propagation of the human polyomavirus, jcv, in human

VIROLOGY 228, 269–277 (1997)ARTICLE NO. VY968409

Propagation of the Human Polyomavirus, JCV, in Human Neuroblastoma Cell Lines

TOSHIYA SHINOHARA,* KAZUO NAGASHIMA,† and EUGENE O. MAJOR*,1

*Laboratory of Molecular Medicine and Neuroscience, National Institute of Neurological Disorders and Stroke, NIH, Bethesda, Maryland 20892;and †Department of Pathology, Hokkaido University School of Medicine, Sapporo, Japan

Received June 12, 1996; returned to author for revision July 22, 1996; accepted December 12, 1996

Susceptibility to infection by the human polyomavirus, JCV, is determined by intracellular mechanisms which controltranscription and replication. Originally thought to propagate well only in human cells of oligodendroglial lineage, JCV hasrecently been shown to infect astrocytes, astrogliomas, and a neuroblastoma cell line. The data reported here describetwo cell types that have been subcultured from a human neuroblastoma cell line, SK-N-SH. The SH-SY5Y subclone displaysneuronal phenotypes and is not susceptible to JCV infection, while the SH-EP subclone displays glial cell phenotypes andis susceptible to infection. Binding of nuclear proteins from the permissive SH-EP cells to the nuclear factor-1 (NF-1) sitein the JCV regulatory DNA sequences results in a gel shift pattern that is different from the nonpermissive SH-SY5Y cellproteins. Northern analysis of mRNA for the four classes of NF-1 proteins showed a predominance of the NF-1/X class inSH-EP cells similar to the highly permissive human fetal glial cells. Very low levels of mRNA for NF-1/X were seen in thenonpermissive SH-SY5Y cells, similar to that seen for the nonpermissive HeLa cells. Several other cell lines tested thatwere permissive for JCV infection also showed synthesis of the NF-1/X class of proteins. SH-EP cells represent a cell linein a glial cell lineage which is susceptible to JCV multiplication. These cells may be a useful cell culture system for theinvestigation of DNA binding factors which correlates with viral susceptibility. q 1997 Academic Press

INTRODUCTION reported that the human neuroblastoma cell line (IMR-32)was susceptible to JCV infection (Akatani et al., 1994).

The human polyomavirus JC virus (JCV) is the etiologic In this report we examined JCV susceptibility of theagent of the human demyelinating disease, progressive human neuroblastoma cell line SK-N-SH, two of its sub-multifocal leukoencephalopathy (PML), a condition re- clones SH-EP and SH-SY5Y, and the GOTO and IMR-32sulting from direct lytic infection of the myelin-producing cell lines (Sekiguchi et al., 1979; Tumilowicz et al., 1970).oligodendrocytes in the central nervous system (CNS) The SK-N-SH neuroblastoma cell line is of particular in-(Astrom et al., 1958). Seroepidemiologic surveys of the terest since it is composed of several phenotypes thathuman population indicate that JCV infection is very com- include N- (neuronal) and S- (glial) type cells (Biedler etmon and occurs worldwide (Walker et al., 1983). Despite al., 1978; Ross et al., 1983). Clonal subtypes have beenthis prevalence, PML occurs predominantly in immuno- isolated from the parent SK-N-SH cell line. SH-SY5Y be-compromised patients (Brown, 1975; Rziha, 1978; Padgett longs to the N-type which has neuroblast characteristicsand Walker, 1983). However the site of JCV latency in such as neurotransmitter biosynthetic enzyme activitiesindividuals remains unclear. Recent evidence suggests and catecholamine uptake (Ross et al., 1983; Rettig etbone marrow involvement as well as the kidney as proba- al., 1987; Ciccarone et al., 1989; Tsokos et al., 1989; Grossble sites of viral latency (Major et al., 1992). et al., 1992). SH-EP cells are similar to immature glial

Since its isolation in 1971 from brain tissue of a PML cells (S type), since they are not neuronal but do producepatient, JCV is propagated most efficiently in human fetal type I and III collagens (Ross, 1983; Rettig et al., 1987;glial cells (HFGC) (Padgett et al., 1971). Two human fetal Ciccarone et al., 1989; Tsokos et al., 1989; Gross et al.,glial cell lines, SVG and POJ (Major et al., 1985; Mandl et 1992). Since the restricted tropism of JCV both in vivoal., 1987) (human fetal astrocytes immortalized by SV40 T and in vitro is most likely due to tissue-specific transcrip-protein or JCV T protein, respectively) are also susceptible tion and replication factors that interact with the regula-to JCV infection. Other cell types such as human fetal tory region, we further compared the DNA binding pro-Schwann cells and the B lymphoma cell lines, Namalwa teins to domain B of the viral regulatory region usingand BJA-B, support only limited JCV multiplication. (Assou- nuclear extracts from permissive and nonpermissive cellline and Major, 1991; Atwood et al., 1992). Recently, Akatani lines. In this paper we describe a permissive cell line

having an immature glial cell phenotype which has differ-ent DNA protein binding characteristics to the JCV regu-1 To whom correspondence and reprint requests should be ad-

dressed. E-mail [email protected]. Fax: 301-594-5799. latory region compared to nonpermissive cells.

