Transcriptional Activation of Odf2/Cenexin by Cell Cycle Arrest and the Stress Activated Signaling Pathway (JNK Pathway)

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

provided by Elsevier - Publisher Connector

Biochimica et Biophysica Acta 1833 (2013) 1338–1346

Contents lists available at SciVerse ScienceDirect

Biochimica et Biophysica Acta

journal homepage: www.elsevier.com/locate/bbamcr

Transcriptional activation of Odf2/Cenexin by cell cycle arrest and the stress activated signaling pathway (JNK pathway)

Nadin Pletz 1, Anja Medack 2, Eva Maria Rieß 3, Kefei Yang, Zahra Basir Kazerouni, Daniela Hüber, Sigrid Hoyer-Fender ⁎

Georg-August-Universität Göttingen, Johann-Friedrich-Blumenbach Institute for Zoology and Anthropology, Developmental Biology, GZMB, Justus-von-Liebig-Weg 11, Göttingen, Germany

article info abstract

Article history: The centrosome/basal body protein ODF2/Cenexin is necessary for the formation of the primary cilium. Primary Received 15 October 2012 cilia are essential organelles that sense and transduce environmental signals. Primary cilia are therefore critical Received in revised form 1 February 2013 for embryonic and postnatal development as well as for tissue homeostasis in adulthood. Impaired function of Accepted 21 February 2013 primary cilia causes severe human diseases. ODF2 deﬁciency prevents formation of the primary cilium and is Available online 28 February 2013 embryonically lethal. To explore the regulation of primary cilia formation we analyzed the promoter region of Odf2 and its transcriptional activity. In cycling cells, Odf2 transcription is depressed but becomes up-regulated Keywords: ODF2/Cenexin in quiescent cells. Low transcriptional activity is mediated by sequences located upstream from the basal pro- Primary cilium moter, and neither transcription factors with predicted binding sites in the Odf2 promoter nor Rfx3 or Foxj, Transcriptional regulation which are known to control ciliary gene expression, could activate Odf2 transcription. However, co-expression G0-arrest of either C/EBPα, c-Jun or c-Jun and its regulator MEKK1 enhances Odf2 transcription in cycling cells. Our results C/EBPα provide the ﬁrst analysis of transcriptional regulation of a ciliary gene. Furthermore, we suggest that transcrip- Stress signaling pathway tion of even more ciliary genes is largely inhibited in cycling cells but could be activated by cell cycle arrest and by the stress signaling JNK pathway. © 2013 Elsevier B.V. All rights reserved.

1. Introduction microtubules [11,12]. Moreover, as shown in F9 cells formation of primary cilia essentially depends on the presence of ODF2 [13]. ODF2 is an essential component of vertebrate cells. Homozygous Primary cilia are solitary organelles found on virtually all verte- mice with a gene trap insertion in exon 9 of the Odf2 gene that results brate cells [14]. They protrude from the cell surface and execute sen- in truncation of the protein are embryonic lethal [1]. A functional sory functions. For instance, the hedgehog pathway essentially ODF2 protein is hence crucial for embryonic development to proceed. depends on primary cilia [15–21]. Primary cilia are therefore pivotal ODF2 was first identified as a major component of the outer dense for embryonic and postnatal development and tissue homeostasis in fibers of the mammalian sperm tail, which are prominent cytoskeletal adulthood, and their disturbance causes severe human diseases, the structures [2–5]. In somatic cells, ODF2 is an essential component of so-called ciliopathies [22–25]. Each cilium is anchored in the cell by the centrosome and the basal body [6–9]. Due to extensive alternative its basal body. The basal body locates at the cell surface underneath splicing diverse protein isoforms of ODF2 exist [2]. The protein the cell membrane. Formation of the ciliary shaft is initiated at the isoform Cenexin is characterized by a 42 amino acid insertion in the distal end of the basal body, which itself originates from the mother N-terminal region that is largely encoded by exon 3b. This “Cenexin centriole. Basal bodies and centrioles are thus structurally related insertion” is necessary for targeting ODF2/Cenexin to the centrosome and can be transformed into each other. Centrioles are generally and the basal body [10,11]. Centrosomes, basal bodies and primary found in a pairwise association in cycling animal cells. They are cilia are microtubule-based structures, and ODF2 associates with surrounded by a halo of pericentriolar material (PCM) to which γ-tubulin complexes are recruited [26–28]. A pair of centrioles with its associated PCM constitutes the centrosome that is the main micro-

⁎ Corresponding author. tubule organizing center (MTOC) of animal cells. In mitotic cells the E-mail address: [email protected] (S. Hoyer-Fender). centrosome plays a central role in organizing the mitotic spindle to 1 Present address: Georg-August-Universität Göttingen, Universitätsmedizin, Department separate chromosomes [29]. The centrosome and its pair of centrioles of Dermatology, Venereology and Allergology, Robert-Koch-Str. 40, Göttingen, Germany. are duplicated once in a single cell cycle. Duplication occurs syn- 2 Present address: University of Lübeck, Medizinische Klinik II, Cardiovasculare Genetics, chronously with the cell cycle and starts with the formation of Ratzeburger Allee 160, Lübeck, Germany. 3 Present address: Max Planck Institute for Biophysical Chemistry, Am Fassberg, procentrioles at the proximal ends of both centrioles of a centrosome Göttingen, Germany. in S-phase [30–32]. Elongation of procentrioles and assembly of PCM

