a novel PHD-containing protein modulating ... - Semantic Scholar

2946

Research Article

SPOC1: a novel PHD-containing protein modulating chromatin structure and mitotic chromosome condensation Sarah Kinkley1, Hannah Staege1, Gerrit Mohrmann1,*, Gabor Rohaly1, Theres Schaub1,‡, Elisabeth Kremmer2, Andreas Winterpacht3 and Hans Will1,§ 1

Heinrich-Pette Institute for Experimental Virology and Immunology, Martinistrasse 52, 20251 Hamburg, Germany Institute of Molecular Immunology, Helmholtz Center Munich, German Center for Environmental Health (GmbH), Marchioninstrasse 25, 81377 Munich, Germany 3 Institute for Human Genetics, Schwabachanlage 10, 91054 Erlangen, Germany 2

*Present address: Labor Arndt, Lademannbogen 61, 22339 Hamburg, Germany ‡ Present address: Institute for Human Genetics, University Hospital Eppendorf (UKE), Martinistrasse 52, 20246 Hamburg, Germany § Author for correspondence ([email protected])

Accepted 11 May 2009 Journal of Cell Science 122, 2946-2956 Published by The Company of Biologists 2009 doi:10.1242/jcs.047365

Summary In this study, we characterize the molecular and functional features of a novel protein called SPOC1. SPOC1 RNA expression was previously reported to be highest in highly proliferating tissues and increased in a subset of ovarian carcinoma patients, which statistically correlated with poor prognosis and residual disease. These observations implied that SPOC1 might play a role in cellular proliferation and oncogenesis. Here we show that the endogenous SPOC1 protein is labile, primarily chromatin associated and its expression as well as localization are regulated throughout the cell cycle. SPOC1 is dynamically regulated during mitosis with increased expression levels and biphasic localization to mitotic chromosomes indicating a functional role of SPOC1 in mitotic processes. Consistent with this postulate, SPOC1 siRNA knockdown experiments resulted in defects in mitotic

Introduction We have recently identified a novel human gene, which contains an open reading frame for a protein with a single PHD domain, coined Survival time associated PHD finger protein in Ovarian Cancer 1 (SPOC1) (Mohrmann et al., 2005). To date, nothing has been reported on SPOC1 protein expression and there is only information on its RNA expression and its chromosomal location. The SPOC1 gene is located in chromosomal region 1p36.3, an area previously identified for chromosomal instability and implicated in tumour development and progression (Van Gele et al., 1998; Varga et al., 2001). Consistent with this possibility, SPOC1 RNA transcript levels were found to be significantly enhanced in a subset of ovarian carcinomas and statistically correlated with poorer survival and residual disease in these patients (Mohrmann et al., 2005). SPOC1 transcripts are most highly expressed in the testis and ovaries and are ubiquitously expressed to a low level in all tissues. Furthermore, SPOC1 RNA was found to be specifically expressed in the highly proliferating cell types of the testis such as the spermatagonia but not in the mature non-proliferating sperm cells (Mohrmann et al., 2005). These data are compatible with a possible role of this protein in cell proliferation and oncogenesis of specific cell types. The human SPOC1 protein is 300 amino acids in length and has a predicted molecular mass of 34 kDa. Except for the single C-

chromosome condensation, alignment and aberrant sister chromatid segregation. Finally, we have been able to show, using micrococcal nuclease (MNase) chromatin-digestion assays that SPOC1 expression levels proportionally influence the degree of chromatin compaction. Collectively, our findings show that SPOC1 modulates chromatin structure and that tight regulation of its expression levels and subcellular localization during mitosis are crucial for proper chromosome condensation and cell division. Supplementary material available online at http://jcs.biologists.org/cgi/content/full/122/16/2946/DC1 Key words: PHD domain, Chromatin, Mitosis, Chromosome condensation

terminal PHD domain and a bipartite nuclear localization sequence, suggesting that it may be a nuclear chromatin affiliated protein, no additional functional domains for the SPOC1 protein have so far been identified. Therefore, based on the current limited knowledge it can only be postulated that SPOC1 is an expressed nuclear localized protein and that it may play a role in cell proliferation and oncogenesis. To understand the involvement of SPOC1 in these processes and to gain insight into its biological function we have characterized it at the protein level. In order to experimentally evaluate SPOC1 we generated expression constructs, antibodies, stable cell lines, siRNA oligos and performed a more detailed ‘in silico’ analysis. Using these tools and various biochemical approaches we show that SPOC1 is a tightly regulated chromatin-associated protein capable of modulating chromatin structure and important for proper mitotic chromosome condensation and cell division. Results SPOC1 is a novel protein conserved across higher eukaryotes and contains several interesting putative functional domains

The evolutionary conservation of a protein is a good indication of biological relevance and implies an important biological function. Therefore we evaluated in detail the SPOC1 protein by ‘in silico’

SPOC1: a novel mitotic player sequence alignment of all SPOC1 proteins currently available in the GenBank. This revealed that SPOC1 is highly conserved across higher eukaryotes and probably originated in evolution in the phylum Chordata, subphylum Vertebrata (supplementary material Fig. S1). The SPOC1 protein (also designated PHF13 in GenBank annotations) is conserved in species ranging from fish to humans and shows total sequence homologies ranging from 62% to 100% (supplementary material Table S1). Further evaluation of the SPOC1 sequences for putative domain structures identified a highly conserved N-terminal region of unknown function (except in Gallus gallus) from amino acid residues 20 to 70 (Mac Vector 7.0), a centrally located bipartite nuclear localization sequence (Prosite motif scan) from residues 110 to 127, a C-terminal PHD domain from residues 232 to 280 (PFAM and SMART servers) and two high scoring PEST domains at amino acid residues 52-88 and 141-190 (www.at.embnet.org/embnet/tools/bio/PESTfind; supplementary material Fig. S1; Fig. 1A). Interestingly, these motifs, with the exception of the PEST domains, show very little sequence diversity and are almost 100% conserved from fish to humans (supplementary material Fig. S1) arguing strongly for the functional conservation of the SPOC1 protein across the vertebrate species in the phylum Chordata. Furthermore, although the sequence length and amino acid composition for the two PEST domains varies in lower

2947

Chordata species, they received similar or better PEST scores by the PESTfind program (data not shown), indicating that these degradation motifs are also conserved. SPOC1 is a novel chromatin-associated protein

The putative domain structure of SPOC1 implies that it may be a labile [PEST domains (Chevaillier, 1993; Garcia-Alai et al., 2006)] nuclear (NLS) protein that is capable of interacting with and modulating chromatin [PHD domain (Aasland et al., 1995; Bienz, 2006; Mellor, 2006; Zhang, 2006)]. To evaluate SPOC1 at the protein level, rabbit polyclonal and rat monoclonal antibodies against SPOC1 were generated and tested for specificity (supplementary material Figs S2-S4). Consistent with its conserved NLS and PHD domain, immunofluorescence localization of both exogenous and endogenous SPOC1 showed that it is in fact a nuclear protein (Fig. 1B). To demonstrate the functionality of the SPOC1 NLS, a FLAG-SPOC1 construct deleted of its NLS was expressed. SPOC1 ΔNLS was found to localize to both the cytoplasm and the nucleus, indicative of the NLS being at least in part responsible for the nuclear localization of SPOC1. The partial nuclear localization of SPOC1 ΔNLS may be explained by either simple diffusion ratios and/or that it has a nuclear binding partner that helps to retain it in the nucleus. In contrast to the NLS, deleting SPOC1 of either of its PEST domains or its PHD

