Ed

RL

ARRAA

KHCD(CE

1

ElmflOad

s

c

oMf

0h

Journal of Virological Methods 185 (2012) 228– 233

Contents lists available at SciVerse ScienceDirect

Journal of Virological Methods

jou rn al h om epa ge: www.elsev ier .com/ locate / jv i romet

valuation of two (semi-)nested VP1 based-PCRs for typing enterovirusesirectly from cerebral spinal fluid samples

.P. Minnaar, G. Koen, K. de Haan, K.C. Wolthers, K.S.M. Benschop ∗

aboratory of Clinical Virology, Department of Medical Microbiology, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands

rticle history:eceived 15 March 2012eceived in revised form 28 June 2012ccepted 4 July 2012vailable online 11 July 2012

eywords:uman enterovirusesSFirect genotyping

Semi-)nested PCRodehopV-A and B specific PCR

a b s t r a c t

Human enteroviruses (EVs) are the leading cause of CNS-associated disease in childhood. Identificationof the EV types that patients are infected with is essential for monitoring outbreaks, the emergence ofnew types or variants, epidemiological surveillance and contributes to patient management. Rapid andsensitive molecular detection methods are frequently used to detect EVs/HPeVs directly from CSF. Thisrequires that sensitive EV typing methods from CSF material need to be developed.

In the present study two nested PCR-based typing assays were evaluated. The performance of the EV-Aand -B specific nested PCR protocol and the Codehop-based PCR protocol were analyzed with severalTCID50-titrated EV-A to D strains and 22 EV positive CSF samples.

The EV-A and -B protocol was found to be more sensitive than the Codehop protocol. The Codehopprotocol showed a high degree of aspecific amplification products when run on a gel, and required addi-tional gel purification. The detection limit of the two protocols varied between the types, ranging from0.1 TCID50/mL sample to 106 TCID50/mL sample. From the 22 EV positive CSF samples, 15 (68%) samples

were typed using either protocol. All samples were characterized as members of species B (E30 (9), CAV9(2), E6 (1), E11 (1), E21 (1), E25 (1)). Three samples (E30 (2) and E25 (1)) could only be typed using theEV-B protocol.

In this study, selected EV strains could be typed using both assays at low virus concentrations, typicallyfound in CSF. However, the EV-A and -B protocol was more sensitive than the Codehop protocol forprimary typing of CSF samples.

. Introduction

Human enteroviruses (EVs) belong to the Picornaviridae family.Vs have a small genome of approximately 7500 nucleotides inength. Their genome is composed of a positive single stranded RNA

olecule encoding for a single open reading frame (ORF) that isanked by an untranslated region (UTR) at the 5′ and 3′ ends. TheRF encompasses four capsid proteins (VP4, VP2, VP3, and VP1)nd seven non-structural proteins, such as viral proteases and RNAependent polymerase (2A-2C, and 3A-3Dpol).

Based on phylogenetic analysis, EVs are classified into fourpecies: EV-A to -D (Hyypia et al., 1997; King et al., 2000),

Abbreviations: EVs, enteroviruses; CBVs, coxsackie B viruses; E, echovirus; CAV,oxsackie A viruses; HRV, human rhinovirus; LLOT, lower limit of typing.∗ Corresponding author. Address: Laboratory of Clinical Virology, Department

f Medical Microbiology, Academic Medical Center, University of Amsterdam,eibergdreef 15, 1105 AZ Amsterdam, The Netherlands. Tel.: +31 20 5665619;

ax: +31 206974005.E-mail address: k.s.m.benschop@gmail.com (K.S.M. Benschop).

166-0934/$ – see front matter © 2012 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.jviromet.2012.07.009

© 2012 Elsevier B.V. All rights reserved.

along with human rhinoviruses (HRV) HRV-A to -C (http://www.picornastudygroup.com/).

Members of the species B, including all Echoviruses (E) andcoxsackie B viruses (CBVs) are the most common cause of asep-tic meningitis cases and outbreaks (e.g. E13 and E30) (Bailly et al.,2002; Leonardi et al., 1993; Mullins et al., 2004; Oberste et al.,1999a; Savolainen et al., 2001). In addition, members of speciesA such as EV71 and coxsackie A virus (CAV)16 are responsible forlarge outbreaks of hand, foot and mouth disease associated withsevere neurological complications.