2690042-6822/97 $25.00Copyright q 1997 by Academic PressAll rights of reproduction in any form reserved.

AID VY 8409 / 6a29$$$$81 02-04-97 20:20:58 viras AP: Virology

brought to you by COREView metadata, citation and similar papers at core.ac.uk

provided by Elsevier - Publisher Connector

https://core.ac.uk/display/82160085?utm_source=pdf&utm_medium=banner&utm_campaign=pdf-decoration-v1

270 SHINOHARA, NAGASHIMA, AND MAJOR

MATERIALS AND METHODS and the HindIII site of pCR-Script (SK) vector was clonedinto the HindIII site of promoterless plasmid pCAT-Basic

Virus and cells Vector (Promega). JC virus minimal regulatory region wascloned into pCR-Script (SK) vector using oligonucleotidesThe 586 strain of JCV (JCV 586) was recovered fromof annealed primers COR TOP 5* TCTAGATTTTTTTAT-owl monkey tumor cells (Major et al., 1987) and maintainsATATACAGGAGGCCGAGGCATGC3 * and COR BOT 5*a Mad-4 regulatory region genotype (Major et al., 1987).GCA TGC CTC GGC CTC CTG TAT ATA TAA AAA AAT C T-U87-MG, SK-N-SH, IMR-32, and HeLa cells were pur-AGA3 *. The fragment between the SphI site within thechased from the American Type Culture Collectionprimers and the HindIII site of the vector was cloned into(ATCC, Rockville, MD). SH-EP and SH-SY5Y cells werepCAT-Basic Vector. Transcription from the JCV promoterkindly provided Fabrizio Ensoli (National Cancer Institute,is in the early orientation.Bethesda, MD). GOTO cells were purchased from the

The plasmid pAT1 was constructed by cloning theJapanese Cancer Research Resource Bank (JCRR). TheEcoRI fragment of the plasmid pBRNF-1AT-1 intoSVG cells were established as an astrocyte cell line de-pcDNA3 (Invitrogen, Sumner et al., 1996).rived from human fetal brain which maintains susceptibil-

ity to JCV infection (Major et al., 1985). These cells wereTransient transfection and CAT assaycultured in Eagle’s minimal essential medium (MEM) or

RPMI 1640 supplemented with 10% fetal bovine serum Cells (6 1 105) were plated on a 60-mm plate and(FBS), L-glutamine (3.0 mg/ml), and gentamicin. The cul- grown overnight. Three hours prior to transfection, thetures were maintained with 5% CO2 at 377. cells were fed with new growth media. Transfections

were carried out with lipofectamin (Life Technologies).Infection and virus passage

Extracts were prepared 48 hr after transfection. The pro-tein concentration of the sample was determined by theVirus inoculations were carried out by the modifiedmethod of Bradford (1976). Equal concentrations of pro-centrifugation–absorption method (Akatani et al., 1994).tein were assayed for CAT enzyme activity using theForty hemaglutination (HA) units per milliliter of JCV werefluor diffusion method of Neuman (1987). Briefly, 50 mgmixed with 1 1 106 cells in serum-free MEM and centri-of protein, diluted in 250 mM Tris–HCL (pH 7.8) wasfuged for 1 hr at 3000 rpm at room temperature. The cellsadded to a reaction mixture containing 0.25 mM chloram-were absorbed with the virus for 1 hr at 377, washedphenicol and 0.5 mCi of 3H-acetyl coenzyme A (DuPont).once with MEM, and seeded into 35-mm plates (Costar,The mix was then overlayed with 5.0 ml liquid scintillationCambridge, MA).cocktail and counted in a b scintillation counter (Beck-For virus passage, infected cells were frozen andman Instruments) for 2.5 hr. Efficiency of transfection wasthawed 31 and then treated with 0.05 mg/ml neurami-monitored with pSV2-b-galactocidase control plasmiddase Type V (Sigma) for 16 hr. The infected cell lysates(Promega) or by slot blot hybridization of DNA isolatedwere treated at 567 for 30 min and centrifuged at 1000gfrom the transfected cells. Three independent trans-for 10 min. The final virus containing supernatant wasfection experiments were performed with duplicatequantitated by hemagglutination (HA of human type Osamples.erythrocytes) (Padgett et al., 1971) and used for virus

inoculations.Immunocytochemical and immunofluorescent analysis

PlasmidsAt various times after infection, cells were washed with

PBS and fixed for 5 min in acetone at 0207. CoverslipsThe plasmid pM1EHL CAT contains the regulatory re-gion of the JCV Mad-1 strain (5113 to 298 bp). The plas- were then incubated with anti-SV40 T antigen antibody

