KEIJO VIIRI

The Sin3-Associated 30 (SAP30) Family of Transcriptional Regulators

ACADEMIC DISSERTATION To be presented, with the permission of the Faculty of Medicine of the University of Tampere, for public discussion in the Small Auditorium of Building B, Medical School of the University of Tampere, Medisiinarinkatu 3, Tampere, on May 8th, 2009, at 12 o’clock.

UNIVERSITY OF TAMPERE ACADEMIC DISSERTATION University of Tampere, Medical School Tampere University Hospital, Department of Paediatrics Tampere Graduate School in Biomedicine and Biotechnology (TGSBB) Finland

Supervised by Reviewed by Professor Markku Mäki Professor Lea Sistonen University of Tampere Åbo Akademi University Finland Finland Olli Lohi, MD, PhD Docent Sami Väisänen University of Tampere University of Kuopio Finland Finland

Distribution Tel. +358 3 3551 6055 Bookshop TAJU Fax +358 3 3551 7685 P.O. Box 617 [email protected] 33014 University of Tampere www.uta.fi/taju Finland http://granum.uta.fi

Cover design by Juha Siro

Acta Universitatis Tamperensis 1396 Acta Electronica Universitatis Tamperensis 824 ISBN 978-951-44-7660-0 (print) ISBN 978-951-44-7661-7 (pdf) ISSN-L 1455-1616 ISSN 1456-954X ISSN 1455-1616 http://acta.uta.fi

Tampereen Yliopistopaino Oy – Juvenes Print Tampere 2009 To my eternal promoters and loved ones Leena, Pyry and Paavo CONTENTS

LIST OF ORIGINAL COMMUNICATIONS ...... 7 ABBREVIATIONS...... 8 ABSTRACT...... 10 TIIVISTELMÄ ...... 12 INTRODUCTION...... 14 REVIEW OF THE LITERATURE ...... 15 1. Cell nucleus...... 15 1.1. Nucleolus...... 16 1.2. Other nuclear bodies and regions...... 16 1.3. Nuclear matrix (nuclear scaffold, nucleoskeleton, karyoskeleton) ...... 18 2. Chromatin...... 20 2.1. Transcriptionally active euchromatin...... 21 2.2. Transcriptionally silent heterochromatin...... 22 3. Posttranslational modifications regulate DNA-related processes...... 24 3.1. Histone tail modification...... 25 3.2. ’Histone code’ hypothesis ...... 26 3.3. Histone and deacetylation ...... 28 3.3.1. Histone deacetylases...... 30 3.3.2. HDACs in yeast genetic screens...... 31 3.3.3. HDAC inhibitors as drugs...... 33 4. The Sin3-HDAC corepressor complex...... 34 4.1. Sin3A protein...... 35 4.2. The core Sin3-HDAC complex...... 36 4.2.1. Sin3 associated protein 30 (SAP30)...... 37 4.3. SAP30/HDAC complexes in diseases ...... 39 4.3.1. SAP30 in cancer...... 39 4.3.2. SAP30 protein is a cofactor in virus transmission...... 40 5. Phosphoinositides (PtdInsP) - messengers of cytosolic and nuclear signaling in the cell...... 41 6. Zinc-dependent protein structures...... 44 AIMS OF THE STUDY...... 45 MATERIALS AND METHODS ...... 46 1. Three-dimensional T84 epithelial cell model for the jejunal crypt-villus axis (I)...... 46 2. Cell cultures and transfections (I - IV) ...... 46 3. RNA isolation and detection methods (I) ...... 47 3.1. RNA isolation and differential display PCR...... 47 3.2. Quantitative PCR...... 47 3.3. Screening of cDNA library for the whole-length transcript...... 47 4. cDNA cloning and protein production...... 48 4.1. cDNA cloning (I - III) ...... 48 4.2. Production of GST-fusion in E. coli (II & III) ...... 48 4.3. Protein production by coupled in vitro transcription/translation (II & III) ...... 48

4 4.4. Protein production and expression in mammalian cells (I – IV)...... 48 5. Protein functional studies...... 49 5.1. Protein detection: immunoblotting and immunofluorescence (I – IV) ...... 49 5.2. Protein-protein interaction studies (II & III) ...... 49 5.2.1. Immunoprecipitations...... 49 5.2.2. GST-pull-downs...... 49 5.3. Protein-nucleic acid interaction studies (III & IV) ...... 50 5.3.1. Electrophoretic mobility shift assays (EMSA) (III)...... 50 5.3.2. Novel ladder-EMSA (L-EMSA) (III)...... 50 5.3.3. Interphase chromatin spreads (III)...... 50 5.3.4. Chromatin isolation/ subcellular fractionation (III & IV)...... 51 5.4. Protein-lipid interaction studies (III)...... 51 5.5. DNA-bending assay/ligation-mediated circulization assay (III) ...... 51 5.6. Nucleosome preparations (III & IV)...... 51 5.7. HDAC activity and repression studies (II & III)...... 51 5.8. Mass spectrometric analysis of the N-terminal SAP30L peptides (III)...... 52 5.9. Protein binding microarray (PBM) experiments and data analysis (III) ...... 52 5.10. Nuclear matrix preparations (IV) ...... 53 6. Phylogenetic and molecular evolution studies (IV)...... 53 6.1. Protein sequence searches, gene loci data retrieval and multiple sequence alignments...... 53 6.2. Phylogenetic analysis and detection of functional divergence...... 53 RESULTS...... 55 1. Identification of SAP30L in differentiated T84 cells (I)...... 55 2. Identification of SAP30L as a member of the Sin3A corepressor complex (II) ...... 56 2.1. SAP30L associates with HDACs and represses transcription ...... 57 3. Identified domains and functional motifs in SAP30L and SAP30 proteins ...... 58 3.1. Nuclear localization signal (NLS) (I & II)...... 58 3.2. Nucleolar localization signal (NoLS) (II) ...... 59 3.3. Protein-protein interaction domain (II) ...... 59 3.4. Zinc-dependent structure (III)...... 60 3.4.1. Sequence-independent DNA binding and bending (III)...... 61 3.4.2. Monophosphoinositides (PtdInsP) binding domain (III)...... 62 3.5. Acidic central domain contributing to histone interaction (III)...... 63 3.6. Nuclear matrix targeting signal (IV) ...... 64

5 4. The subcellular localization, chromatin attachment and repressional activity of SAP30L is regulated by its interactions with DNA and monophosphoinositides (III) ...... 64 5. Evolution of the SAP30 family of transcriptional regulators (IV)...... 65 DISCUSSION ...... 67 1. The domain structure of the SAP30 family proteins indicates nuclear scaffolding and transcriptional regulatory functions (I - IV) ...... 67 2. Evolution of the SAP30 family (IV)...... 69 3. Novel proposed mechanism: Regulation of protein-DNA interactions by nuclear phospholipids (III)...... 70 4. Inhibition of disease-associated HDAC complexes...... 73 CONCLUSIONS AND FUTURE PROSPECTS ...... 74 ACKNOWLEDGEMENTS ...... 77 REFERENCES...... 79 ORIGINAL COMMUNICATIONS...... 95

6 LIST OF ORIGINAL COMMUNICATIONS

This thesis is based on the following original communications, referred to in the text by their Roman numerals I-IV. Publications II and III are reprinted with copyright permissions from Oxford University Press and American Society for Microbiology, respectively

I Lindfors K, Viiri KM, Niittynen M, Heinonen T, Mäki M & Kainulainen H. (2003): TGF-ȕ induces the expression of SAP30L, a novel nuclear protein. BMC Genomics. Dec 18;4(1):53. *

II Viiri KM, Korkeamäki H, Kukkonen M, Nieminen LK, Lindfors K, Peterson P, Mäki M, Kainulainen H & Lohi O. (2006): SAP30L interacts with members of the Sin3A corepressor complex and targets Sin3A to the nucleolus. Nucleic Acids Res. Jul 4;34(11): 3288 - 3298.

III Viiri KM, Jänis J, Siggers T, Heinonen T, Valjakka J, Mäki M, Bulyk ML, Lohi O. (2009): DNA-binding and -bending activities of SAP30L and SAP30 are mediated by a zinc-dependent module and monophosphoinositides. Mol Cell Biol. Jan;29(2): 342-356.

IV Viiri KM, Heinonen T, Mäki M & Lohi O.: Phylogenetic analysis of the SAP30 family of transcriptional regulators reveals functional divergence and conserved nuclear scaffolding function. Submitted.

* This article has been used in the PhD thesis of Katri Lindfors

7 ABBREVIATIONS

ATP adenosine triphosphate BSA bovine serum albumine C1 domain conserved region-1 (from protein kinase C) (binds DAG) cDNA complementary DNA CEF chromatin enriched fraction CIR CBF1-interacting corepressor Co-IP co-immunoprecipitation DAG diacylglycerol DAPI 4',6-diamidino-2-phenylindole DD-PCR differential display- polymerase chain reaction DMSO dimethyl sulfoxide DNA dexoyribonucleic acid DNMT DNA methyltransferase EDTA ethylenediaminetetraacetic acid EMSA electrophoretic mobility shift assay FCS fetal calf serum FYVE ‘Fab1, YOTB, Vac, EEA1’ (shared domain in these proteins) GFP green fluorescent protein GST glutathione-S-transferase GTF general transcription factors HAT histone acetyltransferase HDAC HID histone interacting domain HMG high mobility group HMR Hidden MAT Right hnRNP heterogeneous nuclear ribonucleoproteins Hog1 high osmolarity glycerol response 1 HRP horseradish peroxidase ING inhibitor of growth (proteins) LacZ LacZ encodes beta-galactosidase L-EMSA ladder electrophoretic mobility shift assay LUC luciferase enzyme MeCP2 methyl CpG binding protein 2 Mof maintenance of frame mRNA messenger RNA NAD nicotinamide adenine dinucleotide NcoR nuclear receptor corepressor NES nuclear export signal NLS nuclear localization signal NMTS nuclear matrix targeting signal NMR nuclear magnetic resonance NoLS nucleolar localization signal

8 NOR nucleolar organizer region NPM nucleophosmin OPT domain (Oct1/PTF/Transcription) domains ORF open reading frame PAH paired amphipathic helix (in Sin3 proteins) PBM protein binding microarray PBS phosphate buffered saline PCR polymerase chain reaction PEV position effect variegation PH pleckstrin homology PHD plant homeodomain PI phosphoinositides PIP phosphatidylinositol phosphate PML promyelocytic protein PNC perinucleolar compartment PtdInsP phosphoinositides PTEN phosphatase and tensin homologue on choromosome ten PTM posttranslational modification PX phox-homology RbAp46 retinoblastoma protein-associated protein 46 RbAp48 retinoblastoma protein-associated protein 48 rDNA ribosomal DNA RNA ribonucleic acid RNA pol II RNA polymerase II RNP ribonucleoprotein Rpd reduced potassium dependency rRNA ribosomal RNA RT-PCR reverse transcriptase PCR RVFV rift valley fever virus S/MAR scaffold/matrix-attachment regions SAP18 Sin3-associated protein 18 SAP30 Sin3-associated protein 30 SAP30L Sin3-associated protein 30-like SDI SWI-dependent interconversion SDS suppressor of defective silencing SDS3 suppressor of defective silencing 3 SDS-PAGE sodium dodecyl sulfate polyacrylamide gel electrophoresis SHIP SH2-containing inositol 5’-phosphatase Sin3A SWI-independent 3a Sin3B SWI-independent 3b SIR silent information regulators snoRNP small nucleolar ribonucleoprotein snRNA small nuclear RNA snRNP small nuclear ribonucleoprotein TGF-beta transforming growth factor –beta TLE transducin-like enhancer TRK transport of potassium (K) TSA trichostatin A YY1 Yin Yang 1

9 ABSTRACT

BACKGROUND: The transcription of is influenced by their packing status around the nucleosome: tightly packed DNA is inaccessible for RNA polymerase enzymes whereas loosely packed DNA is more efficiently transcribed. The most appreciated histone modification governing the DNA packing is histone tail acetylation and deacetylation. Deacetylation of plays a fundamental role in gene silencing by producing deacetylated nucleosomes, where DNA is wrapped more tightly. Deacetylation is often mediated by a corepressor complex (consisting of seven to eight proteins) containing SWI-independent 3A (Sin3A) as an essential scaffold protein. Sin3-Associated Protein 30 (SAP30), has previously been suggested to function as a linker molecule between various corepressors. However, the domain structure and the molecular function of SAP30 protein have remained unknown. AIMS: The purpose of this study was to identify regulatory proteins in the differentiating intestinal epithelial cells and to further decipher their function in closer molecular detail. RESULTS: We identified a novel transforming growth factor-beta (TGF-beta)- upregulated mRNA species in a mesenchymal-epithelial cell co-culture model which mimics the intestinal crypt villus axis biology in terms of epithelial cell differentiation. mRNA was ubiquitously expressed in various tissues and encoded a nuclear localizing protein with 70% identity with SAP30 and we thus named it as SAP30-like (SAP30L) (I). SAP30L interacts with several components of the Sin3A corepressor complex and binds to the PAH3/HID (Paired Amphipathic Helix 3/Histone deacetylase Interacting Domain) region of Sin3A. Like SAP30, when tethered to promoters, SAP30L induces strong transcriptional repression in a manner dependent on Sin3A and histone deacetylases. We discovered a functional nucleolar localization signal in SAP30L and showed that SAP30L and SAP30 (hereafter called SAP30 proteins) are able to target Sin3A to the nucleolus (II). In further structure-function mapping of the SAP30 proteins we identified a zinc-coordinating structure which is necessary for DNA binding and found that one consequence of binding is bending of the DNA. We also showed that nuclear signaling lipids, phosphoinositides (PIs), bind to the same region in the SAP30 proteins as DNA. In fact, PIs and DNA evince an antagonizing interrelationship in regard to their binding to the SAP30 proteins, since binding of PIs detaches SAP30 proteins from the chromatin. PI binding also reduces the repressive activity of the SAP30 proteins and affects their translocation from the nucleus to the cytoplasm. In addition, the repressional activity of the SAP30 proteins is partly dependent on their direct interaction with the globular domain of the core histones and nucleosomes (III). Our molecular evolutionary analysis indicated that SAP30L is the ancestral protein of the SAP30 protein family and that SAP30 originated by a segment

10 duplication which occurred after the divergence of Actinopterygii (ray-finned fishes) and Sarcopterygii (flesh-finned fishes) about 450 million years ago. Phylogenetic analysis and biochemical experiments suggested that SAP30 has diverged functionally from the ancestral SAP30L by accumulating mutations causing attenuation of one of the original functions, association with the nuclear matrix (IV). CONCLUSIONS: First, we identified a novel epigenetic regulator protein, SAP30L, in differentiating human intestinal epithelial cells. In general, SAP30L might have a role in recruiting the Sin3-histone deacetylase complex to specific corepressor subcomplexes in response to TGF-beta, thus leading to the silencing of proliferation-driving genes in these differentiating intestinal epithelial cells. Our detailed results suggest that SAP30L and SAP30 mediate gene repression through multiple interactions, i.e. with Sin3A, histone deacetylases (HDACs), DNA and histones/nucleosomes. Furthermore, the ability of the SAP30 family proteins to bend DNA suggests that they are active partners in complexes regulating . This work describes multifarious interactions of SAP30 proteins in the promoter-bound Sin3 repressome which can be utilized in drug design to destabilize SAP30-containing Sin3 complexes in diseases wherever they are presumably implicated (e.g. cancer and viral infections). Moreover, the findings here shed light on an ambiguous, already half-century ago characterized, nuclear phospholipid component of the cell by proposing that nuclear signaling lipids, phosphoinositides (PIs), regulate protein-DNA interactions. Thus, part of the PI signaling-mediated changes in gene expression are probably due to the ability of certain proteins (such as the SAP30 proteins) to sense the nuclear PI content and to detach from chromatin when a threshold concentration of PI is reached. Finally, PI-mediated inhibition of protein-DNA interactions is discussed in terms of its potential as a novel strategy for inhibiting pathogenic protein-chromatin interactions and its utilization in combinatorial therapies with e.g. HDAC inhibitors.

11 TIIVISTELMÄ

TAUSTA: Geenien luennan eli transkription vilkkaus on riippuvainen mm. siitä miten DNA on pakkautunut kromatiiniksi histoni-proteiineista koostuvien nukleosomien ympärille: Geenien lukijaentsyymit, RNA-polymeraasit, lukevat helposti löyhästi pakattua DNA:ta kun taas tiukkaan pakattuihin geeneihin ne eivät pääse hyvin käsiksi. Histoni-häntien modifikaatioiden, kuten asetylaation ja deasetylaation, tiedetään vaikuttavan DNA:n pakkautumiseen nukleosomien ympärille. Deasetyloidun histonin ympärille DNA pakkautuu tiiviimmin ja lähes poikkeuksetta vaimennetut geenit sijaitsevatkin kromosomi-alueilla, joissa histonit ovat deasetyloitu. SWI-independent 3A (Sin3A)-korepressorikompleksi on hyvin tunnettu histonideasetylaatiota välittävä proteiinikompleksi, jonka jäsenet (8-9 kpl) ovat koordinoidusti sitoutuneet Sin3A-proteiinin ympärillä. Sin3 Associated Protein 30 (SAP30) on yksi jäsenistä ja sen on ehdotettu välittävän interaktioita muiden korepressoreiden kanssa. Kuitenkin, SAP30-proteiinin tehtävä ja toiminnallinen domeenirakenne on tuntematon. TARKOITUS: Tämän työn tarkoituksena oli etsiä erilaistuvista suolen epiteelisoluista mahdollisia säätelyproteiineja ja selvittää niiden toimintaa tarkemmin myös molekyylitasolla. TULOKSET: Löysimme uuden lähetti-RNA:n, transformoiva kasvutekijä-betalla (TGF-beta) erilaistetuista epiteelisoluista suolen krypta-villus akseli-solumallissa. Useiden ihmiskudosten havaittiin ekspressoivan ko. lähetti-RNA:a ja sen koodittama proteiini lokalisoitui solun tumaan ja oli 70 % identtinen SAP30:n kanssa ja siksi nimesimme sen ”SAP30-like” (SAP30L) (I). SAP30L:n havaittiin sitoutuvan useiden Sin3A-kompleksin jäsenien kanssa ja itse Sin3A-proteiiniin sen PAH3/HID (Paired Amphipathic Helix 3/Histone deacetylase Interacting Domain) - alueeseen. Kuten SAP30, myös SAP30L kykeni tehokkaasti vaimentamaan geenejä, kun se oli sidottu kohdegeenin säätelyalueeseen (promoottori). Lisäksi SAP30L- välitteinen geenien vaimennus oli riippuvainen Sin3A:sta ja histonideasetylaatio- aktiivisuudesta. Karakterisoimme SAP30L-proteiinista ns. tumajyväslokalisaatiosignaalin ja havaitsimme SAP30L ja SAP30 (kutsutaan tästedes yhteisnimellä SAP30-proteiinit) kohdistavan Sin3A:n tumajyväseen (II). Teimme tarkempaa rakenne-toiminta kartoitusta SAP30-proteiineilla ja havaitsimme niissä sinkki-atomi-koordinoidun rakenteen joka vastasi DNA:n sitomisesta ja taivuttamisesta. Havaitsimme myös, että tietyt tuman signaalilipidit, fosfoinositidit (PI), sitoutuvat samaan paikkaan SAP30-proteiineissa kuin DNA. Itse asiassa, PI- ja DNA-molekyylien välillä havaittiin antagonistinen suhde niiden sitoutumisessa SAP30-proteiineihin koska PI syrjäytti DNA:n ja tämän seurauksena SAP30- proteiinit irtoavat kromatiinista. PI:n sitoutumisesta SAP30-proteiineihin seurasi geenien vaimennuskyvyn heikkeneminen ja SAP30-proteiinit lokalisoituivat tuman sijasta enemmän solulimaan. Lisäksi SAP30-proteiinien repressiokyky oli osittain

12 riippuvainen niiden kyvystä sitoutua histonien ja näistä koostuvien nukleosomien globulaarisiin keskusalueisiin (III). Molekyylievolutiiviset tutkimuksemme osoittavat, että SAP30-proteiini sai alkunsa SAP30L:n sisältävän kromosomi- segmentin monistumisen myötä. Monistuminen ajoittui n. 450 miljoonan vuoden päähän, luokkien Actinopterygii (viuhkaeväiset) ja Sarcopterygii (varsieväiset) eriytymisen jälkeiseen ajankohtaan. Fylogeneettiset analyysimme ehdottavat, että SAP30 on toiminnallisesti eriytynyt kantamuodostaan SAP30L:stä ja kokeellisesti havaitsimmekin, että perheen alkuperäinen funktio, assosioituminen tuman matriksiin oli SAP30:llä merkittävästi heikompaa kuin SAP30L:llä (IV). JOHTOPÄÄTÖKSET: Ensiksi, löysimme uuden epigeneettisen säätelyproteiinin, SAP30L, ihmisen ohutsuolen erilaistuvista epiteelisoluista. Yleisesti ottaen SAP30L-proteiini ehkä värvää Sin3-histonideasetylaatiokomplekseja, TGF-betalla erilaistetuissa suolen epiteelisoluissa, spesifisiin ala-korepressorikomplekseihin, jotka hiljentävät solujen lisääntymiseen osallistuvia geenejä. Yksityiskohtaisemmat tuloksemme osoittavat, että SAP30L ja SAP30 vuorovaikuttavat Sin3A:n, histonideasetylaasien (HDAC), DNA:n ja histoneiden ja nukleosomien kanssa ja tämän moninaisen sitomiskyvyn ansiosta ne kykenevät hiljentämään geenejä. Lisäksi SAP30-proteiinien kyky taivuttaa DNA:ta viittaa siihen, että ne toimivat aktiivisesti geeninsäätelykomplekseissa. Tässä työssä kuvataan SAP30 proteiinien moninaiset interaktiot promoottoriin sitoutuneessa Sin3-repressomissa. Viime aikoina on tullut uutta tietoa SAP30 sisältävien Sin3-kompleksien mahdollisesta osallisuudesta sairauksiin (syöpä ja virusten aiheuttamat infektiot) ja näin ollen tietoa näistä interaktioista voidaan mahdollisesti hyödyntää lääkesuunnittelussa, joiden tarkoituksena on destabiloida SAP30-Sin3-komplekseja. Lisäksi, tämä työ valottaa uudella tavalla tuman fosfolipidien roolia - komponentti joka on tunnettu jo puoli vuosisataa, mutta jonka toiminta on edelleen epäselvä. Ehdotamme, että tuman signaalilipit, fosfoinositidit (PI:t), säätelevät proteiini-DNA-interaktioita. Täten ehkä osa PI signaloinnin aiheuttamista geeniekspression muutoksista on itse asiassa seurausta siitä, että jotkin DNA:ta sitovat geenien luennan säätelyproteiinit (kuten SAP30-proteiinit) kykenevät tunnustelemaan tuman PI:n määrää ja irtoavat kromatiinista kun tietty PI:n kynnyspitoisuus ylittyy. Tässä työssä löydetty PI- välitteinen proteiini-DNA interaktion estäminen on uusi potentiaalinen strategia estää patogeenisia proteiini-kromatiini interaktioita, minkä käyttöä voisi ajatella sairauksien yhdistelmähoitona esim. HDAC-inhibiittoreiden kanssa.

13 INTRODUCTION

In a narrow sense, the term epigenetics is defined as the study of mitotically and/or meiotically heritable changes in phenotype or gene function which cannot be explained by changes in the underlying DNA sequence (Russo et al. 1996, Bird 2007). Hence the name epi - "in addition to" - genetics. Due to its immense growth as a study subject in contemporary biology, the term epigenetics has confronted inflationary pressure and is now often broadly used to define all epigenetic mechanisms despite whether they are passed on to the next cell generation or not. An illustrative example of epigenetics in action is to be found in the process of cellular differentiation. During morphogenesis, totipotent stem cells evolve into the various pluripotent cell lines of the embryo, which in turn become fully differentiated cells. The genomes in all, in stem cells and in various fully differentiated adult cells, are identical (except in cells involved in the immune system e.g. V(D)J recombination). The distinct fates of the cells are demarcated by differential gene expression via epigenetic mechanisms. Epigenetic marks, printed either directly in the DNA or in the histone proteins in which DNA is wrapped, influence gene expression. DNA methylation is the archetypical repressive epigenetic mark which is transmitted by mitotic inheritance in both plants and animals (Goll and Bestor 2005). By contrast, histone modifications such as acetylation, methylation and phosphorylation associated with transcription remain ambiguous with respect to heritability. On the other hand, DNA methylation affects histone acetylation and methylation and thus these modifications can be viewed as heritably epigenetic, albeit indirectly (Klose and Bird 2006). For instance, the link between DNA methylation and histone deacetylation is provided by methyl CpG binding protein 2 (MeCP2) which binds methylated DNA and recruits the Sin3 histone deacetylase corepressor complex to modify histones and repress transcription (Figure 6) (Jones et al. 1998). The purpose of this study was to identify regulatory proteins in the differentiating intestinal epithelial cells upon TGF-beta treatment. The focus in this work is on deciphering the molecular function of a novel epigenetic transcriptional regulator, Sin3 Associated Protein 30 – Like (SAP30L) and also its close homolog, SAP30.

14 REVIEW OF THE LITERATURE

1. Cell nucleus

The nucleus (pl. nuclei; from Latin nucleus or nuculeus, "little nut" or kernel) was the first cell organelle to be discovered, and was first described by Franz Bauer, an Austrian botanical artist, in 1804. The first cue of its important function came from Oscar Hertwig’s studies between 1876 and 1878 on the fertilization of sea urchin eggs, showing that the nucleus of the sperm enters the oocyte and fuses with its nucleus. This observation gave impulse to the conception that an individual develops from a (single) nucleated cell. Since Hertwig’s suggestion confronted vast rebuttals among his contemporary colleagues he had to confirm his observation also in other animal organisms e.q. amphibians and molluscs. When Eduard Strasburger produced the same results for plants (1884) the way was paved to assign the nucleus an important role in heredity. Insight into how the nucleus contributes to heredity came only at the beginning of the 20th century, when the Mendelian rules were rediscovered and the chromosome theory of heredity was developed (Harris 2000). Almost all the DNA (except mitochondrial and chloroplastidial DNA) in the eukaryotic cell is packed in the nucleus, which occupies about 10% of the total cell volume. The nucleus is delimited by a nuclear envelope, which is a concentric lipid bilayer membrane punctured at intervals by large nuclear pores. The nuclear envelope is directly connected to the membranes of the endoplasmic reticulum in the cytosol. Thus, the nucleus and the cytosol are topologically continuous but functionally distinct. This topological link has led evolutionary biologists to the conception that in some ancient bacteria the single DNA molecule was attached to an invagination in the plasma membrane. Subsequently, in a very ancient prokaryotic cell such an invagination was completely enveloped and pinched off from the plasma membrane, producing a nuclear compartment surrounded by a double membrane (Alberts 2002).

15 1.1. Nucleolus

The nucleolus (pl. nucleoli) is the most prominent sub-compartment of the nucleus, as reviewed by Boisvert et al. (2007). It was already found over 200 years ago. Since the nucleolus is not compartmentalized by a membrane, it is not by definition regarded as a cellular organelle. Its main function is the transcription and processing of ribosomal RNA (rRNA) and the assembly of ribosomes, although other functions, such as ribonucleoprotein (RNP) assembly, cell cycle control, messenger RNA (mRNA) maturation, stress response and protein sequestration, have recently been attributed to it (Bernardi et al. 2004, Shaw and Brown 2004). Principally its manifestation in the nucleus is consequent upon its function and therefore nucleoli are very dynamic in numbers ranging from one to ten depending on the step of the cell cycle. The fluctuation in the number of nucleoli per cell is explained by their genesis: at the end of mitosis (cell division) nucleoli form around the tandemly repeated clusters of ribosomal DNA (rDNA) in acrocentric 13, 14, 15, 21 and 22 to form nucleolar organizer regions (NORs), thus yielding ten nucleating NORs in the diploid . The size of the individual nucleolus is subject to change and correlates with the proliferation pace of the cell. In fact, malignant cells more frequently display a larger nucleolus than benign cells (Busch et al. 1963) Typically, a high proliferation rate correlates with large nucleoli and it has been shown that nucleolar size represents a morphological parameter of the cell proliferation rate also in cancer tissues (Derenzini et al. 2000).

1.2. Other nuclear bodies and regions

The perinucleolar compartment (PNC) was found by Ghetti et al. (1992) in immunofluorescent staining studies with heterogeneous nuclear ribonucleoprotein I (hnRNP I), the polypyrimidine tract-binding protein. Huang and colleagues screening the PNC in a large number of human cancer and normal cells, showed that PNCs are much more prevalent in cancer cells than in normal cells (Huang et al. 1997). The function of PNC is currently unknown, but the presence of hnRNP proteins, splicing factors and small nuclear RNAs (snRNAs) transcribed by RNA polymerase III suggests a role for this compartment in RNA processing (Wang et al. 2003). In fact, the presence of RNA is prerequisite for the structural integrity of PNC, since RNAse treatment (Huang 2000) and inhibition of RNApol III activity (Wang et al. 2003) abolishes PNCs. Sam68 nuclear structures were first visualized in HeLa cells using antibodies directed against the Src Associated in Mitosis 68 kDa protein (Sam68) (Chen et al. 1999). HeLa cells usually have two to three Sam68 bodies per cell nucleus. While they are frequently located in close proximity to the nucleolus, which is reminiscent of the PNC, Sam68 and PNC have different nuclear localizations (Wang et al. 2003). Nonetheless, both Sam68 bodies and PNCs share some common characteristics: both are believed to be involved in RNA metabolism and are observed mostly in transformed cells (Huang 2000). Splicing speckles/interchromatin clusters. In 1961 J. Swanson Beck, introduced the term 'speckles' when he examined rat-liver sections which had been immunolabelled with

16 the serum of individuals with autoimmune disorders and saw speckle-like structures in the nuclei (Beck 1961). Although the connection was not made at the time, two years earlier Hewson Swift had identified these speckles at electron-microscope level and named them interchromatin particles (Swift 1959). They are thought to be sites of storage of mRNA splicing factors (Lamond and Spector 2003), and these splicing factors are then recruited from the speckles to the sites of transcription (Misteli et al. 1997). Furthermore, speckles have been purified and proteomic analysis has identified many proteins linked to pre-mRNA splicing (Misteli et al. 1997, Saitoh et al. 2004). Paraspeckles are often found adjacent to speckles and a similar function is proposed since they contain proteins such as PSP1 and p54nrp, which have an evolutionary conserved role in pre-mRNA processing (Fox et al. 2002). Cajal bodies and Gems are typically found together paired or juxtaposed and the current view is that they are two manifestations of the same structure. Cajal bodies participate in the biogenesis of small nuclear ribonucleoproteins (snRNPs) and in the trafficking of small nucleolar ribonucleoproteins (snoRNPs) and snRNPs, which move through the Cajal bodies en route to nucleoli or splicing speckles, respectively (Sleeman and Lamond 1999). The presence of 2'-O-methylation and pseudouridine machinery in Cajal bodies strongly supports the conception that they constitute the site at which snRNA and rRNA maturation occurs (Ogg and Lamond 2002). Cleavage bodies contain several factors specifically involved in the cleavage and polyadenylation steps of pre-mRNA processing and they usually overlap or are adjacent to Cajal bodies (Schul et al. 1996). OPT domains (Oct1/PTF/Transcription) are transcriptionally active sites in the nucleus rich in PTF, Oct1, TBP, SP1, and RNA pol II (Grande et al. 1997, Pombo et al. 1998) reviewed in (Spector 2001). Although the precise function of these bodies is unknown, it has been proposed that OPT domains and nucleoli have an analogous function. The nucleolus is rich in PolI ‘factories’ and each nucleolus can associate with rDNA genes in several chromosomes. Analogously, OPT domains are rich in PolII and III factories and they tend to associate with specific chromosomes (significantly most with chromosomes 6 and 7) (Pombo et al. 1998). Polycomb or PcG bodies are sites of accumulation of polycomb group (PcG) proteins (Alkema et al. 1997), usually localized close to constitutive heterochromatin (Saurin et al. 1998). Polycomb group proteins were first identified in Drosophila melanogaster and they function in maintaining cellular memory for the transcriptional repression of the hometic genes and thus the posterior-anterior axis of developing larvae (Alberts 2002). PML bodies contain promyelocytic leukemia protein (PML), first identified by its fusion to the retinoic acid receptor alpha (RARa) in translocation t(15,17) associated with acute promyelocytic leukemia (Weis et al. 1994). Debate on the function of PML bodies is still vigorous and it is proposed to be involved in transcription, DNA repair, viral defense, stress, cell cycle regulation, proteolysis and apoptosis, as reviewed by Borden (2002). PML-/- mice are viable but show impaired commitment to apoptosis, indicating that PML is a tumor suppressor, mediating multiple apoptotic signals (Wang et al. 1998). Nuclear lamina/inner nuclear membrane is the protein filament scaffold surrounding the nuclear periphery. It is composed mostly of type V intermediate

17 filament proteins lamin A/C and B. Various functions have been attributed to the nuclear lamina, including maintenance of nuclear shape, spatial organization of nuclear pores within the nuclear membrane, regulation of transcription, organization of interphase heterochromatin, as well as a role in DNA replication, as reviewed in (Foisner 2001, Wilson et al. 2001). Lamins are also involved in transcriptional repression: Lamin A/C is responsible for the correct localization of retinoblastoma repressor (RB) protein and A-type lamins protect RB from proteasomal degradation (Johnson et al. 2004). Nuclear lamina components have been proposed to interact with many transcriptional repressors such as HDAC 3 (Somech et al. 2005) and HP1 (Polioudaki et al. 2001). The nuclear pore complex in higher organisms comprises over 100 nucleoporin proteins and it is a complex of 125 MDa in size, as reviewed in (Nakielny and Dreyfuss 1999). The nuclear pore complex machinery is responsible for transporting a vast number of molecules (proteins and RNA-protein complexes, RNPs) in and out of the nucleus in a rapid, accurate and regulated manner. The complex recognizes protein cargo by either nuclear localization signals (NLSs) or nuclear export signals. The majority of the NLSs are arginine-lysine rich regions in proteins, as previously discussed in (Dingwall and Laskey 1998, Gorlich 1998). Chromosome territories. When stained with chromosome-specific fluorescent probes, each chromosome is confined to a discrete region and thus has a spatially limited volume instead of a diffuse appearance in the nucleus, referred to as a chromosome territory. The pattern of chromosome territories in a single nucleus is probabilistic rather than absolute (Meaburn and Misteli 2007). Remarkably, small chromosomes have a tendency to be located at the center of the nucleus, while large chromosomes are distributed at the nuclear periphery. Also, the functional pattern of chromosome territories is emerging: gene-poor chromatin domains form a layer beneath the nuclear envelope, while gene-dense chromatin is enriched in the nuclear interior (Bolzer et al. 2005). There is also evidence that chromosome territories are differentially manifested depending on the differentiation status of the cell. For example, in CD4+ thymocytic T-cells the position of chromosome 6 is shifted towards the center of the nucleus, whereas in CD8+ cells it is shifted significantly towards the periphery (Kim et al. 2004). Numerous studies have shown a correlation between the transcriptional repression of mammalian genes and their positioning at the nuclear periphery, further providing evidence of a gene repression function for the nuclear lamina (Kosak et al. 2002, Hewitt et al. 2004, Zink et al. 2004, Chuang et al. 2006, Williams et al. 2006). Histone deacetylation appears to have a crucial role in nuclear lamina-mediated gene silencing (see above), since genes associated with the nuclear lamina are hypo-acetylated at histone H4 (Somech et al. 2005, Pickersgill et al. 2006, Reddy et al. 2008).