Previously, EV detection and typing from CSF have relied onthe use of virus isolation by cell culture followed by neutraliza-tion against serum pools (Kapsenberg, 1988; Rotbart and Romero,1995; Terletskaia-Ladwig et al., 2008; Van Doornum et al., 2007).However, this method is labor-intensive, time-consuming andnot always sensitive enough for detecting viruses in CSF samples(Rotbart, 2000). No single cell line is optimal for replication ofall EV types, resulting in the requirement of multiple cell lines

(Benschop et al., 2010; Van Doornum et al., 2007). Furthermore,since the serum pools used for serotyping consist of antibodiesagainst known prototypic types, new EV types and variants willnot be identified (Bendig and Earl, 2005).

ologica

t(T2rsdist(sie2

dtpsttaaaEcge

ppeiweat9

sw

2

2

CHsSCtwaT51p1

P

R.P. Minnaar et al. / Journal of Vir

At present, rapid molecular methods with increased sensi-ivity are the preferred method for detection of EVs from CSFEspy et al., 2006; Gorgievski-Hrisoho et al., 1998; Rotbart, 1995;avakoli et al., 2008; Van Doornum et al., 2007; Wolthers et al.,008). In order to detect all known types in a single PCR reaction,eal time PCRs assay have been developed which target the con-erved 5′UTR (Beld et al., 2004; Benschop et al., 2008). However,ue to the conservation in this region, type identification is lim-

ted to this region. Therefore, typing methods generally involveequencing of the capsid gene VP1 (Oberste et al., 1999b). Geno-yping based on this region correlates well with serotyping resultsOberste et al., 1999b). Molecular typing directly from a clinicalample is advantageous since no culture step is required allow-ng for the detection of otherwise unculturable viruses (Benschopt al., 2010; Mirand et al., 2008; Nix et al., 2006; Tavakoli et al.,008).

To type directly from clinical samples various assays have beeneveloped using highly degenerate primers to detect all knownypes in a single assay (Oberste et al., 2003). However, whenrimers attain a high degree of degeneracy this adversely affects theensitivity of the assay. Furthermore, since these primers requirehe use of low annealing temperatures non-specific viral amplifica-ion can be problematic. To overcome this Nix et al. (2006) designed

reverse transcription-seminested PCR (RT-snPCR) reaction, using consensus degenerate hybrid oligonucleotide primer (Codehop)pproach targeting the VP3/VP1 region. Using the Codehop assay,Vs from culture as well as from clinical specimens, including CSF,ould be detected. Since HRVs are classified within the enterovirusenus the assay is therefore also capable of detecting HRVs (Oberstet al., 2010).

Detection methods to type specific EVs directly from CSF sam-les have been developed to reduce the use of highly degeneraterimers (Iturriza-Gomara et al., 2006; Kilpatrick et al., 1998; Mirandt al., 2008; Tavakoli et al., 2008;). Unfortunately, to achieve max-mum assay sensitivity, additional rounds of testing or retesting

ith type specific primers are needed (Mirand et al., 2008; Tavakolit al., 2008). McWilliam Leitch et al. (2009b) have developed a EV-And B specific nested PCR assay to type directly from CSF, avoidinghe use of multiple assays. The detection sensitivity was as high as7%.

To validate the performance of the two EV typing assays (Fig. 1)everal TCID50-titrated EV types and EV positive CSF samples (2010)ere analyzed.

. Materials and methods

.1. Virus strains and clinical specimens

Culture isolates containing EV strains; CAV16 and EV71 (EV-A),AV9, CBV3, E7, 9, 11, and 30 (EV-B) (Benschop et al., 2010), andPeV1 and 3 (Benschop et al., 2006) were obtained from clinical

pecimens (Table 1). HRV16 was a kind gift from Dr. K. van derluijs, Laboratory of Medical Immunology, AMC. Prototype strainsAV13, CAV20, and CAV21 (EV-C); and EV68 (EV-D), including pro-otype strains CAV16 (EV-A); CAV9, CBV3, and E7, 9, 11, 30 (EV-B)ere kindly provided by the Dutch National Institute of Health

nd Environment (RIVM) (Beld et al., 2004; Benschop et al., 2008).he isolates were titrated by Tissue Culture Infectious Dose at0% (TCID50) using the Reed–Muench method (Reed and Muench,938) and characterized by serotyping with EV specific antisera

ools A-G, H-R, CBV, and PV obtained from the RIVM (Kapsenberg,988).

In 2010, 22 CSF samples were found positive for EV by real-timeCR. Samples were stored at −80 ◦C.

l Methods 185 (2012) 228– 233 229

2.2. RNA extraction and 5′UTR real time PCR

Culture isolates (20 �l) and CSF samples (200 �l) were extractedby automatic extraction using the total nucleic acid isolation kitwith the MagnaPure LC instrument (Roche Diagnostics, Almere, theNetherlands). RNA was eluted in 50 �l elution buffer and reversetranscribed as previously described (Benschop et al., 2006). Asinternal control (IC), 10 �l of an Equine Arteritis Virus (EAV) cul-ture (100 TCID50/ml) was co-extracted. For direct genotyping no ICwas added. Five microliters of cDNA was used for real-time PCRusing the LC480 (Roche Diagnostics). EV was detected using anEV-specific duplex assay detecting both EV target and the IC asdescribed previously (Benschop et al., 2008; Jansen et al., 2011).

