Journal of Structural Biology 173 (2011) 213–218

Contents lists available at ScienceDirect

Journal of Structural Biology

journal homepage: www.elsevier .com/locate /y jsbi

Replication timing of pseudo-NORs

Evgeny Smirnov a,b,⇑, Dušan Cmarko a,b, Lubomír Kovácik a,b, Guy M. Hagen a,b, Alexey Popov a,b,Otakar Raška c, José-Luis Prieto d, Boris Ryabchenko e, Filipa Amim a,b, Brian McStay d, Ivan Raška a,b

a Institute of Physiology, Academy of Sciences of the Czech Republic v.v.i., Department of Cell Biology, Albertov 4, 128 01 Prague, Czech Republicb Charles University in Prague, First Faculty of Medicine, Institute of Cellular Biology and Pathology, Albertov 4, 12801 Prague 2, Czech Republicc Charles University in Prague, Third Faculty of Medicine, Department of Normal, Pathological and Clinical Physiology, Ke Karlovu 2, 12000 Prague 2, Czech Republicd Department of Biochemistry National University of Ireland, Galway, University Road, Galway, Irelande Charles University in Prague, Faculty of Science, Laboratory of Molecular Virology, Vinicna 5, 12844 Prague 2, Czech Republic

a r t i c l e i n f o a b s t r a c t

Article history:Received 18 March 2010Received in revised form 23 November 2010Accepted 29 November 2010Available online 3 December 2010

Keywords:UBFPseudo-NORsrDNAReplicationChromatin structure

1047-8477/$ - see front matter � 2010 Elsevier Inc. Adoi:10.1016/j.jsb.2010.11.023

⇑ Corresponding author at: Charles UniversityMedicine, Institute of Cellular Biology and PathologyCzech Republic. Fax: +420 224917418.

E-mail address: esmir@lf1.cuni.cz (E. Smirnov).

In mammalian cells, transcriptionally active ribosomal genes are replicated in the early S phase, and thesilent ribosomal genes in the late S phase, though mechanisms of this timing remain unknown. UBF(Upstream Binding Factor), a DNA binding protein and component of the pol I transcription machinery,is considered to be responsible for the loose chromatin structure of the active rDNA. Here we questionwhether such structure alone can ensure early replication of DNA. We investigate this problem on themodel of pseudo-NORs, the tandem arrays of heterologous DNA sequence with high affinity for UBF,introduced into human chromosomes. Such arrays are not transcribed, yet efficiently bind UBF and mimicthe chromatin structure of active rDNA. In our study, a human derived stable cell line containing onepseudo-NOR on the chromosome 10 was transiently transfected with UBF-GFP and PCNA-RFP, whichallowed us to observe in vivo the growth of pseudo-NORs resulted from their replication. We found thatreplication of pseudo-NORs is not restricted to the early S phase, but continues in the late S phase at asignificant level. These results were confirmed in the experiments with incorporation of thymidin analogEdU and BrdU ChIP assay. Similar results were obtained with another cell line containing pseudo-NOR onthe chromosome 7. Our data indicate that the specific loose structure of chromatin, produced by thearchitect protein UBF, is not sufficient for the early replication.

� 2010 Elsevier Inc. All rights reserved.

1. Introduction

In eukaryotic cells, replication timing of various genes corre-lates with their transcriptional activity, chromatin structure, nucle-ar position, and is regarded presently as an important epigeneticmark (reviewed in Gilbert, 2002; Berezney, 2002; Lucas and Feng,2003; Hiratani and Gilbert, 2009; Méndez, 2009; Ryba et al., 2010;Pope et al., 2010). Ribosomal genes coding 5.8S, 18S, and 28S rRNArepresent special case. These genes exist in numerous copies, someof which are characterized by the ‘closed’ structure of chromatinand permanent transcriptional silence (Chen and Pikaard, 1997;Raška et al., 2006; Santoro, 2005; Birch and Zomerdijk, 2008; Sanijand Hannan, 2009). It has been found that in mammalian cellstranscriptionally active ribosomal genes are replicated predomi-nantly in the early S phase, and the silent ribosomal genes in thelate S phase (Berger et al., 1997; Li et al., 2005). However, it re-mains unclear what determines such timing. One possibility is that

ll rights reserved.

in Prague, First Faculty of, Albertov 4, 12801 Prague 2,

the structure of decondensed chromatin characteristic for activeribosomal genes creates favorable conditions for their early repli-cation (Berger et al., 1997; Li et al., 2005).