Fig. 1. Predicted domain structure and subcellular localization of SPOC1. (A) A schematic depiction of the predicted domain structure and putative GSK3β phosphorylation sites (P) of the human SPOC1 protein. (B) Immunofluorescence localization of endogenous and exogenous SPOC1, FLAG-SPOC1 ΔNLS, SPOC1 ΔPEST1, SPOC1 ΔPEST2, SPOC1 ΔPHD, GFP-β-gal and GFP-β-gal fused to the SPOC1 NLS in U2OS cells. Endogenous SPOC1 (upper row) and transiently expressing SPOC1 (second row) were detected with the rabbit polyclonal CR53 antibody (green). FLAG-SPOC1 ΔNLS was detected with a mouse anti-FLAG antibody. GFPβ-gal and GFP-β-gal SPOC1 NLS were detected by the intrinsic fluorescence of GFP. SPOC1 ΔPEST1 was detected with rat monoclonal 7F8, whereas SPOC1 ΔPEST2 and SPOC1 ΔPHD were detected with the rat mAb 6F6. DNA was stained with Hoechst (blue), and phase contrast (PC) images were captured. Scale bars: 10 μm. (C) Immunoblot detection of SPOC1 from U2OS cells lysed by either a 2% SDS total cell lysis approach (Tot.) or after a differential fractionation procedure producing soluble (Sol.), chromatin (Chr.) and insoluble (Ins.) fractions. SPOC1 was detected with the rat mAb 6F6. Histone H3 (H3) and HP1α served as chromatin fraction markers and p21 served as a soluble fraction marker. (D) Immunoblot detection of exogenous expression of SPOC1, SPOC1 ΔPEST1, SPOC1 ΔPEST2 and SPOC1 ΔPHD in differentially fractionated lysates, soluble (S), moderately chromatin-associated (C1) and tightly chromatinassociated (C2) fractions. Histone H3, HP1α and p21 served as fraction-specific markers. SPOC1, SPOC1 ΔPEST2 and SPOC1 ΔPHD were detected with the rat mAb 6F6 whereas SPOC1 ΔPEST1 was detected with the rat mAb 7F8.

2948

Journal of Cell Science 122 (16)

domain did not alter its nuclear localization (Fig. 1B). To investigate the functionality of the SPOC1 NLS further, it was fused to GFP-βgal, a strictly cytoplasmic protein. This fusion resulted in the nuclear localization of GFP-β-gal, demonstrating that the NLS of SPOC1 is functional (Fig. 1B). Taken together these findings convincingly show that SPOC1 is a nuclear-localized protein and that it contains a functional NLS. To further evaluate which subcellular fraction SPOC1 localizes to, immunoblotting was performed with total and differentially fractionated cellular lysates. Although SPOC1 is predicted to be a 34 kDa protein, it shows aberrant electrophoretic mobility with an apparent molecular mass of 43 kDa. The SPOC1 protein was primarily detected in the chromatin-associated fraction in immunoblots along with the marker proteins histone H3 and HP1α (Fig. 1C) arguing strongly that it is a chromatin-associated protein. However, it is important to note that upon longer exposure SPOC1 was also detected in the soluble fraction along with p21 (soluble nuclear protein), suggesting that a small percentage of SPOC1 is present in the soluble nuclear fraction of asynchronous cells. The fact that SPOC1 primarily localizes to the chromatin-associated fraction is consistent with the prediction of a C-terminal PHD domain, a domain which is known to directly interact with chromatin (Aasland et al., 1995; Bienz, 2006) and implicates a chromatinrelated function for SPOC1. The observations that deleting SPOC1 of either one of its PEST domains or its PHD domain did not alter its nuclear localization as determined by immunofluorescence microscopy (Fig. 1B), prompted us to further investigate whether deleting these regions had any effect on its subnuclear localization. To answer this question, we expressed full length SPOC1, SPOC1 ΔPEST1, SPOC1 ΔPEST2 and SPOC1 ΔPHD in cells, and then subjected the cells to a differential fractionation procedure that produced a soluble fraction (S), a moderately chromatin-associated fraction (C1) using a chromatin buffer without nuclease, and a tightly chromatinassociated fraction (C2) using a chromatin buffer supplemented with benzonase (Fig. 1D). Histone H3, HP1α and p21 served as fraction specific markers for the tightly chromatin-associated (H3 and HP1α), moderately chromatin-associated (HP1α) and soluble (p21) fractions. In this way we were able to determine if deleting any of these motifs resulted in an altered subcellular localization of the protein in comparison to full length SPOC1. Our results show that deleting the PEST1 domain results in no observable change to the subcellular localization of SPOC1. By contrast, however, deleting SPOC1 of its PEST2 domain or its PHD domain did result in a substantial shift in its biochemical localization and appeared to have contrasting effects on the solubility of the protein (Fig. 1D). Interestingly, deleting SPOC1 of its PEST2 domain appeared to make it more tightly associated with chromatin and very little was detected in either the soluble or the moderately chromatin-associated fraction. These findings imply that some region within the PEST2 domain is essential for SPOC1s ability to be released from chromatin into the nucleoplasm. In stark contrast to this, deleting the PHD domain of SPOC1, made the resulting protein significantly more soluble in comparison to the full length protein. These findings strongly support the prediction that the PHD domain of SPOC1 is important for its association with chromatin. SPOC1 is a labile protein that is regulated by GSK3β and by proteasome-mediated degradation

The prediction of two highly conserved and high scoring PEST domains in the SPOC1 protein by the PEST FIND program, suggests

that it may be a labile protein. PEST domains, which are found in less than 10% of all mammalian proteins, can confer susceptibility to degradation and are frequently found in proteins with short halflifes (Rechsteiner and Rogers, 1996). PEST sequences are enriched in proline (P), glutamic acid (E), serine (S), threonine (T) and aspartic acid (D) and have been identified in transcriptional regulators, key metabolic enzymes and cell cycle regulating proteins. To evaluate the significance of the PEST motifs in the SPOC1 protein we first measured the half-life of endogenous SPOC1 in both the soluble and chromatin subcellular fractions by treating cells in culture for varying periods of time with cycloheximide, an inhibitor of protein synthesis (Farber and Roberts, 1971). Endogenous soluble SPOC1 was found to have a very short halflife of approximately 10-15 minutes whereas chromatin-associated SPOC1 was found to be stable across the time frame of our experiment (Fig. 2A). These findings argue that soluble SPOC1 is in fact a labile protein and susceptible to degradation, whereas chromatin-associated SPOC1 is much more stable, further supporting the likelihood of a chromatin-associated function of SPOC1. With these results in mind we decided to further address the functionality of the PEST domains in regulating SPOC1 stability. To do this we measured the half-life of exogenous SPOC1 protein as well as of SPOC1 deleted of each of its PEST domains or of its PHD domain, in the soluble nuclear fraction (Fig. 2B-E). Exogenous SPOC1 was found to have a half-life of approximately 10-15 minutes similar to our findings for endogenous SPOC1 (Fig. 2A,B). Interestingly, however, deleting either one of the PEST domains (SPOC1 ΔPEST1 or SPOC1 ΔPEST2) resulted in a markedly stabilized soluble nuclear protein (Fig. 2C,D). Furthermore, deletion of the C-terminal PHD domain, a domain with no known function in protein stability and of proportional size to the PEST2 domain, resulted in no stabilization of SPOC1 and the protein was found to have a similar half-life to that of both endogenous and exogenous full length SPOC1 (Fig. 2E). These findings strongly suggest that SPOC1 is a labile protein and that the two PEST domains play a role in regulating the stability and half-life of the SPOC1 protein. In order to gain additional evidence that SPOC1 is a short lived protein, we also evaluated the influence of different proteasome inhibitors on SPOC1 stability. We treated one of our inducible SPOC1 stable cell lines (U2OS SPOC1 clone #5) with doxycycline for 24 hours to induce SPOC1 expression and then with either MG132 or Velcade for varying periods of time (Fig. 2F). Consistent with SPOC1 being a labile protein, both Velcade and MG132 increased SPOC1 expression levels in a time-dependent manner. These results further support that SPOC1 is a short-lived protein and that proteasome-mediated degradation also regulates SPOC1 expression levels. The SPOC1 amino acid sequence is also predicted to contain eight GSK3β phosphorylation motifs (SxxxS) as determined by the Elm server (www.elm.eu.org), most of which are evolutionarily conserved and six of which lie directly within the two predicted PEST domains. This observation was intriguing, since GSK3β phosphorylation has been previously reported to play a role in the regulation of protein stability for a variety of labile proteins and in some cases by directly cooperating with the ubiquitin proteasome system (Alao et al., 2006a; Alao et al., 2006b; Zhou et al., 2004). GSK3β is a serine-threonine kinase that requires its substrates to be pre-phosphorylated by an additional priming kinase at the +4 position before it physically binds and phosphorylates its substrate at the N-terminal serine (Meijer et al., 2004; Patel et al., 2004). As

