Journal of Medical Virology 82:494–504 (2010)

Characterization of JC Human PolyomavirusInfection in a Portuguese Population

Ana Matos,1* Vitor Duque,2 Sılvia Beato,3 Joao Poiares da Silva,1 Eugene Major,4

and Antonio Melico-Silvestre5

1Laboratory of Microbiology, Faculty of Pharmacy, University of Coimbra, Coimbra, Portugal2Laboratory of Virology, Infectious Diseases Department, Coimbra’s University Hospital, Coimbra, Portugal3Dr Lopes Dias School of Health, IPCB, Castelo Branco, Portugal4Laboratory of Molecular Medicine and Neuroscience, National Institute of Neurological Disorders and Stroke,National Institutes of Health, Bethesda, Maryland5Infectious Diseases Department, Coimbra’s University Hospital, Coimbra, Portugal

JC virus (JCV) is ubiquitous in the human pop-ulation, infecting children asymptomatically. Afterprimary infection, JCV persists in the host through-out life and is often excreted in the urine. Twohundred thirty-four urine samples and 78 serumsamples, collected from 171 healthy individualsand 63 patients infected with HIV, were used tocharacterize JCV infection in a Portuguese popula-tion. Using PCR, JCV DNA was detected in 38% ofthe urine samples. A significant difference in theexcretion rate was observed between patientsinfected with HIV (51%) and healthy individuals(33%). The frequency of JCV viruria increased withage in healthy individuals, but not in patientsinfectedwithHIV.JCVurinary load wasdeterminedby real-time quantitative PCR and was independentof gender, age, HIV infection, and CD4þ cell count.Overall, the JCV genotype detected most com-monly was 1B, followed by genotypes 2B and 4.The detection and quantitation of JCV-specificantibodies were performed in serum samples byan established enzyme immunoassay (EIA). Anti-bodies to JCV were observed in 91% of the patientstested, irrespective of HIV infection. A positivecorrelation between JCV urinary load and antibodytiters was demonstrated. The present study pro-vides the first characterization of seroprevalenceand urinary excretion of JCV in a Portuguesepopulation and revealed similar results to thoseobserved in other European countries. A compar-ison between healthy individuals and patientsinfected with HIV, despite identical values ofseroprevalence, showed some differences in thepattern of urinary excretion. J. Med. Virol.82:494–504, 2010. � 2010 Wiley-Liss, Inc.

KEY WORDS: JC virus; epidemiology; sero-prevalence; urinary excretion;Portugal

INTRODUCTION

JC virus (JCV) is a member of the genus Polyomavirusfrom the Polyomaviridae family. This polyomavirus isubiquitous in the human population, infecting childrenearly in life and producing antibodies that persistthroughout life [Berger and Major, 1999; Major, 2001].

After primary infection, JCV seems to persist indef-initely. Evidence suggests that JCV may remain latentin a number of sites, notably in the renal tissue and in Blymphocytes [Dorries et al., 1994; Dolei et al., 2000; Poloet al., 2004]. In most healthy adults, reactivation oflatent JCV in the kidney may occur, with progenyviruses excreted into the urine without any associatedclinical symptoms [Chang et al., 2002; Lednicky et al.,2003; Kmieciak et al., 2008].

Although primary infection and persistence of JCVare generally asymptomatic, severe cellular immuno-suppression may induce JCV reactivation and itsdissemination into the central nervous system (CNS).Once in the CNS, JCV could infect lytically oligoden-drocytes and result in the development of progressivemultifocal leukoencephalopathy, a rare fatal demyeli-nating disease, which affects mainly severely immuno-suppressed individuals [Major et al., 1992; Hou andMajor, 2000; Major, 2001].

Before the AIDS epidemic, progressive multifocalleukoencephalopathy was a very rare disease, observedespecially in older individuals and patients withlymphoproliferative disorders. Currently, HIV infectionrepresents the most common underlying cause of

*Correspondence to: Ana Matos, Laboratorio de Microbiologia,Faculdade de Farmacia, Universidade de Coimbra, Polo dasCiencias da Saude, Azinhaga de Santa Comba, 3000-548 Coimbra,Portugal. E-mail: anamatos@ci.uc.pt

Accepted 19 October 2009

DOI 10.1002/jmv.21710

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

� 2010 WILEY-LISS, INC.

immunosuppression associated with the development ofprogressive multifocal leukoencephalopathy, which hasbecome consequently more frequent, affecting up to 5%of AIDS patients [de Luca et al., 1996; Berger and Major,1999; Berger, 2000].

Seroepidemiological studies have shown that morethan 70% of the world population has specific antibodiesto JCV [Major et al., 1992; Rollison et al., 2003; Stoltet al., 2003].

All JCV strains are members of a single serotype[Major, 2001]; however, sequence analysis of the genomehas led to the identification of eight genotypes andvarious subtypes [Agostini et al., 2001b]. The analysis ofa 215-bp DNA sequence on the 50 end portion of the VP1major capsid protein gene proved to be reliable foridentifying the majority of those genotypes and sub-types. Thus, this fragment is used most frequently forJCV genotyping [Ryschkewitsch et al., 2000; Agostiniet al., 2001b; Pagani et al., 2003; Schaffer et al., 2006].JCV is endemic in most regions of the world, but the viralgenotypes vary considerably in different geographicalareas [Sugimoto et al., 1997; Agostini et al., 2001a;Ikegaya and Iwase, 2004]. Although several reportshave described the distribution of the various JCVgenotypes in different countries, no data are availablefor Portugal. Noteworthy, there is also no available dataregarding the incidence of JCV-specific antibodiesamong Portuguese people.

In order to characterize the prevalence of JCV urinaryexcretion in a Portuguese population, urine sampleswere screened for the presence and subsequent quanti-tation of JCV DNA. Genotype identification was carriedout on positive urine samples, to determine its distribu-tion among the population studied. Possible correlationsbetween JCV urinary viral load and the genotype of theexcreted strain, age, gender, and immune status of thevolunteers were also ascertained.

The immune status of Portuguese persons in respectto JCV infection was also determined in the presentstudy. For this purpose, serum samples were assayed forthe presence and quantitation of JCV-specific anti-bodies.

Although JCV is a ubiquitous virus, the clinicalprogression of JCV infection depends on the immunesystem, varying from an asymptomatic persistentinfection to a fatal, although rare, CNS disease,progressive multifocal leukoencephalopathy. SinceHIV infection represents the most frequent conditionassociated with the development of progressive multi-focal leukoencephalopathy [Major et al., 1992; Berger

and Major, 1999], the present study aimed to comparethe pattern of JCV infection in healthy volunteers andthose infected with HIV, in order to determine anydifferences which could be related to the differentevolution of the JCV infection in these patients.

JCV infection was studied in a Portuguesepopulation, to determine the seroprevalence of JCV,and the frequency of JCV viruria and respectivegenotypes in healthy individuals and patients infectedwith HIV.

MATERIALS AND METHODS

Study Subjects and Samples

Two hundred thirty-four Portuguese persons wereenrolled in this study. The subjects were divided into twogroups. The first group included 171 healthy individu-als, comprising blood donors from Jose Maria GrandeHospital, Portalegre, and healthy volunteers attendingthe Clinical Laboratory of the Faculty of Pharmacy atthe University of Coimbra for routine clinical analysis.The second group included 63 patients infected withHIV followed at the Infectious Diseases Department ofCoimbra University Hospital.

