Nuclear Import of Moloney Murine Leukemia Virus DNA...

JOURNAL OF VIROLOGY,0022-538X/00/$04.0010

Jan. 2000, p. 721–734 Vol. 74, No. 2

Copyright © 2000, American Society for Microbiology. All Rights Reserved.

Nuclear Import of Moloney Murine Leukemia Virus DNAMediated by Adenovirus Preterminal Protein Is Not Sufficient

for Efficient Retroviral Transduction in Nondividing CellsANDRE LIEBER,1* MARK A. KAY,2 AND ZONG-YI LI1

Division of Medical Genetics, University of Washington, Seattle, Washington 98195,1 and Departments of Pediatricsand Genetics, Stanford University, Stanford, California 94305-52082

Received 28 July 1999/Accepted 7 October 1999

Moloney murine leukemia virus (MoMLV)-derived vectors require cell division for efficient transduction,which may be related to an inability of the viral DNA-protein complex to cross the nuclear membrane. Incontrast, adenoviruses (Ad) can efficiently infect nondividing cells. This property may be due to the presenceof multiple nuclear translocation signals in a number of Ad proteins, which are associated with the incomingviral genomes. Of particular interest is the Ad preterminal protein (pTP), which binds alone or in complex withthe Ad polymerase to specific sequences in the Ad inverted terminal repeat. The goal of this study was to testwhether coexpression of pTP with retroviral DNA carrying pTP-binding sites would facilitate nuclear importof the viral preintegration complex and transduction of quiescent cells. In preliminary experiments, wedemonstrated that the karyophylic pTP can coimport plasmid DNA into the nuclei of growth-arrested cells.Retroviral transduction studies were performed with G1/S-arrested LTA cells or stationary-phase humanprimary fibroblasts. These studies demonstrated that pTP or pTP-Ad polymerase conferred nuclear import ofretroviral DNA upon arrested cells when the retrovirus vector contained the corresponding binding motifs.However, pTP-mediated nuclear translocation of MoMLV DNA in nondividing cells was not sufficient for stabletransduction. Additional cellular factors activated during S phase or DNA repair synthesis were required forefficient retroviral integration.

Integration of viral DNA into the host chromosome is anessential step in the retrovirus life cycle. With their remarkablyefficient ability to integrate, retrovirus vectors are an importanttool in obtaining stable gene expression in vitro and in vivo (37,65). However, the most commonly used and as yet best-char-acterized vectors based on mammalian C-type retrovirusescannot transduce quiescent cells, which often represent thetargets for gene therapy approaches (20, 38, 43, 63).

After entry of the retroviral core into the cytoplasm, theretroviral genome is reverse transcribed by using enzymaticactivities associated with the incoming virion. The resultingdouble-stranded linear viral DNA is associated with viral pro-teins, forming a large nucleoprotein complex. To complete theretroviral life cycle, this preintegration complex must be trans-located to the nucleus. The mechanisms of nuclear transportappear to differ among different retrovirus subfamilies andhost cells (for a review, see reference 8). For Moloney murineleukemia virus (MoMLV)-derived vectors, it is believed thatbreakdown of the nuclear membrane during mitosis is neces-sary to allow access of the preintegration complex to chromo-somal DNA. This is supported by two lines of evidence: (i)inhibitors, which delay the onset of mitosis, delay integration aswell (3, 33, 38); and (ii) MoMLV proviruses segregate into onlyone daughter cell during the first mitotic division after infec-tion, implying that integration occurs after DNA replication(19, 43). The MoMLV preintegration complex contains linearDNA associated with the CA, IN, and possibly RT and NCproteins (8). This complex with a sedimentation constant of;160S is too large to pass through the nuclear membrane by

simple diffusion. To enter the nucleus, larger proteins (.40 to60 kDa) require nuclear localization signals (NLS) (for a re-view, see reference 55). NLS are currently classified into sev-eral classes, including those consisting of a single stretch ofbasic residues, bipartite NLS composed of two clusters of basicamino acids separated by a spacer of 10 to 12 amino acids, andthose resembling the NLS of the yeast homeodomain proteinMata2 (55). These NLS interact with cytoplasmic receptorsvariously named importin a/b, importin 58/97, or karyopherina/b, initiating an energy-dependent multistep translocationinto the nucleus. Importantly, the rate of active nuclear importof proteins is independent of cell proliferation and is similarbetween arrested and dividing cells (14). Karyophilic proteinscontaining NLS can function as a nuclear import shuttle forother proteins, RNA, or DNA (11, 39, 51).

In contrast to MoMLV-based retrovirus vectors, adenovirustype 5-derived vectors (Ad) can efficiently transduce nondivid-ing cells in vitro and in vivo (for a review, see reference 24).The efficiency of Ad infection relies to a large degree onefficient targeting of the Ad genome to the host cell nucleus.Ad DNA is packaged together with viral core proteins andpTP/TP, the terminal protein, into virions. After entry into thehost cell, the virion is uncoated and the Ad DNA is transportedinto the nucleus where its replication occurs. It is generallythought that the NLS in the pTP/TP and the core protein Vplay a crucial role in directing this complex to the nucleus.Both the 80-kDa precursor to the terminal protein (pTP) andthe 55-kDa terminal protein (TP) contain an NLS (RLPVRRRRRRVP, corresponding to residues 362 and 373 within theTP), which is well conserved among all Ad serotypes (48). Itwas shown that this motif can mediate nuclear import whentransplanted into other proteins (42). Once in the nucleus, oneof the first early viral proteins synthesized is the pTP, whichbinds with its N terminus to bp 9 to 18 of the origin for Ad

* Corresponding author. Mailing address: Division of Medical Ge-netics, Box 357720, University of Washington, Seattle, WA 98195.Phone: (206) 221-3973. Fax: (206) 685-8675. E-mail: [email protected]

721

at SE

RIA

LS C

ON

TR

OL Lane M

edical Library on October 10, 2007

jvi.asm.org

Dow

nloaded from

http://jvi.asm.org

replication found at each end of the linear DNA genomewithin the inverted terminal repeats (62, 68). During lytic in-fection, pTP forms a stable heterodimer with the Ad polymer-ase (Pol), which is translocated to the nucleus by using the NLSof pTP. The pTP-Pol complex binds with increased affinity andspecificity to the origin. This binding is further enhanced byinteraction of the heterodimer with the cellular factors NFIand OCT-1, which bind at nucleotides nt 25 to 50 of the origin(60). pTP functions as a protein primer for Ad DNA replica-tion. After binding, pTP is covalently linked to dCTP, provid-ing a free 39-hydroxyl group to begin the synthesis of a daugh-ter DNA strand. Furthermore, pTP serves as the site ofprimary attachment of the viral DNA to a specific protein(s) inthe nuclear matrix, forming replicative complexes (5, 17). Latein infection, pTP is proteolytically cleaved by the viral pro-tease, generating the 55-kDa terminal protein.

The goal of this study was to incorporate the pTP-based, Ad,nuclear import machinery into MoMLV vectors and to testwhether this would allow retroviral transduction of nondividingcells. To approach this hypothesis, we first analyzed whetherpTP could mediate the nuclear import of plasmid DNA carry-ing pTP-binding sites into arrested cells. Based on these pre-liminary results, MoMLV-based vectors containing pTP-bind-ing sites were generated and tested for the ability to transducenondividing cells expressing pTP.

MATERIALS AND METHODS

Plasmids. The pTP coding sequence (Ad5 bp 8533 to 10589), including a smallexon around Ad map unit 39, which contained the initiation codon, was providedby Jerry Schaack, University of Colorado, Denver, Colo. The pTP gene wascloned under the control of different promoters into pDE1sp1A derivatives(Microbix, Toronto, Canada). To generate pRSV-pTP or pPGK-pTP, the pTPgene was inserted as 1.2-kb HindIII-EcoRI fragment into pAd.RSV or pAd.PGK(26). To generate pCMV-pTP, a cytomegalovirus (CMV) promoter-pTP genebovine growth hormone polyadenylation signal (bpA) containing the 4.1-kbNruI-SmaI fragment (provided by J. Schaack) was inserted into pDE1aSp1. Togenerate pMT-pTP, the pTP gene was first cloned into pMRENeo (53) andthen transferred as a NotI-XbaI fragment into pDE1sp1A. The Ad Pol expressionplasmid pCMVpol was also provided by J. Schaack. The Pol cDNA (Ad bp 5187to 8357) contains a mutation at the C terminus to create a SphI site that does notimpair enzymatic activity.

To generate the test plasmid for pTP-mediated nuclear transport (pITR-hAAT), a 0.8-kb fragment of pFG140 from bp 2382 to 1452 (Microbix) con-taining two head-to-head joined Ad inverted terminal repeats (provided by JimNelson, Stanford University) was inserted as a pTP-binding motif into the XhoIsite of pBS-RSV.hAAT (26) in front of the human a1-antitrypsin (hAAT) ex-pression cassette.

Plasmid DNA was purified by ultracentrifugation in two CsCl gradients.Ad vectors. All Ad vectors were generated by homologous recombination with

pJM17 (Microbix) in 293 cells. The shuttle vectors containing pTP expressioncassettes based on pDE1sp1A (Microbix) were cotransfected with pJM17 intolow-passage 293 cells by calcium phosphate coprecipitation as previously de-scribed (35). The plaque titers of all viruses were determined on 293 cells. Thepresence of replication-competent Ad and contamination with endotoxin in viruspreparations were excluded by tests described previously (35). Viruses with atiter of 5 3 1011 PFU/ml were stored at 280°C in 10 mM Tris-Cl (pH 8.0)–1 mMMgCl2–10% glycerol. Attempts to generate an Ad vector expressing Ad Polunder a CMV promoter were not successful.