(Pab 416, Oncogene Science, New York) and anti-SV40mid p586EHL CAT contains the regulatory region of theJCV 586 strain and the plasmid pM1EH28 CAT contains VP-1 antibody (provided by Dr. L. Norkin) for 30 min at

377C. The coverslips were washed 31 in PBS and incu-the minimal regulatory region (1 to 28 bp) for the negativecontrol. The cloning procedures utilized the pCAT-Basic bated with secondary anti-mouse or anti-rabbit antibody

conjugated with biotin (Zymed) for 30 min. The complexVector (Promega) as follows: JC virus promoter wascloned into pCR-Script (SK) vector by amplifying the plas- was incubated for 30 min with strepavidin–biotin,

washed, and then mixed with diaminobenzidine (DAB)mid pM1TC and pBRJCV586 as templates for the regula-tory regions of either JCV Mad 1 strain or JCV 586 strain in the presence of H2O2 . T and V antigen-positive JCV-

infected cells were counted by light microscopic exami-respectively. The primers were LCS5*GACACGTCGA-CTC CCT ATT CAG CAC TTT GTC CAC TTT A G CT3 * and nation. For immunofluorescent analysis, coverslips were

fixed with 50% methanol/50% acetone for 5 min at roomLCX 5 * GACAC - TC - TAGATACGTGACAGCTGGCGAAG -AACCTTGGCCA3 *.The fragment which contains the JC temperature, incubated with antibodies, washed 31 with

PBS, and incubated with fluorescein-conjugated mousevirus regulatory region between the HindIII site of JCV


271PROPAGATION OF JCV

TABLE 1or rabbit antibody for 30 min. Positive cells were countedusing a Zeiss, ICM 405 epiflourescent microscope. Yield of JCV in Neuroblastoma Cell Lines

Preparation of nuclear extracts JCV586

Cultures of human fetal glial cells, HeLa, SH-EP, and Cell lines Input HA HA titer a

SH-SY5Y were grown in 125-cm2 culture flasks (Corning).IMR-32 40 2560Nuclear extracts were prepared by a modification of theGOTO 40 320procedure of Schreiber (Schreiber et al., 1989). The pro-U87-MG 40 8

tein concentration of the sample was determined by the SK-N-SH 40 2560method of Bradford (1976). SH-EP b 40 10,240

SH-SY5Y b 40 õ2HeLa 40 õ2Oligonucleotide probe and gel shift assays

The following oligonucleotides were synthesized by a 28 days after infection.b Subclone of SK-N-SH.Gibco-BRL (Gaithersburg, MD). NF-134:5*-GATCTGGAA-

GGGATGG CT GC CA GCCAAGCATGA-3 * and NF-134m:5* - GATCTGGAAGGGATGGCTGCCAGCCAGCCA -

labeled probe was added to the hybridization solution.AGCATGA-3 *. Double-stranded oligonucleotides wereHybridization was carried out at 557 for 20 hr. The filterlabeled with [g-32P]ATP. The labeled probe (11 104 cpm)was washed in two changes of 21 SSC/0.1% SDS atwas incubated with 10 mg of nuclear extract in the pres-room temperature for 15 min each and in two changesence or the absence of a 50-fold molar excess of anof 0.21 SSC/0.1% SDS at 557 for 15 min each. After wash-unlabeled, homologous probe or an unlabeled mutanting, the filter was exposed to Kodak X-Omat film from 1probe. The DNA protein binding reaction also containedto several days.10 mM Tris (pH 7.9), 50 mM NaCl, 5 mM MgCl2, 0.5 mM

EDTA, 1.0 mM DTT, 10% glycerol (v/v), and 4 mg of polyRESULTS(dI-dC) (Pharmacia). The reaction was carried out at room

temperature for 15 min and electrophoresed through a JCV infection in human neuronal and glial tumor cell6% polyacrylamide Tris–glycine gel. The gel was dried, linesand samples were visualized by autoradiography follow-

Forty HA units of JCV 586 per 106 cells were inoculateding exposure of Kodak X-Omat film with an intensifyinginto human neuroblastoma cell lines IMR-32, GOTO, andscreen for 24 hr.SK-N-SH, the human glioma cell line U87-MG, and HeLacells. After 4 weeks there was no significant cytopatho-Probelogic effect (CPE) in any of the infected cultures. Virus