UBF (Upstream Binding Factor), a DNA binding protein andcomponent of the pol I transcription machinery (Roussel et al.,1996; Hannan et al., 1998; Sirri et al., 2000; Stefanovsky et al.,2001; Russell and Zomerdijk, 2006; Smirnov et al., 2006), isbelieved to be the major architect of this particular structure(O’Sullivan et al., 2002; Mais et al., 2005; Prieto and McStay,2007; Wright et al., 2006). We question whether UBF binding alonecan determine early replication of DNA, independent of its se-quence and transcription status. We investigate this problem onthe model of pseudo-NORs (reviewed in Prieto and McStay,2008). The pseudo-NORs are tandem arrays of heterologous non-transcribed DNA sequence with high affinity for UBF introducedinto human chromosomes. Such arrays are not transcribed, yetefficiently recruit UBF to sites outside the nucleolus and, duringmetaphase, form novel silver positive secondary constrictions.Furthermore, when UBF binds to DNA it recruits pol I and othercomponents of the rDNA transcription machinery, as well as somefactors of rDNA processing (Mais et al., 2005; Prieto and McStay,2007). Thus pseudo-NORs in several aspects mimic the structure

214 E. Smirnov et al. / Journal of Structural Biology 173 (2011) 213–218

of the natural clusters of ribosomal genes, or Nucleolus OrganizerRegions (NORs).

In the present study, we determined replication timing of pseu-do-NORs in two human derived cell lines 3D-1 and 5E-2 with dif-ferent location of XEn arrays (on the chromosome 10 and 7,respectively) (Mais et al., 2005). We found that replication of thesearrays begins in the early, but continues in the late S phase. Ourdata indicate that the specific loose chromatin structure, producedby UBF binding, does not ensure early replication.

2. Material and methods

2.1. Cell lines and culture

The pseudo-NOR containing cell line 3D-1 is derived from thehuman fibrosarcoma cell line HT1080 and contains a 1.5-Mb arrayof XEn sequences, derived from the intergenic spacer of Xenopuslaevis rDNA, on the long arm of chromosome 10 (Prieto andMcStay, 2008). We used also another similar cell line 5E-2 in whichthe same XEn sequences are located on the long arm of chromo-some 7 (Mais et al., 2005). The cells were cultivated in Dulbecco’sMEM (+Glutamax, sodium pyruvate, and 4.5 g/L glucose; GIBCO)supplemented with 10% fetal bovine serum and 0.1% gentamycin,in atmosphere supplemented with 5% CO2.

2.2. Plasmids and XEn FISH probe

Plasmids RFP-PCNA and GFP-UBF were kindly provided byDr. T. Misteli (National Cancer Institute, NIH, Bethesda, MD). Theplasmids were transfected into 3D-1 cells using FuGENE (Rochediagnostics).

The plasmid pXEn8 (Mais et al., 2005) contains a tandem arrayof eight copies of the 625-bp X. laevis enhancer cloned as aSalI–XhoI fragment in a modified pGEM3 vector (Promega).

Applying CLB-Transfection™ Device (Lonza), we introduced theplasmid in E. coli cells for amplification. The biotin-labeled XEnprobe was prepared using nick-translation kit BIONICK LabelingSystem (GIBCO-BRL, Invitrogen). The probe was stored in hybrid-ization mixture containing 25 ng of probe, 0.5 mg/ml sonicatedsalmon sperm DNA, 50% deionized formamide, 2� SSC and 10%dextran sulfate.

2.3. BrdU ChIP

In this assay 3D-1 cells were synchronized by double block with1 lg/ml aphidicolin (Fluka). Two and five hours after release fromthe block, the cells were pulse labeled with 30 lM 50-BrdU (Sigma)for 30 min. Control cells were not labelled. Genomic DNA from thelabelled and control cells was isolated using AllPrepDNA/RNAPro-tein Mini Kit (Qiagen) and sonicated to �800 bp fragments. 2 lgof heat-denatured DNA from each sample was incubated with1.5 lg anti-BrdU antibody in PBS (50 lL) for 30 min at room tem-perature. 10 lg of anti-mouse IgGs in PBS (20 lL) was added andafter a further 30 min incubation the mix was centrifuged at11,000g for 10 min. Precipitates were washed twice with 100 lLPBS at room temperature. The final pellets were dissolved by boil-ing in TE (25 lL). Real Time PCR reactions were performed in dupli-cate with 2.5 lL of each sample using a DYNAmo HS SYBR GreenqPCR mastermix (New England Biolabs) and an MJ research Opti-con 2 Thermocycler. The primer pairs employed are as follows:

j rDNA promoterj Forward Primer: 50 GTGTGTCCTGGGGTTGACC 30

j Reverse Primer:50 GCAGGCGGCTCAAGCAGGAG 30

j U1 snRNA gene

j Forward Primer: 50 TTACCTGGCAGGGGAGATAC 30

j Reverse Primer: 50 GCAGTCGAGTTTCCCACATT 30

j Xen sequencesj Forward Primer: 50 GACCGGGAGTTCCAGGAG 30

j Reverse Primer:50 CAGGGCAGGGGGACGAG 30

2.4. UBF and pol I immunocytochemistry (IC)

Cells were rinsed in PBS and fixed in 2% PFA (formaldehydefreshly prepared from from paraformaldehyde) for 10 min at RT,and permeabilized with 0.2% Triton X-100. Primary antibodiesagainst human UBF and pol I were kindly provided by Dr. U. Scheer(Biocenter of the University of Wurzburg). We also used monoclo-nal (mouse) anti-UBF antibody (Santa Cruz Biotechnology, Inc.),which binds human UBF. Secondary anti-human and anti-mouseantibodies were labeled with Cy3 or FITC (Jackson ImmunoRe-search Laboratories). Coverslips were mounted in Mowiol.

2.5. EdU labeling of replication combined with UBFimmunofluorescence

For the labeling of replication, we used the thymidine analogueEdU (5-ethynyl-20-deoxyuridine) provided by Invitrogen. Incontrast to BrdU, detection of EdU requires no DNA denaturation,thus allowing better preservation of the nuclear structure. EdUwas administered to the intact living cells in concentration 10 lMfor 10 min. The cells were fixed in PFA and processed for UBF ICwith FITC-tagged secondary antibody (Jackson ImmunoResearchLaboratories). After rinsing in PBS, the replication signal was visu-alized using EdU Alexa Fluor� 647 Imaging Kit (Invitrogen). We alsolabeled replication sites with 20 lM BrdU (Sigma, Aldrich). Thesignal was visualized by using mouse anti-BrdU antibody (Roche).

2.6. Combined UBF immunofluorescence and XEn FISH

After UBF immunolabeling the cells were postfixed with meth-anol/acetic acid (3:1) overnight at �20 �C, then the regular FISHfollowed (Pliss et al., 2005). The UBF signal was well preserveddue to this procedure. For denaturation, cells were placed in 70%formamide/2� SSC for 3 min at 73 �C followed by 1 min in 50%formamide/2� SSC at 73 �C. Hybridization proceeded in wet cham-ber with 50% formamide at 37� overnight. Post hybridization wash-ing was performed after Harnicarová et al. (2006). Namely, thecells were washed in 50% formamide in 2� SSC, pH 7, for 15 minat 43 �C, in 0.1% Tween-20/2� SSC for 8 min at 43 �C; in 0.1% Igepal(ICN Biomedicals, Inc)/4� SSC for 3 � 4 min at 37 �C, in PBS3 � 3 min at RT. XEn FISH probe was visualized using rabbit anti-biotin antibody (Bethyl).

2.7. Image analysis

For assessing colocalization between UBF and EdU signals, con-focal image stacks were recorded under conditions of optimal sam-pling (Heintzmann, 2006), then deconvolved using Huygensprofessional software (Scientific Volume Imaging) with a calcu-lated point spread function (PSF) and the classical constrainedmaximum likelihood estimation (CMLE) algorithm. We also exam-ined our samples using a newly constructed widefield microscopysystem consisting of an inverted microscope (IX71 with a 100�,1.35NA PlanApo objective, Olympus, Hamburg, Germany), EMCCDcamera (Ixon DU885, Andor), Piezo Z stage (Nanoscan Z, Prior Sci-entific), and appropriate filters sets for FITC and Cy5. The acquisi-tion system is controlled by IQ software (Andor). The microscopeis supported on a large, actively isolated optical table (Thor Labs)

E. Smirnov et al. / Journal of Structural Biology 173 (2011) 213–218 215

to prevent vibrations. We then deconvolved image stacks acquiredwith this system using Huygens software as above.

2.8. 4D imaging of pseudo-NORs

4D imaging was performed on confocal microscope SP-5 (Leica)with objective HCX PL APO lambda blue 63�, 1.4 NA. A 3D imagestack of 25–30 optical sections, spanning the entire cell volume,was recorded at each time point. Low light imaging conditions, res-onance scanner regime, and the period of 1 h between the sequen-tial stacks were chosen, which allowed us to follow the changes inpseudo-NORs without impairing the ability of the cell to progressthrough the cell cycle, and without bleaching of the signal.

