down-regulation of cyclin b1 gene transcription in terminally differentiated skeletal muscle cells...

Down-regulation of cyclin B1 gene transcription in terminally di�erentiatedskeletal muscle cells is associated with loss of functional CCAAT-bindingNF-Y complex

A Farina1,3, I Manni1, G Fontemaggi1, M Tiainen1,4, C Cenciarelli1,5, M Bellorini2, R Mantovani2,A Sacchi1 and G Piaggio*,1

1Laboratorio di Oncogenesi Molecolare, CRS-IRE, Rome, Italy; 2Dipartimento di Genetica e Biologia dei Microrganismi,Universita' di Milano, 20133 Milan, Italy

The observation that cyclin B1 protein and mRNAs aredown-regulated in terminally di�erentiated (TD) C2C12cells, suggested us to investigate the transcriptionalregulation of the cyclin B1 gene in these cells.Transfections of cyclin B1 promoter constructs indicatethat two CCAAT boxes support cyclin B1 promoteractivity in proliferating cells. EMSAs demonstrate thatboth CCAAT boxes are recognized by the trimeric NF-Ycomplex in proliferating but not in TD cells. Transfect-ing a dominant-negative mutant of NF-YA we provideevidence that NF-Y is required for maximal promoteractivity. Addition of recombinant NF-YA to TD C2C12nuclear extracts restores binding activity in vitro, thusindicating that the loss of NF-YA in TD cells isresponsible for the lack of the NF-Y binding to theCCAAT boxes. Consistent with this, we found that theNF-YA protein is absent in TD C2C12 cells. Inconclusion, our data demonstrate that NF-Y is requiredfor cyclin B1 promoter activity. We also demonstratethat cyclin B1 expression is regulated at the transcrip-tional level in TD C2C12 cells and that the switch-o� ofcyclin B1 promoter activity in di�erentiated cells dependsupon the loss of a functional NF-Y complex. Inparticular the loss of NF-YA protein is most likelyresponsible for its inactivation.

Keywords: cyclin B1 promoter; muscle di�erentiation;CCAAT-binding factor NF-Y

Introduction

In mammalian cells the G2/M transition is controlledby a speci®c set of cell cycle genes: the mitotic cyclins(A, B1, B2), the mitotic kinase cdk1 alias cdc2p34 andthe cdc25C phosphatase (Draetta et al., 1987; Murrayet al., 1989; Pines and Hunter, 1990). The expression ofthese genes is tightly controlled, during the cell cycle,

both at transcriptional (Dalton, 1992; Hwang et al.,1995; Henglein et al., 1994; Katula et al., 1997;Lucibello et al., 1995; Piaggio et al., 1995; Pines andHunter, 1990; Zwicher et al., 1995) and translational(Brandeis and Hunt, 1996; Draetta et al., 1989; Millaret al., 1991; Murray et al., 1989; Strausfeld et al., 1991)levels and impairment in their regulated expression isthought to contribute to the malignant phenotype(Buckley et al., 1993; Keyomarsi and Pardee, 1993).

It has been reported that transcriptional regulationof cyclin B1 gene is, at least in part, responsible for theloss of expression in quiescent cells (Hwang et al.,1995; Piaggio et al., 1995). We have previouslyreported that a palindromic E-box sequence,CACGTG, located at position 7124/7119 in thepromoter region, plays a crucial role as a quiescenceresponsive element (Farina et al., 1996). Interestingly,it has been reported that the mRNA of cyclin B1 isalso down-regulated in myoblast upon induction ofterminal di�erentiation (Millar et al., 1991). We askedwhether the same molecular mechanisms previouslyobserved in quiescent cells, could be responsible for thedown-regulation of the cyclin B1 expression interminally di�erentiated muscle cells. To answer thisquestion we used skeletal muscle C2C12 cells.Proliferating precursors, or myoblast, can be inducedto di�erentiate by growth factor withdrawal. Growthfactor-starved myoblasts irreversibly exit from the cellcycle and express muscle-speci®c genes, while down-regulating a number of cell cycle regulators amongwhich are mitotic cyclins and the mitotic kinase cdk1(Ohkubo et al., 1994; Tiainen et al., 1996). Theresulting terminally di�erentiated (TD) myocyteseventually fuse with one another to generate multi-nucleated cells termed myotubes (Okazaki and Holter,1966).

The results presented here indicate that the down-regulation of the cyclin B1 expression in terminallydi�erentiated muscle cells is mediated by molecularmechanisms di�erent from those previously observed inquiescent cells. Two CCAAT boxes located close to theCap site, but not the E-box sequence located atposition 7124/7119 in the promoter region,mediated the down-regulation of cyclin B1 promoteractivity in terminally di�erentiated muscle cells. TheseCCAAT boxes are essential for the transcriptionalregulation of cyclin B1 gene and their inactivation isresponsible for the loss of promoter activity. CCAATboxes are bound with high a�nity and speci®city by anubiquitous trimeric complex, NF-Y. This complex iscomposed of three highly conserved subunits NF-YA,

*Correspondence: G Piaggio, Laboratorio di Oncogenesi Molecolare,Istituto Regina Elena, Centro della Ricerca Sperimentale, (CRS-IRE), Via delle Messi D'Oro 156, 00158 Rome, ItalyThe ®rst two authors contributed equally to this workCurrent addresses: 3NIH, NICHHD, Bethesda, Maryland 20892,USA; 4University of Helsinki, Haartman Institute, Cell cyclelaboratory, Haartmaninkatu 3, Helsinki, Finland; 5New YorkUniversity, Medical Center, Department of Pediatrics, 550 FirstAvenue, 10016 New York, NY, USAReceived 20 July 1998; revised 16 September 1998; accepted 6October 1998

Oncogene (1999) 18, 2818 ± 2827ã 1999 Stockton Press All rights reserved 0950 ± 9232/99 $12.00

http://www.stockton-press.co.uk/onc

NF-YB, and NF-YC (Bellorini et al., 1997; Hooft vanHuijsduijnen et al., 1990), all required for DNAbinding (Sinha et al., 1995). In the MHC class II Eapromoter, which has no functional TATA box, NF-Yappears to play an important role in correctlypositioning the start site together with an initiatorelement (Dorn et al., 1987; Mantovani et al., 1993).Functional analysis has shown that NF-Y can beimportant in transcription re-initiation, suggesting arole in basal transcription (Mantovani et al., 1992). Inthe promoters of the genes encoding the mitotic cyclinA as well as the CDK1 kinase and the CDC25Cphosphatase and other cell cycle regulated genes,several CCAAT elements are present, that have beenreported to be recognized by the NF-Y complex(Zwicher et al., 1995).