2.3. Direct genotyping

Using the protocol published by McWilliam Leitch et al. (2009b),6 �l of extracted RNA was amplified by a combined RT and firstround PCR using Superscript III (Invitrogen, Gaithersburg, MD, USA)and outer primers EV-A-OS and EV-A-OAS for EV species A and EV-B-OS and EV-B-OAS (Fig. 1) for species B as published (McWilliamLeitch et al., 2009b). This resulted in the amplification of an 841 bp(A) and 1181 (B) bp fragment encompassing the VP3 and VP1 genes.One microliter of the RT-PCR reaction was used in a second ampli-fication reaction. For species A, the outer forward primers EV-A-OSwas used in combination with the inner reverse primer, EV-A-IAS,in a semi-nested format resulting in a 748 bp fragment encompass-ing the VP3 and VP1 genes (Fig. 1). For species B, the inner nestedprimers EV-B-IS and EV-B-IAS were used as described (McWilliamLeitch et al., 2009b), resulting in a 1085 bp fragment of the VP1 gene(Fig. 1). The amplicons were sequenced directly from PCR usingthe second amplification reaction primers with the BigDye Ter-minator reaction kit on an ABI 3730/3100 DNA analyzer (AppliedBiosystems, Nieuwerkerk a/d IJssel, the Netherlands).

Using the Codehop protocol published by Nix et al. (2006), cDNAsynthesis of extracted RNA was adapted and performed as previ-ously described (Benschop et al., 2006) using random hexamers.Following cDNA synthesis, 5 �l of the RT mixture was used in thefirst PCR using primers 224 and 222, after which 2 �l was used in thesecond PCR with primers AN89 and AN88 (Fig. 1). Both PCR 1 and 2were performed as described (Nix et al., 2006). The first round PCRinvolved the amplification of a 992 bp fragment encompassing theVP3 and VP1 genes. The second round resulted in the amplificationof a 350–400 bp fragment of the VP1 gene (Fig. 1). The ampliconswere purified from 1% agarose gels and sequenced using primersAN89 and AN88 with the BigDye Terminator reaction kit on an ABI3730/3100 DNA analyzer (Applied Biosystems).

2.4. Phylogenetic analysis

Sequences were analyzed utilizing the RIVM genotyping tool(Kroneman et al., 2011) for type characterization. Genetic distanceswere calculated with MEGA5.0 (Kumar et al., 2004) based on JCdistances (Jukes and Cantor, 1969).

3. Results

3.1. Adaption of the assays

The second amplification reaction of the EV-A specific PCR wasadapted to a semi-nested format, using the outer forward primerEV-A-OS instead of the published inner forward EV-A-IS (Fig. 1).

The use of the forward outer primer yielded better amplificationproducts for sequencing.

For the Codehop protocol the cDNA step was adapted to con-form to the random hexamer cDNA method used (Benschop et al.,

230 R.P. Minnaar et al. / Journal of Virological Methods 185 (2012) 228– 233

F l and

t regioi ength

2ttror

3

eECH

ettato

ig. 1. Schematic representation of the primer locations of the EV-A and B protocohe outer primers is given as a broken line. The amplified VP1 region depicting thenner forward primer was omitted and is depicted as a grey arrow. In brackets the l

006). No difference in performance was found when comparingo the originally published protocol (data not shown). In addition,he AN89 and AN88 primers were used in the second amplificationound for sequencing. Use of either the AN232 and AN233 primersr the AN89 and AN88 primers showed no differences in sequencingesults.

.2. Assay sensitivity

The EV-A and EV-B protocol and the Codehop protocol werevaluated on 12 TCID50-titrated cultures of EV strains (CAV16 andV71 (EV-A); CAV9, CBV3, and E7, E9, E11, E30 (EV-B); CAV13,AV20, and CAV21 (EV-C); and EV68 (EV-D)); HPeV1 and 3; andRV16 (Table 1).

The range at which the EV strains could be genotyped byither protocol was analyzed by testing serial 10-fold dilution ofhe TCID50-titrated cultures (Table 1). A sample was considered

ypeable if the amplicons could be visualized on an agarose gelnd sequenced for genotype characterization. The lowest concen-ration at which the PCR was positive is defined as the lower limitf typing (LLOT).

Table 1Lower limit of typing of the EV-A and B protocol and the

type

Virusconcentration(TCID50/mL) 106 10 5

A CAV16 107.0 15.44 21.06EV71 105.2

B 10 CAV9 8.2 22.96 27.8610CBV3 8.0 22.28 25.41

E7 107.6 20.33 23.71 E9 107.3 22.18 23.62E11 106.1 17.46E30 107.9 22.18 26.65

C 10 CAV13 6.5 20.6910CAV20 3.7

CAV21 106.0 21.73

D EV68 105.9 32.49 35.51

HPeV 10 HPeV1 8.1 nd nd 10HPeV3 5.5

HRV HRV16 105.3

Note 1. The Cp values are based on the 5′UTR real time PCNote 2. Blue denotes the typing range of the EV-A and BCodehop protocol. Green denotes the typing range of bot

the Codehop protocol. The amplified VP3-VP1 region depicting the outer region ofn that is used for typing is depicted as a continuous line. For the EV-A protocol the

of the outer–inner region is given.