0167-4889/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.bbamcr.2013.02.023 N. Pletz et al. / Biochimica et Biophysica Acta 1833 (2013) 1338–1346 1339 during G2-phase of the cell cycle eventually resulted in two centro- containing 0.15% bovine serum albumin, 0.5% Triton X-100), incubated somes each consisting of a pair of centrioles and its associated PCM. with the primary antibody anti-ODF2 [2] and anti-γ-tubulin (GTU 88; Based on this conservative centriole duplication each centrosome Sigma-Aldrich, St. Louis, USA), respectively. For detection of the primary consists of one older centriole generated at least one cell cycle before, cilium cells were fixed in 3.7% paraformaldehyde in PBS and incubated which is the mature or mother centriole, and one younger centriole, with anti-acetylated-tubulin antibody (clone 6-11B-1; Sigma-Aldrich, the daughter centriole. Mother and daughter centrioles differ not St. Louis, USA). Primary antibodies were detected with the secondary only in age but also in structure, protein composition, and function. antibodies anti-rabbit or anti-mouse IgG linked to Alexa 488 or The mother centriole is characterized by the presence of distal and MFP590 (Molecular Probes, Eugene). DNA was counterstained with subdistal appendages, and it is the mother centriole exclusively that DAPI. Images were taken by confocal microscopy (LSM 510, Zeiss), and can be transformed into a basal body to grow a cilium. Both append- processed using Adobe Photoshop 5.0. ages are therefore present on the mother centriole as well as on the basal body (reviewed in: [33]). ODF2/Cenexin specifically locates to 2.2. Cell culture and synchronization the subdistal appendages and initiation of ciliogenesis depends on a crucial amount of ODF2 present [7,34]. For synchronization at G0 an exponentially growing cell culture Although primary cilia have already been found in proliferating was kept in 10% serum-containing medium for 24 h after they cells the percentage of ciliated cells increases in G0-phase arrested reached confluence, followed by serum starvation (0.5% FCS in cells [14,35–40]. Therefore, medium with reduced serum concen- DMEM) for 48 h. For cell cycle arrest at G2/M transition NIH3T3 tration (serum starvation medium) is generally used to induce cells were grown until they reached confluence followed by treat- ciliogenesis in cultured cells [39]. However, it remains unclear whether ment with 400 ng/ml nocodazole for ~16 h. For G1/S boundary arrest serum reduction induces ciliogenesis by driving the cells into G0-phase a sub-confluent G0 arrested culture was generated first. Approxi- or if deprivation of inhibitory factors causes ciliogenesis. Though abro- mately 18 h after seeding cells were washed twice in PBS and once gation of signaling pathways may be involved in ciliogenesis since inhi- in DMEM containing 0.5% FCS to remove loosely attached, floating bition of PI3K- and MAPK-pathways promotes primary cilia formation cells and residual medium. Cells were then incubated in DMEM in some cell lines [41]. In mammals, two transcription factor (TF) fami- supplemented with 0.5% FCS for 24 h. Subsequently, the medium lies are known to be important players in controlling ciliary gene was changed to DMEM complete containing 5 μg/ml aphidicolin expression. These are the FOXJ1 (forkhead box J1) and RFX (regulatory and cells incubated for 16 h. Cell synchronization was monitored factor X) TF families that fulfill complementary and synergistic by flow-cytometry analysis of propidium iodide-stained cells (FACS functions in regulating ciliary gene expression (for review see: [42]). analysis). Whereas FOXJ1 regulates genes specific for motile cilia or necessary for basal body apical transport, RFX TFs are essential for the assembly 2.3. Real time RT-PCR of primary and motile cilia in metazoans by regulating genes involved in intraflagellar transport [43–48]. RFX3 specifically is necessary for Total RNA was prepared using peqGOLD RNAPure™ (PEQLAB, the growth of primary cilia in the embryonic node and in the endocrine Erlangen, Germany) and digested with RQ1 DNase. cDNA was gener- pancreas [49–51]. ated from 0.5 μg of total RNA using RevertAid™ H Minus First Strand

To explore the regulation of primary cilia formation we analyzed cDNA Synthesis Kit (Fermentas) and oligo(dT)18 primer. Real time PCR the promoter region of Odf2 and its transcriptional activity. Transcrip- was performed on iCycler IQ PCR System (BioRad) with iQ SYBR Green tion of Odf2 in cycling cells is generally very low. But we found by Supermix (BioRad) using cloned Odf2-andHprt-fragments, respective- qRT-PCR and reporter gene assays that transcription of Odf2 is ly, as standards for quantification. Primer sequences for qRT-PCR analy- up-regulated in quiescent cells. We additionally demonstrate that sis are as follows: HPRT (mHPRT-for AGCCCCAAAATGGTTAAGGTTGC ~2 kb of the promoter region is sufficient to drive transcription of and mHPRT-rev TTGCAGATTCAACTTGCGCTCAT), Odf2 (Odf2-for ACCA Odf2 in the same manner as the endogenous gene. Low transcrip- TGAAGGACCGCTCTTC and Odf2-rev CGCACATTCACAGTGTCCCC). The tional activity in cycling cells is mediated by sequences located up- specificity of the SYBR Green assay was verified by melting curve anal- stream of the basal promoter and omitting these sequences strongly yses. Overall, at least three independent qRT-PCR experiments were increased promoter activity. Although several transcription factor performed, and each measurement was done in triplicate. The ratio be- binding sites have been predicted in the Odf2 promoter we could tween the amount of Odf2 transcripts and the amount of Hprt tran- not observe activation of the promoter by those TFs. Additionally, scripts in each cell cycle phase was calculated, and the value obtained transcription factors RFX3 and FOXJ that are known to regulate ciliary for the G0-phase was set as the base level (=1). genes did not affect Odf2 transcription. However, co-expression of either C/EBPα, c-Jun or c-Jun and its regulator MEKK enhances Odf2 2.4. Generation of Odf2 promoter constructs transcription in cycling cells. Our results thus suggest that transcription of even more ciliary genes is inhibited in cycling cells but could A fragment of about 2.2 kb, comprising 1.8 kb upstream of the tran- be activated by cell cycle arrest and by the stress signaling JNK scriptional start site and 358 bp of the transcribed region, was generat- pathway. ed by EcoRI/ApaI double digestion of a genomic mouse Odf2 clone (isolated from strain ola/129 and obtained from the Resource Centre 2. Materials and methods of the German Human Genome Project at the Max-Planck-Institute for Molecular Genetics, Berlin) and cloned into pBluescript. The 3′ 2.1. Immunocytochemistry end of the fragment was then gradually shortened by ExoIII deletion. Truncated promoter fragments were isolated by SacI/KpnI double NIH3T3 and HEK293 cells were cultured in Dulbecco's modified digestion and directionally cloned into pGL3 basic (Promega, Madison, Eagle's medium (DMEM) supplemented with 10% fetal calf serum USA) yielding clones 2.2 (−1805/+358), 7.1 (−1805/−94), 7.6 (FCS), 20 mM L-glutamine, 1000 U/ml penicillin, and 1000 μg/ml (−1805/−1282), and 22.1 (−1805/−1368). Clones 1.5 (−797/+1), streptomycin at 5% CO2 and 37 °C. Cells were grown on cover slips and #1 (−1675/+1), A1 (−797/+358) and 0.5 (−94/+255) were gener- optionally transfected using Transfectin reagent (BioRad) or Xtreme ated by PCR using the genomic clone as template and primer pairs p1 HP (Roche). For immunological detection of endogenous ODF2 or of (cgagctcgcaagcagcagagatag)/p2 (tcccccgggccatttaggcccagcccccag), p3 γ-tubulin cells were fixed in methanol and, after blocking unspecific (ctgtcgaccagaagggcagctaaggg)/p2, T3/Odf2PromSacI (cgagctcgcaagcagc binding sites in PBT (phosphate buffered saline, PBS, at pH 7.5, agagatag), and Odf2-94forSacI (Cgagctcgggcggtggacaggcctcg)/05Smarev 1340 N. Pletz et al. / Biochimica et Biophysica Acta 1833 (2013) 1338–1346

(atcccgggcgcccaacgaggagcgg), respectively. Finally, fragments were 1,2 cloned directionally into pGL3basic. All promoter constructs were validated by sequencing. The Odf2 promoter region of clone #1 (−1675/ 1 +1) was subcloned into promoter-less pEGFP-1 (Clontech, Palo Alto, USA), and the EGFP coding region exchanged by the open reading 0,8 frame of 13.8NC fused to EGFP [11]. Afterwards, the coding region of EGFP was exchanged by the coding region of destabilized EGFP 0,6 from vector pd2EGFP-N1 (Clontech, Palo Alto, USA) yielding clone #1-13.8NC-d2EGFP. Finally, the promoter region was exchanged by fold change 0,4 the promoter region of clone 2.2 that was ampliﬁed using primer pair 2.2f (gcactcgagtgagattatagctatgtac)/2.2r (cagaattccccaacgaggagcgggg) 0,2 yielding clone 2.2–13.8NC-d2EGFP. G0 G1/S G2/M 0