SPOC1: a novel mitotic player

2949

To evaluate whether GSK3β is capable of regulating SPOC1 expression levels, we treated cells in culture with two independent GSK3β inhibitors, LiCl and insulin (Meijer et al., 2004; Patel et al., 2004). LiCl and insulin were found to both independently stabilize SPOC1 expression levels in comparison to the induced only control (Fig. 2H) and further implicate a role of this kinase in the regulation of SPOC1. The observations that both GSK3β and proteasome inhibitors alone were capable of stabilizing SPOC1 expression levels, indicates that both of the corresponding enzymes play a role in SPOC1 degradation. Interestingly when used in combination, LiCl and MG132 had a synergistic effect, supporting the possibility that both GSK3β and proteasomal proteases may act together in regulating SPOC1 stability and half-life (Fig. 2H). Consistent with this possibility, endogenous SPOC1 protein was also similarly stabilized in various human cell lines after treatment with LiCl and MG132 (Fig. 2I). Taken together, these findings show that SPOC1 is a labile protein and that it is regulated in its expression by both GSK3β and the proteasome degradation system. SPOC1 is dynamically regulated during the cell cycle and mitosis

Fig. 2. SPOC1 is a labile protein regulated in part by GSK3β and proteasomal degradation. (A) Immunoblot analysis of endogenous SPOC1 expression levels from fractionated lysates of U2OS cells that were treated for varying periods of time with cycloheximide (CHX). Tubulin and actin served as loading controls for the soluble (Sol.) and chromatin (Chr.) fractions, respectively. (B-E) Immunoblot analysis of soluble lysates from U2OS cells exogenously expressing SPOC1, SPOC1 ΔPEST1, SPOC1 ΔPEST2 or SPOC1 ΔPHD and treated for varying periods of time with cycloheximide (CHX). Tubulin was used as a loading control and GFP as a transfection control. (F) Doxycycline-induced expression of exogenous SPOC1 (24 hours) in a stable U2OS (clone #5) cell line is stabilized by two independent proteasome inhibitors in a time-dependent manner. Tubulin and actin served as loading controls. (G) Endogenous GSK3β co-immunoprecipitates with FLAG-SPOC1 but not from the control vector-transfected lysate when purified with FLAGM2 agarose. IN, input; S, supernatant; IP, immunoprecipitate; WB, western blot. (H) LiCl and MG132 both independently stabilize SPOC1 expression and act synergistically. Doxycycline-induced (24 hours) U2OS SPOC1 clone #5 was left untreated or treated for 6 hours with LiCl, insulin, MG132 or LiCl+MG132 and then analyzed by immunoblotting for SPOC1 expression levels. Tubulin and actin served as loading controls. (I) Endogenous SPOC1 expression is stabilized in various human cell lines (U2OS, HeLa, Ovcar 5 and Ovcar 8) after 6 hours of LiCl+MG132 treatment. Tubulin served as a loading control. SPOC1 was detected in A, B, D, E, G and I with the rat mAb 6F6, in C with the rat mAb 7F8, and in F and H with the rabbit polyclonal CR53 antibody.

a result of this docking property it is often possible to coimmunoprecipitate this kinase with its substrates. Therefore to evaluate whether or not SPOC1 may be a GSK3β substrate, we expressed FLAG-tagged SPOC1 or empty FLAG-vector (as a control) in cells and immunoprecipitated the lysates with antiFLAG-M2 agarose and then eluted the purified proteins using FLAG peptide (Fig. 2G). FLAG-SPOC1 was found to specifically coimmunoprecipitate endogenous GSK3β from cell lysate, implicating that the two proteins interact either directly or indirectly.

Triggered by the observations that SPOC1 is stable when associated with chromatin but labile when soluble in the nucleoplasm, we decided to investigate its cell-cycle-specific expression and subcellular localization profile to see if these properties were evident at specific cell cycle stages. To evaluate this, we synchronized CV1 cells in specific cell cycle stages by growing them in an isoleucinedeficient medium for 40 hours to block them in G0. The cells were then released in normal FBS supplemented medium for 8 hours to bring them to the G1-S phase border and were then collected every 2 hours and analyzed by both FACS profiling and immunoblotting (Fig. 3A,B). Interestingly, these experiments showed that SPOC1 expression levels are significantly decreased both in the total cell lysate and in the chromatin fraction at the G1-S phase transition (0 hours) and during early S phase (2 hours) as was determined by both the FACS profile and the levels of cyclin A and geminin. Furthermore, SPOC1 was detected at these time points in the soluble fraction upon longer exposure (Fig. 3B). In contrast to this, SPOC1 expression levels and its chromatin association are dramatically increased during late G2 and M phases, as determined by both the FACS profile and the expression levels of cyclin A and cyclin B, indicating a potential functional role of SPOC1 during both late G2 and mitosis (Fig. 3B). These findings indicate that both the expression levels and localization of SPOC1 are regulated during the cell cycle, further indicating a potential cell-cycle-specific function of the SPOC1 protein. The results of the synchronization experiments suggested that SPOC1 may have a functional purpose during mitosis. In an effort to analyze the relevance of this we looked at the localization of SPOC1 to mitotic chromosomes during the distinct mitotic phases by indirect immunofluorescence microscopy and immunoblotting (Fig. 3C,D). Consistent with our synchronization data (Fig. 3B), SPOC1 showed greater immunofluorescence intensity in mitotic cells than in interphase cells (white arrow in Fig. 3C) supporting the conclusion that SPOC1 expression is increased during mitosis. Most notably these experiments demonstrated also that SPOC1 is very dynamic in its chromatin association during mitosis. SPOC1 was found to be associated with chromatin during prophase, soluble during metaphase and early anaphase and then associated with chromatin once again during mid to late anaphase and telophase (Fig. 3C). These findings were also confirmed biochemically by

2950


Fig. 3. Cell cycle regulation and mitotic localization of SPOC1. Synchronized CV1 cells were analyzed by FACS (A) and immunoblotting (B) to identify the cell cycle stage-specific expression and localization of SPOC1. Geminin, cyclin A and cyclin B1 were used to control the cell cycle stage and tubulin and actin were used as loading controls. SPOC1 was detected with the rat mAb 6F6. The asterisk denotes that the blot with the soluble fraction was exposed 25 times longer than the chromatin fraction blot. (C) Immunofluorescence localization of endogenous SPOC1 (green) during the distinct mitotic phases was detected with the rabbit peptide polyclonal CR53 antibody. DNA was stained with Hoechst and phase contrast (PC) images were captured. The white arrow indicates an interphase cell that shows weaker fluorescence intensity compared with the mitotic cells. Scale bars: 10 μm. (D) Immunoblot analysis of SPOC1 in fractionated and total cell lysates from nocodazole (Noc)-treated (24 hours) and released U2OS cells. Pre-metaphase (PM), metaphase (M), early anaphase (EA), anaphase (A) and late anaphase-telophase (LA/T) time points were evaluated. Cyclin B1 and histone H3 phosphorylated at serine 10 (H3P10) served to mark the metaphase anaphase transition and anaphase, respectively. Actin served as a loading control. SPOC1 was detected with the rat mAb 6F6.

synchronizing cells in pre-metaphase with nocodazole (microtubule inhibitor) and then releasing them for varying periods of time to follow them through the distinct mitotic phases. SPOC1 localization to chromatin or the cytoplasm/nucleoplasm was determined by fractionating the cell pellets into soluble and chromatin-associated lysates (Fig. 3D). Consistent with the immunofluorescence findings SPOC1 was found to be both soluble and chromatin associated at late prophase/pre-metaphase (0 minutes), primarily soluble during metaphase and early anaphase (15-60 minutes) and then increased again back on chromatin at mid-late anaphase and telophase (90105 minutes; Fig. 3D). Together these data indicate that SPOC1 is very dynamically regulated in its association with chromatin during mitosis and further supports a potential functional role of this protein in the mitotic processes.

phospho 10 (H3P10) and analyzed by immunofluorescence microscopy (Fig. 4). When the mitotic cells of the SPOC1 siRNA knockdown group were analyzed at higher magnification, various mitotic defects were observed, including chromosome misalignments, lagging chromosomes and chromosome bridges (Fig. 4A). Although these defects were also observed occasionally in the control groups, it occurred with a greater frequency in the SPOC1 siRNA treated cells. To statistically evaluate these consequences, 500 mitotic cells from each group were screened and showed that these chromosome abnormalities occurred with a frequency of ~23% in SPOC1 siRNA-treated cells compared with only 4.7% and 7.3% in the control groups (Fig. 4B). Taken together these findings further substantiate the functional importance of SPOC1 in mitotic processes.