Both protocols showed a variable LLOT for the different typesfrom the four EV genogroups tested (Table 1). Using the EV-APCR, the CAV16 and EV71 strains could be typed with a LLOT of1 and 100 TCID50/mL sample, respectively. Based on the Cp-valuesderived from the 5′UTR PCR, the two strains could be typed witha Cp value of <33. With the Codehop protocol the CAV16 strainshowed the highest LLOT of 106 TCID50/mL sample (Cp value 15.44)and the EV71 strain could be typed with the same sensitivity as theEV-A PCR.

The EV-B PCR assay showed LLOT ranging between 0.1 and100 TCID50/mL for the CAV9, CBV3, E7, E9, E11 and E30 strains.With the exception of E11, the EV-B strains could be typed at a Cpvalue of 37. Interestingly, the E7 strain could be typed even whenthe 5′UTR PCR was found to be negative. The E11 strain showedthe highest LLOT of 100 TCID50/mL sample, with a Cp value <30.In contrast, the Codehop protocol was 1-log more sensitive for theE11 strain but also with a Cp value of around 30. The LLOT for theCodehop protocol of the other EV-B strains was higher compared

to the EV-B PCR ranging from 10 to 103 TCID50/mL. Similar to theE11 strain, the Cp value was around 30. However, with only 1–2 logdifference in TCID50/mL, the E30 strain could be typed at a CT valueof 37.

Codehop protocol.

104 10 3 10 2 110 10-1

23.86 27.10 29.84 31.04 32.37 32.6826.93 29.46 32.93 36.01 - -

30.53 31.92 33.18 35.39 35.88 36.4327.97 31.24 30.93 32.80 32.90 34.8027.02 30.85 34.25 37.86 - -27.49 30.48 32.81 33.96 35.14 35.2921.80 25.16 28.29 30.78 35.10 37.8529.63 31.92 35.76 37.17 - -

24.18 31.6027.05 33.39 34.0534.09 20.38 23.93 29.09 31.99 34.21

26.36 29.86 33.59 34.82 35.94 37.11

37.20 40.00 - - - -

nd nd ndnd nd ndnd nd ndnd nd nd

nd nd nd nd nd ndR (Benschop et al., 2008).

protocol. Yellow denotes the typing range of theh protocols. Nd – not detected.

R.P. Minnaar et al. / Journal of Virological Methods 185 (2012) 228– 233 231

F and Ci

stCcuctgn

ttEP1T

rsrNCuHdsacs

3

pmb(ti

c

when new antiviral drugs will again become available (Wildenbeestet al., 2010, 2011).

For molecular typing of enterovirus strains in CSF, two (semi-)nested assays recently published by Nix et al. (2006) and

ig. 2. Agarose gel of 22 EV positive CSF samples available for typing using the EV-B,n comparison to the Codehop protocol.

Using both protocols the prototype strains of CAV16 and thepecies B strains, CAV9, CVB3, and E9, E11, E30 (EV-B) could beyped with similar LLOTs (data not shown). However, using theodehop protocol the E7 prototype strain isolated in the mid-60sould not be typed. RNA could be detected in all the dilutionssing the 5′UTR PCR. With the EV-B PCR, the prototype E7 strainould be typed at all dilutions tested. The genetic variation betweenhe E7 clinical and prototype strain was in the same range as theenetic variation of the other clinical and prototype strains (dataot shown).

Species C and D strains were analyzed using the Codehop pro-ocol alone. The CAV20 (EV-C), and EV68 (EV-D) strain could beyped at a low LLOT of 0.1 TCID50/mL. Similar to the E7 strain, theV68 strain was found positive in the typing PCR whilst the 5′UTRCR was negative. The LLOT for the CAV13 and CAV21 strains was03 TCID50/mL, having the highest LLOT of all four EV species tested.he two strains could be typed at a Cp value <30.

The primers from the EV-A and B protocol showed no cross-eactivity to the EV-C and D strains or the HPeV1, HPeV3 and HRV16trains. The primers of the Codehop protocol were found to cross-eact with the HRV16 strain at a low LLOT of 0.1 TCID50/mL sample.o cross reactivity with the HPeV strains was found. Using theodehop protocol a high degree of aspecific amplification prod-cts was observed, as seen in Fig. 2. This was not reduced whenPLC purified primers were used, nor did the HPLC primers pro-uce a better limit of typing (data not shown). Although the cDNAtep of the Codehop protocol was adapted the protocol also showedspecific amplification when tested as described. In contrast, ampli-ons generated by the EV-A and EV-B protocol could be directlyequenced following a second PCR round.