2.5. Promoter analysis Fig. 1. Real-time RT-PCR of Odf2 in synchronized NIH3T3 cells. Odf2 and Hprt transcript levels were quantified by real-time RT-PCR and the Odf2–Hprt-transcript ratio calculated. The activity of promoter fragments was measured using the firefly The value obtained for G0-arrested cells was set as the base level (=1), and the ratio luciferase encoded by pGL3 basic as reporter. Each promoter construct obtained for G1/S or G2/M arrested cells quoted as fold change. Real time RT-PCR revealed was co-transfected into NIH3T3 or HEK293 cells together with that transcription of Odf2 is highest in G0-arrested cells. phRL-SV40 (Promega, Madison, USA), which was used as internal control, and the luminescence measured 24 h later (Dual-Glo Luciferase of ODF2 to the basal body in G0-phase arrested cells has already been Assay System, Promega, USA) using Centro LB 960 luminometer demonstrated [9]. Co-staining for γ-tubulin validated the centrosome (Berthold Technologies, Germany). The ratio between the lumines- (Fig. 2 in green). cence of firefly luciferase and the luminescence of Renilla luciferase (encoded by phRL-SV40) was calculated resulting in relative lumines- 3.2. Repression of the Odf2 promoter in cycling cells cence. Each experiment was performed in triplicate and repeated at least 3 times. Expression vector constructs for E2F1-5, Sp1, Extensive alternative splicing of Odf2 transcripts resulted in vari- Sp3, c-Myc, RFX3, FOXJ1, FOXO3A, C/EBPα, Fos were kindly provided ous isoforms including ODF2 and Cenexin [3,10]. However, tran- by Steven Brody (St. Louis), Matthias Dobbelstein (Göttingen), scripts of all isoforms start either in exon 1 or in exon 2a (Fig. 3). Kristian Helin (Copenhagen), Walter Reith (Genf), and Guntram Specifically, all somatic transcripts isolated from mouse NIH3T3 cells Suske (Marburg), or were from addgene (Addgene plasmids 10914, including the isoform Cenexin (and clone DH13.8) start in exon 1 24213, 12550, 8966). Expression vector plasmids for c-Jun (pFA2- [10]. The expected promoter region therefore seems to be upstream c-Jun, aa 1–223) and MEKK1 (MAP3K1; pFC-MEKK, aa 380–672) of exon 2a and exon 1. were from Stratagene (PathDetect c-Jun trans-Reporting System; In silico analyses of Odf2 sequences upstream of the transcrip- Stratagene, La Jolla, USA). MEK inhibitor U0126 (Enzo Life Sciences, tional start site in exon 1 using ECR (http://ecrbrowser.dcode.org/), Germany) was dissolved in ethanol (final concentration 2 mg/ml), TFSearch (http://www.cbrc.jp/research/db/TFSEARCH.html), and and used as 100 μM final concentration in reporter gene assays. ENCODE (http://genome.ucsc.edu/ENCODE/) predicted a huge number of putative transcription factor binding sites including E2F, 3. Results c-Myc, Sp1, and AP-1. In order to characterize the promoter region in more detail, we generated a fragment of 2163 base pairs (bp) 3.1. Up-regulation of Odf2 transcription in G0-phase that encompasses 1805 bp upstream of the transcriptional start site (defined as the transcriptional start of Odf2/1; Fig. 3) in addition to Exploration of cell-cycle dependent transcription of Odf2 in the 358 bp downstream including the first exon and the first intron mouse fibroblast cell line NIH3T3 was performed by qRT-PCR on (clone 2.2). Next, this upstream region was subdivided into several total RNA isolated from synchronized cells. NIH3T3 cells were syn- shorter fragments yielding clones 7.1 (−1805/−94), 7.6 (−1805/ chronized at G0-phase by serum starvation for 48 h, at G2/M-phase −1282), 22.1 (−1805/−1368), 1.5 (−797/+1), #1 (−1675/+1), A1 of the cell cycle by nocodazole treatment for 16 h, and at G1/S bound- (−797/+358), and 0.5 (−94/+255). Only clones A1 and 0.5 contained ary by 16 h aphidicolin treatment of a G0-arrested cell culture. FACS the transcribed exon 1 sequences (Fig. 4). All promoter fragments analyses of propidium iodide-stained cells confirmed successful were cloned directionally into promoter-less pGL3 basic and verified by synchronization at G1/S- or G2/M-phase (Fig. S1). G0 arrest was sequencing. confirmed immunocytologically by the incidence of a primary cilium Promoter constructs were co-transfected with phRL-SV40, which (not shown). Quantification of Odf2 transcripts was performed by encodes Renilla luciferase driven by the SV40 promoter, into NIH3T3 real-time RT-PCR (qRT-PCR) using cloned Odf2- and Hprt-fragments, or HEK293 cells, and luciferase activities measured 24 h later. The respectively, as standards. The ratio between the amount of Odf2 ratios of firefly luciferase activity to Renilla luciferase activity were transcripts and the amount of Hprt transcripts in each cell cycle calculated resulting in the relative luminescence. Promoter constructs phase was calculated, and the value obtained for the G0-phase was 7.1 (−1805/−94), 7.6 (−1805/−1282), and 22.1 (−1805/−1368) set as the base level (fold change = 1). Our results showed compara- revealed almost no activity at all whereas promoter regions 1.5 ble Odf2 transcript levels in G1/S boundary synchronized cells, and in (−797/+1) and #1 (−1675/+1) showed increased but much lower G2/M-phase arrested cells (Fig. 1). However, in cells arrested in the activity than 2.2 (−1805/+358) (Fig. 5A). Since promoter construct G0-phase the Odf2 transcript level is approximately twofold increased #1 (−1675/+1) is nearly identical to construct 7.1 (−1805/−94) compared to G1/S-phase and G2/M-phase arrested cells (Fig. 1). except for 130 bp missing at the most distal end and the presence of The endogenous ODF2 protein could be detected throughout the 94 bp immediately upstream of the transcription start site but other- cell cycle. In G1-phase one single centrosome is present stressed by wise showed a much higher transcriptional activity than construct a single ODF2 positive dot. In G2, the centrosome has already been 7.1, we expected the basal promoter immediately upstream of the tran- duplicated demonstrated by two ODF2 positive dots slightly apart scriptional start in exon 1. Construct 1.5 (−797/+1) comprises 797 bp from each other, and in mitosis ODF2 locates to the spindle poles, immediately upstream of the transcriptional start site but misses which are generated by the centrosome (Fig. 2 in red). The localization 878 bp more distally that are otherwise present in #1. Since promoter N. Pletz et al. / Biochimica et Biophysica Acta 1833 (2013) 1338–1346 1341

Fig. 2. Immunological detection of endogenous ODF2 during the NIH3T3 cell cycle. Anti ODF2 antibody (in red) and anti-γ-tubulin antibody (in green) highlight the centrosome in G1 and G2 cells, and the spindle pole in metaphase (in red). Bars are of 5 μmor10μm.

