Journal of Medical Virology 71:313–319 (2003)

Molecular Characterisation of Varicella-ZosterVirus Strains in Germany and DifferentiationFrom the Oka Vaccine Strain

A. Sauerbrei,* U. Eichhorn, S. Gawellek, R. Egerer, M. Schacke, and P. Wutzler

Institute of Virology and Therapy, Friedrich-Schiller University of Jena, Jena, Germany

With the introduction of varicella vaccination,surveillance of varicella-zoster virus (VZV) strainsoccurring in cases of chickenpox or zoster shouldbe considered. Differentiating Oka vaccine strainfrom wild-type VZV can be achieved only usingmolecular genotyping. In the present study, theVZV genotype was examined in 53 VZV strainsisolated frompatientswith varicella or zoster andin 73 samples from skin eruptions, cerebrospinalfluid, and throat swabs obtained from patientswith VZV infections in Germany. The polymerasechain reaction and restriction fragment lengthpolymorphisms analysis using DNA fragmentsof the open reading frames 38, 54, 62, and theR5 repeat region were used. Whereas all VZVisolates could be typed, direct genotyping ofviral DNA in patients’ samples was achieved in63 of 73 cases (86.3%). The dominant genotypeof VZV found in 88.8% of 116 patients had thewild-type pattern PstIþ BglI� R5A followed bythe wild-genotype PstIþ BglIþ R5A in 6.0%, thewild-genotype PstIþ BglI� R5B in 3.4%, the wild-genotype PstIþ BglI� R5C and the Oka vaccinegenotypePstI�BglIþR5B in0.9%ofpatients each.BglI� wild-types were found in 90.7% of patientswith zoster and in 9.3% of patients with varicella.By contrast, theBglIþwild-typewas diagnosed infive patients with varicella and in two patientswith zoster. In conclusion, VZV strains found inGermany are similar to strains circulating in theUnitedStates and theUnitedKingdom.VZVwild-type strains containing a BglI restriction site inORF 54 as well as Oka vaccine strains can rarelybe detected. J.Med.Virol. 71:313–319, 2003.� 2003 Wiley-Liss, Inc.

KEY WORDS: Varicella-zoster virus; genotyp-ing;Okastrains;wild-typestrains

INTRODUCTION

In recent years, the vaccination against the primaryvaricella-zoster virus (VZV) infection has been intro-

duced in some industrialised countries. While vaccina-tion is routine in the United States, Japan, and SouthKorea, other countries such as Germany or the UnitedKingdom (UK) recommend the use of the vaccine ingroups at risk of severe chickenpox, e.g., seronegativeimmunocompromised patients and seronegativewomenwho may be considering pregnancy. In Germany,varicella vaccination has also been recommended foradolescents at the age of 12 to 15 yearswithout history ofchickenpox or vaccination and in seronegative healthworkers [Current vaccination committee of the Robert-Koch Institute, 2001].

With the introduction of varicella vaccination, sur-veillance of VZV strains in cases of chickenpox or zosterappears to be necessary.Differentiating such cases frominfections with wild-type VZV is important epidemiolo-gically and can be achieved only using moleculargenotyping methods. Most of these genotyping systemshave been developed in Japan and the United States.Techniques reported commonly are based on amplifica-tion and restriction fragment length polymorphisms(RFLP) analysis using DNA fragments of the openreading frames (ORF) 38, 54, 62, the R5 repeat region,and theR2 repeat region [LaRussaet al., 1992;Hawramiet al., 1996; Loparev et al., 2000]. In the United States,especially outside Japanese communities, wild-type andvaccine strains are distinguished by amplification ofVZV DNA across a region in ORF 38 that is known to bepolymorphic for a PstI restriction site [LaRussa et al.,1992; Mori et al., 1998]. Studies from the UnitedKingdom [Hawrami and Breuer, 1997] and the UnitedStates [Gershon and Takahashi, 1999] revealed thePstIsite in all wild-type strains investigated. Furthermore,

*Correspondence to: Dr. Andreas Sauerbrei, Institute ofVirology and Therapy, Friedrich-Schiller University of Jena,Winzerlaer Straße 10, D-07745 Jena, Germany.E-mail: Andreas.Sauerbrei@med.uni-jena.de

Accepted 28 April 2003

DOI 10.1002/jmv.10485

Published online in Wiley InterScience(www.interscience.wiley.com)