.3. Detection and typing of EV in CSF samples

In 2010, 22 CSF samples were found positive for EV. Fifteen sam-les (68%) were successfully genotyped directly from the clinicalaterial by either the Codehop protocol or the EV-B PCR as mem-

ers of species B as E30 (9), CAV9 (2), E6 (1), E11 (1), E21 (1), E25 (1)Fig. 2). Three of the 16 samples (E25 (n = 1) and E30 (n = 2)) could be

yped using the EV-B protocol alone. None of the samples showednfection with a species A enterovirus.

The EV-B protocol gave a specific band of 1085 bp and the ampli-on generated in the second PCR round could be sequenced without

odehop protocol. In bold are given types identified additionally by the EV-B protocol

any purification steps (Fig. 2). In contrast, additional purificationsteps had to be performed when using the Codehop protocol.

A clear segregation was observed between samples that couldbe genotyped versus samples that remained untyped based on theirCp values. The mean Cp values of samples that could be genotypedwas 33.2 and 32.9 for the EV-B and Codehop protocols, respec-tively. The mean Cp value of samples that could not be genotypedby either assay was 36, however, this was not significant based onthe unpaired T-test (Fig. 3, p = 0.604 (EV-B)/0.319 (Codehop)).

4. Discussion

EVs have a worldwide distribution marked by seasonal peaksthat can vary between the virus species as well as between thetypes (Benschop et al., 2010; Khetsuriani et al., 2006; Roth et al.,2007). Therefore, it is important to determine what types patientsare infected with, in particular, when major outbreaks of meningitisinfections occur (Bailly et al., 2002; Savolainen et al., 2001; Thoelenet al., 2003). Such epidemiological surveillance is of immense valuewhen monitoring for the eradication of polioviruses and especially

Fig. 3. Cp values of EV positive CSF samples that could be typed using the EV-B, andCodehop protocol. Grey circles/triangles (negative by Codehop protocol) indicate 3samples that were found positive by EV-B protocol.

2 ologic

Mo

fssCs

apiuttwtp(pUmt

CaaEef

1f1

ftsi

CWchp

Eatbdtl

E2csSltpistb

32 R.P. Minnaar et al. / Journal of Vir

cWilliam Leitch et al. (2009b) were adopted. The performancef the assays was evaluated on several EV strains of species A–D.

When considering sensitivity the EV-A and -B protocol per-ormed better when compared to the Codehop protocol in thistudy. However, the limiting dilutions of the EV types analyzedhowed that the LLOT can vary between types. For CAV16, CAV9,BV3, and E7, the EV-A and B protocol was found to be more sen-itive than the Codehop protocol.

Both clinical isolates as well as the prototype strains of CAV16nd CAV9, CBV3, E9, E11 and E30 could be detected using eitherrotocol. However, while the E7 clinical isolate could be typed eas-

ly, the prototype E7 strain isolated in 1965 could not be typedsing Codehop. The genetic variation of the prototype E7 strain tohe clinical isolate was similar to that found for the other proto-ype strains and their clinical counterparts. As the first round PCRas negative, it could be speculated that the failure to amplify

he prototype strain was due to primer mismatches of the outerrimers. Based on the sequence generated by the EV-B protocolFig. 1), no significant primer mismatch differences between therototype and the clinical strain for primer 222 were identified.nfortunately, it could not be determined whether primer mis-atches within the outer primer 224 was the reason for the failure

o type the virus strain successfully.For EV71 (EV-A), and E30 (EV-B) the LLOT was similar, while the

odehop protocol was more sensitive for E11 typing. As the EV-And B protocol is group specific, typing of members of species Cnd D could only be analyzed for Codehop. For CAV20 (EV-C) andV68 (EV-D) typing the Codehop protocol performed well, how-ver, for CAV13 and CAV21 (EV-C) the typing limit was at leastourfold higher.

The LLOT for both protocols was determined to be as low as00–1000 TCID50/mL sample, with the exception of CAV16. Thisalls in the viral titre range predicted to be present in CSF of0–1000 TCID50/mL (Rotbart and Romero, 1995).

The LLOT has been analyzed for EV-prototypes that are foundrequently in CSF. Because of the variation in performance, addi-ional testing is needed to assess whether the protocols areufficient to detect other different variants or EV types rarely seenn CSF.

The EV-A and B protocol yielded a better performance than theodehop protocol in terms of amplification products seen on gel.hile the Codehop protocol gave many aspecific smaller ampli-

ons, the EV-A and B PCR gave EV-specific bands. This limitedands-on time as no gel purification was required, and the PCRroduct from the second PCR could be used for sequencing directly.