1.5 showed higher transcriptional activity than #1, we anticipated the the activity of the Odf2 promoter in each cell cycle phase was related. presence of repressor binding sites from −797 bp upstream. However, Our data showed that the Odf2 promoter construct 2.2 has a similar promoter constructs A1 (−797/+358) and 0.5 (−94/+255) have a cell cycle dependent activation profile as the endogenous Odf2 gene much stronger transcriptional activity than 2.2 (−1805/+358) (fold with highest transcriptional activation during G0-phase (Fig. 6). Cyto- change = 1) with 0.5 revealing an approximate 2000 fold increase logically Odf2 promoter (2.2) driven expression of ODF2 fused to and A1 a 108 fold increase (Fig. 5B). Therefore, sequences upstream destabilized GFP (construct 2.2–13.8NC-d2EGFP) could be detected from the basal promoter seem to repress Odf2 transcription. at the centrosome in G1/S-phase synchronized cells as well as at the basal body and the primary cilium in G0-phase arrested cells 3.3. Transcriptional activity of the Odf2 promoter 2.2 during the cell cycle (Fig. 7). The centrosome was visualized by anti-γ-tubulin staining, and the basal body and the primary cilium by detection of acetylated We then investigated whether 1.8 kb upstream of the transcrip- tubulin (Fig. 7). tional start site are sufficient to regulate transcription of Odf2 cell cycle dependent. The reporter vector encoding firefly luciferase driv- 3.4. Transcriptional activation of Odf2 by C/EBPα and by the stress en by the Odf2 promoter 2.2 (−1805/+358) was co-transfected with response signaling pathway (JNK) the control vector phRL-SV40 into synchronized NIH3T3 cells. Lucifer- ase activities were measured 24 h post-transfection and the relative Although binding sites of several transcription factors have been luminescences calculated. As internal reference we co-transfected predicted in the Odf2 promoter by in silico search most TFs did not the vector pGL3 control, in which transcription of firefly luciferase is activate the promoter. Luciferase assays were performed using Odf2 under the control of the SV40 promoter, together with phRL-SV40 promoter construct 2.2 (−1805/+358) co-transfected with expres- into synchronized cells. This experiment served as control, to which sion vectors encoding either E2F1, E2F2, E2F3, c-Myc, Sp1, Sp3 1342 N. Pletz et al. / Biochimica et Biophysica Acta 1833 (2013) 1338–1346

A 70,0

60,0 2.2

50,0

40,0 1.5

30,0

rel. luminescencerel. 20,0 #1

10,0 22.1

7.1 7.6 0,0 B Fig. 3. Transcriptional start sites of Odf2. The exon–intron structure of the N-terminal 2500 region of the Odf2 gene from Mus musculus is shown on top. Translation start codons are asterisked. For the testicular Odf2/1 as well as for all transcripts isolated from mouse 1978 NIH3T3 cells (DH13.8 and “Cenexin”) transcription starts in exon 1. The sole exception 2000 up to now is the transcript of testicular Odf2/2 that starts in exon 2a. Transcription starts are marked by arrows. The term “Cenexin” is used here to indicate those transcripts comprising exon 3b encoding the Cenexin-insertion [10]. 1500

(Fig. S2), E2F4, E2F5, FOXO3A (not shown), RFX3 or FOXJ1 (Fig. S3) 1000

but none activated the Odf2 promoter. However, our results have change fold shown that the endogenous Odf2 promoter as well as the reporter construct 2.2 are both transcriptionally activated in quiescent cells. 500 Moreover, two AP-1 binding sites were predicted in the Odf2 pro- 108 moter, one in the basal promoter region up to −94 bp and a second 1 at approximately −1330 bp (Fig. 4 asterisks). Reporter gene assays 0 2. 2 A1 0.5 performed in cycling cells using the Odf2 promoter construct 2.2 revealed that co-expression of MEKK1 (MAP3K1) did not activate Fig. 5. Activity of Odf2 promoter constructs in exponentially growing NIH3T3 cells. the promoter whereas co-expression of c-Jun resulted in a transcrip- A) Activity of Odf2 promoter constructs 2.2 (−1805/+358), 7.1 (−1805/−94), 7.6 tional increase of ~50%. Furthermore, co-expression of both MEKK1 (−1805/−1282), 22.1 (−1805/−1368), 1.5 (−797/+1), and #1 (−1675/+1) in − − and c-Jun strongly increased the promoter activity (~2 fold) whereas NIH3T3 cells. B) Activity of Odf2 promoter constructs 2.2 ( 1805/+358), A1 ( 797/ +358), and 0.5 (−94/+255). Transcription of the reporter gene firefly luciferase was co-expression of c-Fos had no obvious impact on promoter activity. driven by Odf2 promoter constructs. Constructs were transfected into NIH3T3 cells in con- Addition of the small molecule inhibitor U0126 reduced the Odf2 pro- junction with the control vector phRL-SV40. Luciferase activities were measured 24 h later. moter activity to less than 50%. Besides we identified C/EBPα as a The ratios firefly luciferase activity/Renilla luciferase activity were calculated resulting in transcriptional activator of Odf2 (Fig. 8). the relative luminescence. Standard deviations are included.

4. Discussion

ODF2/Cenexin is a centrosome/basal body protein in vertebrates [9] and locates at the centrosome throughout the cell cycle (Fig. 2). Although its transcriptional activity in cycling cells is very low, quan- titative RT-PCR revealed that Odf2 is twofold up-regulated in cultured mouse fibroblast cells that are driven into G0-phase by serum starvation compared to G1/S-phase or G2/M-phase arrested cells. There is essentially no difference in transcription level of Odf2 in G1/S- and G2/M-phase arrested cells. So far, there is only one report demonstrating cell cycle dependent transcription of Odf2 and its up-regulation at G2/M-phase [52]. Since Whitfield et al. analyzed the periodically transcription in the HeLa cell cycle (by exclusion of the G0-phase) up-regulation of Odf2 at G2/M-phase could be specific for this human tumor cell line. Analyses of the promoter region of the mouse Odf2 gene demonstrated that approximately 1.8 kb upstream of the transcriptional start site along with the first exon and the first intron of the Odf2 gene Fig. 4. Promoter constructs used for reporter gene assays. Construct 2.2 comprises (promoter construct 2.2) is sufficient to drive reporter gene transcrip- 1805 bp upstream of exon 1 in addition to exon 1 (E1) and intron 1 (I1). All other con- tion essentially identical to the endogenous Odf2 with about twofold structs (7.1, 7.6, 22.1, 1.5, #1) are truncated versions of the upstream region neither up-regulation in G0-phase. Reporter gene assays, however, additionally including exon 1 nor intron 1. Only constructs A1 and 0.5 contain exon 1 and parts revealed that transcription is about twice in G2/M-phase arrested of intron 1 sequences. The numerical legend denotes nucleotides upstream (−)or downstream (+) from the transcriptional start site (+1) in exon 1 of Odf2/1. Predicted cells compared to those in G1/S-phase. This result therefore supported AP-1 binding sites are marked by asterisks. the data of Whitfield [52] that showed G2/M-phase dependent N. Pletz et al. / Biochimica et Biophysica Acta 1833 (2013) 1338–1346 1343

