9 3 ( 0 1 ) 0 0 3 3 0 - 3

100

T. Watkins-Riedel et al. / Diagnostic Microbiology and Infectious Disease 42 (2002) 99 –105

isolation procedure. An initial single-round RT-PCR (1°LCPCR) for rapid diagnostic screening of EV infections was optimized by a single-capillary, one-step RT/amplification and hybridization protocol. Subsequently, EV LC-PCR conditions were adapted to the demands of our dual-labeled fluorogenic TaqMan-probe for maximum EV-specific online detection of LC-PCR products. For low EV load samples and/or confirmation of negative 1°LC-PCR results an additional 30-min rapid-cycle reamplification step (2°LCPCR) using a novel primer-probe design was developed. For RNA isolation, 200 ␮L of CSF, serum or plasma was extracted with High Pure Viral Nucleic Acid Kit (Roche Diagnostic Systems, Inc., Branchburg, N.J.) according to the manufacturer’s instructions. Prior to RNA extraction 100 ␮L of a stool sample were resuspended in 1 mL of phosphate-buffered saline (PBS) or TRIS/EDTA-buffer (TE), and centrifuged at room temperature for 5 min at 5000 rpm. 200 ␮L of the turbid supernatant were subjected to viral RNA preparation (see above). Throat swabs were swirled in 250 ␮L PBS or TE prior to RNA purification. All samples were subjected to RT-PCR within two hours or transferred to – 80°C for long term storage. Aliquots were frozen and thawed only once. For evaluation purposes our LC-PCR approach was compared throughout to AMPLICOR EV Test and an in-house PCR assay, respectively (primers/probes see Table 1; Rotbart et al., 1994; Steininger et al., 2001). For both 1°LC-PCR and first round in-house PCR primers E1 and E2 (obtained from ViennaLab, Vienna, Austria), modifications (see Table 1) of the published AMPLICOR EV Test primers, were used, located within the highly conserved 5⬘-noncoding region (5⬘-NCR) (Rotbart, 1990; Rotbart et al., 1994; Rotbart, 1997). For on-line detection of EV-specific PCR products a 5⬘-6FAM/3⬘TAMRA labeled TaqMan probe (see Table 1) was designed (purchased from VBC-Genomics, Vienna, Austria) using the “BLASTN” search tool on the world wide web (National Center for Biotechnology Information [NCBI]; Bethesda, Md., USA; http://www.ncbi.nlm.nih.gov/blast). Because of the high degree of homology to human rhinoviruses within the 5⬘-NCR common to all Picornaviridae (Halonen et al., 1995; Melnick, 1996; Steininger et al., 2001), 5⬘-NCR sequence alignments of all known EVs were compared consequently to the closely related rhinoviruses by “Clustal W” (http://www.ebi.ac.uk/clustalw/; EMBL European Bioinformatics Institute). A short, extremely high conserved region within this amplicon revealed to be optimal for EV-specific positioning of the TaqMan probe allowing exclusion of cross-hybridization with rhinoviruses. Negative and positive PCR run controls consisting of stock virus preparation aliquots with a final 50% tissue culture-infective dose (TCID50) of 103 per mL were included permanently in all EV PCR runs, i.e., one rhinovirus 14 specificity control, and both one coxsackievirus B3 and one poliovirus 3 sensitivity control (Fig. 1, Table 2). EV stocks were obtained from the American Type Culture Collection (ATCC). Two additional contamination controls