To analyze the performance of both assays on clinical samples 22V positive CSF samples from 2010 were tested. Neither assay gave

100% typing coverage. A majority of the samples that could not beyped fell above the Cp-value 32 (Benschop et al., 2010). It shoulde noted that molecular typing of a sample based on Cp value is notefinitive, and is merely indicative of its success rate. In addition,he Cp cutoff can vary between types and strains as shown in theimiting dilutions of the clinical isolates.

In concordance with previous studies on summer meningitis,30 was found in a majority of Dutch CSF samples (Bailly et al.,002; Khetsuriani et al., 2006; Leonardi et al., 1993). All 9 strainsould be characterized as genogroup VII (data not shown), respon-ible for a large outbreak in Finland in 2009 (Savolainen et al., 2001).imilar strains were identified in outbreaks across Europe in ear-ier years (Bailly et al., 2009; McWilliam Leitch et al., 2009a). Usinghe EV-B PCR three additional samples could be typed when com-ared to using Codehop PCR. The Cp values of these samples did not

ndicate a low viral load and were well within the range of the E30trains analyzed. In addition, the two E30 strains were not foundo be genetically divergent from the other E30 strains typed withoth protocols.

al Methods 185 (2012) 228– 233

In conclusion, based on the performance of the assays and theclinical evaluation, the EV-A and B protocol can be used successfullyfor identifying EV types in CSF positive samples. Sequentially, theCodehop protocol can be used when samples are negative in the EV-A and B protocol for further identification of virus strains belongingto species C and D.

Funding

This work was supported by a grant from the NetherlandsOrganisation for Health Research and Development’s Clinical Fel-lowship.

Competing interest

None.

Ethical approval

Not required.

Acknowledgements

We thank the Netherlands National Institute of Public Healthand the Environment (RIVM), for providing the prototype EVstrains. We would also like to thank Dr. R. Molenkamp for criticallyreviewing the manuscript.

References

Bailly, J.L., Brosson, D., Archimbaud, C., Chambon, M., Henquell, C., Peigue-Lafeuille,H., 2002. Genetic diversity of echovirus 30 during a meningitis outbreak, demon-strated by direct molecular typing from cerebrospinal fluid. Journal of MedicalVirology 68, 558–567.

Bailly, J.L., Mirand, A., Henquell, C., Archimbaud, C., Chambon, M., Charbonne, F.,Traore, O., Peigue-Lafeuille, H., 2009. Phylogeography of circulating populationsof human echovirus 30 over 50 years: nucleotide polymorphism and signatureof purifying selection in the VP1 capsid protein gene. Infection, Genetics andEvolution 9, 699–708.

Beld, M., Minnaar, R., Weel, J., Sol, C., Damen, M., van der Avoort, H., Wertheim-VanDillen, P.M., van Breda, A., Boom, R., 2004. Highly sensitive assay for detection ofenterovirus in clinical specimens by reverse transcription-PCR with an armoredRNA internal control. Journal of Clinical Microbiology 42, 3059–3064.

Bendig, J., Earl, P., 2005. The Lim Benyesh-Melnick antiserum pools for serotyp-ing human enterovirus cell culture isolates—still useful, but may fail to identifycurrent strains of echovirus 18. Journal of Virological Methods 127, 96–99.

Benschop, K., Minnaar, R., Koen, G., van Eijk, H., Dijkman, K., Westerhuis, B.,Molenkamp, R., Wolthers, K., 2010. Detection of human enterovirus and humanparechovirus (HPeV) genotypes from clinical stool samples: polymerase chainreaction and direct molecular typing, culture characteristics, and serotyping.Diagnostic Microbiology and Infectious Disease 68, 166–173.

Benschop, K., Molenkamp, R., van der Ham, A., Wolthers, K., Beld, M., 2008. Rapiddetection of human parechoviruses in clinical samples by real-time PCR. Journalof Clinical Virology 41, 69–74.

Benschop, K.S., Schinkel, J., Minnaar, R.P., Pajkrt, D., Spanjerberg, L., Kraakman, H.C.,Berkhout, B., Zaaijer, H.L., Beld, M.G., Wolthers, K.C., 2006. Human parechovirusinfections in Dutch children and the association between serotype and diseaseseverity. Clinical Infectious Diseases 42, 204–210.

Espy, M.J., Uhl, J.R., Sloan, L.M., Buckwalter, S.P., Jones, M.F., Vetter, E.A., Yao, J.D.,Wengenack, N.L., Rosenblatt, J.E., Cockerill, F.R., Smith, T.F., 2006. Real-timePCR in clinical microbiology: applications for routine laboratory testing. ClinicalMicrobiology Reviews 19, 165–256.

Gorgievski-Hrisoho, M., Schumacher, J.D., Vilimonovic, N., Germann, D., Matter, L.,1998. Detection by PCR of enteroviruses in cerebrospinal fluid during a summeroutbreak of aseptic meningitis in Switzerland. Journal of Clinical Microbiology36, 2408–2412.