1.3. Nuclear matrix (nuclear scaffold, nucleoskeleton, karyoskeleton)

Early light-microscopic studies revealed very few visible structures in the nucleus and it was proposed that these sparse structures are suspended in a liquid medium called ‘the nuclear sap’ or ‘karyophylum’. Since the advent of electron-microscopic

18 techniques, the nucleus has been revealed to be much more highly structured. In 1966 Don Fawcett defined the nuclear matrix as the non-chromatin structure of the nucleus, which he observed in unextracted cells under the electron microscope (Nickerson 2001). Also, a nuclear matrix had been discovered by biochemical methods as a nuclease and salt-resistant “non-chromatin structural carcass” and patented by Russian investigators as early as 1948 (Pederson 2000). Since then, some researchers have argued that the nuclear matrix is merely a global aggregation phenomenon consequent upon the preparation methods (high salt and detergent content) rather than a real in vivo structure (Pederson 1998, Hancock 2000, Pederson 2000). However, the main facts supporting the conception that nuclear matrix is a real structure rather than a mere concept are: 1) the protein network of the nuclear matrix in the unfractionated nucleus can be observed by analytical electron-spectroscopic imaging (Hendzel et al. 1999) 2) the nuclear matrix can be isolated at physiological salt concentrations (Jackson et al. 1990a) 3) the existence of chromatin loops bound to a non-chromatin network (after the stripping of histones) was inferred from biochemical studies (Benyajati and Worcel 1976) and visualized by microscopy (Vogelstein et al. 1980). The nuclear matrix consists of two electron-microscopically visible parts: the nuclear lamina and an internal nuclear matrix connected to the lamina. The internal nuclear matrix is a highly branched and fibrogranular network of ribonucleoproteins (RNPs) (Smetana 1963). The 10-nm filaments of the matrix are organized in a three- dimensional anastomosing network in which nucleoli are enmeshed. Nuclear matrix preparations examined by 2D gel electrophoresis typically reveal 200 major protein spots, and the principal proteins have been identified as hnRNP proteins and the nucleolar protein nucleophosmin/B23 (Capco et al. 1982). The matrix is composed of proteins and RNA, and intriguingly phospholipid content is also repeatedly reported (Berezney and Coffey 1974, Cocco et al. 1980). In fact, phospholipids appear to serve a structural function, since hydrolyzing them with phospholipase C releases the RNA and destroys the matrix fibrils (Cocco et al. 1980). Almost all of the nuclear phosphoinositide signaling lipid species (discussed in greater detail in chapter 5.) are also localized in the nuclear matrix (Gonzales and Anderson 2006). Although many internal nuclear matrix proteins have been identified (~ 400 in the nuclear matrix database (Mika and Rost 2005), it is not currently clear how these proteins assemble to form the filaments of the internal nuclear matrix. Chromatin contains DNA sequences called matrix-attachment regions (MARs) or scaffold-attachment regions (SARs). S/MARs are defined as chromatin elements which bind specifically to the nuclear matrix and as DNA fragments which copurify with the nuclear matrix (Michalowski et al. 1999). In the Drosophila melanogaster genome MARs are interspersed at the intervals of 26-112 kb (Mirkovitch et al. 1986), which is consistent with the estimated sizes of chromatin loops in flies and mammals (Benyajati and Worcel 1976, Vogelstein et al. 1980, Jackson et al. 1990b, Razin et al. 1995). In vertebrates, predicted MARs are significantly conserved in evolution between human and mouse and, intriguingly, MARs in the 5’ intergenic regions especially so, suggesting their possible involvement in gene regulation (Glazko et al. 2003). MARs are believed to be control elements in maintaining the independent realms of gene activity. Thus each particular MAR creates a unique

19 nuclear microenvironment where different regulatory proteins assemble and usually enhance transcription but can sometimes also repress gene expression (Boulikas 1995).

2. Chromatin

Chromatin was already visualized using basophilic aniline dyes around 1840 and given this name by Walther Flemming. Chromatin is a complex of DNA, histone proteins and other non-histone proteins. It was already observed in the late 19th century that chromatin can ‘transform’ (condense) into chromosomes during mitosis and decondense after cell division. In 1928 Emil Heitz named the faintly and brightly stained chromatin euchromatin and heterochromatin, respectively. Heitz also formulated the hypothesis (albeit no longer entirely valid) that “euchromatin is genicly active”, whereas “heterochromatin is genicly passive” and that “heterochromatin chromosomes or pieces of chromosomes contain no genes or somehow passive genes”, as reviewed by Trojer and Reinberg (2007). In the 1970s it was proposed that the chromatin structure is based on a repeating unit of eight histone molecules and about 200 DNA base pairs (Figure 1) (Kornberg 1974). When the X-ray crystal structure of the nucleosome core particle of chromatin was solved in atomic detail, it was verified that 146 base pairs of DNA are organized around the histone protein octamer (nucleosome). Roughly two superhelical turns of DNA are wrapped around an octamer of core histone proteins: H3-H4 tetramer and two H2A-H2B dimers (Luger et al. 1997). Histones are small basic proteins and extremely conserved through evolution. They consist of a globular domain and a more flexible and charged NH2-terminus (the histone “tail”) which protrudes from the nucleosome. The function of core histones and their tails will be discussed with greater detail in chapter 3. Linker histones (H1/H5 family) instead represent a diverse family of proteins that are bigger in size and bind to nucleosomes and bring them together to form a 30-nm chromatin fiber (Belikov and Karpov 1998).

20 Figure 1. Chromatin packing. The model depicts the many levels of chromatin packing. As a net result, DNA is packaged into a mitotic chromosome 10 000-fold shorter than its extended length. The diameter of the molecules is shown on the right margin. Figure modified from (Alberts 2002).

2.1. Transcriptionally active euchromatin

Euchromatin is the 11-nm chromatin fiber where nucleosomes are regularly spaced on a DNA template. Although in vitro experiments have shown that 11-nm chromatin is a poor template for transcription, it is still regarded as the template for transcription in vivo (Knezetic and Luse 1986, Lorch et al. 1987). However, active promoter regions are usually devoid of nucleosomes (Yuan et al. 2005, Ozsolak et al. 2007), as the promoter-associated nucleosomes have usually been repositioned by the action of the ATP-dependent chromatin remodeling enzymes. Overall, the nucleosome architecture imposes structural obstacles on RNA polymerase II (RNA

21 polII)-mediated transcription. Results from in vitro transcription experiments using 11-nm chromatin as template are an oversimplification, since transcription initiation in vivo requires at least histone modifications, chromatin remodeling, histone variant incorporation and histone eviction, as reviewed by (Li et al. 2007). Typically RNA transcription commences with the binding of activators upstream of the core promoter, including the TATA box and the transcription start site, but they can also bind downstream of the promoter (Maston et al. 2006). Probably the most exhaustively studied example is the yeast gene activator protein Gal4, which is divided into a DNA binding domain and an activation domain. The DNA binding domain in activator proteins recognizes and binds on the specific DNA sequence upstream of the gene, and the distance between the binding site and the activated gene can be very long. The DNA looping brings the distant activator protein and the activated gene promoter into close proximity with each other. The activator domain attracts adaptor complexes such as SAGA (Green 2005) or Swi/Snf and Mediator, all of which in turn facilitate the binding of general transcription factors (GTFs) (Thomas and Chiang 2006). RNA pol II is positioned at the core promoter with the help of numerous GTFs to form the preinitiation complex. One of the GTFs, TFIIH, melts 11-15 bp of DNA and thus positions the single-stranded template in RNA polII. Subsequently the carboxy terminal domain of RNA polII is phoshorylated by TFIIH during the first 30 bp of transcription and loses its contacts with GTFs and proceeds to the elongation stage (Alberts 2002, Buratowski 2003).

2.2.Transcriptionally silent heterochromatin

Heterochromatin is the ³ 30-nm chromatin classically divided into facultative and constitutive heterochromatin. Both are transcriptionally silent, but facultative heterochromatin retains the potential to decondense and thus to interconvent between hetero- and euchromatin. Facultative heterochromatin can be molecularly defined as a condensed and silent chromatin which decondenses and allows transcription in a temporal, spatial and parental/heritable manner. Telomeric and centromeric regions are constitutively heterochromatinized for the reason that this is important for genome integrity (Grunstein 1998, Grewal and Jia 2007). Genomes of the higher eukaryotes contain more repetitive and non-coding sequences and therefore a larger proportion of the genome is constitutively heterochromatinized in these organisms (Craig 2005, Grewal and Jia 2007). When chromatin is condensed into constitutive heterochromatin this does not mean that the transcription is totally blocked. The level of transcription is, however, low compared to protein coding genes in the euchromatin regions. A heterochromatinized region can be of any size, from a gene promoter through bands of pericentric heterochromatin or whole chromosome up to a whole genome of terminally differentiated erythrocyte nuclei (Craig 2005). A classical example of whole chromosome being facultatively heterochromatinized is the inactivation of the X chromosome in female mammalian organisms in dosage compensation (Ohno et al. 1959, Lyon 1961, Beutler et al.

22 1962). In females, one X chromosome is stably silenced in the preimplantation stage embryo but is reactivated at the blastocyst stage. Again, before gastrulation, one randomly chosen X chromosome is subjected to whole-chromosome condensation. Neither the conformation of facultative heterochromatin nor the entire number of factors affecting its establishment and maintenance is precisely known. Nevertheless, according to Trojer and Reinberg (2007), the interconvention process between euchromatin and facultative heterochromatin includes at least: 1) incorporation of specific or alternate components in the chromatin 2) modulation of the chromatin 3) intervention of trans-acting factors and 4) subnuclear localization. These processes are explained in greater detail in the legend to Figure 2.

Figure 2. Euchromatin – Facultative heterochromatin interconvention process. Transcriptionally active 11-nm euchromatin is converted to silent and compacted heterochromatin via multiple factors: 1) Exchange of chromatin components, including the incorporation of linker histone H1 and switch of canonical core histone H2A with the variant macroH2A; 2) Chromatin modulation involves covalent modifications of histones (discussed in greater detail in chapter 3) and DNA; common histone modifications are histone acetylation (in euchromatin) and histone deacetylation (in facultative heterochromatin) by histone acetyltransferase (HAT) and histone deacetylase (HDAC) enzymes, respectively; alteration in positioning by ATP-dependent chromatin remodelers in the nucleosome clearance process in active transcription in euchromatin; DNA methylation by DNA methyltransferases (DNMTs) is usually needed in silent heterochromatin formation; 3) Chromatin trans-acting factors include various non-coding RNAs, and many trans-acting proteins are also needed for facultative heterochromatin maintenance e.g.: heterochromatin protein 1 (HP1) and polycomb group (PcG) proteins; 4) Subnuclear position of the genomic locus is also reported to affect facultative heterochromatin formation. The figure is a modified version from Trojer and Reinberg (2007).

It is now widely accepted that reversible and heritable changes in gene expression can occur without alterations in DNA sequence (Jenuwein and Allis 2001). Pioneering studies on X-ray-induced chromosomal translocations in the fruitfly Drosophila melanogaster provided some of the earliest evidence that genes are in either “on” or “off” state depending largely on whether they are close to the euchromatin or to the heterochromatin region, respectively (Muller and Gershenson

23 1935). This phenomenon, where active euchromatic genes are translocated adjacent to the heterochromatin and are therefore silenced, is known as position-effect variegation (PEV). PEV has been adopted as a versatile tool for genetic screens in D. melanogaster (Reuter and Spierer 1992) and in yeast Schizosaccharomyces pombe (Thon and Klar 1992, Allshire et al. 1994), to identify genes involved in modifying PEV and thus potentially regulating the chromatin structure. Use of these screens is also discussed in section 3.3.2.

3. Posttranslational histone modifications regulate DNA-related processes

In vitro transcription experiments with 11-nm chromatin fiber (Figure 1) have shown that nucleosomes on the DNA template impede transcription. Consistent with this, further in vivo studies showed that removal of entire nucleosomes (Han and Grunstein 1988) or only their basic tails (Kayne et al. 1988) exerts specific effects on gene transcription. When the first nuclear histone modification enzyme, histone acetyltransferase (HAT) (Brownell et al. 1996), and the first chromatin remodeling complex (Swi/Snf) (Cote et al. 1994, Imbalzano et al. 1994, Kwon et al. 1994) were biochemically isolated and characterized, it became clear that the chromatin structure and the non-histone proteins regulating it impose a profound and ubiquitous effect on almost all DNA-related metabolic processes. These processes include transcription, recombination, DNA repair, replication, kinetochore and centromere formation. The simplified mechanism whereby histone modification enzyme complexes affect transcription is depicted in Figure 3 and explained in the figure legend.

24 Figure 3. Model for transcriptional repression and activation mediated by histone modification. A) In the off state, the repressor (REP) binds to the upstream repressor site (URS) in a sequence- specific manner and recruits negative modifiers such as histone deacetylase (HDAC) either directly or via a co-repressor complex (CO-REP) such as Sin3 (see chapter 4). HDAC removes acetyl (ac) groups from the histone N-terminal tails, leading to condensation of the chromatin, which is inaccessible to RNA polymerase II (RNApol II) to initiate transcription, and the gene is silenced. B) In the on state, the DNA-bound activator (ACT) at the upstream activator site (UAS) recruits positive modifiers such as histone acetyltransferase (HAT) either directly or via a co-activator complex (CO- ACT). HAT transfers the acetyl groups to the histones, which leads to loosened conformation of the chromatin, which is thus accessible to RNApol II to start transcription of the open reading frame (ORF) of the gene. Figure adapted from (Berger 2007).

3.1. Histone tail modification

Post-translational histone tail modifications (PTMs) are thought to have an influence on gene expression in two ways: 1) The degree of chromatin packing is altered directly by a change in electrostatic charge or through internuclesomal contacts and 2) attached chemical moieties alter the nucleosomal surface and attract a different set of chromatin-binding proteins (chromatin trans-acting factors). Both outcomes are probably equally important, as previously reviewed (Hansen et al. 1998, Wolffe and Hayes 1999). For example, acetylation of lysine 16 in histone H4 (H4K16) in a naive chromatin array assembled from bacterial recombinant histone results in relaxation of the array (Shogren-Knaak et al. 2006). Additionally, bromodomains in nearly all HAT-associated transcriptional co-activators can interact specifically with

25 acetylated lysine in the histone H3 and H4 tail sequences (Dhalluin et al. 1999). Histone tail modifications are often termed “epigenetic” marks; as yet, however, the relationship of these modification to stable epigenetic marks passed to subsequent cell generations, epigenetic inheritance, remains obscure (Berger 2007). After extensive mass spectrometric studies, as reported for example by (Zhang et al. 2003) it is reasonable to assume that the most prevalent PTMs in the core histones are now known. Canonical core histones H2A, H2B, H3 and H4 (and some of their variants) can be decorated with various covalent modifications such as acetylation, methylation, phosphorylation, citrullination, ADP-ribosylation, ubiquitination and sumoylation (Figure 4) (Strahl and Allis 2000, Berger 2007). The majority of modifications are localized in the amino-terminal histone tails, while a few are localized in the carboxy-terminal histone tails. Only a few of the modified residues are located in the central globular domains in the histones, as depicted in Figure 4. For example, one peptide fingerprint study identified over 60 different histone modifications: 31 , at least 20 methylations, at least 4 phosphorylations and 2 ubiquitinations (Zhang et al. 2003).

Figure 4. Identified protein modifications (methyl-, acetyl- and phosphate) in histone H3 as an example. The majority of the post-translational histone modifications occur in the N-terminal tail region in histones. Modifications (mod): M, methylation; P, phosphorylation; A, acetylation. Amino acids (aa): R, arginine; T, tyrosine; K, lysine; S, serine. Numbers indicate the amino acid positions (pos). Data obtained from (Zhang et al. 2003).

3.2. ’Histone code’ hypothesis

The ‘Histone code’ hypothesis (Strahl and Allis 2000, Jenuwein and Allis 2001) emerged soon after it was experimentally established that histone tails and their modifications play a crucial role in recruiting certain trans-acting proteins for transcriptional regulation. According to this hypothesis: “distinct histone modifications, on one or more tails, act sequentially or in combination to form ‘histone code’ that is, read by other proteins to bring about distinct downstream

26 events” (Strahl and Allis 2000). This hypothesis claims that histones and their PTMs provide an additional layer of indexing potential, but it also predicts that all PTMs can be predictive of biological function. Increase in indexing potential by PTMs is an accepted argument due to the number of domains discovered in effector proteins specifically detecting these histone marks, e.g. bromodomains for acetylated lysine in HAT-associated co-activators (Dhalluin et al. 1999, Winston and Allis 1999, Owen et al. 2000), chromodomains for H3K9me in HP1 protein (Bannister et al. 2001, Lachner et al. 2001) and plant homeodomains (PHD) which preferentially bind to H3K4me3 in ING family proteins (Zhang 2006). Current debate on the histone code hypothesis focuses on the fact that the outcomes of many histone modifications are so ambiguous that the word ‘code’ is not warranted at least when set against the genetic code which dictates protein composition. In the following, four specific examples are presented in order to piece together the problems inherent in ‘histone code’ over-generalization. First, histone hypoacetylation and hyperacetylation correlate most often with transcriptionally silent and active chromatin, respectively. However, acetylation of H4K12 has been reported to be a hallmark of heterochromatin formation, which contradicts this general theme (Turner et al. 1992). A second example is provided by H3K4me3 histone modification. This PTM is typically associated with active transcription, since its recognition facilitates subsequent activation events such as histone acetylation (Sims and Reinberg 2006, Ruthenburg et al. 2007). Contradictory to this, under conditions of DNA damage, H3K4me3 recruits a repressor complex and silences transcription. Apart from gene expression, H3K4me3 recognition also facilitates V(D)J recombination in vertebrates, in which segments of genes encoding specific proteins important in the immune system are assembled (Matthews et al. 2007). Therefore, since the effect of H3K4me3 modification is something between the regulation of DNA recombination and activation or repression of transcription, it is difficult to predict the outcome of this modification without considering the cellular context. A third example is provided by histone H3K9me1, which is associated with constitutive heterochromatin and thus with transcriptional silencing, but, surprisingly, is also present at several actively transcribed genes (Vakoc et al. 2005). Therefore, H3K9me1 modification can be of particularly diverse significance, depending on the chromosomal location of the modification. A fourth example comes from studies with phosphorylation at histone H3 (H3S10ph). This PTM is a canonical marker for the chromosomal condensation at the onset of mitosis and meiosis. It also has a role in transcriptional activation, since it is induced at gene promoters (Johansen and Johansen 2006). Hence, H3S10ph modification cannot be used solely for outcome prediction, because it is involved in the broad compaction of whole chromosomes and local decompaction of the active gene promoters. These examples suggest that single histone PTM alone is rarely (if ever) predictive of the state of chromatin. In order to make predictions, one must know the chromosomal location of the modification, the cellular context e.g. cell cycle step), neighboring histone modifications and the combinatorial effect of these histone marks. Thus, with the increasing complexity of the histone PTMs and their contextuality the term ‘histone code’ is currently held to be an over-generalizing

27 concept (Sims and Reinberg 2006). The less stringent term ‘chromatin language’ is now proposed to model the chancy outcomes of histone PTMs (Berger 2007).

3.3. Histone acetylation and deacetylation

In mammals, acetylation occurs at at least 31 different sites in four histones, and especially in histones H3 and H4 these lysines are conserved in eukaryotes (Zhang et al. 2003, Ekwall 2005). Of the histone modifications listed above, histone acetylation has been the most widely studied and best known (Grunstein 1997). In 1964 Allfrey and coworkers discovered that histone acetylation levels correlate with gene activity (Allfrey et al. 1964). More detailed mechanistic insights came from findings that the N-terminal tails of histone H4 exert specific effects on gene transcription (Kayne et al. 1988) and that reversibly acetylated lysines (at positions 5, 8, 12, and 16) encompass a region required for the activation of transcription (Durrin et al. 1991). When the first HAT was discovered in Tetrahymena and seen to be strikingly similar to transcriptional adaptor protein Gcn5 in yeast, it became clear that histone acetylation is a targeted in vivo phenomenon in gene activation (Brownell et al. 1996). In the same year, a counteracting enzyme, histone deacetylase (HDAC), was purified from mammalian cells and was shown to be markedly similar to yeast transcription regulator Rpd3 (from the reduced potassium dependency screen ) (Taunton et al. 1996). In general, histone deacetylation leads to compact chromatin formation and gene repression, and this holds true on a global scale (Robyr et al. 2002). However, there are many examples of particular genes in which a decrease in histone acetylation in promoters is associated with transcription induction (Bernstein et al. 2000, Deckert and Struhl 2001, Wang et al. 2002). For instance, during osmotic stress Hog1 recruits a yeast homolog of the Sin3A-HDAC corepressor complex (Sin3-Rpd3) to certain osmoresponsive promoters, and this leads to deacetylation of histone H3 and H4 and, surprisingly, transcriptional activation (De Nadal et al. 2004). The authors in question concluded, however, that they could not exclude the possibility that other non-histone substrates of HDACs may contribute to gene induction. The effect of an acetyl group in the histone is also dependent on the position of the modified nucleosome, being either in the promoter or within the open reading frame (ORF) of the gene. For instance, HDAC hos2 in S. cerevisiae is bound to the ORFs of highly expressed genes and it deacetylates H4K12ac in the coding regions only when the target gene is transcribed (Wang et al. 2002). The authors hypothesized that deacetylation of H4 in ORFs by Hos2 is required for gene activity in that it elongates RNA polymerase and reverts disrupted chromatin to the original permissive state required for efficient transcription. At most locations studied in the yeast genome, the level of acetylation can be raised or lowered by deleting a particular HDAC or HAT respectively. Thus, apart from the promoter targeted-manner dependent on sequence-specific-DNA-binding proteins, HDACs and HATs also function in a global manner in maintaining acetylation status throughout the genome (Krebs et al. 2000, Kuo et al. 2000,

28 Vogelauer et al. 2000). The function of global histone acetylation is to modulate basal transcription. For example, in the case of the PHO5 gene, knock-out of yeast HDAC (Rpd3) leads to increased basal expression of PHO5 in the absence of activating signals. A second function is to allow gene repression or activation to be rapidly reversed. The rapid turnover of acetyl groups at most nucleosomes as a result of global HAT and HDAC activities may allow chromatin to revert to the initial default acetylation state when repressive or activating targeting is removed (Vogelauer et al. 2000). It is not currently known how global acetylation and deacetylation occur, but for example in yeast HDAC, Rpd3, the complex does bind globally also to non-promoter sequences in the yeast genome (Kurdistani et al. 2002). HATs and HDACs with specificities for different sites of acetylation affect common chromatin regions. Hence, it has been suggested that HAT and HDAC activity can coordinate biologically related processes by creating distinguishable histone surfaces (Ekwall 2005). For example, H4K8ac and H4K12ac modifications are positively correlated, whereas H3K18ac and H4K16ac are negatively correlated in the yeast genome. Furthermore, different clusters of acetylation patterns have been shown to correspond to different groups of co-expressed genes such as those co-expressed during nitrogen starvation or the cell cycle (Kurdistani et al. 2004). In the work in question the authors also demonstrate that the double bromodomain- containing protein Bdf1 is mostly associated with hyperacetylated genome regions but specifically avoids H4K16ac marks and thus presumably regulates these biologically related chromatin processes. Apart from their canonical role in gene transcription, HAT- and HDAC- mediated chromatin processes also have an effect on other DNA processes such as replication, repair and heterochromatin formation. DNA replication originates from multiple sites in the genome and the acetylation status of the chromatin in replication origins correlates with the early onset and efficient firing of replication in yeast (Vogelauer et al. 2002). Also for DNA double-strand break repair to occur, certain N-terminal lysines in histones H4 and H3 must be acetylated (Bird et al. 2002, Qin and Parthun 2002). Histone acetylation presumably generates an open chromatin structure which renders the damaged site more accessible to the DNA- repair machinery and/or generates appropriate binding surfaces for this machinery to be recruited (Kurdistani and Grunstein 2003). Nicotinamide adenine dinucleotide (NAD) –dependent class III HDAC, silent information regulator 2 (Sir2), is responsible for heterochromatin formation for example in the chromosome ends, the telomeres. Heterochromatin spreading at the telomeres is a self-perpetuating process in which Sir2 deacetylates K16 on histone H4 and Sir3 binds to deacetylated H4K16 and recruits Sir4, which then recruits more Sir2 to start the next cycle (Hoppe et al. 2002, Luo et al. 2002, Rusche et al. 2002). This process would eventually heterochromatinize the whole chromosome if the MYST-like HATs did not counteract this by acetylating H4K16 and in so doing, set boundaries between the euchromatin and heterochromatin (Kimura et al. 2002, Suka et al. 2002).

29 3.3.1. Histone deacetylases

HDACs form an enzyme group responsible for the removal of acetyl groups from the lysine residues in the histone tails. HDACs are divided into two distinct protein families based on their co-factor dependency: the classical Zn-dependent HDAC family and the NAD+ -dependent class III Sir2 family of HDACs (Table 1). The classical HDACs are further divided into phylogenetic classes I, II and IV. Mechanistically, classes I and II both require the presence of a zinc ion for hydrolysis of the acetyl group. Upon deacetylation both class I and II HDACs release the acetyl group in the form of acetate (Figure 5). In contrast, class III Sir2 requires the presence of the metabolic co-factor NAD+ for deacetylation, and releases the acetyl group in a manner reminiscent of ADP-ribosyl transferases, which transfer the acetyl group to ADP ribose upon NAD+ catalysis (Ekwall 2005).

Table 1 Different classes of histone deacetylases (HDACs) and their functions. Genomewide functional analysis of HDACs in Schizosaccharomyces pombe has revealed that 1) Clr6 (class I) is the principal enzyme in promoter-targeted repression, 2) Sir2 (class III) and Hos2 (class I) prevent nucleosome loss, 3) Clr3 (class II) acts cooperatively with Sir2 e.g. in rDNA, centromeres and telomeres (Wiren et al. 2005). References for substrate specificity studies: a, (Rundlett et al. 1998); b, (Kadosh and Struhl 1998); c, (Suka et al. 2001); d, (Wu et al. 2001); e, (Imai et al. 2000); f, (Borra et al. 2004).

Class Human S. cerevisiae S. pombe Substrate specificity Class I HDAC1 = Rpd3 Clr6 (Rpd3) deacetylates all sites of acetylation HDAC2 on histones H4, H3, H2A and H2B HDAC3 = Hos2 Hos2 except H4K16 (a,b,c) HDAC8 Hos1 Class II HDAC4 Hos3 HDAC5 Hda1 Clr3 (Hda1) deacetylates all sites of acetylation HDAC6 on histones H3 and H2B only (d) HDAC7 HDAC9 HDAC10 Class III SIRT1 Hst2 Hst2 SIRT2 Hst3 SIRT3 Hst4 Hst4 SIRT4 Sir2 Sir2 (ScSir2) prefers H4K16, H4K8, H3K9, SIRT5 Hst1 H3K14 and H4K12 acetylated substrates SIRT6 in descending order (e,f) SIRT7 Class IV HDAC11

30 Figure 5. Equilibrium of steady-state histone acetylation is maintained by opposing enzymatic activities of histone acetyltransferases (HAT) and deacetylases (HDAC). Acetyl coenzyme A is the acetyl donor. HATs transfer the acetyl group (shaded box) to the e-NH3+ group of lysine residues in the N-terminal tails of histone proteins. Reversal reaction is catalyzed by HDACs. Gray barrels represent nucleosomes. Figure is modified from (Kuo and Allis 1998).

HDAC enzymes are in fact evolutionarily more ancient than their histone substrates. Class I and II ancestors, for instance in Mycoplana ramose bacteria, which lack histones, are acetylpolyamine amidohydrolases which are involved in the degradation of acetylpolyamines via a deacetylation mechanism similar to that of eukaryotic HDACs (Sakurada et al. 1996, Leipe and Landsman 1997). HDACs are also found in archae bacteria, which have histones (White and Bell 2002). Acetylation is not restricted to histones, as acetylation modification occurs in ~85% of eukaryotic proteins, which in turn are also likely substrates for HDACs (Su et al. 2008).

3.3.2. HDACs in yeast genetic screens

The first HDAC encoding gene, yeast RPD3, was isolated genetically in a screen for mutants with reduced potassium dependency (rpd mutants) (Vidal et al. 1990). At that time it was not known that RPD3 is an HDAC. In the same screen, RPD2, which corresponds to the TRansport of potassium (K) TRK2 gene, together with RPD1 (Sin3 in mammals) was also isolated. TRK2 mutant rpd2 was shown to be epistatic to both rpd3 and rpd1 mutants. Since the effects of rpd1 and rpd3

31 mutations were not additive, the authors concluded that Rpd3 and Rpd1 might function at different steps in a single pathway or as subunits of a single negative regulator of the TRK2 gene. The latter hypothesis turn out to be correct, since we now know that RPD3 (HDAC) and RPD1 (Sin3) together, with some other proteins, form a Sin3A-HDAC corepressor complex, the global regulator of gene expression (Laherty et al. 1997). Subsequent study showed that RPD3 together with RPD1 is required for both the full repression and the full activation of transcription of many target genes (Vidal and Gaber 1991). Five years later, Taunton and coworkers, using a Trapoxin (HDAC inhibitor) affinity matrix, isolated a nuclear protein ~46 kDa in size, from the bovine thymus which copurified with HDAC activity (Taunton et al. 1996). This protein was enzymatically active and 60% identical with yeast Rpd3 transcriptional repressor. Since the HAT enzyme had already been found a year earlier (Kleff et al. 1995), the molecular entities needed for histone (de)acetylation- dependent transcriptional regulation were established and research within this field began to accelerate exponentially. The RPD3 gene was isolated in four independent mutant suppressor screens designed to identify transcriptional repressors, and was sometimes named after the corresponding function: SDI2 (SWI-dependent interconversion), which partially suppresses the requirement for SWI5 and which causes daughter cells to express the target gene (Nasmyth et al. 1987, Stillman et al. 1994). McKenzie et al. (1993) found that the RPD3 mutant relieves gene repression in the CBF1 mutant transcriptional regulator strain. Bowdish and Mitchell (1993) discovered that the RPD3 mutant yeast strain was able to express meiotic genes in an IME1 mutant strain (expression of meiotic genes depends on the IME1). Although silent heterochromatin is usually hypoacetylated, some studies have attributed a counteractive function to RPD3 in heterochromatin formation. RPD3 was in fact, identified as a factor generating chromatin permissive to transcription in a screen to identify factors affecting transcriptional silencing at the HMR mating- type locus in yeast SDS6 (suppressor of defective silencing) (Sussel et al. 1995). Similarly, both a centromeric PEV screen in Drosophila and the telomeric position- effect (TPE) in yeast identified histone deacetylase RPD3 as a protein which counteracts genomic silencing (De Rubertis et al. 1996). The authors concluded that this function of RPD3, which is in striking contrast to the general correlation between histone acetylation and increased transcription, might be due to a specialized chromatin structure at silent loci in centromeric and telomeric heterochromatin. A possible biochemical mechanism for this unorthodox anti- silencing function of HDAC was recently provided by Raisner and Madhani (2008), who showed that RPD3 negatively regulates sirtuins, class III HDACs with an important function in telomere silencing (Rusche et al. 2003). One suggested possibility is that RPD3 antagonizes direct acetylation of the Sir2 complex itself (Raisner and Madhani 2008) and indeed, N-terminal acetylation of Sir3 has been reported to promote silencing (Wang et al. 2004). Some yeast genetic screens have attributed other functions, apart from that of transcriptional repressor, to RPD3 and subsequently named them after corresponding functions. For example, Meskauskas et al. (2003) found that mof6 (maintenance of frame, promotes increased efficiencies of programmed -1

32 ribosomal frameshifting) is an RPD3 allele and proposed that HDAC has a function in ribosome biogenesis. Esposito and Brown (1990) found an REC3 mutant which is defective in mitotic recombination and later demonstrated that REC3 is an allele of RPD3 (Dora et al. 1999). The role of HDAC in ribosome biogenesis remains elusive; one possibility is that HDACs regulate the snoRNAs synthesis responsible for rRNA maturation rather than being physically directly linked to ribosome biogenesis (Meskauskas et al. 2003). Also the function of the RPD3 in mitotic recombination is best explained by changes in chromatin architecture in an RPD3 mutant strain. Change in chromatin structure is also the best explanation for the finding that RPD3 deletion leads to increased mobility of retrotransposons in yeast (Nyswaner et al. 2008).

