7) 102–113www.elsevier.com/locate/yviro

Virology 357 (200

Adeno-associated virus interactions with B23/Nucleophosmin:Identification of sub-nucleolar virion regions

Joyce M. Bevington, Patrick G. Needham 1, Kristin C. Verrill, Roy F. Collaco,Venkatesh Basrur, James P. Trempe ⁎

Department of Biochemistry and Cancer Biology, University of Toledo College of Medicine, 3035 Arlington Ave., Toledo, OH 43614-5804, USA

Received 15 June 2006; returned to author for revision 18 July 2006; accepted 24 July 2006Available online 7 September 2006

Abstract

Adeno-associated virus (AAV) is a human parvovirus that normally requires a helper virus such as adenovirus (Ad) for replication. The fourreplication proteins (Rep78, 68, 52 and 40) encoded by AAV are pleiotropic effectors of virus integration, replication, transcription and virionassembly. Using Rep68 column chromatography and mass spectrometry, we have identified the nucleolar, B23/Nucleophosmin (NPM) protein asan Rep-interacting partner. Rep–NPM interactions were verified by co-immunofluorescence and chemical cross-linking studies. We have foundthat there is demonstrable, but limited co-localization between Rep and NPM in co-infected cells. In contrast, there was significant co-localizationbetween NPM and AAV Cap proteins. In vitro experiments using purified MBPRep78 and NPM show that NPM stimulates MBPRep78interactions with the AAV ITR as well as endonuclease activity. These studies suggest that NPM plays a role in AAV amplification affecting Repfunction and virion assembly.© 2006 Elsevier Inc. All rights reserved.

Keyword: AAV–Nucleophosmin co-localization

Introduction

Adeno-associated virus (AAV) is a non-pathogenic mem-ber of the Parvovirus family and the Dependovirus genus(Muzyczka and Berns, 2001). As a Dependovirus, AAV needsanother virus, such as adenovirus, to efficiently replicate insidea host cell. AAV has a linear single-stranded DNA genome of4780 nucleotides (Muzyczka and Berns, 2001). The genomecontains two translation open reading frames (ORF) encodingthree structural and four non-structural proteins and is flanked atboth ends by inverted terminal repeat (ITR) sequences that serveas origins of replication (Lusby et al., 1980; Srivastava et al.,1983). The ORF on the left side encodes four non-structuralproteins, or replication (Rep) proteins designated Rep78,

⁎ Corresponding author. Fax: +1 419 383 6228.E-mail addresses: NeedhamP@niddk.nih.gov (P.G. Needham),

james.trempe@utoledo.edu (J.P. Trempe).1 Present address: Laboratory of Biochemistry and Genetics, NIDDK, NIH,

Bldg. 8, Rm. 407. Bethesda, MD 20892, USA.

0042-6822/$ - see front matter © 2006 Elsevier Inc. All rights reserved.doi:10.1016/j.virol.2006.07.050

Rep68, Rep52 and Rep40 based on their apparent molecularweight in SDS-PAGE gels (Mendelson et al., 1986). Rep78 andRep68 are translated from mRNAs originating from atranscription promoter at map unit 5 (p5). Rep52 and Rep40are translated from mRNAs originating from a transcriptionpromoter at map unit 19 (p19). Rep68 and Rep40 differ fromRep78 and Rep52 as a result of mRNA splicing that replaces 92amino acids from the carboxyl terminus with 9 amino acidresidues. Rep78/68 are required for viral DNA replication,regulation of AAV gene expression and site-specific integrationinto human chromosome 19, which occurs in the absence ofhelper virus infection (Kotin et al., 1990). The smaller Repproteins, Rep52/40, play roles in virus assembly (Chejanovskyand Carter, 1989; King et al., 2001). Rep78 and Rep68 bothinteract with a Rep-binding site (RBS) found in the A-stem ofthe AAV ITR. Both larger Rep proteins also possess ATPase,helicase and site-specific strand-specific endonuclease activ-ities that are important for viral replication (Chiorini et al.,1994; Im and Muzyczka, 1990, 1992). Rep52 and Rep40 arenot endonucleases but share Rep78/68's ATPase and helicase

Table 1

Rep68 interacting partners a

HMG-1SET/TAF-1Acidic Leu-rich nuclear phosphoprotein 32 family member (Anp32)Nucleolin (C23)Nucleophosmin (B23/NPM)a Hela nuclear extracts were passed over an Rep68 column. Adherent proteins

were eluted with KCl, separated by SDS-PAGE and identified by massspectrometry.

103J.M. Bevington et al. / Virology 357 (2007) 102–113

activities (Collaco et al., 2003; Im and Muzyczka, 1992; Smithand Kotin, 1998). Rep78 and Rep68 also have DNA ligaseactivity (Smith and Kotin, 2000). Since there is extensivesequence identity, the two large or two small Rep proteins arenearly interchangeable in terms of function (Collaco et al.,2003; Im and Muzyczka, 1990, 1992; Smith and Kotin, 1998).Three structural, or capsid (Cap or VP), proteins are encoded onthe right side of the genome. A transcription promoter at mapunit 40 (p40) directs the transcription of differentially splicedmRNAs that are translated into the three structural proteinsVP1–3.

AAV and Ad replicate and assemble their genomes in thenucleus of the co-infected cell. AAV Rep and Cap proteins co-localize with the Ad, E2a, single-stranded DNA binding proteinin replication centers (Hunter and Samulski, 1992; Weitzman etal., 1996). AAV capsid proteins also localize in the nucleolus atearly stages of infection, and Rep protein expression is requiredfor capsid proteins to leave the nucleolus (Wistuba et al., 1997).Furthermore, Rep proteins transiently exist in the nucleolus(Wistuba et al., 1997). While searching for cellular factors thatinteract with AAV Rep proteins, we observed in vitro asso-ciations with the abundant nucleolar protein B23/Nucleophos-min (NPM).

NPM is a nucleolar protein with many functions (Okuda,2002). NPM is involved in ribosome biogenesis (Savkur andOlson, 1998; Yung et al., 1985), duplication of centrosomes(Okuda, 2002; Okuda et al., 2000) and shuttling of proteins tothe nucleus (Szebeni et al., 1995, 1997 and has chaperoneprotein characteristics (Szebeni et al., 2003; Szebeni and Olson,1999). Two forms of the protein, referred to as B23.1 and B23.2,arise from differential splicing of mRNA. B23.1, and to a lesserextent B23.2, has ribonuclease activity that can cleave tRNAand mRNA but has specificity for rRNA (Herrera et al., 1995;Savkur and Olson, 1998). Only B23.1 nonspecifically binds tosingle-stranded DNA, double-stranded DNA and RNA (Dum-bar et al., 1989; Herrera et al., 1996; Wang et al., 1994). TheB23/NPM gene is often targeted in chromosomal translocationsassociated with acute myeloid leukemia (AML) resulting inexpression of oncogenic NPM fusion proteins (Redner, 2002;Yoneda-Kato et al., 1996). NPM exerts other effects on cellproliferation in that it associates with Rb, p53 HDM2 andp14ARF (Bertwistle et al., 2004; Colombo et al., 2002; Kurki etal., 2004b; Takemura et al., 1999). These associations arebelieved to play pivotal roles in cellular DNA damage responseand cancer (Kurki et al., 2004a).

Using co-immunofluorescence and co-immunoprecipita-tions, we demonstrate that a portion of the AAV Rep proteinin the co-infected cell associates with NPM in intact nucleoli andin punctate extra-nucleolar structures in the infected nucleus. Atearly stages of infection, Cap proteins are found in punctatereplication/assembly structures that are affiliated with NPM.Using purified Rep78/68 and NPM proteins, we demonstratethat NPM stimulates Rep-specific binding to the AAV ITR andsite-specific endonuclease activities. Our observations that NPMassociates with the AAVCap and Rep proteins expand the cast ofNPM-interacting partners and provide new insights into theAAV replication cycle.

Results

Identification of cellular Rep-interacting partners

The wide array of functions performed by the Rep proteinsduring AAV infection suggests that these relatively smallproteins must interact with cellular proteins to facilitate the virusamplification cycle. Most of the studies that have identifiedRep-interacting partners have used yeast two-hybrid method-ologies. We chose a different approach resulting in theidentification of several novel partners. Rep68, purified fromE. coli, was covalently attached to CnBr-activated sepharoseto form a Rep-column. HeLa nuclear extracts were prepared andfractionated over the column followed by several wash steps toeliminate nonspecific binding. Proteins were eluted from thecolumn in 1 M KCl and separated by SDS-PAGE. The gel wasstained with Coomassie, gel fragments excised and treated withtrypsin to elute peptides for mass spectrometry identification(Table 1). One of the proteins that interacted with the Rep68column was nonhistone HMG-1. Detection of HMG-1 validatesour approach because it has already been identified as a Rep-interacting partner (Costello et al., 1997). The SET/TAF-1 andAnp32 proteins are members of the acidic leucine-rich nuclearphosphoprotein 32 (Anp32) family of protein phosphatase 2A(PPP2, formerly PP2A) inhibitors (Santa-Coloma, 2003). SET/TAF-1 has also been identified as a Rep-interacting partner thatstimulates AAV replication (Pegoraro et al., 2006). SET/TAF-1is also a phosphatase inhibitor that is reported to be necessaryfor Ad DNA replication in vitro (Matsumoto et al., 1995).Nucleolin (C23) and NPM are nucleolar proteins. C23 interactswith NPM and has roles in ribosomal synthesis, can act as a cellsurface receptor, and along with NPM, shuttle proteins from thecytoplasm to the nucleus (Li et al., 1996; Srivastava and Pollard,1999). A role for C23/nucleolin in AAVamplification has beensuggested because it was co-purified with intact AAV2 capsid(Qiu and Brown, 1999).