Hyypia, T., Hovi, T., Knowles, N.J., Stanway, G., 1997. Classification of enterovirusesbased on molecular and biological properties. Journal of General Virology 78 (Pt1), 1–11.

Iturriza-Gomara, M., Megson, B., Gray, J., 2006. Molecular detection and characteri-zation of human enteroviruses directly from clinical samples using RT-PCR and

DNA sequencing. Journal of Medical Virology 78, 243–253.

Jansen, R.R., Schinkel, J., Koekkoek, S., Pajkrt, D., Beld, M., de Jong, M.D., Molenkamp,R., 2011. Development and evaluation of a four-tube real time multiplex PCRassay covering fourteen respiratory viruses, and comparison to its correspond-ing single target counterparts. Journal of Clinical Virology 51, 179–185.

ologica

J

K

K

K

K

K

K

L

M

M

M

M

N

O

R.P. Minnaar et al. / Journal of Vir

ukes, T., Cantor, C., 1969. Evolution of protein molecules. In: Munro, H.N. (Ed.),Mammalian Protein Metabolism. Academic Press, New York, pp. 21–123.

apsenberg, J.G., 1988. Picornaviridae; the enteroviruses (polioviruses, coxsack-ieviruses, echoviruses). In: Balows, A., Hausler, W.J., Lennette, E. (Eds.),Laboratory Diagnosis of Infectious Diseases. Principles and Practice. Springer-Verlag, New York, pp. 692–722.

hetsuriani, N., Lamonte-Fowlkes, A., Oberst, S., Pallansch, M.A., 2006. Enterovirussurveillance–United States, 1970–2005. MMWR Surveillance Summaries 55,1–20.

ilpatrick, D.R., Nottay, B., Yang, C.F., Yang, S.J., Da, S.E., Penaranda, S., Pallansch,M., Kew, O., 1998. Serotype-specific identification of polioviruses by PCR usingprimers containing mixed-base or deoxyinosine residues at positions of codondegeneracy. Journal of Clinical Microbiology 36, 352–357.

ing, A.M.Q., Brown, F., Christian, P., Hovi, T., Hyypiä, T., Knowles, N.J., Lemon, S.M.,Minor, P.D., Palmenberg, A.C., Skern, T., Stanway, G., 2000. Picornaviridae. In:Van Regenmortel, M.H.V., Fauquet, C.M., Calisher, C.H., Carsten, E.B., Estes, M.K.,Lemon, S.M., Maniloff, J., Mayo, M.A., McGeoch, D.J., Pringle, C.R., Wickner, R.B.(Eds.), Virus Taxonomy. Classification and Nomenclature of Viruses, SeventhReport of the ICTV. Academic Press, New York, pp. 657–673.

roneman, A., Vennema, H., Deforche, K., Avoort, H., Penaranda, S., Oberste, M.S.,Vinje, J., Koopmans, M., 2011. An automated genotyping tool for enterovirusesand noroviruses. Journal of Clinical Virology 51, 121–125.

umar, S., Tamura, K., Nei, M., 2004. MEGA3: integrated software for molecular evo-lutionary genetics analysis and sequence alignment. Briefings in Bioinformatics5, 150–163.

eonardi, G.P., Greenberg, A.J., Costello, P., Szabo, K., 1993. Echovirus type 30 infec-tion associated with aseptic meningitis in Nassau County, New York, USA.Intervirology 36, 53–56.

cWilliam Leitch, E.C., Bendig, J., Cabrerizo, M., Cardosa, J., Hyypia, T., Ivanova,O.E., Kelly, A., Kroes, A.C., Lukashev, A., MacAdam, A., McMinn, P., Roivainen,M., Trallero, G., Evans, D.J., Simmonds, P., 2009a. Transmission networks andpopulation turnover of echovirus 30. Journal of Virology 83, 2109–2118.

cWilliam Leitch, E.C., Harvala, H., Robertson, I., Ubillos, I., Templeton, K., Sim-monds, P., 2009b. Direct identification of human enterovirus serotypes incerebrospinal fluid by amplification and sequencing of the VP1 region. Journalof Clinical Virology 44, 119–124.

irand, A., Henquell, C., Archimbaud, C., Chambon, M., Charbonne, F., Peigue-Lafeuille, H., Bailly, J.L., 2008. Prospective identification of enterovirusesinvolved in meningitis in 2006 through direct genotyping in cerebrospinal fluid.Journal of Clinical Microbiology 46, 87–96.

ullins, J.A., Khetsuriani, N., Nix, W.A., Oberste, M.S., Lamonte, A., Kilpatrick, D.R.,Dunn, J., Langer, J., McMinn, P., Huang, Q.S., Grimwood, K., Huang, C., Pallansch,M.A., 2004. Emergence of echovirus type 13 as a prominent enterovirus. ClinicalInfectious Diseases 38, 70–77.