Information on age and sex of each individual wascollected. According to age, participants were includedinto one of the seven groups (divided by decades intoages <20, 20–29, 30–39, 40–49, 50–59, 60–69, �70)(Table I). Additional information on CD4þ cell counts atthe time of specimen collection was obtained for patientsinfected with HIV.

The healthy group included 72 men and 99 womenaged between 3 and 75 years (mean age 40.3 years). Thegroup infected with HIV comprised 48 men and 15women, aged between 25 and 75 years (mean age43.6 years). The CD4þ cell count in the patients infectedwith HIV ranged from 6 to 1,325�106 cells/L (median294�106 cells/L).

To investigate JCV shedding, single urine sampleswere collected from all individuals of both groups andstored at �708C until tested.

The seroprevalence of JCV was evaluated usingserum samples obtained at the same time as urinespecimens. These samples were available from 78individuals, 23 healthy volunteers, and 55 patientsinfected with HIV, comprising 27 women and 51 men,aged from 25 to 75 years (mean age 45.9 years). Wholeblood was collected by peripheral venipuncture, and theserum was separated out by low speed centrifugation.Serum samples were stored at �708C until testing.

J. Med. Virol. DOI 10.1002/jmv

TABLE I. Demographic Characterization of the Population Studied

Population group Total

Gender Age groups (years)

F M <20 20–29 30–39 40–49 50–59 60–69 �70

Healthy 171 99 72 12 41 34 27 32 18 7HIV infected 63 15 48 0 7 19 18 14 3 2Total 234 114 120 12 48 53 45 46 21 9

Epidemiology of JC Virus in Portugal 495

Informed consent was obtained from each patient (orparents, if the donors were minors), and the study wasapproved by the Ethics Committee of the CoimbraUniversity Hospital, Portugal.

DNA Extraction

Urinary viral DNA was extracted as describedpreviously [Agostini et al., 1997b, 2001a; Ryschkewitschet al., 2000]. Briefly, 10 ml of urine samples wascentrifuged at 4,300g for 10 min. The cell pellet wasresuspended in 10 ml of sterile phosphate-bufferedsaline and re-centrifuged. The supernatant was dis-carded and viral DNA was then extracted usingthe commercial kit QIAamp1Viral RNA Mini kit(Qiagen1, Izasa, Carnaxide, Portugal) according to themanufacturer’s instructions. Viral DNA was elutedfrom the column with 80 ml of elution buffer and storedat �208C until required for PCR.

Detection of JCV DNA in Urine Samples

For detection and genotyping of JCV DNA in urinesamples a 215-bp fragment was amplified from the VP1major capsid protein gene using the primers describedpreviously: JLP-15 (nt. 1710–1734 50-ACAGTGTGGC-CAGAATTCCACTACC-30) and JLP-16 (nt. 1924–190250-TAAAGCCTCCCCCCCAACAGAAA-30) [Agostiniet al., 2001a,b].

The amplification was carried out in a total volume of100 ml containing: 300 nM of each primer, 2.5 U ofGoTaq1 DNA polymerase (Promega1, VWR Interna-tional, Carnaxide, Portugal), 20 ml of 5� ColorlessGoTaq1 Flexi PCR buffer (Promega1, VWR Interna-tional), 200 mM of each dNTP (Promega1, VWRInternational), 1.5 mM of MgCl2 (Promega1, VWRInternational), and 10 ml of DNA extract. The reactionvolume was adjusted to 100 ml with PCR-grade water.

After an initial denaturation at 958C for 5 min, thesamples were subjected to 50 cycles of a three-step PCRprogramme: 1 min at 958C, 1.5 min at 638C, and 1 min at728C. These cycles were followed by an additionalprimer extension step for 10 min at 728C.

Amplification products were electrophoresed throughan ethidium bromide-containing 1.2% agarose geland visualized under UV light. A positive result wasindicated by the presence of a 215-bp band.

One positive and two negative controls were includedin each batch of PCRs. One negative control consisted ofthe elute obtained from nucleic acid extraction protocolperformed on sterilized water instead of urine sample.The other negative control consisted of the PCR mixturecontaining water instead of DNA template. Tissueculture supernatant of JCV Mad-4-infected SVG cells,with a concentration of 128 hemagglutination units(HAU)/50ml, was used as the positive control.

The analytical sensitivity of this PCR was evaluatedby testing 10-fold serial dilutions of a quantifiedplasmid containing full-length JCV genome (AdvancedBiotechnologies, Inc., Columbia, MD). Amplification

conditions were the same as for the clinical samplesand controls. All dilutions and samples were assayed induplicate.

DNA Sequencing

The JCV VP1-specific amplification product visual-ized on agarose gel was excised and purified usingQIAquick1 Gel Extraction Kit (Qiagen1, Izasa) inaccordance with the procedure described by the manu-facturer. PCR-amplified fragments were used as atemplate for nucleotide automatic sequencing withprimers JLP-15 and JLP-16.

JCV genotypes were identified by scrutiny of the DNAsequence in the VP1 gene and comparison to prototypesequences of the known genotypes (Table IV), asdescribed previously [Agostini et al., 1996, 2001b;Ryschkewitsch et al., 2000]. DNA sequencing number-ing of the coding region is that of Frisque et al. [1984].

The sequences were analyzed using the Basic LocalAlignment Search Tool (BLAST) program via thewebsite of the National Center for BiotechnologyInformation of the United States (http://ncbi.nlm.nih.gov, accessed on May 2009).

Quantitation of JCV DNA in Urine Samples

Viral load of JCV DNA in urine samples wasdetermined by quantitative real-time PCR performedwith hybridization probes and FRET chemistry usingthe LightCycler 1.5 Instrument (Roche1, Roche Diag-nostica, Amadora, Portugal).

For this purpose a set of two amplification primerswas used (JC.For.2: 50-GGA TGT TGC CTT TAC TTTTAG GGT-30 and JC.Rev.2: 50-TAA AAA GCA TTC TACCTC TGT AAT TGA-30) and two FRET probes (JC.FRET.up: 50-GGG TCC TTC CTT TCT CCT TTT CTT TT-(Fluorescein)-30; and JC.FRET.dn: 50-[Red(640)]-GTTGGG GCC ATC TTC ATA TGC TTC AAG-(phosphate)-30) specific for the sequence of the JCV major capsidprotein gene.

A LightCycler-FastStart DNA Master Plus Hybrid-ization Probes (Roche1, Roche Diagnostica) containingFastStart Taq DNA polymerase, reaction buffer, deox-ynucleoside triphosphates, and MgCl2 was used as thebasis for the reaction mixture.

Each reaction was performed in a final volume of 20 mlcontaining 10 ml of extracted DNA, 10 ml of reactionmixture with 1� LightCycler-FastStart DNA MasterPlus Hybridization Probes (Roche Diagnostica), primersat concentrations of 1.0 mM, and FRET probes atconcentrations of 0.2 mM.

For the thermal cycling protocol an initial 8-minincubation at 958C to activate the DNA polymerase, andmelt double-stranded DNA, was followed by 45 cycles of5 sec at 958C, 10 sec at 628C, and 20 sec at 728C. Meltingcurve analysis was performed following PCR amplifica-tion using LightCycler software.