NPM is not known to play any role in AAV replication.However, NPM is involved in the replication cycles of otherviruses. NPM is redirected to adenovirus replication centersduring adenovirus infection and NPM is also important in invitro DNA replication of the Ad genome (Matthews, 2001;Okuwaki et al., 2001a; Walton et al., 1989). The nucleolus isimportant for the replication of the related Parvovirus, minutevirus of mice (MVM) (Walton et al., 1989). NPM associateswith HTLV-1 Rex, HIV-1 Rev and hepatitis delta antigens(Adachi et al., 1993; Fankhauser et al., 1991; Huang et al.,

104 J.M. Bevington et al. / Virology 357 (2007) 102–113

2001). Given NPM's role in the replication cycle of otherviruses and its involvement in the cellular DNA damageresponse, we investigated its interactions with AAV proteins.

NPM co-localization with Cap and Rep proteins

During the early–middle stages of AAVand Ad co-infection,Rep proteins localize to punctate replication centers in the

Fig. 1. Co-localization of NPM and AAV proteins. HeLa cells were co-infected with A(G–I) and fixed 20 to 30 h post-infection. Fixed cells were stained for AAV Rep (A),yellow or white spots. Fixed cells were also stained for: Ad hexon protein (G and J)Nuclei were stained with DAPI and are depicted in blue in the merged images (C, F, IAAV- and Ad-infected HeLa cells (M–P). Immunofluorescent staining of Cap (M andPanels M–P and O–P are of the same field respectively and are separated by 1.5 μm. Y(For interpretation of the references to colour in this figure legend, the reader is refe

nucleus and co-localize with Cap proteins (Hunter andSamulski, 1992; Weitzman et al., 1996; Wistuba et al., 1997).As replication progresses, the Rep and Cap proteins spreadthroughout the nucleus (Weitzman et al., 1996; Wistuba et al.,1997). Co-immunofluorescence experiments were conducted todetermine if Rep and NPM proteins co-localize in AAV- andAd-co-infected HeLa cells. Immunofluorescence with anti-Repshowed staining throughout the nucleus as well as in punctate

d5 (10 m.o.i.) and AAV2 (250 m.o.i.) (A–F and J–L) or infected with Ad5 aloneCap (D) and NPM (B and E). Merged images (C and F) show co-localization as, NPM protein (H), and AAV Cap (K). These images are also merged (I and L).and L). Scale bar is equivalent to 15 μm. Confocal microscopy was performed onN) and Rep (O and P) is shown in green. NPM staining (M–N) is shown in red.ellow staining is the result of image merging and is indicative of co-localization.rred to the web version of this article.)

105J.M. Bevington et al. / Virology 357 (2007) 102–113

nuclear regions (Fig. 1A). Merging Rep- and NPM-stainedimages showed limited areas of co-localization (Fig. 1C). Aminority of the cells that are Rep positive have some degree ofco-localization between Rep and NPM at early stages ofinfection. Staining with AAVanti-Cap showed similar punctatestructures (Fig. 1D). However, merged Cap- and NPM-stainedimages showed a much greater degree of co-localization (Fig.1F). All of the cells that had punctate Cap staining also had co-localized NPM staining. Moreover, all of the punctate, Cap-stained structures also showed NPM staining. As a negativecontrol for these experiments, co-infected cultures were stainedfor NPM and Ad hexon protein (Figs. 1G–I). Ad hexon is themost abundant outer structural protein. We also stained for AAVCap and hexon (Figs. 1J–L). There was no detectable co-localization of NPM with hexon and only diffuse nuclear andcytoplasmic co-localization between Cap and hexon.

The results shown in Figs. 1A–L were obtained with a stan-dard immunofluorescence microscope. To more conclusivelydemonstrate that Rep and Cap co-localize with NPM, confocalmicroscopy was used. Figs. 1M and N are from the same field,and the images are separated by 1.5 μm. Co-localization of Capand NPM is demonstrated by the yellow color in these images. Itis clear that all of the nucleoli and punctate structuresdemonstrate co-localization. Panels O and P show Rep andNPM staining with co-localization in the swollen nucleoli and inthe punctate structures. These results are consistent with theresults shown in the first part of Fig. 1 but provide moreconvincing evidence of co-localization.

NPM staining in AAV- and Ad-co-infected cells was foundin two structures, numerous small punctate dots throughout thenucleus and large rounded structures that appear to be swollennucleoli (Figs. 2B–D). These structures suggest that nucleoliare remodeled in the co-infected cell. The punctate structureswere often smaller than normal nucleoli, and we have observed

Fig. 2. Nucleolar remodeling during AAVand Ad co-infection. AAV- and Ad-co-infecwere fixed 20 p.i. NPM was stained with AlexaFlour568 (A–D) or FITC (E–H). Astained with AlexaFlour568 (F–H). AlexaFlour568 is a red fluorophore, whereas FITDAPI. Swollen nuclei and small punctate subnuclear structures are evident in paneinterpretation of the references to colour in this figure legend, the reader is referred

as many as fifty per nucleus (Fig. 2D). Cells infected with Adalone will often have multiple nucleoli-like structures, but theyare not as numerous as those seen in AAV- and Ad-co-infectednuclei. Swollen nucleoli have been described previously(Wistuba et al., 1997). These nucleoli are enlarged in Rep-expressing infected cells and had a more intense fluorescence ascompared to uninfected (Rep-negative) cells (Wistuba et al.,1997). Cap co-localization in the swollen nucleoli was observedthroughout the structure (Figs. 1F, M and N). Nearly all of thepunctate structures stain with both NPM and Cap antibodies,suggesting that these may be sites of capsid assembly. Ad-infected cells did not show the abundant sub-nucleolar, NPM-stained structures (Figs. 2F–H).

AAV Cap, Rep and NPM association detected by chemicalcross-linking and immunoprecipitation

To further investigate if there is a physical interactionbetween AAV proteins and NPM, we co-immunoprecipitatedthese proteins with specific antibodies. Our initial attempts todetect an association between NPM and the AAV Rep proteinsusing co-immunoprecipitations from AAV- and Ad-co-infectedcells were unsuccessful, therefore we used dithiobis (succini-midyl) propionate (DSP) to chemically cross-link weakly inter-acting proteins. Nuclei were isolated from virus-infectedcultures and cross-linked with DSP. Nuclear extracts wereprepared and immunoprecipitated with antisera against Rep, Capand NPM. The cross-links were reversed and proteins separatedby gel electrophoresis and immunoblotted. When extracts wereimmunoprecipitated with anti-Rep, NPM was co-immunopreci-pitated from the co-infected culture, but only when chemicallycross-linked (Fig. 3A, lane 5). NPM was not precipitated fromthe uncross-linked extracts (Fig. 3A, lane 4). Conversely,Rep proteins were co-immunoprecipitated with anti-NPM

ted HeLa cells (B–D), Ad infected HeLa cells (F–H) and uninfected cells (A, E)AV Cap proteins were stained with AlexaFlour488 (B–D), and Ad hexon wasC and AlexaFlour488 are green fluorophores. Nuclei were counterstained withls B–D. All panels are merged images. Scale bar is equivalent to 15 μm. (Forto the web version of this article.)

Fig. 3. Co-immunoprecipitation of NPM and AAV proteins. Ad- or Ad- andAAV-co-infected HeLa cells were harvested 22 h post-infection, nuclei wereisolated and chemically cross-linked with 2 mM DSP. Antibodies to Rep andNPMwere used for immunoprecipitation, the cross-linked were reversed and theprecipitates separated by SDS-PAGE. The separated proteins were analyzed byWestern blot using NPM, Rep or Cap antibodies. The extracts (Ext.) used werefrom Ad-infected (Ad) or AAV- and Ad-co-infected (Co) cells. The use of theDSP cross-linking agent (XL) is indicated by ‘+’ or ‘−’. Lanes 1 and 2 in eachpanel are crude extracts loaded on the gel without prior immunoprecipitation.