3,5 one AP1-site (score 85.6), one CdxA site (score 89.7), and two Sp1 sites (with scores of 89.0 and of 94.5) using TFSEARCH (ver. 1.3). 3 In silico analyses of extended human core promoter regions covering [−70, +60] segment relative to the transcription start site have 2,5 shown that the most common promoter elements found include E2F-1, AP-2, and CdxA among others [53]. Moreover, using 2 TFSEARCH (ver. 1.3) we found two predicted E2F binding sites, one in intron 1, and the other upstream of −94 (both with scores 1,5 of 86.2). One putative c-Myc binding site (score 86.8) was found at

fold change −1368 to −1675. Additionally, further Sp1 binding sites were pre- 1 dicted: three in intron 1 with scores of 87.7 and of 90.4, one in exon 1 (score of 91.8), and two upstream of −94 with scores of 87.7 and 0,5 86.3. A putative TATA box with a score of 87.1 was found upstream of the predicted basal promoter [−1to −94] at position −422 to −437. E2F transcription factors are essential for the initiation of 0 S-phase in regulating the activation of genes whose products are im- G0 plicated in cell proliferation and DNA replication [54]. The E2F family G1/S G2/M members are divided into two subgroups based on the fact that E2F1, control E2F2, and E2F3 are all capable of inducing S-phase in serum-starved Fig. 6. Cell cycle dependent activity of Odf2 promoter 2.2. NIH3T3 cells were synchro- fibroblasts, but E2F4 and E2F5 are not [54,55]. It has also been demon- nized to yield G0-, G1/S-, and G2/M-phase cells. 24 h prior to luminescence measure- strated that centrosome duplication requires the activation of E2F − fi fl ment cells were co-transfected with Odf2 promoter 2.2 ( 1805/+358) driving re y transcription factors and Cdk2-cyclin A activity [56]. Since ODF2 is a luciferase, and phRL-SV40 as internal control. As control for general promoter activity, synchronized cells were co-transfected with the control vector pGL3 basic (firefly lucifer- component of centrosomes its transcriptional activation by E2F TFs ase under the control of SV40 promoter) and phRL-SV40 (control). The relative lumines- could be envisioned. C-Myc is necessary to induce and maintain cell cence was calculated and related to the relative luminescence of the control experiment in proliferation (for review see: [57]), and therefore might also affect each individual cell cycle phase that was set as 1 (100%). The relative luminescence of Odf2 the transcription of genes encoding centrosomal proteins. Transcrip- promoter activity was then calculated as fold change related to the control experiment in the same cell cycle phase. Standard deviations are included. tion factor Sp1 binds selectively to GC-rich sequences, known as the “GC box”, that are recurring motifs in promoters of mammalian genes, and regulates the expression of a vast number of genes. Sp3 up-regulation of endogenous Odf2 in HeLa cells. The generally very is one of several closely related, but distinct, factors [58]. Sp1 has low transcription rate of endogenous Odf2 in NIH3T3 cells thus might been described as a transcriptional activator as well as a represssor hamper the detection of a periodically transcriptional increase by depending on the promoter it binds to and its binding partners [59]. qRT-PCR. The low transcriptional activity of the Odf2 promoter was Overexpression or overactivation of Sp1 has been reported for a also prominent in the reporter gene assay. Compared to the control vec- number of cancers suggesting that Sp1 activity plays a critical role tor pGL3 control (Promega), in which the reporter firefly luciferase is in tumor formation or in late stage carcinogenesis. For example, Sp1 under the control of the SV40 promoter, the activity of firefly luciferase activity is increased in breast carcinoma that is also characterized driven by the Odf2 promoter 2.2 (−1805/+358) is roughly of one-third by overexpression of Odf2 ([60,61]; http://www.oncomine.org/). only (not shown). The Odf2 promoter 2.2 furthermore drives the However, reporter gene assays using Odf2 promoter construct 2.2 expression of an ODF2 fusion to destabilized EGFP (Odf2 promoter (−1805/+358) demonstrated that neither anyone of these TFs nor 2.2–13.8NC-d2GFP). The fusion protein was detected in G1/S-phase RFX3 or FOXJ could activate the promoter in cycling cells. Since neg- cells at the centrosome, and in G0-phase cells at the basal body and ative regulation by E2F transcription factors has also been reported the primary cilium. Location to the primary cilium might be caused by [54,62,63], the possibility that E2F TFs might be involved in the tran- either overexpression of the fusion protein or by improved detection scriptional repression of the Odf2 promoter has to be considered. Our since endogenous ODF2 could not be found in the primary cilium reporter gene assays indeed indicate a repressive function of c-Myc, (unpublished). However, we did not found the fusion protein in E2F and Sp TFs on Odf2 promoter activity (supplementary Fig. S2). G2/M-phase cells. Since the destabilized EGFP contains a PEST degrada- Although transcriptional activity of Odf2 is generally low in cycling tion domain at its C-terminus the fusion protein is rapidly turned over cells, we found significant up-regulation in quiescent mouse fibro- in mammalian cell culture with a half-life of 2 h only (Promega). Detec- blasts that have been driven into G0-phase by serum starvation. tion of the fusion protein thus points to a continuous translation in Serum starvation is widely used to induce the formation of primary G1/S-, and G0-phase cells. The absence of the fusion protein in cilia [14]. Therefore, the transcriptional activation of genes involved G2/M-phase cells however might indicate a delay in translation and in primary cilia formation, as e.g. Odf2, in G0-phase fits very well. hence translation regulation during G2/M-phase. The requirement of serum reduction to induce transcription of Odf2, Reporter gene assays using Odf2 promoter deletion constructs furthermore suggested that either repressors are then no longer pres- revealed that the basal promoter is immediately upstream of the ent at the Odf2 promoter, or their function is overcome by activators. transcriptional start in exon 1 (up to −94) and further upstream By co-expression of either c-Jun or c-Jun and MEKK1 in cycling cells sequences repress the basal promoter activity. Additionally, the con- we found that the Odf2 promoter becomes strongly activated (up to struct comprising 1.8 kb of the upstream region in conjunction with 2 fold). C-Jun is a component of the activating protein-1 family exon 1 and intron 1 (Odf2 promoter 2.2) has all necessary regulatory (AP-1) of transcription factors that controls the expression of numer- sequences to drive expression essentially identical as the endogenous ous target genes. The transcriptional activity of c-Jun is stimulated promoter. Promoter constructs that missed exon 1 and intron 1 se- by c-Jun N-terminal kinases (JNK) mediated phosphorylation. The quences (Odf2 promoter constructs 1.5 and #1) showed a reduction most potent JNK-activating kinase is MEKK1 that itself is activated in transcriptional activity to about 60% and 30%, respectively, com- by several stress signals [64] . We also observed that overexpression pared to construct 2.2. Exon 1 and intron 1 sequences therefore com- of MEKK1 without concurrent overexpression of c-Jun strongly prise important regulatory sequences for transcriptional activation. repressed the Odf2 promoter indicating the activation of transcrip- The putative basal Odf2 promoter in between −1 and −94 has tional repressors. However, the significant up-regulation of Odf2 tran- two predicted USF binding sites (with scores of 85.0 and of 86.8), scription by MEKK1/c-Jun overexpression suggested that activation of 1344 N. Pletz et al. / Biochimica et Biophysica Acta 1833 (2013) 1338–1346