A 464-bp 3 * probe was obtained from BSNF-A which production was determined by HA assay at 4 weekswas subcloned by RT–PCR and then sequenced (Sum- postinfection. HA titers were normalized to 106 cells. Thener et al., 1996). A 1547-bp 3 * probe was obtained from results are shown in Table 1: 2560 HA units in IMR-32;NF-1/Red (ATCC) after digestion of the plasmid with 160 HA units to 320 HA units in GOTO; 2560 HA units inEcoRI. A 1023-bp 3 * probe was obtained from pCTF1 SK-N-SH; 8 HA units in U87-MG; and õ2 HA units (lysate(kindly provided from Dr. Tijian) after digestion of the of 105 cells) in HeLa. The two subclones of the SK-N-SHplasmid with Sal I and EcoRI. A 842-bp 3 * probe was cell line, SH-EP and SH-SY5Y, were also tested for JCVobtained from pBRNF1/AT1 after digestion of the plasmid infectivity. Forty units per 106 cells of JCV 586 were inocu-with Afl I and EcoRI. These fragments were purified from lated into SH-EP and SH-SY5Y and maintained for 4agarose gel and were labeled with [32P]dCTP by the ran- weeks with no significant CPE. However, there was adom primer labeling method. significant difference in virus production between SH-EP

and SH-SY5Y: 10240 HA units per 106 cells in SH-EP butNorthern analysis õ2 HA units of 106 cell lysate in SH-SY5Y (Table 1).

Total RNA was isolated by a modified single-stepSusceptibility in SH-EP compared with that of SVG

method of RNA isolation by acid–guanidium thiocya-cells

nate–phenol–chloroform extraction (Chromczynski andSacchi, 1987). Twenty micrograms of total RNA was elec- The extent of JCV infection in SH-EP cells was com-

pared with SVG cells, an immortalized human fetaltrophoresed through a 1% formaldehyde gel and trans-ferred to nylon filters. The filter was prehybridized in 50% astrocyte cell line (Major, 1985) and examined for virion

multiplication. JCV 586 was inoculated onto coverslips offormamide, 51 SSPE, 51 Denhardt’s, 100 mg/ml herringsperm DNA, 0.1% SDS at 557 for 1 hr. Then a [P32]dCTP- SH-EP, SH-SY5Y, and SVG cells. At 1, 2, and 3 weeks



FIG. 1. Immunocytochemical staining of SH-EP cells for T protein at 21 days postinoculation with JCV 586.

postinfection, coverslips were removed and fixed or cells SH-SY5Y cells at any time point. The extent of JCV multi-plication from either SH-EP or SVG cells was comparablewere harvested for virus assays. Viral T and virion pro-

teins were analyzed using specific antibodies in immuno- as measured by the final HA titer as shown in Table 2.cytochemical assays. Virion multiplication was tested byHA assay. In the SH-EP cells, JCV T protein was detected Identification of antigenic phenotype of SH-EPin 12.6% of the cells by 2 weeks. By 21 days postinfectionwith JCV 586, 30.7% of SH-EP cells were stained for T We established the phenotype of SH-EP and SH-SY5Y

using immunofluorescent analysis with the following an-protein (Fig. 1) and 12.6% stained for V protein (Fig. 2).However, there was no detectable virus production in tibodies: anti-NF, anti-GFA, anti-GC, anti-vimentin, O4,

FIG. 2. Immunocytochemical staining of SH-EP cells for V capsid protein at 21 days postinoculation with JCV 586.



TABLE 2 tition with homologous, unlabeled NF-1 probe but not bymutated NF-1 probe. In lane 10, nuclear extract fromKinetics of JCV 586 Virion Multiplication in SVG and SH-EP CellsHeLa cells yielded much slower migrating bands. In lane

HA titer 7, the nuclear extracts from HFG resulted in faster butslightly smeared migrating bands with a more intense

Days PI SH-EP SVG signal. In SH-EP and SH-SY5Y, NF-1-specific bands wereseen in lanes 1 and 4. There were strong signals in both7 õ2 1280bands with a significant difference in the slower shifted14 5120 10,240

21 40,960 40,960 band. To clarify the difference between the shifted bands,extended electrophoresis was performed under thesame conditions as shown in Fig. 3B. NF-1-shifted bands

and A2B5. SH-EP cells did not stain with antibodies to were seen in positions marked I to VI. Complexes gener-GFA, NF, or GC but did react with vimentin and A2B5 ated with nuclear extracts from SH-EP cells demon-antibodies. SH-SY5Y was positive for neurofilaments. strated two bands (I and III), whereas complexes fromThis result confirms that SH-SY5Y has a neuronal pheno- SH-SY5Y cells had bands migrating at different positionstype, while SH-EP was not positive for neuronal, differen- (II and IV). HeLa cells demonstrated (V and VI).tiated glial, and differentiated oligodendrocyte markers.