106 J.M. Bevington et al. / Virology 357 (2007) 102–113

from cross-linked extracts whereas no Rep co-immunoprecipita-tion was observed without cross-linking (Fig. 3B, compare lanes4 and 5). To detect Cap and NPM interactions, infected extractswere immunoprecipitated with anti-NPM. This experimentshowed that NPM antibody co-imunoprecipitated AAV struc-tural proteins from both untreated and cross-linked co-infectednuclear extracts (Fig. 3C, lanes 4 and 5). We have been unable toco-immunoprecipitate NPM with either a polyclonal anti-Capantibody obtained from denatured VP3 or the A20 monoclonalantibody that recognizes native capsid (results not shown).These results demonstrate a close association, if not a physicalinteraction, between NPM and AAV proteins.

NPM stimulates MBPRep78 binding to the AAV ITR RBS

Immunofluorescence and chemical cross-linking showed co-localization and close association between Cap and NPMproteins. Since Rep proteins associate with the maturing virionvia noncovalent interactions, and covalently, via linkage to the5′ end of the viral DNA, it is not possible to determine if there isa direct interaction between NPM and the Rep proteins. In vitroEMSA analyses were performed to investigate potential Rep–NPM interaction. AAV Rep78/68 proteins interact with the RBSfound in the A-stem of the AAV ITR element (Muzyczka andBerns, 2001). An A–D stem DNA fragment was radiolabeledand incubated with purified MBPRep78 and NPM. Fig. 4 showsthat NPM stimulated MBPRep78 binding to a DNA probecontaining its cognate binding site. Both a His-tagged and aGST-tagged NPM stimulated MBPRep78 binding. Thus, theincreased binding does not depend on the tag used to purifyNPM (Fig. 4A; compare lanes 5 and 6 to lane 4). AlthoughNPM stimulated binding, no new complexes were observed.This suggests that NPM stimulated binding in a transient mannerand that the chaperone activity of NPM may be involved in thisbinding.

NPM did not show any detectable binding to AAV DNA byitself (Fig. 4A, lanes 2 and 3) even though it nonspecificallybinds to DNA (Dumbar et al., 1989; Wang et al., 1994). Aprobable reason for this is that conditions used in thisexperiment were not optimal for NPM binding to DNA sinceNaCl concentration in the buffer was 0.05 M and NPM binds toDNA at lower salt concentrations (Dumbar et al., 1989; Herreraet al., 1996; Wang et al., 1994).

To further demonstrate the enhanced binding of MBPRep78,a dose–response assay was performed with increasing amountsof NPM or BSA. Fig. 4B shows that, with increasing concen-trations of NPM, MBPRep78 interaction with the AAV RBSincreases. However, BSA did not stimulate binding (Fig. 4C).Similar molar ratios of MBPRep78 to BSAwere used in EMSAassays to verify that NPM stimulation, rather than anunspecified stimulatory effect, results in increased MBPRep78binding. This verifies that increased binding of MBP-Rep78 isdependent upon NPM.

Increased Rep-mediated nicking at the AAV trs site with NPM

Rep78 or Rep68 interaction with the AAV ITR is required forsite-specific nicking of the covalently closed end of the viralgenome. Upon binding and oligomerization, Rep78 or Rep68make a site-specific, strand-specific nick in the terminalresolution site (trs) enabling completion of DNA synthesis atthe ends of the viral DNA. To determine if NPM affects Rep-mediated nicking, endonuclease assays were performed withincreasing amounts of NPM or BSA. Purified MBPRep78 andGSTNPM were incubated with a radiolabeled 182 nt ITRendonuclease substrate. Fig. 5A shows that increasing amountsof NPM stimulate nicking, yielding the appropriate, 73 ntproduct. Equimolar concentrations of BSA had no effect onMBPRep78 nicking (Fig. 5B). This result is consistent with theEMSA analyses because stimulation of binding to the A–D

Fig. 4. NPM stimulates MBPRep78 interaction with the AAV ITR. A 65-bp DNA fragment from the A–D component of the AAV ITR was radiolabeled and incubatedwith NPM and MBPRep78 (designated Rep78 in the figure). Protein–DNA complexes were separated by non-denaturing gel electrophoresis and exposed to X-rayfilm. (A) 3.4 nMMBPRep78 was used in lanes 4, 5 and 6, 14.3 nM GSTNPM was used in lanes 2 and 5 and 13.8 nM HisNPM was used in lanes 3 and 6. (B) 8.9 nMMBPRep78 was used in lanes 3 to 8, and 4.7, 9.4, 18.7, 37.5, 74.9 and 93.5 nM GSTNPM were used in lanes 4, 5, 6, 7, 8 and 2, respectively. (C) 3.4 nM MBPRep78was used in lanes 3 to 8. 1.8, 3.5, 7, 14, 28 and 28 nM of BSA were used in lanes 4, 5, 6, 7, 8 and 2, respectively. Equimolar amounts of BSA:MBPRep78 andGSTNPM:MBPRep78 were used in the respective lanes (lanes 4 to 8) of panels B and C.

107J.M. Bevington et al. / Virology 357 (2007) 102–113

oligonucleotide would be expected to result in increased nickingactivity.

Discussion

The minimal coding capacity of the Parvovirus genomedictates that numerous cellular proteins are required to supportvirus amplification. Several cellular proteins have beenidentified that interact with AAV proteins. Most of the proteinsthat interact with Rep play roles in mRNA transcription. Rep78/68 interact with: Sp1 (Hermonat et al., 1996; Pereira andMuzyczka, 1997), HMG-1 (Costello et al., 1997), the transcrip-

Fig. 5. NPM stimulates MBPRep78 endonuclease activity. MBPRep78(1.34 nM, designated Rep78) was incubated with radiolabeled ITR with ATPto induce endonuclease activity. The reactions were separated by denaturingpolyacrylamide gel electrophoresis and exposed to X-ray film. (A) Lanes 4, 5, 6,7 and 2 contain to 0.8, 4, 20, 40 and 61 nM of HisNPM, respectively. (B) Lanes3, 4, 5, 6 and 1 contain 0.8, 4, 20, 40 and 60 nM of BSA, respectively.

tional co-activator PC4 (Weger et al., 1999), TATA bindingprotein (Hermonat et al., 1998; Needham et al., 2006), the p53and topoisomerase binding protein, Topors (Weger et al., 2002),a putative protein kinase, protein kinase X (PKX) and proteinkinase A (PKA) (Chiorini et al., 1998; Di Pasquale and Chiorini,2003). The biological effect of the PKX association is inhibitionof the steady-state levels of cAMP-responsive-element-bindingprotein (CREB) and cyclin A protein. Nucleolin, also known asC23, co-purifies with AAV2 and was found to associate withintact AAV2 capsid by immunoprecipitation and immuno-fluorescence techniques (Qiu and Brown, 1999). C23/nucleolinis a ubiquitous nucleolar protein implicated in nucleartransport, organization of nucleolar chromatin, packaging ofpre-RNA, rDNA transcription and ribosome assembly (Srivas-tava and Pollard, 1999). Thus, there are only a limited numberof cellular proteins known to interact with AAV-encodedproteins.

To identify additional cellular proteins that interact withRep78/68, we attached purified Rep68 to activated sepharosecreating a Rep affinity chromatography resin. HeLa nuclearextracts were passed over the column and eluted with high salt.Mass spectrometry analyses identified several new potentialRep partners. One of the proteins identified was HMG-1. Thisobservation verifies the validity of our approach becauseHMG-1 has already been identified as a Rep-interacting partner(Costello et al., 1997). Two of the identified proteins areinvolved in protein phosphatase activity: protein SET/TAF-1and Anp32. The SET/TAF-1 and Anp32 proteins are membersof the acidic leucine-rich nuclear phosphoprotein 32 family ofprotein phosphatase 2A (PPP2, formerly PP2A) inhibitors(Santa-Coloma, 2003). SET/TAF-1 was first identified as a genefused to the CAN gene in a patient with acute undifferentiatedleukemia (von Lindern et al., 1992). SET was subsequentlyfound to be identical to template activation factor I (TAF-1), acellular protein necessary for DNA replication of the Adgenome in vitro (Matsumoto et al., 1995). C23/nucleolin andB23/NPM were also identified. Recently, SET/TAF-1 was

108 J.M. Bevington et al. / Virology 357 (2007) 102–113

shown to be a Rep-interacting partner that regulates AAV DNAreplication (Pegoraro et al., 2006). We focused on NPM becauseof its wide array of functions in normal and transformed cells.