3.3.3. HDAC inhibitors as drugs

Traditionally cancer has been considered a disease of genetic defects such as gene mutations and deletions and chromosomal abnormalities, all of which result in loss of function of tumor-suppressor and/or gain of function or hyperactivation of oncogenes (Bolden et al. 2006). However, recent advances in our understanding of the function of histone and DNA modifications have indicated that gene expression governed by these epigenetic changes is also crucial in the triggering and progression of cancer (Bolden et al. 2006). The first conceptions of HDAC inhibitors as anti-cancer drugs stemmed from the discovery of aberrant recruitment of HDACs to promoters in human malignancies. This aberration is due to the interactions of HDACs with oncogenic DNA-binding fusion proteins which result from chromosomal translocations, or due to overexpression of repressive transcription factor interacting with HDACs. For example, PML-RARa, PLFZ-RARa and AML-ETO fusion proteins induce acute promyelocytic leukemia and acute myeloid leukemia by recruiting HDAC- corepressor complexes to repress their target promoters (Lin et al. 2001, Pandolfi 2001). HDAC inhibitors have been used in combination with retinoids to treat acute promyelocytic and myeloid leukemias (Cote et al. 2002). Another example is offered by diffuse large B-cell lymphoma, in which in 40% of cases BCL6 transcription factor is overexpressed. BCL6 recruits HDAC2 to repress growth- regulatory genes such as p21 and thus promotes malignant growth in diffuse large B-cell lymphoma. By HDAC inhibition, p21 expression can be rescued, increasing tumor cell apoptosis (Pasqualucci et al. 2003). Since it is widely recognized that HDACs are promising targets for therapeutic interventions aiming to reverse cancer-associated abnormal epigenetic states (Baylin and Ohm 2006), there has been considerable effort to develop HDAC inhibitors. Some of these inhibitors, for example SAHA/vorinostat and CI-994, have reached phase III in clinical trials. On 2006, the U.S. Food and Drug Administration granted regular approval to vorinostat (Zolinza®; Merck & Co., Inc.), a histone deacetylase inhibitor, for the treatment of cutaneous manifestations of cutaneous T-cell lymphoma (CTCL) in patients with progressive, persistent, or recurrent disease on

33 or following two systemic therapies (Mann et al. 2007). Inhibitors developed to date are able to kill cancer cells mainly through apoptosis and besides their intrinsic effects on tumor cells, they might also affect neoplastic growth and survival by regulating host immune responses and tumor vasculature. However, many preclinical studies have also indicated that the effect of HDAC inhibition can be broader and more complicated than originally understood (Bolden et al. 2006). HDAC inhibition by vorinostat for instance, is also reported to lead to genomic instability by a variety of mechanisms (Eot-Houllier et al. 2009). This is not surprising considering that HDAC in yeast (RPD3) has been caught in several genetic screens (as discussed in section 3.3.2.), indicating that Rpd3 complexes regulate (repress) many target genes. In addition, besides being involved in deacetylation of the target promoters, HDACs also have a global genomewide deacetylation function in non-promoter targets (Kurdistani et al. 2002) (see section 3.3.), which has to be taken into account when considering the specificity of HDAC inhibition. Finally, considering that histones comprise only one substrate among many non-histone protein substrates of HDACs, as 85% of eukaryotic proteins can be modified by lysine acetylation (Su et al. 2008), the effect of inhibitors could have a much broader effect on cellular physiology than originally anticipated. To date, the precise molecular mechanisms by which HDAC inhibitors induce cell death is not fully clear and the roles of individual HDAC inhibitors have not been identified (Pan et al. 2007). On the other hand, broad effect of HDAC inhibitors makes their use in combinatorial therapies possible. Several research groups have reported that HDAC inhibition act as a potent radiosensitizer. The ability of pre-treatment with vorinostat to sensitize cells to ionizing radiation appears to result in part due to its effects on the transcription rate of genes involved in DNA damage repair. In addition, HDAC inhibition-induced chromatin relaxation is required for increased efficacy of DNA- targeted chemotherapeutics. Vorinostat is currently being tested in combination with several DNA-targeted chemotherapeutics in metastatic non-small cell lung cancer and anticancer activity has been observed. HDAC inhibition by vorinostat has also been combined with DNA methyltransferase inhibitor and proteasome inhibitor therapies aiming to treat human malignancies. (Richon et al. 2009).

4. The Sin3-HDAC corepressor complex

Early genetic screens in yeast already suggested that RPD3/HDAC works together with the Sin3A protein complex (Nasmyth et al. 1987, Vidal et al. 1990, Sussel et al. 1995). Chromatographic studies with Sin3 in mammals revealed that it is manifested in multiple different HDAC-containing complexes varying in protein composition and in molecular size (Zhang et al. 1997). Sin3 is considered a canonical transcriptional corepressor protein. Since many studies have shown that reporter genes can be repressed when Sin3 is recruited to their promoters, it is widely accepted that the major role for Sin3 is to act as a

34 transcriptional corepressor. The Sin3-HDAC complex mediated histone deacetylation is also important for the genomic integrity in general. Namely, in S. pombe it has been shown that Sin3 targets HDACs to centromeres to repress their transcription and underacetylated status of the centromeres is necessary for sister chromatid cohesion (Silverstein et al. 2003) and for proper chromosome segregation (Ekwall et al. 1997). However, there are examples demonstrating a capability of Sin3 also to participate in gene activation. In yeast Sin3 was not only necessary for the full repression of certain genes but was also required to achieve maximal transcription when those genes were induced (Vidal et al. 1991). SIN3 knock-out screens in yeast and fruitfly have also demonstrated that a significant number of genes were also down-regulated, indicating that these genes are positively regulated by SIN3 (Bernstein et al. 2000, Pile et al. 2003). It has also been demonstrated that Hog1 recruits Sin3-HDAC directly to the target promoter and induces transcription (De Nadal et al. 2004). However, these studies are always subject to caveats, as screening experiments alone cannot exclude possible indirect effects, and as non- histone substrates of HDACs may contribute to gene induction. The Sin3 corepressor complex is a well-known platform with HDAC enzymatic functions, but it is noteworthy that the core complex can harbor other catalytic modules as well. The Sin3-HDAC complex is reported to bind and possess enzymatic activities such as Swi/Snf nucleosome remodeling (Sif et al. 2001), monosaccharide transferase (see OGT in Figure 6) (Yang et al. 2002) and histone methyltransferase activities (Yang et al. 2003). In fact, Nakamura et al. (2002) purified an over 2MDa supercomplex consisting of over 29 proteins, including the Sin3-HDAC complex, and further showed that the complex remodels, acetylates, deacetylates, and methylates nucleosomes and/or free histones.

4.1. Sin3A protein

SIN3 was first isolated in genetic screens investigating mating-type switching in S. cerevisae. The SIN3 (SWI-independent 3) mutant strain could bypass SWI5 transcriptional activator dependent mating type switching and was thus proposed to act as a transcriptional repressor (Sternberg et al. 1987). SIN3 was genetically isolated seven times and named accordingly: five times as negative regulator of transcription i.e. SDI1 (Nasmyth et al. 1987), RPD1 (Vidal et al. 1990), UME4 (Strich et al. 1989), CPE1 (Hudak et al. 1994), once as a positive regulator of transcription i.e. GAM2 (Yoshimoto et al. 1992), and once as an enhancer of silencing i.e. SDS16 (Sussel et al. 1995). Remarkably, yeast genetic screens where HDAC was caught often evinced SIN3 as an accompanying protein (SDI, RPD and SDS screens above). Sin3 protein thus came into focus in close molecular research on how it co-operates with HDAC to regulate transcription in mammalian cells. Yeast Sin3 is a large (1536 amino acids) acidic protein. In mammals, it exists in two major isoforms, Sin3A and Sin3B, encoded from the human chromosome bands 15q24 and 19p13.1, respectively. Human Sin3A is composed of 1273 amino acids,

35 whereas Sin3B isoform has a shorter amino terminal region and is 1162 amino acids in length. Multiple splicing variants of both Sin3A and Sin3B in both humans and mice have been reported (Alland et al. 1997, Yang et al. 2000). The most prominent feature is the existence of four paired amphipathic a-helices (PAH1-PAH4) (Wang et al. 1990). Since PAHs were structurally similar to the helix-loop-helix protein dimerization domains in the Myc family of transcription factors, it was predicted and later experimentally demonstrated that Sin3 is involved in multiple protein- protein interactions. Sin3 possesses no DNA-binding motifs or enzymatic activities and it is suggested to serve as a platform for the proteins involved in chromatin- level gene regulation to assemble (Silverstein and Ekwall 2005). Furthermore, two additional regions show evolutionary conservation: the histone deacetylase interacting domain (HID) situated between PAH3 and PAH4, and the C-terminal highly conserved region (HCR) (Wang and Stillman 1993) (Figure 6).

4.2. The core Sin3-HDAC complex

In 1997, three laboratories simultaneously isolated a complex from mammalian cells which is now considered to constitute the core Sin3 complex (Hassig et al. 1997, Laherty et al. 1997, Zhang et al. 1997). It is currently thought to be composed of eight proteins, Sin3, HDAC1, HDAC2, RbAp46, RbAp48, SAP30 and SAP18, and additional studies (Vannier et al. 1996, Dorland et al. 2000, Lechner et al. 2000) also legitimate SDS3 as belonging to the core complex (Figure 6). In early studies it was demonstrated that Sin3 could repress transcription when tethered to the gene promoters (Wang and Stillman 1993). Sin3A protein cannot bind DNA in vitro (Wang and Stillman 1990) and thus it needs accessory DNA targeting proteins in order to repress transcription. When it was demonstrated that yeast RPD3 is a histone deasetylase (Taunton et al. 1996), a subsequent study proved that gene repression by Sin3 is HDAC-dependent (Laherty et al. 1997). At amino acid level, mammalian histone deacetylases HDAC1 and HDAC2 are both ~60% identical with the yeast ortholog RPD3 (Taunton et al. 1996). HDAC1 and HDAC2 bind to the highly conserved HID region between PAH3 and PAH4. The Sin3-HDAC complex is probably able to function widely in regard to acetyl substrates, since RPD3 has been shown to deacetylate all sites of acetylation on histones H4, H3, H2A and H2B (except H4K16) (see Table 1). Retinoblastoma-associated protein 48 (RbAp48) was originally copurified with human HDAC1 in chromatographic studies using an HDAC inhibitor affinity matrix, and it was demonstrated to be required for HDAC targeting (Taunton et al. 1996). RbAp46 is a close homolog of RbAp48, and both have been copurified with the Sin3 complex (Zhang et al. 1997). RbAp46 and RbAp48 are WD repeat proteins (Neer et al. 1994) originally shown to interact with retinoblastoma (Rb) protein fragments (Qian et al. 1993, Qian and Lee 1995). Both Rb-associated proteins can interact with histone H4 (Verreault et al. 1998) and are thus predicted to stabilize the interaction between the Sin3-HDAC complex and histones. Such a conception is supported by the fact that some of the Rb-associated proteins are also subunits in other complexes working on histone templates: the HAT complex (Parthun et al.

36 1996, Verreault et al. 1998), the chromatin assembly complex CAF1 (Tyler et al. 1996, Verreault et al. 1996) and the nucleosome remodeling complex Snf2 (Wade et al. 1998). SAP18 has been shown to copurify with the mammalian Sin3 complex and to interact directly with Sin3 and HDAC1. SAP18 enhances Sin3-mediated transcriptional repression and is therefore proposed to stabilize HDAC1-Sin3 interaction and/or enhance HDAC1 activity (Zhang et al. 1997). In addition to regulating transcription, it also participates in mRNA processing through its association with the apoptosis- and splicing-associated protein (ASAP) complex (Schwerk et al. 2003). In yeast, SDS3 regulates the expression of the same set of genes as SIN3 and RPD3 (HDAC) (Vannier et al. 1996, Dorland et al. 2000, Lechner et al. 2000). A yeast strain lacking SDS3 possesses only residual Sin3A- associated HDAC activity and the physical interaction between Sin3 and RPD3 is severely destabilized implying that SDS3 promotes the integrity of the complex (Lechner et al. 2000). Other components such as SAP25, SAP130 and SAP180 have been reported to be associated with the Sin3A-HDAC complex, but their roles in it are currently unknown (Fleischer et al. 2003, Shiio et al. 2006).

4.2.1. Sin3 associated protein 30 (SAP30)

SAP30 (Sin3-associated protein 30) was originally identified as a co- immunopurifying protein with the Sin3 complex (Zhang et al. 1997) and was further characterized as a conserved member of the Sin3A corepressor complex (Laherty et al. 1998, Zhang et al. 1998). In yeast, SAP30 was demonstrated to be important for normal cell growth, but not essential for cell viability (Zhang et al. 1998). Human SAP30 is a small 220 amino acid long protein composed mainly of basic residues. The biochemical function of SAP30 is unknown and the amino acid sequence possesses no similarities with any known proteins or any identifiable domains. The C-terminus of SAP30 binds to the PAH3 region in Sin3A and Sin3B, whereas the N-terminus binds N-CoR corepressor (Figure 6). It has been proposed that SAP30 functions as a linker molecule between these two corepressors (Laherty et al. 1998). SAP30 represses transcription in an HDAC-dependent manner when tethered to promoters and the C-terminal Sin3 interacting region in SAP30 is necessary for its ability to repress transcription. Results of SAP30 antibody microinjection experiments, however, have suggested that it is not needed for Sin3’s intrinsic repressive activity but is involved in Sin3-mediated N-CoR repression. SAP30 is probably required for repression driven by a specific subset of corepressor complexes (Laherty et al. 1998). It is reported to bind directly to HDAC1 and RbAp48 and is also able to repress transcription in a Sin3-deleted yeast strain (Zhang et al. 1998). There is thus strong evidence to suggest that SAP30 can form a repressome independently of Sin3 and in fact, transcription regulator Yin Yang 1 (YY1), which binds DNA sequence specifically, recruits SAP30-HDAC1 without Sin3 to its target promoter to repress transcription (Huang et al. 2003). Although accumulating evidence suggests that SAP30 can act Sin3- independently, co-purification of SAP30 with the core members of the Sin3

37 complex and the yeast genetic data would imply that co-operation exists. In yeast SAP30 counteracts genomic silencing in telomeres, HMR and rDNA in a manner similar to Rpd3/HDAC and Sin3 (Zhang et al. 1998, Smith et al. 1999, Sun and Hampsey 1999, Loewith et al. 2001). Evidence is accumulating to indicate that Sin3-HDAC-SAP30-mediated counteracting of genomic silencing in telomeric and rDNA loci is to be explained by their ability to counteract Sir2, a class III HDAC silencing these loci (Smith et al. 1999, Sun and Hampsey 1999). In fact, an independent screen recently identified SAP30, Rpd3 and Sin3 as negative regulators of sirtuin spreading (Raisner and Madhani 2008). The anti-silencing function of SAP30 seems to be important for telomere length maintenance, since in yeast strains where either Sap30 or Sin3 is deleted, the lengths of the telomeres are 50-150 bp shorter as compared to the wild-type strain (Askree et al. 2004). On a local one-promoter scale, disruption of yeast Sap30 usually causes a transcriptional derepression which is milder than the derepression of the Rpd3 and Sin3. On the other hand, some promoters are not derepressed by the disruption of Sap30, while they are derepressed when Rpd3 and Sin3 are disrupted. These findings offer an explanation why Sap30 did not emerge in the initial selection which revealed Rpd3 and Sin3 as transcriptional repressors, and they also show that Sap30 operates in a promoter-dependent manner (Zhang et al. 1998). The co- operation between Sap30, Rpd3 and Sin3 is also demonstrated in a yeast mof-screen (see section 3.3.2. HDACs in yeast genetic screens), which would imply an unorthodox function for this complex in rRNA processing (Meskauskas et al. 2003). SAP30 is predicted to stabilize the interaction between HDAC1 and Sin3 (Silverstein and Ekwall 2005), but it evidently increases the modularity of the Sin3 complex by bridging interactions. The human SAP30 has been reported to interact with a number of other proteins such as the retinoblastoma-binding protein 1 (RBP1) (Lai et al. 2001), the CBF1-interacting corepressor (CIR) (Hsieh et al. 1999), and the inhibitor of growth 1b (ING1b) tumor suppressor protein (Skowyra et al. 2001, Kuzmichev et al. 2002). The latter interaction is particularly interesting, since it has been demonstrated that the ability of p33ING1b to inhibit cell growth depends on its interaction with the Sin3–HDAC complex through SAP30. Another possible implication for SAP30 in human pathogenesis is provided by reports that human SAP30 plays a role in the transmission and propagation of certain viruses (Krithivas et al. 2000, Le May et al. 2008)

38 Figure 6. The Sin3 corepressor complex. The eight core components (unfilled circles) of the Sin3 complex are assembled in PAH3 (Paired Amphipathic Helix 3) and HID (Histone deacetylase Interacting Region). Some of the occasionally interacting components are depicted as filled circles. Opi1 repressor regulates the transcription of structural genes of phospholipid biosynthesis (Wagner et al. 2001). Pf1 links the Transducin-Like Enhancer (TLE) corepressor with Sin3 (Yochum and Ayer 2001). The mammalian Sin3 corepressor was originally found in a screen for proteins interacting with the mammalian Mad1 protein, a transcriptional repressor (Ayer et al. 1995). In yeast, Sin3 and HDAC1 are specifically required for transcriptional repression by Ume6, a DNA-binding protein which regulates genes involved in meiosis (Kadosh and Struhl 1997). The gray barrel depicts the nucleosome and interaction of RbAP46 and 48 proteins with histone H4. Figure modified from (Silverstein and Ekwall 2005).

4.3. SAP30/HDAC complexes in diseases

4.3.1. SAP30 in cancer

Moderate cell cycle progression is assured by the regulatory function of two classes of genes: proto-oncogenes which promote cellular growth and tumor suppressor genes which inhibit it. Mutations in these genes can result in alterations in cell growth and, as a consequence, neoplastic transformation (Fearon and Vogelstein 1990, Weinberg 1991). The identification of deletions at specific loci in a tumoral sample suggests the presence of a tumor suppressor gene within the deleted region. Some tumor suppressor genes are commonly involved in different types of tumors, while others are preferentially associated with the carcinogenesis of specific tissues (Ponder 1988, Fearon and Vogelstein 1990, Lasko et al. 1991). The first indications that SAP30 could act as a tumor suppressor came from a study in which a loss of heterozygosity screen was conducted for 19 cases of skin basal cell carcinoma (Sironi et al. 2004). The minimal deleted region in basal cell carcinoma samples was assessed to be 4q32-35 -harboring SAP30 and p33ING2/ING1L genes. The ING (inhibitor of growth) family of genes are tumor suppressor genes and at least p33ING1 is a component of the p53 signaling pathway

39 and cooperates with p53 in the negative regulation of cell proliferation and promotion of apoptosis (Garkavtsev et al. 1996). One particular alternative transcript from an ING family member, p33ING1b, associates through its N- terminal sequence with SAP30 and represses transcription (Figure 6) (Skowyra et al. 2001). Furthermore, it has been demonstrated that the ability of p33ING1b to inhibit cell growth depends on its interaction with the Sin3–HDAC complex through SAP30 (Kuzmichev et al. 2002). Recently, chromosomal abnormalities in the 4q region have also been recognized in another class of skin cancer, squamous cell carcinomas. In an array comparative genomic hybridization screen, the 4q33-34 chromosomal region was detected as a locus with gain of fragments of genetic material (Salgado et al. 2008). Congruently, a loss of heterozygosity screen with head and neck squamous cell carcinoma samples also identified frequent deletion in the 4q34-35 chromosomal region (Cetin et al. 2008). In melanoma, patients having missense mutations in the ING1 gene had a 50% higher risk of dying from the disease within five years compared to patients with no ING1 mutation (18%) (Campos et al. 2004). In the same study it was further shown that 20% of the melanoma primaries contained missense mutations in the SAP30-interacting region in the ING1 protein. It is worthy of note that there are several other genes in the 4q32–35 region which could be implicated in the above-mentioned cancers. However, many of the proteins encoded by these genes (e.g. SWI/SNF complex, ING2, SAP30) have been shown elsewhere and in this study to play important roles in gene expression. The deletion in 4q32–35 appears functionally significant in involving tumor suppressor genes whose loss could impair gene silencing by epigenetic modifications and consequently perturbed cell growth control. The involvement of SAP30 protein in skin carcinogenesis is so far based on circumstantial evidence and it seems to be co- factored by ING-mediated growth suppression, which is dependent on the p53 pathway (Garkavtsev et al. 1998).

4.3.2. SAP30 protein is a cofactor in virus transmission

Recently, SAP30-mediated transcriptional repression was shown to play a role in the transmission and propagation of certain viruses. The first relevant finding was that the human herpes virus 8 LANA interacts with proteins of the Sin3 corepressor complex via SAP30 and negatively regulates Epstein-Barr virus gene expression in dually infected primary effusion lymphoma cells (Krithivas et al. 2000). The latest finding was made in work with Rift Valley Fever Virus (RVFV), where it was shown that RVFV non-structural (NSs) protein recruits HDAC complexes via SAP30 and YY1 to repress the interferon-beta (IFN-b) gene and thus to counteract host cytokine defense against viral infection. To ascertain the role of SAP30, the authors produced a recombinant RVFV in which the interacting domain in NSs was deleted. The virus was unable to inhibit the IFN-b response and was avirulent for mice.

40 Another link between viral transmission and the SAP30-Sin3A-HDAC complex comes from the maintenance of frame screen (Meskauskas et al. 2003) described in section 3.3.2. “HDACs in yeast genetic screens”. Programmed ribosomal frameshifting (PRF) is used by many viruses to regulate the production of structural and enzymatic proteins from the corresponding overlapping genes in the viral genomes (Dinman et al. 1998). Meskauskas et al. (2003) showed that the SAP30- Sin3A-HDAC complex is required for control of the wild-type levels of frameshifting and virus maintenance. To conclude, SAP30-mediated HDAC activity seems to be beneficial for viral transmission in two ways: i) silencing host cytokine defense and ii) manipulating the translational apparatus to maintain a proper level of PRF necessary to the morphogenesis of RNA viruses.

5. Phosphoinositides (PtdInsP) - messengers of cytosolic and nuclear signaling in the cell

Eukaryotic cells are able to sense and react to changes in their environment. Phosphoinositides (PtdInsP) have a crucial role in transferring information from the exterior through the plasma membrane to the inside of the cell where the information can be processed and reacted on. Sometimes information is further transferred within the nucleus to respond to changes in gene transcription, which eventually leads to changes in cell growth and differentiation (Hammond et al. 2004, Di Paolo and De Camilli 2006, Lemmon 2008). In the early 1950s Hokin and Hokin noted that stimulation of exocrine tissues causes changes in the turnover of membrane phospholipids. This so-called “phospholipid effect” was confirmed as universal phenomenon in a variety of tissues and attributed to changes in the turnover of phosphatidylinositols (PtdIns) and its phosphorylated derivatives phosphoinositides (PtdInsP) (Hokin 1985) (Table 2).

Table 2. Phospholipid values in mammalian cells. Reprinted with the permission of the Nature Publishing Group from (Lemmon 2008).

Fold increase on Lipid Relative level (%) stimulation Phosphatidylserine 8.5 1 Phosphatidic acid 1.5 1 Phosphatidylinositol 1.0 1 PtdIns3P 0.002 1 PtdIns4P 0.05 0.7 PtdIns5P 0.002 3-20* PtdIns(4,5)P2 0.05 0.7 PtdIns(3,4)P2 0.0001 10 PtdIns(3,5)P2 0.0001 2-30

* in response to thrombin stimulation (Morris et al. 2000), cellular stress and during the cell cycle (Pendaries et al. 2005)

41 PtdIns, the precursor of all phosphoinositides, is primarily synthesized in the endoplasmic reticulum and is delivered to other membranes either by vesicular transport or with the help of cytosolic PtdIns transfer proteins. Phosphoinositides are concentrated on the cytosolic side of the cell membrane and reversible phosphorylation of its sugar head group, the inositol ring at positions 3, 4 and 5 results in the generation of seven phosphoinositides species (Figure 7). PtdIns(4,5)P2 alone regulates exocytosis, endocytosis, phagocytosis, macropinocytosis, cell motility, signal transduction, ion channels, cell adhesion and membrane microtubule capture. It is thus justifiable to say that phosphoinositides have been implicated in almost all aspects of cellular function, as reviewed by (Di Paolo and De Camilli 2006). Many protein domains which recognize specific species of phosphoinositides have been identified (Figure 7). Specific recognition is in some cases vital, since for instance amino acid substitution in the pleckstrin homology (PH) domain of Bruton’s Tyrosine Kinase (BTK) protein causes severe signaling defects, as seen in X-linked agammaglobulianemia (Lindvall et al. 2005).

Figure 7. Protein domains binding specific phosphoinositide lipid targets. The action of phospholipases, lipid kinases and lipid phosphatases in response to extracellular signals leads to remodeling of the phosphoinositide profile, which in turn is sensed by proteins with various domains to execute the response. Shown are the structure of unphosphorylated PtdIns and interconversion reactions for all phosphoinositides found in mammalian cells. DAG, diacylglyserol. Enzymes: PTEN, phosphatase and tensin homologue on chromosome ten; SHIP, SH2-containing inositol 5’- phosphatase. Domains: C1, conserved region-1 (from protein kinase C); PH, pleckstrin homology; PROPPINs, beta propellers which bind phospoinositides; FYVE, ‘Fab1, YOTB, Vac, EEA1’; PX, Phox-homology; PHD, Plant homeodomain. The PHD fingers, FYVE and C1 domains contain Zn2+ ions which are crucial for their structure. The PH and FYVE domains contain conserved lysine (K) and arginine (R) residues which form most interactions with phosphate groups in phosphoinositides. Many PH domain-bearing kinases such as BTK (Bruton’s Tyrosine Kinase) undergo dramatic transient relocalization to the plasma membrane on signal-dependent activation of phosphoinositide 3-kinase (PI3K) and a consequent increase in PtdIns(3,4)P2 and/or PtdIns(3,4,5)P3. Fab1 is a FYVE domain containing PtdIns3P 5-kinase responsible for the production of PtdIns(3,5)P2 lipid species essential for endosome-trafficking. The status of the PHD fingers as putative PtdIns5P effectors is not clear. Figure modified from (Lemmon 2008).

42 The most fundamental difference between the nuclear and cytosolic phosphoinositides is that in the nucleus they are located mainly outside the membrane bilayers (Hammond et al. 2004). In fact, early isolations of the nuclear matrix already showed its phospholipid component to be necessary for the integrity of the nuclear matrix (Berezney and Coffey 1974, Cocco et al. 1980). These studies covered all the phospholipids, i.e. phosphatidylinositol, phosphatidylserine, phosphatidylethanolamine, sphingomyelin etc., which are hereafter referred to as phospholipids. When studied in greater molecular detail, it was shown that cell nuclei stripped of their envelopes contained phosphoinositides (Cocco et al. 1987). Subsequent studies have shown that cells evince an intranuclear phosphoinositide metabolism utilizing enzymes and substrates equivalent to those found in cytosol and plasma membrane (Vann et al. 1997). The nuclear bulk of phospholipids co- purified with non-histone chromosomal proteins, while DNA and histone fractions did not reveal the presence of lipids (Manzoli et al. 1976). The fact that the nucleus is inhabited by non-membraneous lipids, resistant to detergents, would imply that certain nuclear matrix proteins with hydrophobic pockets are able to accommodate these lipids and their fatty acid tails. In 1963, Rees and coworkers found phosphorus-associated lipids in intranuclear fractions from rat liver nuclei and concluded further that most lipid-rich material may be in the heterochromatin associated with the nucleoli (Rees et al. 1963). It has also been demonstrated by histochemical techniques that chromatin contains phospholipids (La Cour et al. 1958) and further by electron-microscopic autoradiography after a pulse-chase with the lipid precursor 3H-glycerol (Rose and Frenster 1965). Moreover, (La Cour et al. 1958) suggested that phospholipids are associated with the chromosomes through mitosis, whereafter they dissociate from interphase chromatin, with the exception of heterochromatin. On the other hand, modern transmission electron-microscopic studies suggest that intranuclear phospholipids colocalize with RNA in the regions between the hetero- and euchromatin (Fraschini et al. 1992) rather than in the heterochromatin itself. These regions are called perichromatin fibrils or interchromatin granules, and they have been indicated as the sites of RNA transcription (Fakan 1986). Consistent with this, phospholipids quantitatively change parallel to transcriptional activity during the cell cycle (Fraschini et al. 1999). To summarize the literature discussed in this section, it seems evident that there is an intranuclear phospholipid metabolism in cells and that lipids are associated with non-histone chromosomal proteins, and this association is probably dependent on the cell-cycle phase and the status of the chromatin. Furthermore, phospholipids seem to be associated with active chromatin. Nuclear processes, such as DNA repair, transcription regulation and RNA dynamics, have been shown to be co- factored by these lipids (Hammond et al. 2004).

43 6. Zinc-dependent protein structures

Protein-nucleic acid interaction is a crucial event in the regulation of gene expression. Many DNA-binding proteins contain independently folded domains for the recognition of DNA, and these in turn belong to a number of domain families such as leucine zipper, the helix-turn-helix and families (Aravind et al. 2005, Deppmann et al. 2006). Zinc fingers, as the name implies, are finger-shaped structures in which a small group of conserved amino acids bind a zinc ion (Figure 8). The bound Zn2+ ion is structurally important, and the ability to nucleate the protein structure obviates the need for a large hydrophobic core for protein folding (Lemmon 2008). These fingers were first identified as a DNA-binding motif in transcription factor TFIIIA in Xenopus laevis. The DNA-binding motif in TFIIIA is composed of nine tandem units, each consisting of approximately 30 residues and containing two invariant pairs of cysteines and histidines as C2H2, which tetrahedrally co-ordinate one zinc ion each (Miller et al. 1985). Such C2H2 is the most typical signature, but Zn2+ can be co-ordinated by many other permutants of cysteines and histidines. Most zinc fingers are thought to interact with DNA. However, they are now also known to bind RNA, proteins and lipids (e.g. C1, PHD and FYVE domains contain Zn2+, Figure 7) (Matthews and Sunde 2002, Brown 2005). A zinc finger consists of two anti-parallel ȕ strands and an Į helix which is responsible for the DNA binding. Usually, a single zinc finger does not bind DNA with high specificity as it can only recognize two or three base pairs but when two zinc fingers are concatenated it is usually enough to bring sequence specificity as is the case with proteins of the nuclear receptor family (Claessens and Gewirth 2004). There are some naturally occurring proteins where 60 zinc fingers are in tandem, they bind more tightly and can specifically recognize very long DNA sequences (Branden and Tooze 1999).

Figure 8. Typical DNA-binding C2H2 zinc finger motif. The C2H2 zinc finger motif (2 Cys and 2 His residues bonded tetrahedrally to a Zn2+ ion) consists of a short antiparallel ß-sheet formed by two ß-strands and a hairpin turn, followed by an a-helix which forms the main contact surface with DNA. Figure adapted from (Lee et al. 1989).

44 AIMS OF THE STUDY

1. To find differentially expressed genes in differentiating intestinal epithelial cells

2. To characterize the function of one upregulated gene (SAP30L) in molecular detail:

2.1 To study the role of SAP30L in the Sin3A corepressor complex.

2.2 To study the domain structure of proteins of the SAP30 family in order to understand their role in the Sin3A corepressor complex.

2.3 To study the molecular evolution of the SAP30 protein family.

45 MATERIALS AND METHODS

1. Three-dimensional T84 epithelial cell model for the jejunal crypt- villus axis (I)

Human intestinal epithelial T84 cells (CCL 248) from the American Type Culture Collection (ATCC) (Rockville, MD) were cultured in Dulbecco's modified Eagle medium (DMEM) and Ham's F-12 (1:1) (Gibco BRL, Paisley, Scotland) supplemented with 5% fetal calf serum (FCS) and antibiotics (500 IU/ml penicillin and 100 ȝg/mL streptomycin; Gibco). Three-dimensional type I collagen gel cultures were established as previously described (Halttunen et al. 1996). Differentiation of T84 cells was induced by adding 20 ng/ml of human recombinant TGF-ȕ1 (hTGF-ȕ1, R&D Systems Europe Oxon, UK) to the cultures and the cultures were kept in 5% CO2 at 37°C for seven days.

2. Cell cultures and transfections (I - IV)

The human embryonic lung fibroblast cell line IMR-90 (CCL 186) was purchased from the ATCC. The cells were cultured in basal medium (Eagle, Gibco) supplemented with 10% FCS, 0.075% NaHCO3 and 2 mmol/l glutamine. Human embryonal kidney epithelial cells HEK293T (ATCC) were cultured in DMEM (Gibco), 5% FCS, 1 mM sodium pyruvate and 50 µg/ml of uridine. HeLa cells were cultured in RPMI1640 (Gibco) supplemented with 10% FCS and L-glutamine. Additionally, MCF7 epithelial breast cancer, COS-7 kidney fibroblasts and Daudi B lymphoblast cells were used in protein localization studies and these were cultivated according to the instructions from ATCC. All cell cultivation media were supplemented with penicillin and streptomycin antibiotics.

46 3. RNA isolation and detection methods (I)

3.1. RNA isolation and differential display PCR

Total RNA was isolated from control and hTGF-ȕ1-treated three-dimensionally cultured T84 cells with TRIzol reagent (Gibco) according to the manufacturer’s instructions and subjected to DNase I (Roche Molecular Biochemicals, Indianapolis, IN) treatment, after which RNA was extracted with phenol-chloroform- isoamylalcohol (Sigma Chemical Co., St. Louis, MO, USA). DD-PCR was performed according to the RNAmap™ protocol (GenHunter Corporation Nashville, TN, USA) with arbitrary 5' primers and anchoring 3' primers. The reactions were repeated twice with independently purified RNA in order to confirm the reproducibility of the results. The differentially expressed transcripts were recovered from the gel and sequenced using the ABI PRISM Dye Terminator Cycle Sequencing Ready Reaction Kit (Perkin Elmer, Foster City, CA, USA) as instructed by the manufacturer.

3.2. Quantitative PCR

Differential expression was confirmed using LightCycler technology in three independent RNA populations. One microgram of the Dnase I-treated total RNA was reverse-transcribed to cDNA using SuperScript II reverse transcriptase (Gibco) with 0.5 ȝg of oligo(dT) primer. This cDNA was then subjected to PCR using a LightCycler Fast Start Cyber Green kit (Roche Diagnostics, Espoo, Finland) according to manufacturer's instructions. The primers 3EX3S and 3EX4AS (Table 1, I) were used at a concentration of 0.5 ȝM. The cycling conditions were as follows; 96°C 10 min followed by 45 cycles at 96°C 10 s, 57°C 10 s and 72°C 10 s. The relative amounts of the blind-selected samples (control and TGF-ȕ-treated) were calculated by setting their cross points to the standard curve generated by a serial dilution of cDNA produced from T84 cells. The expression level of SAP30L mRNA in undifferentiated and differentiated T84 cells was normalized by the housekeeping gene glyceraldehydes 3-phosphate dehydrogenase.

3.3. Screening of cDNA library for the whole-length transcript

A human heart cDNA library (Rapid-Screen Arrayed cDNA Library Panel; OriGene Technologies, Rockville, MD, USA) was screened by PCR using primers 3EX3S and 3EX4AS (Table 1, I). The conditions of the PCR amplification for both Master Plate and Sub-plates were as follows: 95°C for 5 min, followed by 40 cycles of denaturation at 95°C for 45 s, annealing at 57°C for 30 s, and extension at 72°C for 60 s with a final extension at 72°C for 5 minutes. For the third round of screening, PCR was performed on single bacterial colonies. DNA from positive clones was

47 sequenced using both vector- and gene-specific primers indicated in Table I in original publication I. The accession number for SAP30L is AY341060.