In the present paper we report that: (i) the twoCCAAT boxes located close to the Cap site supportcyclin B1 basal promoter activity; (ii) the loss of afunctional NF-Y complex is responsible for the down-regulation of the promoter activity; (iii) NF-YAprotein is the limiting factor for NF-Y binding activityin TD cells; (iv) the lack of NFY-A protein in TD cellsis, at least in part, responsible for the down-regulationof cyclin B1 promoter activity in these cells.

Results

Down-regulation of cyclin B1 expression in TD C2C12cells occurs at transcriptional level

In di�erent cell types, cyclin B1 expression seems to berestricted to the proliferating state, being down-regulated in growth arrested or di�erentiated cells(Doerner et al., 1996; Fromm and Overbeek, 1996;Hanley-Hyde et al., 1992; Zhao et al., 1995). It hasbeen reported that the cyclin B1 mRNA is down-regulated in C2C12 myoblasts upon induction ofterminal di�erentiation (Ohkubo et al., 1994). Weasked whether cyclin B1 protein expression is alsodown-regulated. To answer this question, the levels ofboth cyclin B1 protein and mRNAs were evaluated inproliferating and TD C2C12 cells. Northern blothybridization experiments (Figure 1a), and Westernblot assays (Figure 1b), show that the amount of cyclinB1 transcripts and protein, dramatically decreases inpost-mitotic myotubes.

To investigate whether down-regulation of cyclin B1mRNA occurs at the transcriptional level, C2C12 cellswere stably cotransfected with CAT reporter constructsdriven by two di�erent cyclin B1 promoter fragments(p332B1CAT and p240B1CAT) (Piaggio et al., 1995).Upstream to the CAT reporter gene, the plasmids carrythe wild-type cyclin B1 promoter fragment from 7154to +182 bp (p332B1CAT), or a 5' deletion (757 to+182 bp) of this promoter (p240B1CAT). We havepreviously reported that both promoter fragmentsdrive comparable transcription in proliferatingNIH3T3 cells (Piaggio et al., 1995). Similar resultswere obtained in proliferating C2C12 cells (data notshown). As a control, a C2C12 RSVCAT transfectedpolyclonal population (a kind gift from Silvia Soddu)was employed. Figure 2 shows that both cyclin B1promoter constructs are capable of driving transcrip-tion in C2C12 proliferating cells. In contrast, no

transcriptional activity is detected in TD cells withboth constructs (Figure 2), whereas the transcriptionalactivity of the viral RSV promoter is not down-regulated by terminal di�erentiation (data not shown).

Taken together these results demonstrate that cyclinB1 protein, mRNA, and promoter activity are down-regulated in TD C2C12 cells, indicating that theexpression of cyclin B1 gene in these conditions isregulated, at least in part, at transcriptional level.Moreover, the ®nding that the shorter construct(p240B1CAT) becomes inactive in TD cells indicatesthat the region from 757 bp to +182 bp upstream ofthe start site contains elements required for the down-regulation of cyclin B1 promoter activity in TD cells.

Figure 1 Cyclin B1 expression is down-regulated in TD C2C12.(a) Northern blot analysis was performed on total RNA extractedfrom proliferating, P, and TD C2C12 cells. 25 mg of total RNAswere size-fractionated on a 1% agarose gel, blotted on a nylonmembrane, hybridized with 32P-labelled cyclin B1 probe andassessed by auto-radiography. To normalize RNA loading, themembrane was re-hybridized with the 32P-labelled cDNA of thehouse-keeping gene GAPDH. (b) Western blot analysis wasperformed on total cell lysates from proliferating, P, and TDC2C12 cells. The extracts were probed with a monoclonalantibody anti-cyclin B1. To normalize protein loading, the ®lterwas stained with a rabbit polyclonal antiserum raised againstCREB protein

Inactivation of cyclin B1 promoter in differentiated cellsA Farina et al

2819

The CCAAT boxes in the cyclin B1 promoter play afunctional role on its basal transcriptional activity

Computer assisted analysis of the sequence fromnucleotides 757 to +182 reveals the presence of twoputative CCAAT boxes located at position 714 to717 and +16 to +20 of the cyclin B1 promoter. Inorder to investigate the functional contribution of theseelements to the promoter activity, mutations wereintroduced in the wild-type promoter (CCAAT toCTCGA). We generated three CAT reporter constructscarrying mutations in the upstream (pmtupCCAAT),downstream (pmtdownCCAAT) or in both (pmtup/downCCAAT) boxes in the context of the shorterconstruct p240B1CAT. Mutated and wild-type con-structs were transiently transfected in proliferatingC2C12 cells. The CAT activity of p240B1CATconstruct was made 100%, and the relative activity ofthe CCAAT mutated plasmids was calculated. Asshown in Figure 3, the single mutation of eitherCCAAT box does not signi®cantly a�ect promoteractivity; while the mutations of both boxes reduce by80% the promoter activity, indicating that the twoCCAAT boxes are key sequences for the cyclin B1promoter activity. Interestingly, these data also indicatethat in these experiments each CCAAT box on its ownis su�cient to support basal promoter activity.

NF-Y protein complex binds the CCAAT boxes inproliferating but not in TD C2C12 cells

By EMSAs we veri®ed the presence of nuclear factorsbinding to the CCAAT boxes in proliferating and TD

cells. Electromobility shift assays reveal one proteincomplex binding to two radiolabelled probes, spanningnt 726 to +1 and 73 to +29 respectively in thecyclin B1 promoter (B1upCCAAT and B1down-CCAAT) (Figure 4a). The binding activity of thecomplex present in nuclear extracts from proliferatingcells (lanes 2 and 12) indicates that the CCAAT-binding complex is highly abundant in proliferatingcells. By contrast, the complex is almost undetectablein the nuclear extracts prepared from terminallydi�erentiated cells (lanes 7 and 17). Competitionassays were then designed with both cyclin B1-derivedprobes to asses the binding speci®city of the complexto the CCAAT boxes. As shown in Figure 4a, thecomplex is speci®cally inhibited by a 200-fold molarexcess of unlabelled probes (lanes 3 and 13), but not bya 1000-fold molar excess of a similar oligonucleotidescontaining a mutated CCAAT box (lanes 5 and 15), orby an unrelated probe containing an E-box sequence(lanes 6 and 16). Although in these experiments a nonreproducible faster migrating complex is also barelydetectable with both probes (lanes 2 and 12), it is