B23/NPM is a nucleolar protein with many proposedfunctions (Okuda et al., 2000). It exists in two forms, B23.1and B23.2, which are nearly identical except at the C-terminalend (Chang and Olson, 1989, 1990). The protein sequence ishighly conserved among rat, mouse and human. B23.2, which isthe alternatively spliced form of B23.1, replaces 36 amino acidsat the C-terminus of B23.1 with 2 amino acids (Chang andOlson, 1989, 1990). The more abundant B23.1 protein existspredominantly in the nucleolus (Okuwaki et al., 2002; Wanget al., 1993). NPM has a variety of proposed functions includingribosome biogenesis (Okuwaki et al., 2002; Savkur and Olson,1998; Yung et al., 1985), histone chaperone (Okuwaki et al.,2001b), duplication of centrosomes (Okuda, 2002; Okuda et al.,2000), shuttling proteins to the nucleus (Borer et al., 1989;Szebeni et al., 1995, 1997, 2003) and has general chaperoneprotein characteristics (Szebeni and Olson, 1999). B23.1, and toa lesser extent B23.2, has ribonuclease activity that cleavestRNA and mRNA but has specificity for rRNA (Herrera et al.,1995; Savkur and Olson, 1998). Only B23.1 nonspecificallybinds to single-stranded DNA, double-stranded DNA and RNA(Dumbar et al., 1989; Herrera et al., 1996; Wang et al., 1994).NPM has also been implicated in the initial cellular response toenvironmental stressors. DNA damaging agents such as UVlight induces expression (Weber et al., 2000) and nuclear re-localization of NPM (Kurki et al., 2004a) thus stimulating DNArepair and escape from apoptosis (Wu et al., 2002). NPM is partof a large protein complex that includes the p53 tumor sup-pressor HDM2 and p19Arf (Colombo et al., 2002; Itahana et al.,2003; Korgaonkar et al., 2005; Kurki et al., 2004b). NPMincreases the stability and transcriptional potential of p53 via itschaperone activity and inhibition of Mdm2 ubiquitin ligaseactivity (Colombo et al., 2002; Kurki et al., 2004b).

Our co-immunofluorescence and chemical cross-linkingstudies demonstrate that NPM is found in close proximity toAAV Rep and Cap proteins. Intra-nuclear Cap and NPM co-localization was prevalent with domains showing highconcentrations of Cap enriched in NPM. However, Rep andNPM co-localization was not as prevalent as that observed withCap proteins. Immunofluorescence using confocal microscopyfurther substantiated the hypothesis that AAV proteins co-localize with NPM. From our chemical cross-linking experi-ments, the relatively short length of the DSP cross-linker (12 Å)suggests that NPM exists in close proximity to Cap or Repproteins. Since Rep proteins associate with the virion or viralDNA by noncovalent and covalent interactions, respectively(Dubielzig et al., 1999; Prasad and Trempe, 1995), it is notpossible to verify that NPM directly interacts with both viralproteins in the co-infected cell. That AAV Cap proteins are co-immunoprecipitated with anti-NPM from non-cross-linkedextracts suggests that there is a stronger affinity between theseproteins than between NPM and Rep. Our inability to co-immunoprecipitate NPM with Cap-specific antibodies suggeststhat NPM interactions may block the structural epitopesrequired for immunoprecipitations.

We performed in vitro studies with purified Rep and NPM tofurther investigate whether these proteins interact. NPMstimulated MBPRep78 interactions with the AAV ITR andRep endonuclease activity. NPM by itself did not interact withthe ITR, even though it has DNA binding ability. NPM purifiedas either a His-tagged and GST-tagged fusion protein stimulatedITR binding. Neither version of the protein appeared to becomepart of the MBPRep78–ITR complex in that no new specieswere found in the EMSA. Stimulation of MBPRep78 binding tothe ITR would be expected to result in increased endonucleaseactivity as we observed here. Although NPM stimulatedMBPRep78–ITR binding and endonuclease activities, it didnot stimulate Rep-specific ATPase activity (data not shown).These experiments suggest that NPM stimulates Rep bindingvia its chaperone activity (Szebeni and Olson, 1999). In theinfected cell, NPM may also interact with Rep in a transientmanner.

It will be interesting to determine which region of the Repproteins is responsible for the interaction with Nucleophosmin(it is presumably not the C-terminus since both Rep 68 and 78interact) and if mutant studies will yield an idea of thesignificance of the Rep–Nucleophosmin interaction.

Early in infection Rep proteins are primarily nuclear and areobserved in a punctate pattern. As infection progresses, Repproteins pass transiently through the nucleolus followed by co-localization with AAV DNA as demonstrated by in situhybridization. Cap proteins also transiently pass through thenucleolus. AAV DNA, Rep and Cap proteins co-localize inpunctate replication/encapsidation centers (Hunter andSamulski, 1992; Weitzman et al., 1996; Wistuba et al., 1997).Expression of the AAV Cap gene in the absence of Rep or AAVDNA replication in HeLa cells resulted in Cap enrichment in thenucleolus (Weger et al., 1997; Wistuba et al., 1997). It has beenproposed that Rep expression may be required for assembledcapsid to escape the nucleoli (Wistuba et al., 1997). Aninteresting observation from the immunofluorescence studiesis the apparent nucleolar remodeling that occurs in AAV- andAd-co-infected cells. In addition to the appearance of swollennucleoli, there were numerous punctate structures that weresmaller than nucleoli from uninfected cells. The near totalcongruence of Cap and NPM staining in these structuressuggests that capsid assembly and/or genome encapsidation mayoccur at these sub-nucleolar structures.

It is unclear which of the numerous functions of NPM isinvolved in AAVamplification. Our in vitro studies suggest thatthe NPM chaperone activities may play a role in Rep proteinfunction in vivo. A transient interaction between Rep and NPMmay also be reflected in the minimal co-localization betweenthese proteins observed in our co-immunofluorescence studies.The prevalence of NPM and Cap co-localization suggests thatthese proteins are more intimately involved. The accumulationof Cap proteins in the nucleolus in the absence of Rep protein orAAV DNA suggests that capsid assembly occurs in thenucleolus (Wistuba et al., 1997). Our results, and thosedescribed above, support a model of AAV assembly in whichCap proteins accumulate and assemble in the nucleolus wherethey associate with NPM and perhaps other nucleolar proteins.

109J.M. Bevington et al. / Virology 357 (2007) 102–113

Passage of Rep proteins through the nucleolus allows for exit ofRep–Cap–NPM complexes from the nucleolus and into thenucleoplasm where viral DNA encapsidation occurs. Little isknown about AAV capsid assembly in vivo (Timpe et al., 2005).It will be informative to determine if NPM plays a role in AAVassembly.

Methods

Cells and virus preparation

HeLa cells (American Type Culture Collection) were grownas a monolayer at 37 °C (5% CO2 atmosphere) in Eagle'sminimum essential medium (MEM) supplemented with 10%(v/v) fetal bovine serum (FBS), 2 mM L-glutamine, 25 U/mLpenicillin, 25 μg/mL streptomycin, 2.5 μg/mL amphotericin Band 100 μg/mL gentamicin. AAV2 and Ad5 were prepared bythe method previously described (Casper et al., 2005; Wintersand Russell, 1971).

Protein purification

An amino-terminal His6 tagged NPM (B23.1 version) in apQE30 prokaryotic expression vector (pHis-B23.1) inSG13009 cells was kindly provided by Dr. Mark Olson(Umekawa et al., 2001). A 5 mL overnight Luria-Bertani(LB) starter culture was used to inoculate a 1 L LB cultureand incubated at 37 °C until an A600 of ∼0.6 was reached.The culture was induced with 1 mM IPTG for 2.5 h at 37 °C.The cells were pelleted at 9000×g for 15 min at 4 °C andthen washed once with 30 mL phosphate buffered saline(PBS). Cells were pelleted again at 9000×g for 10 min at 4 °C.Cells were lysed in 20 mL of lysis buffer (50 mM NaPO4,500 mM NaCl, 10 mM imidazole, 1 mM β-mercaptoethanol,40 μg/mL lysozyme, pH 8.0) on ice for 20 min. Cells weresonicated using a Fisher Sonic Dismembrator 550 on ice level3 with 1 min on and 10 s off three times. The lysate waspelleted for 10 min 12,000×g at 4 °C. The supernatant wasapplied to 0.5 mL column of Ni-NTA Superflow (Qiagen)equilibrated in lysis buffer. The column was washed with ninecolumn volumes of cold wash buffer (50 mM NaPO4,500 mM NaCl, 50 mM imidazole, 1 mM β-mercaptoethanol,pH 8.0). HisNPM was eluted off the column with cold elutionbuffer (50 mM NaPO4, 500 mM NaCl, 250 mM imidazole,1 mM β-mercaptoethanol, pH 8.0). The purified protein wasdialyzed overnight in 20 mM Tris pH 8.0, 10% glycerol,200 mM NaCl.