with diethylpyrocarbonate (DEPC)-treated nuclease-free H2O (see Table 2) and EV PCR-negative CSF aliquots were also subjected to each PCR run. All negative controls assayed negative by LC-PCR and AMPLICOR EV Test, with the in-house method rhinovirus 14 was detected at low rates (see Fig. 1). Therefore, a rhinovirus specific RT-PCR assay has recently been developed to differentiate subsequently between EVs and rhinoviruses (Steininger et al., 2001). Each 1°LC-PCR assay contained 16 ␮L master mix (1⫻ LC RNA Master Hybridization Probes (Roche), 5.5 mM Mn2⫹, 0.4 ␮M primers (each) and 0.2 ␮M EV-TaqMan probe) and 4 ␮L of sample RNA. Real-time PCR was performed on the LightCycler instrument (Roche, Mannheim, Germany) with the following temperature program: 20 min reverse transcription at 61°C, 2 min denaturation at 95°C and 60 PCR cycles with a denaturation at 95°C for 5 s, an annealing step with touch-down temperature profiling 10 sec from 70°C to 62°C for 10 s (step size 0.2°C per cycle, with single fluorescence measurement) and an elongation for 10 s at 72°C. For samples containing minimal amounts of enteroviral RNA and/or sequence polymorphism probably affecting the 1°LC-PCR performance, an additional rapid-cycle reamplification step by a semi-nested approach (2°LC-PCR) was developed. To overcome the problem of different amplification efficiencies with genetically distant EV serotypes (i.e., coxsackieviruses and polioviruses), a novel primerprobe combination for this 2°LC-PCR was designed. Limiting dilution testing of cell culture stock preparations with known TCID50 for a given serotype led to a mix (purchased from VBC-Genomics, Vienna, Austria) of three primers CoxF, PolF and EntR3 (Table 1) proved to be optimal for pan-specific generation of a 85 bp (coxsackievirus 3) or 87 bp (poliovirus 3) amplicon including the target sequence for the EV-TaqMan probe. Aliquots from the 1°LC-PCR were diluted 1:50 in nuclease-free distilled water and 2 ␮L of the dilution were combined with 18 ␮L LC-PCR master mix (1⫻ LC-Fast Start DNA Master Hybridization Probes (Roche), 4 mM Mg2⫹ (final concentration), 0.4 ␮M primer EntR3, 0.2 ␮M primers CoxF and PolF each and 0.2 ␮M EV-TaqMan probe). The LC-PCR temperature profile for the second round included a 10-min denaturation step at 95°C followed by 30 cycles of denaturation (95°C/5 s), annealing (65°C/ 10s, with single fluorescence measurement) and an elongation at 72°C for 10 s. Uracyl-DNA Glycosylase (UNG) carry-over prevention (Cat. No. 1 775 367; Roche Molecular Biochemicals, Mannheim, Germany) was done for both our LightCycler (LC) PCR and in-house PCR by adding 1 ␮L UNG to the respective master mix. Nevertheless, we found not a single discrepancy in a series of eight combined 1°⫹2°LC-PCR runs (data not shown) by testing 24 in-house PCR negative samples in parallel with and without UNG (i.e., a total of 48 control reactions) including at least 3 EV positive samples within each run to “provoke” crossover contamination. All

Real-time LightCycler (LC) PCR First round LC RT-PCR (1°LC-PCR) Primer E1 Primer E2 Amplicon Second round semi-nested LC-PCR (2°LC-PCR) Primer CoxF Primer PolF Primer EntR3 Amplicon Amplicon 1°⫹2°LC-PCR fluorescence detection Probe In-house nested RT-PCR First-round RT-PCR Second round nested PCR Primer Primer Amplicon AMPLICOR Enterovirus PCR (AMPLICOR EV Test) a

Primer and probe sequence

Length (nucleotides)

EV serotyp

Positions amplified

Forward Reverse

5⬘-CCC CTG AAT GCG GCT AAT CC-3 5⬘-CAA TTG TCA CCA TAA GCA GCC A-3⬘

20 22 149

Cox B3 Cox B3 Cox B3

453-472 601-580 453-601

Forward Forward Reverse

5⬘-GTA ACG GGC AAC TCT GCA GC-3⬘ 5⬘-CGT AAC GCG CAA GTC TGT GG-3⬘ 5⬘-ATT GTC ACC ATA AGC AGC CA-3⬘

20 20 20

Cox B3 Polio 3 Cox B3 Polio 3 Cox B3 Polio 3

515-534 516-535 599-580 602-583 515-599 516-602

26

Cox B3 Polio 3

560-535 562-537

20 22 109

Cox B3 Cox B3 Cox B3

465-484 573-552 465-573

85 87 EntTM1

TaqMan

5⬘-[6FAM-]CAD GGA CAC CCA AAG TAG TCG GTT CC [TAMRA]TP-3⬘ Identical to 1°LC-PCR

E3 E4

Forward Reverse

5⬘-TGA ATG CGG CTA ATC C(CT)A AC-3⬘ 5⬘-TGA AAC ACG GAC ACC CAA AGT A-3⬘

EV1b EV2b EV3

Reverse Forward Probe

Rotbart et al. 1994 J Clin. Microbiol. 32:2590–2592.

GenBank accession numbers: NC_001473.1 for coxsackievirus B3 (Cox B3) and L76405.1 for poliovirus 3 (Polio 3).

T. Watkins-Riedel et al. / Diagnostic Microbiology and Infectious Disease 42 (2002) 99 –105

Table 1 Oligonucleotide primers/probe locations and sequences for EV RT-PCR assaysa

101

102

T. Watkins-Riedel et al. / Diagnostic Microbiology and Infectious Disease 42 (2002) 99 –105

Fig. 1. Specificity and sensitivity testing of EV RT-PCR assays.