Retrovirus vectors. All retrovirus vectors were based on MSCVneoEB (21).The murine stem cell virus (MSCV)-series vectors were derived from LN andMESV vectors. They contain the extended packaging signal from LN vectors forhigh viral titers with a mutated MoMLV gag start codon and an upstream regionderived from Moloney murine sarcoma virus. A 1.4-kb hAAT cDNA containingthe EcoRI-EcoRI fragment derived from pBS-RSV.hAAT was inserted into theEcoRI site of pMSCVneoEB in front of the PGKneo cassette. The hAAT cDNAis under control of a variant long terminal repeat from the retrovirus mutantPCMV (PCC4 embryonal carcinoma cell-passaged myeloproliferative sar-coma virus) (21). The pTP-binding sites were synthesized as oligonucleotidesand cloned as XhoI-BglII fragments downstream of the hAAT cDNA intopMSCVneoEB. The 18-mer pTP-binding site (Ad nt 1 to 18) used for RV.18-hAAT was assembled by annealing oligonucleotides (60) 59TCGAGCATCATCAATAATATACCTTAGA and 59GATCTCTAAGGTATATTATTGATGATGC. The mutated 18-mer pTP-binding site used for RV.D18-hAAT wasassembled by annealing oligonucleotides (59) 59TCGAGCATCATCAGCGGCGCGTTTTAGA and 59GATCTCTAAAACGCGCCGCTGATGATGC. The

90-mer pTP binding site (Ad nt 1 to 90) used for RV.90-hAAT was assembled byannealing oligonucleotides (62) 59TCGAGCATCATCAATAATATA, 59CCTTATTTTGGATTGAAGCCAATATGATAATGAGGA, 59GATCTCCTCATTATCATATTGGCTT, and 59CAATCCAAAATAAGGTATATTATTGATGATGC.

The correctness of all constructs was confirmed by sequencing. Retrovirusvectors were generated as described previously (34). Ecotropic virus generatedfollowing transient transfection of the vectors into PE501 cells was used to infectthe amphotropic packaging line PA317. Transduced PA317 cells were selectedwith G418 (700 mg/ml) for 4 weeks. For each retrovirus vector, 40 individualG418-resistant clones were screened for hAAT expression. The clone with thehighest transgene expression for each vector was amplified and subjected to titerdetermination on mouse LTA cells by counting the number of G418-resistantcolonies. The titers of the corresponding amphotropic viruses were as follows:RV.hAAT, 5 3 105 CFU/ml; RV.18-hAAT, 4 3 105 CFU/ml; RV.D18-hAAT,8.5 3 105 CFU/ml; and RV.90-hAAT, 4 3 105 CFU/ml.

Western blot analysis. After pTP plasmid transfection or Ad pTP gene trans-fer, cell pellets were lysed on ice for 30 min in 20 mM HEPES (pH 7.5)–2 mMEGTA–10% glycerol–1% Triton X-100–0.1 M dithiothreitol–and protease inhib-itors. After 5 min of boiling, 80 mg of total protein in 13 Laemmli buffer with 4%b-mercaptoethanol was separated on a sodium dodecyl sulfate–10% polyacryl-amide gel. After electrotransfer and blocking, the filters were incubated withmonoclonal antibodies against pTP (gift from Sarah Jones, University of St.Andrews, St. Andrews, United Kingdom) at a dilution of 1:40 and then incubatedwith peroxidase-labeled anti-mouse immunoglobulin antibodies (1:1,000). Thefilters were developed with the ECL detection kit (Amersham, Burlington,Mass.).

Cell lines. 293 cells (Microbix), LTA cells (American Type Culture Collection,Rockville, Md.) (61), and neonatal primary human foreskin fibroblasts (46) weregrown in Dulbecco’s modified Eagle’s medium (DMEM) plus 10% fetal calfserum (FCS). Stationary-phase fibroblasts were maintained in DMEM plus 5%FCS and 1 mM dexamethasone (46). For cell cycle arrest, LTA cells werepresynchronized by serum starvation and then incubated with either 10 mMhydroxurea or 2 mg of aphidicolin per ml. Selection for G418 resistance wasperformed by trypsinizing the cell cultures in 6-cm dishes and plating them assingle-cell suspensions in 10-cm dishes in the presence of 600 mg of active G418per ml. For studies with the ZnSO4-inducible pMRE-pTP construct, FCS waspretreated to eliminate potentially interfering heavy-metal traces by filtrationthrough Chelex 100 resin (Bio-Rad, Hercules, Calif.). A 50-ml volume of FCSwas applied to a column matrix formed of 5 g of resin. The procedure wasrepeated 5 times. The in situ cell death detection kit (Boehringer Mannheim)was used to quantify apoptosis in LTA cells as specified by the manufacturer.

Plasmid transfection and virus infection. For standard transfections, 5 3 105

LTA cells in 6-cm dishes were transfected by calcium phosphate coprecipitationwith 10 mg of plasmid DNA. The transfection efficiency was between 30 and 35%as determined by X-Gal (5-bromo-4-chloro-3-indolyl-b-D-galactopyranoside)staining of a parallel transfection with 10 mg of test plasmid plus 1 mg ofpCMV-lacZ. If the transfection efficiency was less than 30%, the test plasmid wasreprepared and transfection was repeated with double CsCl-banded DNA. ForAd infections, cells in 6-cm dishes or 12-well plates were incubated overnightwith virus in 2 or 0.5 ml of DMEM plus FCS, respectively. The multiplicity ofinfection (MOI) that infected 100% of cells was 1,000 for LTA cells and 2,000 forhuman fibroblast as determined by infection with Ad.RSVbGal (26) and subse-quent X-Gal staining. For retroviral infection, cells in 6-cm dishes or 12-wellplates were incubated for 2 h at an MOI of 1 in a total volume of 2 or 0.5 ml,respectively. Fresh supernatant from retroviral packaging cells was filteredthough 0.45-mm-pore-size filters, diluted with DMEM plus FCS, supplementedwith 4 mg of Polybrene per ml, and immediately used for infection. Test cellswere incubated with retrovirus for 2 h and then extensively washed.

Southern analysis. Nuclei were isolated from transfected cells by Nonidet P-40cell lysis and centrifugation though a sucrose step gradient as described byFitzgerald et al. (15). Extraction of genomic DNA and Southern analysis wereperformed as described previously (35). Loading differences were adjusted byrehybridization of the filters with a fragment of the mouse metallothionein gene.The following DNA fragments were used as labeled probes: a 1.4-kb fragment ofthe hAAT cDNA (EcoRI fragment of pAd.RSVhAAT [26]), and a 2-kb frag-ment of the mouse metallothionein gene (HindIII-EcoRI fragment of pmMMT[67]).

PCR. For semiquantitative PCR, a specific competitor plasmid was con-structed by inserting a blunted 2.3-kb HindIII l-DNA fragment into the EcoRVsite of the hAAT cDNA in pRSV.hAAT (26). A 500-ng portion of genomic DNAfrom isolated nuclei was spiked with competitor plasmid DNA corresponding to0.1 copy per cell and subjected to PCR with hAAT-specific primers (59 ATGCCGTCTTCTGTCTCGTGG and 59 GCACGGCCTTGGAGAGCTTC) andPCR buffer containing 1.5 mM MgCl2 and 2.5 U of Taq polymerase (Perkin-Elmer) in a total volume of 100 ml. The PCR was run for 20 or 30 cycles (1 minat 95°C, 1 min at 60°C, and 1 min at 72°C). Then 10 ml of the reaction productwas analyzed by electrophoresis in a 0.8% agarose gel.

BrdU labeling. Bromodeoxyuridine (BrdU) labeling reagent (Amersham) wasadded to the culture medium (1:1,000 dilution) for a specific period (1, 4, or 24 h[see the figure legends]). After metabolic labeling, the cells were washed twicewith phosphate-buffered saline and fixed with acetic alcohol for 30 min at room

722 LIEBER ET AL. J. VIROL.

at SE

RIA

LS C

ON

TR

OL Lane M


jvi.asm.org

Dow

nloaded from

http://jvi.asm.org

temperature. Endogenous peroxidase activity was blocked by incubation with0.03% methanol for 30 min at room temperature. The cells were incubated in 1.5N HCl for 15 min at 37°C, blocked with 10% FCS, and incubated with anti-BrdUantibodies (DAKO; 1:50 diluted in 10% FCS). Specific antibody binding wasenhanced and developed with Vectastain ABC kit (Vector Laboratories, Bur-lingame, Calif.). To quantify S-phase cells, the number of BrdU-positive cells per1,000 cells was counted from random fields.

ELISA. hAAT concentrations were determined by enzyme-linked immunosor-bent assay (ELISA) as previously described (26). The detection limit of the assaywas 500 pg/ml. Culture supernatants were used undiluted for hAAT detection.