A standard curve for the quantitation of JCV DNAwas constructed using serial decimal dilutions of aplasmid containing the entire genome of JCV

J. Med. Virol. DOI 10.1002/jmv

496 Matos et al.

(Advanced Biotechnologies, Inc.) ranging from 1.4 to1.4� 105 genomic copies of JCV DNA per PCR reaction.

As in conventional PCR methods, and in real-timePCR assays, one positive and two negative controls(described above) were included in all experiments.

JCV Antibodies Detection by VLP-Based ELISA

An ELISA based on purified virus-like particles(VLPs) was used to detect and quantify specific anti-bodies to JCV in serum samples. JCV VLPs based on themajor capsid protein, VP1, are generated in insect cellstransfected with a recombinant baculovirus expressingthat JCV protein. The recombinant JCV VP1 (rJCVVp1)used as an antigen in the ELISA was kindly provided byPeter Jensen (Laboratory of Molecular Medicine andNeuroscience, NINDS, NIH).

VLP-ELISA was performed in 96-well, flat-bottom,immunolon-4 microtiter plates as described previouslywith minor modifications [Hamilton et al., 2000].Briefly, 100ml of 25,000 HAU/ml of rJCVVp1 antigenwas added to each plate well. The plates were incubatedovernight at 48C to adsorb rJCVVp1 to the plate wellsurface. After washing the wells to remove unadsorbedantigen, blocking buffer was added, and the plate wasincubated for 1 hr at room temperature. Serum samples,starting at 1:40 dilution, were diluted in plate wells inserial fourfold increments and incubated at 378C for30 min and washed. The following reagents were addedsequentially with plate washing between each reagentaddition: biotin-labeled goat anti-human immunoglo-bulin G diluted at 1:30,000, for 1 hr at 378C; peroxidase-labeled streptavidin diluted at 1:75,000, for 30 min at378C; and 3,30,5,50-tetramethylbenzidene substrate for30 min at room temperature in the dark. The substratereaction was stopped by the addition of stop solution,and sample absorbance was measured at 450 nm. Serumsamples with an absorbance of 0.05 higher than that ofappropriated serum controls were considered positive.Serum antibody titer corresponded to the highest seradilution with a positive absorbance result. Serumsamples presenting titers of �1:640 were consideredpositive for antibodies to JCV. All sera were tested induplicate. Two serum samples with different antibodiestiters (1:2,560 and 1:40) were tested in each batch ofdeterminations to determine the reliability of theELISA.

Statistical Analysis

Statistical analysis was performed using the chi-square test (w2) for between-groups comparisons of

categorical variables. If any cell of the contingency tablecontained fewer than five expected observations, theFisher’s exact test was applied. Mann–Whitney U-testwas used for comparison between two unrelated groups,while Kruskal–Wallis test was used for comparison ofmore than two unrelated groups.

For all statistical analyses, differences were consid-ered statistically significant when P< 0.05.

Nucleotide Sequence Accession Numbers

The JCV prototype sequences used in comparisonanalysis were available at GenBank and included(GenBank accession numbers are indicated in brackets):JCV type 1A [J02227] [Frisque et al., 1984]; type 1B[AF015527] [Agostini et al., 1998a], type 2A [AF015531,Agostini et al., 1998b; AB372037], type 2B [AF015532][Agostini et al., 1998b], type 3A [U73500] [Agostini et al.,1997a], and type 4 [AF015528, Agostini et al., 1998a;AF281622, AB048566, AB048565].

The new JCV sequences reported in this study havebeen deposited in the GenBank database under theaccession numbers FJ984323–FJ984329.

RESULTS

Urinary Excretion of JCV

Two hundred thirty-four Portuguese persons, dividedinto two groups: healthy individuals and patientsinfected with HIV, respectively, were examined forurinary excretion of JCV (Table I). Detection of JCV inurine was performed by conventional PCR amplifying a215-bp fragment of the VP1 major capsid protein gene.The sensitivity of this conventional PCR technique wasdetermined by using a commercial quantitative plasmidcontaining full-length JCV genome. The detection limitof the assay was 1.4–14 genome copies per reaction.

The summary of the prevalence of JCV urinaryexcretion in the Portuguese population studied is shownin Table II.

Eighty-nine of the urine samples tested (38%)revealed the presence of JCV DNA. Urinary JCV wasdetected in 33% of healthy individuals; a detection ratesignificantly lower than that observed for patientsinfected with HIV, which was of 51% (P¼0.015).

In healthy individuals, the rate of JCV shedding intourine did not vary between gender, with urinary JCVDNA being detected in 33 (33%) women and 24 (33%)men. However, in patients infected with HIV, JCV DNAwas detected more frequently in urine samples obtained

J. Med. Virol. DOI 10.1002/jmv

TABLE II. Prevalence of JCV Urinary Excretion in a Portuguese Population

Population group No. tested No. positive P-value

No. positive (%)

P-valueMale Female

Healthy 171 57 (33%)0.015

24 (33%) 33 (33%) 1.000HIV infected 63 32 (51%) 28 (58%) 4 (27%) 0.041Total 234 89 (38%) — 52 (43%) 37 (32%) 0.087

Epidemiology of JC Virus in Portugal 497

from men (58%), compared with samples obtained fromwomen (27%) (P¼0.041).

To determine the variation on the frequency of JCVexcretion according to the age of the donor, the donorswere divided into seven age groups: <20, 20–29, 30–39,40–49, 50–59, 60–69, and �70 (Table I). JCV urinaryexcretion was associated with age in healthy individu-als, being more frequent in older age groups (P<0.001)(Fig. 1A). Among patients infected with HIV, despite ahigher frequency of JCV excretion was also observedamong older individuals (>60 years), no significantdifferences related with age were observed on the rate ofJCV excretion (P¼0.341) (Fig. 1B).

A correlation was explored between the shedding ofJCV in urine and CD4þ cell count in patients infectedwith HIV. For this purpose, patients infected with HIVwere classified into three subgroups according to theirCD4þ cell counts: <200, 200–500, and >500�106 cells/L, respectively. JCV DNA was detected in 46% of thepatients with CD4þ cell count below 200�106 cells/L, in54% of the patients with CD4þ cell count ranging from200 to 500�106 cells/L, and in 53% of the patientspresenting more than 500�106 cells/L, revealing thatno significant differences existed in the frequency of JCVurinary excretion in relation to CD4þ cell count inpatients infected with HIV (P¼0.801).

Urinary Viral Load of JCV

JCV urinary viral load was determined by real-timequantitative PCR. A commercial quantitative plasmidcontaining full-length JCV genome was used to con-struct the standard curve as well as to determine thesensitivity of the technique. The detection limit of thisquantitative PCR was found to be 1.4–14 genome copiesper reaction.

Only samples positive by conventional PCR wereassayed by real-time PCR. All of the 89 urine sampleswith detectable JCV DNA were quantified.

JCV urinary viral load ranged from 0.4 to 6.9 log -copies/ml of urine (median 5.81 log copies/ml). Althoughmedian JCV urinary viral load was slightly higher inhealthy individuals (5.27 log copies/ml) than in patientsinfected with HIV (5.06 log copies/ml), the difference inthe excreted JCV DNA quantity was not statisticallysignificant (P¼ 0.647).