4. cDNA cloning and protein production

4.1. cDNA cloning (I - III)

Complementary DNA cloning was performed by producing cDNA inserts by PCR amplification or oligo annealing, then by restriction enzyme digestion followed by ligation of the insert to the desired vector. The authenticity of the constructs was confirmed by sequencing. Point mutations were created using the QuikChange® Site Directed Mutagenesis Kit (Stratagene, La Jolla, CA, USA) according to manufacturer's instructions.

4.2. Production of GST-fusion proteins in E. coli (II & III)

GST-SAP30 and GST-SAP30L fusion proteins were produced in Escherichia coli (BL-21 strain) and purified with Glutathione Sepharose 4B beads (GE healthcare, UK) according to manufacturer's instructions.

4.3. Protein production by coupled in vitro transcription/translation (II & III)

In vitro transcription and translation was carried out with the TnT® Quick Coupled Transcription/Translation System (Promega, UK) according to the manufacturer's protocols.

4.4. Protein production and expression in mammalian cells (I – IV)

DNA was transfected using FuGENE 6 (for HEK293T cells) or FuGENE HD (for HeLa cells) reagents (Roche) according to the manufacturer's protocol. DNA was delivered into IMR-90 cells using Tfx-50 reagent (Promega).

48 5. Protein functional studies

5.1. Protein detection: immunoblotting and immunofluorescence (I – IV)

For SDS–PAGE, lysed cells or protein samples or immunoprecipitants were boiled in Laemmli buffer and resolved on SDS-PAGE. Proteins were transferred to a nitrocellulose membrane (Amersham Biosciences) and blotted with the primary antibodies and HRP-conjugated secondary antibodies (as indicated in original publications II-IV). Proteins were detected with the ECL Plus Western Blotting Detection System (Amersham Biosciences). Band intensities were quantified using the ImageQuantTMTL program (Amersham Biosciences). HEK293T, HeLa, IMR-90, MCF-7, COS-7 or Daudi cells were fixed with 4% paraformaldehyde in PBS [1x PBS (137 mM NaCl)] for 20 min and then washed with PBS and permeabilized for 10 min with 0.2% Triton X-100 in PBS. Unspecific binding of the antibodies was blocked by 1% BSA in PBS for 30-60 min before incubation of the cells with primary antibody usually at 1-2 mg/ml dilutions for 60 min at 37°C. After washes with PBS, the cells were incubated with secondary fluorophor-conjugated antibody as described in original publications I-IV. Slides were analyzed and photographed with a confocal microscope (Ultraview Confocal Imaging System, Perkin Elmer Life Sciences Inc., Boston, MA, USA).

5.2. Protein-protein interaction studies (II & III)

5.2.1. Immunoprecipitations

For the immunoprecipitation experiments, HEK293T cells were lysed in RIPA lysis buffer [1x phosphate buffered saline (PBS) containing 137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 2 mM KH2PO4), 1% Igepal-CA630, 0.5% sodium deoxycholate and 0.1% SDS] with freshly added protease inhibitors (Roche). Lysates were passed several times through a 21-gauge needle or sonicated to sheer DNA, incubated for 30 min on ice and centrifuged at 12 000 g for 20 min at 4°C. Supernatants were collected. Immunoprecipitations were carried out in end-over-end rotation overnight at 4°C with agarose-conjugated antibodies as indicated in original publications II and III. Precipitants were washed six to eight times either with RIPA lysis buffer containing 500 mM NaCl and 0.5% Igepal-CA630 or PBS containing 500 mM NaCl and 0.5% Igepal-CA630. Immunoprecipitants were analyzed by immunoblotting.

5.2.2. GST-pull-downs

For GST pull-downs, 1 µg of GST or GST fusion proteins coupled to beads were incubated with 3–36 µl 35S-labeled in vitro-translated proteins in binding buffer [1xPBS (137 mM NaCl), 0.1% Igepal-CA630 and freshly added protease inhibitors

49 (Roche)] in end-over-end rotation overnight at 4°C. The beads were washed six times with the binding buffer containing 200 mM NaCl. Protein complexes were subjected to SDS-PAGE and autoradiography as instructed in [TnT® Quick Coupled Transcription/Translation System (Promega)].

5.3. Protein-nucleic acid interaction studies (III & IV)

5.3.1. Electrophoretic mobility shift assays (EMSA) (III)

The [ -32P] ATP-labeled mtDNA tRNA-Leu(UUR) 150 bp probe is described elsewhere (Hyvarinen et al. 2007). The probe was incubated with 0.5 mg of GST- fusion protein on ice for 30 min in a bandshift buffer containing 50 mM Tris-HCl, pH 7.5, 125 mM NaCl, 2.5mM DTT, 0.5 mM EDTA, 1 mM MgCl2 and 4% glycerol. The reaction products were analyzed on 6% non-denaturing polyacrylamide gel, dried and autoradiographed.

5.3.2. Novel ladder-EMSA (L-EMSA) (III)

Because of the non-sequence-specific binding nature of the SAP30L and SAP30 proteins, a faster and simplified assay, L-EMSA, was developed for the protein- DNA interaction studies: 5 mg of fusion protein was incubated with 0.25 mg of 1 kb DNA ladder (GeneRulerTM, Fermentas, MD, USA) in PBS for 10 min at room temperature and protein-DNA complexes were run in EtBr containing 1% agarose gel with standard DNA gel loading buffer. Prior to use, this method was validated by comparing the DNA bandshifts in L-EMSA to shifts in conventional EMSA. The GST-fusion proteins used in Figure 1A in the original publication generated identical shifts in both assays (data not shown). Where indicated, PtdIns and PtdIns5P were added after the protein-DNA complex formation.

5.3.3. Interphase chromatin spreads (III)

Chromatin spread preparation for the SAP30L-GFP transfected HEK293T cells were performed as previously described (McGuinness et al. 2005) with the following modifications. Nocodazole was not added (because SAP30 family proteins do not associate with mitotic chromosomes, data not shown) and collected; PBS washed, hypotonically swollen cells were dropped from a height (1.5 m) onto the tilted microscope glass. Unfixed dried drop was counterstained with DAPI and photographed under the confocal microscope.

50 5.3.4. Chromatin isolation/ subcellular fractionation (III & IV)

Chromatin isolation/subcellular fractionation was performed as described elsewhere (Mendez and Stillman 2000).

5.4. Protein-lipid interaction studies (III)

PIP strips and arrays were purchased from Echelon Biosciences (Logan, Utah, USA). Protein-lipid blot assays were carried out by adding 0.5 mg/ml of GST-fusion proteins and were further processed as described in the manufacturer’s protocol. Each protein-lipid blot experiment was repeated at least once.

5.5. DNA-bending assay/ligation-mediated circulization assay (III)

The DNA-bending assay was performed essentially as earlier elsewhere (Paull et al. 1993). See also the legend to Figure 3 in original publication III.

5.6. Nucleosome preparations (III & IV)

Intact nucleosomes and tailless globular nucleosomes were prepared as described earlier (Macfarlan et al. 2005). The presence of solubilized nucleosomes was confirmed by DNA agarose gel electrophoresis and SDS-PAGE followed by Coomassie staining (see Supplementary Figure 4A, III). Histone proteins from calf thymus were purchased from Roche. For the GST-fusion pull-down experiments, 30 mg of nucleosomes, tailless nucleosomes or histone proteins were used. For the initial screening experiment (Figure 2B in III) 100 mg of histones were used.

5.7. HDAC activity and gene repression studies (II & III)

Histone deacetylase activity was measured using a Fluor de Lys AK-500-kit (Biomol) according to manufacturer's protocol. The HDAC inhibitor Trichostatin A (TSA) was added in control reactions at 1 µM concentration. In order to explore the role of class III HDAC enzymes, NAD+ coenzyme (N1511, Sigma-Aldrich) was added to reactions at 200 µM. Fluorescence was measured at 460 nm with a VICTOR2 1420 multilabel counter (Wallac, Perkin Elmer, Life Sciences). For the repression analysis, HEK293T cells were transfected with Gal4DBD- SAP30, Gal4DBD-SAP30L or Gal4DBD-SAP30L mutants along with 5xGal4-TK- LUC or 5xGal4-14D-LUC luciferase reporter plasmids as indicated. Transfection efficiency was normalized by measuring the activity of the ß-gal produced from the cotransfected pcDNA3.1-LacZ (Invitrogen). Twenty-four-hour post-transfection cells were split into two dishes and treated with either TSA at 200 nM or DMSO for

51 24 h. In cases where increased levels of nuclear PtdIns5P were desired, cells were treated with 500 mM H2O2 for 15 min and washed and incubated in normal growth media for 4 h. Thereafter, cells were harvested and luciferase activity was measured using the Luciferase Assay System (Promega). Measurements were done in duplicate from two independent experiments and the range of observed values is reported.

5.8. Mass spectrometric analysis of the N-terminal SAP30L peptides (III)

Wild-type and C29S, C30S, C38S, H70A, C74S and H77A mutants of SAP30L 1- 92 peptide were cleaved from GST by prothrombin yielding SAP30L 1-94 peptides with two additional amino acids from the GST vector. The samples were desalted by PD-10 columns (Amersham Biosciences) and concentrations were measured from –1 –1 the absorbance at 280 nm using e280 = 2560 cm M . Prior to measurements, the samples were further diluted with appropriate solvents (CH3CN/H2O/HOAc (49.5:49.5:1.0 v/v, pH 3.2) for denaturing solution conditions and NH4OAc buffer (10 mM, pH 6.8) for non-denaturing solution conditions). Mass spectrometric experiments were performed with a 4.7-T hybrid quadrupole Fourier-transform ion cyclotron resonance (Q-FT-ICR) instrument (APEX-Qe; Bruker Daltonics, Billerica, MA, USA), interfaced to an external electrospray ionization (ESI) source (Apollo-IIÔ). The samples were infused directly at a flow rate of 1.5 mL min–1, with dry N2 serving as the drying (10 psi, 200 °C) and nebulizing gas. ESI-generated ions were externally accumulated in a hexapole ion trap for 0.5-1.0 s and transferred to an Infinity ICR cell for SidekickÔ trapping, conventional “RF-chirp” excitation and broadband detection. A total of up to 256 co-added 1-Megaword time-domain transients were fast Fourier-transformed prior to magnitude calculation and external frequency-to-m/z calibration with respect to the ions of an ES Tuning Mix (Agilent Technologies, Santa Clara, CA, USA). All data were acquired and processed with Bruker XMASS 7.0.8 software.

5.9. Protein binding microarray (PBM) experiments and data analysis (III)

PBM experiments and analysis were performed as described by Berger et al. (2006) for four different GST-tagged protein constructs: full-length SAP30, full-length SAP30L, SAP30 residues 1-131 (SAP30 1-131), SAP30L residues 1-92 (SAP30L 1- 92), and a GST control (protein concentration ~1 mM). A fluorophore Alexa488- conjugated anti-GST antibody was applied to the protein-bound microarray to detect bound protein. The feature set of ~44,000 oligonucleotides present on the custom- designed (Agilent, Santa Clara, CA, USA) microarrays followed the design described by Berger et al. (2006), but were incorporated on the Agilent ‘4x44K’ array platform, allowing four independent PBM experiments to be performed simultaneously on the same microarray. To identify DNA binding site motifs, two approaches were used: 1) the approach described by (Berger et al. 2006) based on perturbations of the highest ranked 8-bp sequence, 2) de novo motif search on the

52 sequences from the top 20, 30 and 50 brightest microarray probes using the program MEME (Bailey et al. 2006).

5.10. Nuclear matrix preparations (IV)

Nuclear matrix preparations were done as previously described (Zeng et al. 1997). Subsequently the cells were fixed and stained as described in section 5.1.

6. Phylogenetic and molecular evolution studies (IV)

6.1. Protein sequence searches, gene loci data retrieval and multiple sequence alignments

Protein Psi-Blast (Altschul et al. 1997) searches with the full-length human SAP30L sequence were performed at the NCBI Web site (http://www.ncbi.nlm.nih. gov/BLAST/) on the non-redundant protein sequence database available on December 3, 2007. After six rounds of iteration, SAP30 and SAP30L orthologs below E-value 0.005 (except for C. elegans and P. nodorum, for which the E-values were 0.88 and 0.011, respectively) within the Metazoan, Plant and Fungi kingdoms were selected, and all redundant sequences were excluded. SAP30 and SAP30L proteins are encoded in four exons and variable usage of these exons is reported to yield multiple splicing variants (Korkeamaki et al. 2008). It is also predicted (USCS database) that the longer SAP30 and SAP30L cDNAs are composed of additional spliced-in exons upstream of these four. For the sake of clarity, only the full four exon-encoded proteins were included. All protein sequences were collected in FASTA format for further analysis (Table 1, IV). The SAP30 and SAP30L sequences were aligned using the MegAlign 5.06Ó program (DNASTAR Inc) with Clustal V (Higgins and Sharp 1989) or W (Thompson et al. 1994) default settings. The alignments were then shaded using the multiple sequence alignment editor GENEDOC (http://www.nrbsc.org/gfx/genedoc/index.html). Gene loci data were retrieved from a NCBI Map viewer (http://www.ncbi. nlm.nih.gov/mapview/).

6.2. Phylogenetic analysis and detection of functional divergence

PHYLIP version 3.67 (Felsenstein 1989) was used for the phylogenetic analyses. Distance, parsimony and likelihood analyses were performed using the protein alignment as input. Bootstrap values were obtained using SEQBOOT and creating 100 “delete-half jackknife” data sets. The distance analysis was performed using PROTDIST and subsequently NEIGHBOR with standard parameters, and the parsimony analysis using PROTPARS with standard parameters. The likelihood

53 analysis was made using PROML with standard parameters. In all cases, the "M" option for the analysis of multiple data sets created with SEQBOOT was invoked. DIVERGE version 2.0 (Gu and Vander Velden 2002) was used to detect type-I (Gu 1999) and type-II (Gu 2006) functional divergence. Clustal W alignments of the arthropodan and sarcopterygian clades for SAP30L and the sarcopterygian clade for SAP30 were created, and a distance analysis with 100 “delete-half jackknife” data set tests was performed using PHYLIP as described above. The alignment and the neighborjoining tree were used as input for the functional divergence analyses. P- values were derived from the ș and standard error values using the Z-score.

54 RESULTS

1. Identification of SAP30L in differentiated T84 cells (I)

Our group has previously set up an in vitro mesenchymal-epithelial cell co-culture model to mimic the intestinal crypt villus axis biology in terms of epithelial cell differentiation (Halttunen et al. 1996). In this model the fibroblast-induced epithelial cell differentiation from secretory crypt cells to absorptive enterocytes is mediated via transforming growth factor-ȕ (TGF-ȕ), the major inhibitory regulator of epithelial cell proliferation known to induce differentiation in intestinal epithelial cells. Using both differential display PCR (DD-PCR) and quantitative RT-PCR (LightCyclerÒ) it was shown that TGF-ȕ1 induces consistent and reproducible upregulation of a novel transcript denoted SAP30L (Figure 1 in I). The differentiated TGF-ȕ-treated cells expressed this transcript 2.0 times more than the unstimulated T84 cells. Screening of a heart cDNA library yielded a whole-length transcript 1.3 kb in size. Sequence analysis of this transcript showed that SAP30L is identical to an mRNA transcribed from the gene FLJ11526 located in chromosome 5q33.2. SAP30L was deposited in the Gene bank with access number AY341060.

1.1. SAP30L gene, mRNA and protein

The human FLJ11526 gene was named SAP30-like (SAP30L) because the encoded protein was found to be 70% identical to the human SAP30 protein. Amino acid comparison of SAP30L with SAP30 showed that SAP30 possesses in its N-terminus a 38-amino-acid stretch which was absent in SAP30L. The SAP30L gene has four exons (as does SAP30) and the expected size of the transcribed mRNA is 1281 nucleotides. Indeed, hybridization to a multi-tissue northern blot showed that a SAP30L-specific probe recognized an mRNA approximately 1.3 kb in size (Figure 2 in I). The mRNA was expressed in all tissues examined, with somewhat weaker expression in the liver and muscle and particularly abundant expression in the testis and placenta. According to the USCS gene expression database (Karolchik et al. 2008), the most prominent expression of SAP30L and SAP30 mRNA is in tissues of hematopoietic origin. Interestingly, there was also a larger transcript 6.5 kb in size which was abundantly expressed in brain and lung but not at all in liver and stomach. In fact, larger transcripts, with more than four exons, are predicted in the USCS database (Karolchik et al. 2008), but their authenticity remains to be established.

55 2. Identification of SAP30L as a member of the Sin3A corepressor complex (II)

SAP30 was originally characterized as a conserved member of the Sin3A corepressor complex (Zhang et al. 1997, Laherty et al. 1998, Zhang et al. 1998). Since the novel SAP30L showed 70% amino acid identity with SAP30 and thus is a homolog for SAP30, it was sought to establish whether SAP30L can also associate with the Sin3A corepressor complex. SAP30 protein was used as a positive control throughout the experiments in original work II. GST pull-down studies with in vitro transcribed and translated Sin3A proteins revealed that SAP30L associates with Sin3A and that the interaction requires the PAH3/HID region of Sin3A protein (Figure 9).

Figure 9. SAP30L binds directly to the PAH3/HID region in Sin3A. The Sin3A constructs used are illustrated on the left side of the experimental panel. The Sin3A proteins were produced by a coupled in vitro transcription/translation system and labeled with 35S-Methionine before subjection to pull- down experiments with GST-fusion proteins as indicated. SDS–PAGE was subjected to autoradiography in order to visualize Sin3A polypeptides. PAH, paired amhipathic helix; HID, histone deacetylase interacting domain.

It was further proved that the interaction of SAP30L with Sin3A also occurs in vivo by showing that myc-tagged mouse Sin3A co-immunoprecipitated (co-IPed) with myc-his-tagged SAP30L in transiently transfected HEK293T cells. Similarly, green fluorescent protein (GFP)-tagged SAP30L co-IPed with the myc-tagged Sin3A, while GFP alone was unable to co-IP with Sin3A. These results confirm that the interactions are independent of the tag used. In GST pull-down experiments with nuclear lysates of HEK293T cells, SAP30L associated with endogenous human Sin3A similarly to SAP30. A co-IP experiment with truncated SAP30L mutants demonstrated that the C-terminus of SAP30L is critical for the interaction with Sin3A (Figure 1 in II). Moreover, the transfected myc-his-tagged SAP30 and SAP30L were able to relocate transfected myc-tagged Sin3A (and endogenous Sin3A, data not shown) to the nucleolus: 42% and 7% of the SAP30L- and SAP30-

56 transfected cells, respectively, showed Sin3A-positive nucleoli, whereas none of the control vector-transfected cells showed Sin3A positive nucleoli. Congruent with IP experiments, cells transfected with SAP30L lacking C-terminus showed only 1% positive Sin3A nucleoli (Figure 6 in II). To summarize the interaction data discussed in this section, the C-terminus of SAP30L interacts directly with the PAH3/HID region in Sin3A and this mode of interaction is reminiscent of that of SAP30, suggesting that both proteins are assembled to the core-Sin3A co-repressor complex in an identical manner. Both in vitro pull-down and in vivo IP and immunofluorescence relocalization experiments strongly suggested interaction of SAP30 and SAP30L with Sin3A.

2.1. SAP30L associates with HDACs and represses transcription

In order to study whether SAP30L associates with HDACs, HDAC activity measurements from GST pull-down precipitants from the HEK293T cell nuclear extracts were performed. GST-SAP30L pulled down HDAC activity comparably with that of GST-SAP30, and this activity was sensitive to TSA, an inhibitor of class I and II HDACs. Addition of NAD+, which is an essential cofactor for the activity of class III HDACs, did not increase HDAC activity, further suggesting that class III HDACs do not contribute to the HDAC activity associated with SAP30 proteins in this assay (Figure 10A). An intact C-terminus of SAP30L was necessary to associate HDAC activity, as shown by a series of mutants of SAP30L (Figure 3B in II). When detected by Western blotting, the pull-down experiments demonstrated that GST- SAP30 and GST-SAP30L interacted with class I HDACs 1–3 (Figure 3C in II). The results suggested that SAP30L forms a complex with Sin3A possessing HDAC activity. We therefore studied whether this complex can execute transcriptional repression when tethered to different promoters. The transcriptional repression activity of SAP30 has previously been shown by utilizing the luciferase reporter assay, where fusion of SAP30 and the Gal4 DNA-binding domain was tethered in front of the luciferase gene to promoters harboring five Gal4-binding sites (Laherty et al. 1998). When SAP30L was tethered to 14D (Figure 10B) and thymidine kinase promoters (Supplementary Figure 1 in II), it was shown to repress transcription 23- and 10-fold, respectively, compared to Gal4 alone. It was also noted that in both promoters SAP30L was able to repress transcription 1.6 to 2.0 times more efficiently than SAP30. TSA treatment greatly diminished the repressive activity of SAP proteins, suggesting that HDAC activity plays an important role in mediating the repressive capability (Figure 10B). Experiments with mutant versions of SAP30L showed that an intact C-terminus of SAP30L was needed for full repression. Taken together, these findings suggest that SAP30L represses transcription, and that this repression involves the recruitment of Sin3A and HDACs (Figure 4B in II).

57 Figure 10. SAP30L associates with histone deacetylase activity and represses transcription. A) GST-fusion pull-downs from the HEK293T nuclear extracts were performed and HDAC activities were measured by fluorescence-based assay (Fluor de Lys kit). GST and GST-SAP30 were used as negative and positive control, respectively. The basal level of fluorescence (blank) and sensitivity in the experiment were defined by measuring fluorescence from the assay buffer and from the 1 µM deacetylated standard respectively (white bars). NAD+ coenzyme and TSA were added when indicated. Arbitrary fluorescence units are represented as HDAC activity units. B) HEK293T cells were cotransfected with the 5xGal4-14D luciferase reporter vector (contains five binding sites for the Gal4 DNA-binding domain), Gal4DBD (Gal4 DNA-binding domain) fusions and LacZ-vector as indicated. Twenty-four hour post-transfection cells were treated either with TSA or DMSO (vehicle) for another 24 h. Lysed cells were analyzed for luciferase and ß-gal activity. The results were normalized by the activity of the ß-gal produced. The histogram illustrates the average fold- repressions of the Gal4DBD-fusions compared with Gal4 alone. A&B) Shown are the means of two experiments performed in duplicate and the error bars represent the range of measurements.

3. Identified domains and functional motifs in SAP30L and SAP30 proteins

3.1. Nuclear localization signal (NLS) (I & II)

In transient transfection experiments on IMR-90 fibroblasts, we were able to show that the wild-type SAP30L-GFP fusion protein is nuclear. The best understood system to transport proteins to the nucleus is mediated by their classical basic NLS whose amino acids bind nuclear importin proteins (Lange et al. 2007). We identified a canonical basic NLS in the middle of the SAP30L protein responsible for its nuclear targeting (Figure 5A in I). Transfection of the GFP fusion protein, which had only the putative nuclear localization and six flanking amino acids on either side (pEGFP-NLS), also resulted in nuclear localization of the protein (Figure 5B in I), providing further evidence for the functionality of the signal. Mutation in the NLS (KRKRK ĺ KSNRK) disturbed this nuclear localization to some extent, causing some cytosolic retention of the SAP30L-GFP fusion protein. Also myc-his-tagged

58 wt SAP30L was found in the nucleus of the studied cell lines (MCF-7, COS-7, IMR- 90, T84, Daudi, HEK293T and HeLa), indicating that SAP30L is a nuclear protein when transiently expressed, independent of the tag used. Furthermore, NLS does not seem to contribute to the nucleolar targeting of SAP30L (see next section), since the (KRKRK ĺ KAAAK) mutant is still nucleolar (Figure 5 in II).

3.2. Nucleolar localization signal (NoLS) (II)

Transfected SAP30L showed a patchy staining pattern which only partially colocalized with PML bodies (Figure 5D in I). When transfected cells were stained with a nucleolar marker nucleophosmin (NPM or B23), marked colocalization was detected (Figure 5A in II). SAP30L was found to harbor a stretch of basic residues consistent with a proposed NoLS consensus sequence (R/K-R/K-x-R/K) (Horke et al. 2004) in its C-terminal region. Insertion of eight alanines in the 120-127 region abolished the nucleolar localization entirely (Figure 5C, D in II). Many nuclear and nucleolar proteins such as HSP70, EBNA-5 (Pokrovskaja et al. 2001), p53 and MDM2 (Klibanov et al. 2001) are known to accumulate in the nucleolus under proteotoxic stress caused by proteasome inhibitor MG132, and we observed that SAP30 and SAP30L also accumulated to the nucleolus upon proteotoxic stress. Furthermore, it was found that overexpressed SAP30L and SAP30 are able to target Sin3A to the nucleolus.

3.3. Protein-protein interaction domain (II)

In SAP30, a C-terminal region has been shown to be critical for Sin3A interaction (Laherty et al. 1998). Amino acids from position 120 to 140 in SAP30L (159-179 in SAP30) were deemed critical for Sin3A interaction (Figure 1D in II). This region was also sufficient for the self-association of SAP30L, suggesting that the C- terminus is a common protein-protein interaction element (Figure 2 in II). This same region also harbors the NoLS (Figure 5 in II). However, the full-length C- terminus (residues 120-183) in SAP30L was needed for full self-association, association with HDACs and repressive activity (Figures 2B, 3B and 4B in II, respectively), suggesting that protein-protein interactions are critical for the correct subnuclear localization and function of SAP30L. In addition, the fast turnover of the SAP30L protein seems to correlate with its ability to interact with other proteins (Figure 7 in II).

59 3.4. Zinc-dependent structure (III)

It was noted that SAP30L lacking the C-terminal part (aa 1-120) evinced increased and the N-terminally truncated (aa 61-183) SAP30L decreased stability (over five fold) compared to the full-length SAP30L (Figure 7 in II). Moreover, proteasome- dependent degradation of the N-terminally truncated SAP30L was observed. This led us to hypothesize that the N-terminal part of SAP30L contains a co-factor essential for its folding and stability and that the 61-183 mutant lacks crucial residues for co-factor coordination, leading to misfolding and proteasomal degradation of the apo-form protein. In a stretch of 49 residues aa 29-77 in human SAP30L, we identified four cysteine and two histidine residues, which suggested the possibility of a metal coordinating motif such as for zinc or copper. These residues are completely conserved in a phylogenetic comparison of SAP30/SAP30L sequences from several species, including fruitfly and human (Figure 2A in III) and throughout the animal kingdom except in the nematode Caenorhabditis elegans (Figure 2 in IV). The N- terminal peptide of SAP30L (aa 1-92) and mutants in which the putative metal- coordinating cysteine residues were replaced by serines (C29S, C30S, C38S, C74S), and histidines by alanines (H70A, H77A), one at a time, were produced as a GST fusion in E. coli. It was noted that the SAP30L mutants C29S, C38S, C74S and H77A completely degraded into small peptide fragments when expressed in E. coli (Figure 2C in III) and when transiently transfected into mammalian cells as myc- tagged fusions (data not shown). This suggested that the N-terminal peptide of the proteins of the SAP30 family contains a prosthetic group which is co-ordinated by four residues (C-C-C-H), a pattern reminiscent of that in some zinc-dependent structures (Carballo et al. 1998). To establish whether SAP30L binds zinc, we determined ESI Q-FT-ICR mass spectra for an N-terminal peptide of SAP30L (aa 1-92) in denaturing and non- denaturing conditions. In non-denaturing conditions, a 62.94-Da increase in the mass of the SAP30L peptide was detected, consistent with the binding of one Zn2+ 2+ cation (Figure 2B in III and Table 3). The binding of Zn (maverage = 65 Da) by zinc finger domains is always accompanied by loss of two protons (deprotonation of two coordinating cysteines), which gives a theoretical 62.92-Da increase in mass (Fabris et al. 1999). The calculated and determined masses for the peptides SAP30L 1-94, C30S and H70A are listed in Table 3. The SAP30L mutants C29S, C38S, C74S and H77A had completely degraded into small peptide fragments and could not be measured (Figure 2C and Supplementary Figure. 2B in III).

60 Table 3. The calculated and experimentally determined masses for the wild-type SAP30L 1-94 peptide (contains two residues from the vector), and the mutants C30S and H70A.

mexp Dmexp-calc a b c d peptide (Da) mcalc (Da) (Da) Elemental composition

apo-w.t. 10413.33 10413.28 +0.05 C447H726N140O137S5

holo-w.t. 10476.25 10476.20 +0.05 C447H724N140O137S5Zn1

apo-C30S 10397.31 10397.35 +0.04 C447H726N140O138S4

holo-C30S 10460.24 10460.22 +0.02 C447H724N140O138S4Zn1

apo-H70A 10347.26 10347.30 +0.04 C444H724N138O137S5

holo-H70A 10410.18 10410.32 +0.14 C444H722N138O137S5Zn1 a The data are presented only for the peptide variants comprising residues 1-94. b Experimentally determined, most abundant isotopic mass. c Calculated, most abundant isotopic mass based on the sequence-derived elemental composition. d In the case of zinc binding, a loss of two protons was considered.

3.4.1. Sequence-independent DNA binding and bending (III)

Since the best recognized ligand for zinc fingers is DNA (Klug 1999), we analyzed whether the zinc-dependent structure in proteins of the SAP30 family also bind DNA. An electrophoretic mobility shift assay (EMSA) was carried out using various GST-SAP30L fusion protein constructs incubated in the presence of radiolabeled mitochondrial DNA. EMSAs established that SAP30 family proteins bind DNA and that the N-terminal zinc-dependent structure is critical for this binding (Figure 1A in III). However, it was shown that the zinc-dependent structure is not capable of sequence-specific DNA binding, as demonstrated in a protein binding microarray (PBM) experiment (Figure 1B and Supplementary Figure 1 in III). To further define the residues responsible for DNA binding in SAP30L, we developed a simplified EMSA assay which utilizes a commercial DNA ladder (L-EMSA) (see Materials and Methods section 5.3.2). Using L-EMSA it was demonstrated that the zinc- dependent structure alone (aa. 1-77) does not bind DNA, while it needs the following hydrophobic region (aa. 78-84) and polybasic region (i.e. NLS signal, aa. 85-92) for sufficient and maximal binding, respectively (Figure 1D in III). The NLS mutant (KAAAK) which disrupts the polybasic region of SAP30L showed severely impaired DNA binding, further emphasizing the role of the polybasic region in DNA binding. Intriguingly, a short peptide spanning residues 78 to 92 of SAP30L alone was sufficient for DNA binding. On the other hand, the zinc-dependent structure was needed for the stability and correct folding of the DNA-binding domain, as it was demonstrated that either disruption of the zinc-dependent structure or depletion of zinc cation abolished DNA binding (Figure 1E and Supplementary Figure 2C in III). To address the DNA binding of SAP30L in vivo, interphase chromatin spreads were prepared from SAP30L-GFP transfected cells. GFP-tagged SAP30L associated with chromatin in vivo when hypotonically swollen HEK293 cells were splashed on a microscope slide and counterstained with DAPI (Figure 6A in III). It should be

61 noted that SAP30L is not a component of chromatin in the same way as histones, since it is not present in mitotic chromosomes. Furthermore, in subcellular fractionation experiments, wild-type SAP30L associated strongly with the chromatin-enriched fraction (CEF), whereas the NLS mutant (KAAAK) showed significantly compromised CEF association (Figure 6B in III). The high mobility group (HMG) proteins provide a well-known example of a protein family in which sequence-independent DNA binding exist accompanied by bending of the DNA (Hock et al. 2007). Using a T4 DNA ligase-mediated circulization assay, we showed that SAP30L is able to induce significant bending of the DNA, and the zinc-dependent structure alone was sufficient for DNA bending (Figure 3 in III).

3.4.2. Monophosphoinositides (PtdInsP) binding domain (III)

In addition to their role as a DNA-binding module, zinc-dependent structures have recently been shown also to mediate protein-lipid interactions (Matthews and Sunde 2002, Gozani et al. 2003, Kaadige and Ayer 2006). Pf1 and ING2 are both Sin3A- binding proteins (Figure 6), having a phosphoinositide (PI)-binding polybasic region (PBR) following the first PHD zinc finger (Kaadige and Ayer 2006). We identified a similar modular organization in SAP30L, which also contains a zinc-binding element followed by a PBR (85RNKRKRK91) (Figure 5A in III). Interestingly, in SAP30L the PBR motif was also shown to act as an NLS (see chapter 3.1.). To investigate the PI binding of SAP30L and SAP30, the GST fusion proteins were tested for binding to a variety of immobilized lipids, as depicted in Figure 5B-E in original publication III). Both GST-SAP30L and GST-SAP30 bound the monophosphorylated phosphoinositides PtdIns3P, PtdIns4P and PtdIns5P (Figure 11).

Figure 11. SAP30L and SAP30 bind monophosphoinositides. The indicated GST fusion proteins (0.5 µg/ml) were incubated with a PIP array as described in the Materials and Methods section. Shown are the pmol quantities of the indicated phosphoinositides.

The full-length GST-SAP30L bound most tightly to PtdIns5P, followed by PtdIns3P and PtdIns4P. The PtdIns5P-binding of GST-SAP30L was four-fold higher compared to PtdIns3P, and eight-fold higher compared to PtdIns4P. The mapping

62 studies with a mutant and truncated version of GST-SAP30L revealed that the same residues dictate the DNA and PIP interaction (Figure 5E in III and Table 4).

Table 4. Summary of mapping results on DNA and PIP interactions.

SAP30L DNA PIP Construct interactiona interactiona wt 1-183 +++ +++ 61-183 - - 40-183 - NA. 35-183 NA. - 25-183 +++ +++ 1-77 - - 1-84 ++ ++ 1-92 +++ +++ 1-120 +++ +++ 84-92 - - 78-92 +++ +++ del50-69 - - 87KAAAK91 + + a NA., not available; -, no interaction; +, weak interaction; ++, moderate interaction; +++, strong interaction.

To conclude chapter 3.4., proteins of the SAP30 family contain an N-terminal zinc- dependent structure which together with the following hydrophobic and polybasic region is able to bend and interact sequence-independently with DNA and, in addition, this very same region also interacts specifically with nuclear signaling lipids, monophosphoinositides.