SPOC1 siRNA-mediated reduction results in pleiotropic mitotic defects

SPOC1 modulates chromatin structure and is important for mitotic chromosome condensation

Based on our aforementioned results we speculated that SPOC1 may play an important mitotic role. To evaluate this possibility we examined the efficacy of mitosis in U2OS cells after SPOC1 siRNA mediated knockdown, using two independent oligos. To this end, asynchronous U2OS cells were either left untreated, or transfected with a control or SPOC1-specific siRNA oligos (768 or 941) for 72 hours. Mitotic cells were then stained with histone H3 serine

With the observations that SPOC1 associates with mitotic chromosomes during prophase and anaphase (Fig. 3C,D), the two stages of the mitotic cell cycle where chromosome condensation occurs, and since defects in mitotic chromosome condensation have been previously reported to lead to the pleiotropic mitotic defects (Cimini et al., 2003; Inoue et al., 2008; Mora-Bermudez et al., 2007; Nasmyth, 2002; Yong-Gonzalez et al., 2007) that we have observed

SPOC1: a novel mitotic player

2951

Fig. 4. SiRNA-mediated reduction of SPOC1 results in various mitotic chromosome defects. (A) Immunofluorescence detection of histone H3 phosphophorylated at serine 10 (H3P10) in wild-type U2OS cells, control or SPOC1 siRNA-768 or -941transfected cells (green). DNA (blue) was stained with Hoechst. Scale bars: 5 μm. (B) Quantification of chromosome misalignments, lagging chromosomes and chromosome bridges (aberrant mitosis) of wildtype U2OS cells (Wt), control (si-Cont.) and SPOC1 siRNA (si-768, si-941)-transfected cells. 500 mitotic cells were counted for each sample. (C) Immunoblotting of total cell lysates from wildtype U2OS cells (Wt) or U2OS cells treated with either a control siRNA (C) or with SPOC1-specific siRNAs (768, 941). SPOC1 was detected with the rat mAb 6F6 and tubulin was used as a loading control.

in our SPOC1 siRNA knockdown cells, we further questioned whether reducing SPOC1 expression levels interfered with proper mitotic chromosome condensation. In order to investigate this possibility, we designed a functional assay in which SPOC1 expression was knocked down in cells for 48 hours by 768 siRNA transfection and then treated with nocodazole for 12 hours to evaluate the efficiency and ability of the cells to form typical premetaphase condensed structures. These cells were evaluated by comparative immunofluorescence microscopy against control cells treated only with nocodazole for 12 hours or that were first transfected with a control siRNA for 48 hours before the addition of nocodazole for 12 hours (Fig. 5A-C). To identify the mitotic nuclei we stained the cells for Ki-67, a proliferation marker which is localized to the nucleolus during interphase and labels mitotic chromosomes during mitosis (Endl and Gerdes, 2000; Scholzen and Gerdes, 2000). These experiments revealed that the majority of mitotic cells from the wild-type and control siRNA cohorts had the typical rounded-up pre-metaphase morphology that is expected after nocodazole treatment. By contrast, however, the mitotic cells of the SPOC1 siRNA cohort showed a more prophase-like morphology after nocodazole treatment (Fig. 5B). In order to quantify the significance of this, 500 mitotic cells were counted from each group

and classified as either prophase-like or pre-metaphase-like (Table 1). Cells were scored into these two categories based on the size and shape of their mitotic nuclei and on the degree of visible chromosome condensation. We observed that unlike the control cells that were primarily blocked at pre-metaphase (72.4-74.4%), SPOC1 768 siRNA-transfected cells appeared to a much greater extent to block in prophase (59.4%) showing significantly less condensed mitotic chromosomes (Fig. 5B, Table 1). Similar results were also obtained using a second SPOC1 siRNA (941) oligo (supplementary material Fig. S5A,B). These findings suggest that reducing SPOC1 expression levels may result in defective prophase chromosome condensation. To analyze whether or not the observed downstream chromosome abnormalities occur as a secondary consequence to defective chromosome condensation we used a high throughput modular microscope (ScanR, Olympus). Using this imaging system we were able to measure and capture all cells on a coverslip and divide them into different categories based on DNA content (cell cycle profile), Ki-67 staining intensities (mitotic cells) and cell size (prophase versus pre-metaphase cells). We were therefore able to demonstrate that the pre-metaphase cells of the SPOC1 siRNA cohort (40.6%) were significantly larger and less condensed than the pre-metaphase cells in the control cohorts (Fig. 5C).

2952


Fig. 5. SPOC1 protein expression levels modulate chromatin structure and condensation. (A) Immunoblotting of total cell lysates from wild-type (Wt) U2OS cells or cells transfected for 48 hours with control siRNA (Cont.) or SPOC1 siRNA 768. SPOC1 was detected with the rat mAb 6F6 and actin was used as a loading control. (B) Immunofluorescence analysis of wild-type (Wt) U2OS cells or U2OS cells transfected with control or SPOC1-specific 768 siRNA for 36 hours and then treated with nocodazole for 12 hours. Ki67 (green), DNA (blue) and phase contrast (PC) imaging was used to detect mitotic cells and pre-metaphase cells. Scale bar: 10 μm. (C) Pre-metaphase cells from each experimental group were identified using a scanning microscope and SCAN-R software. (D,E) U2OS cells were either left untransfected (Wt) or transfected with a SPOC1 siRNA (768 or 941), full length SPOC1 or SPOC1 ΔPHD constructs for 48 hours. SPOC1 levels were determined by immunoblotting (upper panels). SPOC1 was detected with rat mAb 6F6 and tubulin served as a loading control. Chromatin sensitivity assays (lower panels) were performed either using an RNase treatment before digestion with 40 U of MNase for the indicated periods of time (D) or digestion with 80 U of MNase for the indicated periods of time followed by an RNase treatment (E). DNA was separated on a 1% agarose gel. Mono-, di- and tri-nucleosomes are denoted.

Furthermore, using both SPOC1 siRNAs we observed a significant increase in the percentage of pre-metaphase cells with misalignments (~35% compared with the ~10% in the control cells; supplementary material Fig. S5C,D). These findings demonstrate that SPOC1 is important for mitotic chromosome architecture and for proper mitotic chromosome condensation. Furthermore, it is conceivable that condensation defects can result in downstream mitotic chromosomal abnormalities and therefore it is tempting to speculate that a disruption to proper prophase condensation may induce the observed alignment and segregation defects in SPOC1 siRNA-treated cells.