RESULTS

pTP-mediated nuclear import of transfected plasmid DNAinto arrested cells. Transfection of plasmid DNA into cellsrepresents a simple model system for studying nuclear importof DNA in nondividing cells. It is generally known that theefficiency of plasmid transfection by calcium phosphate copre-cipitation depends on the proliferative stage of target cells andis inefficient in confluent or growth-arrested cells (18). A num-ber of studies with synchronized cells or regenerating tissues invivo have demonstrated that mitosis with nuclear membranebreakdown is a prerequisite for efficient transfection measuredby reporter gene expression (58, 66, 71). A simple diffusion ofplasmid DNA through the nuclear membrane is impossibledue to the high molecular mass (1 kb 5 618 kDa) and the largegyration radius (12, 45). Previously, it was hypothesized thatkaryophilic, DNA-binding proteins would improve plasmidtransfection into nondividing cells (22, 29). To test this, weinvestigated whether pTP can mediate nuclear transport ofplasmids carrying pTP-binding sites after transfection intonondividing cells.

To express pTP in test cells, we generated a number ofconstructs with the pTP gene under the control of promoters,which varied in their activity. This was done because a priori itwas not clear which level of pTP expression would allow fornuclear import of DNA and would, at the same time, avoidcytotoxicity. Earlier attempts to establish stable, pTP-express-ing cell lines indicated that at a certain expression level, pTPcan induce cell cycle arrest or apoptotic cell death (28, 49). Toprovide pTP expression at different levels, the strong CMV andRous sarcoma virus (RSV) promoters, the relatively weakphosphoglycerate kinase (PGK) promoter (26), and a metal-inducible promoter (MRE/MT) (53) were used. All constructswere assembled in pDE1sp1A, a shuttle plasmid subsequentlyused for production of first-generation, E1-deleted Ad vectors.

Ad vectors with pTP expression cassettes were generated tofacilitate pTP gene transfer in vitro into nondividing cells.

In preliminary experiments, we selected the murine fibro-blast cell line LTA (61) as an in vitro test cell system becausethis cell line can be transfected by calcium phosphate copre-cipitation with a relatively high efficiency (.30%) and caneasily be cell cycle arrested. pTP was expressed in mouse LTAcells after transfection of pTP expression plasmids or Ad genetransfer. On day 3 after transfection or infection, the amountof expressed pTP was analyzed by Western blotting with pTP-specific antibodies (Fig. 1). To correlate pTP expression levelswith cytotoxicity, the percentage of apoptotic cells measured bya terminal deoxynucleotidyl transferase-mediated dUTP nickend labeling (TUNEL) assay was determined in a parallel setof pTP-expressing test cells. Generally, pTP (77kDa) expres-sion levels were higher after Ad gene transfer than after plas-mid transfection. About half of the pTP expressed after Adgene transfer was converted into TP (55 kDa), probably by theAd protease expressed in transduced LTA cells. Notably, a lowlevel of pTP expression was observed in LTA cells infectedwith a first-generation control Ad vector (Ad.CMV-bGal), in-dicating that the E2 promoter is active in these cells. RSV andCMV promoters yielded similarly high levels of pTP expres-sion after transfection or infection. In comparison, pTP levelswere about threefold lower when the PGK promoter was used.Expression of pTP was maintained for at least 7 days aftertransfection or infection (data not shown). Basal, noninducedexpression from the transfected pMRE-pTP construct wasbarely detectable. HAAT expression was ;15-fold induced inthe presence of ZnSO4 added 36 h before protein analysis.After removal of ZnSO4, pTP expression returned to baselinelevels within 24 h (data not shown). In cells infected with theMRE promoter-containing Ad vector (Ad.MRE-pTP), therewas a high level of basal expression, which was comparable tothat from Ad.PGK-pTP (data not shown). We recently re-ported that the MRE promoter in Ad vectors is transactivatedby viral enhancers (D. Steinwaerder and A. Lieber, submittedfor publication). Because of this undesirable interference,Ad.MRE-pTP was excluded from further experiments.

High-level pTP expression in cells transfected with pCMV-pTP and pRSV-pTP or after infection with pTP-expressing Adwas associated with cytotoxicity (Fig. 1, bottom). Importantly,in cells transfected with pMRE-pTP, noninduced pTP expres-sion and transient induction of pTP expression over 36 h had

FIG. 1. pTP expression and pTP-associated cytotoxicity after plasmid transfection or Ad infection of LTA cells. pTP was expressed in mouse LTA cells aftertransfection of expression plasmids (left) or Ad gene transfer (right). On day 3 after transfection or infection, the amount of expressed pTP was analyzed in cell lysatesby Western blotting with pTP-specific antibodies. The molecular masses of pTP and TP are 77 and 55 kDa, respectively. pBHG10 (Microbix) contains the pTP geneunder its endogenous E2 promoter. LTA cells transfected with the ZnSO4-inducible construct pMRE-pTP were cultured in DMEM plus 10% pretreated FCS. At 36 hbefore protein analysis, 100 mM ZnSO4 was added to one set of cells (1Zn). To analyze cytotoxic effects associated with pTP expression, the percentage ofTUNEL-positive cells was determined on day 3 after transfection or infection by using the in situ cell death detection kit from Boehringer Mannheim. SEM, standarderror of the mean (n 5 3).

VOL. 74, 2000 RETROVIRAL TRANSDUCTION OF NONDIVIDING CELLS 723

at SE

RIA

LS C

ON

TR

OL Lane M


jvi.asm.org

Dow

nloaded from

http://jvi.asm.org

FIG. 2. pTP-mediated nuclear import of transfected plasmid DNA. (A) Test plasmids to study pTP-mediated nuclear import of plasmid DNA. (B) Scheme of theexperiment (see Results for a detailed description). Cells were transfected with pMRE-pTP in a proliferative stage. One set of cells was arrested in the cell cycle byadding 10 mM hydroxyurea (HU) to serum-starved cells 48 h before transfection of test plasmids. Nucleus-localized plasmid DNA was analyzed by Southern blotting24 h after transfection. At 48 h after transfection, the hAAT concentrations in culture supernatants were measured


at SE

RIA

LS C

ON

TR

OL Lane M


jvi.asm.org

Dow

nloaded from

http://jvi.asm.org

no toxic effects. However, attempts to establish pTP-express-ing, stable cell lines after cotransfection of pMRE-pTP, to-gether with a selectable marker, were not successful. Takentogether, the data show that pTP expression exerted cytotoxicside effects in a concentration-dependent manner, which com-plicated the experimental strategies. Therefore, all further ex-periments studying the effect of pTP on retroviral transductionwere designed by using transiently induced pTP expressionafter transfection of pMRE-pTP or short-term assays afterinfection with Ad.PGK-pTP.

To generate the “target” plasmid for pTP-mediated nucleartransport, a fragment of Ad DNA containing two head-to-head-joined Ad ITRs was used as a pTP-binding motif andcloned in front of a reporter gene (hAAT) expression cassette(pITR-hAAT) (Fig. 2A). The full-length Ad ITR was chosenas the pTP-binding motif because it represented the nativestructure used during the Ad life cycle. The control plasmid(phAAT) contained the same transgene cassette without theAd ITRs.

pTP-mediated nuclear import of plasmid DNA was investi-gated in parallel in dividing and in arrested cells (Fig. 2B).Cells were arrested in G1/S phase by serum starvation andtreatment with hydroxyurea, a nonspecific inhibitor of DNAsynthesis (6). Cell cycle arrest was assessed by quantifying thenumber of cells in S phase based on BrdU incorporation.During the proliferative stage, 78% of cells were BrdU posi-tive, whereas in arrested cells, only 7% of all cells passed the Sphase during the 4 h of exposure to BrdU. To analyze pTP-mediated nuclear import of plasmid DNA, LTA cells weretransfected with pMT-pTP and pTP expression was induced inone set of cells. At 36 h after induction of pTP expression,proliferating or arrested cells were transfected with the testplasmid (phAAT or pITR-hAAT). At 24 h after pITR-hAATor phAAT transfection, the concentration of plasmid DNAtranslocated to the nucleus was measured. To do this, genomicDNA extracted from isolated nuclei was analyzed by Southernblotting with a hAAT cDNA-specific probe (Fig. 2C). All pro-liferating cells had a strong transgene-specific signal from nu-clear DNA 24 h after transfection; in arrested cells, efficientnuclear localization was observed only in pTP-expressing cellstransfected with pITR-hAAT. Transfection with pITR-hAATalone without pTP coexpression and pTP expression in com-bination with the control plasmid, phAAT, did not confernuclear localization upon plasmid DNA in arrested cells. Plas-mid DNA detected from isolated nuclei could result fromnuclear import or from plasmid DNA associated with the outersurface of the nuclear membrane, without translocation. Todistinguish between these possibilities, hAAT expression wasanalyzed after transfection (Fig. 2D). Transgene expressioncould occur only if the corresponding expression plasmid hadbeen imported into the nucleus, where the transcriptional ma-chinery is located. In arrested cells, high-level hAAT expres-sion was observed only in pTP-expressing cells transfected withpITR-hAAT. The level of transgene expression in this case wascomparable to that from transfected proliferating cells. Takentogether, this demonstrates that nuclear import of plasmidDNA is hampered in nondividing cells, confirming the re-ported observation that mitosis is a prerequisite for efficient

plasmid transfection (by calcium phosphate coprecipitation).Importantly, transient pTP expression can mediate nuclearimport of plasmid DNA carrying pTP-binding sites.