An amino-terminal glutathione-S-transferase (GST) taggedB23.1 (GSTNPM) prokaryotic expression vector (pGST-B23.1)was constructed as follows. The B23.1 cDNAwas excised frompHis-B23.1 with BamHI and inserted into the BamHI multi-cloning site of the pGEX-6P-2 (Amersham Biosciences) vector.For protein expression, BL21 Star (DE3) was transformed withthe plasmid. Purification of GSTNPM was as follows. A 5 mLovernight LB starter culture was used to inoculate a 250 mL LBculture and incubated at 37 °C until an A600 of ∼0.6 wasreached. The culture was induced with 1 mM IPTG for 2.5 h at

37 °C. The cells were pelleted at 10,000×g for 10 min at 4 °Cand then washed once with 30 mL PBS. The pellet wasresuspended in 5 mL cold PBS. The cells were lysed using aFisher Sonic Dismembrator 550 on ice level 4 for 30 s. TritonX-100 was added for a final concentration of 1%. The lysatewas pelleted for 5 min 4 °C at 16,000×g. The supernatant wasadded to 1 mL of 50% glutathione–agarose (Sigma) bead slurry.The 50% glutathione–agarose bead slurry was preparedfollowing manufacturers' instructions. The protein extract wasmixed with the slurry for 2 min at 4 °C. The beads were pelletedat 500×g for 1 min at 4 °C. The bound protein was washedtwice with 50 mL cold PBS+1 M NaCl and then twice with50 mL cold PBS. GSTNPM was eluted with 500 μL cold50 mM Tris pH 7.5+5 mM reduced glutathione. Glycerol wasadded for a final concentration of 10%.

An amino-terminal His6 tagged Rep68 (HisRep68) in apQE70 prokaryotic expression vector was kindly provided byDr. R.J. Samulski (Young et al., 2000). Purified bacteriallyexpressed HisRep68 was prepared as described previously(Casper et al., 2005).

An amino-terminal maltose binding protein (MBP) taggedRep78 (MBPRep78) in a pPR997 prokaryotic expression vectorwas kindly provided by Dr. R.M. Kotin (Chiorini et al., 1994).Purified bacterially expressed MBPRep78 was prepared asdescribed previously (Needham et al., 2006).

Rep-column chromatography

HisRep68 was attached to the activated CNBr-sepharose(Sigma catalog #C5338) by the procedure described by themanufacturer with slight modifications. 1.5 mg of HisRep68 in1 mL volume was dialyzed for 4 h in 100 mL of 100 mMNaHCO3, 500 mM NaCl, pH 8.3. The buffer was discarded anddialysis continued for 2 h in 100 mL fresh buffer. The CNBr-resin was prepared by swelling in 1 mM HCl, washed withdistilled water and then with coupling buffer (0.1 M NaHCO3,0.5 M NaCl, pH 8.3). HisRep68 from dialysis was addedimmediately to 300 μL of the prepared resin and allowed torotate overnight at 4 °C. The following day the resin waswashed twice in cold coupling buffer and then mixed for 2 hwith blocking buffer (100 mM Tris–Cl, 500 mM NaCl, pH8.0). The blocking buffer was washed away and the resin wasequilibrated in HeLa extract Buffer D (20 mM HEPES pH 7.9,100 mM KCl, 0.2 mM EDTA, 20% glycerol). HeLa nuclearextracts were prepared following the method described(Dignam et al., 1983). To look for Rep-interacting proteins,400 μL of HeLa nuclear extract (in Buffer D) was mixed with100 μL of the Rep-column resin and rotated at 4 °C for 2 h. Theresin was allowed to settle, and the supernatant was removed.The resin was washed three times with 1 mL of Buffer D+300 mM KCl and the supernatant saved. This was followedby a wash of 250 μL of Buffer D+500 mM KCl and 250 μL ofBuffer D+1 M KCl. Samples of these fractions were separatedby SDS-PAGE gels and silver stained. Fractions of interestwere separated by SDS-PAGE and stained with colloidalCoomassie G-250. Bands were cut from the gel and sent formass spectrometry analysis.

110 J.M. Bevington et al. / Virology 357 (2007) 102–113

Protein identification by LC–tandem MS

The proteins were separated on SDS-PAGE and visualizedwith colloidal Coomassie stain (Invitrogen). Protein bands wereexcised and destained with 30% methanol for 3 h at RT. In gelproteolysis with modified, sequencing grade trypsin (Promega,Madison, WI) was carried out essentially as described pre-viously (Basrur et al., 2003). Briefly, gel slices were furtherwashed with 150 μL of 50% acetonitrile in 0.1 M ammoniumbicarbonate buffer, pH 8.0, for 30 min, trypsin (0.5 μg,Promega) was added in a minimal volume of 0.1 M ammoniumbicarbonate buffer and the digestion was carried out for 16 h at37 °C with an additional aliquot of trypsin (0.25 μg) added after12 h. Peptides were extracted sequentially with 150 μL of 60%acetonitrile containing 0.1% TFA for 30 min and 100 μL ofacetonitrile containing 0.1% TFA. All extracts were pooled andconcentrated using Vacufuge to a final volume of 15 μL. Twomicroliters of the digest was separated on a reverse phasecolumn (Aquasil C18, 15 μm tip×75 μm id×5 cm Picofritcolumn, New Objectives, Woburn, MA) using acetonitrile/1%acetic acid gradient system (5–75% acetonitrile over 35 minfollowed by 95% acetonitrile wash for 5 min) at a flow rate of∼250 nL/min. Peptides were introduced into an in-line, ion trapmass spectrometer (LCQ Deca XP Plus, ThermoFinnigan)equipped with a nano-spray source. The mass spectrometer wasset for analyzing the positive ions and acquiring a full MS scanand a collision induced dissociation (CID) spectrum on the mostabundant ion from the full MS scan (relative collision energy∼30%). Dynamic exclusion was set to collect 3 CID spectra onthe most abundant ion and then exclude it for 3 min. Databasesearch against an indexed, non-redundant human proteindatabase was performed using TurboSEQUEST software(BioworksBrowser v 3.0, ThermoFinnigan). CIDs whichshowed Xcorr and μCn values of >2.0 and >0.2, respectively,for a +2 charged peptide, were considered positive. All CIDspectra were also verified manually using the MS-Digest andMS-Product provision of Protein Prospector (http://prospector.ucsf.edu).

Immunofluorescence

Antibodies used in these experiments were as follows: rabbitpolyclonal antibody affinity purified against all four Repproteins (Trempe et al., 1987) and Cap proteins (dataunpublished), anti-nucleophosmin (Zymed cat.#32-5200), goatpolyclonal anti-hexon (American Research Products, Inc. cat.#12-6235-1), goat anti-mouse AlexaFlour568 (MolecularProbes cat.#A-11031), donkey anti-goat AlexaFlour568 (Mole-cular Probes cat.#A-11057), donkey anti-rabbit AlexaFlour488(Molecular Probes cat.#A21206), donkey anti-mouse FITC and4′,6-diamidino-2-phenylindole, dihydrochloride (DAPI) (Mole-cular Probes). HeLa cells were plated 24 h prior to infection at1×104 cells/well in each well of an 8-chamber slide (LAB-TEKBrand, Nalge Nunc International). Cells were infected or mockinfected for 1 h in 200 μL of 2% (v/v) FBS MEM medium(supplemented with L-glutamine) with Ad5 and AAV2 at amultiplicity of infection (m.o.i.) of 10 and 250, respectively.

After 1 h, an equivalent amount of 18% (v/v) FBS MEMmedium (supplemented with L-glutamine and antibiotics) wasadded to the cells. At 20–30 h, cells were washed twice withPBS. Cells were then fixed and permeabilized with 100% coldmethanol for 10 min on ice and then air-dried for 10 min. Theslides were washed twice with 2% bovine serum albumin (BSA)in PBS (BSA–PBS) and blocked in the same solution for20 min. The slide chambers were removed leaving the gasket.The slides were incubated in RIPA buffer (50 mM Tris pH 8,150 mM NaCl, 0.5% deoxycholate, 0.1% SDS, 1% nonidetP40) for 10 min and then washed three times with BSA–PBS.Slides were incubated with primary antibody in BSA–PBS for1 h followed by three washes with BSA–PBS. Slides wereincubated with secondary antibody in BSA–PBS for 1 hfollowed by three washes with BSA–PBS. Nuclei were stainedwith 150 nM DAPI for 5 min followed by three washes withBSA–PBS. A final three washes with just PBS were performedbefore removing the slide gasket and mounting the slide withDAKO Fluorescent Mounting Medium (DAKO Corporation).All antibody and wash incubations were done at room tempe-rature. A Nikon eclipse E800 fluorescent microscope was usedfor visualization of co-localization. A BioRad Radiance 2000Laser Scanning system mounted on an Olympus BX51WImicroscope was used to confirm co-localization.