In both subject groups, no significant differencesrelated to age (P¼0.654 in healthy group andP¼0.922 in HIV-infected group) or gender (P¼0.663in healthy group and P¼ 0.569 in HIV-infected group)were observed in JCV urinary viral load.

In patients infected with HIV, the concentration ofJCV DNA in the urine was also independent of the CD4þcell count (P¼ 0.466).

Identification of JCV Genotypes

JCV genotypes and subtypes excreted by Portuguesepersons were determined by nucleotide sequence anal-ysis of the 215-bp VP1 fragment amplified from the urinesamples. Sequenced PCR fragments were analyzed byBLAST and compared with a library of JCV sequences(GenBank). JCV genotypes were identified by compar-ison with known DNA sequences of prototype strains.

It was possible to determine the sequence on 87 of the89 JCV isolates, 55 from healthy individuals and 32 frompatients infected with HIV. Sequence analysis of theamplified products confirmed that the DNA sequenceswere specific for JCV and revealed the presence of16 different sequences.

Overall the most prevalent JCV type excreted byPortuguese persons was JCV type 1 (56%), followed bytype 2 (23%), type 4 (16%), and type 3 (5%).

Ninety-eight percent of JCV genotype 1 belonged tosubtype 1B, and only one strain (2%) was identified assubtype 1A. Among the JCV type 2 isolates, 95%of the strains were of subtype 2B and only a singleone (5%) was identified as subtype 2A. All the type3 strains detected were identified as belonging togenotype 3A.

The frequency of different JCV genotypes and sub-types was slightly different between the two groups ofthe individuals studied (Table III). The predominantJCV type excreted by both groups was type 1B,representing 54% and 57% of the JCV isolated strainsfrom healthy and persons infected with HIV, respec-tively. In healthy individuals the second most prevalentgenotypes excreted were types 4 (22%) and 2B (20%),followed by types 3A (2%) and 1A (2%). In contrast,among patients infected with HIV, JCV type 1B wasfollowed in frequency by the genotypes 2B (25%), 3A(9%), 4 (6%), and 2A (3%). However, the differencesobserved in the frequency of genotypes 4 (P¼0.057), 2B(P¼0.586), and 3A (P¼0.105), between healthy indi-viduals and patients infected with HIV did not reach astatistical significant level.

J. Med. Virol. DOI 10.1002/jmv

Fig. 1. Age-dependent prevalence of JCV antibodies and viruria in(A) healthy individuals and in (B) patients infected with HIV.

498 Matos et al.

No significant correlation between excreted JCVgenotype and viral load in urine samples was detected(P¼0.295). Also, no significant differences wereobserved in JCV genotype excreted according to gender(P¼0.695) or age (P¼ 0.155) of the subjects sheddingJCV into urine. However, the analysis of each genotypeseparately, and independently of HIV infection, showedthat individuals excreting JCV type 3A were signifi-cantly younger than those excreting JCV type 1B(P¼0.013), type 2B (P¼0.007), or type 4 (P¼0.034).

The alignment of the JCV VP1 sequences found in theurine samples and the comparison with previouslypublished sequences are shown in Table IV. Severalpoint mutations were detected, some of which have beenpreviously described in other JCV strains.

JCV VP1 sequences from the majority of genotype 1Bstrains isolated from Portuguese individuals aligned toGenBank accession number AF015527, described byAgostini et al. [1998a]. A new variant of subtype 1B witha nucleotide substitution (C!T) at position 1834 wasfound in JCV strains excreted by two healthy individ-uals (GenBank accession no. FJ984323). Two patientsinfected with HIV excreted another new variant of

subtype 1B with a substitution (T!C) at position 1819(GenBank accession no. FJ984324). One genotype 1Bstrain detected in a healthy individual showed two newnucleotide substitutions at positions 1804 (T!A) and1810 (G!A) (GenBank accession no. FJ984325).Another healthy individual excreted another newvariant of JCV genotype 1B with nucleotide substitu-tions at positions 1790 (T!G) and 1822 (G!A)(GenBank accession no. FJ984326). Another variant ofgenotype 1B with a nucleotide substitution at position1858 (C!T) was detected in the urine of one healthyindividual (GenBank accession no. FJ984327). All thesefive new genotype 1B variants have not been previouslydescribed.

Only one strain of JCV genotype 2A was found amongthe individuals studied. This strain was detected in apatient infected with HIV who presented a pointmutation (T!C) at position 1794, which aligned withGenBank accession number AB372037, described in theurine of a non-immunocompromised person fromGreece.

Among JCV isolates belonging to genotype 2B, allbut one aligned with GenBank accession number

J. Med. Virol. DOI 10.1002/jmv

TABLE III. JCV Genotype Prevalence in a Portuguese Population

Population groupNo. of JCV isolates

examined

No. of isolates belonging to the genotype (%)

1A 1B 2A 2B 3A 4

Healthy 55 1 (2%) 30 (54%) — 11 (20%) 1 (2%) 12 (22%)HIV infected 32 — 18 (57%) 1 (3%) 8 (25%) 3 (9%) 2 (6%)Total 87 1 (1%) 48 (55%) 1 (1%) 19 (22%) 4 (5%) 14 (16%)

TABLE IV. Identification of JCV Genotypes and Subtypes

JCV No.

Nucleotide position

1753 1756 1771 1786 1790 1794 1804 1810 1818 1819 1822 1834 1837 1843 1850 1851 1852 1858 1869 1870

Type 1A(J02227)

1 A C C G T T T G G T G C T G A C T C G G

Type 1B(AF015527)

41 A C C G T T T G G T G C T T G C T C G G

FJ984323 2 � � � � � � � � � � � T � � � � � � � �FJ984324 2 � � � � � � � � � C � � � � � � � � � �FJ984325 1 � � � � � � A A � � � � � � � � � � � �FJ984326 1 � � � � G � � � � � A � � � � � � � � �FJ984327 1 � � � � � � � � � � � � � � � � � T � �

Type 2A(AF015531)

0 A C A G T T T G C T G C T T A C T C G A

AB372037 1 � � � � � C � � � � � � � � � � � � � �Type 2B

(AF015532)18 A C A T T T T G C T G C C T G C T C G A

FJ984328 1 � T � � � � � � � � � � � � � � � � � �Type 3A

(U73500)3 T C A A T T C G C T G C T T A C T C C A

FJ984329 1 � � � � � � � � � � � � � � � � C � � �Type 4

(AF015528)8 A C C G T T T G C T G C T G A C C C C A

AF281622 3 � � � � � � � � � � � � � � G � � � � GAB048565 2 � � � � � � � � � � � � � � � G � � � �AB048566 1 � � � � � � � � � � � � � � G � � � � �

Sequences of the prototype for each genotype are indicated in the upper line, in bold type with the number of isolates indicated. The remaining linesrepresent variants in VP1 fragment sequence, showing nucleotides where differences with respect to prototypes were found. The correspondingGenBank accession number is given.

Epidemiology of JC Virus in Portugal 499

AF015532, described by Agostini et al. [1998b]. Theremaining strain represented a new genotype 2Bvariant with a nucleotide substitution (C!T) atposition 1756 (GenBank accession no. FJ984328).