� 2003 WILEY-LISS, INC.

using amplification of a DNA fragment located inORF 54, wild-type VZV can be differentiated in BglIþ

and in BglI� strains. About 20% of wild-type strainsin the United States and the United Kingdom havebeen found to be positive for the BglI restriction site[Hawrami and Breuer, 1997; Gershon and Takahashi,1999] which can also be detected in Oka-like viralstrains. In the United Kingdom,BglIþwild-type strainswere found with increasing frequency over time amongcases of varicella. Additionally, this genetic variationwas strongly associated with zoster in persons who hadimmigrated from countries with low adult immunity tovaricella [Hawrami et al., 1997].

Characterisation of the PstI marker in ORF 38 hasbeen limited to wild-type strains from Japan andprobably some other Asian regions as well as Japanesecommunities outside Japan due to the circulation ofviruses related to the Oka vaccine strain. In comparisonto the Oka parental virus, some mutations have beenreported in ORF 62 of the vaccine strain [Gomi et al.,2001]. Thus, Oka-like and non-Oka-like wild-typestrains can be discriminated from Oka vaccine strainby amplification and SmaI cleavage of ORF 62 frag-ments [Loparev et al., 2000]. This method might beuniversally applied for discrimination between Okavaccine VZV and other VZV strains.

In Germany, the molecular epidemiology of VZVinfections including ORF 38, 54, 62, and the R5 repeatregion has so far not been investigated. Therefore, theaim of the present study was to analyse the VZVgenotype in 53 VZV strains isolated from patients withvaricella or zoster. Furthermore, VZV should be typeddirectly in 73 samples from patients with VZV infec-tions. The results obtained should provide informationabout the distribution of VZV genotypes found in VZVinfections in Germany.

MATERIALS AND METHODS

Varicella-Zoster Virus Strains

In this study, 53 VZV strains isolated from vesiclefluid of patients between 1997 and 2001 were included.Specimens from the patients were submitted to theGerman Reference Centre for a-herpesviruses for diag-nosing VZV infection. The technique of viral isolationand typing has been described previously [Sauerbreiet al., 1999]. Forty-four strains were obtained frompatients with zoster and 9 strains were isolated frompatients with varicella. The ages of patients with zosterranged between 2 and 90 years (mean: 53 years) and thepatients with varicella were between 5 months and39 years old (mean: 11 years). An underlying immuno-suppressive disease was known in 16 cases (12 withzoster and 4 with varicella). One patient developedzoster after varicella vaccination. No information wasavailable from the other patients about their vaccina-tion history against varicella.

The following VZV prototype strains from the collec-tion of the German Reference Laboratory for VZVInfections were used as controls:

1. VZV wild-type strain YS/R.2. VZV OKA strain American Type Culture Collection

(ATCC) VR-795.3. VZV OKA vaccine strain I. This strain was isolated

from the vaccine Varilrix1 distributed by Smith-Kline Beecham, Munich, Germany, in 1991.

4. VZV OKA vaccine strain II. This strain was isolatedfrom the vaccine Varilrix1 (Ch.-B.: VA 212 A44B-1)produced by GlaxoSmithKline, Uxbridge, UK, in1999.

All VZV strains were propagated in human thyroidcells or human embryonal lung fibroblasts. The thyroidcells were maintained in a mixed culture mediumconsisting of 50% lactalbumin hydrolysate (GibcoTM,Invitrogen, Paisley, UK), 40% medium 199 Hanks(Biochrom, Berlin, Germany), and 10% Leibovitz’sL15 medium (Biochrom). For the cultivation of lungfibroblasts, a mixed culture medium consisting of 50%lactalbumin hydrolysate (GibcoTM) and 50% Leibovitz’sL15 medium was used. Both media were supplementedwith 10% fetal calf serum, 100,000 IU/l penicillin and0.1 g/l streptomycin sulfate.