Figure 2 The down-regulation of cyclin B1 expression in TDC2C12 cells occurs at transcriptional level. Cyclin B1 promoterCAT-reporter constructs p332B1CAT and p240B1CAT (Piaggioet al., 1995), were stably transfected by calcium-phosphate inC2C12 cells. The obtained poly-clonal cell lines were analysed forthe CAT activity. CAT assay was performed as described(Desvernie et al., 1991). The values in TD cells are expressed,on Y axis, as percentages of the CAT activity obtained fromproliferating cells (100%). Results represent one typical experi-ment. A schematic representation of the two used reporterconstructs is represented

Figure 3 The CCAAT boxes, present in the cyclin B1 promoter,play a functional role on its basal transcriptional activity. CyclinB1 promoter-CAT reporter constructs (10 mg), p240B1CAT,pmtupCCAAT, pmtdownCCAAT or pmtup/downCCAAT werecontransfected by the calcium-phosphate method into proliferat-ing C2C12 cells along with the CMVb-gal reporter construct(1 mg). CAT assays were performed as described (Desvernie et al.,1991). The CAT activity of p240B1CAT was made 100%, and therelative activities of the mutated constructs are indicated. Resultsrepresent the mean of three independent experiments eachperformed in duplicate. A schematic representation of the usedreported constructs is represented


2820

inhibited by all tested competitors indicating that itdoes not re¯ect speci®c binding. This ®nding demon-strates that one protein complex speci®cally binds boththe B1upCCAAT and B1downCCAAT probes in vitro.The observation that CREB-binding activity does notchange in di�erent cell populations (proliferatingversus terminally di�erentiated, Figure 4b) con®rmsthat the loss of CCAAT binding complex in TD cellscan not be attributed to trivial technical problems.

It has been reported that both CTF/NF1 (Zorbas etal., 1992) and C/EBP (Osada et al., 1996), bindCCAAT-related sequences, while NF-Y binds aCCAAT consensus sequence that ®ts very well withthe statistically derived one (Bucher, 1990). In order togain information about the identity of the proteins thatbind the CCAAT boxes of the cyclin B1 promoter, weperformed competition experiments using bothB1upCCAAT and B1downCCAAT radiolabelledprobes and cold competitor oligonucleotides contain-ing the speci®c cognate sequences for the NF1/CTFfamily of proteins, for C/EBP, and for CBF/NF-Y.Only the oligonucleotide containing the CBF/NF-Yrecognition sequence is able to compete out the cyclinB1 CCAAT-binding complexes (data not shown),indicating that NF-Y complexes bind the CCAATboxes of the cyclin B1 promoter. Direct evidence of theNF-Y binding was obtained through the use of speci®cantibodies against either NF-YA or NF-YB in thebinding reactions with both radio-labelled probes.C2C12 nuclear extracts were pre-incubated withantibodies against the NF-YA or NF-YB subunitsbut not against NF-YC protein because this was stillunavailable. As shown in Figure 4c NF-YA antibodyretards the migration of the assembled complex withboth probes (lanes 3 and 7); the NF-YB antibody, butnot a pre-immune serum (lanes 2 and 6), retards, aswell as interferes with, the migration of the complex(lanes 4 and 8). Altogether, these results provideevidence that, at least in vitro, NF-Y recognizes thetwo cyclin B1 CCAAT boxes.

Figure 4 (a) Protein complexes bind the two CCAAT boxes ofthe cyclin B1 promoter only in proliferating cells. Gel mobilityretardation assays were performed with two double-stranded, 32P-mlabelled oligonucleotides, spanning position 726 to +1(B1upCCAAT) and 73 to +29 (B1downCCAAT) containingthe two CCAAT boxes of the cyclin B1 promoter. On bothprobes, addition of nuclear extracts from proliferating (C2C12

PNE, lanes 2 and 12), but not from TD (C2C12 TDNE, lanes 7and 17) C2C12 cells resulted in a retarded band indicated by thearrows. Lanes 1 and 11 represent the B1downCCAAT andB1upCCAAT free probes respectively. The formation of thecomplex is speci®cally inhibited by a 200-fold molar excess ofunlabelled probe (lanes 3 and 13) but not by a 1000-fold molarexcess of oligonucleotides containing a mutated CCAAT box inwhich this sequence is changed into CTCGA (lanes 5 and 15) or a1000-fold molar excess of an unrelated probe (lanes 6 and 16). (b)Gel mobility retardation assays were performed with a double-stranded, 32P-labelled oligonucleotide, containing the CREelement of the somatostatin promoter. Lane 1 represents theCRE free probe, lanes 2 and 3 represent the retarded bandsobtained with addition of nuclear extracts from proliferating andTD C2C12 cells, respectively. (c) NF-Y complex binds the twoCCAAT boxes of the cyclin B1 promoter. Gel mobilityretardation assays were performed with 32P-labelled probes as in(a). Nuclear extracts from proliferating cells (P NE), before assaywith probes, were preincubated with a pre-immune serum (lanes2, 6) a monoclonal antibody anti-NF-YA (mAb YA7)(Mantovani et al., 1992) (lanes 3, 7) and a polyclonal antiserumraised aganst the recombinant NF-YB protein (pRaYB)(Mantovani et al., 1992) (lanes 4, 8). Both antibodies super-shiftthe complex. The supershifted bands are indicated by the arrows


2821

NF-Y plays an essential role in the cyclin B1transcription in vivo

To evaluate the functional signi®cance of NF-Ycomplex on the cyclin B1 promoter activity in vivowe over-expressed NF-YA mutant cDNA in proliferat-ing C2C12 cells. The cells were cotransfected with thecyclin B1 promoter and di�erent amounts of NF-YAmutant expression vector. It is known that the NF-YAmutant inhibits NF-Y dependent transcription bysequestering the NF-YB and NF-YC subunits indefective complexes, which are unable to bind theDNA, thus functioning as a dominant negative protein(Mantovani, 1998). In all experiments, the basalactivity (100%) was assessed by co-transfecting theempty vector along with the CAT constructp240B1CAT (lane 1). Repeated experiments wereperformed and similar results were observed; never-theless, minimal variations were observed in di�erentcell lysates. Analysis of CAT activity reported inFigure 5 demonstrates that (a) the over-expression ofmutant NF-YA protein completely abrogate cyclin B1promoter activity as expected for a dominant negativeprotein, and (b) its e�ect is dose-dependent. As shownin Figure 5, 2.5 mg of NF-YA mutant-expressingplasmid decreases by 70% the transcriptional activityof the p240B1CAT construct whereas the higher doseof plasmid (5 mg) decreases transcription by 98%. Thespeci®city of the e�ects generated by mutant NF-YAprotein on cyclin B1 promoter activity was alsocon®rmed by co-transfecting a reporter gene drivenby SV40 promoter (pSV2CAT). The over-expression ofNF-YA dominant negative protein does not signifi-cantly a�ect the SV40 promoter activity (Figure 5).Taken together these results demonstrate that thetransfection of a dominant negative analogue of NF-YA speci®cally inhibits cyclin B1 promoter activity in adose-dependent manner, thus giving direct evidence

that a functional NF-Y complex is required formaximal promoter activity.