In order to gain some mechanistic insight into the ability of SPOC1 to modulate mitotic chromosome architecture and condensation we performed MNase chromatin sensitivity assays to evaluate directly the consequence of SPOC1-mediated reduction and SPOC1 transient overexpression on chromatin architecture and condensation. These experiments were performed using chromatin templates that were first treated with RNase (Fig. 5D) followed by the MNase treatment or vice versa (Fig. 5E) to preserve chromatin structure. The expression levels of SPOC1 were determined by immunoblotting (Fig. 5D,E). In these experiments we analyzed the MNase sensitivity of chromatin from wild-type U2OS cells in

SPOC1: a novel mitotic player Table 1. Percentage of prophase and pre-metaphase cells % of cells Cell type U2OS Control siRNA SPOC1 siRNA

Prophase

Pre-metaphase

27.6 25.6 59.4

72.4 74.4 40.6

The percentage of prophase and pre-metaphase cells were counted in the U2OS, control siRNA and SPOC1 siRNA cohorts after 12 hours of nocodazole treatment.

comparison to chromatin from U2OS cells transfected with a SPOC1-specific siRNA (768 or 941) or chromatin from U2OS cells transfected with either a SPOC1 or SPOC1 ΔPHD expression vector for 48 hours (Fig. 5D,E). We found, regardless of the experimental approach, that SPOC1 siRNA-mediated reduction resulted in a greater sensitivity of the chromatin to MNase digestion compared with the wild-type U2OS cells (Fig. 5D,E). This is evident from the increase in mono-, di- and tri-nucleosomes after equivalent lengths of MNase treatment (Fig. 5D,E) and by the reduced amount of undigested chromatin in SPOC1 siRNA-treated samples compared with the wild type control (Fig. 5E). In contrast to this, overexpression of SPOC1 led to a reduced sensitivity of the chromatin to MNase digestion compared with the wild-type U2OS cells, as is indicated by the absence mononucleosomes (Fig. 5D) or the substantial increase in undigested chromatin (Fig. 5E) after equivalent lengths of MNase treatment. Furthermore as a negative control, over expression of SPOC1 deleted of its PHD domain showed equivalent sensitivity to MNase digestion as the wild-type control U2OS cells (Fig. 5D). Therefore, we have been able to show here that over expression of wild-type SPOC1 results in a more condensed and less accessible chromatin architecture which is dependent on its PHD domain whereas siRNA-mediated reduction of SPOC1 results in less condensed and a more accessible chromatin architecture than in wild-type U2OS cells (Fig. 5D,E). These findings demonstrate that SPOC1 expression levels are important for chromatin architecture and show that depletion of SPOC1 results in a less condensed chromatin state, whereas its over expression increases the condensation state of chromatin. Discussion In this study we have described the expression and biological function of a novel protein called SPOC1. Using the predicted and conserved domain motifs of SPOC1 as a platform for our experimental evaluation, we have systematically dissected SPOC1 at the protein level. Based on previous work that suggested a relationship of enhanced SPOC1 RNA levels with tumourigenesis, we deemed it important to try and elucidate the physiological function of the SPOC1 protein. Lending credence to the biological importance of this protein is the fact that it is highly conserved across higher eukaryotes, it is tightly regulated in both its expression and localization during the cell cycle and because both its enhanced and reduced expression has been associated with proliferation, genomic instability and/or cancer. Our findings support, as is predicted by its C-terminal PHD domain, that SPOC1 is a chromatin-associated protein with an important role in the regulation of chromatin structure and function. PHD domains are found in less than 150 mammalian proteins with functions ranging from transcriptional activators, repressors to chromatin remodelling enzymes (Aasland et al., 1995; Bienz, 2006;

2953

Eberharter et al., 2004; Nourani et al., 2001; Papait et al., 2008), for example HATs (Nourani et al., 2001; Ullah et al., 2008), HDACs (Li et al., 2007; Yochum and Ayer, 2001), demethylases (Iwase et al., 2007) and methyltransferases (Tanaka et al., 2008). Recently, PHD domains have been identified as a molecular reader of the epigenetic modifications H3K4(me3) and H3K36(me3) both of which normally denote actively transcribed euchromatin (Li et al., 2006; Mellor, 2006; Ramon-Maiques et al., 2007; Shi et al., 2007; Zhang, 2006). PHD-containing proteins therefore participate in the epigenetic regulation of chromatin by dictating its structure and function, and how chromatin is interpreted by other molecular players. Based on SPOC1’s increased expression and dynamic biphasic chromatin localization during mitosis, we speculated that the protein may have a functional role in the regulation of chromatin structure and function during mitosis. Furthermore, since SPOC1 was specifically localized to chromatin during prophase and then later during anaphase, the stages of the mitotic cell cycle where chromosome condensation occurs, we further speculated that it plays a role in the regulation of mitotic chromosome condensation. Our results support these postulates. Functional assays employing siRNA-mediated reduction of SPOC1 were designed to measure the impact of SPOC1 knockdown on mitotic chromatin structure and function. Reducing SPOC1 protein levels resulted in various mitotic defects such as hypocondensed mitotic chromosomes, misalignment and segregation defects. Similar mitotic defects have been previously described after the disruption of other important factors involved in mitotic chromosome condensation (Cimini et al., 2003; Nasmyth, 2002; Somma et al., 2003) such as, components of the condensin complex (Yong-Gonzalez et al., 2007), chromokinesins (Mazumdar and Misteli, 2005; Mazumdar et al., 2004) topoisomerase IIα (Dawlaty et al., 2008; Downes et al., 1991), aurora kinase (Mora-Bermudez et al., 2007) and HP1α (Inoue et al., 2008). This can be explained by the fact that the failure of a mitotic cell to properly condense it chromosomes can result in unresolved sister chromatids, which can preclude successful metaphase alignment, often resulting in mitotic delays owing to activated checkpoints and ultimately leading to various segregation defects, such as lagging chromosomes, chromosome bridges and aneuploidy or polyploidy (Arlot-Bonnemains et al., 2001; Cortes and Pastor, 2003; Fu et al., 2007; Giet and Glover, 2001; Hodges et al., 2005; Storchova et al., 2006). We found that the mitotic chromosome structure was significantly altered after SPOC1 siRNA-mediated reduction, being larger and less condensed than normal mitotic chromosomes. These findings strongly indicate that SPOC1 does play a role in chromatin organization during mitosis and that reducing its expression levels can have deleterious consequences to cellular proliferation and genomic stability. Further evidence supporting a functional role of SPOC1 in modulating chromatin structure and condensation, was obtained from chromatin MNase sensitivity assays which demonstrated that SPOC1 overexpression results in a more compacted chromatin architecture in a PHD-domain-dependent manner. Therefore, based on all of these data we believe that SPOC1 plays a role in the condensation of mitotic chromosomes by directly altering the chromatin structure and by favouring a more condensed state. We postulate that SPOC1 directly tethers itself to specific histone residues in chromatin via its PHD domain and that when it is concentrated on the chromatin, as it is during early mitosis and later during anaphase, that this aids in condensing the chromatin architecture.

2954


Owing to their important biological roles, PHD-containing proteins have often either tumour suppressor functions or oncogenic potentials (Coles and Jones, 2009; Mohrmann et al., 2005; Shi et al., 2006; Ythier et al., 2008). As a consequence of their chromatin regulatory functions, disruptive aberrations to PHD-fingercontaining proteins can have severe genomic consequences, and therefore have been found to be involved in various human pathological conditions, ranging from mental retardation to malignancy (Badens et al., 2006; Barrett et al., 2007; Bronner et al., 2007; Campos et al., 2004; Chen et al., 2008; Coles and Jones, 2009; Ythier et al., 2008). The PHF23 protein is the most closely related protein to SPOC1, containing a predicted N-terminal domain, two predicted PEST motifs, a NLS and a PHD domain with strong sequence homologies (supplementary material Fig. S6). Although the function of PHF23 is not yet known, chromosomal translocation of the PHD finger of PHF23 to nucleoporin protein 98 has been recently identified in acute myeloid leukaemia and the corresponding predicted fusion protein was proposed to act as an aberrant transcription factor (Reader et al., 2007). Since SPOC1 appears to be important for cell division and proliferation it is conceivable that aberrant SPOC1 expression levels or localization may also favour oncogenic transformation which may explain its tight cellular regulation. Consistent with this assumption, we have shown here that reducing SPOC1 expression levels led to mitotic chromosome alignment and segregation defects. When left unresolved these mitotic defects lead to genomic instability and aneuploid phenotypes, which is the hallmark of almost all known malignancies. Furthermore, it was previously shown that increased SPOC1 RNA expression levels were found in a subset of ovarian carcinomas and resulted in a significantly poorer prognosis (Mohrmann et al., 2005). It will be important in the future to evaluate SPOC1 protein expression levels not only in ovarian carcinomas but also in other malignancies, to study the biological consequences of SPOC1 overexpression to genomic stability. Lastly, it will also be important to perform specific epigenetic studies to evaluate which chromatin modifications SPOC1 recognizes, in what context and if it is part of a greater molecular chromatin remodelling complex. In summation, we show here that the SPOC1 protein is an expressed chromatinassociated protein that is tightly regulated and is important for proper mitotic chromosome structure and function and therefore cell division processes.