Three of the four isolated JCV type 3A strains alignedwith GenBank accession number U73500, described byAgostini et al. [1997a]. The other JCV type 3A strain,isolated from the urine of one patient infected withHIV, revealed a nucleotide substitution (T!C) atposition 1852, which has not been described previously(GenBank accession no. FJ984329).

In 6 of the 14 JCV isolates that belonged to type 4,some nucleotide substitutions described in otherEuropean countries were found. Three strains hadnucleotide substitutions at positions 1850 (A!G) and1870 (A!G) and were found to align with GenBankaccession number AF281622, identified in Germany.Two strains showing a nucleotide substitution (C!G)at position 1851 aligned to GenBank accession numberAB048565, identified in a strain from Netherlands. Onestrain with a nucleotide substitution at position 1850(A!G) aligned with GenBank accession numberAB048566, described in United Kingdom and Spain.The remaining eight JCV type 4 isolated strains alignedwith GenBank accession number AF015528, describedby Agostini et al. [1998a].

Prevalence of JCV Antibodies

To determine the seroprevalence of JCV among thestudied groups of individuals, JCV-specific IgG anti-bodies were quantified using a VLP-based ELISA, asdescribed previously [Hamilton et al., 2000].

Although no standard cut-off values have beenestablished firmly for seropositivity to JCV usingenzyme immunoassay (EIA), Hamilton et al. [2000]indicated a correlation between the positive titer of1:40 using the hemagglutination inhibition methodswith the 1:2,560 result using EIA. However, likeRollison et al. [2003], it was more convenient touse the titer of �1:640 as the cut-off value for sero-positive results. This value was chosen because two ofthe individuals who excreted JCV into urine presentedJCV antibody titers of 1:640.

Serum samples were available for 78 of the 234 sub-jects included in the present study, comprising23 healthy individuals and 55 patients infected withHIV. IgG antibodies to JCV VLPs were detected in 71 ofthe samples, resulting in an overall seroprevalence of91%. The frequency of JCV antibody detection was the

same in both groups of individuals (P¼1.000) and didnot differ with gender (P¼ 1.000) (Table V).

Despite the median age of the seronegative individu-als for JCV being lower than that of seropositivesubjects, the difference observed was not statisticallysignificant (P¼ 0.097). Comparing antibody responsesin different age groups, we found that IgG seropreva-lence in healthy individuals increased with age, reach-ing a maximum value in the age group of 40–49 years,which is maintained across all other age groups(Fig. 1A). However, this variation was not statisticallysignificant (P¼0.250). In patients infected with HIV,despite higher seroprevalence of JCV also beingdetected in older individuals, no significant differenceswere also observed in the frequency of JCV antibodiesaccording to age (P¼ 0.318) (Fig. 1B). Due to the absenceof serum samples from children and adolescents, it wasnot possible to evaluate the age at which the serocon-version usually occurs in Portugal.

JCV antibody titers ranged from 1:640 to 1:1,310,720(median 1:2,560). In healthy individuals, the titerranged from 1:640 to 1:163,840 with a median value of1:10,240, while in patients infected with HIV, theantibody titers varied from 1:640 to 1:1,310,720, with amedian value of 1:2,560. Despite these values, JCVantibody titers were not found to be significantlydifferent between the two groups of subjects (P¼ 0.970).

Fifty-six percent of the individuals seropositive forJCV also excreted the virus into urine.

The comparison of JCV antibody titers according tothe presence of JCV urinary excretion revealed thatindividuals with JCV detectable in urine presentedantibody titers significantly higher than those who donot excrete JCV in the urine (P< 0.001) (Fig. 2).

JCV IgG titers according to the load of JCV shed inurine revealed significantly higher values of JCV anti-body titers in individuals excreting larger amounts ofJCV in urine (P<0.001) (Fig. 3).

DISCUSSION

The present study aimed to characterize JCV infec-tion in Portuguese persons, by determining the fre-quency and pattern of JCV urinary excretion andseroprevalence. As progressive multifocal leukoence-phalopathy is a CNS disease, caused by the reactivationof JCV, which occurs most frequently in immunosup-pressed patients, especially in those infected with HIV, acomparison between healthy individuals and patientsinfected with HIV was also ascertained.

J. Med. Virol. DOI 10.1002/jmv

TABLE V. JCV Seroprevalence in a Portuguese Population

Population group No. tested No. positive P-value

No. positive (%)

P-valueMale Female

Healthy 23 21 (91%)1.000

7 (88%) 14 (93%) 1.000HIV-infected 55 50 (91%) 39 (91%) 11 (92%) 1.000Total 78 71 (91%) — 46 (90%) 25 (93%) 1.000

500 Matos et al.

The overall rate of JCV urinary excretion was foundto be of 38% among all Portuguese persons included inthis study.

Thirty-three percent of healthy individuals haddetectable JCV DNA in their urine. This excretion rateis slightly higher than that published previously fromanother group of healthy Portuguese individuals[Rodrigues et al., 2007]. It is believed that this differencein excretion rate could be related to technical differencesin assays, as well as to demographic characteristics ofthe groups studied. In other European countries, thereported frequency of JCV excretion in immunocompe-tent individuals varied from 21% in Ireland [Schafferet al., 2006] to 47–54% in Spain [Agostini et al., 2001a;Polo et al., 2004], 32% in Germany [Agostini et al.,2001a], 52% in Hungary [Agostini et al., 2001a], 39–46%in Poland [Agostini et al., 2001a; Kmieciak et al., 2008],40–47% in Italy [Ferrante et al., 2001; Pagani et al.,2003; Rossi et al., 2007], and 19% in Switzerland [Egli

et al., 2009]. Thus, the frequency of urinary excretion inthe Portuguese healthy cohort population is within thatreported previously across Europe.

Most of these reports show no significant differencesbetween the frequency of JCV shedding in men andwomen [Kitamura et al., 1990; Ling et al., 2003; Paganiet al., 2003; Zhong et al., 2007; Kmieciak et al., 2008], aswas observed in the present study.

The findings of the present study revealed an ageinfluence on the frequency of JCV urinary excretionamong healthy individuals. Although excretion of JCVwas not observed in the youngest group of donors, thefrequency of excretion increased with age, reaching 50%in the age group of 50–59 years, and of nearly 90% inindividuals older than 70 years. These results are inaccordance with the majority of other reports [Linget al., 2003; Schaffer et al., 2006; Rodrigues et al., 2007;Zhong et al., 2007; Kmieciak et al., 2008; Egli et al.,2009]. Despite this positive correlation, a slight decreasein excretion rate was observed in the age group of 60–69 years, when compared with the age group of 50–59 years. However, this difference was not statisticallysignificant, and a larger number of samples will berequired to confirm this tendency. Agostini et al. [1996]also noted a similar trend in the frequency of JCVexcretion.

The pattern of JCV urinary excretion among Portu-guese patients infected with HIV revealed some differ-ences with respect to those observed in healthyindividuals.