The expression of NF-Y genes is down-regulated inTD C2C12 cells

Next we evaluated whether the lack of binding of NF-Y complexes to the CCAAT boxes on the cyclin B1promoter in TD C2C12 cells could be attributed to alack of the expression of the three NF-Y subunits.Northern blot analysis of total RNAs prepared fromproliferating and TD C2C12 cells was performed.Figure 6a shows that the levels of the three NF-Ysubunits transcripts decrease in TD C2C12 cells. Asassessed by densitometric analysis, the di�erentiation-associated decrease in mRNA levels was of 59 and56% for NF-YB and NF-YC respectively and 70% forNF-YA mRNA. To assess whether the expression ofthe NF-Y proteins parallels that of mRNAs Westernblot analysis was performed. Figure 6b shows that theNF-YA protein is absent in TD cells, whereas the NF-YB protein is signi®cantly reduced. The reduction inthe level of expression of NF-YB protein in TD cellsparallels the down-regulation of its mRNA, suggestinga transcriptional regulation of this protein during thedi�erentiation process. By contrast, the lack of NF-YAprotein expression in TD C2C12 cells does not matchwith the presence of detectable levels of NFY-AmRNA. This ®nding strongly suggests that theexpression of NF-YA protein during di�erentiation isregulated both at transcriptional and translationallevels. Furthermore, the signi®cant reduction of NF-YA protein in TD C2C12 cells has been interpreted asthe event responsible for the lack of NF-Y binding tothe CCAAT boxes of the cyclin B1 promoter and theinhibition of the promoter activity in post-mitotic cells.

NF-YA protein is the limiting factor for the binding ofNF-Y complex in TD C2C12 cells

To assess whether loss of the binding of NF-Y complexto the CCAAT boxes of cyclin B1 promoter in TDC2C12 cells could be attributed to loss of NF-YAprotein we performed electromobility shift assaysadding decreasing amounts of recombinant puri®edNF-YA protein to the nuclear extracts from these cells.As shown in Figure 7, the addition of 3 ng of puri®edNF-YA protein to TD nuclear extract is su�cient tofully restore in vitro binding activity of the NF-Ycomplex to both the B1upCCAAT (lane 6) andB1downCCAAT (data not shown) probes in vitro.The addition of puri®ed NF-YA protein to nuclearextracts from proliferating cells does not substantiallymodify the complex mobility, but it induces asigni®cant increase in the complex formation (lane 7and data not shown). The addition of recombinantpuri®ed NF-YB/NF-YC dimer to the nuclear extractsobtained from TD C2C12 cells does not restore thebinding activity of the NF-Y complex to bothB1upCCAAT (lane 11) and B1downCCAAT (datanot shown) nor increase the complex formation innuclear extracts from proliferating cells (lane 9 anddata not shown). The observation that the addition ofrecombinant puri®ed NF-YA/NF-YB/NF-YC trimerto the nuclear extracts obtained from proliferating andTD C2C12 cells fully restores binding activity of the

Figure 5 NF-Y plays an essential role on the cyclin B1transcription in vivo. Proliferating C2C12 cells were cotransfectedwith 5 mg of cyclin B1 promoter-CAT reporter constructp240B1CAT or 5 mg of SV40 promoter-CAT reporter constructpSV2CAT, 2.5 mg or 5 mg of eucariotic vector expressing NF-YAdominant negative protein, and 0.5 mg of CMVb-gal reporterconstruct. The CAT values, normalized against b-galactosidase,are expressed as percentages of the basal activity (100%) assessedby cotransfecting the empty vector along with each CATconstruct. Results represent the mean of three independentexperiments, each performed in duplicate


2822

NF-Y complex to both B1upCCAAT (lanes 12 and 13)and B1downCCAAT (data not shown) demonstratesthat the recombinant NF-YB/NF-YC dimer is active,and the fact that it does not restore the binding activityof the NF-Y complex in TD cells can not be attributedto trivial technical problems.

Altogether these results clearly demonstrate that theNF-YA protein is a limiting factor for binding to the

CCAAT box in post-mitotic cells at least in vitro, andsuggest a functional role for the loss of NF-YA proteinin the regulation of cyclin B1 promoter activity in TDcells. Moreover, the observation that the amount ofNF-YA in proliferating cells determines the levels ofDNA binding activity of the NF-Y complex (comparelane 2 with lane 7) is in good agreement with the®nding that NF-YB and C are expressed at higherlevels than NF-YA in proliferating cells (RM,unpublished observation).

Overexpression of NF-YA in vivo is switched o� inTD C2C12 cells

To investigate whether loss of NF-YA protein in TDC2C12 is responsible for the down-regulation of cyclinB1 promoter activity in vivo, we transient transfectedNF-YA cDNA along with cyclin B1 promoter inC2C12 proliferating cells and, after serum withdrawal,we evaluated CAT activity in TD cells. Despite severalattempts we were not able to restore cyclin B1promoter activity in TD C2C12. Western blot analysiswas performed to assess the level of NF-YA expressionachieved after transfection of exogenous gene. Nuclearextracts from NF-YA transfected myoblasts andmyotubes and, as control, from NF-YB transfectedcells were prepared and the results are reported inFigure 8. Densitometric analysis demonstrate that in

Figure 6 The expression of NF-Y genes is down-regulated in TDC2C12 cells. (a) Northern blot analysis was performed on totalRNA extracted from proliferating, P, and TD C2C12 cells. 10 mgof total RNAs were size-fractionated on a 1% agarose gel, blottedon a nylon membrane, hybridized with 32P-labelled cyclin B1probe and assessed by auto-radiography. To normalize RNAloading, the membrane was re-hybridized with the 32P-labelledcDNA of the house-keeping gene GAPDH. The percentage ofdecrease amount of each mRNA has been evaluated bydensitometric analysis performed as in Figure 5a. The sizes ofthe transcripts are indicated. (b) Western blot analysis wasperformed on total cell lysates from P and TD C2C12 cells. Theextracts were probed with two polyclonal antiserum raised againsta C-terminal peptide of NF-YA (pRaYA/C) (Mantovani et al.,1992) or against the recombinant NF-YB protein (pRaYB)(Mantovani et al., 1992) respectively. To normalize proteinloading, the ®lter was stained with a rabbit polyclonal antiserumraised against CREB protein