Effect of pTP-mediated nuclear import of viral DNA ontransduction of nondividing cells. Preliminary plasmid trans-fection studies supported our hypothesis that pTP may allowthe nuclear import of MoMLV viral DNA carrying pTP-bind-ing sites in nondividing cells. In the next step, we generatedMoMLV-based retroviruses containing the hAAT cDNA ex-pression cassette together with a neo expression unit (RV.hAAT/neo) (Fig. 3). Attempts to introduce the Ad ITR frag-ment into retrovirus vectors resulted in viruses with very lowtiters (,102 CFU/ml), indicating that these elements affectedretroviral replication. In contrast, amphotropic retrovirus vec-tors containing only the minimal pTP-binding motif, an 18-meroligonucleotide (60), could be produced at high titers (RV.18-hAAT/neo). The pTP-binding site was placed in the center ofthe recombinant vector genome between hAAT and the neocassette to avoid potential interference of pTP binding with theformation and function of the retroviral preintegration com-plex, which is thought to be associated with the LTRs. Todifferentiate between effects mediated by pTP binding andsubsequent nuclear localization and other possible nonspecificeffects of pTP on retroviral transduction (e.g., increased intra-cellular half-life), a control vector that contained a mutatedpTP binding site (RVD18.hAAT/neo) was generated. It hasbeen previously shown that this sequence is not recognized bypTP or TP (59, 60). pTP binds to the Ad origin with higheraffinity in complex with Ad Pol and NFI. DNA binding of thiscomplex involves nt 1 to 90 of the Ad ITR. To test whether astrengthened interaction of pTP can improve nuclear DNAimport, MoMLV vectors containing the 90-mer binding motifwere generated (62). All retrovirus vectors were produced inamphotropic packaging cell lines with titers of ;4 3 105 to 8 3105 CFU/ml (measured on LTA cells based on G418-resistantcolonies).

In all further experiments, we used aphidicolin instead ofhydroxyurea to obtain a more complete cell cycle arrest.Aphidicolin is a more potent inhibitor of DNA polymerase a A6-h treatment with aphidicolin arrested most of the cells (2%BrdU-positive cells) when LTA cells were presynchronized byserum starvation (Fig. 4). According to a previous report, cellcycle arrest by aphidicolin occurs at the G1/S border (43). Cellviability as measured by plating efficiency after trypsinizationwas decreased by less than 10% when cells were treated for24 h with aphidicolin prior to trypsinization (data not shown).This is in agreement with earlier reports demonstrating thataphidicolin incubation for 16 h (43) or 24 h (16, 56) does notsignificantly affect cell viability. For retroviral transduction ex-periments, it was important to determine the kinetics of cellcycle progression after removal of aphidicolin. Therefore, afterremoval of aphidicolin, the cells were pulse-labeled with BrdUfor 1 h intervals (Fig. 4). The percentage of cells undergoingDNA synthesis peaked at approximately 4 h and declined by6 h after aphidicolin removal, indicating that most cells hadpassed through S phase and had entered the G2/M phase bythis time.

To determine how long retroviruses that have successfully

by ELISA. (C) After transfection, cells were lysed and nuclei were purified by centrifugation though a sucrose step gradient. Genomic DNA was isolated from purifiednuclei and subjected to Southern analysis with a 32P-labeled (1.4-kb) probe specific to the hAAT cDNA. For Southern analysis, 10 mg of genomic DNA was loadedper lane. All blots were rehybridized with a 32P-labeled fragment of the mouse metallothionein gene (67) to adjust for loading differences. Southern blots from threeindependent experiments were quantified by PhosphorImager analysis, and adjusted signals were expressed as arbitrary units based on plasmid concentration standardsloaded on each gel. (D) At 48 h after transfection of the test plasmids, a separate set of transfected cells was analyzed by ELISA for hAAT expression. 1 pTP, pTPexpression was induced by ZnSO4. Means and standard deviations of three determinations are given.


at SE

RIA

LS C

ON

TR

OL Lane M


jvi.asm.org

Dow

nloaded from

http://jvi.asm.org

entered the target cell retained their ability to integrate, theintracellular stability of retroviruses was estimated for each testretrovirus vector in arrested target LTA cells. The ability tointegrate and express the transgene is a function of the timebetween virus entry and the next mitosis (4). This period wasextended from ;6 to ;30 h by varying the lengths of G1/Sarrest mediated by aphidicolin (Fig. 5). The total lengths ofthese periods include the time of aphidicolin treatment plusthe ;6 h required by LTA cells to enter the next mitosis afteraphidicolin is removed. Retroviral transduction, as measuredby neo gene expression resulting in G418-resistant colonies,decreased to ,1% when aphidicolin-mediated cell cycle arrestwas extended for 4 h or more after completion of retrovirusinfection. This implies that nearly all of the transduction-com-petent virus was inactivated or degraded when mitosis wasdelayed for more than 10 h. This appears to be in agreementwith the results of an earlier study in which, by a differentmethod, the intracellular half-life of retrovirus vectors wasmeasured to be 6.4 h (4). The data shown for RV.18-hAAT/neo was representative of all four test retroviruses.

A number of important parameters obtained in these studieswere used in our central experiments aimed to analyze theeffect of coexpressed pTP on retrovirus transduction. Theseparameters include the following: (i) 2 h after the removal of

aphidicolin, cells reentered the S phase of the cell cycle, and(ii) treatment with aphidicolin for 4 h or more after retrovirusinfection prevented transduction, because all the virus wasinactivated or degraded before the cells entered the next mi-tosis.

To test whether pTP can mediate the nuclear import of viralDNA carrying pTP-binding motifs and to analyze whether thiseffect would be more pronounced with pTP in complex with AdPol, proliferating LTA cells were transfected with pTP, pTPplus Pol, or control plasmids (Fig. 6A). After transfection, thecells were arrested by serum starvation and aphidicolin treat-ment. Retrovirus infection was performed at an MOI of 1 inarrested cells. Our first experimental design was based on anextended cell cycle arrest by incubation with aphidicolin thatwas continued for 4 h after retrovirus infection (scheme I),which was shown earlier to prevent transduction because allviral preintegration complexes were inactivated or degradedbefore the infected cells could enter the next mitosis. However,in this scheme, transduction-competent virus was still presentwhen cells entered the S phase (Fig. 6A).

After retroviral infection of arrested LTA cells followed byaphidicolin treatment extended for 4 h, pTP-mediated nuclearimport of viral DNA was analyzed by competitive PCR ofgenomic DNA isolated from purified nuclei (Fig. 6B). PCR

FIG. 3. Retrovirus vectors carrying pTP-binding sites. All vectors were based on MSCVneoEB (21) and contained the hAAT cDNA under the control of theretroviral LTR and the neo gene under the control of the PGK promoter. pTP- or pTP-Pol-NFI-binding sites were inserted into the center of the retroviral genometo avoid potential interference with the formation and stability of the preintegration complex. RV.D18-hAAT/neo contained a mutated pTP-binding motif that is notrecognized by pTP. The underlined positions were mutated. The nucleotide numbers 1 to 18 and 1 to 90 refer to positions in Ad5 DNA. All viruses were producedin amphotropic packaging cells at titers ranging from 4 3 105 to 8 3 105 CFU/ml.

FIG. 4. Analysis of cell cycle progression after arrest by aphidicolin. LTA cells (;70% confluent) were presynchronized by serum starvation for 2 days, resultingin 14% of cells passing the S phase during the 4-h period of BrdU labeling. Cell cycle arrest was completed by incubation with aphidicolin for 6 h, resulting in only2% of the cells replicating DNA in a 4-h interval. To monitor cell cycle progression after removal of aphidicolin, different dishes with cells were pulse-labeled with BrdUfor 1-h intervals (hours 0 to 1, 1 to 2, 3 to 4, 5 to 6, and 7 to 8 after the removal of aphidicolin). S-phase cells were counted for selected intervals. SEM, standard errorof the mean (n 5 3).


at SE

RIA

LS C

ON

TR

OL Lane M


jvi.asm.org

Dow

nloaded from

http://jvi.asm.org

was performed with hAAT-specific primers. A strong vector-specific signal appeared in nuclear DNA from cells coexpress-ing pTP and viral DNA with pTP-binding sites. This signal wasstronger in cells expressing both pTP and Pol and the RV.90-hAAT/neo DNA containing the extended binding motif forpTP, Pol, and NFI. This indicates that pTP can mediate thenuclear import of viral DNA carrying pTP-binding sites intoarrested cells and that this process is enhanced by Ad Pol(probably due to increased pTP-binding affinity or possibly useof the NLS present in Ad Pol). Vector-specific backgroundsignals were visible in all control lanes. These signals probablyoriginated from the transduction of the small percentage ofnonarrested cells, which were present at the time of retroviralinfection.

Retroviral transduction was assessed based on transgene(hAAT and neo) expression (Fig. 6C). hAAT expression wasanalyzed 3 days after infection. The number of stably trans-duced cells was determined based on the number of G418-resistant colonies present after 3 weeks of selection. SignificanthAAT expression was observed only in cells expressing pTP orpTP plus Pol and containing viral DNA with the correspondingbinding motifs. Combined expression of pTP and Pol yieldedhigher transgene expression in cells infected with RV.90-hAAT/neo compared to cells containing the pTP-binding motifalone or cells infected with RV.90-hAAT/neo without coex-pressed Ad Pol. The number of G418-resistant colonies mir-rored the transduction data obtained based on hAAT expres-sion, suggesting that pTP-mediated transduction is associatedwith stable vector integration. For comparison, transduction ofproliferating cells with the same MOI of RV.18-hAAT/neoyielded about 50 times more G418-resistant colonies than inarrested cells with pTP-supported import of retroviral DNA.Clearly, there was a significant stimulation of transductionsupported by pTP; however, this mechanism was not as effi-cient as transduction during cell division.