48 “contamination” controls reacted negative throughout. Thus, in our routine laboratory setting we use UNG prevention for LC-PCR only if some contamination might be suspected, because of additional costs and handling steps. To optimize LC-PCR assay conditions, EV serotypespecific detection limits were determined (1) by testing tenfold limiting dilution series (i.e., 102 to 10⫺3 TCID50) of cell culture stocks with EV serotypes representing one member of each major EV group (Fig. 1), (2) by participating in the international ‘Second (2nd; delivered in 1999) and Third (3rd; delivered in 2001) European Union Quality Control Concerted Action (EU-QCCA) EV proficiency panel’ (Table 2), and (3) by EV detection in a total of 60 clinical specimens submitted from the University Hospital Vienna (Table 3). EV stocks used in our study were either obtained from the ATCC or by cell culture propagation of EVs isolated from clinical stool samples followed by neutralizing serotyping (antisera obtained from ATCC). As shown in Fig. 1, we evaluated in our study serial tenfold TCID50 EV RT-PCR detection rates defined as the percentage (%) of positive PCR results at a given TCID50 and respective serotype per number of test runs (no. ⫽ 8). All LC-PCR results were compared to AMPLICOR EV Test and in-house PCR (Fig. 1, Tables 2 and 3). The 2°LC-PCR reamplification step utilizing the novel semi-nested primer mix resulted in an EV type-specific 10- to 100-fold im-

provement regarding the TCID50 detection threshold compared to 1°LC-PCR and AMPLICOR EV Test for all serotypes tested (Fig. 1). The EU-QCCA EV proficiency panel revealed a similar ⱖtenfold increase in sensitivity by realtime LC-PCR compared to AMPLICOR EV Test (Table 2). Both the endpoint titration experiments and EU-QCCA testing revealed comparable sensitivity of LC-PCR to in-house nested PCR, but a noticeably higher specificity regarding rhinovirus “cross-detection” was seen in our study due to the highly specific TaqMan format (Fig. 1, Table 2). To evaluate specimen/material-dependent 1°/2°LC-PCR assay precision, identical EV RNA preparations of tenfold serial dilutions of cell culture supernatant, EV-spiked negative human plasma (NHP; component of all COBAS AMPLICOR PCR Test kits; Roche), and spiked PCR-negative CSF containing 105 to 10⫺4 TCID50, each, were tested eight times on different days. No differences regarding the application of different sources of EV containing material, i.e., cell culture medium, NHP or CSF, could be observed (data not shown). To validate the relevance of this novel LC-PCR approach in a clinical setting we tested in a comparative study a total of 60 clinical samples (i.e., 12 stool, 38 CSF, 8 serum and 2 throat swabs) obtained from 50 patients with clinical signs of meningitis or symptoms possibly due to EV infections. Compared to virus isolation done from stool specimen, all 6

Table 2 Second and Third European Union Quality Control Concerted Action (EU-QCCA) EV proficiency panela and standardized permanent eV RT-PCR run controls External and internal quality control (QC) testing

Number (No.) of positive results at a given TCID50/no. of tests (% detection rate) Real-time LC/TaqManc

2nd EU-QCCA EV panel [codes: yr. 1999/2000] EV-B05 EV-B06 EV-B07 EV-B11 EV-B08 EV-B01 EV-B09 3rd EU-QCCA EV panel [codes: yr. 2001] EV-C03 EV-C02 EV-C01 EV-C07 EV-C05 EV-C09 EV-C11 EV-C06 EV-C01 Negative run control “Contamination” control “Contamination” control Positive run control Positive run control

Serotyped

b

TCID50

In-house nested PCR

1°LC-PCR

2°LC-PCR

No.(%)

No.(%)

No.(%)

No.(%)

Cox A9 Cox A9 Cox A9 Cox A16 Cox B5 Cox B5 Entero 71

3.6 0.36 0.036 0.25 320 32 560

8/8 (100) 2/8 (25) 0/8 (0) 8/8 (100) 8/8 (100) 8/8 (100) 8/8 (100)

8/8 (100) 6/8 (75) 2/8 (25) 8/8 (100) 8/8 (100) 8/8 (100) 8/8 (100)

7/8 (87.5) 1/8 (12.5) 0/8 (0) 8/8 (100) 8/8 (100) 5/8 (62.5) 8/8 (100)

8/8 (100) 4/8 (50) 0/8 (0) 8/8 (100) 8/8 (100) 8/8 (100) 8/8 (100)

Cox Cox Cox Cox

0.36 0.036 0.0036 320

8/8 (100) 5/8 (62.5) 0/8 (0) 8/8 (100)

8/8 (100) 8/8 (100) 6/8 (75) 8/8 (100)

cNA cNA cNA cNA

8/8 (100) 8/8 (100) 4/8 (50) 8/8 (100)

8/8 (100) 8/8 (100) 4/8 (50) 2/8 (25) 8/8 (100) 0/8 (0) 0/8 (0) 0/8 (0) 8/8 (100) 8/8 (100)