Figure 7 NF-YA protein is the limiting factor for the binding ofNF-Y complex in TDC2C12 cells. Gel mobility retardation assayswere performed with B1upCCAAT 32P-labelled oligonucleotide.Decreasing amounts of recombinant puri®ed NF-YA protein(rNF-YApp) from 0.15 ± 0.003 mg (lanes 4 to 6) or 0.15 mg ofrecombinant NF-YB and 0.15 mg of recombinant NF-YC (rNF-YB/YCpp, lane 11) or 0.15 mg of recombinant NF-YB, 0.15 mg ofrecombinant NF-YC and 0.15 mg of recombinant NF-YA (lane12) have been added to the nuclear extract from TD C2C12 cells.Lanes 2 and 3, and lanes 8 and 10 represent the shifts obtainedwhen nuclear extracts derived from proliferating (PNE) and TDC2C12 (TDNE) cells were directly assayed with the probe. In lane7, 1.5 mg of recombinant puri®ed NF-YA protein has been addedto the nuclear extract from proliferating C2C12 cells. In lane 9and 13 recombinant puri®ed NF-YB and NF-YC or NF-YA,NF-YB, and NF-YC proteins have been added to the nuclearextract from proliferating C2C12 cells


2823

proliferating cells the amount of NF-YA (lane 3) andNF-YB (lane 2) subunits in transfected cells was abouttenfold higher than that observed in untransfected cells(lane 1). The same cells, however, after serumwithdrawal (TD), show barely detectable levels ofNF-YA protein independently whether or not the cellswere previously transfected with the exogenous gene(lanes 4, 6). By contrast NF-YB protein in TD cells isalways expressed and, as expected, the level ofexpression is higher in transfected (lane 5) than inuntransfected cells (lane 4). To exclude that di�erentamounts of loaded protein might be responsible for theobserved di�erences, the same blot was checked withan anti-CREB antibody. The observation that thelevels of expression of the CREB protein are similar indi�erent samples indicate that the loss of NF-YAexpression in TD C2C12 cells can not be attributed totechnical problems. We therefore conclude that over-expression of an exogenous NF-YA protein in TD cellsis prevented and that the switch-o� of its expression isspeci®c. Indeed we do not observe similar down-regulation of exogenous NF-YB protein.

Discussion

Although it has been reported that cyclin B1 expressionis down-regulated in several di�erentiated cell types(Doerner et al., 1996; Fromm and Overbeek, 1996;

Hanley-Hyde et al., 1992; Ohkubo et al., 1994; Zhao etal., 1995), the molecular mechanisms underlying thismodulation are still unknown. The aim of this study isto elucidate the mechanism(s) controlling the expres-sion of the cyclin B1 gene in the TD skeletal musclecell line C2C12. This is the ®rst report which identi®esmolecular complexes involved in the down-regulationof cyclin B1 expression in terminally di�erentiatedcells. The data reported here suggest a model wherebythe transcription of cyclin B1 is supported by NF-Y inproliferating cells and is switched o� in post-mitoticcells, concomitantly with the loss of a functional NF-Ycomplex.

We here report that in TD C2C12 cells cyclin B1gene expression is down-regulated at protein level. Thisobservation is in good agreement with the ®nding thatthe proteolytic machinery of the mitotic-type cyclins isactive in TD C2C12 cells (Brandeis and Hunt, 1996). Ithas been previously reported that also cyclin B1mRNA is down-regulated in terminally di�erentiatedcells (Ohkubo et al., 1994). The ®nding that in TD cellsboth protein and mRNA are down-regulated indicatesthat cyclin B1 expression in these cells is regulated bothat post-transcriptional and transcriptional level. Ac-cordingly, we ®nd that cyclin B1 promoter activity isalso down-regulated under these conditions.

The functional analysis of a truncated cyclin B1promoter construct in stably transfected cells showsthat 57 bp of the upstream sequence and 182 bp of thetranscribed non-translated sequence are su�cient toconfer di�erentiation-dependent repression of a CATreporter gene in post-mitotic C2C12 cells. Interestingly,we previously demonstrated that this region is notsu�cient to confer growth arrest-dependent repressionin quiescent NIH3T3 cells necessary for the transcrip-tional inactivation of the cyclin B1 promoter in theregion from nt 7158 to 758 (Piaggio et al., 1995;Farina et al., 1996). Thus, di�erent molecular pathwaysact to down-regulate cyclin B1 promoter in the twodi�erent post-mitotic states (terminal di�erentiationversus quiescence).

The promoter region su�cient to confer differentia-tion-dependent down-regulated activity in TD cellscontains two CCAAT boxes. Mutations of bothCCAAT boxes lead to a dramatic decrease in thereporter gene expression in proliferating C2C12 cells.Interestingly, when each CCAAT box is individuallymutated the expression of the reporter gene is nota�ected. These data suggest that each one of theCCAAT elements is su�cient in supporting the basaltranscription of the cyclin B1 promoter.

By DNA-binding competition and antibody super-shift experiments, we unambiguously identi®ed NF-Yas the CCAAT-binding factor interacting with boththese elements. The direct evidence that NF-Y isrequired for the activity of cyclin B1 promoter wasobtained using a dominant negative mutant. Intransient transfection experiments we demonstratethat the dominant negative mutant of NF-Y is ableto completely inhibit cyclin B1 promoter activity. Theseresults establish that NF-Y is a rate-limiting transcrip-tion factor for the overall cyclin B1 promoter activity.Since the binding activity of NF-Y to CCAAT boxes ishighly e�cient in proliferating cells, while it isdramatically reduced in TD cells, it is inferred thatthe loss of a functional NF-Y complex might be

Figure 8 Overexpression of NF-YA in vivo is switched o� in TDC2C12 cells. Western blot analysis was performed on nuclearextracts from untransfected (lanes 1 and 4), NF-YB (lanes 2 and5), and NF-YA (lanes 3 and 6) transfected P and TD C2C12 cellsrespectively. The extracts were probed with two polyclonalantiserum raised against a C-terminal peptide of NF-YA(pRaYA/C) (Mantovani et al., 1992) or against the recombinantNF-YB protein (pRaYB) (Mantovani et al., 1992) respectively.To normalize protein loading, the ®lter was stained with a rabbitpolyclonal antiserum raised against CREB protein


2824

responsible for the down-regulation of cyclin B1transcription in TD cells.