Fig.2. Monotonous enlargement of pseudo-NORs in the course of S phase. A, B:Changes of pseudo-NORs observed in two 3D-1 cells transfected with GFP-UBF (A)and RFP-PCNA (B). The time is counted from the point when both cells begin toexhibit the pattern of PCNA typical for the late S phase (here the observation started3.5 h prior to this point). Pseudo-NORs continued to grow after the transition fromearly to late S phase. C: Changes in the intensity of the UBF signal in pseudo-NORs ofthree selected 3D-1 cells. The dashed vertical line indicates transition to the late Sphase. Scale bar: 4 lm.

3. Results

3.1. Intensities of UBF IC and XEn FISH signals strongly correlate

We used XEn DNA probe (see Section 2) to label pseudo-NORs in3D-1 and 5E-2 cells. On the preparations of fixed non-synchronizedcells, we performed UBF IC combined with XEn FISH. Most cells con-tained one pseudo-NOR situated most frequently in the midst ofnucleoplasm, but sometimes also close to the lamina or imbedded,more or less deeply, into one of the nucleoli. The IC and FISH signalson the pseudo-NORs were represented in all stages of interphase byone single dot per cell, and well colocalized with each other(Fig. 1A–C, Suppl. I). Additionally, we measured integral intensitiesof both signals on confocal sections, using ImageJ software. In theinterphase cells, we found strong correlation, with correlation coef-ficient r = 0.98 for 3D-1 cells (Fig. 1D), and r = 0.89 for 5E-2 cells,between intensity of XEn FISH and UBF antibody signal in the inter-phase pseudo-NORs. Since the intensity of FISH fluorescence ontandem arrays of DNA strongly correlates with the number ofrepeats in these arrays (Leitch et al., 1992; Mellink et al., 1994;Suzuki et al., 1996), our data show that UBF immunofluorescencesignal can be used for assessing the changes in the DNA contentsof pseudo-NORs in the course of their replication.

3.2. In 3D-1 cells transfected with UBF-GFP and PCNA-RFP, the size ofpseudo-NORs continues to increase in the late S phase

To observe in vivo the dynamics of pseudo-NORs, namely in-crease of their size during S-phase, we transiently transfected3D-1 cells with UBF-GFP and PCNA-RFP constructs. PCNA was pre-viously shown to be a good marker for topographic patterns ofDNA synthesis (Raška et al., 1989, 1991). The transfected cells con-tinued to divide and expressed both signals for more than one cellcycle. In such cells RFP-PCNA signal well colocalized with replica-tion signal obtained after 10 min pulse incorporation of EdU (see

Fig.1. Pseudo-NORs in two 3D-1 cells. UBF IC (A), XEn FISH (B) and merge (C). Correlation between the intensity of UBF and XEn FISH signals (D), measured in 50 cells (thesize of the pseudo-NORs varies in different cells so much, that the minimum and maximum values on the graph differ 3–4 times, though in each individual cell the signalintensities in earliest and latest stages differ no more than 1:2; as seen in Fig. 2). Scale bar: 4 lm.

216 E. Smirnov et al. / Journal of Structural Biology 173 (2011) 213–218

Section 2), which allowed us to observe directly the transition fromearly to late S phase (Fig. 2A, B). We selected for our observationsindividual cells with the pattern of PCNA signal corresponding toearly S phase. Such cells were then watched for 6–8 h, or untilentering G2 and mitosis. This time interval was chosen becausethe average length of S phase in the 3D-1 cell line, assessed afterpulse – chase incorporation of BrdU and EdU, was about 7 h (datanot shown). Confocal 3D images were obtained every hour. In allstudied cells, the intensity of UBF-GFP signal gradually increasedbefore transition from early to late S phase (with regression coeffi-cient 5.5 ± 2.0), and also after the transition (with regression coef-ficient 2.3 ± 1.4) (Fig. 2C, Table 1). These data show that replicationof all observed pseudo-NORs began in the early S phase and contin-ued after transition to the late S phase, though at a slower rate.