The frequency of JCV urinary excretion in patientsinfected with HIV was significantly higher than thatobserved in healthy subjects. Although this finding is inagreement with earlier studies conducted in Irish andItalian persons infected with HIV [Ferrante et al., 2001;Schaffer et al., 2006], other reports described similarrates of JCV excretion for both patients infected withHIV and healthy individuals [Shah et al., 1997;Lednicky et al., 2003; Engels et al., 2005]. Furthermore,and in contrast to that observed in healthy individuals,excretion rate was significantly higher among malesthan females infected with HIV. No association betweenthe frequency of JCV urinary excretion and age wasnoted. Although these observations are similar to thosereported by Lednicky et al. [2003] and Schaffer et al.[2006], the younger age groups enrolled in the presentstudy included only a small number of patients infectedwith HIV, suggesting that more subjects will be neededto clarify this tendency. The frequency of JCV excretionas well as the urinary JCV load in patients infectedwith HIV was not dependent on the degree of immuno-deficiency as measured by total CD4þ cell count. Thislack of association was also reported in a previousstudy by Behzad-Behbahani et al. [2004] and maysuggest that JCV urinary excretion is not influencedby the immunodeficiency induced by HIV. However,conclusions can only be drawn by examining largerand more representative groups. Noteworthy, studiesconducted by Lednicky et al. [2003] and Schaffer et al.[2006] reported higher frequencies of JCV urinary

J. Med. Virol. DOI 10.1002/jmv

Fig. 3. Box plots showing JCV urinary load according to the antibodytiters for JCV.

Fig. 2. Frequency of JCV antibody titer according to the detection ofJCV in urine samples (P< 0.001).

Epidemiology of JC Virus in Portugal 501

excretion in patients infected with HIV with lowerCD4þ cell counts.

No significant differences on urinary viral load wereobserved between healthy individuals and patientsinfected with HIV excreting JCV. The median viral loadobserved in this cohort population was similar to thatobserved in Swiss healthy blood donors [Egli et al., 2009]and American recipients of lung transplants [Thomaset al., 2007]. Despite the higher frequency of JCVexcretion in older individuals, no association betweenthe urinary viral load and age was observed, as was alsodescribed by Thomas et al. [2007] in lung transplantrecipients. In the present study, the quantity of JCVexcreted into urine was also not influenced by gender.

Several genotypes and subtypes of JCV have beendescribed across the world. The presence and origin ofeach genotype have been related to specific geographicalareas. Type 1 is the major genotype in European andAmerican countries, types 2 and 7 are Asian genotypes,types 3 and 6 have been identified as African in origin[Agostini et al., 1996, 1998b, 2001a; Schaffer et al.,2006], type 4 is found in USA and Europe [Agostini et al.,1998b; Schaffer et al., 2006], type 5 was shown to be arecombinant between types 2B and 6 [Hatwell andSharp, 2000], and JCV type 8 has been described inPapua New Guinea [Whiley et al., 2004; Schaffer et al.,2006]. Although these genotypes and subtypes may haveoriginated from a particular region, human migrationhas led to their dispersal throughout the world [Ledn-icky et al., 2003; Whiley et al., 2004]. For example, inEurope, the most common genotype is type 1, but types4 and 2 are also frequently found [Dubois et al., 2001;Pagani et al., 2003; Schaffer et al., 2006]. Despite suchreports, to date no data were available for Portugal.Therefore, using the results of the DNA sequencing,we were able to determine the incidence and profile ofthe various JCV types and subtypes in the Portuguesepopulation studied. Although the number of urinesamples tested is not representative of the generalpopulation, the distribution of JCV genotypes inPortuguese persons was similar to that observed inother European countries. In Portugal, as in Italy[Pagani et al., 2003; Rossi et al., 2007], Ireland [Schafferet al., 2006], Poland [Agostini et al., 2001a; Kmieciaket al., 2008], Hungary [Agostini et al., 2001a], Germany[Agostini et al., 2001a], and France [Dubois et al., 2001],genotype 1 was the JCV strain detected most frequently.

JCV genotype 1 can be subdivided into two subtypes,types 1A and 1B. The relative proportion of thesesubtypes changes from Eastern to Southwestern Europe[Dubois et al., 2001; Agostini et al., 2001a; Kmieciaket al., 2008]. Type 1A is the predominant JCV genotypein East Europe (Poland and Hungary), whereas type 1Bis more frequent in the Southwest, namely in Spain,France [Jobes et al., 1998; Agostini et al., 2001a], andPortugal, where 98% of the isolated type 1 strains wereidentified as type 1B and represented 55% of the overallisolated strains.

In some European countries, such as Ireland [Schafferet al., 2006], Italy [Pagani et al., 2003], and France

[Dubois et al., 2001], the second most common JCVgenotype detected in healthy individuals is type 4,followed by type 2. In the Portuguese healthy personsenrolled in the present study, both genotypes 2 and4 were detected in almost identical frequencies, asoccurs in Germany [Agostini et al., 2001a].

JCV genotype 2 comprises two major subtypes, 2A and2B. Subtype 2A is present mostly in northeastern Asia.Subtype 2B, which was related originally to the EastAsia, is now also found in Europe and the USA [Agostiniet al., 2001a; Schaffer et al., 2006]. Subtype 2B has beenthe genotype 2 detected most frequently in Europeancountries such as Germany [Agostini et al., 2001a], Italy[Pagani et al., 2003], France [Dubois et al., 2001], andPortugal.

One interesting finding of the present study was thedetection of 5% of JCV type 3, among all isolated strains.A similar finding was also observed by Dubois et al.[2001], in a study carried out in France. JCV genotype3 is characteristically of African origin. Portugal’shistorical relation with several countries of the Africancontinent could be the explanation for this finding.However, a larger number of samples obtained fromindividuals with specific demographic characteristicswill be needed to confirm this possible relationship.

Lednicky et al. [2003] and Schaffer et al. [2006]reported that distribution of JCV genotypes varybetween healthy and HIV-infected individuals. Identi-cal findings were observed in the present study as well.JCV genotypes 1B, 4, and 2B were the most frequentamong healthy individuals, whereas types 1B, 2B, and3 were the most common genotypes identified amongpatients infected with HIV. But the small number ofpatients infected with HIV included in the present studycould be a limitation to reach definitive conclusionsabout this pattern of JCV genotype excretion.

In agreement with Kmieciak et al. [2008] and Paganiet al. [2003], the results of the current study alsorevealed that there are no statistical differences in theexcretion of different JCV genotypes according to genderor age of the subjects. The only exception was individ-uals excreting JCV type 3, which were significantlyyounger than individuals excreting other genotypes.However, only four subjects excreted this genotype, thusa larger number of subjects will be needed to confirm thistendency.

In the present study, seven new variants of the JCVVP1 gene were identified that had not been describedpreviously, some being detected in more than oneindividual. Several investigators argue that urinaryJCV DNA could be used to trace human migrations[Sugimotoetal., 1997].Due to the long historyofPortugalin the discovery of new lands, the analysis of a largernumber of urine samples from Portuguese persons, aswell as from related people living in the formerPortuguese colonies distributed worldwide, could be aninteresting in trying to understand the natural historyand epidemiology of the infection by this virus.

A VLP-based ELISA was used to detect and quantifyIgG-specific antibodies to JCV in serum samples from

J. Med. Virol. DOI 10.1002/jmv

502 Matos et al.

healthy and Portuguese subjects infected with HIV. Theoverall JCV seroprevalence in Portuguese persons was91% and did not differ between the two studied groups.Also, no significant differences were observed in theantibody titers between healthy individuals andpatients infected with HIV.