To elucidate the molecular mechanism responsiblefor the loss of binding activity we investigated theexpression of NF-Y subunits in TD C2C12 cells. Ourobservations indicate that NF-YA protein is comple-tely down-regulated in terminally di�erentiated cells. Adecrease in the mRNA levels of NF-YA, NF-YB andNF-YC is also observed in terminally di�erentiatedC2C12 cells, but only NF-YA protein is undetectablein these conditions. This is the ®rst report indicatingthat NFY-A is down-regulated in TD cells, never-theless a modulation of NF-YA protein has beenpreviously reported. Reduction of NF-YA expressionhas been reported in IMR-90 ®broblasts after serum-deprivation (Chang and Liu, 1994) whereas an increasein NFY-A expression was observed during maturationof human peripheral monocytes to macrophages(Marziali et al., 1997). The observations that duringmaturation of monocytes to macrophages NFY-Aprotein expression increases whereas in terminallydi�erentiated C2C12 cells NFY-A protein is undetect-able, can be reconciled considering that in macro-phages several genes require NF-Y function. MHCclass II and other genes involved in antigen presenta-tion are critically dependent upon their CCAATsequences. By contrast, none of the muscle speci®cpromoters cloned so far has been shown to dependupon NF-Y activity. In fact, among the 120 di�erentpromoters activated by NF-Y we found nonespeci®cally active in myotubes (Mantovani, 1998).Therefore it might be speculated that the absence ofNFY-A relies on a tissue-speci®c di�erentiation statusof the cells.

We were able to reconstitute NF-Y binding to bothcyclin B1 CCAAT boxes by adding recombinant NF-YA to nuclear extracts from TD C2C12 cells, thusproving that the NF-YB/NF-YC dimers still present inTD cells cannot bind DNA in the absence of theircompanion subunit (NFY-A). Furthermore, evidencethat the loss of CCAAT binding activity is due to thelack of NF-YA in the nuclear extracts from TD C2C12cells is given. In TD myotubes the addition of NF-YArestores binding activity at the same level as occurs innuclear extracts from proliferating myoblasts (Figure7). This observation suggests that the amount of NF-YA in proliferating cells determines the levels of DNAbinding activity of the NF-Y complex and, at least invitro, the amount of NF-YB/NF-YC dimers stillpresent in myotubes is comparable to the amountthat in myoblasts is involved in the binding with NF-YA.

Based on the observation that NF-YA is a limitingfactor for the binding activity of NF-Y complex invitro, we tried to restore cyclin B1 promoter activity inTD C2C12 cells by the constitutive over-expression ofNF-YA protein. Surprisingly, we found that the over-expression of exogenous NF-YA is not allowed interminal di�erentiated C2C12 cells. One possibleexplanation is that cells transfected with NF-YA diewhen switched to di�erentiation medium thus indicat-ing an important role of this gene in the switch fromproliferation to di�erentiation. Otherwise, the fact thatwe were unable to express exogenous NFY-A proteinin terminally di�erentiated cells has to be attributed todi�erent protein stability between proliferating and TD

cells or to post-translational mechanisms present in TDcells. Indeed NF-YA cDNA is driven by SV40constitutive promoter and the constructs do notcontain untranslated 5' or 3' regions. Moreover, thesame construct generates signi®cant expression ofNFYA protein in proliferating cells (Figure 5). Ingood agreement with this hypothesis the expression ofsigni®cant amounts of exogenous NF-YB protein wasachieved in terminally di�erentiated cells using thesame vector. According to these results we speculatethat similar mechanism/s are also responsible for thedown-regulation of the endogenous NF-YA proteinobserved in Figure 6b.

Finally, the data presented indicate that the histone-fold containing NF-YB/NF-YC dimers are constitu-tively distributed in the cells independently on theirfunctional status. In fact their expression is largelymaintained upon terminal di�erentiation, albeit atreduced levels. The activity of NFY complexapparently is regulated by the NFY-A subunit.Elucidation of the regulatory circuits governing NF-YA protein levels, possibly by control of translationand/or proteolysis, will shed further light on thebiology of this conserved protein.

It is interesting to notice that the promoters ofseveral cell cycle regulatory genes such as murine cyclinB2, human cyclin A, CDK1, and CDC25C do allcontain CCAAT boxes. Furthermore, NF-Y complexhas been reported to bind these sequences (Zwicher etal., 1995 and RM, unpublished data). For some ofthese genes, the role of NF-Y in the basal transcriptionhas yet to be demonstrated (Zwicher et al., 1995). Thisevidence, together with our ®nding that cyclin A andCDK1 promoters are down-regulated in TD C2C12cells (data not shown) support the hypothesis that thetranscription of these genes in post-mitotic skeletalmuscle cells is down-regulated by the same commonmechanism presently described for the cyclin B1 gene.

Materials and methods

Reporter plasmid constructs

p332B1CAT and p240B1CAT have been already described(Piaggio et al., 1995). The reporter plasmids containing amutation in one or both of the CCAAT box elements weregenerated using the QuikChangeTM Site-Directed Mutagen-esis Kit (Stratagene). The primers, used are the following(mutated nucleotides are underlined): 5'-GCGACC-GCAGCCGCTCGAGGGAAGGGAGTGAGTGC-3' andits complementary sequence to generate the pmtupC-CAAT reporter plasmid; 5'-GAGTGCCACGAACAG-GCTCGAAAGGAGGGAGCAGTGCGG-3' and its com-plementary sequence to generate the pmtdownCCAATreporter plasmid; and both couples of the above primersto generate the pmtup/downCCAAT reporter plasmid. Themutations were con®rmed by DNA sequencing.

The plasmid CMVbgal, expressing a b-galactosidasecDNA under the control of the cytomegalovirus immediate-early promoter, has been used as an internal control intransient transfection experiments.

Cell culture, DNA transfections and CAT assay

The mouse myoblast line C2C12 (Blau et al., 1985) a clonederived from C2 cell line (Ya�e and Saxel, 1977) wascultured in DMEM containing 10% foetal bovine serum


2825

(FBS) and antibiotics. Di�erentiation was induced byplating the cells into collagen-coated dishes and switchingthem to serum free (SF) medium for 48 h: DMEMsupplemented with Redu-Ser (Upstate BiotechnologyIncorporated, Lake Placid, NY, USA) to a ®nal concentra-tion of 5 mg/ml human insulin, 5 mg/ml human (holo)transferrin, and 5 ng/ml sodium selenite. Fifty mM Ara-Cwas added to the SF medium to eliminate undi�erentiatedcells. As previously described (Tiainen et al., 1996), morethan 90% of the cells become terminally di�erentiatedunder these conditions.