3.3. Intensity of the UBF signal in pseudo-NORs, measured at differentstages of S-phase, shows a gradual increase through the most of Sphase

Alternatively, we measured integral intensity of the UBF signalin pseudo-NORs in the fixed non-synchronized 3D-1 and 5E-2 cellsafter 10 min pulse incorporation of EdU. We arranged the data ingroups according to four sequential stages of S phase, which canbe distinguished by the typical pattern of the replication signal,as described in the legend to Fig. 3A. The first of these stages isthe longest, the last is the shortest. In three independent experi-ments, we found that the UBF signal in pseudo-NORs increasedin the course of S phase (Fig. 3B). The increase of the integral inten-sity value was significant between the stages I and II (by 38% inaverage), as well as between the stages II and III (by 21% in aver-age). Only from the stage III to the stage IV the UBF signal changedinsignificantly (by 5% in average), apparently because replication ofpseudo-NORs is mainly finished between these stages. The inten-sity of the XEn FISH signal in the course of the S phase changedin a similar manner (Fig. 3C). These data confirm our observationsin the live cells and show that replication of pseudo-NORs contin-ues through the most of S phase.

Fig.3. Size of pseudo-NORs at four sequential stages of S phase. A: stages of S phase distiphase; replication foci are fine, numerous and evenly distributed in the nucleoplasm. Stalarge foci concentrated along the nuclear lamina and around the nucleoli. Stage III represebetween them, there are also pronounced fringes of coarse replication foci along the lamlarge, sparsely arranged, distributed without any particular pattern, and the nucleus appethe UBF signal in pseudo-NORs at different stages of S phase. Data of five independentintensity of XEn FISH signal in pseudo-NORs at different stages of S phase.

3.4. Replication signal within the area of pseudo-NORs is detectable inall stages of S phase, except the very end of it

We also observed colocalization of UBF and EdU signals in pseu-do-NORs. After deconvolution of the images, the EdU signal wasdetected inside the pseudo-NORs in both early and late S phase(Fig. 4A, B, Suppl. II). These data, obtained on 3D-1 and 5E-2 cells,together with the results of in vivo observations (Fig. 2), indicatethat replication of each pseudo-NOR proceeds from the beginningof S phase to its late stage. Only in rare cases, the signal of replica-tion appeared as separate focus entirely enclosed within pseudo-NOR, but usually this signal was a continuation or extension of alarger adjacent (nucleoplasmic) focus (Fig 4B). It should be also no-ticed, that the EdU signal in pseudo-NOR was less intensive thanmany replication foci in the nucleoplasm, apparently reflectingthe loose chromatin structure of pseudo-NORs. Indeed, such lowintensity of the replication signal is typical for the UBF associatedchromatin in the nucleoli. In the cells expressing the pattern ofvery late replication (stage IV in Fig. 3A) the EdU signal was usuallyundetectable (Fig 4C), which agrees with the small increase in theintensity of UBF signal between the stages III and IV (Fig. 3B).

3.5. XEn sequences incorporate BrdU both in early and late S phase

In order to support our morphological findings by an alternativeapproach, we performed BrdU ChIP assay on synchronized 3D-1cells. About 95% cells were in the early or late S phase, respectively,2 and 5 h after release from the aphidicolin block (data not shown).After pulse labeling with BrdU for 30 min, DNA from the labeledand control cells was isolated and processed for the ChIP. We foundthat XEn sequences incorporated BrdU in both early and late Sphase with comparable intensities (Fig. 5). No signal was observedwith IPs from the control DNA. Furthermore, using the U1 snRNAgene as a convenient marker for early replicating DNA, weobserved more than tenfold decrease of the real time PCR valuebetween the early and late S phase. Our IP results on rDNA agreewith the published data about replication of the ribosomal genesin both early and late S phase (Berger et al., 1997; Li et al., 2005),

nguished by the pattern of EdU or BrdU incorporation. Stage I corresponds to early Sge II corresponds to middle S phase; most replication foci are fine, but there are alsonts typical pattern of late S phase; replication foci are mostly large, with wide spacesina and around the nucleoli. Stage IV is the very end of S phase; replication foci arears mostly dark. Stage I is the longest, stage IV – the shortest. B: Average intensity ofexperiments with 3D-1 cells (curves 1–3), and 5E-2 cells (curves 4–5). C: Average

Fig.4. Replication signal in pseudo-NORs at different stages of S phase. Incorporation of EdU, red; UBF, green and arrowheads in the inserts. Intensity profiles (IP): profiles ofintegral intensity within the broad bands indicated in the ‘‘merge’’ panel by the dashed lines. A: Early S phase (stage I). B: Late S phase (stage III). C: Very late S phase (stageIV). There are distinct replication signals in both early and late S phase (arrows), only in the very late S phase the signal is not detectable. Scale bar: 1.5 lm. (For interpretationof the references to colour in this figure legend, the reader is referred to the web version of this article.)