GST-SPOC1. Lou/c rats were immunized twice intraperitoneally and subcutaneously with approximately 50 μg of affinity-purified GST-SPOC1 protein. After fusion of the myeloma cell line P3X63 Ag 8.653 with immune rat spleen cells, positive clones were identified with a solid phase enzyme linked immunosorbent assay (ELISA) using cell extracts of E. coli expressing the GST-SPOC1 fusion protein. An irrelevant GST fusion protein was used as a negative control. Positive clones were then further confirmed by immunoblotting, siRNA knockdown and transient expression experiments (see supplementary material Figs S2-S4). The rat monoclonal 6F6 and 7F8 antibodies were used in a 1:50 dilution for immunoblotting and undiluted for immunofluorescence staining of cells. SPOC1 stable inducible cell lines were generated using the Tet-On system commercially available from Invitrogen. Briefly a stable tetracycline (Tet) repressorexpressing cell line under blasticidine selection (Invivogen) was generated using the pcDNA6/TR expression vector. pCDNA-4TO-SPOC1 was then co-transfected into a Tet-repressor-expressing cell line and placed under the selection markers zeocin (0.5 mg/ml) and blasticidine (1 μg/ml). Surviving colonies were then isolated, grown and tested for doxycycline (Sigma)-inducible SPOC1 expression. Doxycycline was used at a concentration of 1 μg/ml and was added to the culture medium for 16-24 hours.

Cell lines and antibodies Human osteosarcoma cells (U2OS), cervical carcinoma cells (HeLa), two ovarian cancer cell lines (OVCAR 5 and OVCAR 8) and an African green monkey kidney cell line (CV1) were used in our experiments. All cell lines were cultured under standard conditions and grown in DMEM medium supplemented with 10% FCS, 1⫻ penicillin-streptomycin (PAA) and 10 mM Hepes (Gibco). Rabbit monoclonal antihistone H3 (Epitomics), rabbit polyclonal anti-geminin (Santa Cruz Biotechnology), anti-cyclin A (Santa Cruz Biotechnology) and anti-histone H3 serine phospho 10 (Upstate) and mouse monoclonal anti-HP1α (Upstate), anti-p21 (Santa Cruz Biotechnology), anti-tubulin (Sigma), anti-β-actin (Sigma), anti-GFP (Roche), anticyclin B1 (Santa Cruz Biotechnology), anti-FLAG (Sigma) and anti-Ki67 (Dako) antibodies were purchased from the respective companies and used according to manufacturer’s instructions. All HRPO-conjugated and fluorescently labelled secondary antibodies were purchased from Dianova and Invitrogen and used according to manufacturer’s instructions.

Extraction protocol and immunoblotting SPOC1 protein was extracted from cell pellets using either (1) a SDS total cell lysate extraction buffer (50 mM Tris, pH 6.8, 2% SDS and 10% glycerol) or (2) a fractionated lysate protocol. (1) In brief, the cell pellet (5⫻106 cells) was resuspended in 150 μl of SDS total cell lysis buffer supplemented with 1⫻ protease inhibitor cocktail (Roche), 1 mM NaF (Sigma), 1 mM Na3VO4 (Sigma) and 1 mM PMSF (Calbiochem), heated at 99°C for 5 minutes, supplemented with Laemmli buffer and then reheated for an additional 5 minutes at 99°C before loading on a SDS polyacrylamide gel. (2) The fractionated lysate protocol was performed as follows: soluble cellular proteins from approximately 5⫻106 cells were extracted with 150 μl of a standard NP40 CSK buffer (10 mM PIPES pH 7.0, 100 mM NaCl, 300 mM sucrose, 3 mM MgCl2, 0.1% NP40) supplemented with 1⫻ protease inhibitor cocktail (Roche), 10 μM MG132 proteasome inhibitor and 1 mM PMSF (Calbiochem) for 10 minutes on ice. Chromatin-associated proteins were extracted from the remaining pellet in 150 μl of a chromatin extraction buffer (20 mM Tris-HCl pH 7.5, 350 mM NaCl, 1% Triton X-100, 1 mM β-glycerophosphate, 1 mM Na3VO4, 1 mM NaF) supplemented with 1⫻ protease cocktail inhibitor, 1 mM PMSF, 2.5 mM MgCl2 and 25 units of Benzonase (Novagen) for 30 minutes on ice. SDS-PAGE and immunoblotting were performed using standard protocols.

Materials and Methods Plasmids

SiRNA knockdown

Full length human SPOC1 constructs were generated in the pCDNA4-TO (Invitrogen), pFLAG-CMV-4 (Sigma) and GST pGEX4-T1 (Amersham) expression vectors. Human SPOC1 constructs deleted of their PEST1 domain, PEST2 domain or PHD domain were also generated in the pCDNA4-TO expression vector, and SPOC1 deleted of its NLS in the pFLAG-CMV4 vector. The GFP–β-gal (β-galactosidase) fusion construct (pHM830) was kindly provided by Frank Fackelmayer (Biochemical Research Institute, Ioannina, Greece) and was used to generate the GFP–β-galSPOC1NLS construct. All constructs were generated using standard PCR and cloning approaches and were verified by sequencing. Unless otherwise stated, all transfections were performed using Fugene 6 (Roche) according to the manufacturer’s instructions.

SPOC1 protein was knocked down using the 768 siRNA oligo 5⬘-UCACCUGUCCUGUGCGAAA-3⬘ or the 941 siRNA oligo 5⬘-GAUCAGGUCAAAGAAAUAAA3⬘. Briefly, cells were allowed to grow to 80% confluence and then transfected with 100 nM end concentration of SPOC1 siRNA using Lipofectamine 2000 (Invitrogen) according to manufacturer’s instructions. Cells were then grown at 37°C for an additional 24-72 hours before harvesting (total cell lysate) for protein detection. A control siRNA 5⬘-AGGUAGUGUAAUCGCCUUG-3⬘ was purchased from Dharmacon.

Generation of SPOC1 antibodies and stable inducible cell lines SPOC1 antibodies were generated using standard approaches. In brief, the SPOC1 rabbit peptide polyclonal CR53 antibody was generated by immunizing rabbits three times with 100 μg of a KLH-coupled C-terminal peptide (CRDSKFDIRRSNRSRT) in complete (1⫻) and incomplete (2⫻) Freund’s adjuvant (Sigma). The antibody was affinity purified by chromatography using HiTrap NHS activated HP resin coupled to the peptide (Amersham Biosciences). The CR53 antibody was used at a 1:3000 and 1:500 dilution for immunoblotting and immunofluorescence experiments, respectively. Rat monoclonal antibodies were generated using recombinant full length

Immunofluorescence microscopy Immunofluorescence analysis of SPOC1 was performed according to standard protocols. In brief, cells were grown on coverslips, fixed with 4% paraformaldehyde in PBS, permeabilized with 0.5% Triton X-100, washed three times in PBS and then blocked overnight with 3% BSA in PBS. Coverslips were then incubated with primary antibodies according to the manufacturer’s described dilution for 1 hour, washed three times with PBS and then incubated with the appropriate secondary antibodies for 1 hour. The coverslips were then washed three more times in PBS and mounted on slides using Mowiol (Calbiochem). The cells were analyzed by indirect immunofluorescence microscopy on a Zeiss Axiophot or a high content screening cytometry microscope (ScanR Olympus).

SPOC1: a novel mitotic player Scan^R analysis Cells were seeded in six-well dishes, left untransfected or were transfected with either a control siRNA or a SPOC1-specific siRNA (768) for 36 hours. Following this, the cells were treated for 12 hours with 100 ng/ml nocodazole. Cells were analyzed on a high throughput scanning Olympus microscope (www.olympus.de/microscopy/ 22_scan_R.htm) for comparative analysis of mitotic nuclei. In brief, this combined microscope and computer platform scans and captures images from all cells in all wells for comparative analysis. Then, using an integrated FACS profiling function it is capable of generating a cell cycle profile from all cells based on a DNA stain (Hoechst). Using this FACS profile, cells in the G2-M phase were selected for in the first gate. From this G2-M population a second gate selecting for mitotic cells was generated based on Ki-67-Alexa Fluor 488 staining intensity. Lastly, based on the combined differences in DNA and nuclear volume, prophase cells were differentiated from pre-metaphase cells to generate our final gate of pre-metaphase only cells. All gates were constant for all experimental conditions and generated sub-galleries of cells that were then directly comparable with the same gate from other treatment conditions.