Clinical Specimens

In addition, specimens from 73 patients who haddeveloped VZV infections were investigated. Thesespecimens collected consecutively between 1999 and2001 were sent to the German Reference Centre fora-herpesviruses for diagnosing VZV infection. Samplesfrom skin eruptions were taken from 68 patients, cere-brospinal fluid was obtained from three patients andthroat swabs were taken from two patients. In general,65 patients suffered from zoster and in 8 cases varicellawas diagnosed clinically. The ages of the patients withzoster were 2–88 years (mean: 38 years) and the ages ofpatients with varicella ranged between 4 months and15 years (mean: 7 years). The distribution of differentage groups including patients from whom VZV wasisolated is summarised in the Table I. An underlying

TABLE I. Distribution of Age Groups in Patients Enrolled inThis Study

Age groups(years)

Number of patents (n¼ 126)

Herpes zoster(n¼ 109)

Varicella(n¼ 17)

0–9 4 910–19 12 620–29 9 030–39 7 140–49 5 050–59 19 060–69 23 070–79 14 080–89 8 090–99 1 0Age not known 7 1

314 Sauerbrei et al.

immunodeficiency was known in 25 patients (24 withzoster and 1 with varicella). There was no informationabout vaccination history against varicella.

Using primers of VZV ORF 28 [Sauerbrei et al., 1999]or nested primers of ORF 4 [Puchhammer-Stockl et al.,1991], VZVDNAwasdetected in all patients’ samples bypolymerase chain reaction (PCR). Until these investiga-tions, the speciments were stored at �208C and �808C,respectively.

Polymerase Chain Reaction and RestrictionFragment Length Polymorphisms Analysis UsingDNA Fragments of the Open Reading Frames 38,

54, 62 and the R5 Repeat Region

DNA was prepared and purified from virus-infectedcells or clinical specimens by QIAamp1 Blood Kit(Qiagen, Hilden, Germany). For molecular character-isation of wild-type and vaccine VZV, DNA fragmentsof ORF 38, 54, and 62 were amplified by PCR. Theamplified products were subsequently specified byRFLP analysis. Additionally, length polymorphisms ofthe R5 repeat region (A, B, or C) were determined[Sauerbrei et al., 2003]. To assess the specificity ofthe methods used, DNA of the prototype strains YS/R,Oka ATCC VR-795, as well as the Oka vaccine strainsVarilrix1 I and II served as controls.

Multiplex PCR and RFLP analysis of DNA fragmentsusing ORF 38 and 54 were performed as described byLaRussa et al. [1992]. Briefly, two DNA fragments withthe size of 222 base pairs (bp) (ORF 54) and 350 bp (ORF38) were amplified using two sets of oligonucleotideprimer pairs and were detected by 2% agarose gelelectrophoresis [Sauerbrei et al., 2003]. PCR productswere subjected to restriction endonuclease digestionwith BglI and PstI (Roche Diagnostics, Mannheim,Germany), respectively, and the digests were directlyseparated by electrophoresis in a 4% agarose gel. Ingeneral, VZV wild-type DNA harbours a specific PstIrestriction site for the 350-bp PCR product, which is cutto give a 250-bp fragment and a 100-bp fragment. Bycontrast, the 350-bp PCR-amplified product of Okastrains including Oka vaccine and Oka-like wild strainscannot be digested by PstI endonuclease. BglI digestionof the 222-bp PCR product of Oka strains results in twofragments of 137- and 85-bp size.

A 647-bp fragment was amplified using another setof two specific primers of the ORF 38. The PCR pro-duct and the DNA fragments obtained after digestionwith PstI were analysed directly by 2% agarose gelelectrophoresis [Mori et al., 1998; Sauerbrei et al., 2003].In accordancewith the results described above, theDNAfragment from Oka strains cannot be digested by PstI,whereas the PstI endonuclease cleaves the wild-typeDNA in two fragments of 357- and 290-bp size.

PCR was also carried out with oligonucleotide primerpairs specific for the ORF 62 according to the methoddescribed previously [Loparev et al., 2000; Sauerbreiet al., 2003]. AmplifiedDNA fragmentswith a size of 268bp visualised by 2% agarose gel electrophoresis were

digested subsequently by the endonuclease SmaI andthe generated fragments were differentiated electro-phoretically in a 4% agarose gel. In wild-type VZV in-cluding Oka-like wild-type strains, cleavage withSmaI results in a set of DNA fragments consisting of153-, 79-, and 36-bp fragments. In contrast, DNA ofthe Oka vaccine strain is cleaved in a set of 112-, 79-,and 41/36-bp fragments.