Cross-linking and immunoprecipitation

HeLa cells were grown in 150 mm plates to 95% confluenceand infected. The infection was performed in serum free mediumwith Ad5 and/or AAV2 at an m.o.i. of 10 and 100, respectively.After 2 h of incubation at 37 °C, the medium was replaced with10% (v/v) FBS MEM supplemented with L-glutamine andantibiotics. Twenty-two hours later, the cells were scraped fromthe plate and spun down at 1500×g for 5 min at 4 °C. Thepelleted cells were washed twice with 5 ml cold PBS+5 mMMgCl2. The washed pellets were re-suspended and incubated onice for 10 min in 500 μL cold Buffer A (10 mM HEPES (pH7.8), 10 mM KCl, 1.5 mM MgCl2 0.4% Triton X-100, 0.34 Msucrose, 10% (v/v) glycerol, 1 μM leupeptin, 1 μg/mL pepstatinA, 1 mMPMSF). The lysate was pelleted at 1500×g for 5 min at4 °C. The supernatant (cytoplasmic extract) was carefullyremoved, the nuclear pellet was re-suspended in 500 μL coldBuffer A+200 mM NaCl and vortexed for 5 s. Dithiobis(succinimidyl) propionate (DSP) (Pierce Chemical Company)dissolved in dimethyl sulfoxide (stock=50 mM) was added tothe re-suspended pellet to a final concentration of 2 mM andvortexed for 5 s. This treated nuclear fraction was rotated for20 min at room temperature. The cross-linking was stopped bythe addition of 50 mM glycine and further rotation for 10 minat room temperature. EDTA was added to a final concentrationof 20 mM; the cross-linked nuclear fraction was vortexedvigorously for 30 s and incubated on ice for 1 h. The lysedcross-linked extract was sonicated using a Fisher SonicDismembrator 550 on ice (level 4, one-second pulses, for45 s). SDS was added (final concentration of 1% (wt/v)), andthe lysed extract was denatured at 65 °C for 10 min. Meanwhile,for each immunoprecipitation, 75 μL of Immunopure protein A

111J.M. Bevington et al. / Virology 357 (2007) 102–113

agarose bead slurry (Pierce) was washed with 500 μL of IPPbuffer (50 mM Tris (pH 8.0), 20 mM EDTA (pH 8.0), 150 mMNaCl, 0.5% NP-40). The beads were rotated at 4 °C for 5 min,pelleted at 500×g for 2 min and the supernatant was aspirated.Five hundred microliters of IPP buffer was added to theseequilibrated beads followed by 100 μL of the earlier heattreated, denatured extracts. The extracts were pre-cleared byrotation for 1 h at 4 °C. The beads were then pelleted at 500×gfor 2 min, and the supernatant containing the cleared extractswas carefully added to 75 μL of freshly equilibrated beads(washed as above). Appropriate antibodies were added, and theimmunoprecipitation was performed by rotation at 4 °C for16 h. The antibodies used in the immunoprecipitations were:anti-NPM, anti-Rep and rabbit anti-Cap raised against dena-tured VP3 (described above). The beads were pelleted (500×gfor 2 min at 4 °C) and washed three times with IPP buffer. Theproteins were eluted from the beads and cross-links broken bythe addition of loading buffer (1.5× SDS-PAGE Sample buffer,10% (v/v) β-mercaptoethanol) and heating to 100 °C for10 min. Eluted proteins were run on a 10% SDS-PAGE gel andtransferred to a nitrocellulose membrane (Immobilon-FisherScientific Co.). Transferred proteins were Western blotted asindicated and detected by chemiluminescence.

EMSA probe preparation

The A–D stem sequence used for Fig. 4B is as follows:TCCAGGAACCCCTAGTGATGGAGTTGGCCACTCCCTC-TCTGCGCGCTCGCTCGCTCACTGAGGC and Phos-GCCTCA-GTGAGCGAGCGAGCGCGCAGAGAGGGAGTGGCCAA-CTCCATCACTAGGGGTTCCTGGA (InvitrogenCustomPrimers).The A–D stem sequence used for Figs. 4A and C is as follows:CCTAGTGATGGAGTTGGCCACTCCCTCTCTGCGCGCT-CGCTCGCTCACTGAGGC and Phos-GCCTCAGTGAGC-GAGCGAGCGCGCAGAGAGGGAGTGGCCAACTCCAT-CACTAGG (Invitrogen Custom Primers). Both A–D stemsequences contain the Rep binding element (RBE) and the trsnicking site (Chiorini et al., 1995; Muzyczka and Berns, 2001).The oligos were annealed following the manufacturers' recom-mendations. Briefly, the oligos were annealed together in a 50 μLreaction (100 mM Tris pH 7.5, 500 mM NaCl, 50 mM EDTA),which was heated in sand bath at 65 °C for 10 min and thenallowed to cool slowly to room temperature. One picomole ofannealed oligos was incubated with 20 U of T4 polynucleotidekinase and 50 μCi [γ-32P] ATP. The reaction went 16 h 37 °C andpurified over a 1 mL G-25 Sephadex column in TE.

Electrophoretic mobility shift assay (EMSA)

EMSAs were performed following method previouslydescribed (Needham et al., 2006). Briefly, the radiolabeledA–D substrate was incubated in a 20 μL reaction with purifiedMBPRep78 and/or purified GSTNPM (B23.1) or BSA inbinding buffer (10 mM Tris pH 7.5, 50 mM NaCl, 4% (v/v)glycerol, 1 mM MgCl2, 0.44 mM EDTA, 0.5 mM ditiothreitol(DTT), 12.5 μg polydeoxyinosinic-deoxycitidylic (Sigma) permL, 50 μg of BSA per mL) for 20 min at room temperature. The

mixture was run on a non-denaturing 4% polyacrylamide TBEgel and vacuum dried. The gels were exposed to BioMax MRFilm (Kodak).

Endonuclease activity assay

The hairpin probe was prepared as follows. The AAV ITRswere excised from psub201 (Samulski et al., 1989) with XbaIand PvuII and calf intestinal phosphatase (CIP) treated. Onepicomole of the purified CIP treated fragment was incubatedwith 20 U of T4 polynucleotide kinase and 50 μCi [γ-32P] ATP.The reaction went 16 h 37 °C and purified over a 1 mL G-25Sephadex column in TE. The labeled fragment was heated in aboiling water bath for 5 min and then snap cooled on iceforming the hairpin probe. The endonuclease activity assay wasperformed as described previously with slight modifications (Imand Muzyczka, 1992; Li et al., 2003). Briefly, the labeledhairpin probe (5 fmol) was incubated with the purifiedMBPRep78 (1.34 nM) and/or purified HisNPM or BSA in a20 μL reaction containing: 25 mM HEPES·KOH pH 7.5,10 mM NaCl, 5.5 mM MgCl2, 0.5 mM ATP, 0.2 mM DTT,0.25% Tween 20, and 10 μg of BSA per mL. The reactionswere incubated at 37 °C for 1 h. The reaction was stopped with20 μg of proteinase K incubated for 30 min at 37 °C. Theproducts were phenol/chloroform extracted and then ethanolprecipitated. The pellet was resuspended in 2 μL of 10× agaroseloading buffer and 18 μL of dH2O and run on a 10% denaturingpolyacrylamide gel containing 50% urea and vacuum dried. Thegels were exposed to BioMax MR Film (Kodak). The cleavageproduct is 73 nucleotides.

Acknowledgments

Wewould like to thankChristian Peters andDr. JeanOvermeyerfor valuable assistance and insights on the immunofluorescenceexperiments. This work was supported in part by NIH GM64765and AI51471.

References

Adachi, Y., Copeland, T.D., Hatanaka, M., Oroszlan, S., 1993. Nucleolartargeting signal of Rex protein of human T-cell leukemia virus type Ispecifically binds to nucleolar shuttle protein B-23. J. Biol. Chem. 268 (19),13930–13934.

Basrur, V., Yang, F., Kushimoto, T., Higashimoto, Y., Yasumoto, K., Valencia,J., Muller, J., Vieira, W.D., Watabe, H., Shabanowitz, J., Hearing, V.J., Hunt,D.F., Appella, E., 2003. Proteomic analysis of early melanosomes:identification of novel melanosomal proteins. J. Proteome Res. 2 (1), 69–79.

Bertwistle, D., Sugimoto, M., Sherr, C.J., 2004. Physical and functionalinteractions of the Arf tumor suppressor protein with nucleophosmin/B23.Mol. Cell. Biol. 24 (3), 985–996.

Borer, R.A., Lehner, C.F., Eppenberger, H.M., Nigg, E.A., 1989.Major nucleolarproteins shuttle between nucleus and cytoplasm. Cell 56 (3), 379–390.

Casper, J.M., Timpe, J.M., Dignam, J.D., Trempe, J.P., 2005. Identification of anadeno-associated virus Rep protein binding site in the adenovirus E2apromoter. J. Virol. 79 (1), 28–38.

Chang, J.H., Olson, M.O., 1989. A single gene codes for two forms of ratnucleolar protein B23 mRNA. J. Biol. Chem. 264 (20), 11732–11737.

Chang, J.H., Olson, M.O., 1990. Structure of the gene for rat nucleolar proteinB23. J. Biol. Chem. 265 (30), 18227–18233.

112 J.M. Bevington et al. / Virology 357 (2007) 102–113

Chejanovsky, N., Carter, B.J., 1989. Mutagenesis of an AUG codon in theadeno-associated virus rep gene: effects on viral DNA replication. Virology173, 120–128.