ix, W.A., Oberste, M.S., Pallansch, M.A., 2006. Sensitive, seminested PCR amplifica-tion of VP1 sequences for direct identification of all enterovirus serotypes from

original clinical specimens. Journal of Clinical Microbiology 44, 2698–2704.

berste, M.S., Maher, K., Kennett, M.L., Campbell, J.J., Carpenter, M.S., Schnurr, D., Pal-lansch, M.A., 1999a. Molecular epidemiology and genetic diversity of echovirustype 30 (E30): genotypes correlate with temporal dynamics of E30 isolation.Journal of Clinical Microbiology 37, 3928–3933.

l Methods 185 (2012) 228– 233 233

Oberste, M.S., Maher, K., Kilpatrick, D.R., Pallansch, M.A., 1999b. Molecularevolution of the human enteroviruses: correlation of serotype with VP1sequence and application to picornavirus classification. Journal of Virology 73,1941–1948.

Oberste, M.S., Nix, W.A., Maher, K., Pallansch, M.A., 2003. Improved molecular identi-fication of enteroviruses by RT-PCR and amplicon sequencing. Journal of ClinicalVirology 26, 375–377.

Oberste, M.S., Penaranda, S., Rogers, S.L., Henderson, E., Nix, W.A., 2010. Com-parative evaluation of Taqman real-time PCR and semi-nested VP1 PCR fordetection of enteroviruses in clinical specimens. Journal of Clinical Virology 49,73–74.

Reed, L.J., Muench, H., 1938. A simple method of estimating fifty percent endpoints.American Journal of Hygiene 27, 493–497.

Rotbart, H.A., Romero, J.R., 1995. Laboratory diagnosis of enteroviral infections. In:Rotbart, H. (Ed.), Human Enterovirus Infections. ASM Press, Washington, DC, pp.401–418.

Rotbart, H.A., 1995. Enteroviral infections of the central nervous system. ClinicalInfectious Diseases 20, 971–981.

Rotbart, H.A., 2000. Viral meningitis. Seminars in Neurology 20, 277–292.Roth, B., Enders, M., Arents, A., Pfitzner, A., Terletskaia-Ladwig, E., 2007. Epidemi-

ologic aspects and laboratory features of enterovirus infections in WesternGermany, 2000–2005. Journal of Medical Virology 79, 956–962.

Savolainen, C., Hovi, T., Mulders, M.N., 2001. Molecular epidemiology of echovirus 30in Europe: succession of dominant sublineages within a single major genotype.Archives of Virology 146, 521–537.

Tavakoli, N.P., Wang, H., Nattanmai, S., Dupuis, M., Fusco, H., Hull, R., 2008. Detectionand typing of enteroviruses from CSF specimens from patients diagnosed withmeningitis/encephalitis. Journal of Clinical Virology 43, 207–211.

Terletskaia-Ladwig, E., Meier, S., Hahn, R., Leinmuller, M., Schneider, F., Enders, M.,2008. A convenient rapid culture assay for the detection of enteroviruses inclinical samples: comparison with conventional cell culture and RT-PCR. Journalof Medical Microbiology 57, 1000–1006.

Thoelen, I., Lemey, P., van der Donck, I., Beuselinck, K., Lindberg, A.M., van Ranst, M.,2003. Molecular typing and epidemiology of enteroviruses identified from anoutbreak of aseptic meningitis in Belgium during the summer of 2000. Journalof Medical Virology 70, 420–429.

Van Doornum, G.J., Schutten, M., Voermans, J., Guldemeester, G.J., Niesters, H.G.,2007. Development and implementation of real-time nucleic acid amplificationfor the detection of enterovirus infections in comparison to rapid culture ofvarious clinical specimens. Journal of Medical Virology 79, 1868–1876.

Wildenbeest, J.G., Harvala, H., Pajkrt, D., Wolthers, K.C., 2010. The need for treatmentagainst human parechoviruses: how, why and when? Expert Review of Anti-infective Therapy 8, 1417–1429.

Wildenbeest, J.G., van den Broek, P.J., Benschop, K.S., Koen, G., Wierenga, P.C., Vossen,A.C., Kuijpers, T.W., Wolthers, K.C., 2011. Pleconaril revisited: clinical courseof chronic enterovirus meningoencephalitis after treatment correlates with

in vitro susceptibility. Antiviral Therapy 17, 459–466.

Wolthers, K.C., Benschop, K.S., Schinkel, J., Molenkamp, R., Bergevoet, R.M., Spijker-man, I.J., Kraakman, H.C., Pajkrt, D., 2008. Human parechoviruses as an importantviral cause of sepsislike illness and meningitis in young children. Clinical Infec-tious Diseases 47, 358–363.