These data indicate that pTP can mediate the nuclear im-port of viral DNA and that this is sufficient for integration.

However, in experimental scheme I, transduction-competentvirus was still present when cells entered the S phase. Toanalyze whether in addition to pTP-mediated nuclear import,events occurring during S phase are critical for vector integra-tion, treatment with aphidicolin was continued for 24 h(scheme II) after retrovirus infection (Fig. 7A). During thisperiod, all the virus was degraded or inactivated long beforethe infected cells could enter either the S or M phase accordingto data obtained earlier. Analysis of nucleus-localized viralDNA and retroviral transduction was performed as describedfor scheme I. No vector-specific signal was detectable by com-petitive PCR after 20 PCR cycles (data not shown). After 30PCR cycles, vector-specific signals that were slightly strongerthan background signals appeared in cells expressing pTP orpTP plus Pol after infection with RV.18-hAAT/neo or RV.90-hAAT/neo (Fig. 7B). This indicates that most viral DNA hadnot integrated and was degraded before analysis. It is thoughtthat nonintegrated viral DNA is not protected from degrada-tion by nucleases, which determines the short intracellularhalf-life of transduction-competent virus (4, 71). There was alow level of transduction (based on hAAT expression andformation of G418-resistant colonies) in cells coexpressingpTP or pTP plus Pol and viral DNA with the correspondingbinding sites. However, in comparison, hAAT expression andformation of G418-resistant colonies was ;10-fold less effi-cient in scheme II (24 h of aphidicolin) than in scheme I (4 hof aphidicolin).

Ad-mediated pTP gene transfer allows the expression ofpTP in 100% of test cells at a higher level than transfectionwith pTP expression plasmids. To test whether this propertywould change the outcome of retroviral transduction studies inarrested cells, LTA cells were infected with Ad.PGK-pTP orcontrol virus (Ad.Co). Subsequently, arrested cells were in-fected with retrovirus, after which incubation with aphidicolinwas continued for 4 or 24 h according to the experimentaldesigns developed for Fig. 6 (scheme I) and Fig. 7 (scheme II).Retroviral transduction was evaluated based on hAAT expres-

FIG. 5. Retroviral life span in arrested cells. Cells were arrested as described for Fig. 4. Arrested cells were infected with retrovirus (RV18.hAAT/neo) at an MOIof 1 for 2 h. After infection, cell cycle arrest by aphidicolin was continued for different periods, after which the cells were trypsinized and subjected to G418 selection.The number of G418-resistant colonies was determined after 3 weeks of selection. The selected periods of cell cycle arrest were based on the consideration that thelag phase between aphidicolin removal and next mitosis was assumed to be at least 6 h (Fig. 4). The period between virus entry and mitosis (0 h of incubation withaphidicolin) includes the 6 h that LTA cells require to enter the next mitosis after aphidicolin was removed. This period was lengthened by aphidicolin-mediated cellcycle arrest, which was extended for 0 to 24 h after virus infection. SEM, standard error of the mean (n 5 3).


at SE

RIA

LS C

ON

TR

OL Lane M


jvi.asm.org

Dow

nloaded from

http://jvi.asm.org

FIG. 6. Effect of pTP expression on retroviral nuclear import and transduction (scheme I: treatment with aphidicolin extended for 4 h). (A) Scheme of theexperiment. A total of 8 3 105 LTA cells (70% confluent) were transfected with 10 mg of pcDNA3 (Invitrogen) as control plasmid (Co), 5 mg of pMRE-pTP plus 5mg of pcDNA3 (pTP), or 5 mg of pMRE-pTP plus 5 mg of pCMV-pol (pTP 1 Pol). After transfection, the cells were synchronized by serum starvation and arrestedin the cell cycle by aphidicolin. Arrested cells were infected with retroviruses (RV.hAAT/neo, RV.D18-hAAT/neo, RV.18-hAAT/neo, or RV.90-hAAT/neo) at an MOIof 1 for 2 h. pTP expression was induced for the time of retrovirus infection by addition of ZnSO4. Cell cycle arrest was continued for 4 h after retrovirus infection.At this time point, one set of infected cells was trypsinized and lysed, and genomic DNA was extracted from isolated nuclei (see Materials and Methods). Another setof dishes was analyzed for hAAT expression 3 days after infection and then subjected to G418 selection. The number of G418-resistant colonies was counted after 3


at SE

RIA

LS C

ON

TR

OL Lane M


jvi.asm.org

Dow

nloaded from

http://jvi.asm.org

sion on day 3 postinfection (Fig. 8). Analysis for formation ofG418-resistant colonies was not possible due to the cytotoxicitycaused by the high level of pTP expression after Ad genetransfer. In agreement with the previously obtained data, inscheme I (4 h of aphidicolin) significant retroviral transductionin arrested cells was detected only in cells expressing pTPafter Ad.PGK-pTP gene transfer, which were subsequentlyinfected with vectors carrying pTP-binding sites (RV.18-hAAT/neo, RV.90-hAAT/neo). Transduction rates were ap-proximately 20-fold lower with these retrovirus vectors in cellsafter infection with the first-generation control vector (Ad.Co),which had yielded a low level of pTP expression in transducedLTA cells (Fig. 1). Transgene expression was not detectable incells not infected with Ad or in cells infected with the controlretroviruses (RV.hAAT/neo or RV.D18h-AAT/neo). Pro-longed cell cycle arrest after retrovirus infection (24 h ofaphidicolin) resulted in ;20-fold-lower retroviral transductionrates in pTP-expressing cells infected with retroviruses carryingpTP-binding sites.

In conclusion, it appears that pTP can support the nuclearimport of viral DNA carrying pTP-binding sites. Coexpressionof Pol enhances this process. However, pTP-mediated nuclearimport is not sufficient for transduction in long-term-arrestedcells. Our data indicates that in addition to nuclear localiza-tion, events occurring during S phase are required for MoMLVintegration and stable transduction.

So far, for convenience, we have used artificially arrestedtest cells for retroviral transduction studies. To exclude thepossibility that aphidicolin nonspecifically affects cellular fac-tors necessary for retroviral integration, another cell culturesystem was used as a model to test the pTP-mediated retroviraltransduction of nondividing cells. Primary neonatal humanforeskin fibroblast cultures contain only 2 to 4% of cells pass-ing through the S phase each day when maintained as conflu-ent monolayers at a reduced serum concentration (46, 47).However, transfection of these cells is extremely inefficient(data not shown). Therefore, for transduction studies, pTPexpression was provided by infection with Ad.PGK-pTP at anMOI of 2,000, which allowed the transduction of all cells.Retrovirus vectors were applied to arrested cells (at an MOI of1) 2 days after Ad infection, and hAAT expression was mea-sured on day 3 after retrovirus infection to assess retroviraltransduction. hAAT levels after infection of stationary humanfibroblasts were detectable only in pTP-expressing cells afterinfection with retroviruses carrying pTP-binding sites (Fig.9A). However, transgene expression was about 300-fold lowerthan retroviral infection of dividing fibroblasts (ranging from145 to 167 ng/ml for RV.hAAT/neo, RV.D18-hAAT/neo,RV.18-hAAT/neo, and RV.90-hAAT/neo). This underscoresour conclusion that pTP-supported nuclear import is not suf-ficient to allow significant transduction of nondividing cells.

This model for nondividing cells allowed us to study whetherin addition to pTP-mediated nuclear import, pretreatmentwith agents that stimulates DNA repair synthesis would in-crease retroviral transduction. Previous studies with recombi-

nant adeno-associated virus vectors have demonstrated thatDNA synthesis inhibitors or DNA-damaging agents can induceunscheduled DNA synthesis or DNA repair pathways, result-ing in an increased transduction of nondividing cells by rAAVvectors (2, 46). To test whether this had an effect on oursystem, stationary fibroblasts infected with Ad.pTP or Ad.Cowere exposed to [3H]thymidine (10 mCi/ml), as a DNA-dam-aging agent (2), or the DNA synthesis inhibitors aphidicolin ordistamycin A (72) prior to retrovirus infection. As shown inFig. 9, pretreatment with these agents, which were able toinduce cellular DNA repair enzymes, stimulated retroviraltransduction; however, this occurred only in combination withpTP-mediated nuclear import of viral DNA. Notably, trans-duction rates as measured based on transient hAAT expressionwere still 20 to 30 times lower than infection of dividing cells.