E. Smirnov et al. / Journal of Structural Biology 173 (2011) 213–218 217

and thus provide additional control for the replication timing ofpseudo-NORs.

4. Discussion

Our approach based on the correlation between the intensity ofXEn FISH and UBF signals, does not allow us to quantify replicationprecisely. Nevertheless, we found that in 3D-1 cells transfectedwith UBF-GFP and PCNA-RFP, the size of pseudo-NORs increasedin the late S phase as well as in the early S phase (Fig. 2). Moreover,

Fig.5. XEn sequences incorporate BrdU throughout S phase. Pseudo-NOR contain-ing cells, (clone 3D-1), were synchronized, released and then pulse labeled withBrdU during early or late S phase. DNA recovered from anti-BrdU IPs was used inReal Time PCR with XEn, rDNA and U1 primer pairs. An anti-BrdU IP of genomicDNA from unlabeled 3D-1 cells was used as a control. The Y-axis represents therelative proportions of target DNA in each sample.

in non-transfected 3D-1 and 5E-2 cells, intensity of the UBF signalin pseudo-NORs, measured after the incorporation of EdU, wasgrowing in both early and late S phase (Fig. 3). Performing BrdUChIP assay, we found a considerable incorporation of BrdU in XEnsequences in the late S phase (Fig. 5). Thus, the results obtainedby three alternative methods show that replication of pseudo-NORs is not restricted to the early S phase, but continues in the lateS phase at a significant level. Detection of the replication signalwithin the area occupied by pseudo-NORs (Fig. 4) further confirmsthese results.

Pseudo-NOR arrays used in the present study have the totallength of about 1.5 Mb, while average human NOR is roughly twiceas large (Prieto and McStay, 2008). Still these arrays are apparentlytoo large to be fully replicated from the external origins. The dataof Hyrien et al. (1995) also suggest, though indirectly, presence ofthe internal replication origins within pseudo-NORs. In this re-spect, pseudo-NORs resemble true NORs, which are replicated atleast partially from their endogenous origins (Little et al., 1993;Yoon et al., 1995). Remarkably, replication of all transcriptionallyactive ribosomal genes in the cell, which constitute roughly onehalf of the entire rDNA, i.e. about 15 Mb in the human cells (Sant-oro and Grummt, 2002; Grummt and Pikaard, 2003; Moss et al.,2007; Sanij and Hannan, 2009), is restricted to the early S phase,whereas transcriptionally silent genes are replicated in the late Sphase (Berger et al., 1997; Li et al., 2005). But according to our data(Fig. 2), replication of each individual pseudo-NOR, involving abouttenfold less DNA than the active ribosomal genes, requires a longerperiod that begins at the onset of S phase and ends at its final stagewhen most part of the genome is already duplicated.

The mechanisms ensuring the early replication, which is typicalfor transcriptionally active DNA, remain unknown. Our dataindicate that the structure of chromatin produced by UBF, and imi-tating the structure of nucleolar chromatin (Prieto and McStay,

Table 1Regression coefficients illustrate dynamics of the integral intensities of pseudo-NORs in the course of the S phase. The data from 10 living 3D-1 cells transfected with GFP-UBF andRFP- PCNA, as in Fig. 2. Mean values and standard deviations are shown in the last two columns. The positive value of the regression indicates that the intensity of UBF-GFP signalgradually increased both in early and late S phase.

S-phase Cell number Mean std.

1 2 3 4 5 6 7 8 9 10

Early 2.9 1.7 5.3 4.3 6.5 3.9 6.0 8.6 5.1 10.8 5.5 2.0Late 0.9 1.2 1.3 1.4 1.0 1.5 2.1 4.8 4.0 4.6 2.3 1.4

218 E. Smirnov et al. / Journal of Structural Biology 173 (2011) 213–218

2008), is not sufficient for mobilization of those mechanisms. Onthe other hand, our results (Fig. 3, Table 1) suggest that replicationof pseudo-NORs reaches a high level of intensity in the early Sphase, when the numerous replication foci are evenly distributedin the nucleoplasm, and is afterwards decreased towards the verylate S phase. Moreover, we rarely found large replication domains,such as are scattered over the nucleus in the late S phase, belong-ing exclusively to pseudo-NORs. In these areas, the replication sig-nals (Fig. 4) were usually represented by parts and extensions ofthe large replication foci situated in the adjacent nucleoplasm. Thisargues for the absence of special mechanisms regulating the timingof replication within the decondensed chromatin area of pseudo-NOR. Such mechanisms would, probably, require more sophisti-cated genetic structure than the array of short XEn repeats.