3.5. Acidic central domain contributing to histone interaction (III)

A domain architecture of SAP30L and SAP30 similar to that of the HMG proteins was noted, both having an N-terminal DNA-binding/bending domain followed by an acidic domain (Figure 4A in III). As some reports demonstrate that the acidic region of HMG proteins mediates interactions with H1 or core histones (Carballo et al. 1983, Bernues et al. 1986), we set out to examine whether SAP30L may also associate with core histones and nucleosomes. In a pull-down experiment, GST- fused SAP30L was able to interact with histones 2A and 2B in purified DNA-less form, in DNA-containing nucleosomal form and in tailless globular nucleosomal form (Figure 4B&C in III). The central acidic domain in SAP30L contributed to each of the three types of histone interaction. Nonetheless, the presence of DNA abolished the affinity of certain mutants (SAP30L 1-120 and del109-113), reflecting that the residues responsible for the interaction with nucleosomes and naked histones are slightly different. Furthermore, in confocal microscopy, colocalization

63 of histone 2B and SAP30L was observed, with simultaneous relocalization of histone 2B around the nucleolus in response to overexpression of SAP30L (Figure 4D in III). Coexpression with wild-type SAP30L, but not with the SAP30Ldel109- 113 mutant, increased the perinucleolar localization of H2B from 10% to over 80%. To conclude this section, proteins of the SAP30 family functionally bind the globular domain of the core histones and nucleosomes, based on the results from 1) in vitro GST pull-downs and 2) in vivo relocalization of H2B.

3.6. Nuclear matrix targeting signal (IV)

Our previous subcellular fractionation experiments showed that nuclear retention of SAP30L is achieved by interaction with DNA through the N-terminal domain (see section 3.4.1.). The same experiments also demonstrated that the C-terminus has a role in nuclear retention, since C-terminally truncated mutant of SAP30L leaked to the cytoplasm in subcellular fractionation experiments (Figure 6B in III). When the nuclear matrix was isolated, we observed that staining of the perinucleolar ring was resistant to Triton-X and DNAse I treatments, indicating that the proteins of the SAP30 family remained attached to the nuclear matrix in the perinucleolar ring region (Figure 6A in IV). Furthermore, as shown in Figure 6C in original publication IV, solubilization of chromatin with micrococcal nuclease does not detach wt SAP30 or wt SAP30L from the nuclear matrix. To conclude, attachment of proteins of the SAP30 family is dependent on an intact C-terminus, which thus constitutes a nuclear matrix targeting signal (NMTS). The C-terminal NMTS is in respect of hydropathic properties similar to that of the NMTS in proteins of the Runx family (Javed et al. 2005) (Figure 6D in IV).

4. The subcellular localization, chromatin attachment and repressional activity of SAP30L is regulated by its interactions with DNA and monophosphoinositides (III)

It is of note that the same region in SAP30L interacts with both DNA and PIs, as summarized in Table 4. We therefore asked whether the association of SAP30L with chromatin is regulated by PIs, which could compete for the same binding sites and thus detach SAP30L from chromatin. This issue was first addressed with L-EMSA in vitro, with pure protein, lipid and DNA components. The mobility shift generated by binding of SAP30L to DNA in the L-EMSA assay was greatly diminished after addition of equivalent molar amounts of PtdIns5P, but not when PtdIns was added (Figure 12). The interaction of SAP30L with nucleosomes and Sin3A remained unchanged after addition of PtdIns5P, indicating that monophosphoinositides blocks specifically the DNA binding of SAP30L (Figure 6F in III).

64 Figure 12. PtdIns5P competes with DNA on binding to SAP30L. A L-EMSA with GST-SAP30L 1- 92 in the presence of equivalent molar quantities of PtdIns or PtdIns5P. This is an inverted color image of the agarose gel where DNA is stained with ethidium bromide.

Our in vivo approach utilized exposure of the cells to hypoxic conditions with H2O2 treatment, which has previously been shown to increase the amount of intranuclear PtdIns5P (Jones et al. 2006). Brief treatment of cells with H2O2 led to a significant detachment (five-fold) of myc-tagged SAP30L from the chromatin-enriched fraction as assayed by subcellular fractionation in HEK293 cells (Figure 6D in III). Detachment of SAP30L from chromatin was also visualized with confocal microscopy of HeLa cells; 9% of non-treated and 41% of H2O2-treated cells expressed cytoplasmic GFP-SAP30L (Figure 6E in III). Finally, the detachment of SAP30L from chromatin subsequently led to a attenuated repression activity as assayed by a Gal4 fusion system. As shown in Figure 6G (III), reduced repressive activity was observed both in the PBR/NLS mutant (SAP30L KAAAK), which lacks DNA binding and mimics PtdInsP binding, and after H2O2 treatment, which increases nuclear PtdIns5P. Taken together, the association of SAP30L with chromatin is dependent on intact PIP/DNA-binding domains, and PtdIns5P disrupt this association, leading to decreased transcriptional repression through SAP30L.

5. Evolution of the SAP30 family of transcriptional regulators (IV)

SAP30L seems essential to eukaryotic biology, as it is found in animals, plants and fungi, as well as in some taxa of unicellular eukaryotes. When viewing the human SAP30 and SAP30L genes located in chromosome bands 4q34.1 and 5q33.2, respectively, it is noteworthy that similar genes (GALNT and HAND gene families) flank the SAP30 and SAP30L genes in their respective loci (Figure 1 in IV). In fact, these two chromosomes are known to share duplicated segments (Friedman and Hughes 2003). The SAP30L harbored 400 kb microsynteny region has been

65 pinpointed as an interchromosomally duplicated block in a study where the sequence of the whole chromosome five was analyzed (Schmutz et al. 2004). The GALNT-SAP microsynteny has been preserved between fish and human chromosomes, and between human chromosomes 4 and 5, and is thus at least 450 million years old. Phylogenetic trees were generated from the Clustal W alignment of 63 members of the SAP30 protein family presented in Figure 2 and Supplementary Table 1 in original publication IV. All three phylogenetic tree-constructing methods (distance, parsimony and likelihood) gave identical tree topologies. Furthermore, in all trees SAP30 proteins clearly fall into one monophyletic group with statistical significance (Figures 3 & 4 and Supplementary Figure 4 in IV). The presence of SAP30L and the absence of SAP30 in the fish (Danio rerio and Tetraodon nigroviridis) genomes indicates that the SAP30 gene originated from the ancestral SAP30L gene by duplication of a chromosome segment after the appearance of fishes (Actinopterygii, ray-finned fishes) but before the appearance of amphibians (Sarcopterygii, lobe- finned fishes). When analyzing the distance tree, in which the branch lengths correspond to evolutionary relationships, it is evident that SAP30L is the ancestral protein (Figure 4 in IV). Animal SAP30 proteins form a peripheral cluster in the tree, whereas animal SAP30L proteins settle closer to the plant, yeast and mycetozoan members of the family. In addition, it is noteworthy that SAP30 orthologs from frogs to humans (Sarcopterygii) are much more widely dispersed than are the SAP30L orthologs in the corresponding species. These findings suggest that since their divergence by segmental duplication from a common ancestor, the evolutionary rate in SAP30 proteins has been much higher than in SAP30L proteins. This is what is canonically thought to occur in duplicated genes, where the new copy will evolve unencumbered by the selective constraints imposed on its progenitor (Ohno 1970). Interestingly, milder selective constraints in the branch leading to SAP30 did not predispose ancestral SAP30 to random mutations but rather to mutations with a pattern reminiscent of that in typical functional divergence. Both site-specific shifts in evolutionary rate and cluster and site-specific amino acid property shifts (see Materials and Methods, section 6.2.) were detected with statistical significance (Figure 5 in IV). As shown in original publication IV, phylogenetic analysis and biochemical experiments suggest that SAP30 has diverged functionally from the ancestral SAP30L by accumulating mutations which have caused attenuation of one of the original functions, association with the nuclear matrix. Further, these findings show that proteins of the SAP30 family possess many characteristics typical of nuclear scaffolding proteins (Zaidi et al. 2007): They are able to interact with co-repressors (e.g. Sin3A, N-CoR) and chromatin (both naked and nucleosomal DNA) and associate with the nuclear matrix.

66 DISCUSSION

1. The domain structure of the SAP30 family proteins indicates nuclear scaffolding and transcriptional regulatory functions (I - IV)

In this work a novel component of the Sin3 corepressor complex, SAP30L, was discovered. SAP30L, together with SAP30 protein, constitutes a conserved protein family in which SAP30L is the ancestral protein. Proteins of the SAP30 family bind to the PAH3/HID region of Sin3A through their C-terminal part (Figure 9). SAP30 family proteins induce transcriptional repression via recruitment of Sin3A and HDACs (II). Originally, SAP30 was identified as a stabilizing or “bridging” component of the Sin3A complex (Zhang et al. 1997, Laherty et al. 1998, Zhang et al. 1998). However, the domain structure and the molecular function of SAP30 proteins remained unknown. We therefore carried out a structure-function mapping of SAP30 and SAP30L in order to elucidate their role in the Sin3a corepressor complex (III). A series of mutants of SAP30L revealed that the N-terminal region is critical for the stability of SAP30 proteins. We found that these proteins have an N-terminal zinc-dependent module in which a zinc ion proved to be critical for the stability of the protein (Figure 13b). Most typical zinc-finger structures (as in TFIIIA) consist of approximately 30 residues with two pairs of cysteines and histidines (C2H2) which tetrahedrally co-ordinate one zinc ion (Miller et al. 1985). The zinc-dependent structure in SAP30 proteins deviates from this rule, as it is 50 residues long and seems to be of C2CH-type. In fact, while the writing process of this doctoral thesis was under way He et al. (2009) published the nuclear magnetic resonance (NMR) structure for SAP30 and confirmed our results in (III) that SAP30 proteins are of C2CH-type large zinc fingers. There is, however, a precedent for a large zinc- binding module, namely THAP domains, which are conserved zinc-dependent modules capable of sequence-specific DNA binding and are 44-59 residues in length (Clouaire et al. 2005). The THAP domain, however, contains other conserved elements in addition to the C2CH module, making it distinct from the zinc-binding motif in SAP30L. By reason of its DNA-binding behavior, the zinc dependent structure in SAP30 proteins seems to be a typical ‘zinc finger’ in function (Figure 13). One zinc finger is able to interact only with two to three nucleotides in the DNA and in line with this, we detected no sequence-specific binding of SAP30 proteins (III). Many studies have shown that SAP30 copurifies with other components of the Sin3A complex with roughly equivalent stoichiometry (Zhang et al. 1997, Laherty et al. 1998, Zhang et al. 1998, Lai et al. 2001, Skowyra et al. 2001, Kuzmichev et al.

67 2002, Fleischer et al. 2003) and hence is concluded to be a core member of the Sin3A complex. One may surmise that SAP30 family proteins, as core components, stabilize the Sin3A repressome to sites determined by transcription factors capable of sequence-specific DNA binding. In fact, transcription factor YY1-mediated repression has been shown to be enhanced by the presence of SAP30 in a dose- dependent manner (Huang et al. 2003). The fact that SAP30L was able to induce significant bending in the DNA suggests that they are also able to stimulate and assist the binding of transcription factor to DNA target sites, as is the case with the proteins of the HMGB family, which also bind and bend DNA in a sequence- independent manner (Grasser et al. 2007). Mutations in SAP30L which attenuated DNA binding correlated with decreased repression activity. This further supported the conception of an active stabilizing role for SAP30 proteins in the Sin3A complex in vivo (III). Proper DNA binding of SAP30 proteins was also responsible for their nuclear localization, since mutating NLS significantly reduced their affinity to DNA and SAP30 proteins became more soluble and leaked to the cytoplasm (II & III). We found further analogies in structures between the SAP30 and HMG proteins: proteins in both families contain an N-terminal DNA-binding/bending domain followed by an acidic region which contributes to histone binding (Figure 13), (III), (Carballo et al. 1983, Bernues et al. 1986). This would imply that SAP30 proteins, together with naked DNA interactions, can stabilize the Sin3A complex further by interactions also with nucleosomally chromatinized DNA. Such a conception is supported by the finding that SAP30L-GFP associated with chromatin in vivo when interphase (but not mitotic) chromatin spreads were prepared (III). The absence of a nucleosome association of the SAP30Ldel109-113 mutant leads to the impaired in vivo repressional ability of SAP30L, as it is halved compared to wild-type SAP30L (Korkeamaki et al. 2008). The last domain/motif we experimentally identified in SAP30 proteins was the nuclear matrix targeting signal in the C-terminal region (Figure 13), (IV). NMTS in SAP30 proteins is similar in hydropathic properties to NMTS in the RUNX family of transcription factors. NMTS in both families comprises a stretch of hydrophobic residues flanked by hydrophilic residues. Interestingly, C-terminal NMTS is evolutionarily the oldest region, as it is conserved from yeast to human. In the literature, Sin3A has also been reported to associate with the nuclear matrix (Imai et al. 2004). Possibly SAP30 proteins provide more contacts with the nuclear meshwork and as a result further stabilize the Sin3A complex in order to obtain its maximal repression.

68 Figure 13. Various domains of SAP30L identified in original publications I – IV. A) Zn, zinc- coordinating motif; DNAbd, DNA-binding domain; PIPbd, PIP-binding domain; NLS, nuclear localization signal; acidic region, a central region contributing to histone binding; NoLS, Nucleolar localization signal; protein bd, protein-binding domain; NMTS, Nuclear matrix targeting signal. B) Schematic representation of the N-terminal zinc-coordinating motif of SAP30L.

The domain structure of the SAP30 family proteins shows many characteristics typical of nuclear scaffolding proteins (Zaidi et al. 2007). The latter bind DNA, associate with the nuclear matrix, and interact with co-repressors (such as Sin3A) and are thus nuclear transcription regulatory factors which assemble in focally organized nuclear microenvironments associated with the nuclear matrix. In the case of the SAP30 family proteins, the NoLS is a part of the NMTS motif and targets the SAP30-Sin3A complex to the perinucleolar matrix. Through interactions with DNA/chromatin, histones, nuclear matrix and Sin3A complex, SAP30 family proteins form a functional and stable repressome regulating gene expression by transcriptional repression (Figure 15). The importance of these interactions is demonstrated in Table 5, showing that mutants defective in various interactions lead to impaired repression of transcription in Gal4-guided promoter tethering assays.

Table 5. Defects in the nuclear scaffolding function of SAP30L affecting its repressive activity.

Repressive Construct Defect activity % ref. wtSAP30L ʚ 100 II SAP30L 87KAAAK91 DNA&PIP binding 83 III SAP30Ldel109-113 histone binding 58 Korkeamaki (2008) & III SAP30L 1-120 Sin3a binding & nuclear matrix association 17 II SAP30L 1-140 nuclear matrix association 54 II & IV

2. Evolution of the SAP30 family (IV)

Regulation of gene transcription by histone acetylation and deacetylation is an evolutionarily conserved mechanism. Perhaps the most common histone deacetylation-mediating complex is the Sin3/HDAC corepressor complex. This complex consists of seven to eight core proteins, many of which are conserved from yeast to man (Silverstein and Ekwall 2005). The main exception is the Sin3/HDAC

69 complex in S. pombe (belonging to the Taphrinomycotina subphylum), which is reported to lack SAP18 and SDS3 protein components (Ekwall 2005). Similarly, we found that the ancestral SAP30 family member SAP30L is absent in S. pombe, whereas it is present (as well as SAP18, Viiri unpublished observation) in antecedent taxa such as Chlorophyta. Probably the loss of SAP18, SDS3 and SAP30 family components in S. pombe is a superimposition of their common task in chromatin biology of which Taphrinomycotinas are adapted to deal with in an alternative way. The SAP30 protein family emerged from a single chromosome block duplication event from the ancestral SAP30L gene in “flesh-finned fishes” about 450 million years ago. Gene duplication is generally regarded as the prime factor in evolution, especially from fish to mammals (Ohno et al. 1968). Gene duplication is a convenient way to provide raw material for evolution in that at the same time the function of the ancestral copy is preserved. In the duplicated paralog, SAP30, the evolutionary rate of amino acid substitution has been significantly higher compared to the ancestral template, SAP30L. However, substitutions were not random, suggesting possible neofunctionalization (Zhang 2003) of the SAP30 paralog. By our methods we were able to detect attenuation of one of the original functions, association with the nuclear matrix. We also found that SAP30 is a 1.6- to 2.0-fold weaker repressor than SAP30L. Thus, neofunctionalization was not experimentally proved but rather that the original functions were weaker in the duplicated paralog, SAP30. On the other hand, the expression pattern of the proteins of the SAP30 family shows striking differences: the full-length, 30 kD four-exon-isoform of SAP30 is expressed throughout the tested cancer cell lines, whereas SAP30L is expressed in the same cell lines only as a short 18 kD isoform and the 28-kD four- exon-isoform of SAP30L has so far been detected only in the nuclear blood cells (Viiri et al. unpublished observation). Interestingly, Fleischer et al. (2003) found a novel 28-kD protein in the Sin3A complex in a lymphoblast K562 cell line which they failed to identify. Based on its molecular mass, the 28 kDa protein probably represents SAP30L. To conclude, SAP30 and SAP30L are functionally divergent and show differences in expression pattern. Tissue-specific expression suggests that Sin3A complexes are alternatively furnished by SAP30 isoforms to provide tissue-specific gene expression.

3. Novel proposed mechanism: Regulation of protein-DNA interactions by nuclear phospholipids (III)

One intriguing finding in this work was the ability of nuclear phosphoinositides (PI) to displace DNA from proteins of the SAP30 family. Our In vitro mapping data showed that PIs compete with DNA for the same binding sites in the zinc-dependent structure in SAP30L protein. Afterwards this same conclusion was also drawn from the independent NMR-study by He et al. (2009) where authors show that the same

70 subset of SAP30 Zinc finger resonances that are perturbed upon PI binding are also affected by DNA binding. Furthermore, our preliminary in vivo data suggest that displacement of DNA by PIs leads to reduced repression of transcription. The interaction of proteins of the SAP30 family with DNA and PI is presumably based on the positions and the interdistances of the negatively charged phosphate groups in the sugar and sugar alcohol (polyol) rings in the DNA and PI, respectively (Figure 14).

Figure 14. The position of phosphate groups presumably dictates the binding specificity of the SAP30 proteins to DNA and monophosphoinositides. Segments of lines emphasize the distance resemblance between negatively charged phosphate groups in the DNA and PtdIns5P molecules.

The sugar moiety in the DNA (deoxyribonucleic acid) is composed of pentose sugar (deoxyribose), where phosphate groups are coordinated between the consecutive sugar rings by their fifth and third carbon atoms. The sugar alcohol moiety in PI (i.e. glycerophosphoinositol monophosphates) is composed of a myoinositol ring

71 inhabited by one invariable phosphate group between the diacylglycerol and myoinositol and one variable phosphate group in either the third, fourth or fifth carbon. Comparison of the molecular geometry of DNA and PI reveals that in the PtdIns5P molecule, to which the SAP30 proteins have the highest affinity, the distance between negatively charged phosphate groups resembles that in the DNA molecule (Figure 14). This may explain the antagonizing interrelationship between DNA and PtdIns5P molecules in respect of their binding to proteins of the SAP30 family. Our data also suggest that DNA and PtdIns5P maintain an antagonizing interrelationship in live cells. In hypoxic conditions the nuclear PtdIns5P content increases (Jones et al. 2006), leading to relocalization of SAP30L to the cytoplasm and reduced repressive activity. Because these in vivo phenotypes were identical with DNA binding-deficient mutant (SAP30L-KAAAK), the most plausible conclusion is that PtdIns5P displaces DNA from SAP30L also in vivo. Phosphatidylinositol monophosphates or monophosphoinositides (PtdIns3P, PtdIns4P and PtdIns5P) were initially considered to be only intermediate metabolites for polyphosphoinositides, but they are now regarded as important signaling molecules as well (Hammond et al. 2004). In addition, the amount of certain phosphoinositides in mammalian cells can be stimulated by physiological ligands or by cellular stresses (Table 2) (Lemmon 2008), and this also applies to the nuclear phosphoinositides (Jones et al. 2006). Moreover, PtdIns5P in the nucleus of murine erythroleukemia cells has been found to increase 20-fold during the G1 phase, bespeaking a potential role for PtdIns5P in cell-cycle progression (Clarke et al. 2001). Components of the phosphatidylinositol signaling pathways colocalize with components of the mRNA-processing machinery in nuclear speckles (Boronenkov et al. 1998). Capitani et al. (1986) showed that addition of phospholipids to purified nuclei could affect in vitro transcription and replication of DNA. Furthermore in vitro-added negatively charged lipids lead to chromatin decondensation, whereas positively charged lipids have the opposite effect (Kuvichkin 2002). This fits well with the data presented in this thesis to the effect that PtdIns5P as a negatively charged phospholipid abolishes negative regulators, SAP30 proteins, from the chromatin, and as a consequence, transcription is increased presumably due to the increased acetylation status and hence decondensed chromatin on the promoter. Involvement of PIs and their derivates the inositol phosphates (IPs) in the function of chromatin-modifying complexes such as SAP30-Sin3A-HDAC is not unprecedented. A yeast genetic screen, in an attempt to identify PHO5 gene transcriptional regulators, surprisingly identified nuclear inositol polyphosphate kinase (IPK2) as one. In IPK2 mutant strains, remodeling of the PHO5 promoter chromatin was impaired, and the ATP-dependent chromatin-remodeling complexes SWI/SNF and INO80 were not efficiently recruited to the PHO5 promoters (Steger et al. 2003). An independent study confirmed these data and found that IPs modulate several classes of chromatin remodeling complexes (NURF, ISW2, INO80 and SWI/SNF) in eukaryotes in vivo and in vitro (Shen et al. 2003). The same studies further showed that IPs inhibited the nucleosome mobilization which was due to the inhibition of ATPase activity of the NURF, ISW2 and INO80 complexes. On the

72 other hand, IPs were able to stimulate nucleosome mobilization by the SWI/SNF complex, the other ATP-dependent nucleosome remodeler. Modulation of transcriptional regulation is not restricted to nuclear IPs but extends to nuclear PtdIns. Another SWI/SNF-like chromatin remodeling complex, BAF, is targeted to chromatin and the nuclear matrix specifically by a PtdIns(4,5)P2-dependent mechanism upon lymphocyte activation (Zhao et al. 1998). A further example is the Sin3A-binding tumor suppressor, ING2, which binds PtdIns3P, PtdIns4P and PtdIns5P. In response to cellular stress by UV irradiation or hydrogen peroxide, ING2 associates with chromatin through a PtdIns5P-mediated mechanism (Gozani et al. 2003). This study opens of a new prospect for the mechanistic understanding of the way PIs accomplish a plethora of chromatin-related functions such as DNA repair, transcription regulation and RNA dynamics (Hammond et al. 2004). For the first time, a simple antagonizing interrelationship between the monophosphoinositides and DNA, in regard to protein binding, has been described and protected by patent application (Viiri et al. 2008). It is tempting to speculate that this is a universal mechanism whereby DNA-binding zinc fingers are regulated. Further in vivo studies are needed to corroborate our findings. Such studies would potentially include chromatin immunoprecipitations of SAP30 proteins from their target promoter, after RNAi silencing of components of the PtdIns5P biosynthesis machinery .

4. Inhibition of disease-associated HDAC complexes

In endeavors to design new drugs for cancer, shut down of the whole HDAC machinery causes a “broad-effect” problem as discussed in section 3.3.3. “HDAC inhibitors as drugs”. One should also consider further that HDACs are usually cofactored by other proteins which are also implicated in certain diseases as discussed in section 4.3. “SAP30/HDAC complexes in diseases”. Perhaps further efforts should focus more on the design of drugs to inhibit HDAC subcomplexes, where the deacetylation function is directed to specific targets by protein cofactors such as SAP30 and ING proteins. Inhibition of protein cofactors together with traditional HDAC inhibitors would ideally potentiate the effect of the latter. In order to attain such 2nd generation function-specific HDAC inhibitors, it is important to understand the molecular function of these protein cofactors. Since it is shown in this work, for example, that the function of the SAP30 cofactors is to stabilize the Sin3-HDAC complexes to chromatin by contacts with DNA, nucleosome and nuclear matrix, it is easy to conceive of the destabilization by specific drugs for therapeutic purposes. In fact, such a destabilizing agent may be endogenous, as this work suggests that inducible nuclear PIs destabilize the SAP30-Sin3-HDAC complex and attenuate its repressive ability.

73 CONCLUSIONS AND FUTURE PROSPECTS

In this work, a novel transcriptional corepressor Sin3A associated protein 30 like (SAP30L) was identified in differentiating T84 epithelial cells (I). SAP30 and SAP30L together constitute a well-conserved SAP30 protein family in which SAP30L is the ancestral protein. Phylogenetic analysis and biochemical experiments suggest that SAP30 has diverged functionally from the ancestral SAP30L by accumulating mutations which have caused attenuation of one of the original functions, i.e. association with the nuclear matrix (IV). SAP30L interacts with several components of the Sin3A corepressor complex and induces transcriptional repression via recruitment of Sin3A and histone deacetylases (II). Since the function of the SAP30 proteins was unknown, the functional domain and motif structure of SAP30 family members were investigated (III) (Figure 13). We found that SAP30 proteins have sequence-independent contact with DNA by their N-terminal zinc-dependent module. The acidic central region contributed to histone and nucleosome interactions and the C-terminal region was responsible for the interaction with Sin3A and nuclear matrix targeting of proteins of the SAP30 family. The domain structure indicates that SAP30 family proteins are intimately involved in Sin3-dependent regulation of gene expression. Furthermore, various contact surfaces of the SAP30 proteins described above suggest for them a nuclear scaffolding function in assembling the functional Sin3 repressome in target promoters. Stabilization provided by the SAP30 proteins is consequently evinced by stronger repressive activity as it has been shown that SAP30 can enhance YY1- mediated repression (Huang et al. 2003). We propose a model in which SAP30L/SAP30 are actively involved in the multiple protein-protein and protein- DNA interactions which modulate transcriptional repression. We suggest that the DNA-binding activity plays a role in anchoring the Sin3A complex to nucleosomal and/or linker DNA in chromatin, and that this binding is further strenghtened by interaction with core histone 2A/2B dimer. One consequence of DNA binding is bending of the DNA, and we envisage that this leads to enhanced accessibility of nucleosomes and histone tails to deacetylating enzymes (Figure 15).

74 Figure 15. The proposed model. 1) When the histones are acetylated, the DNA is loosely packed and therefore accessible to RNA polymerase II. 2) A sequence-specific transcriptional repressor (TF) recruits the Sin3A complex to its target promoter (green bar). SAP30 or SAP30L (SAP30 f, SAP30 family proteins) stabilizes the complex through interactions with DNA, histones 2A/2B, Sin3A and nuclear matrix (red bars). The interaction of SAP30/SAP30L with DNA induces bending of the DNA, as a result of which the nucleosomes are more accessible to HDAC enzymes, and the repressome is fully formed. 3) Nuclear PtdInsPs interact with the N-terminal domain of SAP30/SAP30L, displacing the DNA, which leads to relocalization of SAP30/SAP30L to the cytoplasm and derepression of the promoter. Blue lattice represents the nuclear matrix.

Besides shedding light on the function of the SAP30 family, this work also suggests a novel role for the nuclear signaling lipids. We found that DNA binding of SAP30 proteins is regulated by nuclear phosphoinositides (PI). Namely, DNA and PI seem to stand in a mutually antagonizing interrelationship in regard to their interaction with SAP30 proteins. Interaction of SAP30 proteins with nuclear PIs leads to transcriptional derepression and relocalization of SAP30 proteins to the cytoplasm (Figure 15).

75 Further efforts should be made to clarify whether this antagonizing interrelationship between PIs and DNA is a general theme on DNA binding zinc finger proteins. If it appears to be the case, destabilization of zinc finger DNA interactions by elevating nuclear PI concentration could be a new potential strategy for combinatorial therapies with HDAC inhibitors such as vorinostat. The reason why the PI metabolic machinery has not been caught earlier in yeast genetic screens, designed to identify factors regulating transcription, may simply be because the PI metabolism is different in yeast compared to humans. In fact, PtdIns5P species, for which SAP30 proteins have the highest affinity, is lacking in S. cerevisae (Pettitt et al. 2006). Congruently, several enzymes required for PtdIns5P interconversion are reported to be missing in the fungi in whole Ascomycota phylum (Lecompte et al. 2008). Thus, further endeavours should be made to study the antagonizing interrelationship between PtdIns5P and DNA in conditions where the PI metabolic machinery is fully evolved, e.g. in mammalian cells. Strong evidence suggests that SAP30 proteins are implicated in viral transmission through a YY1-mediated mechanism (Le May et al. 2008). Indirect evidence also indicates that SAP30 proteins are probably involved in certain types of cancers (Campos et al. 2004, Sironi et al. 2004, Cetin et al. 2008, Salgado et al. 2008). It has been unambiguously demonstrated that SAP30 is involved in cell growth control by interacting with certain ING proteins which have been shown to inhibit cell growth in a manner dependent on their interaction with SAP30 protein (Skowyra et al. 2001, Kuzmichev et al. 2002). This thesis suggests that the function of SAP30 proteins is to stabilize the Sin3-HDAC complex by multiple interactions to achieve maximal transcriptional repression. It is therefore likely that the implication of SAP30 in both, cofactoring of viral transmission and inhibition of cell growth, comes about by their ability to stabilize the Sin3 complex in target promoters. The domain structure and mode of action of the SAP30 proteins described here will potentially be of assistance in designing drugs for diseases wherever the Sin3 complex is implicated and medical intervention is needed.

76 ACKNOWLEDGEMENTS

This study was carried out at the Paediatric Research Centre, Department of Pediatrics, Medical School, University of Tampere during years 2003-2009. The research was made possible by financial support from the Academy of Finland Research Council for Health (funding decision numbers 201361 and 115260), the Foundation for Paediatric Research in Finland, the Medical Research Memorial Foundation, the Competitive Research Funding of the Pirkanmaa Hospital District (EVO), the Nona and Kullero Väre Foundation, the Päivikki and Sakari Sohlberg Foundation, the Finnish Coeliac Society and the Tampere Graduate School in Biomedicine and Biotechnology (TGSBB). First, I want to thank my supervisors Professor Markku Mäki and Olli Lohi, MD, PhD for their guidance and support through this thesis project. Markku has been an excellent supervisor and I appreciate his commitment to this project, although it moonlighted beyond the crypt-villus axis as we did not get back to the gut within the frames of this thesis. He has also taught me the valuable lesson to “not spoil your research by publishing, instead publish it twice”. Thanks to Olli’s professional and firm supervision and particular ability to make things “unbearably easy”, truly big part of this work became possible. When you joined the project things started to accelerate. Also, regarding to our shared off-duty interest, thanks for beating me by 8.7 seconds in half-marathon! That will keep me in motion for decades to come. The reviewers of my thesis, Professor Lea Sistonen and Docent Sami Väisänen are warmly acknowledged for their constructive criticism and comments which truly improved this thesis. Professors Heikki Kainulainen and Pärt Peterson are thanked for participating in my thesis committee and also for hands-on guidance at the beginning of this project. I wish to express my gratitude to all my coauthors and collaborators for all their contributions regarding the original publications of this thesis: Hanna Korkeamäki, MSc, Laura Nieminen, MSc, Mari Kukkonen, MSc, Marjo Niittynen, MSc and Katri Lindfors, PhD in our lab in Tampere; Docent Janne Jänis, PhD, Jarkko Valjakka, PhD in Joensuu for performing mass spectrometry part of the work; Trevor Siggers, PhD and Associate Professor Martha Bulyk, PhD, in Boston for providing the protein binding microarray data. I’m also thankful to coauthor and roommate Taisto Heinonen, PhD, for duty- and sometimes off-duty-related lively discussions. You also provided important hold-your-horses-type of mentoring and in-house-reviewing what comes to language and scientific content of the material produced by the room 105. In addition, I thank Robert McGilleon, MA, for revising the language of this thesis. I owe sincere thanks to past and current members in Markku’s Coeliac Disease Study Group and Olli’s Hemato-oncology Research Group. It was really rewarding

77 to have this chance to work in both harbors where also “kelitädit” from Service laboratory provided some extra spice. I’m also grateful to Jokke, who has cultivated for me an amount of cells which probably approach a number of cells in one human body. Ernesto Zanotto is thanked for a great company during the summerschools…after PhD I’m afraid you have to accelerate again! Friends constitute one edge in the triangle of life together with family and work. Tornio-gubbarna, core staff being Mika, Janne, Tommi N. and Moku, more or less annually reunited, has been an important possibility for me to stretch the scope of life. I’m also grateful for the membership in fishing squad which is usually lined-up members such as Arttu, Petteri, Jukka, Sampo, Mikko, Tomi and Matti N. Those annual (and sometimes semiannual) fishing trips have always been very mind- ventilating for me. Rakkaat kiitokset tuesta osoitan siskolleni Hennalle ja vanhemmilleni Ullalle ja Tuomakselle Tornioon. Vanhemmilleni olen ikuisesti kiitollinen siitä henkisestä ja taloudellisesta tuesta mitä olen teiltä elämääni ja opintoihini saanut. Olen myös kiitollinen siitä, että piditte minut kurissaJ ja aina jaksoitte painottaa opiskelun tärkeyttä. Se tehtävä ei ole aivan helppo teinivuosina jolloin nuori elää vain tässä ja nyt. Siinäpä sitä on samaa sarkaa itselleni omien poikiemme kanssa. Appivanhempiani Ritvaa ja Taunoa kiitän ennen kaikkea Pyryn ja Paavon hoitoavusta, mikä on osaltaan edistänyt Leenan ja minun väitöskirjojen syntymistä. Arvostan myös suuresti sitä perheemme arkielämän henkistä ja taloudellista tukea mitä esim. mökkeily pitopalvelun kera jos joku nimenomaan on. Finally, I want to express my blessed gratitude to my dear family, Leena, Pyry and Paavo. Family has taught me the dimensions of life and without you all this would be nothing but a bare frame. To you my beloved wife Leena I owe my deepest gratitude for your constant support, love and friendship. Without your care of the household and looking after our vivid boys this thesis project would not have been possible and incredibly, you made your own thesis simultaneously. I feel so fortunate and impatient for entering our next adventure with you and our boys by myside! Well and truly...nothing else matters.