DISCUSSION

Transduction with MoMLV-based retrovirus vectors doesnot occur in nondividing cells. It is thought that this phenom-enon is related to an inability of the MoMLV preintegrationcomplex to cross the nuclear membrane due to the absence ofcompetent NLS in viral proteins associated with the incomingvirion. In contrast to MoMLV, Ad efficiently infects nondivid-ing cells, and this property is probably due to the presence ofNLS in viral proteins, in particular the terminal protein. Fur-thermore, there is increasing evidence from other viral systemsthat NLS within viral proteins are crucial for nuclear translo-cation of viral genomes. For influenza virus, the nuclear entryof the viral genome is mediated by an NLS in the nucleopro-tein component of the viral RNA-protein complex (10). Asimilar mechanism of nuclear localization of the viral genomeseems to exist for simian virus 40, mediated by its small struc-tural proteins Vp2 and Vp3 (7). Importantly, for both influ-enza virus and simian virus 40, quiescent cells are among theirnatural targets. Members of another retrovirus subfamily, thelentiviruses, do not require mitosis for transduction. It isthought that this property is based on active transport of thepreintegration complex into the nucleus of an infected cellwithout the requirement for nuclear envelope breakdown dur-ing cell division (23, 40, 41).

pTP mediates nuclear import of plasmid DNA with pTP-binding sites in nondividing cells. The hypothesis underlyingthis study was that coexpression of the Ad pTP with viralMoMLV DNA carrying pTP-binding sites would facilitate nu-clear import of the preintegration complex in nondividing cellsand that this would be sufficient to allow retroviral transduc-tion. To support this hypothesis, we demonstrated with ar-rested cells that pTP mediated the efficient nuclear import oftransfected plasmid DNA carrying pTP-binding sites. This is inagreement with a study from 1984 demonstrating that trans-fection of Ad DNA attached to pTP or TP was several ordersof magnitude more efficient than transfection of viral DNAwithout pTP or TP (22). It underscores the point that nucleartransport is a limiting step for transfection and that nuclear

weeks of selection. As shown in Fig. 5, treatment with aphidicolin for 4 h after retrovirus infection prevented transduction, because all retroviral DNA was inactivatedor degraded before the infected cells could enter the next mitosis. However, transduction-competent virus was still present when cells entered the S phase. (B) Detectionof nucleus-localized viral DNA by competitive PCR. A 500-ng portion of genomic DNA isolated from purified nuclei was spiked with ;0.1 copy per cell of competitorplasmid DNA and subjected to PCR (20 cycles) with hAAT-specific primers. The PCR product specific for the vector DNA is 1.1 kb. The PCR product derived fromthe competitor template is 3.4 kb. PCR products were separated by electrophoresis in 0.8% agarose gels. (C) Transgene expression after pTP-mediated nuclear importof viral DNA in arrested cells. hAAT expression in culture supernatants was analyzed by ELISA 3 days after retrovirus infection. To determine the number ofG418-resistant, neo expressing clones, infected cells were trypsinized and subjected to G418 selection for 3 weeks. Co, cells transfected with control plasmid; pTP, cellstransfected with pMRE-pTP with subsequent induction by ZnSO4; pTP1Pol, cells transfected with pMRE-pTP and pCMV-pol and induction with ZnSO4. Means andstandard deviations of three determinations are given.


at SE

RIA

LS C

ON

TR

OL Lane M


jvi.asm.org

Dow

nloaded from

http://jvi.asm.org

FIG. 7. Effect of pTP expression on retroviral nuclear import and transduction (scheme II: treatment with aphidicolin for 24 h). (A) Scheme of the experiment. Cellswere transfected with pTP-Pol expression plasmids, arrested, and infected with retroviruses as described for Fig. 6. After retrovirus infection, treatment with aphidicolinwas continued for 24 h. During this period, all virus was degraded or inactivated long before the infected cells could enter either the S or M phase. (B) Detection ofnucleus-localized viral DNA by competitive PCR. PCR was performed for 30 cycles under the conditions described for Fig. 6B. (C) Transgene expression afterpTP-mediated nuclear import of viral DNA in arrested cells. hAAT and long-term neo expression were analyzed as described for Fig. 6C.


at SE

RIA

LS C

ON

TR

OL Lane M


jvi.asm.org

Dow

nloaded from

http://jvi.asm.org

entry of plasmid DNA could be improved by DNA-boundkaryophilic proteins. Moreover, a recent report noted thatplasmid DNA is rapidly degraded by cytosolic nucleases andthat transfection is more efficient the faster the DNA is trans-located to the nucleus (29).

pTP mediates nuclear import of viral MoMLV DNA con-taining pTP-binding sites. In arrested cells, pTP mediated thenuclear import of MLV DNA carrying pTP-binding sites. Thisallowed retroviral transduction in arrested cells that were laterreleased into S phase while transduction-competent virus wasstill present. However, pTP-mediated transduction rates in ar-rested cells were still lower than in proliferating cells. Theremay be several explanations for this. For example, the affinityof pTP binding to viral DNA may be too low, which might limitthe efficiency of nuclear translocation. This is supported by ourobservation that pTP-mediated nuclear import and transduc-tion was increased by Ad Pol, which presumably increases theaffinity of pTP binding to DNA. On the other hand, weak pTPbinding may be advantageous by allowing rapid dissociation.Otherwise, this complex could be tightly associated with thenuclear matrix via pTP, which may decrease the efficiency ofintegration. Alternatively, pTP binding within the central re-gion of viral DNA may affect the composition, stability, oractivity of the preintegration complex, for example, resultingperhaps in the loss of viral integrase. However, it is thoughtthat the preintegration complex, or the so-called intasome, isorganized around several hundred base pairs at each end of theviral DNA (69, 70), which makes it unlikely that pTP bound inthe middle of the viral DNA could critically affect integration.Finally, the concentration of nucleotides available for reversetranscription of retroviral genomes may be a limiting factor innondividing cells.

Requirement of S phase or DNA repair for MoMLV inte-gration. From the data obtained in this study, we concludedthat nuclear translocation of MoMLV DNA is not sufficient forstable transduction and that additional cellular factors acti-vated during S phase or DNA repair are required to mediateintegration and stable transduction of nondividing cells. It isnow generally accepted that cellular proteins are required forretroviral integration, and it is speculated that the activity ofsome of these accessory cellular proteins is regulated in a cellcycle-specific manner. Although the purified integrase proteinis sufficient to carry out 39 processing and DNA strand transferin cell-free systems, viral integration is several orders of mag-nitude more efficient in cells, indicating that cellular factors areinvolved in the process of integration (70). Among the hostproteins that are potentially involved in MoMLV integrationare BAF (30), INI 1 (25), HMG 1 (1), HMGI(Y) (13), andcellular DNA repair enzymes. During the process of MoMLVintegration, the 59 ends of the viral DNA and the 39 ends of thetarget DNA remain unjoined, and cellular repair enzymes arebelieved to be responsible for degradation of the unpairednucleotides at the 59 ends of the viral DNA, filling in thesingle-strand gaps, and for the subsequent ligation to completethe integration process (8). A recent study has demonstratedthat retroviral integration intermediates are detected as DNAdamage by the host cell that and completion of the integrationprocess requires the DNA-dependent protein kinase-mediatedDNA repair pathway (9). Importantly, DNA-dependent pro-tein kinase is regulated in a cell cycle-dependent manner, withpeaks of activity during the G1 and early S phases (31). In lightof this, we speculate that viral DNA imported to the nucleus bypTP symport is rapidly degraded when not integrated and that

FIG. 8. Retroviral transduction in arrested LTA cells after Ad-mediated pTP gene transfer. Arrested LTA cells without prior Ad infection (no Ad), transduced withAd.RSVbGal (34) (Ad.Co), or transduced with Ad.PGK-pTP (Ad.pTP) were infected with retroviruses as described for Fig. 6. One set of cells was incubated withaphidicolin for 4 h after retrovirus infection (scheme I), and another set was incubated for 24 h (scheme II). hAAT levels were measured on day 3 after retrovirusinfection. Means and standard deviations of three determinations are given.


at SE

RIA

LS C

ON

TR

OL Lane M


jvi.asm.org

Dow

nloaded from

http://jvi.asm.org

integration requires cellular factors, which are not activated inartificially arrested LTA cells or quiescent fibroblasts.

Alternatively, serum starvation or aphidicolin treatment mayperturb cellular metabolism, which may affect retrovirus inte-gration independently of the cell cycle requirements. More-over, retroviral integration may require a specific chromatinstructure, which is not present in arrested cells. There are anumber of reports stating that MoMLV integration dependson the methylation or heterochromatinization state of chro-mosomal DNA (27, 44, 50, 64).

At this point, the question arises of how lentiviruses inte-grate in nondividing cells. It is thought that viral proteins canmediate the nuclear translocation of the human immunodefi-ciency virus type 1 preintegration complex. Vpr appears to playa central role in this process (40, 41). Interestingly, Vpr isknown to interfere with cell cycle checkpoint control and isassociated with the induction of chromosomal breaks andDNA repair synthesis (36, 54). The potential property of len-tivirus proteins to induce cell cycle-dependent host proteins,together with a longer intracellular half-life for lentivirus

FIG. 9. Retroviral transduction of primary human fibroblasts after pTP expression and induction of DNA repair synthesis. (A) Stationary-phase human fibroblastcultures without prior Ad infection (no Ad), transduced with Ad.RSVbGal (34) (Ad.Co), or transduced with Ad.PGK-pTP (Ad.pTP) were infected with retrovirusesat an MOI of 1. At the time of retrovirus infection, less than 3% of the cells were in S phase. At 48 h after retrovirus infection, hAAT levels in culture supernatantswere measured by ELISA. hAAT levels detected in pTP-expressing cells (Ad.pTP) were statistically different from those in Ad.Co-infected cells (P # 0.002) (asdetermined by the Student t test assuming a normal distribution). (B) Stationary-phase human fibroblast cultures were pretreated overnight with distamycin A (5 31026 M), aphidicolin (2 mg/ml), or [methyl-3H]thymidine (10 mCi/ml). After distamycin or aphidicolin treatment, the cells were washed and infected with retrovirus.Incubation with [methyl-3H]thymidine was continued during retrovirus infection. At 48 h after retrovirus infection, hAAT levels in culture supernatants were measuredby ELISA. Means and standard deviations of three determinations are given.


at SE

RIA

LS C

ON

TR

OL Lane M


jvi.asm.org

Dow

nloaded from

http://jvi.asm.org

DNA, may be beneficial for integration in nondividing cells. Inthis context, a number of reports noted that although mitosis isnot required for transduction by lentiviruses or lentivirus vec-tors, cell cycle progression through G1/S significantly increasedtransduction efficiencies (32, 52, 57).