Transient DNA transfections were performed using thecalcium phosphate precipitation technique (Graham and VanDer Eb, 1973) modi®ed by Okayama (Chen and Okayama,1987). In each 60-mm plate, 1.56105 cells were transfectedwith precipitates containing: 10 mg of p240B1CAT, pmtupC-CAAT, pmtdownCCAAT plasmid or pmtup/downCCAAT;1 mg of CMVbgal plasmid as an internal control fortransfection e�ciency. For cotransfection experiments theprecipitates contained: (i) 5 mg of p240B1CAT, or SV2CAT;(ii) 0.5 mg of CMVbgal plasmid as an internal control fortransfection e�ciency; and only in cotransfection experi-ments, (iii) 2.5 mg or 5 mg of an eucariotic expression vectorfor NF-YA dominant negative (Mantovani et al., 1994).Sixteen hours later, the precipitate was removed, and freshmedium was added. Cells were harvested from 24 ± 48 h aftertransfection and CAT activity was assayed in whole-cellextracts as described (Desvernie et al., 1991).

Stably transfected C2C12 cells were generated using thecalcium phosphate precipitation technique (Graham and VanDer Eb, 1973) modi®ed by Okayama (Chen and Okayama,1987). Cells were cotransfected with p332B1CAT orp240B1CAT reporter plasmids and a vector containing thehygromycin resistance gene under the control of the SV40promoter (pBabeHygro). The culture medium was replaced12 h after seeding, and the cells were exposed to the selectionmedium after an additional 48 h (DMEM plus hygromycin300 mg/ml).

Northern blot analysis

C2C12 cells were plated at 56106 cells/collagen-coated,150-mm dish and induced to di�erentiate in the presence ofAra-C. Total cellular RNA was extracted by theguanidinium thiocyanate/phenol procedure (Chomczynskiand Sacchi, 1987) from myotubes as well as fromproliferating C2C12 cells. Samples were loaded (25 mg/lane) on formaldehyde gels, then separated and blotted.Filters were hybridized with murine NF-YA, human NF-YB, NF-YC, cyclin B1, and GAPDH cDNA probesaccording to standard protocols (Sambrook et al., 1989).The probes were labelled using a-32P-dATP and the randompriming procedure according to the manufacturer'sspeci®cations (Boehringer, Mannheim, Germany).

Total-cell lysate and Western blot analysis

Total-cell lysate was obtained from proliferating and TDC2C12 cells by freeze-thaw in 0.4 M NaCl bu�er containing50 mM Tris-HCl (pH 8), 25% glycerol 1 mM EDTA,2.5 mM DTT, 10 mg/ml leupeptin, 4 mg/ml pepstatin,5 mg/ml aprotinin, 50 mM NaF, 1 mM orthovanadate and1 mM PMSF. To compensate for the higher structuralprotein content in myotubes compared to myoblasts, 50and 100 mg of proteins obtained from proliferating and TDcells respectively were loaded and separated by SDS ±PAGE (12% polyacrylamide) and electro-blotted ontonitrocellulose. After staining in 0.2% Ponceau S in 3%

TCA, the ®lter was washed twice in PBS and proteinbinding sites blocked in 5% non-fat dried milk in PBS. The®lter was washed twice in PBS and treated with 0.3 mg/mlof polyclonal antiserum pRaYA/C, or pRaYB (Marziali etal., 1997), or 33 ng/ml of anti cyclin B1 mouse monoclonalantibody (Santa Cruz Biotechnology, Inc. CA, USA), or50 ng/ml of anti CREB rabbit polyclonal antibody (NewEngland Biolabs), in 3% BSA /PBS for 2 h at roomtemperature. After four washes in TBS (150 mM NaCl,50 mM Tris-HCl, pH 7.9) the ®lter was incubated with thesecondary antibody (anti-rabbit or anti-mouse Ig con-jugated with peroxidase), in 3% BSA/TBS for 1 h at roomtemperature. The ®lter was washed four times as above andblots were developed using the ECL procedure (Amersham,Little Chalfont, UK).

Electromobility shift assays and nuclear extracts

Electromobility shift assays were performed in a 25 mlDNA binding reaction which contained 5 ± 10 mg of C2C12nuclear extracts, 4 fmol of labelled duplex oligonucleotide,binding bu�er (20 mM Tris-HCl pH 7.8, 60 mM KCl,0.5 mM EDTA, 0.1 mM DTT, 3 mM MgCl2), 1.5 mg ofpoly (dl-dC), and 10 mM spermidine. The reaction wascarried out on ice for 30 min and the protein-DNAcomplexes were subjected to native electrophoresis on 5%polyacrylamide, 0.56TBE gels. The following oligonucleo-tides were used as probes and competitors (consensussites are underlined): B1upCCAAT: 5'-CCGCAGCCGCCAATGGGAAGGGAGTGA; B1downCCAAT: 5'-GT-GAGTGCCACGAACAGGCCAATAAGGAGGGAG;B1upCTCGA: 5'-GCGACCGCAGCCGCTCGAGGGAA-GGGAGTGAGTGC; B1downCTCGA: 5'-GAGTGCC-ACGAACAGGCTCGAAAGGAGGGAGCAGTGCGG;CRE: 5'-CTTGGCTGACGTCAGAGAGA. In supershiftexperiments the following antibodies were added to themixture before the labelled oligonucleotide and incubatedfor 10 ± 20 min: anti-NF-YA mouse monoclonal antibodymAbYA7 or anti-NF-YB rabbit polyclonal antibodypRaYB (Mantovani et al., 1993). Nuclear proteins wereextracted from proliferating or TD C2C12 cells asdescribed (Digna et al., 1983) with the followingmodi®cations: cell pellets were resuspended in a bu�ercontaining 10 mM HEPES (pH 7.9), 1.5 mM MgCl2, 10 mM

KCl and 0.5 mM DTT, leupeptin 10 mg/ml, pepstatin 4 mg/ml, aprotinin 5 mg/ml, 50 mM NaF, 1 mM orthovanadateand 1 mM PMSF. Cells were broken in a Douncehomogenizer, nuclei separated by centrifugation, andcytoplasmic extracts were stored in aliquots at 7808C.Proteins from nuclei were extracted in the following bu�er:20 mM HEPES (pH 7.9), 25% glycerol, 0.42 M NaCl,1.5 mM MgCl2, 0.2 mM EDTA, 0.5 mM DTT. Theproduction of NF-YA, NF-YB, and NF-YC recombinantprotein has been described (Mantovani et al., 1992).

AcknowledgementsThe authors would like to thank Stephen Dalton forhuman cdk1 promoter, and Pidder Yansen-Durr for cyclinA promoter; Patrizia Lavia and Silvia Bacchetti for criticalreading of the manuscript and helpful discussions; MariaPia Gentileschi and Giulio Tibursi for technical advice andoligonucleotide synthesis, and Vincenzo Giusti for comput-ing assistance. This work has been partially supported bygrants from AIRC to GP and RM, Telethon, andMinistero della Sanita'. AF has been a recipient of anAIRC fellowship.