Tampere, March 27th 2009

Keijo Viiri

78 REFERENCES

Alberts BJ, A. Lewis, J. Raff, M. Roberts, K. Walter, P (2002): Molecular Biology Of The Cell. In: pp Ed. Garland Science, Alkema MJ, Bronk M, Verhoeven E, Otte A, van 't Veer LJ, Berns A and van Lohuizen M (1997): Identification of Bmi1-interacting proteins as constituents of a multimeric mammalian polycomb complex. Genes Dev 11: 226-40. Alland L, Muhle R, Hou H, Jr., Potes J, Chin L, Schreiber-Agus N and DePinho RA (1997): Role for N-CoR and histone deacetylase in Sin3-mediated transcriptional repression. Nature 387: 49-55. Allfrey VG, Faulkner R and Mirsky AE (1964): Acetylation and Methylation of Histones and Their Possible Role in the Regulation of Rna Synthesis. Proc Natl Acad Sci U S A 51: 786-94. Allshire RC, Javerzat JP, Redhead NJ and Cranston G (1994): Position effect variegation at fission yeast centromeres. Cell 76: 157-69. Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W and Lipman DJ (1997): Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25: 3389-402. Aravind L, Anantharaman V, Balaji S, Babu MM and Iyer LM (2005): The many faces of the helix-turn-helix domain: transcription regulation and beyond. FEMS Microbiol Rev 29: 231-62. Askree SH, Yehuda T, Smolikov S, Gurevich R, Hawk J, Coker C, Krauskopf A, Kupiec M and McEachern MJ (2004): A genome-wide screen for Saccharomyces cerevisiae deletion mutants that affect telomere length. Proc Natl Acad Sci U S A 101: 8658-63. Ayer DE, Lawrence QA and Eisenman RN (1995): Mad-Max transcriptional repression is mediated by ternary complex formation with mammalian homologs of yeast repressor Sin3. Cell 80: 767-76. Bailey TL, Williams N, Misleh C and Li WW (2006): MEME: discovering and analyzing DNA and protein sequence motifs. Nucleic Acids Res 34: W369- 73. Bannister AJ, Zegerman P, Partridge JF, Miska EA, Thomas JO, Allshire RC and Kouzarides T (2001): Selective recognition of methylated lysine 9 on histone H3 by the HP1 chromo domain. Nature 410: 120-4. Baylin SB and Ohm JE (2006): Epigenetic gene silencing in cancer - a mechanism for early oncogenic pathway addiction? Nat Rev Cancer 6: 107-16. Beck JS (1961): Variations in the morphological patterns of "autoimmune" nuclear fluorescence. Lancet 1: 1203-5. Belikov S and Karpov V (1998): Linker histones: paradigm lost but questions remain. FEBS Lett 441: 161-4. Benyajati C and Worcel A (1976): Isolation, characterization, and structure of the folded interphase genome of Drosophila melanogaster. Cell 9: 393-407. Berezney R and Coffey DS (1974): Identification of a nuclear protein matrix. Biochem Biophys Res Commun 60: 1410-7. Berger MF, Philippakis AA, Qureshi AM, He FS, Estep PW, 3rd and Bulyk ML (2006): Compact, universal DNA microarrays to comprehensively determine transcription-factor binding site specificities. Nat Biotechnol 24: 1429-35.

79 Berger SL (2007): The complex language of chromatin regulation during transcription. Nature 447: 407-12. Bernardi R, Scaglioni PP, Bergmann S, Horn HF, Vousden KH and Pandolfi PP (2004): PML regulates p53 stability by sequestering Mdm2 to the nucleolus. Nat Cell Biol 6: 665-72. Bernstein BE, Tong JK and Schreiber SL (2000): Genomewide studies of histone deacetylase function in yeast. Proc Natl Acad Sci U S A 97: 13708-13. Bernues J, Espel E and Querol E (1986): Identification of the core-histone-binding domains of HMG1 and HMG2. Biochim Biophys Acta 866: 242-51. Beutler E, Yeh M and Fairbanks VF (1962): The normal human female as a mosaic of X-chromosome activity: studies using the gene for C-6-PD-deficiency as a marker. Proc Natl Acad Sci U S A 48: 9-16. Bird A (2007): Perceptions of epigenetics. Nature 447: 396-8. Bird AW, Yu DY, Pray-Grant MG, Qiu Q, Harmon KE, Megee PC, Grant PA, Smith MM and Christman MF (2002): Acetylation of histone H4 by Esa1 is required for DNA double-strand break repair. Nature 419: 411-5. Boisvert FM, van Koningsbruggen S, Navascues J and Lamond AI (2007): The multifunctional nucleolus. Nat Rev Mol Cell Biol 8: 574-85. Bolden JE, Peart MJ and Johnstone RW (2006): Anticancer activities of histone deacetylase inhibitors. Nat Rev Drug Discov 5: 769-84. Bolzer A, Kreth G, Solovei I, Koehler D, Saracoglu K, Fauth C, Muller S, Eils R, Cremer C, Speicher MR and Cremer T (2005): Three-dimensional maps of all chromosomes in human male fibroblast nuclei and prometaphase rosettes. PLoS Biol 3: e157. Borden KL (2002): Pondering the promyelocytic leukemia protein (PML) puzzle: possible functions for PML nuclear bodies. Mol Cell Biol 22: 5259-69. Boronenkov IV, Loijens JC, Umeda M and Anderson RA (1998): Phosphoinositide signaling pathways in nuclei are associated with nuclear speckles containing pre-mRNA processing factors. Mol Biol Cell 9: 3547-60. Borra MT, Langer MR, Slama JT and Denu JM (2004): Substrate specificity and kinetic mechanism of the Sir2 family of NAD+-dependent histone/protein deacetylases. Biochemistry 43: 9877-87. Boulikas T (1995): Chromatin domains and prediction of MAR sequences. Int Rev Cytol 162A: 279-388. Bowdish KS and Mitchell AP (1993): Bipartite structure of an early meiotic upstream activation sequence from Saccharomyces cerevisiae. Mol Cell Biol 13: 2172-81. Branden C and Tooze J (1999): Introduction to protein structure. In: pp Ed. Garland Science Publishing, NY, USA., Brown RS (2005): Zinc finger proteins: getting a grip on RNA. Curr Opin Struct Biol 15: 94-8. Brownell JE, Zhou J, Ranalli T, Kobayashi R, Edmondson DG, Roth SY and Allis CD (1996): Tetrahymena histone acetyltransferase A: a homolog to yeast Gcn5p linking histone acetylation to gene activation. Cell 84: 843-51. Buratowski S (2003): The CTD code. Nat Struct Biol 10: 679-80. Busch H, Byvoet P and Smetana K (1963): The nucleolus of the cancer cell: a review. Cancer Res 23: 313-39. Campos EI, Martinka M, Mitchell DL, Dai DL 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. Capco DG, Wan KM and Penman S (1982): The nuclear matrix: three-dimensional architecture and protein composition. Cell 29: 847-58. Capitani S, Cocco L, Maraldi NM, Papa S and Manzoli FA (1986): Effect of phospholipids on transcription and ribonucleoprotein processing in isolated nuclei. Adv Enzyme Regul 25: 425-38. Carballo E, Lai WS and Blackshear PJ (1998): Feedback inhibition of macrophage tumor necrosis factor-alpha production by tristetraprolin. Science 281: 1001- 5.

80 Carballo M, Puigdomenech P and Palau J (1983): DNA and histone H1 interact with different domains of HMG 1 and 2 proteins. Embo J 2: 1759-64. Cetin E, Cengiz B, Gunduz E, Gunduz M, Nagatsuka H, Bekir-Beder L, Fukushima K, Pehlivan D, N MO, Nishizaki K, Shimizu K and Nagai N (2008): Deletion mapping of chromosome 4q22-35 and identification of four frequently deleted regions in head and neck cancers. Neoplasma 55: 299- 304. Chen T, Boisvert FM, Bazett-Jones DP and Richard S (1999): A role for the GSG domain in localizing Sam68 to novel nuclear structures in cancer cell lines. Mol Biol Cell 10: 3015-33. Chuang CH, Carpenter AE, Fuchsova B, Johnson T, de Lanerolle P and Belmont AS (2006): Long-range directional movement of an interphase chromosome site. Curr Biol 16: 825-31. Claessens F and Gewirth DT (2004): DNA recognition by nuclear receptors. Essays Biochem 40: 59-72. Clarke JH, Letcher AJ, D'Santos C S, Halstead JR, Irvine RF and Divecha N (2001): Inositol lipids are regulated during cell cycle progression in the nuclei of murine erythroleukaemia cells. Biochem J 357: 905-10. Clouaire T, Roussigne M, Ecochard V, Mathe C, Amalric F and Girard JP (2005): The THAP domain of THAP1 is a large C2CH module with zinc-dependent sequence-specific DNA-binding activity. Proc Natl Acad Sci U S A 102: 6907-12. Cocco L, Maraldi NM, Manzoli FA, Gilmour RS and Lang A (1980): Phospholipid interactions in rat liver nuclear matrix. Biochem Biophys Res Commun 96: 890-8. Cocco L, Gilmour RS, Ognibene A, Letcher AJ, Manzoli FA and Irvine RF (1987): Synthesis of polyphosphoinositides in nuclei of Friend cells. Evidence for polyphosphoinositide metabolism inside the nucleus which changes with cell differentiation. Biochem J 248: 765-70. Cote J, Quinn J, Workman JL and Peterson CL (1994): Stimulation of GAL4 derivative binding to nucleosomal DNA by the yeast SWI/SNF complex. Science 265: 53-60. Cote S, Rosenauer A, Bianchini A, Seiter K, Vandewiele J, Nervi C and Miller WH, Jr. (2002): Response to histone deacetylase inhibition of novel PML/RARalpha mutants detected in retinoic acid-resistant APL cells. Blood 100: 2586-96. Craig JM (2005): Heterochromatin--many flavours, common themes. Bioessays 27: 17-28. De Nadal E, Zapater M, Alepuz PM, Sumoy L, Mas G and Posas F (2004): The MAPK Hog1 recruits Rpd3 histone deacetylase to activate osmoresponsive genes. Nature 427: 370-4. De Rubertis F, Kadosh D, Henchoz S, Pauli D, Reuter G, Struhl K and Spierer P (1996): The histone deacetylase RPD3 counteracts genomic silencing in Drosophila and yeast. Nature 384: 589-91. Deckert J and Struhl K (2001): Histone acetylation at promoters is differentially affected by specific activators and repressors. Mol Cell Biol 21: 2726-35. Deppmann CD, Alvania RS and Taparowsky EJ (2006): Cross-species annotation of basic leucine zipper factor interactions: Insight into the evolution of closed interaction networks. Mol Biol Evol 23: 1480-92. Derenzini M, Trere D, Pession A, Govoni M, Sirri V and Chieco P (2000): Nucleolar size indicates the rapidity of cell proliferation in cancer tissues. J Pathol 191: 181-6. Dhalluin C, Carlson JE, Zeng L, He C, Aggarwal AK and Zhou MM (1999): Structure and ligand of a histone acetyltransferase bromodomain. Nature 399: 491-6. Di Paolo G and De Camilli P (2006): Phosphoinositides in cell regulation and membrane dynamics. Nature 443: 651-7.

81 Dingwall C and Laskey RA (1998): Nuclear import: a tale of two sites. Curr Biol 8: R922-4. Dinman JD, Ruiz-Echevarria MJ and Peltz SW (1998): Translating old drugs into new treatments: ribosomal frameshifting as a target for antiviral agents. Trends in Biotechnology 16: 190-196. Dora EG, Rudin N, Martell JR, Esposito MS and Ramirez RM (1999): RPD3 (REC3) mutations affect mitotic recombination in Saccharomyces cerevisiae. Curr Genet 35: 68-76. Dorland S, Deegenaars ML and Stillman DJ (2000): Roles for the Saccharomyces cerevisiae SDS3, CBK1 and HYM1 genes in transcriptional repression by SIN3. Genetics 154: 573-86. Durrin LK, Mann RK, Kayne PS and Grunstein M (1991): Yeast histone H4 N- terminal sequence is required for promoter activation in vivo. Cell 65: 1023- 31. Ekwall K, Olsson T, Turner BM, Cranston G and Allshire RC (1997): Transient inhibition of histone deacetylation alters the structural and functional imprint at fission yeast centromeres. Cell 91: 1021-32. Ekwall K (2005): Genome-wide analysis of HDAC function. Trends Genet 21: 608- 15. Eot-Houllier G, Fulcrand G, Magnaghi-Jaulin L and Jaulin C (2009): Histone deacetylase inhibitors and genomic instability. Cancer Lett 274: 169-76. Esposito MS and Brown JT (1990): Conditional hyporecombination mutants of three REC genes of Saccharomyces cerevisiae. Curr Genet 17: 7-12. Fabris D, Hathout Y and Fenselau C (1999): Investigation of Zinc Chelation in Zinc-Finger Arrays by Electrospray Mass Spectrometry. Inorg Chem 38: 1322-1325. Fakan S (1986): Structural support for RNA synthesis in the cell nucleus. Methods Achiev Exp Pathol 12: 105-40. Fearon ER and Vogelstein B (1990): A genetic model for colorectal tumorigenesis. Cell 61: 759-67. Felsenstein J (1989): PHYLIP (phylogeny inference package). Version 3.2. Cladistics 5: 164-166. Fleischer TC, Yun UJ and Ayer DE (2003): Identification and characterization of three new components of the mSin3A corepressor complex. Mol Cell Biol 23: 3456-67. Foisner R (2001): Inner nuclear membrane proteins and the nuclear lamina. J Cell Sci 114: 3791-2. Fox AH, Lam YW, Leung AK, Lyon CE, Andersen J, Mann M and Lamond AI (2002): Paraspeckles: a novel nuclear domain. Curr Biol 12: 13-25. Fraschini A, Albi E, Gahan PB and Viola-Magni MP (1992): TEM cytochemical study of the localization of phospholipids in interphase chromatin in rat hepatocytes. Histochemistry 97: 225-35. Fraschini A, Biggiogera M, Bottone MG and Martin TE (1999): Nuclear phospholipids in human lymphocytes activated by phytohemagglutinin. Eur J Cell Biol 78: 416-23. Friedman R and Hughes AL (2003): The temporal distribution of gene duplication events in a set of highly conserved human gene families. Mol Biol Evol 20: 154-61. Garkavtsev I, Kazarov A, Gudkov A and Riabowol K (1996): Suppression of the novel growth inhibitor p33ING1 promotes neoplastic transformation. Nat Genet 14: 415-20. Garkavtsev I, Grigorian IA, Ossovskaya VS, Chernov MV, Chumakov PM and Gudkov AV (1998): The candidate tumour suppressor p33ING1 cooperates with p53 in cell growth control. Nature 391: 295-8. Ghetti A, Pinol-Roma S, Michael WM, Morandi C and Dreyfuss G (1992): hnRNP I, the polypyrimidine tract-binding protein: distinct nuclear localization and association with hnRNAs. Nucleic Acids Res 20: 3671-8.

82 Glazko GV, Koonin EV, Rogozin IB and Shabalina SA (2003): A significant fraction of conserved noncoding DNA in human and mouse consists of predicted matrix attachment regions. Trends Genet 19: 119-24. Goll MG and Bestor TH (2005): Eukaryotic cytosine methyltransferases. Annu Rev Biochem 74: 481-514. Gonzales ML and Anderson RA (2006): Nuclear phosphoinositide kinases and inositol phospholipids. J Cell Biochem 97: 252-60. Gorlich D (1998): Transport into and out of the cell nucleus. Embo J 17: 2721-7. Gozani O, Karuman P, Jones DR, Ivanov D, Cha J, Lugovskoy AA, Baird CL, Zhu H, Field SJ, Lessnick SL, Villasenor J, Mehrotra B, Chen J, Rao VR, Brugge JS, Ferguson CG, Payrastre B, Myszka DG, Cantley LC, Wagner G, Divecha N, Prestwich GD and Yuan J (2003): The PHD finger of the chromatin- associated protein ING2 functions as a nuclear phosphoinositide receptor. Cell 114: 99-111. Grande MA, van der Kraan I, de Jong L and van Driel R (1997): Nuclear distribution of transcription factors in relation to sites of transcription and RNA polymerase II. J Cell Sci 110 ( Pt 15): 1781-91. Grasser M, Christensen JM, Peterhansel C and Grasser KD (2007): Basic and acidic regions flanking the HMG-box domain of maize HMGB1 and HMGB5 modulate the stimulatory effect on the DNA binding of transcription factor Dof2. Biochemistry 46: 6375-82. Green MR (2005): Eukaryotic transcription activation: right on target. Mol Cell 18: 399-402. Grewal SI and Jia S (2007): Heterochromatin revisited. Nat Rev Genet 8: 35-46. Grunstein M (1997): Histone acetylation in chromatin structure and transcription. Nature 389: 349-52. Grunstein M (1998): Yeast heterochromatin: regulation of its assembly and inheritance by histones. Cell 93: 325-8. Gu X (1999): Statistical methods for testing functional divergence after gene duplication. Mol Biol Evol 16: 1664-74. Gu X and Vander Velden K (2002): DIVERGE: phylogeny-based analysis for functional-structural divergence of a protein family. Bioinformatics 18: 500- 1. Gu X (2006): A simple statistical method for estimating type-II (cluster-specific) functional divergence of protein sequences. Mol Biol Evol 23: 1937-45. Halttunen T, Marttinen A, Rantala I, Kainulainen H and Maki M (1996): Fibroblasts and transforming growth factor beta induce organization and differentiation of T84 human epithelial cells. Gastroenterology 111: 1252-62. Hammond G, Thomas CL and Schiavo G (2004): Nuclear phosphoinositides and their functions. Curr Top Microbiol Immunol 282: 177-206. Han M and Grunstein M (1988): Nucleosome loss activates yeast downstream promoters in vivo. Cell 55: 1137-45. Hancock R (2000): A new look at the nuclear matrix. Chromosoma 109: 219-25. Hansen JC, Tse C and Wolffe AP (1998): Structure and function of the core histone N-termini: more than meets the eye. Biochemistry 37: 17637-41. Harris H (2000): The Birth of the Cell. In: pp Ed. Yale University Press, Hassig CA, Fleischer TC, Billin AN, Schreiber SL and Ayer DE (1997): Histone deacetylase activity is required for full transcriptional repression by mSin3A. Cell 89: 341-7. He Y, Imhoff R, Sahu A and Radhakrishnan I (2009): Solution structure of a novel zinc finger motif in the SAP30 polypeptide of the Sin3 corepressor complex and its potential role in nucleic acid recognition. Nucleic Acids Res Hendzel MJ, Boisvert F and Bazett-Jones DP (1999): Direct visualization of a protein nuclear architecture. Mol Biol Cell 10: 2051-62. Hewitt SL, High FA, Reiner SL, Fisher AG and Merkenschlager M (2004): Nuclear repositioning marks the selective exclusion of lineage-inappropriate transcription factor loci during T helper cell differentiation. Eur J Immunol 34: 3604-13.

83 Higgins DG and Sharp PM (1989): Fast and sensitive multiple sequence alignments on a microcomputer. Comput Appl Biosci 5: 151-3. Hock R, Furusawa T, Ueda T and Bustin M (2007): HMG chromosomal proteins in development and disease. Trends Cell Biol 17: 72-9. Hokin LE (1985): Receptors and phosphoinositide-generated second messengers. Annu Rev Biochem 54: 205-35. Hoppe GJ, Tanny JC, Rudner AD, Gerber SA, Danaie S, Gygi SP and Moazed D (2002): Steps in assembly of silent chromatin in yeast: Sir3-independent binding of a Sir2/Sir4 complex to silencers and role for Sir2-dependent deacetylation. Mol Cell Biol 22: 4167-80. Horke S, Reumann K, Schweizer M, Will H and Heise T (2004): Nuclear trafficking of La protein depends on a newly identified nucleolar localization signal and the ability to bind RNA. J Biol Chem 279: 26563-70. Hsieh JJ, Zhou S, Chen L, Young DB and Hayward SD (1999): CIR, a corepressor linking the DNA binding factor CBF1 to the histone deacetylase complex. Proc Natl Acad Sci U S A 96: 23-8. Huang NE, Lin CH, Lin YS and Yu WC (2003): Modulation of YY1 activity by SAP30. Biochem Biophys Res Commun 306: 267-75. Huang S, Deerinck TJ, Ellisman MH and Spector DL (1997): The dynamic organization of the perinucleolar compartment in the cell nucleus. J Cell Biol 137: 965-74. Huang S (2000): Review: perinucleolar structures. J Struct Biol 129: 233-40. Hudak KA, Lopes JM and Henry SA (1994): A pleiotropic phospholipid biosynthetic regulatory mutation in Saccharomyces cerevisiae is allelic to sin3 (sdi1, ume4, rpd1). Genetics 136: 475-83. Hyvarinen AK, Pohjoismaki JL, Reyes A, Wanrooij S, Yasukawa T, Karhunen PJ, Spelbrink JN, Holt IJ and Jacobs HT (2007): The mitochondrial transcription termination factor mTERF modulates replication pausing in human mitochondrial DNA. Nucleic Acids Res 35: 6458-74. Imai S, Armstrong CM, Kaeberlein M and Guarente L (2000): Transcriptional silencing and longevity protein Sir2 is an NAD-dependent histone deacetylase. Nature 403: 795-800. Imai Y, Kurokawa M, Yamaguchi Y, Izutsu K, Nitta E, Mitani K, Satake M, Noda T, Ito Y and Hirai H (2004): The corepressor mSin3A regulates phosphorylation-induced activation, intranuclear location, and stability of AML1. Mol Cell Biol 24: 1033-43. Imbalzano AN, Kwon H, Green MR and Kingston RE (1994): Facilitated binding of TATA-binding protein to nucleosomal DNA. Nature 370: 481-5. Jackson DA, Dickinson P and Cook PR (1990a): Attachment of DNA to the nucleoskeleton of HeLa cells examined using physiological conditions. Nucleic Acids Res 18: 4385-93. Jackson DA, Dickinson P and Cook PR (1990b): The size of chromatin loops in HeLa cells. Embo J 9: 567-71. Javed A, Barnes GL, Pratap J, Antkowiak T, Gerstenfeld LC, van Wijnen AJ, Stein JL, Lian JB and Stein GS (2005): Impaired intranuclear trafficking of Runx2 (AML3/CBFA1) transcription factors in breast cancer cells inhibits osteolysis in vivo. Proc Natl Acad Sci U S A 102: 1454-9. Jenuwein T and Allis CD (2001): Translating the histone code. Science 293: 1074- 80. Johansen KM and Johansen J (2006): Regulation of chromatin structure by histone H3S10 phosphorylation. Chromosome Res 14: 393-404. Johnson BR, Nitta RT, Frock RL, Mounkes L, Barbie DA, Stewart CL, Harlow E and Kennedy BK (2004): A-type lamins regulate retinoblastoma protein function by promoting subnuclear localization and preventing proteasomal degradation. Proc Natl Acad Sci U S A 101: 9677-82. Jones DR, Bultsma Y, Keune WJ, Halstead JR, Elouarrat D, Mohammed S, Heck AJ, D'Santos CS and Divecha N (2006): Nuclear PtdIns5P as a transducer of stress signaling: an in vivo role for PIP4Kbeta. Mol Cell 23: 685-95.

84 Jones PL, Veenstra GJ, Wade PA, Vermaak D, Kass SU, Landsberger N, Strouboulis J and Wolffe AP (1998): Methylated DNA and MeCP2 recruit histone deacetylase to repress transcription. Nat Genet 19: 187-91. Kaadige MR and Ayer DE (2006): The polybasic region that follows the plant homeodomain zinc finger 1 of Pf1 is necessary and sufficient for specific phosphoinositide binding. J Biol Chem 281: 28831-6. Kadosh D and Struhl K (1997): Repression by Ume6 involves recruitment of a complex containing Sin3 corepressor and Rpd3 histone deacetylase to target promoters. Cell 89: 365-71. Kadosh D and Struhl K (1998): Targeted recruitment of the Sin3-Rpd3 histone deacetylase complex generates a highly localized domain of repressed chromatin in vivo. Mol Cell Biol 18: 5121-7. Karolchik D, Kuhn RM, Baertsch R, Barber GP, Clawson H, Diekhans M, Giardine B, Harte RA, Hinrichs AS, Hsu F, Kober KM, Miller W, Pedersen JS, Pohl A, Raney BJ, Rhead B, Rosenbloom KR, Smith KE, Stanke M, Thakkapallayil A, Trumbower H, Wang T, Zweig AS, Haussler D and Kent WJ (2008): The UCSC Genome Browser Database: 2008 update. Nucleic Acids Res 36: D773-9. Kayne PS, Kim UJ, Han M, Mullen JR, Yoshizaki F and Grunstein M (1988): Extremely conserved histone H4 N terminus is dispensable for growth but essential for repressing the silent mating loci in yeast. Cell 55: 27-39. Kim SH, McQueen PG, Lichtman MK, Shevach EM, Parada LA and Misteli T (2004): Spatial genome organization during T-cell differentiation. Cytogenet Genome Res 105: 292-301. Kimura A, Umehara T and Horikoshi M (2002): Chromosomal gradient of histone acetylation established by Sas2p and Sir2p functions as a shield against gene silencing. Nat Genet 32: 370-7. Kleff S, Andrulis ED, Anderson CW and Sternglanz R (1995): Identification of a gene encoding a yeast histone H4 acetyltransferase. J Biol Chem 270: 24674-7. Klibanov SA, O'Hagan HM and Ljungman M (2001): Accumulation of soluble and nucleolar-associated p53 proteins following cellular stress. J Cell Sci 114: 1867-73. Klose RJ and Bird AP (2006): Genomic DNA methylation: the mark and its mediators. Trends Biochem Sci 31: 89-97. Klug A (1999): Zinc finger peptides for the regulation of gene expression. J Mol Biol 293: 215-8. Knezetic JA and Luse DS (1986): The presence of nucleosomes on a DNA template prevents initiation by RNA polymerase II in vitro. Cell 45: 95-104. Korkeamaki H, Viiri K, Kukkonen MK, Maki M and Lohi O (2008): Alternative mRNA splicing of SAP30L regulates its transcriptional repression activity. FEBS Lett 582: 379-84. Kornberg RD (1974): Chromatin structure: a repeating unit of histones and DNA. Science 184: 868-71. Kosak ST, Skok JA, Medina KL, Riblet R, Le Beau MM, Fisher AG and Singh H (2002): Subnuclear compartmentalization of immunoglobulin loci during lymphocyte development. Science 296: 158-62. Krebs JE, Fry CJ, Samuels ML and Peterson CL (2000): Global role for chromatin remodeling enzymes in mitotic gene expression. Cell 102: 587-98. Krithivas A, Young DB, Liao G, Greene D and Hayward SD (2000): Human herpesvirus 8 LANA interacts with proteins of the mSin3 corepressor complex and negatively regulates Epstein-Barr virus gene expression in dually infected PEL cells. J Virol 74: 9637-45. Kuo MH and Allis CD (1998): Roles of histone acetyltransferases and deacetylases in gene regulation. Bioessays 20: 615-26. Kuo MH, vom Baur E, Struhl K and Allis CD (2000): Gcn4 activator targets Gcn5 histone acetyltransferase to specific promoters independently of transcription. Mol Cell 6: 1309-20.

85 Kurdistani SK, Robyr D, Tavazoie S and Grunstein M (2002): Genome-wide binding map of the histone deacetylase Rpd3 in yeast. Nat Genet 31: 248-54. Kurdistani SK and Grunstein M (2003): Histone acetylation and deacetylation in yeast. Nat Rev Mol Cell Biol 4: 276-84. Kurdistani SK, Tavazoie S and Grunstein M (2004): Mapping global histone acetylation patterns to gene expression. Cell 117: 721-33. Kuvichkin VV (2002): DNA-lipid interactions in vitro and in vivo. Bioelectrochemistry 58: 3-12. Kuzmichev A, Zhang Y, Erdjument-Bromage H, Tempst P and Reinberg D (2002): Role of the Sin3-histone deacetylase complex in growth regulation by the candidate tumor suppressor p33(ING1). Mol Cell Biol 22: 835-48. Kwon H, Imbalzano AN, Khavari PA, Kingston RE and Green MR (1994): Nucleosome disruption and enhancement of activator binding by a human SW1/SNF complex. Nature 370: 477-81. La Cour LF, Chayen J and Gahan PS (1958): Evidence for lipid material in chromosomes. Exp Cell Res 14: 469-85. Lachner M, O'Carroll D, Rea S, Mechtler K and Jenuwein T (2001): Methylation of histone H3 lysine 9 creates a binding site for HP1 proteins. Nature 410: 116- 20. Laherty CD, Yang WM, Sun JM, Davie JR, Seto E and Eisenman RN (1997): Histone deacetylases associated with the mSin3 corepressor mediate mad transcriptional repression. Cell 89: 349-56. Laherty CD, Billin AN, Lavinsky RM, Yochum GS, Bush AC, Sun JM, Mullen TM, Davie JR, Rose DW, Glass CK, Rosenfeld MG, Ayer DE and Eisenman RN (1998): SAP30, a component of the mSin3 corepressor complex involved in N-CoR-mediated repression by specific transcription factors. Mol Cell 2: 33- 42. Lai A, Kennedy BK, Barbie DA, Bertos NR, Yang XJ, Theberge MC, Tsai SC, Seto E, Zhang Y, Kuzmichev A, Lane WS, Reinberg D, Harlow E and Branton PE (2001): RBP1 recruits the mSIN3-histone deacetylase complex to the pocket of retinoblastoma tumor suppressor family proteins found in limited discrete regions of the nucleus at growth arrest. Mol Cell Biol 21: 2918-32. Lamond AI and Spector DL (2003): Nuclear speckles: a model for nuclear organelles. Nat Rev Mol Cell Biol 4: 605-12. Lange A, Mills RE, Lange CJ, Stewart M, Devine SE and Corbett AH (2007): Classical nuclear localization signals: definition, function, and interaction with importin alpha. J Biol Chem 282: 5101-5. Lasko D, Cavenee W and Nordenskjold M (1991): Loss of constitutional heterozygosity in human cancer. Annu Rev Genet 25: 281-314. Le May N, Mansuroglu Z, Leger P, Josse T, Blot G, Billecocq A, Flick R, Jacob Y, Bonnefoy E and Bouloy M (2008): A SAP30 complex inhibits IFN-beta expression in Rift Valley fever virus infected cells. PLoS Pathog 4: e13. Lechner T, Carrozza MJ, Yu Y, Grant PA, Eberharter A, Vannier D, Brosch G, Stillman DJ, Shore D and Workman JL (2000): Sds3 (suppressor of defective silencing 3) is an integral component of the yeast Sin3[middle dot]Rpd3 histone deacetylase complex and is required for histone deacetylase activity. J Biol Chem 275: 40961-6. Lecompte O, Poch O and Laporte J (2008): PtdIns5P regulation through evolution: roles in membrane trafficking? Trends Biochem Sci 33: 453-60. Lee MS, Gippert GP, Soman KV, Case DA and Wright PE (1989): Three- dimensional solution structure of a single zinc finger DNA-binding domain. Science 245: 635-7. Leipe DD and Landsman D (1997): Histone deacetylases, acetoin utilization proteins and acetylpolyamine amidohydrolases are members of an ancient protein superfamily. Nucleic Acids Res 25: 3693-7. Lemmon MA (2008): Membrane recognition by phospholipid-binding domains. Nat Rev Mol Cell Biol 9: 99-111.

86 Li B, Carey M and Workman JL (2007): The role of chromatin during transcription. Cell 128: 707-19. Lin RJ, Sternsdorf T, Tini M and Evans RM (2001): Transcriptional regulation in acute promyelocytic leukemia. Oncogene 20: 7204-15. Lindvall JM, Blomberg KE, Valiaho J, Vargas L, Heinonen JE, Berglof A, Mohamed AJ, Nore BF, Vihinen M and Smith CI (2005): Bruton's tyrosine kinase: cell biology, sequence conservation, mutation spectrum, siRNA modifications, and expression profiling. Immunol Rev 203: 200-15. Loewith R, Smith JS, Meijer M, Williams TJ, Bachman N, Boeke JD and Young D (2001): Pho23 is associated with the Rpd3 histone deacetylase and is required for its normal function in regulation of gene expression and silencing in Saccharomyces cerevisiae. J Biol Chem 276: 24068-74. Lorch Y, LaPointe JW and Kornberg RD (1987): Nucleosomes inhibit the initiation of transcription but allow chain elongation with the displacement of histones. Cell 49: 203-10. Luger K, Mader AW, Richmond RK, Sargent DF and Richmond TJ (1997): Crystal structure of the nucleosome core particle at 2.8 A resolution. Nature 389: 251-60. Luo K, Vega-Palas MA and Grunstein M (2002): Rap1-Sir4 binding independent of other Sir, yKu, or histone interactions initiates the assembly of telomeric heterochromatin in yeast. Genes Dev 16: 1528-39. Lyon MF (1961): Gene action in the X-chromosome of the mouse (Mus musculus L.). Nature 190: 372-3. Macfarlan T, Kutney S, Altman B, Montross R, Yu J and Chakravarti D (2005): Human THAP7 is a chromatin-associated, histone tail-binding protein that represses transcription via recruitment of HDAC3 and nuclear hormone receptor corepressor. J Biol Chem 280: 7346-58. Mann BS, Johnson JR, Cohen MH, Justice R and Pazdur R (2007): FDA approval summary: vorinostat for treatment of advanced primary cutaneous T-cell lymphoma. Oncologist 12: 1247-52. Manzoli FA, Cocco L, Facchini A, Casali AM, Maraldi NM and Grossi CE (1976): Phospholipids bound to acidic nuclear proteins in human B and T lymphocytes. Mol Cell Biochem 12: 67-71. Maston GA, Evans SK and Green MR (2006): Transcriptional regulatory elements in the human genome. Annu Rev Genomics Hum Genet 7: 29-59. Matthews AG, Kuo AJ, Ramon-Maiques S, Han S, Champagne KS, Ivanov D, Gallardo M, Carney D, Cheung P, Ciccone DN, Walter KL, Utz PJ, Shi Y, Kutateladze TG, Yang W, Gozani O and Oettinger MA (2007): RAG2 PHD finger couples histone H3 lysine 4 trimethylation with V(D)J recombination. Nature 450: 1106-10. Matthews JM and Sunde M (2002): Zinc fingers--folds for many occasions. IUBMB Life 54: 351-5. McGuinness BE, Hirota T, Kudo NR, Peters JM and Nasmyth K (2005): Shugoshin prevents dissociation of cohesin from centromeres during mitosis in vertebrate cells. PLoS Biol 3: e86. McKenzie EA, Kent NA, Dowell SJ, Moreno F, Bird LE and Mellor J (1993): The centromere and promoter factor, 1, CPF1, of Saccharomyces cerevisiae modulates gene activity through a family of factors including SPT21, RPD1 (SIN3), RPD3 and CCR4. Mol Gen Genet 240: 374-86. Meaburn KJ and Misteli T (2007): Cell biology: chromosome territories. Nature 445: 379-781. Mendez J and Stillman B (2000): Chromatin association of human origin recognition complex, cdc6, and minichromosome maintenance proteins during the cell cycle: assembly of prereplication complexes in late mitosis. Mol Cell Biol 20: 8602-12. Meskauskas A, Baxter JL, Carr EA, Yasenchak J, Gallagher JE, Baserga SJ and Dinman JD (2003): Delayed rRNA processing results in significant ribosome biogenesis and functional defects. Mol Cell Biol 23: 1602-13.