Our study demonstrates that in arrested cells, the karyo-philic, DNA-binding Ad protein pTP mediated the nuclearimport of plasmid or MoMLV DNA carrying pTP-bindingsites. This observation may provide a rationale for improvingplasmid transfection techniques or nonviral gene transfer, par-ticularly in nondividing cells. Instead of pTP, which exertscytotoxic side effects, synthetic proteins, which contain strong,sequence-specific DNA-binding domains fused to strong NLS,may be a better alternative to include in nonviral deliverysystems. Furthermore, our finding that nuclear import ofMoMLV viral DNA is not sufficient for integration and ap-pears to require cell cycle-dependent cellular factors contrib-utes to a better understanding of retroviral transduction.

ACKNOWLEDGMENTS

We thank Chen-Yi He for technical assistance and Cheryl Carlsonfor critical discussion. We are grateful to David Russell (University ofWashington) for providing human primary fibroblasts.

This work was supported by the Cystic Fibrosis Foundation (A.L.)and NIH-DK49022 (M.A.K.).

REFERENCES

1. Aiyar, A., P. Hindmarsh, A. M. Skalka, and J. Leis. 1996. Concerted inte-gration of linear retroviral DNA by the avian sarcoma virus integrase in vitro:dependence on both long terminal repeat termini. J. Virol. 70:3571–3580.

2. Alexander, I. E., D. W. Russell, and A. D. Miller. 1994. DNA-damagingagents greatly increase the transduction of nondividing cells by adeno-asso-ciated virus vectors. J. Virol. 68:8282–8287.

3. Andreadis, S., A. O. Fuller, and B. O. Palsson. 1998. Cell-cycle dependenceof retroviral transduction: an issue of overlapping time scales. Biotechnol.Bioeng. 58:272–281.

4. Andreadis, S. T., D. Brott, A. O. Fuller, and B. O. Palsson. 1997. MoloneyMurine Leukemia Virus-derived retroviral vectors decay intracellularly witha half-life in the range of 5.5 to 7.5 hours. J. Virol. 71:7541–7538.

5. Angeletti, P. C., and J. A. Engler. 1996. Tyrosine kinase-dependent release ofan adenovirus preterminal protein complex from the nuclear matrix. J. Virol.70:3060–3067.

6. Brooks, R. F. 1976. Regulation of fibroblast cell cycle by serum. Nature260:248–250.

7. Clever, J., and H. Kasamatsu. 1991. Simian virus 40 VP2/3 small structuralproteins habor their own nuclear transport signal. Virology 181:78–90.

8. Coffin, J. M. 1996. Retroviridae: the viruses and their replication, 3rd ed., p.1767–1848. In B. N. Fields, D. M. Knipe, and P. M. Howley (ed.), Fieldsvirology, vol. 2. Lippincott-Raven Publishers, Philadelphia, Pa.

9. Daniel, R., R. A. Katz, and A. M. Skalka. 1999. A role for DNA-PK inretroviral DNA integration. Science 284:644–647.

10. Davey, J., N. J. Dimmock, and A. Colman. 1985. Identification of the se-quence responsible for the nuclear accumulation of the influenza virus nu-cleoprotein in Xenopus oocytes. Cell 40:667–675.

11. Dean, D. A. 1997. Import of plasmid DNA into the nucleus is sequencespecific. Exp. Cell Res. 230:293–302.

12. Dowty, M. E., P. Williams, G. Zhang, J. E. Hagstrom, and J. A. Wolff. 1995.Plasmid DNA entry into postmitotic nuclei of primary rat myotubes. Proc.Natl. Acad. Sci. USA 92:4572–4576.

13. Farnet, C. M., and F. D. Bushman. 1997. HIV-1 cDNA integration: require-ment for HMGI (Y) protein for function of preintegration complexes invitro. Cell 88:483–492.

14. Feldherr, C. M., and D. Akin. 1994. Variations in signal-mediated nucleartransport during the cell cycle in BALB/c 3T3 cells. Exp. Cell Res. 215:206–210.

15. Fitzgerald, M. E., E. Webber, J. Donovan, and N. Fausto. 1995. Rapid DNAbinding by nuclear factor kappaB in hepatocytes at the start of liver regen-eration. Cell Growth Differ. 6:417–427.

16. Fox, M. H., R. A. Read, and J. S. Bedford. 1987. Comparison of synchronizedChinese Hamster Ovary cells obtained by mitotic shake-off, hydroxyurea,aphidicolin, or methotrexate. Cytometry 8:315–320.

17. Fredman, J. N., and J. A. Engler. 1993. Adenovirus precursor to terminalprotein interacts with the nuclear matrix in vitro and in vivo. J. Virol.67:3384–3395.

18. Graham, F. L., and A. J. van der Eb. 1973. A new technique for the assay of

infectivity of human adenovirus 5 DNA. Virology 52:456.19. Hajihosseini, M., L. Iavachev, and J. Price. 1993. Evidence that retroviruses

integrate into post-replication DNA. EMBO J. 12:4969–4974.20. Harel, J., E. Rassart, and P. Jolicoeur. 1981. Cell cycle dependence of

synthesis of unintegrated viral DNA in mouse cells newly infected withmurine leukemia virus. Virology 110:202–207.

21. Hawley, R. G., F. H. L. Lieu, A. Z. C. Fong, and T. S. Hawley. 1994. Versatileretroviral vectors for potential use in gene therapy. Gene Ther. 1:136–138.

22. Hay, R. T., N. D. Stow, and I. M. McDougall. 1984. Replication of adenovirusmini-chromosomes. J. Mol. Biol. 175:493–510.

23. Heinzinger, N. K., M. I. Bukrinsky, S. A. Haggerty, V. Ragland, M. A.Kewalramani, H. E. Lee, L. Gendelmann, L. Ratner, M. Stevenson, and M.Emerman. 1994. The Vpr protein of HIV-1 influences nuclear localization ofviral nucleic acids in non-dividing host cells. Proc. Natl. Acad. Sci. USA91:7311–7315.

24. Hitt, M. M., C. L. Addison, and F. L. Graham. 1997. Human adenoviralvectors for gene transfer into mammalian cells. Adv. Pharmacol. 40:137–205.

25. Kalpana, G. V., S. Marmon, W. Wang, G. R. Crabtree, and S. P. Goff. 1994.Binding and stimulation of HIV-1 integrase by a human homolog of yeasttranscription factor SNF5. Science 266:2002–2006.

26. Kay, M. A., F. Graham, F. Leland, and S. L. Woo. 1995. Therapeutic serumconcentrations of human alpha1-antitrypsin after adenoviral-mediated genetransfer into mouse hepatocytes. Hepatology 21:815–819.

27. Kitamura, Y., Y. M. H. Lee, and J. M. Coffin. 1992. Nonrandom integrationof retroviral DNA in vitro: effect of CpG methylation. Proc. Natl. Acad. Sci.USA 89:5532–5536.

28. Langer, S. J., and J. Schaak. 1996. 293 cell lines that inducibly express highlevels of adenovirus type 5 precurser terminal protein. Virology 221:172–179.

29. Lechardeur, D., K.-J. Sohn, M. Haardt, P. B. Joshi, M. Monck, R. W.Graham, B. Beatty, J. Squire, H. O’Brodovich, and G. L. Lukacs. 1999.Metabolic instability of plasmid DNA in the cytosol: a potential barrier togene transfer. Gene Ther. 6:482–497.

30. Lee, M. S., and R. Craigie. 1998. A previously unidentified host proteinprotects retroviral DNA from autointegration. Proc. Natl. Acad. Sci. USA95:1528–1533.

31. Lee, S. E., R. A. Mitchell, A. Cheng, and E. A. Hendrickson. 1997. Evidencefor DNA-PK-dependent and -independent DNA double-strand break repairpathways in mammalian cells as a function of the cell cycle. Mol. Cell. Biol.17:1425–1433.

32. Lewis, P., M. Hensel, and M. Emerman. 1994. Human immunodeficiencyvirus infection of cells arrested in cell cycle. EMBO J. 11:3053–3058.

33. Lewis, P. F., and M. Emerman. 1994. Passage though mitosis is required foroncoviruses but not for the human immunodeficiency virus. J. Virol. 68:510–516.

34. Lieber, A., M. J. Vrancken Peeters, and M. Kay. 1995. Adenovirus-mediatedtransfer of the amphotropic retrovirus receptor cDNA increases retroviraltransduction in cultured cells. Hum. Gene Ther. 6:5–11.