Cell synchronization The African green monkey kidney cells CV-1 were synchronized as previously described (Rohaly et al., 2005). In brief, the cells were starved of isoleucine for 3640 hours to block the cells in G0. The cells were then released back into medium containing normal serum for 8 hours before collecting at the first time point (0 hours) at the G1-S transition point. Cells were collected at 2-hour intervals up to 14 hours for FACS and immunoblot analysis. The DNA was stained with propidium iodide for FACS analysis. Cells were synchronized in mitosis at premetaphase by treating the cells in culture with a 100 ng/ml nocodazole (Sigma) for 12-24 hours. A mitotic shake-off was then performed and the cells were collected and replated.

In culture drug treatments Cycloheximide experiments were performed by incubating cells in culture with 10 μg/ml of cycloheximide (Sigma) for varying lengths of time. Inhibition of proteasomal activity was performed by incubating cells in culture for up to 6 hours with 10 μM MG132 (Calbiochem) or 1 μM Velcade (kindly provided by Sven Behrens, University of Halle-Wittenberg, Germany). GSK3β was inhibited by incubating cells in culture for 6 hours with either 40 mM LiCl (Merck) or with 100 nM insulin (Sigma).

MNase digestion assay In brief, cell pellets were lysed in a hypotonic buffer (10 mM Tris-HCl pH 7.4, 10 mM KCl, 15 mM MgCl2) on ice for 10 minutes. Nuclei were pelleted by centrifugation and resuspended in MNase digestion buffer (50 mM Tris-HCl pH 7.9, 5 mM CaCl2) supplemented with RNase and incubated at 37°C for 30 minutes. The DNA was then pelleted once again by centrifugation and resuspended in MNase digestion buffer supplemented with 100 μg/ml BSA and 40 IU MNase and incubated at room temperature for varying periods of time. The MNase reaction was stopped by the addition of 10 mM EDTA. Alternatively, nuclei were first treated with 80 IU of MNase at 37°C for varying periods of time. The MNase reaction was stopped by the addition of 10 mM EDTA followed by centrifugation. The pellet was then resuspensed in MNase digestion buffer supplemented with 10 mM EDTA and RNase and incubated at 37°C for 30 minutes. The DNA was extracted and purified by standard procedures.

This work was supported by a grant from the Deutsche Krebshilfe #10-2243-Wi4 and the Müggenburg Foundation. The Heinrich-Pette Institute is supported by the Freie und Hansestadt Hamburg and the German Federal Ministry of Health and Social Security. We would also like to thank Nicole Lohrengel, Urte Matschl and Kerstin Reumann for their excellent technical support.

References Aasland, R., Gibson, T. J. and Stewart, A. F. (1995). The PHD finger: implications for chromatin-mediated transcriptional regulation. Trends Biochem. Sci. 20, 56-59. Alao, J. P., Gamble, S. C., Stavropoulou, A. V., Pomeranz, K. M., Lam, E. W., Coombes, R. C. and Vigushin, D. M. (2006a). The cyclin D1 proto-oncogene is sequestered in the cytoplasm of mammalian cancer cell lines. Mol. Cancer 5, 7. Alao, J. P., Stavropoulou, A. V., Lam, E. W. and Coombes, R. C. (2006b). Role of glycogen synthase kinase 3 beta (GSK3beta) in mediating the cytotoxic effects of the histone deacetylase inhibitor trichostatin A (TSA) in MCF-7 breast cancer cells. Mol. Cancer 5, 40. Arlot-Bonnemains, Y., Klotzbucher, A., Giet, R., Uzbekov, R., Bihan, R. and Prigent, C. (2001). Identification of a functional destruction box in the Xenopus laevis auroraA kinase pEg2. FEBS Lett. 508, 149-152. Badens, C., Lacoste, C., Philip, N., Martini, N., Courrier, S., Giuliano, F., Verloes, A., Munnich, A., Leheup, B., Burglen, L. et al. (2006). Mutations in PHD-like domain of the ATRX gene correlate with severe psychomotor impairment and severe urogenital abnormalities in patients with ATRX syndrome. Clin. Genet. 70, 57-62. Barrett, A., Santangelo, S., Tan, K., Catchpole, S., Roberts, K., Spencer-Dene, B., Hall, D., Scibetta, A., Burchell, J., Verdin, E. et al. (2007). Breast cancer associated

2955

transcriptional repressor PLU-1/JARID1B interacts directly with histone deacetylases. Int. J. Cancer 121, 265-275. Bienz, M. (2006). The PHD finger: a nuclear protein-interaction domain. Trends Biochem. Sci. 31, 35-40. Bronner, C., Achour, M., Arima, Y., Chataigneau, T., Saya, H. and Schini-Kerth, V. B. (2007). The UHRF family: oncogenes that are drugable targets for cancer therapy in the near future? Pharmacol. Ther. 115, 419-434. Campos, E. I., Martinka, M., Mitchell, D. L., Dai, D. L. and Li, G. (2004). Mutations of the ING1 tumor suppressor gene detected in human melanoma abrogate nucleotide excision repair. Int. J. Oncol. 25, 73-80. Chen, J., Santillan, D. A., Koonce, M., Wei, W., Luo, R., Thirman, M. J., ZeleznikLe, N. J. and Diaz, M. O. (2008). Loss of MLL PHD finger 3 is necessary for MLLENL-induced hematopoietic stem cell immortalization. Cancer Res. 68, 6199-6207. Chevaillier, P. (1993). Pest sequences in nuclear proteins. Int. J. Biochem. 25, 479-482. Cimini, D., Mattiuzzo, M., Torosantucci, L. and Degrassi, F. (2003). Histone hyperacetylation in mitosis prevents sister chromatid separation and produces chromosome segregation defects. Mol. Biol. Cell 14, 3821-3833. Coles, A. H. and Jones, S. N. (2009). The ING gene family in the regulation of cell growth and tumorigenesis. J. Cell Physiol. 218, 45-57. Cortes, F. and Pastor, N. (2003). Induction of endoreduplication by topoisomerase II catalytic inhibitors. Mutagenesis 18, 105-112. Dawlaty, M. M., Malureanu, L., Jeganathan, K. B., Kao, E., Sustmann, C., Tahk, S., Shuai, K., Grosschedl, R. and van Deursen, J. M. (2008). Resolution of sister centromeres requires RanBP2-mediated SUMOylation of topoisomerase IIalpha. Cell 133, 103-115. Downes, C. S., Mullinger, A. M. and Johnson, R. T. (1991). Inhibitors of DNA topoisomerase II prevent chromatid separation in mammalian cells but do not prevent exit from mitosis. Proc. Natl. Acad. Sci. USA 88, 8895-8899. Eberharter, A., Vetter, I., Ferreira, R. and Becker, P. B. (2004). ACF1 improves the effectiveness of nucleosome mobilization by ISWI through PHD-histone contacts. EMBO J. 23, 4029-4039. Endl, E. and Gerdes, J. (2000). The Ki-67 protein: fascinating forms and an unknown function. Exp. Cell Res. 257, 231-237. Farber, E. and Roberts, J. J. (1971). Response of cells to inhibition of synthesis of DNA, RNA and protein. Biochem. Pharmacol 20, 1023-1028. Fu, J., Bian, M., Jiang, Q. and Zhang, C. (2007). Roles of Aurora kinases in mitosis and tumorigenesis. Mol. Cancer Res. 5, 1-10. Garcia-Alai, M. M., Gallo, M., Salame, M., Wetzler, D. E., McBride, A. A., Paci, M., Cicero, D. O. and de Prat-Gay, G. (2006). Molecular basis for phosphorylationdependent, PEST-mediated protein turnover. Structure 14, 309-319. Giet, R. and Glover, D. M. (2001). Drosophila aurora B kinase is required for histone H3 phosphorylation and condensin recruitment during chromosome condensation and to organize the central spindle during cytokinesis. J. Cell Biol. 152, 669-682. Hodges, C. A., Revenkova, E., Jessberger, R., Hassold, T. J. and Hunt, P. A. (2005). SMC1beta-deficient female mice provide evidence that cohesins are a missing link in age-related nondisjunction. Nat. Genet. 37, 1351-1355. Inoue, A., Hyle, J., Lechner, M. S. and Lahti, J. M. (2008). Perturbation of HP1 localization and chromatin binding ability causes defects in sister-chromatid cohesion. Mutat. Res. 657, 48-55. Iwase, S., Lan, F., Bayliss, P., de la Torre-Ubieta, L., Huarte, M., Qi, H. H., Whetstine, J. R., Bonni, A., Roberts, T. M. and Shi, Y. (2007). The X-linked mental retardation gene SMCX/JARID1C defines a family of histone H3 lysine 4 demethylases. Cell 128, 1077-1088. Li, B., Gogol, M., Carey, M., Lee, D., Seidel, C. and Workman, J. L. (2007). Combined action of PHD and chromo domains directs the Rpd3S HDAC to transcribed chromatin. Science 316, 1050-1054. Li, H., Ilin, S., Wang, W., Duncan, E. M., Wysocka, J., Allis, C. D. and Patel, D. J. (2006). Molecular basis for site-specific read-out of histone H3K4me3 by the BPTF PHD finger of NURF. Nature 442, 91-95. Mazumdar, M. and Misteli, T. (2005). Chromokinesins: multitalented players in mitosis. Trends Cell Biol. 15, 349-355. Mazumdar, M., Sundareshan, S. and Misteli, T. (2004). Human chromokinesin KIF4A functions in chromosome condensation and segregation. J. Cell Biol. 166, 613-620. Meijer, L., Flajolet, M. and Greengard, P. (2004). Pharmacological inhibitors of glycogen synthase kinase 3. Trends Pharmacol. Sci. 25, 471-480. Mellor, J. (2006). It takes a PHD to read the histone code. Cell 126, 22-24. Mohrmann, G., Hengstler, J. G., Hofmann, T. G., Endele, S. U., Lee, B., Stelzer, C., Zabel, B., Brieger, J., Hasenclever, D., Tanner, B. et al. (2005). SPOC1, a novel PHDfinger protein: association with residual disease and survival in ovarian cancer. Int. J. Cancer 116, 547-554. Mora-Bermudez, F., Gerlich, D. and Ellenberg, J. (2007). Maximal chromosome compaction occurs by axial shortening in anaphase and depends on Aurora kinase. Nat. Cell Biol. 9, 822-831. Nasmyth, K. (2002). Segregating sister genomes: the molecular biology of chromosome separation. Science 297, 559-565. Nourani, A., Doyon, Y., Utley, R. T., Allard, S., Lane, W. S. and Cote, J. (2001). Role of an ING1 growth regulator in transcriptional activation and targeted histone acetylation by the NuA4 complex. Mol. Cell. Biol. 21, 7629-7640. Papait, R., Pistore, C., Grazini, U., Babbio, F., Cogliati, S., Pecoraro, D., Brino, L., Morand, A. L., Dechampesme, A. M., Spada, F. et al. (2008). The PHD domain of Np95 (mUHRF1) is involved in large-scale reorganization of pericentromeric heterochromatin. Mol. Biol. Cell 19, 3554-3563. Patel, S., Doble, B. and Woodgett, J. R. (2004). Glycogen synthase kinase-3 in insulin and Wnt signalling: a double-edged sword? Biochem. Soc. Trans. 32, 803-808.