For determination of the tandem direct reiterationof the R5 region, VZV DNA was first amplified usingprimer pairs described by Hawrami et al. [1996]. Theelectrophoretic analysis of amplified products in 2%agarose gel will reveal a main single band whose sizecan be calculated with the following formula, when nrepresents the number of repeating units:

Size of DNA fragment ðbpÞ ¼ 24nþ 88ðnþ 1Þ þ 159:

For confirmation of these results, a second PCRmethoddescribed by Yoshida and Tamura [1999] was used. Theelectrophoretic patterns of amplified products in 0.7%agarose gel consist of multiple bands. The numberof bands reduced by one corresponds to the number ofrepeating units. When n stands for the number ofrepeating units, the size of themain band containing thehighest molecular weight was calculated as (Fig. 1):

Size of DNA fragment ðbpÞ ¼ 24nþ 88ðnþ 1Þ þ 840:

RESULTS

All 53 VZV isolates were typed by PCR techniquesusing primers specific forDNA fragments of theORF38,54,62,andthevariableR5region, respectively (TableII).In comparison, direct genotyping of viral DNA inpatients’ sampleswas only successful in 63 of 73 samples(86.3%)whenprimers ofORF38/54wereused.PCRwithsingle primers of the ORF 38 revealed positive resultsin 62 cases (84.9%) and PCR for amplification of ORF 62was positive in 59 cases (80.8%). With 65 of 73 samples(89%), the highest number of positive results wasachieved by PCR amplification of R5 sequences. Bothmethods used in this study for characterisation of R5repeating units showed comparable findings. In VZVisolates as well as patients’ samples, there were con-siderably more weakly positive results using ORF 62primers compared with themethods for amplification ofDNA fragments of ORF 38, 54 and R5 region. Negativesamples using all genotyping methods included sixsamples from skin eruptions and two samples of cere-brospinal fluid.

Thedominantgenotype ofVZV found in this studyhadthe wild-type pattern PstIþ BglI� R5A (Table III). ThisVZV genotype was detected in 103 of 116 patients(88.8%) at the mean age of 45 years (Table IV). Ninety-four (91.3%) of these patients have developed zoster and9 (8.7%) suffered from varicella. Wild-type VZV char-acterised by the genotype PstIþ BglI� R5B were foundin four patients (3.4%; mean age: 40 years), three withzoster and one with varicella whereas the genotypePstIþBglI�R5Cwas only diagnosed in one patient with

Molecular Differentiation of VZV 315

zoster (0.9%; age: 70 years). In seven cases (6.0%; meanage: 14 years), the wild-genotype PstIþ BglIþ R5A couldbe detected. Five of these patients had chickenpox andtwo patients developed zoster. Using the analysis ofORF 62, 112 of 115 PstIþ-positive strains exhibitedtypical molecular pattern of wild-type VZV. In threePstIþ-positive strains, PCR for amplification of ORF62 revealed negative results. The Oka genotype PstI�

BglIþ R5B could be found in one patient suffering fromzoster 16 months after varicella vaccination at the age

of two years. In this strain, the characteristic pattern ofthe vaccine strainOka could be detected aftermolecularanalysis of ORF 62.

The VZV prototype strains served as controls showedthe following genotypes:

1. VZV wild-type strain YS/R: wild-type PstIþ BglI�

R5A,2. VZV OKA strain ATCC VR-795: Oka vaccine-type

PstI� BglIþ R5B,

Fig. 1. DNA fragments after amplification bypolymerase chain reaction based onR5region.Lanes1,10:Molecular weight standard XIV (100 bp ladder) Roche Diagnostics; lane 2: wild-type VZV, R5A allele;lane 3: Oka vaccine-type isolated from a child with zoster after varicella vaccination, R5B allele; lane 4:wild-type VZV, R5C allele; lane 5: Oka vaccine strain I, R5A allele; lanes 6,7: negative controls; lane 8:VZV Oka vaccine strain II, R5B allele; lane 9: VZV Oka ATCC VR-795, R5B allele.

316 Sauerbrei et al.

3. VZV OKA vaccine strain I: Oka vaccine-type PstI�

BglIþ R5A,4. VZV OKA vaccine strain II: Oka vaccine-type PstI�

BglIþ R5B (Fig. 1).