Expression of NF-1 mRNAHowever, the cells were positive for oligodendroglial pre-cursor markers (data not shown). NF-1 DNA binding proteins consist of a family of pro-

teins which are categorized into four classes A, B, C, andDifference in the binding protein for the regulation

D (Inoue et al., 1980; Paonessa et al., 1988; Gill et al.,of JCV gene expression and replication in SH-EP

1988; Santoro et al., 1988; Apt et al., 1994). The amino-and SH-SY5Y

terminal end of each class member contains the DNArecognition site and is shared by all classes. The differentTo determine if there is a difference in DNA binding

proteins between SH-EP and SH-SY5Y that could ac- carboxy termini of the proteins separate the classes intogroups. We investigated whether the different shiftedcount for viral susceptibility, gel shift assays were per-

formed. We focused on domain B in the JCV regulatory bands correlate with the difference in NF-1 classes atthe mRNA level. Figure 4 shows the result of Northernregion which has been shown to be essential for viral

multiplication. We used the NF-1 binding site in domain analysis using probes specific for each class of NF-1proteins. The reported size of the NF-1/A mRNAs are 8.6,B as a probe in the gel shift assay using nuclear protein

extracts from SH-EP, SH-SY5Y, HFG, and HeLa cells (Fig. 4.5, 4.3, 1.8, 1.7, 1.6, 0.95, and 0.6 kb depending upondistribution in different tissues (Paonessa et al., 1988).3A). Specific gel-shifted bands were detected by compe-

FIG. 3. DNA binding profiles of NF-1 consensus sequence of JCV from SH-EP, SH-SY5Y, HeLa, and HFG cells. (A) Bandshift experiment usingan g-32P-labeled NF-1 binding site oligonucleotide with nuclear extracts from SHEP (1–3 lanes), SH-SY5Y (4–6 lanes), HFG (7–9 lanes), and HeLa(10–12 lanes). (B) Band shift experiment using an g-32P-labeled NF-1 binding site oligonucleotide under the same condition as with A, except longerelectrophoresis shows the shifted bands (I to VI). NS refers to a non-specific band.



FIG. 4. NF-1 mRNA abundance in neuroblastoma cells. The Northernblot filters contained 20 mg RNA from SK-N-SH (lane 1), SH-SY5Y (2),SH-EP (3), GOTO (4) IMR-32 (5), HFG (6), and HeLa (7). Filters wereprobed with random labeled fragments of the transcriptional domain FIG. 5. Northern analysis of NF-1/X mRNA in transient transfectedof NF-1/A (A), NF-1/B (B), NF-1/C (C), NF-1/X (D), and b-actin as the HeLa cells. Four micrograms of NF-1/X effector plasmid (pAT1) orRNA concentration control. pcDNA3 was transfected into HeLa cells. Extracts were prepared 48

hr after transfection. The Northern blot filter contained 20 mg total RNAfrom HFG (lane 1), HeLa transfected with the plasmid pcDNA3 (2), andHeLa transfected with the plasmid pAT1 (3). Upper and lower arrowsEach cell line showed an 8.6-kb band. In addition, theshow the transcript which was detected by the NF-1/X-specific probe.GOTO cells also had migrating bands at 4.3 and 0.95 kb.(Bottom) The transcript of actin as the RNA control.

Therefore, smaller bands in the GOTO cells would alsobe specific for the NF-1/A mRNA. However, mRNA levelswere reduced in both SH-EP and SH-SY5Y cells; NF-1/B than the NF-1/X expressed in HFGC cells because ofmRNA was found in all cell lines. In SH-EP cells, how- the cDNA clone originally isolated (Sumner et al.,ever, the NF-1/B 7.5-kb band was approximately 10-fold 1996). Reporter gene constructs using the CAT geneless than the SH-SY5Y cells. The expression of the NF- (Fig. 6 and described under Materials and Methods)1/C-CTF mRNA at 8.6 and 6.5 kb was found in HeLa cells were made using the JCV promoter preserving the NF-but much reduced in SH-EP and SH-SY5Y cells. However, 1 binding sites. These constructs were transfected intoNF-1/D-X was expressed in IMR-32, SH-EP, and human HeLa cells either with or without the NF-1/X expressionfetal glial cells which are susceptible to JCV infection.The NF-1/D-X was not detected in HeLa cells and waslow in SH-SY5Y cells; both cell types are not permissivefor JCV infection.

Transcriptional activation of the JCV early promoterby NF-1/X in HeLa cells

In order to assess its functional significance, weasked whether NF-1/X could activate the JCV early pro-moter in a cell type that is nonpermissive to infection.HeLa cells were selected since they are not permissiveto JCV and also do not express NF-1/X in significantamounts. To be certain that NF-1/X could be expressedin HeLa cells, Northern analysis was done followingtransfection of an expression clone of NF-1/X madefrom HFGC called AT-1 (Sumner et al., 1996). Figure 5shows that the NF-1/X clone is expressed in HeLa cells FIG. 6. Schematic of promoter structure. Open boxes represent tan-using the AT-1 containing vector as a probe (Sumner dem repeat sequences, the black box is the TATA box sequence, and

the arrowheads mark the NF-1 consensus sequence.et al., 1996). The AT-1 expression vector is smaller



second repeat of the regulatory region which is 79 bpbecause of the deletion of the TATA box sequences. Asa result of the TATA box deletions, protein interactionsmay be changed at this NF-1 site. Interestingly, mostisolates of JCV from cell culture or from PML brain tissuehave the second TATA box sequence deleted. There isno variant of JCV with alterations at either NF-1 site.Consequently, NF-1 is thought to be important for suc-cessful JCV infection. A number of studies have charac-terized the NF-1 binding region of JCV, described as do-main B, in the regulatory region (Tamura et al., 1988;Amemiya et al., 1989, 1992; Kumar et al., 1993; Sock etal., 1991). For example, there are reported differences ingel shift patterns using NF-1 probes between susceptiblehuman fetal glial cells and nonpermisive HeLa cells(Amemiya et al., 1989, 1992). A similar observation has