The results obtained in our study indicate that the specificstructure of loose chromatin, produced by the architect proteinUBF, cannot guarantee early replication; other factors, such asthe ongoing transcription and presence of certain DNA sequences,may be necessary for that.

Acknowledgments

This work was supported by the grants from The Ministry ofEducation, Youth and Sports of the Czech RepublicMSM0021620806 and LC535, from The Academy of Sciences ofthe Czech Republic AV0Z50110509, from The Grant Agency of theCzech Republic: 304/09/1047, and the Science Foundation IrelandPrincipal Investigator award 07/IN.1/B924.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.jsb.2010.11.023.

References

Berezney, R., 2002. Regulating the mammalian genome: the role of nucleararchitecture. Adv. Enzyme Regul. 42, 39–52.

Berger, C., Horlebein, A., Gögel, E., Grummt, F., 1997. Temporal order of replicationof mouse ribosomal RNA genes during the cell cycle. Chromosoma 106,479–484.

Birch, J.L., Zomerdijk, J.C., 2008. Structure and function of ribosomal RNA genechromatin. Biochem. Soc. Trans. 36, 619–624.

Chen, Z.J., Pikaard, C.S., 1997. Epigenetic silencing of RNA polymerase Itranscription: a role for DNA methylation and histone modification innucleolar dominance. Genes Dev. 11, 2124–2136.

Gilbert, D.M., 2002. Replication timing and transcriptional control: beyond causeand effect. Curr Opin Cell Biol. 14, 377–383.

Grummt, I., Pikaard, C.S., 2003. Epigenetic silencing of RNA polymerase Itranscription. Nat. Rev. Mol. Cell Biol. 4, 641–649.

Hannan, K.M., Hannan, R.D., Rothblum, L.I., 1998. Transcription by RNA polymeraseI. Front. Biosci. 3, d376–398.

Harnicarová, A., Kozubek, S., Pacherník, J., Krejcí, J., Bártová, E., 2006. Distinctnuclear arrangement of active and inactive c-myc genes in control anddifferentiated colon carcinoma cells. Exp. Cell Res. 312, 4019–4035.

Heintzmann, R., 2006. Band limit and appropriate sampling in microscopy. In:Simons, K., Small, J.V., Hunter, T., Shotton, D., Celis, J.E., Carter, N. (Eds.),Cell Biology: A Laboratory Handbook. Elsevier Academic Press, Amsterdam,pp. 29–36.

Hiratani, I., Gilbert, D.M., 2009. Replication timing as an epigenetic mark.Epigenetics 4, 93–97.

Hyrien, O., Maric, C., Méchali, M., 1995. Transition in specification of embryonicmetazoan DNA replication origins. Science 270, 994–997.

Leitch, A.R., Mosgöller, W., Shi, M., Heslop-Harrison, J.S., 1992. Different patterns ofrDNA organization at interphase in nuclei of wheat and rye. J. Cell Sci. 101,751–757.

Li, J., Santoro, R., Koberna, K., Grummt, I., 2005. The chromatin remodeling complexNoRC controls replication timing of rRNA genes. EMBO J. 24, 120–127.

Little, R.D., Platt, T.H., Schildkraut, C.L., 1993. Initiation and termination of DNAreplication in human rRNA genes. Mol. Cell Biol. 13, 6600–6613.

Lucas, I., Feng, W., 2003. The essence of replication timing: determinants andsignificance. Cell Cycle 2, 560–563.

Mais, C., Wright, J.E., Prieto, J.L., Raggett, S.L., McStay, B., 2005. UBF-binding sitearrays form pseudo-NORs and sequester the RNA polymerase I transcriptionmachinery. Genes Dev. 19, 50–64.

Mellink, C.H., Bosma, A.A., De Haan, N.A., 1994. Variation in size of Ag-NORs andfluorescent rDNA in situ hybridization signals in six breeds of domestic pig.Hereditas 120, 141–149.

Méndez, J., 2009. Temporal regulation of DNA replication in mammalian cells. Crit.Rev. Biochem. Mol. Biol. 44, 343–351.

Moss, T., Langlois, F., Gagnon-Kugler, T., Stefanovsky, V., 2007. A housekeeper withpower of attorney: the rRNA genes in ribosome biogenesis. Cell. Mol. Life Sci. 64,29–49.