87 Michalowski SM, Allen GC, Hall GE, Jr., Thompson WF and Spiker S (1999): Characterization of randomly-obtained matrix attachment regions (MARs) from higher plants. Biochemistry 38: 12795-804. Mika S and Rost B (2005): NMPdb: Database of Nuclear Matrix Proteins. Nucleic Acids Res 33: D160-3. Miller J, McLachlan AD and Klug A (1985): Repetitive zinc-binding domains in the protein transcription factor IIIA from Xenopus oocytes. Embo J 4: 1609-14. Mirkovitch J, Spierer P and Laemmli UK (1986): Genes and loops in 320,000 base- pairs of the Drosophila melanogaster chromosome. J Mol Biol 190: 255-8. Misteli T, Caceres JF and Spector DL (1997): The dynamics of a pre-mRNA splicing factor in living cells. Nature 387: 523-7. Morris JB, Hinchliffe KA, Ciruela A, Letcher AJ and Irvine RF (2000): Thrombin stimulation of platelets causes an increase in phosphatidylinositol 5- phosphate revealed by mass assay. FEBS Lett 475: 57-60. Muller HJ and Gershenson SM (1935): Inert Regions of Chromosomes as the Temporary Products of Individual Genes. Proc Natl Acad Sci U S A 21: 69- 75. Nakamura T, Mori T, Tada S, Krajewski W, Rozovskaia T, Wassell R, Dubois G, Mazo A, Croce CM and Canaani E (2002): ALL-1 is a histone methyltransferase that assembles a supercomplex of proteins involved in transcriptional regulation. Mol Cell 10: 1119-28. Nakielny S and Dreyfuss G (1999): Transport of proteins and RNAs in and out of the nucleus. Cell 99: 677-90. Nasmyth K, Stillman D and Kipling D (1987): Both positive and negative regulators of HO transcription are required for mother-cell-specific mating-type switching in yeast. Cell 48: 579-87. Neer EJ, Schmidt CJ, Nambudripad R and Smith TF (1994): The ancient regulatory- protein family of WD-repeat proteins. Nature 371: 297-300. Nickerson J (2001): Experimental observations of a nuclear matrix. J Cell Sci 114: 463-74. Nyswaner KM, Checkley MA, Yi M, Stephens RM and Garfinkel DJ (2008): Chromatin-associated genes protect the yeast genome from Ty1 insertional mutagenesis. Genetics 178: 197-214. Ogg SC and Lamond AI (2002): Cajal bodies and coilin--moving towards function. J Cell Biol 159: 17-21. Ohno S, Kaplan WD and Kinosita R (1959): Formation of the sex chromatin by a single X-chromosome in liver cells of Rattus norvegicus. Exp Cell Res 18: 415-8. Ohno S, Wolf U and Atkin NB (1968): Evolution from fish to mammals by gene duplication. Hereditas 59: 169-87. Ohno S (1970): Evolution by Gene Duplication. In: pp Ed. Springer-Verlag, Berlin Owen DJ, Ornaghi P, Yang JC, Lowe N, Evans PR, Ballario P, Neuhaus D, Filetici P and Travers AA (2000): The structural basis for the recognition of acetylated histone H4 by the bromodomain of histone acetyltransferase gcn5p. Embo J 19: 6141-9. Ozsolak F, Song JS, Liu XS and Fisher DE (2007): High-throughput mapping of the chromatin structure of human promoters. Nat Biotechnol 25: 244-8. Pan LN, Lu J and Huang B (2007): HDAC inhibitors: a potential new category of anti-tumor agents. Cell Mol Immunol 4: 337-43. Pandolfi PP (2001): Histone deacetylases and transcriptional therapy with their inhibitors. Cancer Chemother Pharmacol 48 Suppl 1: S17-9. Parthun MR, Widom J and Gottschling DE (1996): The major cytoplasmic histone acetyltransferase in yeast: links to chromatin replication and histone metabolism. Cell 87: 85-94. Pasqualucci L, Bereschenko O, Niu H, Klein U, Basso K, Guglielmino R, Cattoretti G and Dalla-Favera R (2003): Molecular pathogenesis of non-Hodgkin's lymphoma: the role of Bcl-6. Leuk Lymphoma 44 Suppl 3: S5-12.

88 Paull TT, Haykinson MJ and Johnson RC (1993): The nonspecific DNA-binding and -bending proteins HMG1 and HMG2 promote the assembly of complex nucleoprotein structures. Genes Dev 7: 1521-34. Pederson T (1998): Thinking about a nuclear matrix. J Mol Biol 277: 147-59. Pederson T (2000): Half a century of "the nuclear matrix". Mol Biol Cell 11: 799- 805. Pendaries C, Tronchere H, Racaud-Sultan C, Gaits-Iacovoni F, Coronas S, Manenti S, Gratacap MP, Plantavid M and Payrastre B (2005): Emerging roles of phosphatidylinositol monophosphates in cellular signaling and trafficking. Adv Enzyme Regul 45: 201-14. Pettitt TR, Dove SK, Lubben A, Calaminus SD and Wakelam MJ (2006): Analysis of intact phosphoinositides in biological samples. J Lipid Res 47: 1588-96. Pickersgill H, Kalverda B, de Wit E, Talhout W, Fornerod M and van Steensel B (2006): Characterization of the Drosophila melanogaster genome at the nuclear lamina. Nat Genet 38: 1005-14. Pile LA, Spellman PT, Katzenberger RJ and Wassarman DA (2003): The SIN3 deacetylase complex represses genes encoding mitochondrial proteins: implications for the regulation of energy metabolism. J Biol Chem 278: 37840-8. Pokrovskaja K, Mattsson K, Kashuba E, Klein G and Szekely L (2001): Proteasome inhibitor induces nucleolar translocation of Epstein-Barr virus-encoded EBNA-5. J Gen Virol 82: 345-58. Polioudaki H, Kourmouli N, Drosou V, Bakou A, Theodoropoulos PA, Singh PB, Giannakouros T and Georgatos SD (2001): Histones H3/H4 form a tight complex with the inner nuclear membrane protein LBR and heterochromatin protein 1. EMBO Rep 2: 920-5. Pombo A, Cuello P, Schul W, Yoon JB, Roeder RG, Cook PR and Murphy S (1998): Regional and temporal specialization in the nucleus: a transcriptionally-active nuclear domain rich in PTF, Oct1 and PIKA antigens associates with specific chromosomes early in the cell cycle. Embo J 17: 1768-78. Ponder B (1988): Cancer. Gene losses in human tumours. Nature 335: 400-2. Qian YW, Wang YC, Hollingsworth RE, Jr., Jones D, Ling N and Lee EY (1993): A retinoblastoma-binding protein related to a negative regulator of Ras in yeast. Nature 364: 648-52. Qian YW and Lee EY (1995): Dual retinoblastoma-binding proteins with properties related to a negative regulator of ras in yeast. J Biol Chem 270: 25507-13. Qin S and Parthun MR (2002): Histone H3 and the histone acetyltransferase Hat1p contribute to DNA double-strand break repair. Mol Cell Biol 22: 8353-65. Raisner RM and Madhani HD (2008): Genomewide screen for negative regulators of sirtuin activity in Saccharomyces cerevisiae reveals 40 loci and links to metabolism. Genetics 179: 1933-44. Razin SV, Gromova, II and Iarovaia OV (1995): Specificity and functional significance of DNA interaction with the nuclear matrix: new approaches to clarify the old questions. Int Rev Cytol 162B: 405-48. Reddy KL, Zullo JM, Bertolino E and Singh H (2008): Transcriptional repression mediated by repositioning of genes to the nuclear lamina. Nature 452: 243-7. Rees KR, Rowland GF and Varcoe JS (1963): The metabolism of isolated rat-liver nucleoli and other subnuclear fractions. The active site of amino acid incorporation in the nucleus. Biochem J 86: 130-6. Reuter G and Spierer P (1992): Position effect variegation and chromatin proteins. Bioessays 14: 605-12. Richon VM, Garcia-Vargas J and Hardwick JS (2009): Development of vorinostat: Current applications and future perspectives for cancer therapy. Cancer Lett Robyr D, Suka Y, Xenarios I, Kurdistani SK, Wang A, Suka N and Grunstein M (2002): Microarray deacetylation maps determine genome-wide functions for yeast histone deacetylases. Cell 109: 437-46.

89 Rose HG and Frenster JH (1965): Composition and metabolism of lipids within repressed and active chromatin of interphase lymphocytes. Biochim Biophys Acta 106: 577-91. Rundlett SE, Carmen AA, Suka N, Turner BM and Grunstein M (1998): Transcriptional repression by UME6 involves deacetylation of lysine 5 of histone H4 by RPD3. Nature 392: 831-5. Rusche LN, Kirchmaier AL and Rine J (2002): Ordered nucleation and spreading of silenced chromatin in Saccharomyces cerevisiae. Mol Biol Cell 13: 2207-22. Rusche LN, Kirchmaier AL and Rine J (2003): The establishment, inheritance, and function of silenced chromatin in Saccharomyces cerevisiae. Annu Rev Biochem 72: 481-516. Russo VEA, Martienssen RA and Riggs AD (1996): Epigenetic Mechanisms of Gene Regulation. Cold Spring Harbor Laboratory Press, Woodbury Ruthenburg AJ, Allis CD and Wysocka J (2007): Methylation of lysine 4 on histone H3: intricacy of writing and reading a single epigenetic mark. Mol Cell 25: 15-30. Saitoh N, Spahr CS, Patterson SD, Bubulya P, Neuwald AF and Spector DL (2004): Proteomic analysis of interchromatin granule clusters. Mol Biol Cell 15: 3876-90. Sakurada K, Ohta T, Fujishiro K, Hasegawa M and Aisaka K (1996): Acetylpolyamine amidohydrolase from Mycoplana ramosa: gene cloning and characterization of the metal-substituted enzyme. J Bacteriol 178: 5781- 6. Salgado R, Toll A, Espinet B, Gonzalez-Roca E, Barranco CL, Serrano S, Sole F and Pujol RM (2008): [Analysis of cytogenetic abnormalities in squamous cell carcinoma by array comparative genomic hybridization.]. Actas Dermosifiliogr 99: 199-206. Saurin AJ, Shiels C, Williamson J, Satijn DP, Otte AP, Sheer D and Freemont PS (1998): The human polycomb group complex associates with pericentromeric heterochromatin to form a novel nuclear domain. J Cell Biol 142: 887-98. Schmutz J, Martin J, Terry A, Couronne O, Grimwood J, Lowry S, Gordon LA, Scott D, Xie G, Huang W, Hellsten U, Tran-Gyamfi M, She X, Prabhakar S, Aerts A, Altherr M, Bajorek E, Black S, Branscomb E, Caoile C, Challacombe JF, Chan YM, Denys M, Detter JC, Escobar J, Flowers D, Fotopulos D, Glavina T, Gomez M, Gonzales E, Goodstein D, Grigoriev I, Groza M, Hammon N, Hawkins T, Haydu L, Israni S, Jett J, Kadner K, Kimball H, Kobayashi A, Lopez F, Lou Y, Martinez D, Medina C, Morgan J, Nandkeshwar R, Noonan JP, Pitluck S, Pollard M, Predki P, Priest J, Ramirez L, Retterer J, Rodriguez A, Rogers S, Salamov A, Salazar A, Thayer N, Tice H, Tsai M, Ustaszewska A, Vo N, Wheeler J, Wu K, Yang J, Dickson M, Cheng JF, Eichler EE, Olsen A, Pennacchio LA, Rokhsar DS, Richardson P, Lucas SM, Myers RM and Rubin EM (2004): The DNA sequence and comparative analysis of human chromosome 5. Nature 431: 268-74. Schul W, Groenhout B, Koberna K, Takagaki Y, Jenny A, Manders EM, Raska I, van Driel R and de Jong L (1996): The RNA 3' cleavage factors CstF 64 kDa and CPSF 100 kDa are concentrated in nuclear domains closely associated with coiled bodies and newly synthesized RNA. Embo J 15: 2883-92. Schwerk C, Prasad J, Degenhardt K, Erdjument-Bromage H, White E, Tempst P, Kidd VJ, Manley JL, Lahti JM and Reinberg D (2003): ASAP, a novel protein complex involved in RNA processing and apoptosis. Mol Cell Biol 23: 2981-90. Shaw PJ and Brown JW (2004): Plant nuclear bodies. Curr Opin Plant Biol 7: 614- 20. Shen X, Xiao H, Ranallo R, Wu WH and Wu C (2003): Modulation of ATP- dependent chromatin-remodeling complexes by inositol polyphosphates. Science 299: 112-4.

90 Shiio Y, Rose DW, Aur R, Donohoe S, Aebersold R and Eisenman RN (2006): Identification and characterization of SAP25, a novel component of the mSin3 corepressor complex. Mol Cell Biol 26: 1386-97. Shogren-Knaak M, Ishii H, Sun JM, Pazin MJ, Davie JR and Peterson CL (2006): Histone H4-K16 acetylation controls chromatin structure and protein interactions. Science 311: 844-7. Sif S, Saurin AJ, Imbalzano AN and Kingston RE (2001): Purification and characterization of mSin3A-containing Brg1 and hBrm chromatin remodeling complexes. Genes Dev 15: 603-18. Silverstein RA, Richardson W, Levin H, Allshire R and Ekwall K (2003): A new role for the transcriptional corepressor SIN3; regulation of centromeres. Curr Biol 13: 68-72. Silverstein RA and Ekwall K (2005): Sin3: a flexible regulator of global gene expression and genome stability. Curr Genet 47: 1-17. Sims RJ, 3rd and Reinberg D (2006): Histone H3 Lys 4 methylation: caught in a bind? Genes Dev 20: 2779-86. Sironi E, Cerri A, Tomasini D, Sirchia SM, Porta G, Rossella F, Grati FR and Simoni G (2004): Loss of heterozygosity on chromosome 4q32-35 in sporadic basal cell carcinomas: evidence for the involvement of p33ING2/ING1L and SAP30 genes. J Cutan Pathol 31: 318-22. Skowyra D, Zeremski M, Neznanov N, Li M, Choi Y, Uesugi M, Hauser CA, Gu W, Gudkov AV and Qin J (2001): Differential association of products of alternative transcripts of the candidate tumor suppressor ING1 with the mSin3/HDAC1 transcriptional corepressor complex. J Biol Chem 276: 8734- 9. Sleeman JE and Lamond AI (1999): Newly assembled snRNPs associate with coiled bodies before speckles, suggesting a nuclear snRNP maturation pathway. Curr Biol 9: 1065-74. Smetana KS, WJ. Busch, H. (1963): A nuclear ribonucleoprotein network. Exp Cell Res 31: 198-201. Smith JS, Caputo E and Boeke JD (1999): A genetic screen for ribosomal DNA silencing defects identifies multiple DNA replication and chromatin- modulating factors. Mol Cell Biol 19: 3184-97. Somech R, Shaklai S, Geller O, Amariglio N, Simon AJ, Rechavi G and Gal-Yam EN (2005): The nuclear-envelope protein and transcriptional repressor LAP2beta interacts with HDAC3 at the nuclear periphery, and induces histone H4 deacetylation. J Cell Sci 118: 4017-25. Spector DL (2001): Nuclear domains. J Cell Sci 114: 2891-3. Steger DJ, Haswell ES, Miller AL, Wente SR and O'Shea EK (2003): Regulation of chromatin remodeling by inositol polyphosphates. Science 299: 114-6. Sternberg PW, Stern MJ, Clark I and Herskowitz I (1987): Activation of the yeast HO gene by release from multiple negative controls. Cell 48: 567-77. Stillman DJ, Dorland S and Yu Y (1994): Epistasis analysis of suppressor mutations that allow HO expression in the absence of the yeast SW15 transcriptional activator. Genetics 136: 781-8. Strahl BD and Allis CD (2000): The language of covalent histone modifications. Nature 403: 41-5. Strich R, Slater MR and Esposito RE (1989): Identification of negative regulatory genes that govern the expression of early meiotic genes in yeast. Proc Natl Acad Sci U S A 86: 10018-22. Su H, Altucci L and You Q (2008): Competitive or noncompetitive, that's the question: research toward histone deacetylase inhibitors. Mol Cancer Ther 7: 1007-12. Suka N, Suka Y, Carmen AA, Wu J and Grunstein M (2001): Highly specific antibodies determine histone acetylation site usage in yeast heterochromatin and euchromatin. Mol Cell 8: 473-9.

91 Suka N, Luo K and Grunstein M (2002): Sir2p and Sas2p opposingly regulate acetylation of yeast histone H4 lysine16 and spreading of heterochromatin. Nat Genet 32: 378-83. Sun ZW and Hampsey M (1999): A general requirement for the Sin3-Rpd3 histone deacetylase complex in regulating silencing in Saccharomyces cerevisiae. Genetics 152: 921-32. Sussel L, Vannier D and Shore D (1995): Suppressors of defective silencing in yeast: effects on transcriptional repression at the HMR locus, cell growth and telomere structure. Genetics 141: 873-88. Swift H (1959): Studies on nuclear fine structure. Brookhaven Symp Biol 12: 134- 52. Taunton J, Hassig CA and Schreiber SL (1996): A mammalian histone deacetylase related to the yeast transcriptional regulator Rpd3p. Science 272: 408-11. Thomas MC and Chiang CM (2006): The general transcription machinery and general cofactors. Crit Rev Biochem Mol Biol 41: 105-78. Thompson JD, Higgins DG and Gibson TJ (1994): CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22: 4673-80. Thon G and Klar AJ (1992): The clr1 locus regulates the expression of the cryptic mating-type loci of fission yeast. Genetics 131: 287-96. Trojer P and Reinberg D (2007): Facultative heterochromatin: is there a distinctive molecular signature? Mol Cell 28: 1-13. Turner BM, Birley AJ and Lavender J (1992): Histone H4 isoforms acetylated at specific lysine residues define individual chromosomes and chromatin domains in Drosophila polytene nuclei. Cell 69: 375-84. Tyler JK, Bulger M, Kamakaka RT, Kobayashi R and Kadonaga JT (1996): The p55 subunit of Drosophila chromatin assembly factor 1 is homologous to a histone deacetylase-associated protein. Mol Cell Biol 16: 6149-59. Wade PA, Jones PL, Vermaak D and Wolffe AP (1998): A multiple subunit Mi-2 histone deacetylase from Xenopus laevis cofractionates with an associated Snf2 superfamily ATPase. Curr Biol 8: 843-6. Wagner C, Dietz M, Wittmann J, Albrecht A and Schuller HJ (2001): The negative regulator Opi1 of phospholipid biosynthesis in yeast contacts the pleiotropic repressor Sin3 and the transcriptional activator Ino2. Mol Microbiol 41: 155- 66. Vakoc CR, Mandat SA, Olenchock BA and Blobel GA (2005): Histone H3 lysine 9 methylation and HP1gamma are associated with transcription elongation through mammalian chromatin. Mol Cell 19: 381-91. Wang A, Kurdistani SK and Grunstein M (2002): Requirement of Hos2 histone deacetylase for gene activity in yeast. Science 298: 1412-4. Wang C, Politz JC, Pederson T and Huang S (2003): RNA polymerase III transcripts and the PTB protein are essential for the integrity of the perinucleolar compartment. Mol Biol Cell 14: 2425-35. Wang H, Clark I, Nicholson PR, Herskowitz I and Stillman DJ (1990): The Saccharomyces cerevisiae SIN3 gene, a negative regulator of HO, contains four paired amphipathic helix motifs. Mol Cell Biol 10: 5927-36. Wang H and Stillman DJ (1990): In vitro regulation of a SIN3-dependent DNA- binding activity by stimulatory and inhibitory factors. Proc Natl Acad Sci U S A 87: 9761-5. Wang H and Stillman DJ (1993): Transcriptional repression in Saccharomyces cerevisiae by a SIN3-LexA fusion protein. Mol Cell Biol 13: 1805-14. Wang X, Connelly JJ, Wang CL and Sternglanz R (2004): Importance of the Sir3 N terminus and its acetylation for yeast transcriptional silencing. Genetics 168: 547-51. Wang ZG, Ruggero D, Ronchetti S, Zhong S, Gaboli M, Rivi R and Pandolfi PP (1998): PML is essential for multiple apoptotic pathways. Nat Genet 20: 266-72.

92 Vann LR, Wooding FB, Irvine RF and Divecha N (1997): Metabolism and possible compartmentalization of inositol lipids in isolated rat-liver nuclei. Biochem J 327 ( Pt 2): 569-76. Vannier D, Balderes D and Shore D (1996): Evidence that the transcriptional regulators SIN3 and RPD3, and a novel gene (SDS3) with similar functions, are involved in transcriptional silencing in S. cerevisiae. Genetics 144: 1343- 53. Weinberg RA (1991): Tumor suppressor genes. Science 254: 1138-46. Weis K, Rambaud S, Lavau C, Jansen J, Carvalho T, Carmo-Fonseca M, Lamond A and Dejean A (1994): Retinoic acid regulates aberrant nuclear localization of PML-RAR alpha in acute promyelocytic leukemia cells. Cell 76: 345-56. Verreault A, Kaufman PD, Kobayashi R and Stillman B (1996): Nucleosome assembly by a complex of CAF-1 and acetylated histones H3/H4. Cell 87: 95-104. Verreault A, Kaufman PD, Kobayashi R and Stillman B (1998): Nucleosomal DNA regulates the core-histone-binding subunit of the human Hat1 acetyltransferase. Curr Biol 8: 96-108. White MF and Bell SD (2002): Holding it together: chromatin in the Archaea. Trends Genet 18: 621-6. Vidal M, Buckley AM, Hilger F and Gaber RF (1990): Direct selection for mutants with increased K+ transport in Saccharomyces cerevisiae. Genetics 125: 313-20. Vidal M and Gaber RF (1991): RPD3 encodes a second factor required to achieve maximum positive and negative transcriptional states in Saccharomyces cerevisiae. Mol Cell Biol 11: 6317-27. Vidal M, Strich R, Esposito RE and Gaber RF (1991): RPD1 (SIN3/UME4) is required for maximal activation and repression of diverse yeast genes. Mol Cell Biol 11: 6306-16. Viiri K, Mäki M and Lohi O (2008): Patent Application 20086014. In: pp Ed. Finland Williams RR, Azuara V, Perry P, Sauer S, Dvorkina M, Jorgensen H, Roix J, McQueen P, Misteli T, Merkenschlager M and Fisher AG (2006): Neural induction promotes large-scale chromatin reorganisation of the Mash1 locus. J Cell Sci 119: 132-40. Wilson KL, Zastrow MS and Lee KK (2001): Lamins and disease: insights into nuclear infrastructure. Cell 104: 647-50. Winston F and Allis CD (1999): The bromodomain: a chromatin-targeting module? Nat Struct Biol 6: 601-4. Wiren M, Silverstein RA, Sinha I, Walfridsson J, Lee HM, Laurenson P, Pillus L, Robyr D, Grunstein M and Ekwall K (2005): Genomewide analysis of nucleosome density histone acetylation and HDAC function in fission yeast. Embo J 24: 2906-18. Vogelauer M, Wu J, Suka N and Grunstein M (2000): Global histone acetylation and deacetylation in yeast. Nature 408: 495-8. Vogelauer M, Rubbi L, Lucas I, Brewer BJ and Grunstein M (2002): Histone acetylation regulates the time of replication origin firing. Mol Cell 10: 1223- 33. Vogelstein B, Pardoll DM and Coffey DS (1980): Supercoiled loops and eucaryotic DNA replicaton. Cell 22: 79-85. Wolffe AP and Hayes JJ (1999): Chromatin disruption and modification. Nucleic Acids Res 27: 711-20. Wu J, Suka N, Carlson M and Grunstein M (2001): TUP1 utilizes histone H3/H2B- specific HDA1 deacetylase to repress gene activity in yeast. Mol Cell 7: 117- 26. Yang L, Mei Q, Zielinska-Kwiatkowska A, Matsui Y, Blackburn ML, Benedetti D, Krumm AA, Taborsky GJ, Jr. and Chansky HA (2003): An ERG (ets-related gene)-associated histone methyltransferase interacts with histone

93 deacetylases 1/2 and transcription co-repressors mSin3A/B. Biochem J 369: 651-7. Yang Q, Kong Y, Rothermel B, Garry DJ, Bassel-Duby R and Williams RS (2000): The winged-helix/forkhead protein myocyte nuclear factor beta (MNF-beta) forms a co-repressor complex with mammalian sin3B. Biochem J 345 Pt 2: 335-43. Yang X, Zhang F and Kudlow JE (2002): Recruitment of O-GlcNAc transferase to promoters by corepressor mSin3A: coupling protein O-GlcNAcylation to transcriptional repression. Cell 110: 69-80. Yochum GS and Ayer DE (2001): Pf1, a novel PHD zinc finger protein that links the TLE corepressor to the mSin3A-histone deacetylase complex. Mol Cell Biol 21: 4110-8. Yoshimoto H, Ohmae M and Yamashita I (1992): The Saccharomyces cerevisiae GAM2/SIN3 protein plays a role in both activation and repression of transcription. Mol Gen Genet 233: 327-30. Yuan GC, Liu YJ, Dion MF, Slack MD, Wu LF, Altschuler SJ and Rando OJ (2005): Genome-scale identification of nucleosome positions in S. cerevisiae. Science 309: 626-30. Zaidi SK, Young DW, Javed A, Pratap J, Montecino M, van Wijnen A, Lian JB, Stein JL and Stein GS (2007): Nuclear microenvironments in biological control and cancer. Nat Rev Cancer 7: 454-63. Zeng C, van Wijnen AJ, Stein JL, Meyers S, Sun W, Shopland L, Lawrence JB, Penman S, Lian JB, Stein GS and Hiebert SW (1997): Identification of a nuclear matrix targeting signal in the leukemia and bone-related AML/CBF- alpha transcription factors. Proc Natl Acad Sci U S A 94: 6746-51. Zhang J (2003): Evolution by gene duplication: an update. Trends in Ecology & Evolution 16: 292-298. Zhang L, Eugeni EE, Parthun MR and Freitas MA (2003): Identification of novel histone post-translational modifications by peptide mass fingerprinting. Chromosoma 112: 77-86. Zhang Y, Iratni R, Erdjument-Bromage H, Tempst P and Reinberg D (1997): Histone deacetylases and SAP18, a novel polypeptide, are components of a human Sin3 complex. Cell 89: 357-64. Zhang Y, Sun ZW, Iratni R, Erdjument-Bromage H, Tempst P, Hampsey M and Reinberg D (1998): SAP30, a novel protein conserved between human and yeast, is a component of a histone deacetylase complex. Mol Cell 1: 1021- 31. Zhang Y (2006): It takes a PHD to interpret histone methylation. Nat Struct Mol Biol 13: 572-4. Zhao K, Wang W, Rando OJ, Xue Y, Swiderek K, Kuo A and Crabtree GR (1998): Rapid and phosphoinositol-dependent binding of the SWI/SNF-like BAF complex to chromatin after T lymphocyte receptor signaling. Cell 95: 625- 36. Zink D, Amaral MD, Englmann A, Lang S, Clarke LA, Rudolph C, Alt F, Luther K, Braz C, Sadoni N, Rosenecker J and Schindelhauer D (2004): Transcription- dependent spatial arrangements of CFTR and adjacent genes in human cell nuclei. J Cell Biol 166: 815-25.

94 ORIGINAL COMMUNICATIONS

95 BMC Genomics BioMed Central

Research article Open Access TGF-β induces the expression of SAP30L, a novel nuclear protein Katri Lindfors1, Keijo M Viiri1, Marjo Niittynen1, Taisto YK Heinonen1, Markku Mäki*1 and Heikki Kainulainen2

Address: 1Paediatric Research Centre, Tampere University Hospital, Tampere, Finland and 2Institute of Medical Technology, University of Tampere, Tampere, Finland Email: Katri Lindfors - [email protected]; Keijo M Viiri - [email protected]; Marjo Niittynen - [email protected]; Taisto YK Heinonen - [email protected]; Markku Mäki* - [email protected]; Heikki Kainulainen - [email protected] * Corresponding author

Published: 18 December 2003 Received: 29 July 2003 Accepted: 18 December 2003 BMC Genomics 2003, 4:53 This article is available from: http://www.biomedcentral.com/1471-2164/4/53 © 2003 Lindfors et al; licensee BioMed Central Ltd. This is an Open Access article: verbatim copying and redistribution of this article are permitted in all media for any purpose, provided this notice is preserved along with the article's original URL.

Abstract Background: We have previously set up an in vitro mesenchymal-epithelial cell co-culture model which mimics the intestinal crypt villus axis biology in terms of epithelial cell differentiation. In this model the fibroblast-induced epithelial cell differentiation from secretory crypt cells to absorptive enterocytes is mediated via transforming growth factor-β (TGF-β), the major inhibitory regulator of epithelial cell proliferation known to induce differentiation in intestinal epithelial cells. The aim of this study was to identify novel genes whose products would play a role in this TGF-β-induced differentiation. Results: Differential display analysis resulted in the identification of a novel TGF-β upregulated mRNA species, the Sin3-associated protein 30-like, SAP30L. The mRNA is expressed in several human tissues and codes for a nuclear protein of 183 amino acids 70% identical with Sin3 associated protein 30 (SAP30). The predicted nuclear localization signal of SAP30L is sufficient for nuclear transport of the protein although mutating it does not completely remove SAP30L from the nuclei. In the nuclei SAP30L concentrates in small bodies which were shown by immunohistochemistry to colocalize with PML bodies only partially. Conclusions: By reason of its nuclear localization and close homology to SAP30 we believe that SAP30L might have a role in recruiting the Sin3-histone deacetylase complex to specific corepressor complexes in response to TGF-β, leading to the silencing of proliferation-driving genes in the differentiating intestinal epithelial cells.

Background terminal differentiation [1]. This acquisition of differenti- The intestinal epithelium is a constantly renewing popu- ated phenotype in epithelial cells requires finely tuned lation of cells, which arise from the proliferating stem gene expression where the genes that drive proliferation cells in the crypts of Lieberkuhn. In the intestinal mucosa are silenced while genes whose products are essential for the secretory crypt cells differentiate to absorptive entero- a differentiated cell are activated. cytes while migrating along the villus side to the villus tip. One of the most important modulators of intestinal epi- We have previously shown by differential display PCR thelial cell differentiation is transforming growth factor-β (DD-PCR) technique [2] that in a cell culture model, (TGF-β), which affects the cell cycle machinery leading to where differentiation of crypt-like T84 cells to enterocyte-

Page 1 of 8 (page number not for citation purposes) BMC Genomics 2003, 4 http://www.biomedcentral.com/1471-2164/4/53

e

n i t s e t a h t n c y e I l n n a l t e s 2,5 n e l e r c i n r g i m t o n e s c a e l a a n l s d v a o r e i u o i u l m p t e

B C H K L L M P S S S T 6,5 kB 2,0

1,5 1,3 kB

1,8 kB 1,6 kB 1,0 fold difference SAP30LFigure 2mRNA expression in human tissues 0,5 SAP30L mRNA expression in human tissues. Northern hybridization to a multi-tissue northern blot showed that a 1.3 kB transcript is expressed in all tissues examined, with higher expression in the testis and weaker in the liver and lung. A transcript of size 6.5 kB was abundantly expressed in brain and lung but not at all in liver and stomach. The lower panel shows the control hybridization with an β-actin specific probe. l ß o - tr F n G o T C

dues in the histone tails is associated with a more open chromatin state and thus increased DNA accessibility to ferentiatedFigureThe expression 1 T84 ofepithelial SAP30L cells in control and TGF-β-treated dif- transcription factors [reviewed in [8]], while deacetylation The expression of SAP30L in control and TGF-β-treated dif- or histone hypoacetylation is associated with tightly ferentiated T84 epithelial cells. The lower panel shows that in packed chromatin and transcriptional silencing [9,10]. DD-PCR the band representing SAP30L was present solely in Both transcriptional coactivators and corepressors regu- the RNA sample from the differentiated T84 cells. Real time quantitative PCR verified this differential expression to be 2.0 late gene expression by influencing the histone acetyla- times higher (SEM ± 0.13). tion status. Histone deacetylation is indeed a basic and well-conserved mechanism for gene silencing and it involves many corepressor proteins that vary depending on the repressor complex. The corepressor may well estab- lish the target specificity of a given deacetylase complex as like cells is induced by TGF-β [3] the changes in gene in the case of the Rb protein. It recruits a histone deacety- expression parallel those seen in differentiating intestinal lase complex to E2F transcription factors, leading to the epithelial cells in vivo [4]. In addition, a novel gene, apop- repression of transcription of E2F-dependent target genes tosis antagonizing transcription factor (AATF) (accession [11]. number HSA249940), whose expression was downregu- lated by TGF-β, was cloned from the undifferentiated In this paper we now for the first time describe the initial crypt-like cells of the same cell culture system [5]. AATF is characterization of another unknown transcript identified involved in epithelial cell proliferation, since it represses by DD-PCR in differentiating intestinal T84 epithelial the growth suppression function of the retinoblastoma cells. The newly identified gene codes for a protein mark- protein (Rb) [6] by inhibiting the recruitment of the his- edly similar to the histone deacetylase-associated core- tone deacetylase HDAC1 to the Rb/E2F complex [7]. pressor Sin3-associated protein 30 (SAP30), and therefore this novel Sin3-associated protein 30-like, SAP30L, could The acetylation status of the histone proteins mainly well have a role in silencing the genes crucial for prolifer- responsible for the packing of chromatin plays a key role ation in intestinal crypt epithelial cells. in transcriptional regulation. Acetylation of lysine resi-

Page 2 of 8 (page number not for citation purposes) BMC Genomics 2003, 4 http://www.biomedcentral.com/1471-2164/4/53

MNGFSTEEDSREGPPAAPAAAAPGYGQSCCLIEDGERCVRPAGNASFSKRVQKSISQKKLKLDIDKSVRHLYICDFH KNFIQSVRNKRKRKTSDDGGDSPEHDTDIPEVDLFQLQVNTLRRYKRHYKLQTRPGFNKAQLAETVSRHFRNIPVNE KETLAYFIYMVKSNKSRLDQKSEGGKQLE

TheFigure amino 3 acid sequence of SAP30L The amino acid sequence of SAP30L. SAP30L had an open reading frame of 183 amino acids. The PsortII-predicted N-glycosyla- tion sites are in italics, the N-myristoylation site is underlined and the nuclear localization signal is boxed.