The observed value of seroprevalence is slightlyhigher than the majority of serological data reportedworldwide (60–80% of adult population present JCV-specific antibodies) [Chang et al., 2002; Stolt et al., 2003;Engels et al., 2005; Rollison et al., 2006; Egli et al., 2009].Differences in seroprevalence rates may be due todifferent cut-off values considered for positive findings,as well as to the different methods used to detectantibodies. Demographic and socioeconomic differencesbetween the populations may also account for differentrates of JCV infection.

The increasing of JCV seroprevalence with age inhealthy individuals has been described in severalprevious reports [Chang et al., 2002; Knowles et al.,2003; Engels et al., 2005; Egli et al., 2009]. Despite notbeing statistically significant, similar data were alsoobserved among the healthy individuals included in thisstudy. JCV IgG seroprevalence increased from 50% inthe youngest group (20–29 years) to 100% for subjectsolder than 40. These findings support the assumptionthat primary infection with JCV occurs early in life, andthe antibodies developed persist throughout life.

In individuals seropositive for JCV, antibody titers didnot vary with age. This observation is different from thatof Egli et al. [2009], which reported increasingly higherIgG JCV titers with age. Differences in group dimen-sions as well as in the methods used, associated with thelack of standardized cut-off values for the definition ofseropositivity, could be the reasons implicated in thedifferent patterns observed.

Knowles et al. [2003] reported that significantly moremales than females had antibodies to JCV, with nodifferences observed in the antibody titers betweengenders. In the present study, as in two other studiesperformed in Ireland [Egli et al., 2009] and the USA[Engels et al., 2005], respectively, the gender of thesubject had no influence on the frequency nor in the titerof JCV antibodies.

As expected, the titers of JCV antibodies weresignificantly higher in the group of individuals excretingJCV in urine. Furthermore, higher JCV antibody titerswere found in individuals excreting superior amounts ofJCV into urine. These findings may suggest that hightiters of IgG antibodies could be a marker for thepresence of active viral replication. In order to confirmthis observation, it would be advisable to submit eachpatient to serial determinations over time, determiningJCV antibody titers and urinary viral load, in order tomonitor its evolution and possible correlation.

ACKNOWLEDGMENTS

The authors wish to thank Ana Donato and LuciliaSilveira for kindly supplied urine and serum samples

from the healthy individuals. The authors would like toexpress their deeply appreciation to Peter Jensen andGary Fahle for sharing their experience with ELISA andReal-Time PCR protocols. The authors are grateful toJose Mario Morgado for his help with the statisticalanalysis and to Cristina Luxo for helpful suggestionsand critical comments on the manuscript.

REFERENCES

Agostini HT, Ryschkewitsch CF, Stoner GL. 1996. Genotype profile ofhuman polyomavirus JC excreted in urine of immunocompetentindividuals. J Clin Microbiol 34:159–164.

Agostini HT, Ryschkewitsch CF, Brubaker GR, Shao J, Stoner GL.1997a. Five complete genomes of JC virus type 3 from Africans andAfrican Americans. Arch Virol 142:637–655.

Agostini HT, Yanagihara R, Davis V, Ryschkewitsch CF, Stoner GL.1997b. Asian genotypes of JC virus in Native Americans and in aPacific Island population: Markers of viral evolution and humanmigration. Proc Natl Acad Sci USA 94:14542–14546.

Agostini HT, Ryschkewitsch CF, Stoner GL. 1998a. JC virus type 1 hasmultiple subtypes: Three new complete genomes. J Gen Virol79:801–805.

Agostini HT, Shishido-Hara Y, Baumhefner RW, Singer EJ, Ryschke-witsch CF, Stoner GL. 1998b. JC virus type 2: Definition of subtypesbased on DNA sequence analysis of ten complete genomes. J GenVirol 79:1143–1151.

Agostini HT, Deckhut A, Jobes DV, Girones R, Schlunck G, Prost MG,Frias C, Perez-Trallero E, Ryschkewitsch CF, Stoner GL. 2001a.Genotypes of JC virus in East, Central and Southwest Europe. JGen Virol 82:1221–1231.

Agostini HT, Jobes DV, Stoner GL. 2001b. Molecular evolution andepidemiology of JC virus. In: Khalili K, Stoner GL, editors. Humanpolyomaviruses: Molecular and clinical perspectives. New York:John Wiley & Sons, Inc. pp 491–526.

Behzad-Behbahani A, Klapper PE, Vallely PJ, Cleator GM, Khoo SH.2004. Detection of BK virus and JC virus DNA in urine samplesfrom immunocompromised (HIV-infected) and immunocompetent(HIV-non-infected) patients using polymerase chain reaction andmicroplate hybridisation. J Clin Virol 29:224–229.

Berger JR. 2000. Progressive multifocal leukoencephalopathy. CurrTreat Options Neurol 2:361–368.

Berger JR, Major EO. 1999. Progressive multifocal leukoencephalo-pathy. Semin Neurol 19:193–200.

Chang H, Wang M, Tsai RT, Lin HS, Huan JS, Wang WC, Chang D.2002. High incidence of JC viruria in JC-seropositive olderindividuals. J Neurovirol 8:447–451.

de Luca A, Cingolani A, Linzalone A, Ammassari A, Murri RR, GiancolaML, Maiuro G, Antinori A. 1996. Improved detection of JC virusDNA in cerebrospinal fluid for diagnosis of AIDS-related progres-sive multifocal leukoencephalopathy. J Clin Microbiol 34:1343–1346.

Dolei A, Pietropaolo V, Gomes E, Di Taranto C, Ziccheddu M, SpanuMA, Lavorino C, Manca M, Degener AM. 2000. Polyomaviruspersistence in lymphocytes: Prevalence in lymphocytes from blooddonors and healthy personnel of a blood transfusion centre. J GenVirol 81:1967–1973.

Dorries K, Vogel E, Gunther S, Czub S. 1994. Infection of humanpolyomaviruses JC and BK in peripheral blood leukocytes fromimmunocompetent individuals. Virology 198:59–70.

Dubois V, Moret H, Lafon ME, Brodard V, Icart J, Ruffault A, Guist’hauO, Bufffet-Janvresse C, Abbed K, Dussaix E, Ingrand D. 2001. JCvirus genotypes in France: Molecular epidemiology and potentialsignificance for progressive multifocal leukoencephalopathy. JInfect Dis 183:213–217.

Egli A, Infanti L, Dumoulin A, Buser A, Samaridis J, Stebler C, GosertR, Hirsch HH. 2009. Prevalence of polyomavirus BK and JCinfection and replication in 400 healthy blood donors. J Infect Dis199:837–846.

Engels EA, Rollison DE, Hartge P, Baris D, Cerhan JR, Severson RK,Cozen W, Davis S, Biggar RJ, Goedert JJ, Viscidi RP. 2005.Antibodies to JC and BK viruses among persons with non-Hodgkinlymphoma. Int J Cancer 117:1013–1019.