FIG. 7. Effect of NF-1/X on the transcriptional activity of JCV early been reported in gel shift assays using adenovirus frompromoter in HeLa cells. Two micrograms of p586EHL CAT, pM1 EHL site NF-1 as a probe with extracts from permissive epi-CAT, and pM1 EH28 CAT were transfected into HeLa cells with 4 mg thelial cells compared with nonpermissive fibroblastsof the NF-1/X effect plasmid or the parent plasmid pcDNA3. The bottom

(Apt et al., 1994).of the graph shows the fold activation over controls.In this study, we found differences in gel shift pat-

terns between SH-EP and SH-SY5Y cells. In the gelshift assay using oligonucleotide probes for NF-1,

vector. The early promoters from JCV 586 (Mad-4) and there were frequent heterogeneous and smearedMad-1 showed an 8.6- and 4.6-fold activation with the bands (Kruse, 1994; Altman et al., 1994; Apt et al.,NF-1/X clone as demonstrated in Fig. 7. These data 1994). Explanations for this observed pattern with NF-suggest that the NF-1/X plays an important role in posi- 1 probes have included: (1) phosphorylation of thetively regulating JCV expression for permissiveness. binding proteins during cell growth (Yang et al., 1993);

(2) the fact that NF-1 consists of four different classeswith each NF-1 class demonstrating a different spliceDISCUSSIONform with different transcriptional activity (Apt et al.,1994); and (3) a complex of multiple proteins affectsThe restricted tissue tropism of JCV is observed both

in vivo and in vitro. In vitro, JCV has been effectively the migration of the probe as reported with PC-12cells (Adams et al., 1995). We conducted a similarpropagated in human fetal glial cells (Major et al.,

1992) and more recently in a neuroblastoma cell line assay as described using PC-12 cells. There was nochange in migration pattern of the gel-shifted bandsIMR-32 (Akatani et al., 1994). In this study, we found

that JCV was produced in several neuroblastoma cell making different complex of proteins unlikely (datanot shown). However, the expression of NF-1 mRNAlines, IMR-32, GOTO, and SK-N-SH. The latter has two

phenotypes, N-type (neuronal) and S-type (glial). We in SH-EP and SH-SY5Y is different (Fig. 4). Whetherthere is a functionally dominant form of NF-1 in thefound that only the S-subtype (SH-EP) was susceptible

for JCV but not the N-type, nor the neuronal subclone glial phenotype of the neuroblastoma cell line is cur-rently being investigated. Of the four classes of the(SH-SY5Y). The phenotype of the SH-EP cells did not

contain the differentiation markers of mature astro- NF-1 family of proteins, we have determined that theNF-1/X or class D NF-1 mRNA is highly expressed incytes (GFAP) or oligodendrocytes, (galactose cera-

mide). SH-EP cells did express the A2B5 cell surface human fetal glial cells and barely detectable in HeLacells (Sumner et al., 1996). These data suggest thatmarker characteristic of immature or precursor glial

cells (Elder and Major, 1988). selective expression of a factor such as NF-1 maycorrelate with successful infection of cells with JCV.The strain of JCV used in these experiments, JCV 586,

originated from an owl monkey-induced astrocytoma Biochemical characterization of NF-1 binding proteinsto JCV sequences in SH-EP cells and other permissive(Major et al., 1987) and has the JCV Mad-4 genotype.

This is similar to the IMR-32 adapted JCV which also has cells is currently underway. The experiments reportedhere extend the host range for JCV to a novel cell linea Mad-4 type arrangement in its regulatory region (Yogo

et al., 1993). Interestingly, the Mad-4 regulatory region which may be used for JCV propagation. This cell linemay be useful in defining the molecular basis for JCVhas a conserved NF-1 consensus sequence which is

present in the first 98-bp ‘‘tandem repeat’’ as well as the glial tropism.