DISCUSSION

With the introduction of varicella vaccination, amolecular surveillance of circulating VZV strains hasgained more significance. The genetic markers used inour investigationhavebeenconsideredwidely invaccineand epidemiological studies. In the United States, thepolymorphisms of genes 38 and genes 54 have been usedto distinguish VZV vaccine and wild-type strains inclinical isolates obtained from recipients of varicellavaccine who were thought to have vaccine-relatedadverse events [LaRussa and Gershon, 2001; Sharraret al., 2001]. However, the main disadvantage of thistechnique is that it fails to distinguish some Japanesewild-type strains [Hondo et al., 1989]. This appears to bepossible by the analysis of polymorphisms in ORF 62[Loparev et al., 2000]. The characterisation of R5 repeatregion was included since previous results have sug-gested there might be Oka vaccine strains with apolymorph R5 region [Sauerbrei et al., 2003].

Using these DNA fragments, VZV genotype could beverified in all viral strains isolated in cell culture. Bycontrast, direct genotyping of viral DNA in clinicalsamples without viral propagation was only successfulin 63 of 73 (86.3%) cases although VZV DNA could be

detected in all 73 samples by diagnostic PCR technique.These findings can reflect a reduced sensitivity of geno-typing methods in comparison to high-sensitive PCRtechniques established for routine use. Therefore, geno-typing of viral strains isolated in cell culture should beconsidered themethod of choice.When rapid genotypingis essential, direct typing of viral DNA can be triedin patients’ specimens without having to propagatethe virus. In addition, this approach can be useful inpatientswhen fresh vesicle specimens for virus isolationare not available. For amplification of DNA fragmentslocated inORF38, themultiplexPCRwith amplificationof ORF 38 and ORF 54 [LaRussa et al., 1992] should bepreferred to single ORF 38 PCR described by Mori et al.[1998].

As the present results demonstrate, all VZV strainsexcept one were positive for the PstI restriction site ingene 38. This is similar to the strains circulating inUnited States [LaRussa et al., 1992] and the UnitedKingdom [Hawrami and Breuer, 1997] and unlike thewild-type strains circulating in Japan as well as pro-bably in Japanese communities and some other Asianregions. In Japan, up to 30% of wild strains are in-distinguishable from Oka at this marker [Takada et al.,1995]. One strain found in our study did not contain aPstI restriction site in ORF 38 indicating of Oka-likevirus type. By molecular analysis of ORF 62, this straincould be characterised as Oka vaccine virus isolatedfrom zoster occurring 16months after varicella vaccina-tion in a 2-year-old infant [Uebe et al., 2002]. From this,

TABLE II. Polymerase Chain Reaction and Restriction Fragment Length PolymorphismsAnalysis of 53 Varicella-Zoster Virus (VZV) Isolates and Viral DNA in 73 Samples From

Patients With VZV Infections

Results ORF 38/54 ORF 38 ORF 62 R5 region

VZV isolates (n¼ 53)Positive 53 53 51 53Weakly positive 0 0 2 0

Patients’ samples (n¼ 73)Positive 60 60 52 65Weakly positive 3 2 7 0Negative 10 11 14 8

TABLE III. Genotyping of 53 Varicella-Zoster Virus (VZV) Strains and Viral DNA in 63 Samples From PatientsWith VZV Infections

Sample collection/clinical diagnosis

Mean age ofpatients (years)

VZV type/genotype

Wild-type PstIþ

BglI� R5AWild-type PstIþ

BglI� R5BWild-type PstIþ

BglI� R5CWild-type PstIþ

BglIþ R5AOka vaccine-typePstI� BglIþ R5B

VZV isolates(n¼ 53)

45 1 1 5 1

Varicella (n¼ 9) 11 4 0 0 5 0Zoster (n¼ 44) 53 41 1 1 0 1

Patients’ samples(n¼ 63)

58 3 0 2 0

Varicella (n¼ 6) 7 5 1 0 0 0Zoster (n¼ 57) 38 53 2 0 2 0

Summary (n¼ 116) 103 (88.8%) 4 (3.4%) 1 (0.9%) 7 (6.0%) 1 (0.9%)Varicella (n¼ 15) 9 9 1 0 5 0Zoster (n¼ 101) 47 94 3 1 2 1

Molecular Differentiation of VZV 317

it may be concluded that the Oka vaccine strain canonly rarely be detected in patients with VZV-associateddiseases in Germany. This is not surprising since thevaricella vaccine has rarely been administered to datein Germany.