Ferrante P, Mediati M, Caldarelli-Stefano R, Losciale L, Mancuso R,Cagni AE, Maserati R. 2001. Increased frequency of JC virus type

J. Med. Virol. DOI 10.1002/jmv

Epidemiology of JC Virus in Portugal 503

2 and of dual infection with JC virus type 1 and 2 in Italianprogressive multifocal leukoencephalopathy patients. J Neurovirol7:35–42.

Frisque RJ, Bream GL, Cannella MT. 1984. Human polyomavirus JCvirus genome. J Virol 51:458–469.

Hamilton RS, Gravell M, Major EO. 2000. Comparison of antibodytiters determined by hemagglutination inhibition and enzymeimmunoassay for JC virus and BK virus. J Clin Microbiol 38:105–109.

Hatwell JH, Sharp PM. 2000. Evolution of human polyomavirus JC. JGen Virol 81:1191–1200.

Hou J, Major EO. 2000. Progressive multifocal leukoencephalopathy:JC virus induced demyelination in the immune compromised host. JNeurovirol 6:S98–S100.

Ikegaya H, Iwase H. 2004. Trial for the geographical identificationusing JC viral genotyping in Japan. Forensic Sci Int 139:169–172.

Jobes DV, Chima SC, Ryschkewitsch CF, Stoner GL. 1998. Phyloge-netic analysis of 22 complete genomes of the human polyomavirusJC virus. J Gen Virol 79:2491–2498.

Kitamura T, Aso Y, Kuniyoshi N, Hara K, Yogo Y. 1990. High incidenceof urinary JC virus excretion in nonimmunosuppressed olderpatients. J Infect Dis 161:1128–1133.

Kmieciak D, Debicki S, Trzeciak WH. 2008. Occurrence rate andgenotype distribution of the JC virus (JCV) in a sample from thePolish population. J Med Virol 80:1079–1083.

Knowles WA, Pipkin P, Andrews N, Vyse A, Minor P, Brown DW, MillerE. 2003. Population-based study of antibody to the humanpolyomaviruses BKV and JCV and the simian polyomavirusSV40. J Med Virol 71:115–123.

Lednicky JA, Vilchez RA, Keitel WA, Visnegarwala F, White ZS,Kozinetz CA, Lewis DE, Butel JS. 2003. Polyomavirus JCVexcretion and genotype analysis in HIV-infected patients receivinghighly active antiretroviral therapy. AIDS 17:801–807.

Ling PD, Lednicky JA, Keitel WA, Poston DG, White ZS, Peng R, Liu Z,Mehta SK, Pierson DL, Rooney CM, Vilchez RA, Smith EOB, ButelJS. 2003. The dynamics of herpesvirus and polyomavirus reactiva-tion and shedding in healthy adults: A 14-month longitudinal study.J Infect Dis 187:1571–1580.

Major EO. 2001. Human polyomaviruses. In: Knipe DM, Howley PM,Griffin DE, Lam RA, Martin MA, Roizman B, Straus SE, editors.Fields virology, Vol. 2. 4th edition. Philadelphia, PA: LippincottWilliams & Wilkins. pp 2175–2196.

Major EO, Amemiya K, Tornatore CS, Houff SA, Berger JR. 1992.Pathogenesis and molecular biology of progressive multifocalleukoencephalopathy, the JC virus-induced demyelinating diseaseof the human brain. Clin Microbiol Rev 5:49–73.

Pagani E, Delbue S, Mancuso R, Borghi E, Tarantini L, Ferrante P.2003. Molecular analysis of JC virus genotypes circulating amongthe Italian healthy population. J Neurovirol 9:559–566.

Polo C, Perez JL, Mielnichuck A, Fedele CG, Niubo J, Tenorio A. 2004.Prevalence and patterns of polyomavirus urinary excretion in

immunocompetent adults and children. Clin Microbiol Infect10:640–644.

Rodrigues C, Pinto D, Medeiros R. 2007. Molecular epidemiologycharacterization of the urinary excretion of polyomavirus inhealthy individuals from Portugal—A Southern European popula-tion. J Med Virol 79:1194–1198.

Rollison DEM, Helzlsouer KJ, Alberg AJ, Hoffman S, Hou J, Daniel R,Shah KV, Major EO. 2003. Serum antibodies to JC virus, BK virus,Simian virus 40, and the risk of incident adult astrocytic braintumors. Cancer Epidemiol Biomarkers Prev 12:460–463.

Rollison DE, Engels EA, Halsey NA, Shah KV, Viscidi RP, HelzlsouerKJ. 2006. Prediagnostic circulating antibodies to JC and BK humanpolyomaviruses and risk of non-Hodgkin lymphoma. CancerEpidemiol Biomarkers Prev 15:543–550.

Rossi A, Delbue S, Mazziotti R, Valli M, Borghi E, Mancuso R, CalvoMG, Ferrante P. 2007. Presence, quantification and characteriza-tion of JC virus in the urine of Italian immunocompetent subjects. JMed Virol 79:408–412.

Ryschkewitsch CF, Friedlaender JS, Mgone CS, Jobes DV, AgostiniHT, Chima SC, Alpers MP, Koki G, Yanagihara R, Stoner GL. 2000.Human polyomavirus JC variants in Papua New Guinea and Guamreflect ancient population settlement and viral evolution. MicrobesInfect 2:987–996.

Schaffer K, Sheey N, Coughlan S, Bergin C, Hall WW. 2006. JC virus inthe Irish population: Significant increase of genotype 2 inimmunocompromised individuals. J Neurovirol 12:39–46.

Shah KV, Daniel RW, Strickler HD, Goedert JJ. 1997. Investigation ofhuman urine for genomic sequences of the primate polyomavirusessimian virus 40, BK virus, and JC virus. J Infect Dis 176:1618–1621.

Stolt A, Sasnauskas K, Koskela P, Lehtinen M, Dillner J. 2003.Seroepidemiology of the human polyomaviruses. J Gen Virol84:1499–1504.

Sugimoto C, Kitamura T, Guo J, Al-Ahdal MN, Shchelkunov SN, OtovaB, Ondrejka P, Chollet JY, El-Safi S, Ettayebi M, Gresenguet G,Kocagoz T, Chaiyarasamee S, Thant KZ, Thein S, Moe K, KobayashiN, Taguchi F, Yogo Y. 1997. Typing of urinary JC virus DNA offers anovel means of tracing human migrations. Proc Natl Acad Sci USA94:9191–9196.

Thomas LD, Vilchez RA, White ZS, Zanwar P, Milstone AP, Butel JS,Dummer S. 2007. A prospective longitudinal study of polyomavirusshedding in lung-transplant recipients. J Infect Dis 195:442–449.

Whiley DM, Arden KE, Mackay IM, Syrmis MW, Sloots TP. 2004.Simultaneous detection and differentiation of human polyoma-viruses JC and BK by a rapid and sensitive PCR-ELAHA assay anda survey of the JCV subtypes within an Australian population. JMed Virol 72:467–472.

Zhong S, Zheng HY, Suzuki M, Chen Q, Ikegaya H, Aoki N, Usuku S,Kobayashi N, Nukuzuma S, Yasuda Y, Kuniyoshi NN, Yogo YY,Kitamura T. 2007. Age-related urinary excretion of BK polyoma-virus by nonimmunocompromised individuals. J Clin Microbiol45:193–198.

J. Med. Virol. DOI 10.1002/jmv

504 Matos et al.