Fig. 7. Expression of ODF2 fused to destabilized GFP driven by the Odf2 promoter 2.2 (construct 2.2–13.8NC-d2GFP) in synchronized NIH3T3 cells. The fusion protein was visible by GFP auto-fluorescence (in green). The centrosome was detected by anti-γ-tubulin staining (in red), whereas the basal body and the primary cilium are detected by anti- acetylated-tubulin staining (in red). Nuclear counterstaining with DAPI (in blue). Overlay of all images (merge). the stress signaling pathway is involved in the expression of ciliary via MEK1/2 and ERK [65,66]. However, overexpression of c-Fos obvi- genes. ously did not affect Odf2 transcription suggesting that it is not Another component of the AP-1 family of transcription factors is involved in AP-1 mediated transcriptional activation. Inhibition of ERK c-Fos. c-Fos is activated by ERK-mediated phosphorylation. The extra- pathway by the potent MEK1/2 inhibitor UO126 significantly down cellular signal-regulated kinase (ERK) pathway is stimulated by regulated Odf2 transcription in cycling cells indicating the requirement serum and growth factors leading to phosphorylation of substrates of serum induced ERK pathway activation for the maintenance Odf2 expression in proliferating cells. 3 Furthermore, we found the CCAAT/enhancer binding protein C/EBPα to activate Odf2 transcription. The C/EBP family of transcription *** *** 2,5 factors has key roles in controlling cellular proliferation and differentiation. C/EBPα specifically seems to have antiproliferative effects [67–69]. Our results thus suggest that C/EBPα not only promotes cell cycle exit 2 but additionally supports ciliogenesis by activation of ciliary genes. *** *** * In summary, we have identified C/EBPα and the JNK stress response 1,5 pathway as transcriptional activators of Odf2 suggesting, more generally, their involvement in transcriptional control of cilia formation.

fold change 1 *** *** ** Acknowledgement 0,5 We thank our colleagues Ralf Dressel (Göttingen) for FACS analyses, Steven Brody (St. Louis), Matthias Dobbelstein (Göttingen), Kristian 0 Helin (Copenhagen), Walter Reith (Genf), and Guntram Suske (Marburg) for the kind gift of expression plasmids encoding transcription factors Foxj1, c-Myc, E2F1–3, RFX3, and Sp1 and Sp3. This work was sup- 2.2 + Fos

2.2 + cJun ported by a grant of the Deutsche Forschungsgemeinschaft to SHF 2.2 control 2.2 + MEKK 2.2 + UO126 2.2 + C/EBP α (Ho 1440/8-1). 2.2 + cJun Fos 2.2 + Fos MEKK 2.2 + cJun MEKK Appendix A. Supplementary data 2.2 + cJun Fos +MEKK Supplementary data to this article can be found online at http:// α Fig. 8. Transcriptional activation of Odf2 by C/EBP and by the stress response pathway dx.doi.org/10.1016/j.bbamcr.2013.02.023. (JNK). Co-transfection of expression vectors for C/EBPα, c-jun, Fos or/and MEKK, the Odf2 reporter vector (2.2pGL3), and phRL-SV40 into cycling. UO126 was used to inhibit MEK1/2. The ratios between luminescences of 2.2pGL3 and phRL-SV40 were calculated References to obtain the relative luminescences. The values obtained were related to the relative luminescence of the control experiment (exclusively 2.2pGL3/phRL-SV40) which was [1] N.A. Salmon, R.A. Reijo Pera, E.Y. Xu, A gene trap knockout of the abundant sperm set as 1. Standard deviations are shown. For calculation of statistical signiﬁcances the tail protein, outer dense ﬁber 2, results in preimplantation lethality, Genesis 44 paired two sided t-test was used (*p b 0.05; **p b 0.01; ***p b 0.005). (2006) 515–522. N. Pletz et al. / Biochimica et Biophysica Acta 1833 (2013) 1338–1346 1345