Kenney, S., Natarajan, V., Strike, D., Khoury, G., and Salzman, N. P.ACKNOWLEDGMENTS(1984). JC virus enhancer-promoter active in human brain cells. Sci-

The authors thank Pauline Ballew for manuscript preparation and ence 226, 1337–1344.Devora Schoenberg for the technical editing of the manuscript. Khalili, K., Rappaport, J., and Khoury, G. (1988). Nuclear factors in human

brain cells bind specifically to the JCV regulatory region. EMBO J. 7,1205–1210.REFERENCES

Kumar, K. U., Pater, A., and Pater, M. M. (1993). Human JC virus perfectpalindromic nuclear factor 1-binding sequences important for glialAdams, A. D., Choate, D. M., and Thompson, M. A. (1995). NF-1-L is the

DNA-binding component of the protein complex at the peripherin cell-specific expression in differentiating embryonal carcinoma cells.J. Virol. 67, 572–576.negative regulatory element. J. Biol. Chem. 270, 6975–6983.

Akatani, K., Imai, M., Kimura, M., Nagashima, K., and Ikegovmi, N. Major, E. O., Amemiya, K., Tornatore, C., Houff, S., and Berger, J. R.(1992). Pathogenesis and molecular biology of progressive multifocal(1994). Propagation of JC virus in human neuroblastoma cell line

IMR-32. J. Med. Virol. 43, 13 – 19. leukoencephalopathy, the JC virus-induced demyelinating disease ofthe human brain. Clin. Microbiol. Rev. 5, 49–73.Altmann, H., Wendler, W., and Winnacker, E.-L. (1994). Transcriptional

activation by CTF proteins is mediated by a bipartite low-proline Major, E. O., Miller, A. E., Mourrain, P., Traub, R. G., and De Widt, E.(1985). Establishment of a line of human fetal glial cells that supportsdomain. Proc. Natl. Acad. Sci. USA 91, 3901–3905.

Amemiya, K., Traub, R., Durham, L., and Major, E. O. (1989). Interaction JC virus multiplication. Proc. Natl. Acad. Sci. USA 82, 1257–1261.Major, E. O., Vacante, D. A., Traub, R. G., London, W. T., and Sever,of nuclear factor-1-like protein with the regulatory region of the hu-

man polyomavirus JC virus. J. Biol. Chem. 264, 7025–7032. J. L. (1987). Owl monkey astrocytoma cells in culture spontaneouslyproduce infectious JC virus which demonstrates altered biologicalAmemiya, K., Traub, R., Durham, L., and Major, E. O. (1992). Adjacent

nuclear factor-1 and activator protein binding sites in the enhancer properties. J. Virol. 61, 1435–1441.Mandl, C., Walker, D. L., and Frisque, R. J. (1987). Derivation and charac-of the neurotropic JC virus. J. Biol. Chem. 267, 14204–14211.

Apt, D., Chong, T., Liu, Y., and Bernard, H.-U. (1993). Nuclear factor I and terization of POJ cells, transformed human fetal glial cells that retaintheir permissivity for JC virus. J. Virol. 61, 755–763.epithelial cell-specific transcription of human papillomavirus type 16.

J. Virol. 67, 4455–4463. Neuman, J., Morency, C., Russian, K. (1987). A novel rapid assay forchloramphenicol acetyltransferase gene expression. BioTechniquesApt, D., Liu, Y., and Bernard, H.-U. (1994). Cloning and functional analy-

sis of spliced isoforms of human nuclear factor 1-X: Interference with 5, 444–447.Padgett, B., Walker, D. L., Zu Rein, G., Eckroade, R., and Dessel, B.transcriptional activation by NF-1/CTF in a cell-type specific manner.

Nucleic Acids Res. 22, 3825 – 3833. (1971). Cultivation of papova-like virus from human brain with pro-gressive multifocal leucoencephalopathy. Lancet i, 1257–1260.Astrom, K.-E., Mancall, E. L., and Richardson, E. P. (1958). Progressive

multifocal leukoencephalopathy. Brain 81, 93–127. Padgett, B. L., and Walker, D. L. (1983). Virologic and serologic studiesof progressive multifocal leukoencephalopathy. Prog. Clin. Res. 105,Atwood, W. J., Amemiya, K., Traub, R., Harms, J., and Major, E. O. (1992).

Interaction of the human polyomavirus, JCV, with human B-lympho- 107–117.Paonessa, G., Gounari, F., Frank, R., and Cortese, R. (1988). Purificationcytes. Virology 190, 716–723.

Biedler, J. L., Roffler-Tarlov, S., Schachner, M., and Freedman, L. S. of a NF-1-like DNA-binding protein from rat liver and cloning of thecorresponding cDNA. EMBO J. 7, 3115–3123.(1978). Multiple neurotransmitter biosynthesis by human neuro-

blastoma cell lines and clones. Cancer Res. 38, 3751–3757. Raj, G. V., and Khalili, K. (1995). Transcriptional regulation: Lessonsfrom the human neurotropic polyomavirus, JCV. Virology 213, 283–Bradford, M. (1976). A rapid and sensitive method for the quantitation

of microgram quantities of protein utilizing the principle of protein- 291.Rettig, W. J., Spengler, B. A., Chesa, P. G., Old, L. J., and Biedler, J. L.dye binding. Anal. Biochem. 72, 248–254.