In 6% of all cases, VZV wild-type strains containinga BglI restriction site in ORF 54 like Oka could bedetected. In comparison, 19 to 20% of wild-type virusesbeingpositive forBglI site have been found in theUnitedKingdom and the United States [LaRussa et al., 1992;Hawrami and Breuer, 1997]. The identification of thisBglI polymorphism in clinical isolates before the in-troduction of varicella vaccine indicates that this geneticalteration arose in the United States as the result ofa single-base-pair mutation [LaRussa et al., 1992] andhas not been created by recombination between vaccineand wild-type virus, which has been observed in vitro[Dohner et al., 1988] and in vivo [Shiraki et al., 1991].Because of the rare detection of Oka, such recombina-tion is also unlikely in Germany. Studies in the UnitedKingdom have revealed a high prevalence of BglIþ

viruses among patients with zoster who have immi-grated from Asia, Africa, or the Caribbean in parallelwith a rise in prevalence ofBglIþ strains among cases ofchickenpox. These results have supported the hypo-thesis that BglIþ VZV strains can be imported fromcountries with low adult immunity [Hawrami et al.,1997]. In our study with a smaller number of patients,the country in which the patients had grown up was notestablished. However, the low proportion of BglIþ VZVstrains can reflect a smaller part of immigrants fromcountries with low adult immunity in Germany. Sur-prisingly, five of seven patients with BglIþ strains hadvaricella and only two had zoster. In addition, patientswith BglIþ strains were considerably younger thanpatients with BglI� strains. These findings could in-dicate an increase over time in the prevalence of BglIþ

strains in Germany. To date, it is not known whetherthe BglIþ polymorphism is linked to differences of thepathogenesis and virulence of VZV strains.

Similar to findings in the United Kingdom [Hawramiet al., 1997], the R5A allele of VZV was found mostfrequently and theR5Ballelewas rarely observed in thisstudy. In contrast tofindings in theUnitedKingdom, theR5C allele was also detected, however, in one case only.As previous studies have demonstrated, about 75% ofVZV strains in Japan are type R5B and the remainderhasbeen representedby theallelesR5AandR5C [Hondoand Yogo, 1988; Takada et al., 1995]. In the viral strainisolated from zoster after varicella vaccination, PCR

analysis resulted in target sequences R5B. Comparablefindings were achieved using Oka ATCC strain and theOka strain from Varilrix1 vaccine produced in 1999. Incontrast, the VZV vaccine strain from Varilrix1 pro-duced in 1991 had the typical pattern of R5A. Theseresults confirm our recent findings suggesting theremight be Oka vaccine strains with a polymorph R5region which seems to be not suitable for molecularcharacterisation [Sauerbrei et al., 2003]. In principle,Oka vaccine strains manufactured in Japan have twoR5 repeating units resulting in the allele R5B [Yoshidaand Tamura, 1999]. This type corresponds to the VZVstrain isolated from case of zoster after varicella vac-cination as well as the Oka ATCC strain and the Okastrain from theVarilrix1 vaccine produced in 1999. Okavaccine strains manufactured outside Japan may con-tain one R5 repeating unit resulting in the allele R5A[Hawrami and Breuer, 1997].

In summary, the results suggest that VZV strainsfound in Germany are similar to strains circulating inthe United States and the United Kingdom. The mostdominant genotype is the wild-type PstIþ BglI� R5A.VZV wild-type strains containing a BglI restriction sitein ORF 54 as well as Oka vaccine strains can only bedetected in rare cases.

REFERENCES

Current vaccination committee of the Robert-Koch Institute. 2001.Recommendations of the current vaccination committee of theRobert-Koch Institute for vaccination. Epidemiol Bull 28:203–218.

Dohner D, Adams SG, Gelb LD. 1988. Recombination in tissue culturebetween varicella-zoster virus strains. J Med Virol 24:329–341.

Gershon AA, Takahashi M. 1999. Varicella vaccine. In: Stanley A,Plotkin S, editors. Vaccines. 3rd ed. Philadelphia: Saunders. p 475–507.

Gomi Y, Imagawa T, Takahashi M, Yamanishi K. 2001. Comparison ofDNA sequence and transactivation activity of open reading frame62 of Oka varicella vaccine and its parental viruses. Arch Virol(Suppl) 17:49–56.

Hawrami K, Breuer J. 1997. Analysis of United Kingdom wild-typestrains of varicella-zoster virus: differentiation from the Okavaccine strain. J Med Virol 53:60–62.