[2] H. Brohmann, S. Pinnecke, S. Hoyer-Fender, Identification and characterization of [37] Y. Mori, H. Akedo, Y. Tanigaki, K. Tanaka, M. Okada, Ciliogenesis in tissue-cultured new cDNAs encoding outer dense fiber proteins of rat sperm, J. Biol. Chem. 272 cells by the increased density of cell population, Exp. Cell Res. 120 (1979) (1997) 10327–10332. 435–439. [3] S. Hoyer-Fender, C. Petersen, H. Brohmann, K. Rhee, D.J. Wolgemuth, Mouse Odf2 [38] C.L. Rieder, C.G. Jensen, L.C.W. Jensen, The resorption of primary cilia during cDNAs consist of evolutionary conserved as well as highly variable sequences and en- mitosis in a vertebrate (PtK1) cell line, J. Ultrastruct. Res. 68 (1979) 173–185. code outer dense fiber proteins of the sperm tail, Mol. Reprod. Dev. 51 (1998) 167–175. [39] R.W. Tucker, A.B. Pardee, K. Fujiwara, Centriole ciliation is related to quiescence [4] C. Petersen, L. Füzesi, S. Hoyer-Fender, Outer dense fibre proteins from human and DNA synthesis in 3T3 cells, Cell 17 (1979) 527–535. sperm tail: molecular cloning and expression analyses of two cDNA transcripts [40] I.B. Alieva, I.A. Vorobjev, Vertebrate primary cilia: a sensory part of centrosomal encoding proteins of 70 kDa, Mol. Hum. Reprod. 5 (1999) 627–635. complex in tissue, but a “sleeping beauty ” in cultured cells? Cell Biol. Int. 28 [5] U. Schalles, X. Shao, F.A. van der Hoorn, R. Oko, Developmental expression of the (2004) 139–150. 84 kDa ODF sperm protein: localization to both the cortex and medulla of outer [41] E.S. Seeley, C. Carriere, T. Goetze, D.S. Longnecker, M. Korc, Pancreatic cancer and dense fibers and to the connecting piece, Dev. Biol. 199 (1998) 250–260. precursor pancreatic intraepithelial neoplasia lesions are devoid of primary cilia, [6] B.M.H. Lange, K. Gull, A molecular marker for centriole maturation in the mam- Cancer Res. 69 (2009) 422–430. malian cell cycle, J. Cell Biol. 130 (1995) 919–927. [42] J. Thomas, L. Morie, F. Soulavie, A. Laurençon, S. Sagnol, B. Durand, Transcriptional [7] Y. Nakagawa, Y. Yamane, T. Okanoue, S. Tsukita, S. Tsukita, Outer dense fiber 2 is a control of genes involved in ciliogenesis: a first step in making cilia, Biol. Cell. 102 widespread centrosome scaffold component preferentially associated with mother (2010) 499–513. centrioles: its identification from isolated centrosomes, Mol. Biol. Cell 12 (2001) [43] E.N. Blatt, X.H. Yan, M.K. Wuerffel, D.L. Hamilos, S.L. Brody, Forkhead transcription 1687–1697. factor HFH-4 is temporally related to ciliogenesis, Am. J. Respir. Cell Mol. Biol. 21 [8] S. Hoyer-Fender, J. Neesen, J. Szpirer, C. Szpirer, Genomic organisation and chro- (1999) 168–176. mosomal assignment of Odf2 (outer dense fiber 2), encoding the main com- [44] S.L. Brody, X.H. Yan, M.K. Wuerffel, S.K. Song, S.D. Shapiro, Ciliogenesis and left– ponent of sperm tail outer dense fibers and a centrosomal scaffold protein, right axis defects in forkhead factor HFH-4-null mice, Am. J. Respir. Cell Mol. Cytogenet. Genome Res. 103 (2003) 122–127. Biol. 23 (2000) 45–51. [9] S. Schweizer, S. Hoyer-Fender, Mouse Odf2 localizes to centrosomes and basal [45] T. Huang, Y. You, M.S. Spoor, E.J. Richer, V.V. Kudva, R.C. Paige, M.P. Seiler, J.M. bodies in adult tissues and to the photoreceptor primary cilium, Cell Tissue Res. Liebler, J. Zabner, C.G. Plopper, S.L. Brody, Foxj1 is required for apical localization 338 (2009) 295–301. of ezrin in airway epithelial cells, J. Cell Sci. 116 (2003) 4935–4945. [10] D. Hüber, S. Hoyer-Fender, Alternative splicing of exon 3b gives rise to ODF2 and [46] X. Yu, C.P. Ng, H. Habacher, S. Roy, Foxj1 transcription factors are master regula- Cenexin, Cytogenet. Genome Res. 119 (2007) 68–73. tors of the motile ciliogenic program, Nat. Genet. 40 (2008) 1445–1453. [11] D. Hüber, S. Geisler, S. Monecke, S. Hoyer-Fender, Molecular dissection of [47] B.V. Jacquet, R. Salinas-Mondragon, H. Liang, B. Therit, J.D. Buie, M. Dykstra, K. ODF2/Cenexin revealed a short stretch of amino acids necessary for targeting to Campbell, L.E. Ostrowski, S.L. Brody, H.T. Ghashghaei, FOXJ1-dependent gene the centrosome and the primary cilium, Eur. J. Cell Biol. 87 (2008) 137–146. expression is required for differentiation of radial glia into ependymal cells [12] F.F. Donkor, M. Mönnich, E. Czirr, T. Hollemann, S. Hoyer-Fender, Outer dense and a subset of astrocytes in the postnatal brain, Development 136 (2009) fibre protein 2 (ODF2) is a self-interacting centrosomal protein with affinity for 4021–4031. microtubules, Cell Sci. 117 (2004) 4643–4651. [48] R. Jain, J. Pan, J.A. Driscoll, J.W. Wisner, T. Huang, S.P. Gunsten, Y. You, S.L. Brody, The [13] H. Ishikawa, A. Kubo, S. Tsukita, S. Tsukita, Odf2-deficient mother centrioles temporal relationship between primary and motile ciliogenesis in airway epithelial lack distal/subdistal appendages and the ability to generate primary cilia, Nat. cell, Am. J. Respir. Cell Mol. Biol. 43 (2010) 731–739, http://dx.doi.org/10.1165/ Cell Biol. 7 (2005) 517–524. rcmb.2009-0328OC. [14] D.N. Wheatley, A.M. Wang, G.E. Strugnell, Expression of primary cilia in mamma- [49] E. Bonnafe, M. Touka, A. Ait-Lounis, D. Baas, E. Barras, C. Ucla, A. Moreau, F. lian cells, Cell Biol. Int. 20 (1996) 73–81. Flamant, R. Dubruille, P. Couble, J. Collignon, B. Durand, W. Reith, The transcrip- [15] K.C. Corbit, P. Aanstad, V. Singla, A.R. Norman, D.Y. Stainier, J.F. Reiter, Vertebrate tion factor RFX3 directs nodal cilium development and left–right asymmetry smoothened functions at the primary cilium, Nature 437 (2005) 1018–1021. specification, Mol. Cell. Biol. 24 (2004) 4417–4427. [16] C.J. Haycraft, B. Banizs, Y. Aydin-Son, Q. Zhang, E.J. Michaud, B.K. Yoder, Gli2 and [50] A. Ait-Lounis, D. Baas, E. Barras, C. Benadiba, A. Charollais, R. Nlend Nlend, D. gli3 localize to cilia and require the intraflagellar transport protein polaris for Liegeois, P. Meda, B. Durand, W. Reith, Novel function of the ciliogenic transcrip- processing and function, PLoS Genet. 1 (2005) e53. tion factor RFX3 in development of the endocrine pancreas, Diabetes 56 (2007) [17] D. Huangfu, K.V. Anderson, Cilia and Hedgehog responsiveness in the mouse, 950–959. Proc. Natl. Acad. Sci. U. S. A. 102 (2005) 11325–11330. [51] L. El Zein, A. Ait-Lounis, L. Morlé, J. Thomas, B. Chhin, N. Spassky, W. Reith, B. [18] A. Liu, B. Wang, L.A. Niswander, Mouse intrafl agellar transport proteins regulate Durand, RFX3 governs growth and beating efficiency of motile cilia in mouse both the activator and repressor functions of Gli transcription factors, Develop- and controls genes involved in human ciliopathies, J. Cell Sci. 122 (2009) ment 132 (2005) 3103–3111. 