Brown, P., Tsai, T., and Gajdusek, D. C. (1975). Seroepidemiology of (1987). Coordinate changes in neuronal phenotype and surface anti-gen expression in human neuroblastoma cell variants. Cancer Res.human papovaviruses: Discovery of virgin populations and some

unusual patterns of antibody prevalence among remote peoples of 47, 1383–1389.Rupp, R. A., Kruse, U., Multhaup, G., Gobel, U., Beyreuther, K., andthe world. Am. J. Epidemiol. 102, 331–340.

Chomczynski, P., and Sacchi, N. (1987). Single-step method of RNA Sippel, A. E. (1990). Nucleic Acids Res. 18, 2607–2616.Rziha, H. J., Bornkamm, G. M., and ZurHausen, H. (1978). Seroepidemio-isolation by acid guanidinium thiocyanate-phenol-chloroform extrac-

tion. Anal. Biochem. 162, 156 – 159. logical studies and serologic response to viral infection. Med. Micro-biol. Immunol. 165, 73–92.Ciccarone, V., Spengler, B. A., Meyers, M. B., Biiedler, J. L., and Ross,

R. A. (1989). Phenotypic diversification in human neuroblastoma Santoro, C., Mermod, N., Andrews, P. C., and Tjian, R. (1988). A familyof human CCAAT-box-binding proteins active in transcription andcells: expression of distinct neural crest lineages. Cancer Res. 49,

219–225. DNA replication: Cloning and expression of multiple cDNAs. Nature334, 218–224.Elder, G., and Major, E. O. (1988). Early appearance of type II astrocytes

in developing human fetal brain. Dev. Brain Res. 42, 146–150. Schreiber, E., Matthias, P., Muller, M. M., and Schaffner, W. (1989).Rapid detection of octamer binding proteins with ‘mini-extracts’, pre-Gil, G., Smith, J. R., Goldstein, J. L., Slaughter, C. A., Orth, K., Brown,

M. S., and Osborne, T. F. (1988). Multiple genes encode nuclear fac- pared from a small number of cells. Nucleic Acids Res. 17, 6419.Sekiguchi, M., Oota, T., Sakakibara, K., and Inui, N. (1979). Establish-tor 1-like proteins that bind to the promoter for 3-hydroxy-3-methylglu-

taryl-coenzyme A reductase. Proc. Natl. Acad. Sci. USA 85, 8963– ment and characterization of a human neuroblastoma cell line intissue culture. J. Exp. Med. 49, 67–83.8967.

Gross, N., Favre, S., Beck, D., and Meyer, M. (1992). Differentiation- Sock, E., Wegner, M., and Grummt, F. (1991). DNA replication of humanpolyomavirus JC is stimulated by NF-1 in vivo. Virology 182, 298–related expression of adhesion molecules and receptors on human

neuroblastoma tissues, cell lines and variants. Int. J. Cancer 52, 85– 308.Sumner, C., Shinohara, T., Durham, L., Traub, R., Major, E. O., and91.

Inoue, T., Tamura, T., Furuichi, T., and Mikoshiba, K. (1990). Isolation Amemiya, K. (1996). Expression of multiple classes of the nuclearfactor-1 family in the developing human brain: differential expressionof complementary DNAs encoding a cerebellum-enriched nuclear

factor I family that activates transcription from the mouse myelin of two classes of NF-1 genes. J. Neuro Virol. 2, 87–100.Tamura, T., Inoue, T., Nagata, K., and Mikoshiba, K. (1988). Enhancerbasic protein promoter. J. Biol. Chem. 265, 19065–19070.



of human polyoma JC virus contains nuclear factor I-binding se- Ulrich, K., and Sippel, A. E. (1994). Transcription factor nuclear factor Iproteins form stable homo- and heterodimers. FEBS Lett. 348, 46–50.quences; analysis using mouse brain nuclear extracts. Biochem.

Biophys. Res. Commun. 157, 419–425. Walker, D. L., and Padgett, B. L. (1983). The epidemiology of humanpolyomaviruses. In ‘‘Polyomaviruses and Human Neurological Dis-Tsokos, M., Scarpa, S., Ross, R. A., and Triche, T. J. (1987). Differentia-

tion of human neuroblastoma recapitulates neural crest develop- ease’’ (J. L. Sever and D. Madden, Eds.), pp. 99–106. A. R. Liss, NewYork.ment. Am. J. Pathol. 1987, 484 –496.

Tumilowicz, J. J., Nichols, W. W., Cholon, J. J., and Greene, A. E. (1970). Yang, B.-S., Gilbert, J. D., and Freytag, S. O. (1993). Overexpression ofMyc suppresses CCAAT transcription factor/nuclear factor 1-depen-Definition of a continuous human cell line derived from neuro-

blastoma. Cancer Res. 30, 2110–2118. dent promoters in vivo. Mol. Cell. Biol. 13, 3093–3102.


propagation of the human polyomavirus, jcv, in human

Documents