Hawrami K, Harper D, Breuer J. 1996. Typing of varicella zoster-virusby amplification of DNA polymorphisms. J Virol Methods 57:169–174.

Hawrami K, Hart IJ, Pereira F, Argent S, Bannister B, Bovill B,Carrington D, Ogilvie M, Rawstorne S, Tryhorn Y, Breuer J. 1997.Molecular epidemiology of varicella-zoster virus in east London,England, between 1971 and 1995. J Clin Microbiol 35:2807–2809.

Hondo R, Yogo Y. 1988. Strain variation of R5 direct repeats in theright-hand portion of the long unique segment of varicella-zostervirus DNA. J Virol 62:2916–2921.

Hondo R, Yogo Y, Yoshida M, Fujima A, Itoh S. 1989. Distribution ofvaricella-zoster virus strains carrying a PstI-site-less mutation inJapan and DNA change responsible for the mutation. Jpn J ExpMed 59:233–237.

TABLE IV. Genotypes of Varicella-Zoster Virus (VZV) and Mean Ages of Patients

VZV genotypeNumber of patients

(n¼ 116)Mean age of patients

(years)

Wild-type PstIþ BglI� R5A 103 45Wild-type PstIþ BglI� R5B 4 40Wild-type PstIþ BglI� R5C 1 70Wild-type PstIþ BglIþ R5A 7 14Oka vaccine-type PstI� BglIþ R5B 1 2

318 Sauerbrei et al.

LaRussa PS, Gershon AA. 2001. Biologic and geographic differencesbetween vaccine and clinical varicella-zoster virus isolates. ArchVirol 17(Suppl):41–48.

LaRussa P, Lungu O, Hardy I, Gershon A, Steinberg SP, Silverstein S.1992. Restriction fragment length polymorphism of polymerasechain reaction products from vaccine andwild-type varicella-zostervirus isolates. J Virol 66:1016–1020.

Loparev VN, Argaw T, Krause PR, Takayama M, Schmid DS. 2000.Improved identification and differentiation of varicella-zoster virus(VZV) wild-type strains and an attenuated varicella vaccine strainusing a VZV open reading frame 62-based PCR. J Clin Microbiol38:3156–3160.

Mori C, TakaharaR, ToriyamaT,Nagai T, TakahashiM,YamanishiK.1998. Identification of theOka strain of the live attenuatedvaricellavaccine from other clinical isolates by molecular epidemiologicanalysis. J Infect Dis 178:35–38.

Puchhammer-Stockl E, Popow-Kraupp T, Heinz FX, Mandl CW, KunzC. 1991. Detection of varicella-zoster virus DNA by polymerasechain reaction in the cerebrospinal fluid of patients sufferingneurological complications associated with chicken pox or herpeszoster. J Clin Microbiol 29:1513–1516.

Sauerbrei A, Eichhorn U, Schacke M, Wutzler P. 1999. Laboratorydiagnosis of herpes zoster. J Clin Virol 14:31–36.

Sauerbrei A, Uebe B, Wutzler P. 2003. Molecular diagnosis of zosterpost varicella vaccination. J Clin Virol 27:190–199.

Sharrar RG, LaRussa P, Galea SA, Steinberg SP, Sweet AR, KeatleyRM, Wells ME, Stephenson WP, Gershon AA. 2001. The post-marketing safety profile of varicella vaccine. Vaccine 19:916–923.

Shiraki K, Horiuchi K, Asano Y, Yamanishi K, Takahashi M. 1991.Differentiation of Oka varicella vaccine strain from wild varicella-zoster virus strains isolated from vaccinees and household contact.J Med Virol 33:128–132.

Takada M, Suzutani T, Yoshida I, Matoba M, Azuma M. 1995.Identification of varicella-zoster virus strains by PCR analysis ofthree repeat elements and a PstI-site-less region. J Clin Microbiol33:658–660.

Uebe B, Sauerbrei A, Burdach S, Horneff G. 2002. Herpes zoster byreactivated vaccine varicella-zoster virus in a healthy child. EurJ Pediatr 161:442–444.

Yoshida M, Tamura T. 1999. An analytical method for R5 repeatedstructure in varicella-zoster virus DNA by polymerase chainreaction. J Virol Methods 80:213–215.

Molecular Differentiation of VZV 319