2956


Ramon-Maiques, S., Kuo, A. J., Carney, D., Matthews, A. G., Oettinger, M. A., Gozani, O. and Yang, W. (2007). The plant homeodomain finger of RAG2 recognizes histone H3 methylated at both lysine-4 and arginine-2. Proc. Natl. Acad. Sci. USA 104, 1899318998. Reader, J. C., Meekins, J. S., Gojo, I. and Ning, Y. (2007). A novel NUP98-PHF23 fusion resulting from a cryptic translocation t(11;17)(p15;p13) in acute myeloid leukemia. Leukemia 21, 842-844. Rechsteiner, M. and Rogers, S. W. (1996). PEST sequences and regulation by proteolysis. Trends Biochem. Sci. 21, 267-271. Rohaly, G., Chemnitz, J., Dehde, S., Nunez, A. M., Heukeshoven, J., Deppert, W. and Dornreiter, I. (2005). A novel human p53 isoform is an essential element of the ATRintra-S phase checkpoint. Cell 122, 21-32. Scholzen, T. and Gerdes, J. (2000). The Ki-67 protein: from the known and the unknown. J. Cell Physiol. 182, 311-322. Shi, X., Hong, T., Walter, K. L., Ewalt, M., Michishita, E., Hung, T., Carney, D., Pena, P., Lan, F., Kaadige, M. R. et al. (2006). ING2 PHD domain links histone H3 lysine 4 methylation to active gene repression. Nature 442, 96-99. Shi, X., Kachirskaia, I., Walter, K. L., Kuo, J. H., Lake, A., Davrazou, F., Chan, S. M., Martin, D. G., Fingerman, I. M., Briggs, S. D. et al. (2007). Proteome-wide analysis in Saccharomyces cerevisiae identifies several PHD fingers as novel direct and selective binding modules of histone H3 methylated at either lysine 4 or lysine 36. J. Biol. Chem. 282, 2450-2455. Somma, M. P., Fasulo, B., Siriaco, G. and Cenci, G. (2003). Chromosome condensation defects in barren RNA-interfered Drosophila cells. Genetics 165, 1607-1611. Storchova, Z., Breneman, A., Cande, J., Dunn, J., Burbank, K., O’Toole, E. and Pellman, D. (2006). Genome-wide genetic analysis of polyploidy in yeast. Nature 443, 541-547.

Tanaka, Y., Nakayama, Y., Taniguchi, M. and Kioussis, D. (2008). Regulation of early T cell development by the PHD finger of histone lysine methyltransferase ASH1. Biochem. Biophys. Res. Commun. 365, 589-594. Ullah, M., Pelletier, N., Xiao, L., Zhao, S. P., Wang, K., Degerny, C., Tahmasebi, S., Cayrou, C., Doyon, Y., Goh, S. L. et al. (2008). Molecular architecture of quartet MOZ/MORF histone acetyltransferase complexes. Mol. Cell. Biol. 28, 6828-6843. Van Gele, M., Van Roy, N., Ronan, S. G., Messiaen, L., Vandesompele, J., Geerts, M. L., Naeyaert, J. M., Blennow, E., Bar-Am, I., Das Gupta, T. K. et al. (1998). Molecular analysis of 1p36 breakpoints in two Merkel cell carcinomas. Genes Chromosomes Cancer 23, 67-71. Varga, A. E., Dobrovic, A., Webb, G. C. and Hutchinson, R. (2001). Clustering of 1p36 breakpoints distal to 1p36.2 in hematological malignancies. Cancer Genet. Cytogenet. 125, 78-79. Yochum, G. S. and Ayer, D. E. (2001). Pf1, a novel PHD zinc finger protein that links the TLE corepressor to the mSin3A-histone deacetylase complex. Mol. Cell. Biol. 21, 4110-4118. Yong-Gonzalez, V., Wang, B. D., Butylin, P., Ouspenski, I. and Strunnikov, A. (2007). Condensin function at centromere chromatin facilitates proper kinetochore tension and ensures correct mitotic segregation of sister chromatids. Genes Cells 12, 1075-1090. Ythier, D., Larrieu, D., Brambilla, C., Brambilla, E. and Pedeux, R. (2008). The new tumor suppressor genes ING: genomic structure and status in cancer. Int. J. Cancer 123, 1483-1490. Zhang, Y. (2006). It takes a PHD to interpret histone methylation. Nat. Struct. Mol. Biol. 13, 572-574. Zhou, B. P., Deng, J., Xia, W., Xu, J., Li, Y. M., Gunduz, M. and Hung, M. C. (2004). Dual regulation of Snail by GSK3beta-mediated phosphorylation in control of epithelialmesenchymal transition. Nat. Cell Biol. 6, 931-940.