35. Lieber, A., C.-Y. He, I. Kirillova, and M. A. Kay. 1996. Recombinant adeno-viruses with large deletions generated by Cre-mediated excision exhibit dif-ferent biological properties compared with first-generation vectors in vitroand in vivo. J. Virol. 70:8944–8960.

36. Mahalingam, S., V. Ayyavoo, M. Patel, T. Kieber-Emmons, G. D. Kao, R. J.Muschel, and D. B. Weiner. 1998. HIV-1 Vpr interacts with the human 34kDa mov34 homologue, a factor linked to the G2/M phase transition of themammalian cell cycle. Proc. Natl. Acad. Sci. USA 95:3419–3424.

37. Miller, A. D. 1992. Human gene therapy comes of age. Nature 357:455–460.38. Miller, D. G., M. A. Adam, and A. D. Miller. 1990. Gene transfer by retro-

virus vectors occurs only in cells that actively replicate at the time of infec-tion. Mol. Cell. Biol. 10:4239–4242.

39. Pollard, W. W., W. M. Michael, S. Nakielny, M. C. Siomi, F. Wang, and G.Dreyfuss. 1996. A novel receptor-mediated nuclear protein import pathway.Cell 86:985–994.

40. Popov, S., M. Rexach, L. Ratner, G. Blobel, and M. Bukrinsky. 1998. Viralprotein R regulates docking of the HIV-1 preintegration complex to thenuclear pore complex. J. Biol. Chem. 273:13347–13352.

41. Popov, S., M. Rexach, G. Zybarth, N. Reiling, M.-A. Lee, L. Ratner, C. M.Lane, M. S. Moore, G. Blobel, and M. Bukrinsky. 1998. Viral protein Rregulates nuclear import of the HIV-1 pre-integration complex. EMBO J.17:909–917.

42. Ramachandra, M., and R. Padmanabhan. 1995. Expression, nuclear trans-port, and phosphorylation of adenovirus DNA replication proteins, vol. 2.Springer-Verlag KG, Berlin, Germany.

43. Roe, T. Y., T. C. Reynolds, G. Yu, and P. O. Brown. 1993. Integration ofmurine leukemia virus DNA depends on mitosis. EMBO J. 12:2099–2108.

44. Rohdewohld, H., H. Weiher, W. Reik, R. Jaenisch, and M. Breindl. 1987.Retrovirus integration and chromatin structure: Moloney murine leukemiaproviral integration sites map near DNase I-hypersensitivity sites. J. Virol.61:336–343.

45. Rosenkranz, A. A., S. V. Yachmenev, D. A. Jans, N. V. Serebryakova, V. I.Muravev, and A. S. Sobolev. 1992. Receptor-mediated endocytosis and nu-clear transport of a transfecting DNA construct. Exp. Cell Res. 199:323–329.

46. Russell, D., I. E. Alexander, and A. D. Miller. 1995. DNA synthesis and


at SE

RIA

LS C

ON

TR

OL Lane M


jvi.asm.org

Dow

nloaded from

http://jvi.asm.org

topoisomerase inhibitors increase transduction by adeno-associated virusvectors. Proc. Natl. Acad. Sci. USA 92:5719–5723.

47. Russell, D. W., A. D. Miller, and I. E. Alexander. 1994. AAV vectors pref-erentially transduce cells in S phase. Proc. Natl. Acad. Sci. USA 91:8915–8919.

48. Russell, W. C., and G. D. Kemp. 1995. Role of adenovirus structural com-ponents in the regulation of adenovirus infection, p. 81–98. In W. Doerflerand P. Boehm (ed.), The molecular repertoire of adenoviruses, vol. 2.Springer-Verlag KG, Berlin, Germany.

49. Schaak, J., X. Guo, W. Y. Ho, M. Karlok, C. Chen, and D. Ornelles. 1995.Adenovirus type 5 precursor terminal protein-expressing 293 and HeLa celllines. J. Virol. 69:4079–4085.

50. Scherdin, U., K. Rhodes, and M. Breindl. 1990. Transcriptionally activegenome regions are preferred targets for retrovirus integration. J. Virol.64:907–912.

51. Schmidt-Zachmann, M. S., C. Dargemot, L. C. Kuehn, and E. A. Nigg. 1993.Nuclear export of proteins: the role of nuclear retention. Cell 74:493–504.

52. Schuitemaker, H., N. A. Koostra, R. A. M. Fouchier, B. Hooibrink, and F.Miedema. 1994. Productive HIV-1 infection of macrophages restricted to thecell fraction with proliferative capacity. EMBO J. 13:5929–5936.

53. Searle, P. F., G. W. Stuart, and R. D. Palmiter. 1987. Metal regulatoryelements of the mouse metallothionein-I gene. EXS 52:407–414.

54. Shimura, M., Y. Onozuka, T. Yamaguchi, K. Hatake, F. Takaku, and Y.Ishizaka. 1999. Micronuclei formation with chromosomal breaks and geneamplification caused by Vpr, an accessory gene of human immunodeficiencyvirus. Cancer Res. 59:2259–2264.

55. Silver, P. A. 1991. How proteins enter the nucleus. Cell 64:489–497.56. Spadari, S., F. Focher, F. Sala, G. Ciarrocchi, G. Koch, A. Falaschi, and G.

Pedrali-Noy. 1985. Controll of cell division by aphidicolin without adverseeffects upon resting cells. Drug Res. 35:1108–1116.

57. Sutton, R. E., M. J. Reitsma, N. Uchida, and P. O. Brown. 1999. Transduc-tion of human progenitor hematopoietic stem cells by human immunodefi-ciency virus type-1 based vectors is cell cycle dependent. J. Virol. 73:3649–3660.

58. Takeshita, S., D. Gal, G. Leclerc, J. G. Pickering, R. Riessen, L. Weir, andJ. M. Isner. 1994. Increased gene expression after liposome-mediated arte-rial gene transfer associated with intimal smooth muscle cell proliferation.J. Clin. Investig. 93:652–661.

59. Temperley, S. M., C. R. Burrow, T. J. Kelly, and R. T. Hay. 1991. Identifi-cation of two distinct regions within the adenovirus minimal origin of repli-

cation that are required for adenovirus type 4 DNA replication in vitro.J. Virol. 65:5037–5044.

60. Temperley, S. M., and R. T. Hay. 1992. Recognition of the adenovirus type2 origin of DNA replication by the virally encoded DNA polymerase andpreterminal proteins. EMBO J. 11:761–768.

61. Tuck, A. B., S. M. Wilson, F. R. Sergovich, and A. F. Chambers. 1991. Geneexpression and metastasis of somatic cell hybrids between murine fibroblastcell lines of different malignant potential. Somat. Cell Mol. Genet. 17:377–389.

62. van der Vliet, P. 1995. Adenovirus DNA replication, p. 1–31. In P. B. W.Doerfler (ed.), The molecular repertoire of adenoviruses, vol. 2. Springer-Verlag KG, Berlin, Germany.

63. Varmus, H. E., T. Padgett, S. Heasley, G. Simon, and J. M. Bishop. 1977.Cellular functions are required for the synthesis and integration of aviansarcoma virus-specific DNA. Cell 11:307–319.

64. Vijaya, S., D. L. Steffen, and H. L. Robinson. 1986. Acceptor sites forretroviral integrations map near DNase I-hypersensitivity sites in chromatin.J. Virol. 60:683–692.

65. Vile, R. G., and S. J. Russell. 1995. Retroviruses as vectors. Br. Med. Bull.51:12–30.

66. Vitadello, M., M. V. Schiaffino, A. Picard, M. Scarpa, and S. Schiaffino. 1994.Gene transfer in regenerating muscle. Hum. Gene Ther. 5:11–18.

67. Vrancken Peeters, M.-J., A. Lieber, J. Perkins, and M. A. Kay. 1996. Methodfor multiple portal vein infusions in mice: quantification of adenovirus-mediated hepatic gene transfer. BioTechniques 20:278–285.

68. Webster, A., I. R. Leith, and R. T. Hay. 1994. Activation of adenovirus-codedprotease and processing of preterminal protein. J. Virol. 68:7292–7300.

69. Wei, S.-Q., K. Mizuuchi, and R. Craigie. 1998. Footprints on the viral DNAends in Moloney murine leukemia virus preintegration complexes reflect aspecific association with integrase. Proc. Natl. Acad. Sci. USA 95:10535–10540.

70. Wei, S.-Q., K. Mizuuchi, and R. Craigie. 1997. A large nucleoprotein assem-bly at the ends of the viral DNA mediates retroviral DNA integration.EMBO J. 16:7511–7520.

71. Wilke, M., E. Fortunati, M. van den Broek, A. T. Hoogeveen, and B. J.Scholte. 1996. Efficacy of a peptide-based gene delivery system depends onmitotic activity. Gene Ther. 3:1133–1142.

72. Zimmer, C. P., H. Grunicke, P. Chandra, and H. Venner. 1971. Influence ofnetropsin and distamycin A on the secondary structure and template activityof DNA. Eur. J. Biochem. 21:269–278.


at SE

RIA

LS C

ON

TR

OL Lane M


jvi.asm.org

Dow

nloaded from

http://jvi.asm.org

Nuclear Import of Moloney Murine Leukemia Virus DNA...

Documents

Transcript of Nuclear Import of Moloney Murine Leukemia Virus DNA...