O’Sullivan, A.C., Sullivan, G.J., McStay, B., 2002. UBF binding in vivo is not restrictedto regulatory sequences within the vertebrate ribosomal DNA repeat. Mol. CellBiol. 22, 657–668.

Pliss, A., Koberna, K., Vecerová, J., Malínsky, J., Mašata, M., Fialová, M., et al., 2005.Spatio-temporal dynamics at rDNA foci: global switching between DNAreplication and transcription. J. Cell. Biochem. 94, 554–565.

Pope, B.D., Hiratani, I., Gilbert, D.M., 2010. Domain-wide regulation of DNAreplication timing during mammalian development. Chromosome Res. 18,127–136.

Prieto, J.L., McStay, B., 2007. Recruitment of factors linking transcription andprocessing of pre-rRNA to NOR chromatin is UBF-dependent and occursindependent of transcription in human cells. Genes Dev. 21, 2041–2054.

Prieto, J.L., McStay, B., 2008. Pseudo-NORs: a novel model for studying nucleoli.Biochim. Biophys. Acta. 1783, 2116–2123.

Raška, I., Koberna, K., Jarník, M., Petrasovicová, V., Bednár, J., Raška Jr, K., et al., 1989.Ultrastructural immunolocalization of cyclin/PCNA in synchronized 3T3 cells.Exp. Cell. Res. 184, 81–89.

Raška, I., Michel, L.S., Jarnik, M., Dundr, M., Fakan, S., Gasser, S., et al., 1991.Ultrastructural cryoimmunocytochemistry is a convenient tool for the study ofDNA replication in cultured cells. J. Electron Microsc. Tech. 18, 91–105.

Raška, I., Shaw, P.J., Cmarko, D., 2006. New insights into nucleolar architecture andactivity. Int. Rev. Cytol. 255, 177–235.

Russell, J., Zomerdijk, J.C., 2006. The RNA polymerase I transcription machinery.Biochem. Soc. Symp. 73, 203–216.

Roussel, P., Andre, C., Comai, L., Hernandez-Verdun, D., 1996. The rDNAtranscription machinery is assembled during mitosis in active NORs andabsent in inactive NORs. J. Cell Biol. 133, 235–246.

Ryba, T., Hiratani, I., Lu, J., Itoh, M., Kulik, M., Zhang, J., et al., 2010. Evolutionarilyconserved replication timing profiles predict long-range chromatin interactionsand distinguish closely related cell types. Genome Res. 20, 761–770.

Sanij, E., Hannan, R.D., 2009. The role of UBF in regulating the structure anddynamics of transcriptionally active rDNA chromatin. Epigenetics 4, 374–382.

Santoro, R., Li, J., Grummt, I., 2002. The nucleolar remodeling complex NoRCmediates heterochromatin formation and silencing of ribosomal genetranscription. Nat. Genet. 32, 393–396.

Santoro, R., 2005. The silence of the ribosomal RNA genes. Cell. Mol. Life Sci. 62,2067–2079.

Sirri, V., Roussel, P., Hernandez-Verdun, D., 2000. The AgNOR proteins: qualitativeand quantitative changes during the cell cycle. Micron. 31, 121–126.

Smirnov, E., Kalmárová, M., Koberna, K., Zemanová, Z., Malínsky, J., Mašata, M., et al.,2006. NORs and their transcription competence during the cell cycle. Folia Biol.(Praha) 52, 59–70.

Stefanovsky, V.Y., Pelletier, G., Bazett-Jones, D.P., Crane-Robinson, C., Moss, T., 2001.DNA looping in the RNA polymerase I enhancesome is the result of non-cooperative in-phase bending by two UBF molecules. Nucleic Acids Res. 29,3241–3247.

Suzuki, H., Wakana, S., Yonekawa, H., Moriwaki, K., Sakurai, S., Nevo, E., 1996.Variations in ribosomal DNA and mitochondrial DNA among chromosomalspecies of subterranean mole rats. Mol. Biol. Evol. 13, 85–92.

Wright, J.E., Mais, C., Prieto, J.L., McStay, B., 2006. A role for upstream binding factorin organizing ribosomal gene chromatin. Biochem. Soc. Symp. 73, 77–84.

Yoon, Y., Sanchez, J.A., Brun, C., Huberman, J.A., 1995. Mapping of replicationinitiation sites in human ribosomal DNA by nascent-strand abundance analysis.Mol. Cell. Biol. 15, 2482–2489.