3180–3189. [19] S.R. May, A.M. Ashique, M. Karlen, B. Wang, Y. Shen, K. Zarbalis, J. Reiter, J. Ericson, [52] M.L. Whitfield, G. Sherlock, A.J. Saldanha, J.I. Murray, C.A. Ball, K.E. Alexander, J.C. A.S. Peterson, Loss of the retrograde motor for IFT disrupts localization of Smo to Matese, C.M. Perou, M.M. Hurt, P.O. Brown, D. Botstein, Identification of genes cilia and prevents the expression of both activator and repressor functions of Gli, periodically expressed in the human cell cycle and their expression in tumors, Dev. Biol. 287 (2005) 378–389. Mol. Biol. Cell 13 (2002) 1977–2000. [20] R. Rohatgi, L. Milenkovic, M.P. Scott, Patched 1 regulates hedgehog signaling at [53] V.B. Bajic, V. Choudhary, C.K. Hock, Content analysis of the core promoter region the primary cilium, Science 317 (2007) 372–376. of human genes, In Silico Biol. 4 (2004) 109–125. [21] E. Tassouri, K.L. Tucker, Primary cilia and organogenesis: is Hedgehog the only [54] H. Müller, M.C. Moroni, E. Vigo, B.O. Petersen, J. Bartek, K. Helin, Induction of sculptor? Cell Tissue Res. 345 (2011) 21–40. S-phase entry by E2F transcription factors depends on their nuclear localization, [22] J.L. Badano, T.M. Teslovich, N. Katsanis, The centrosome in human genetic disease, Mol. Cell. Biol. 17 (1997) 5508–5520. Nat. Rev. Genet. 6 (2005) 194–205. [55] C. Sardet, M. Vidal, D. Cobrinik, Y. Geng, C. Onufryk, A. Chen, R.A. Weinberg, E2F-4 [23] J.L. Badano, N. Misuma, P.L. Beales, N. Katsanis, The ciliopathies: an emerging class and E2F-5, two members of the E2F family, are expressed in the early phases of of human genetic disorders, Annu. Rev. Genomics Hum. Genet. 7 (2006) 125–148. the cell cycle, Proc. Natl. Acad. Sci. U. S. A. 92 (1995) 2403–2407. [24] J.M. Gerdes, E.E. Davis, N. Katsanis, The vertebrate primary cilium in develop- [56] P. Meraldi, J. Lukas, A.M. Fry, J. Bartek, E.A. Nigg, Centrosome duplication in mam- ment, homeostasis, and disease, Cell 137 (2009) 32–45. malian somatic cells requires E2F and Cdk2-cyclin A, Nat. Cell Biol. 1 (1999) [25] S.C. Goetz, K.V. Anderson, The primary cilium: a signaling center during verte- 88–93. brate development, Nat. Rev. Genet. 11 (2010) 331–344. [57] L.M. Facchini, L.Z. Penn, The molecular role of Myc in growth and transformation: [26] D.R. Kellogg, M. Moritz, B.M. Alberts, The centrosome and cellular organization, recent discoveries lead to new insights, FASEB J. 12 (1998) 633–651. Annu. Rev. Biochem. 63 (1994) 639–674. [58] G. Suske, E. Bruford, S. Philipsen, Mammalian SP/KLF transcription factors: bring [27] S. Doxsey, Re-evaluating centrosome function, Nat. Rev. Mol. Cell Biol. 2 (2001) in the family, Genomics 85 (2005) 551–556. 688–698. [59] A. Doetzlhofer, H. Rotheneder, G. Lagger, M. Koranda, V. Kurtev, G. Brosch, E. [28] M. Bornens, Centrosome composition and microtubule anchoring mechanisms, Wintersberger, C. Seiser, Histone deacetylase I can repress transcription by bind- Curr. Opin. Cell Biol. 14 (2002) 25–34. ing to Sp1, Mol. Cell. Biol. 19 (1999) 5504–5511. [29] H. Varmark, Functional role of centrosomes in spindle assembly and organization, [60] A. Zannetti, S. Del Vecchio, M.V. Carriero, R. Fonti, P. Franco, G. Botti, G. D'Aiuto, J. Cell. Biochem. 91 (2004) 904–914. M.P. Stoppelli, M. Salvatore, Coordinate up-regulation of Sp1 DNA-binding activ- [30] E. Robbins, G. Jentzsch, A. Micali, The centriole cycle in synchronized HeLa cells, ity and urokinase receptor expression in breast carcinoma, Cancer Res. 60 (2000) J. Cell Biol. 36 (1968) 329–339. 1546–1551. [31] I.A. Vorobjev, Y.S. Chentsov, Centrioles in the cell cycle. I. Epithelial cells, J. Cell [61] E. Deniaud, J. Baguet, R. Chalard, B. Blanquier, L. Brinza, J. Meunier, M.C. Michallet, Biol. 98 (1982) 938–949. A. Laugraud, C. Ah-Soon, A. Wierinckx, M. Castellazzi, J. Lachuer, C. Gautier, J. [32] P.L. Alvey, An investigation of the centriole cycle using 3T3 and CHO cells, J. Cell Marvel, Y. Leverrier, Overexpression of transcription factor Sp1 leads to gene Sci. 78 (1985) 147–162. expression perturbations and cell cycle inhibition, PLoS 4 (2009) 1–13. [33] S. Hoyer-Fender, Centriole maturation and transformation to basal body, Semin. [62] E. Lam, W.F. Watson, R.J. Watson, An E2F-binding site mediates cell-cycle regulat- Cell Dev. Biol. 21 (2010) 142–147. ed repression of mouse B-myb transcription, EMBO J. 12 (1993) 2705–2713. [34] C.T. Anderson, T. Stearns, Centriole age underlies asynchronous primary cilium [63] Z. Lu, R.Z. Luo, H. Peng, M. Huang, A. Nishimoto, K.K. Hunt, K. Helin, W.S.L. Liao, Y. growth in mammalian cells, Curr. Biol. 19 (2009) 1498–1502. Yu, E2F–HDAC complexes negatively regulate the tumor suppressor gene ARHI in [35] V.G. Fonte, R.L. Searls, S.R. Hilfer, The relationship of cilia with cell division and breast cancer, Oncogene 25 (2006) 230–239. differentiation, J. Cell Biol. 49 (1971) 226–229. [64] A. Minden, A. Lin, M. McMahon, C. Lange-Carter, B. Derijard, R.J. Davis, G.L. [36] D.N. Wheatley, Cilia in cell-cultured fibroblasts. III. Relationship between mitotic activ- Johnson, M. Karin, Differential activation of ERK and JNK mitogen activated pro- ity and cilium frequency in mouse 3T6 fibroblasts, J. Anat. 110 (1971) 367–382. tein kinases by Raf-1 and MEKK, Science 266 (1994) 1719–1723. 1346 N. Pletz et al. / Biochimica et Biophysica Acta 1833 (2013) 1338–1346

[65] P. Monje, M.J. Marinissen, J.S. Gutkind, Phosphorylation of the carboxyl-terminal [67] D.P. Ramji, P. Foka, CCAAT/enhancer-binding proteins: structure, function and transactivation domain of c-Fos by extracellular signal-regulated kinase mediates regulation, Biochem. J. 365 (2002) 561–575. the transcriptional activation of AP-1 and cellular transformation induced by [68] C. Nerlov, The C/EBP family of transcription factors: a paradigm for interaction platelet-derived growth factor, Mol. Cell. Biol. 23 (2003) 7030–7043. between gene expression and proliferation control, Trends Cell Biol. 17 (2007) [66] L.O. Murphy, S. Smith, R.H. Chen, D.C. Fingar, J. Blenis, Molecular interpretation of 318–324. ERK signal duration by immediate early gene products, Nat. Cell Biol. 4 (2002) [69] P.F. Johnson, Molecular stop signs: regulation of cell-cycle arrest by C/EBP tran- 556–564. scription factors, J. Cell Sci. 118 (2005) 2545–2555.