Chiorini, J.A., Weitzman, M.D., Owens, R.A., Urcelay, E., Safer, B., Kotin,R.M., 1994. Biologically active Rep proteins of adeno-associated virus type2 produced as fusion proteins in Escherichia coli. J. Virol. 68, 797–804.

Chiorini, J.A., Yang, L., Safer, B., Kotin, R.M., 1995. Determination of adeno-associated virus Rep68 and Rep78 binding sites by random sequenceoligonucleotide selection. J. Virol. 69 (11), 7334–7338.

Chiorini, J.A., Zimmermann, B., Yang, L., Smith, R.H., Ahearn, A., Herberg, F.,Kotin, R.M., 1998. Inhibition of PrKX, a novel protein kinase, and the cyclicAMP-dependent protein kinase PKA by the regulatory proteins of adeno-associated virus type 2. Mol. Cell. Biol. 18, 5921–5929.

Collaco, R.F., Kalman-Maltese, V., Smith, A.D., Dignam, J.D., Trempe, J.P.,2003. A biochemical characterization of the adeno-associated virus Rep40helicase. J. Biol. Chem. 278 (36), 34011–34017.

Colombo, E., Marine, J.C., Danovi, D., Falini, B., Pelicci, P.G., 2002.Nucleophosmin regulates the stability and transcriptional activity of p53.Nat. Cell Biol. 4 (7), 529–533.

Costello, E., Saudan, P., Winocour, E., Pizer, L., Beard, P., 1997. High mobilitygroup chromosomal protein 1 binds to the adeno-associated virus replicationprotein (Rep) and promotes Rep-mediated site-specific cleavage of DNA,ATPase activity and transcriptional repression. EMBO J. 16 (19),5943–5954.

Dignam, J.D., Lebovitz, R.M., Roeder, R.G., 1983. Accurate transcriptioninitiation by RNA polymerase II in a soluble extract from isolatedmammalian nuclei. Nucleic Acids Res. 11 (5), 1475–1489.

Di Pasquale, G., Chiorini, J.A., 2003. PKA/PrKX activity is a modulator ofAAV/adenovirus interaction. EMBO J. 22 (7), 1716–1724.

Dubielzig, R., King, J.A., Weger, S., Kern, A., Kleinschmidt, J.A., 1999. Adeno-associated virus type 2 protein interactions: formation of pre-encapsidationcomplexes. J. Virol. 73, 8989–8998.

Dumbar, T.S., Gentry, G.A., Olson, M.O., 1989. Interaction of nucleolarphosphoprotein B23 with nucleic acids. Biochemistry 28 (24), 9495–9501.

Fankhauser, C., Izaurralde, E., Adachi, Y., Wingfield, P., Laemmli, U.K., 1991.Specific complex of human immunodeficiency virus type 1 rev andnucleolar B23 proteins: dissociation by the Rev response element. Mol. Cell.Biol. 11 (5), 2567–2575.

Hermonat, P.L., Snatin, A.D., Batchu, R.B., 1996. The adeno-associated virusRep78 major regulatory/transformation suppressor protein binds cellularSp1 in vitro and evidence of a biological effect. Cancer Res. 56, 5299–5304.

Hermonat, P.L., Santin, A.D., Batchu, R.B., Zhan, D., 1998. The adeno-associated virus Rep78 major regulatory protein binds the cellular TATA-binding protein in vitro and in vivo. Virology 245, 120–127.

Herrera, J.E., Savkur, R., Olson, M.O., 1995. The ribonuclease activity ofnucleolar protein B23. Nucleic Acids Res. 23 (19), 3974–3979.

Herrera, J.E., Correia, J.J., Jones, A.E., Olson, M.O., 1996. Sedimentationanalyses of the salt- and divalent metal ion-induced oligomerization ofnucleolar protein B23. Biochemistry 35 (8), 2668–2673.

Huang, W.H., Yung, B.Y., Syu, W.J., Lee, Y.H., 2001. The nucleolarphosphoprotein B23 interacts with hepatitis delta antigens and modulatesthe hepatitis delta virus RNA replication. J. Biol. Chem. 276 (27),25166–25175.

Hunter, L.A., Samulski, R.J., 1992. Colocalization of adeno-associated virusRep and capsid proteins in the nuclei of infected cells. J. Virol. 66 (1),317–324.

Im, D.S., Muzyczka, N., 1990. The AAV origin binding protein Rep68 is anATP-dependent site-specific endonuclease with DNA helicase activity. Cell61 (3), 447–457.

Im, D.S., Muzyczka, N., 1992. Partial purification of adeno-associated virusRep78, Rep52, and Rep40 and their biochemical characterization. J. Virol.66 (2), 1119–1128.

Itahana, K., Bhat, K.P., Jin, A., Itahana, Y., Hawke, D., Kobayashi, R., Zhang,Y., 2003. Tumor suppressor ARF degrades B23, a nucleolar protein involvedin ribosome biogenesis and cell proliferation. Mol. Cell 12 (5), 1151–1164.

King, J.A., Dubielzig, R., Grimm, D., Kleinschmidt, J.A., 2001. DNA helicase-mediated packaging of adeno-associated virus type 2 genomes intopreformed capsids. EMBO J. 20 (12), 3282–3291.

Korgaonkar, C., Hagen, J., Tompkins, V., Frazier, A.A., Allamargot, C., Quelle,F.W., Quelle, D.E., 2005. Nucleophosmin (B23) targets ARF to nucleoli andinhibits its function. Mol. Cell. Biol. 25 (4), 1258–1271.

Kotin, R.M., Siniscalco, M., Samulski, R.J., Zhu, X.D., Hunter, L., Laughlin,C.A., McLaughlin, S., Muzyczka, N., Rocchi, M., Berns, K.I., 1990. Site-specific integration by adeno-associated virus. Proc. Natl. Acad. Sci. U.S.A.87 (6), 2211–2215.

Kurki, S., Peltonen, K., Laiho, M., 2004a. Nucleophosmin, HDM2 and p53:players in UV damage incited nucleolar stress response. Cell Cycle 3 (8),976–979.

Kurki, S., Peltonen, K., Latonen, L., Kiviharju, T.M., Ojala, P.M., Meek, D.,Laiho, M., 2004b. Nucleolar protein NPM interacts with HDM2 and protectstumor suppressor protein p53 from HDM2-mediated degradation. CancerCell 5 (5), 465–475.

Li, Y.P., Busch, R.K., Valdez, B.C., Busch, H., 1996. C23 interacts with B23, aputative nucleolar-localization-signal-binding protein. Eur. J. Biochem. 237(1), 153–158.

Li, Z., Brister, J.R., Im, D.S., Muzyczka, N., 2003. Characterization of theadenoassociated virus Rep protein complex formed on the viral origin ofDNA replication. Virology 313 (2), 364–376.

Lusby, E., Fife, K.H., Berns, K.I., 1980. Nucleotide sequence of the invertedterminal repetition in adeno-associated virus DNA. J. Virol. 34 (2),402–409.

Matsumoto, K., Okuwaki, M., Kawase, H., Handa, H., Hanaoka, F., Nagata, K.,1995. Stimulation of DNA transcription by the replication factor from theadenovirus genome in a chromatin-like structure. J. Biol. Chem. 270 (16),9645–9650.

Matthews, D.A., 2001. Adenovirus protein V induces redistribution of nucleolinand B23 from nucleolus to cytoplasm. J. Virol. 75 (2), 1031–1038.

Mendelson, E., Trempe, J.P., Carter, B.J., 1986. Identification of the trans-acting Rep proteins of adeno-associated virus by antibodies to a syntheticoligopeptide. J. Virol. 60 (3), 823–832.

Muzyczka, N., Berns, K.I., 2001. Parvoviridae: the viruses and their replication,In: Knipe, D.M., Howley, P.M. (Eds.), Fields Virology, 4th ed. LippincottWilliams and Wilkins, Philadelphia, pp. 2327–2359.

Needham, P.G., Casper, J.M., Kalman-Maltese, V., Verrill, K., Dignam, J.D.,Trempe, J.P., 2006. Adeno-associated virus Rep protein-mediated inhibitionof transcription of the adenovirus major late promoter in vitro. J. Virol. 80(13), 6207–6217.

Okuda, M., 2002. The role of nucleophosmin in centrosome duplication.Oncogene 21 (40), 6170–6174.

Okuda, M., Horn, H.F., Tarapore, P., Tokuyama, Y., Smulian, A.G., Chan, P.K.,Knudsen, E.S., Hofmann, I.A., Snyder, J.D., Bove, K.E., Fukasawa, K.,2000. Nucleophosmin/B23 is a target of CDK2/cyclin E in centrosomeduplication. Cell 103 (1), 127–140.

Okuwaki, M., Iwamatsu, A., Tsujimoto, M., Nagata, K., 2001a. Identification ofnucleophosmin/B23, an acidic nucleolar protein, as a stimulatory factor forin vitro replication of adenovirus DNA complexed with viral basic coreproteins. J. Mol. Biol. 311 (1), 41–55.