2826

References

Bellorini M, Zemzoumi K, Farina A, Berthelsen J, Piaggio Gand Mantovani R. (1997). Gene, 193, 119 ± 125.

Blau HM, Pavlath GK, Hardeman EC, Chiu CP, SilbersteinL, Webster SG, Miller SC and Webster C. (1985). Science,230, 758 ± 766.

Brandeis M and Hunt T. (1996). EMBO J., 15, 5280 ± 5289.Bucher P. (1990). J. Mol. Biol., 212, 563 ± 578.Buckley MF, Sweeney KJE, Hamilton JA, Sini RL, Manning

DL, Nicholson RI, deFazio A, Watts CKW, MusgroveEA and Sutherland RL. (1993). Oncogene, 8, 2127 ± 2133.

Chang ZF and Liu CJ. (1994). J. Biol. Chem., 269, 17893 ±17898.

Chen C and Okayama H. (1987). Mol. Cell Biol., 7, 2745 ±2752.

Chomczynski P and Sacchi N. (1987). Anal. Bioch., 162,156 ± 159.

Dalton S. (1992). EMBO J., 11, 1797 ± 1804.Desvernie B, Petty KJ and Nikodem VM. (1991). J. Biol.

Chem., 266, 1008 ± 1013.Digna JD, Lebovitz RM and Roeder RG. (1983). Nucleic.

Acids. Res., 11, 1475 ± 1489.Doerner P, Jorgensen JE, You R, Steppuhn J and Lamb C.

(1996). Nature, 380, 520 ± 523.Dorn A, Bollekens J, Staub A, Benoist C and Mathis D.

(1987). Cell, 50, 863 ± 872.Draetta G, Brizuela L, Potashkin J and Beach D. (1987).

Cell, 50, 319 ± 325.Draetta G, Luca F, Westendorf J, Brizuela L, Ruderman J

and Beach D. (1989). Cell, 56, 829 ± 838.Graham FL and Van Der Eb AJ. (1973). Virology, 52, 456 ±

467.Farina A, Gaetano C, Crescenzi M, Puccini F, Manni I,

Sacchi A and Piaggio G. (1996). Oncogene, 13, 1287 ±1296.

Fromm L and Overbeek PA. (1996). Oncogene, 12, 69 ± 75.Hanley-Hyde J, Mushinski JF, Sadofsky M, Huppi K, Krall

M, Kozak CA and Mock B. (1992). Genomics, 13, 1018 ±1030.

Hwang A, Maity A, McKenna WG and Muschel RJ. (1995).J. Biol. Chem., 270, 28419 ± 28424.

Henglein B, Chenivesse X, Wang J, Eick D and Brechot C.(1994). Proc. Natl. Acad. Sci. USA, 91, 5490 ± 5494.

Hooft van Huijsduijnen R, Li XY, Black D, Matthes H,Benoist C and Mathis D. (1990). EMBO J., 9, 3119 ± 3127.

Katula KS, Wright KL, Hyacinth P, Surman DR, NuckollsFJ, Smith JW, Ting JPY, Yates J and Cogswell JP. (1997).Cell Growth Di�., 8, 811 ± 820.

Keyomarsi K and Pardee AB. (1993). Proc. Natl. Acad. Sci.USA, 90, 1112 ± 1116.

Lucibello FC, Truss M, Zwicker J, Ehlert F, Beato M andMuller R. (1995). EMBO J., 14, 132 ± 142.

Mantovani R, Pessara U, Tronche F, Li XY, Knapp AM,Pasquali JL, Benoist C and Mathis D. (1992). EMBO J.,11, 3315 ± 3322.

Mantovani R, Tora L, Moncollin V, Egly JM, Benoist C andMathis D. (1993). Nucleic Acids Res., 21, 4873 ± 4878.

Mantovani R, Li XY, Pessara U, Hooft van Huisjduijnen R,Benoist C and Mathis D. (1994). J. Biol. Chem., 32,20340 ± 20346.

Mantovani R. (1998). Nucleic Acids Res., 26, 1135 ± 1143.Marziali G, Perrotti E, Ilari R, Testa U, Coccia EM and

Battistini A. (1997). Mol. Cell. Biol., 17, 1387 ± 1395.Millar JBA, McGowan CH, Lenaers G, Jones R and Russel

P. (1991). EMBO J., 10, 4301 ± 4309.Murray AW, Solomon MJ and Kirschner MW. (1989).

Nature, 339, 280 ± 285.Ohkubo Y, Kishimoto T, Nakata T, Yasuda H and Endo T.

(1994). Exp. Cell Res., 214, 270 ± 278.Okazaki K and Holtzer H. (1966). Proc. Natl. Acad. Sci.

USA, 56, 1484 ± 1490.Osada S, Yamamoto H, Nishihara T and Imagawa M.

(1996). J. Biol. Chem., 271, 3891 ± 3896.Piaggio G, Farina A, Perrotti D, Manni I, Fuschi P, Sacchi A

and Gaetano C. (1995). Exp. Cell Res., 216, 396 ± 402.Pines J and Hunter T. (1990). New Biol., 2, 389 ± 401.Sambrook J, Fritsch EF and Maniatis T. (1989). Molecular

Cloning. A Laboratory Manual. Cold Spring Harbor, NewYork: Cold Spring Harbor Laboratory.

Sinha S, Maity SN, Lu J and de Crombrugghe B. (1995).Proc. Natl. Acad. Sci. USA, 92, 1624 ± 1628.

Strausfeld U, Labbe' JC, Fesquet D, Cavadore JC, Picard A,Sadhu K, Russel P and DoreÂ e M. (1991). Nature, 351,242 ± 245.

TiainenM, Pajalunga D, Ferrantelli F, Soddu S, Salvatori G,Sacchi A and Crescenzi M. (1996). Cell Growth Di�., 7,1039 ± 1050.

Ya�e D and Saxel O. (1977). Nature, 270, 725 ± 727.Zhao J, Nornes HO and Neuman T. (1995). Neurosci. Lett.,

190, 49 ± 52.Zorbas H, Rein T, Krause A, Ho�mann K and Winnaker

EL. (1992). J. Biol. Chem., 267, 8478 ± 8484.Zwicher J, Lucibello FC, Wolfraim LA, Gross C, Truss M,

Engeland K and Muller R. (1995). EMBO J., 14, 4514 ±4522.


2827

down-regulation of cyclin b1 gene transcription in terminally differentiated skeletal muscle cells...

Documents