8/8 (100) 8/8 (100) 8/8 (100) 8/8 (100) 8/8 (100) 0/8 (0) 0/8 (0) 0/8 (0) 8/8 (100) 8/8 (100)

cNA cNA cNA cNA cNA 0/8 (0) 0/8 (0) 0/8 (0) 8/8 (100) 8/8 (100)

8/8 (100) 8/8 (100) 8/8 (100) 8/8 (100) 8/8 (100) 2/8 (25) 0/8 (0) 0/8 (0) 8/8 (100) 8/8 (100)

A9 A9 A9 B5

Echo 6 Echo 11 Echo 11 Echo 11 Entero 71 Rhinovirus 14 DEPC-H2Of Neg. CSFf Cox B3 Polio 3

20000 25000 250 25 56 1000 0 0 1000 1000

a Panel tested in duplicate by three reference laboratories using AMPLICOR EV Test (2nd EU-QCCA EV panel; 1999/2000). TaqMan-PCR (3rd EU-QCCA EV panel; 2001) and in-house PCR (2nd and 3rd eV panel) prior to distribution. b 50% Tissue culture-infective doses (TCID50) per mL. c Real-time LightCycler (LC) RT-PCR: single-round LC/TaqMan-protocol (1°LC-PCR); second-round reamplification with primer mix/TaqMan-protocol (2°LC-PCR). d Cox, coxsackievirus; Echo, echovirus; Entero, enterovirus; Polio, poliovirus. f DEPC-H2O, diethylpyrocarbonate treated nuclease-free water; Neg. CSF, EV-PCR negative cerebrospinal fluid g cNA, Amplicor EV Test commercially not available.

T. Watkins-Riedel et al. / Diagnostic Microbiology and Infectious Disease 42 (2002) 99 –105

QC panels

d

AMPLICOR EV Testg

103

104

T. Watkins-Riedel et al. / Diagnostic Microbiology and Infectious Disease 42 (2002) 99 –105

Table 3 EV detection in clinical specimens by RT-PCR and cell culture virus isolation Specimen source/ clinical material (No. ⫽ 60) Stool Patient 1 Patient 2 Patient 3 Patient 4 Patient 5 Patient 6 Patients 7–9 Patients 10–12 CSF Patients 1–3 Patients 13–18 Patients 7–9 Patients 19–21 Patients 22–29 Patient 36 Patients 37–38 Patients 39–40 Patients 41–50 Serum Patients 2⫹4 Patients 7⫹8 Patients 30–33 Throat swab Patient 34 Patient 35

Cell culture EV-Serotypeb

Number (No.) of positive (⫹) or negative (⫺) EV RT-PCR results Real-time LightCyclera 1°LC-PCR

2°LC-PCR

Cox B3 Cox B4 Cox B5 Echo 7 Echo 30 Cox B3 ⫹ Echo30 Negative Negative

⫹(1) ⫹(1) ⫹(1) ⫹(1) ⫹(1) ⫹(1) ⫺(3) ⫺(3)

ND ND ND ND ND Cox B3 Cox B5 Negative Negative

AMPLICOR EV Test

In-house nested PCR

⫹(1) ⫹(1) ⫹(1) ⫹(1) ⫹(1) ⫹(1) ⫺(3) ⫺(3)

⫹(1) ⫹(1) ⫹(1) ⫹(1) ⫹(1) ⫹(1) ⫺(3) cNA

⫹(1) ⫹(1) ⫹(1) ⫹(1) ⫹(1) ND ⫺(3) ⫺(3)

⫹(3) ⫹(6) ⫺(3) ⫺(3) ⫺(8) ⫹(3) ⫹(3) ⫺(3) ⫺(3)

⫹(3) ⫹(6) ⫺(3) ⫺(3) ⫺(8) ⫹(3) ⫹(3) ⫹(3) ⫺(3)

⫹(3) cNA ⫺(3) ⫺(3) cNA cNA cNA cNA cNA

⫹(3) ⫹(6) ⫺(3) ND ⫺(8) ⫹(3) ⫹(3) ⫺(1)/⫹(2) ⫺(3)

ND ND ND

⫹(2) ⫺(2) ⫺(4)

⫹(2) ⫺(2) ⫺(4)

⫹(2) ⫺(2) cNA

⫹(2) ⫺(2) ⫺(4)

ND ND

⫹(1) ⫺(1)

⫹(1) ⫺(1)

cNA cNA

⫹(1) ⫺(1)

a

Real-time LightCycler (LC) RT-PCR: single-round LC/TaqMan (1°LC-PCR); second-round reamplification with primer mix/TaqMan (2°LC-PCR). EV serotypes were determined by neutralizing antisera (obtained from ATCC) after virus propagation in cell culture; Cox, coxsackievirus; Echo, echovirus. c ⫹, EV RT-PCR positive results (No., number of times positive); ⫺, EV RT-PCR negative results (No. numb...