Okuwaki, M., Matsumoto, K., Tsujimoto, M., Nagata, K., 2001b. Function ofnucleophosmin/B23, a nucleolar acidic protein, as a histone chaperone.FEBS Lett. 506 (3), 272–276.

Okuwaki, M., Tsujimoto, M., Nagata, K., 2002. The RNA binding activity of aribosome biogenesis factor, nucleophosmin/B23, is modulated by phos-phorylation with a cell cycle-dependent kinase and by association with itssubtype. Mol. Biol. Cell 13 (6), 2016–2030.

Pegoraro, G., Marcello, A., Myers, M.P., Giacca, M., 2006. Regulation ofadeno-associated virus DNA replication by the cellular TAF/Set complex.J. Virol. 80 (14), 6855–6864.

Pereira, D.J., Muzyczka, N., 1997. The cellular transcription factor SP1 and anunknown cellular protein are required to mediate Rep protein activation ofthe adeno-associated virus p19 promoter. J. Virol. 71, 1747–1756.

Prasad, K.M.R., Trempe, J.P., 1995. The adeno-associated virus Rep78 proteinis covalently linked to viral DNA in a pre-formed virion. Virology 214(360–370).

Qiu, J., Brown, K.E., 1999. A 110-kDa nuclear shuttle protein, nucleolin,specifically binds to adeno-associated virus type 2 (AAV-2) capsid. Virology257 (2), 373–382.

113J.M. Bevington et al. / Virology 357 (2007) 102–113

Redner, R.L., 2002. Variations on a theme: the alternate translocations in APL.Leukemia 16 (10), 1927–1932.

Samulski, R.J., Chen, L.S., Shenk, T., 1989. Helper-free stocks of recombinantadeno-associated viruses: normal integration does not require viral geneexpression. J. Virol. 63, 3822–3828.

Santa-Coloma, T.A., 2003. Anp32e (Cpd1) and related protein phosphatase 2inhibitors. Cerebellum 2 (4), 310–320.

Savkur, R.S., Olson, M.O., 1998. Preferential cleavage in pre-ribosomal RNAbyprotein B23 endoribonuclease. Nucleic Acids Res. 26 (19), 4508–4515.

Smith, R.H., Kotin, R.M., 1998. The Rep52 gene product of adeno-associatedvirus is a DNA helicase with 3′-to-5′ polarity. J. Virol. 72 (6), 4874–4881.

Smith, R.H., Kotin, R.M., 2000. An adeno-associated virus (AAV) initiatorprotein, Rep78, catalyzes the cleavage and ligation of single-stranded AAVori DNA. J. Virol. 74 (7), 3122–3129.

Srivastava, M., Pollard, H.B., 1999. Molecular dissection of nucleolin's role ingrowth and cell proliferation: new insights. FASEB J. 13 (14), 1911–1922.

Srivastava, A., Lusby, E.W., Berns, K.I., 1983. Nucleotide sequence andorganization of the adeno-associated virus 2 genome. J. Virol. 45 (2),555–564.

Szebeni, A., Olson, M.O., 1999. Nucleolar protein B23 has molecular chaperoneactivities. Protein Sci. 8 (4), 905–912.

Szebeni, A., Herrera, J.E., Olson, M.O., 1995. Interaction of nucleolar proteinB23 with peptides related to nuclear localization signals. Biochemistry 34(25), 8037–8042.

Szebeni, A., Mehrotra, B., Baumann, A., Adam, S.A., Wingfield, P.T., Olson,M.O., 1997. Nucleolar protein B23 stimulates nuclear import of the HIV-1Rev protein and NLS-conjugated albumin. Biochemistry 36 (13),3941–3949.

Szebeni, A., Hingorani, K., Negi, S., Olson, M.O., 2003. Role of protein kinaseCK2 phosphorylation in the molecular chaperone activity of nucleolarprotein b23. J. Biol. Chem. 278 (11), 9107–9115.

Takemura, M., Sato, K., Nishio, M., Akiyama, T., Umekawa, H., Yoshida, S.,1999. Nucleolar protein B23.1 binds to retinoblastoma protein andsynergistically stimulates DNA polymerase alpha activity. J. Biochem.(Tokyo) 125 (5), 904–909.

Timpe, J., Bevington, J., Casper, J., Dignam, J.D., Trempe, J.P., 2005.Mechanisms of adeno-associated virus genome encapsidation. Curr. GeneTher. 5 (3), 273–284.

Trempe, J.P., Mendelson, E., Carter, B.J., 1987. Characterization of adeno-associated virus Rep proteins in human cells by antibodies raised againstRep expressed in Escherichia coli. Virology 161, 18–28.

Umekawa, H., Sato, K., Takemura, M., Watanabe, Y., Usui, S., Takahashi, T.,Yoshida, S., Olson, M.O., Furuichi, Y., 2001. The carboxyl terminalsequence of nucleolar protein B23.1 is important in its DNA polymerasealpha-stimulatory activity. J. Biochem. (Tokyo) 130 (2), 199–205.

von Lindern,M., vanBaal, S.,Wiegant, J., Raap, A., Hagemeijer, A., Grosveld, G.,

1992. Can, a putative oncogene associated with myeloid leukemogenesis,may be activated by fusion of its 3′ half to different genes: characterization ofthe set gene. Mol. Cell. Biol. 12 (8), 3346–3355.

Walton, T.H., Moen Jr., P.T., Fox, E., Bodnar, J.W., 1989. Interactions of minutevirus of mice and adenovirus with host nucleoli. J. Virol. 63 (9), 3651–3660.

Wang, D., Umekawa, H., Olson, M.O., 1993. Expression and subcellularlocations of two forms of nucleolar protein B23 in rat tissues and cells. Cell.Mol. Biol. Res. 39 (1), 33–42.

Wang, D., Baumann, A., Szebeni, A., Olson, M.O., 1994. The nucleic acidbinding activity of nucleolar protein B23.1 resides in its carboxyl-terminalend. J. Biol. Chem. 269 (49), 30994–30998.

Weber, J.D., Jeffers, J.R., Rehg, J.E., Randle, D.H., Lozano, G., Roussel, M.F.,Sherr, C.J., Zambetti, G.P., 2000. p53-independent functions of the p19(ARF) tumor suppressor. Genes Dev. 14 (18), 2358–2365.

Weger, S., Wistuba, A., Grimm, D., Kleinschmidt, J.A., 1997. Control of adeno-associated virus type 2 cap gene expression: relative influence of helpervirus, terminal repeats, and Rep proteins. J. Virol. 71 (11), 8437–8447.

Weger, S., Wendland, M., Kleinschmidt, J.A., Heilbronn, R., 1999. The adeno-associated virus type 2 regulatory proteins rep78 and rep68 interact with thetranscriptional coactivator PC4. J. Virol. 73 (1), 260–269.

Weger, S., Hammer, E., Heilbronn, R., 2002. Topors, a p53 and topoisomerase Ibinding protein, interacts with the adeno-associated virus (AAV-2) Rep78/68proteins and enhances AAV-2 gene expression. J. Gen. Virol. 83 (Pt 3),511–516.

Weitzman, M.D., Fisher, K.J., Wilson, J.M., 1996. Recruitment of wild-typeand recombinant adeno-associated virus into adenovirus replication centers.J. Virol. 70, 1845–1854.

Winters, W.D., Russell, W.C., 1971. Studies on the assembly of adenovirus invitro. J. Gen. Virol. 10, 181–194.

Wistuba, A., Kern, A., Weger, S., Grimm, D., Kleinschmidt, J.A., 1997.Subcellular compartmentalization of adeno-associated virus type 2 assem-bly. J. Virol. 71 (2), 1341–1352.

Wu, M.H., Chang, J.H., Yung, B.Y., 2002. Resistance to UV-induced cell-killingin nucleophosmin/B23 over-expressed NIH 3T3 fibroblasts: enhancement ofDNA repair and up-regulation of PCNA in association with nucleophosmin/B23 over-expression. Carcinogenesis 23 (1), 93–100.

Yoneda-Kato, N., Look, A.T., Kirstein, M.N., Valentine, M.B., Raimondi, S.C.,Cohen, K.J., Carroll, A.J., Morris, S.W., 1996. The t(3;5)(q25.1;q34) ofmyelodysplastic syndrome and acute myeloid leukemia produces a novelfusion gene, NPM-MLF1. Oncogene 12 (2), 265–275.

Young Jr., S.M., McCarty, D.M., Degtyareva, N., Samulski, R.J., 2000. Roles ofadeno-associated virus Rep protein and human chromosome 19 in site-specific recombination. J. Virol. 74 (9), 3953–3966.

Yung, B.Y., Busch, H., Chan, P.K., 1985. Translocation of nucleolarphosphoprotein B23 (37 kDa/pI 5.1) induced by selective inhibitors ofribosome synthesis. Biochim. Biophys. Acta 826 (4), 167–173.