Results homologous proteins showed that SAP30L has ortho- DD-PCR showed that TGF-β1 induced consistent and logues in several species (Figure 4). The corresponding reproducible upregulation of a transcript denoted SAP30L mouse protein is 97% identical with the human SAP30L. (see later) (Figure 1) using the arbitrary 5' primer AP-3 The Xenopus protein is 85% identical along the first 150 and the 3' anchoring primer T12MG. Quantitative RT- amino acids after which they begin to diversify considera- PCR using LightCycler technology verified this induction bly while the Drosophila melanogaster orthologue is fairly by TGF-β in three independent experiments. The differen- identical along the whole protein, the identity being 52%. tiated TGF-β-treated cells expressed this transcript 2.0 Interestingly, there was a human protein called SAP30, times more than the unstimulated T84 cells (Figure 1). which was 70% identical with SAP30L. Amino acid com- parison of SAP30L with SAP30 showed SAP30 to have in Sequence analysis of this transcript showed that SAP30L is its N-terminus a 38-amino-acid stretch that was absent in identical with an mRNA transcribed from the gene SAP30L. The corresponding SAP30 protein was also FLJ11526 located in chromosome 5q33.2. The gene has found in mouse but not in other species. four exons and the expected size of the transcribed mRNA is 1281 base pairs. Indeed, northern hybridization to a In transient transfection experiments on IMR-90 fibrob- multi-tissue northern blot showed that a SAP30L-specific lasts we were able to show that the wild-type SAP30L- probe recognized an mRNA of approximately 1.3 kB (Fig- EGFP fusion protein is indeed nuclear and concentrates in ure 2). The mRNA was expressed in all tissues examined, small dense bodies (Figure 5a). The transfection of the with somewhat weaker expression in the liver and lung EGFP fusion protein, which had only the putative nuclear and particularly abundant expression in the testis. Inter- localization and six flanking amino acids on either side estingly, there was also a transcript of size 6.5 kB which (pEGFP-NLS), also resulted in the nuclear localization of was abundantly expressed in brain and lung but not at all the protein (Figure 5b), thus providing evidence for the in liver and stomach. As the genomic sequence did not functionality of the signal. Mutation in the nuclear locali- predict an mRNA of this size, its identity remains to be zation signal (NLS) (KRKRK → KSNRK) disturbed this established. nuclear localization to some extent, causing the protein to be visible also in the cytosol although it did not com- Screening of a heart cDNA library in order to find the pletely inhibit the protein's nuclear transport (Figure 5c). whole-length transcript resulted in identification of a Immunocytochemical staining of these transfected cells positive clone with an insert of size 1.3 kB. When the with anti-promyelotocytic leukaemia (PML) antibody clone was sequenced and compared to the sequence of an showed these nuclear structures to be other than PML Image clone FLJ11526, the two sequences were found to bodies (Figure 5d). be identical. Both clones code for a protein of 183 amino acids (Figure 3), which was named Sin3-associated pro- Discussion tein 30 like, SAP30L. Prosite scan identified two putative We describe here the cloning of a novel human TGF-β- N-glycosylation sites (NASF, amino acids 44–47 and upregulated mRNA from differentiated T84 epithelial NKSR, amino acids 168–171), one N-myristoylation site cells. The novel mRNA is approximately 1.3 kB long and (GQSCCL, amino acids 26–31) and several phosphoryla- was ubiquitously expressed in all tissues examined. The tion sites for different kinases. The PsortII program pre- protein, called SAP30L, is 183 amino acids in length and dicted the SAP30L protein to be nuclear, the putative located in the nucleus, where it concentrates in small nuclear localization signal being KRKRK, ranging from dense structures other than PML bodies. amino acid 87 to 91 (Figure 3). A database search for

Page 3 of 8 (page number not for citation purposes) BMC Genomics 2003, 4 http://www.biomedcentral.com/1471-2164/4/53

SAP30L -MNGFSTEEDSR------Mus musculus EST -MNGFSTEEDSR------Xenopus laevis -MNGFSTEEDSR------Drosophila melanogaster MNNGFSTGEEDS------Homo sapiens SAP30 -MNGFTPEEMSRGGDAAAAVAAVVAAAAAAASAGNGNAAGGGAEVPGAGA Mus musculus SAP MNGFTPDEMSRGGDAAAAVAAVVAAAAAAASAGNGTGAGTGAEVPGAGA ***:. * .

SAP30L EGPPAAPAAAAPGYGQSCCLIEDGERCVRPAGNASFSKRVQKSISQKKLK Mus musculus EST EGPPAAPAAAP-GYGQSCCLIADGERCVRPAGNASFSKRVQKSISQKKLK Xenopus laevis DGP---PAQAAPFFGQTCCLIDGGERCPRPAGNASFSKRVQKSISQKKLK Drosophila melanogaster ------RGHTDQTCCLIDDMERCRNQAGYASYSKRIQKTVAQKRLK Homo sapiens SAP30 VSASGPPGAAGPGPGQLCCLREDGERCGRAAGNASFSKRIQKSISQKKVK Mus musculus SAP VSAAGPPGAAGPGPGQLCCLREDGERCGRAAGNASFSKRIQKSISQKKVK .* *** . *** . ** **:***:**:::**::*

SAP30L LDIDKSVRHLYICDFHKNFIQSVRNKRKRKTSDD-GGDSPEHDTDIPEV- Mus musculus EST LDIDKSVRHLYICDFHKNFIQSVRNKRKRKASDD-GGDSPEHDADIPEV- Xenopus laevis LDIDKNVRHLYICDFHKNYIQSVRNKRKRKTSDD-GGDSPEHETDIPEV- Drosophila melanogaster LSSDPAAQHIYICDHHKERIQSVRTKRRRKDSED---DSNETDTDLHEFP Homo sapiens SAP30 IELDKSARHLYICDYHKNLIQSVRNRRKRKGSDDDGGDSPVQDIDTPEV- Mus musculus SAP IELDKSARHLYICDYHKNLIQSVRNRRKRKGSDDDGGDSPVQDIDTPEV- :. * .:*:****.**: *****.:*:** *:* ** : * *.

SAP30L DLFQLQVNTLRRYKRHYKLQTRPGFNKAQLAETVSRHFRNIPVNEKETLA Mus musculus EST DLFQLQVNTLRRYKRHYKLQTRPGFNKAQLAETVSRHFRNIPVNEKETLA Xenopus laevis DLFQLQVNTLRRYKRYYKLQTRPGLNKAQLAEVLFNSERTLINVVHETKF Drosophila melanogaster DLYQLGVSTLRRYKRHFKVQTRQGMKRAQLADTIMKHFKTIPIKEKEIIT Homo sapiens SAP30 DLYQLQVNTLRRYKRHFKLPTRPGLNKAQLVEIVGCHFKSIPVNEKDTLT Mus musculus SAP DLYQLQVNTLRRYKRHFKLPTRPGLNKAQLVEIVGCHFRSIPVNEKDTLT **:** *.*******::*: ** *:::***.: : :.: ::

SAP30L YFIYMVKSNKSRLDQKSEGGKQLE Mus musculus EST YFIYMVKSNRSRLDQKSEGSKQLE Xenopus laevis LINKIIKGVVHLSNTFIS------Drosophila melanogaster FFVYMVKMGSNKLDQKNGLGNDTT Homo sapiens SAP30 CFIYSVRNDKNKSDLKADSGVH-- Mus musculus SAP YFIYSVKNDKNKSDLKVDSGVH-- : :: :

FigureMultiple 4alignment of SAP30L with its orthologues and SAP30 proteins of human and mouse Multiple alignment of SAP30L with its orthologues and SAP30 proteins of human and mouse. All six proteins are highly identical except for the 38 amino acids which appear in SAP30 of human, and mouse. Asterisks mark identical amino acids, colons and periods designate conservative substitutions.

SAP30L protein is 70% identical to a protein called extra 38 N-terminal amino acids. The size of for example SAP30, the most prominent difference being the lack of the Drosophila orthologue corresponds better with the size 38 amino acids in the N terminus of SAP30L. At the of SAP30L than with SAP30. genomic level, although their DNA sequences differ, their exon-intron organization is exactly the same, which sug- Based on the extremely high degree of identity in the pri- gests a common evolutionary origin for these genes. It mary structure of SAP30L and SAP30 it is probable that would appear that SAP30L is evolutionarily older than they also share functional similarity. SAP30 is a 30 kD SAP30, since only mammals have the protein with the nuclear protein associated with the Sin3 corepressor

Page 4 of 8 (page number not for citation purposes) BMC Genomics 2003, 4 http://www.biomedcentral.com/1471-2164/4/53

A B

C D

TransfectionFigure 5 of the different EGFP fusion constructs to IMR-90 fibroblasts Transfection of the different EGFP fusion constructs to IMR-90 fibroblasts. A) The wild-type SAP30L concentrates in small dense bodies in the nuclei. B) The nuclear localization of the fusion protein with only the NLS of SAP30L provides evidence for the functionality of the nuclear localization signal. C) Mutating the NLS disrupts the nuclear localization of the protein to some extent. D) Anti-PML-antibody staining (red) of wild-type SAP30L-transfected cells shows that the nuclear concentrates are other than PML bodies.

complex [12,13], which contains at least mSin3, HDAC1 Further evidence for the binding of SAP30L to mSin3a and 2, SAP18, RbAp46 and RbAp48 proteins [14]. The comes from a recent study by Fleischer and associates [18] Sin3 complex facilitates transcriptional repression by where they identified a novel 28 kD protein which binds being recruited to specific sites by different DNA-binding to mSin3a. The identified protein is very probably transcription factors [15,16] such as Mad, Ikaros, p53 and SAP30L. In the Sin3 repressor complex SAP30L might nuclear hormone receptors [17]. SAP30 is involved in the work similarly to SAP30 eg to recruit Sin3a to other repres- interaction at least in the case of nuclear hormone sor complexes than N-CoR. receptors, where it is thought to act as a specificity factor stabilizing or facilitating the interaction between the The recruitment of the Sin3-HDAC repressor complex to DNA-binding N-CoR and Sin3A [14]. SAP30 binds to N- E2-dependent promoters leads to exit from the cell cycle, CoR by its 129 N-terminal amino acids and to Sin3 with thus allowing differentiation to occur [19]. TGF-β pro- amino acids 129–220 [14]. Since the C-terminal part of motes exit from cell cycle in many ways, for example by SAP30L and SAP30 are markedly similar but SAP30L lacks directly inhibiting the expression of c-myc [20], which is 38 amino acids in the N-terminus when compared to mediated by proteins E2F4/5, Smads and p107 [21]. P107 SAP30, SAP30L is likely to bind Sin3A but not N-CoR. is a pocket protein able to bind to HDACs. It is interesting

Page 5 of 8 (page number not for citation purposes) BMC Genomics 2003, 4 http://www.biomedcentral.com/1471-2164/4/53

to speculate that SAP30L could be upregulated by TGF-β Kit (Perkin Elmer, Foster City, CA) as instructed by the in order to fulfil its role in the stabilization of the Sin3 manufacturer. repressor complex in E2F-dependent promoter sites to repress transcription of proliferation-associated genes Primers such as c-myc and may thus have a crucial role in The primers used in different experiments are shown in differentiation. Table 1 and were purchased from either Genset Oligos (Paris, France) or TAG Copenhagen (Copenhagen, Conclusions Denmark). In conclusion, we report here the identification of a novel transcript, SAP30L, which encodes a protein that very Quantitative PCR likely has a role in a histone deacetylase complex. Since Differential expression was confirmed using LightCycler the protein is 70% identical with a previously known technology in three independent RNA populations. One protein SAP30 it might function similarly to SAP30, e.g. in microgram of the Dnase I-treated total RNA was reverse- recruiting Sin3A to a specific repressor complex other than transcribed to cDNA using SuperScript II reverse tran- N-CoR, leading to the silencing of proliferation-driving scriptase (Gibco BRL) with 0.5 µg of oligo(dT) primer. gene(s) and ultimately to the differentiation of the intes- This cDNA was then subjected to PCR using a LightCycler tinal epithelial cells. Fast Start Cyber Green kit (Roche Diagnostics, Espoo, Fin- land) according to manufacturer's instructions. The prim- Methods ers 3EX3S and 3EX4AS (see Table 1) were used at a Cell lines and cultures concentration of 0.5 µM. The cycling conditions were as Human intestinal epithelial T84 cells (CCL 248) were pur- follows; 96C 10 min followed by 45 cycles at 96C 10 s, chased from the American Type Culture Collection (Rock- 57C 10 s and 72C 10 s. The relative amounts of the ville, MD). The cells were cultured in Dulbecco's modified unknown samples (control and TGF-β treated) were cal- Eagle medium and Ham's F-12 (1:1) (Gibco BRL, Paisley, culated by setting their cross points to the standard curve Scotland) supplemented with 5% foetal calf serum (FCS) generated by a serial dilution of cDNA produced from T84 and antibiotics (500 IU/ml penicillin and 100 µg/mL cells. The expression level of SAP30L in undifferentiated streptomycin; Gibco BRL). Three-dimensional type I col- and differentiated T84 cells was normalized by the house- lagen gel cultures were conducted as previously described keeping gene glyceraldehyde dehydrogenase. [3]. Differentiation of T84 cells was induced by adding 20 ng/ml of human recombinant TGF-β1 (hTGF-β1, R&D Screening of cDNA library for the whole length transcript Systems Europe, Oxon, UK) to the cultures and the cul- A human heart cDNA library (Rapid-Screen Arrayed tures were kept in 5% CO2 at 37°C for seven days. cDNA Library Panel; OriGene Technologies, Rockville, MD) was screened by PCR using primers 3EX3S and The human embryonic lung fibroblast cell line IMR-90 3EX4AS. The conditions of the PCR amplification for both (CCL 186) was purchased from the American Type Cul- Master Plate and Sub-plates were as follows: 95°C for 5 ture Collection. The cells were cultured in basal medium min, followed by 40 cycles of denaturation at 95°C for 45 (Eagle) supplemented with 10% FCS, 0.075% NaHCO3 s, annealing at 57°C for 30 s, and extension at 72°C for and 2 mmol/l glutamine. 60 s with a final extension at 72°C for 5 minutes. For the third round of screening, PCR was performed on single RNA isolation and differential display PCR bacterial colonies. DNA from positive clones was Total RNA was isolated from control and hTGF-β1-treated sequenced using both vector- and gene-specific primers three-dimensionally cultured T84 cells with TRIzol rea- indicated in the table. The accession number for SAP30L gent (Gibco BRL) as instructed by the manufacturer and is AY341060 subjected to DNase I (Roche Molecular Biochemicals, Indianapolis, IN) treatment, after which they were Northern hybridization extracted with phenol-chloroform-isoamylalcohol (Sigma A SAP30L specific PCR fragment was labelled with [α- Chemical Co., St. Louis, MO). DD-PCR was done accord- 32P]dATP (Amersham Pharmacia Biotech, Amersham, ing to the RNAmap™ protocol (GenHunter Corporation, UK) using Strip-Ez DNA™ Random Primed StribAble™ Nashville, TN) with arbitrary 5' primers and anchoring 3' DNA probe synthesis and removal kit (Ambion, Austin, primers. The reactions were repeated twice with newly USA) according to manufacturer's instructions and purified RNA in order to confirm the reproducibility of hybridised to human 12 tissue northern blot (Origene the results. The differentially expressed transcripts were Technologies). A β-actin specific probe served as a positive recovered from the gel and sequenced using the ABI control. PRISM Dye Terminator Cycle Sequencing Ready Reaction

Page 6 of 8 (page number not for citation purposes) BMC Genomics 2003, 4 http://www.biomedcentral.com/1471-2164/4/53

Table 1: Primers used in the study.

Primer Sequence (5'→3') Method Company

FLJforward CCCAAGCTTGGGGCGGGGAGATGAACGGCTTC EGFP-PCR Genset Oligos FLJreverse CCGGAATTCTCAAGCTGCTTGCCACCCTCCGA EGFP-PCR Genset Oligos EX1S AGCACGGAGGAGGACAGCCGCGAA Sequencing Genset Oligos EX2S GTAAGGCACCTATATATCTGTGAT Sequencing Genset Oligos EX2AS GTGTCGTGCTCGGGAGAATCTCCG Sequencing Genset Oligos EX3S GTTGATCTGTTCCAGCTGCAGGTG Library screening LightCycler Genset Oligos Northern probe EX3AS TTCTGCTAACTGGGCCTTATTGAA Sequencing Genset Oligos EX4AS TCAAGCTGCTTGCCACCCTCCGA Library screening LightCycler Genset Oligos Northern probe 3G3TNLSfor CGAAATAAAAGTAACAGGAAGACAAGT NLS mutation TAG Copenhagen 3G3TNLSrev ACTTGTCTTCCTGTTACTTTTATTTCG NLS mutation TAG Copenhagen

The table indicates the names and sequences of the primers used in given experiments.

Sequence analysis transfection, whereafter the non-specific binding of The nucleic acid sequence and the deduced amino acid antibodies was blocked with normal serum. The anti- sequence were searched against the NCBI Blast database PML-antibody was diluted 1:50 and incubated for one [22]. PSORTII server [23] was used to predict the subcel- hour. The TRITC-conjugated anti-mouse secondary serum lular localization of the SAP30L protein and to identify (1:200) (Dako A/S, Glostrup, Denmark) was incubated the putative peptide responsible for this localization. for an hour before the slides were dried and mounted with 50% glycerol in PBS. To assess the overlapping of the Construction of EGFP expression vectors and transfection EGFP-emitted green and TRITC-emitted red fluorescence Wild-type SAP30L cDNA was cloned into the pEGFP-C1 the slides were studied under a confocal microscope and (Clontech, Palo Alto CA) expression vector by polymerase the images were merged. chain reaction using primers (Table 1) with EcoRI and HindIII restriction sites at the 5'- and 3' ends, respectively. Abbreviations The mutation to the putative nuclear localization signal TGF-β, transforming growth factor-β; DD-PCR, differen- was generated by PCR using two complementary oligos tial display PCR; AATF, apoptosis antagonising transcrip- (Table 1) bearing the mutated sequence (R88→S and tion factor; Rb, retinoblastoma; SAP30, Sin3-associated K89→N) and the previously mentioned primers with protein 30; SAP30L, Sin3-associated protein 30-like; PML, EcoRI and HindIII restriction sites for cloning the insert promyelocytic leukaemia; FCS, foetal calf serum into the EGFP vector. In addition, for generation of the pEGFP-NLS construct a double-stranded oligo containing Authors' contributions the predicted consensus nuclear localization signal plus KL performed the differential display analysis, analyzed six flanking amino acids on either terminus was ordered the sequences and constructed the EGFP fusion vectors from TAG Copenhagen and cloned into pEGFP-C1. and participated in the transfections, immunocytochemi- cal stainings and also in the design of the study. She also 5 × 104 cells were plated on chamber slides (Nalge Nunc, wrote the manuscript. KMV performed the transfection Rochester, NY) and cultured for 24 hours prior to transfec- experiments and the immunohistochemical stainings. tion with 1 µg of the appropriate EGFP construct and Tfx- MN carried out the real-time quantitative PCR, screened 50 reagent (Promega, Madison, WI) for 2 hour. Detection the cDNA library for the whole-length transcript and of the green fluorescence protein by confocal microscopy performed the northern hybridization. TYKH participated (Ultraview Confocal Imaging System, PerkinElmer Life in the sequence analysis and library screening. MM and Sciences Inc., Boston, MA) was performed 24 hours after HK conceived, coordinated and designed the study. All transfection. authors read and approved the final manuscript.

Immunocytochemistry Expression of the PML protein in transfected IMR-90 fibroblasts was detected using a commercial antibody PG- M3 (Santa Cruz Biotechnology Inc, Santa Cruz, CA). The transfected cells were fixed with methanol 24 hours after

Page 7 of 8 (page number not for citation purposes) BMC Genomics 2003, 4 http://www.biomedcentral.com/1471-2164/4/53

Acknowledgements 18. Fleischer TC, Yun UI, Ayer DE: Identification and characteriza- The authors wish to thank Jorma Kulmala for technical assistance. tion of three new components of the mSin3a corepressor complex. Mol Cell Biol 2003, 23:3456-3467. 19. Lai A, Kennedy BK, Barbie DA, Bertos NR, Yang XJ, Theberge MC, The Coeliac Disease Study Group is supported by the Academy of Finland Tsai SC, Seto E, Zhang Y, Kuzmichev A, Lane WS, Reinberg D, Har- Research Council for Health, funding decision numbers 73489 and 201361, low E, Branton PE: RBP1 recruits the mSIN3-histone deacety- the Päivikki and Sakari Sohlberg Foundation, the Foundation of the Friends lase complex to the pocket of retinoblastoma tumour suppressor family proteins found in limited discrete regions of the University Children's Hospitals in Finland, the Foundation for Paedi- of the nucleus at growth arrest. Mol Cell Biol 2001, 21:2918-2932. atric Research in Finland, the Medical Research Fund of Tampere University 20. Seoane J, Pouponnot C, Staller P, Schader M, Eilers M, Massague J: Hospital and the Commission of the European Communities, specific RTD TGFβ influences Myc, Miz and Smad to control the CDK programme "Quality of Life and Management of Living Resources", QLK1- inhibitor p15Ink4b. Nat Cell Biol 2001, 3:400-408. CT-1999-00037, "Evaluation of the prevalence of coeliac disease and its 21. Chen C-R, Kang Y, Siegel PM, Massague J: E2F4/5 an p107 as Smad cofactors linking the TGFβ receptor to c-myc repression. Cell genetic components in the European population". The study does not nec- 2002, 110:19-32. essarily reflect the Commission's views and in no way anticipates its future 22. Altschul SF, Madden T, Schaffer A, Zhang J, Zhang Z, Miller W, Lipman policy in this area. DJ: Gapped BLAST and PSI-BLAST: A new generation of protein database search programs. Nucleic Acids Res 1997, 25:3389-3402. References 23. Nakai K, Kanehisa M: A knowledge base for predicting protein 1. Dignass A, Tsunekawa S, Podolsky D: Fibroblast growth factor localization sites in eukaryotic cells. Genomics 1992, 14:897-911. modulates epithelial cell growth and migration. Gastroenterol- ogy 1994, 106:1254-1262. 2. Liang P, Pardee A: Differential display of eukaryotic messenger RNA by means of the polymerase chain reaction. Science 1992, 257:967-970. 3. Halttunen T, Marttinen A, Rantala I, Kainulainen H, Mäki M: Fibrob- lasts and transforming growth factor-beta induce organiza- tion and differentiation of T84 human epithelial cells. Gastroenterology 1996, 111:1252-1262. 4. Lindfors K, Halttunen T, Kainulainen H, Mäki M: Differentially expressed CC3/TIP30 and rab11 along in vivo and in vitro intestinal epithelial cell crypt-villus axis. Life Sciences 2001, 69:1363-72. 5. Lindfors K, Halttunen T, Huotari P, Nupponen NN, Vihinen M, Visa- korpi T, Mäki M, Kainulainen H: Identification of novel transcrip- tion factor-like gene from human intestinal cells. Biochem Biophys Res Com 2000, 276:660-666. 6. Fanciulli M, Bruno T, Di Padova M, De Angelis R, Iezzi S, Iacobini C, Floridi A, Passananti C: Identification of a novel partner of RNA polymerase II subunit 11, Che-1, which interacts with and affects the growth suppression function of Rb. FASEB J 2000, 14:904-912. 7. Bruno T, De Angelis R, De Nicola F, Barbato C, Di Padova M, Corbi N, Libri V, Benassi B, Mattei E, Chersi A, Saddu S, Floridi A, Passananti C, Fanciulli M: Che-1 affects cell growth by interfering with the recruitment of HDAC1 by Rb. Cancer Cell 2002, 2:387-399. 8. Grunstein M: Histone acetylation in chromatin structure and transcription. Nature 1997, 389:349-352. 9. Braunstein M, Rose AB, Holmes SG, Allis CD, Broach JR: Transcrip- tional silencing in yeast is associated with reduced nucleo- some acetylation. Genes De 1993, 7:592-604. 10. Turner BM: Decoding the nucleosome. Cell 1993, 75:5-8. 11. Ferreira R, Naguibneva I, Mathieu M, Ait-Si-Ali S, Robin P, Pritchard LL, Harel-Bellan A: Cell cycle-dependent recruitment of HDAC-1 correlates with deacetylation of histone H4 on an Rb-E2F target promoter. EMBO Rep 2001, 2(9):794-799. 12. Hassig CA, Fleischer TC, Billin AN, Schreiber SL, Ayer DE: Histone deacetylase activity is required for full transcriptional repression by mSin3A. Cell 1997, 89:341-347. 13. Zhang Y, Iratni R, Erdjument-Bromage H, Tempst P, Reinberg D: His- Publish with BioMed Central and every tone deacetylases and SAP18, a novel polypeptide, are com- ponents of the human Sin complex. Cell 1997, 89:357-364. scientist can read your work free of charge 14. Laherty CD, Billin AN, Lavinsky RM, Yochum GS, Bush AC, Sun J-M, "BioMed Central will be the most significant development for Mullen T-M, Davie JR, Rose DW, Glass CK, Rosenfield MG, Ayer DE, disseminating the results of biomedical research in our lifetime." Eisenman RN: SAP30, a component of the mSin3 corepressor complex involved in N-CoR-mediated repression by specific Sir Paul Nurse, Cancer Research UK transcription factors. Mol Cell 1998, 2:33-42. Your research papers will be: 15. Kadosh D, Struhl K: Repression by Ume6 involves recruitment of a complex containing Sin3 corepressor and Rpd3 histone available free of charge to the entire biomedical community deacetylase to target promoters. Cell 1997, 89:365-371. peer reviewed and published immediately upon acceptance 16. Rundlett SE, Carmen AA, Suka N, Turner BM, Grunstein M: Tran- scriptional repression by UME6 involves deacetylation of cited in PubMed and archived on PubMed Central lysine 5 of histone H4 by RPD3. Nature 1998, 392:831-835. yours — you keep the copyright 17. Knoepfler PS, Eisenman RN: Sin meets NuRD and other tails of repression. Cell 1999, 99:447-50. Submit your manuscript here: BioMedcentral http://www.biomedcentral.com/info/publishing_adv.asp

Page 8 of 8 (page number not for citation purposes) Published online July 4, 2006

3288–3298 Nucleic Acids Research, 2006, Vol. 34, No. 11 doi:10.1093/nar/gkl401 SAP30L interacts with members of the Sin3A corepressor complex and targets Sin3A to the nucleolus K. M. Viiri1, H. Korkeama¨ki1, M. K. Kukkonen1, L. K. Nieminen1, K. Lindfors1, P. Peterson2,M.Ma¨ki1, H. Kainulainen3,4 and O. Lohi1,*

1Paediatric Research Centre, University of Tampere Medical School and Tampere University Hospital, Tampere, Finland, 2Molecular Pathology, University of Tartu, Tartu, Estonia, 3Institute of Medical Technology and Tampere University Hospital, Tampere, Finland and 4Department of Biology of Physical Activity, University of Jyva¨skyla¨, Finland

Received January 19, 2006; Revised March 24, 2006; Accepted May 11, 2006

ABSTRACT remodeling and DNA methylation work in concert (1,2) and at least in the ribosomal DNA locus (rDNA) these Histone acetylation plays a key role in the regulation epigenetic events occur in this particular hierarchical and of gene expression. The chromatin structure and temporal order (3). The Sin3A-HDAC corepressor complex accessibility of genes to transcription factors is consists of multiple proteins and regulates gene expression regulated by enzymes that acetylate and deacetylate by deacetylating histones. Sin3A itself functions as a scaffold histones. The Sin3A corepressor complex recruits protein that mediates various protein–protein interactions (4). histone deacetylases and in many cases represses HDAC 1 and HDAC 2, class I histone deacetylases, the his- transcription. Here, we report that SAP30L, a close tone binding proteins RbAp46 and RbAp48, SAP18, SAP30, homolog of Sin3-associated protein 30 (SAP30), SDS3, SAP180 and SAP130 are recognized components interacts with several components of the Sin3A of the ‘core’ Sin3A-HDAC corepressor complex (5–8). Of corepressor complex. We show that it binds to the these, SAP30 (Sin3A-Associated Protein 30) is a specific component of the Sin3A-complex since it is lacking in PAH3/HID (Paired Amphipathic Helix 3/Histone dea- other HDAC 1/2-containing complexes such as the NuRD cetylase Interacting Domain) region of mouse Sin3A complex (9). SAP30 is not required for intrinsic repression with residues 120–140 in the C-terminal part of the activity of the Sin3A complex but is involved in Sin3A- protein. We provide evidence that SAP30L induces mediated NCoR-repression by facilitating and stabilizing transcriptional repression, possibly via recruitment the interaction between these two corepressor proteins (10). of Sin3A and histone deacetylases. Finally, we In fact, many studies suggest that SAP30 functions as a brid- characterize a functional nucleolar localization sig- ging and stabilizing molecule between the Sin3A complex nal in SAP30L and show that SAP30L and SAP30 are and other corepressors such as CIR (11) or DNA-binding able to target Sin3A to the nucleolus. transcription factors like YY1 (12). In yeast, the DNA- binding repressor UME6 targets the SIN3–RPD3 complex (Sin3A-HDAC 1 homolog in Saccharomyces cerevisiae)to its target sequence in the promoter and causes highly local- INTRODUCTION ized histone deacetylation, occurring over a range of only It is well established that gene expression is influenced by one to two nucleosomes (13). chromatin structure. The compacted chromatin is a sterically In contrast to yeast, which has only one SAP30 homolog, hindered environment for transcription factors to bind mammals have two proteins, SAP30 and SAP30L (L for and assemble the transcription initiation complex, and is like), which share 70% sequence identity. They are both subject to active remodeling. Histone acetylation and DNA widely expressed in human tissues, with the most prominent demethylation are perceived as prerequisites for the ‘open expression being in tissues of hematopoietic origin (14). In state’ of chromatin, enabling transcription initiation. On the this article, we have begun to characterize the function other hand, histone deacetylation and DNA methylation con- of the mammalian SAP30L protein (15). We report that vert chromatin to a ‘closed state’, leading to the silencing of SAP30L is able to self-associate and interact with Sin3A. gene transcription. Recently, it has become evident that pro- Like SAP30, it has transcriptional repression capability and tein complexes that regulate histone acetylation, chromatin is able to associate with several class I histone deacetylases.

To whom correspondence should be addressed. Tel: +358 3 355 184 05; Fax: +358 3 355 184 02; Email: [email protected]

2006 The Author(s). This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/ by-nc/2.0/uk/) which permits unrestricted non-commerical use, distribution, and reproduction in any medium, provided the original work is properly cited. Nucleic Acids Research, 2006, Vol. 34, No. 11 3289

Furthermore, we have identified a novel and functional out with TnT Quick Coupled Transcription/Translation nucleolar localization signal (NoLS) in SAP30L, and show System (Promega) according to the manufacturer’s protocols. that Sin3A is targeted to the nucleolus by SAP30L and For GST pull-downs, 1 mg of GST or GST fusion proteins SAP30 (herein referred to as SAP proteins). Our results coupled to beads were incubated with 3–36 ml 35S-labeled show that SAP30L is able to associate with several partners in vitro translated proteins in binding buffer [1· phosphate- of Sin3A-HDAC complex and suggest that this complex buffered saline (PBS) (137 mM NaCl), 0.1% Igepal-CA630 may play a role in the nucleolus. and freshly added protease inhibitors (Roche)] in end over end rotation overnight at 4C. The beads were washed six times with the binding buffer containing 200 mM NaCl. MATERIALS AND METHODS GST pull-downs from the HEK293T nuclear lysates were done in a similar manner. Nuclei from the HEK293T cells Cloning and constructs were isolated as described previously (10). Human SAP30 cDNA was obtained from IMAGE clone 4074154. Human SAP30L cDNA has been described previ- Immunoprecipitation ously (15). Full length and deletion mutants of these proteins For the immunoprecipitation experiments, HEK293T whole were cloned into pcDNA3.1–myc-his vector (Invitrogen) cell lysates were prepared by lysing cells in RIPA lysis buffer for mammalian transfection experiments, and into pGEX- [1· PBS (137 mM NaCl), 1% Igepal-CA630, 0.5% sodium 4T1 vector (Amersham Biosciences) for production of deoxycholate and 0.1% SDS] with freshly added protease GST-fusion proteins in bacteria. Point mutations were inhibitors (Roche). Lysates were passed several times through created using the QuikChange Site Directed Mutagenesis a 21-gauge needle to sheer DNA, incubated for 30 min on Kit (Stratagene) according to manufacturer’s instructions. ice and centrifuged in 12 000 g for 20 min at 4C. Super- GAL4-DBD fusion proteins were created in the GAL4- natants were collected. Immunoprecipitations were carried DBD-vector (Stratagene). Luciferase reporter vectors under out in end over end rotation overnight at 4C with agarose- the control of TK(6) and 14D(10) promoters harboring 4 or conjugated antibodies against c-myc (9E10; sc-40AC) or 5xGal4 sites were generously provided by D. Reinberg (NJ, His (H-3; sc-8036AC) (Santa Cruz Biotechnology, Inc.), USA) and D. Ayer (Salt Lake City, USA), respectively. and they were washed six to eight times with RIPA lysis pCS2+MT-mSin3A plasmid (a generous gift from D. Ayer, buffer containing 500 mM NaCl and 0.5% Igepal-CA630. Salt Lake City, USA) was used as a PCR-template for mSin3A constructs created and cloned in pcDNA3.1-Myc- Western blotting His-vector. For in vitro transcription and translation For SDS–PAGE, lysed cells or protein samples were boiled experiments, a stop-codon was introduced into the mSin3A in Laemmli buffer and resolved on Novex pre-cast gels constructs to remove myc-his-tag. Flag-epitope tagged (Invitrogen). Proteins were transferred to a nitrocellulose HDAC1, HDAC2, RbAp46, RbAp48 and YY1 (12) cDNAs membrane (Amersham Biosciences) and blotted with the pri- were obtained from W. Yu (Taipei, Taiwan). HDAC3 mary antibodies indicated and HRP-conjugated secondary cDNA was obtained from U. Mahlknecht (Heidelberg, antibodies. Detection was performed with the ECL Plus Germany) and used as a template for PCR-cloning into Western Blotting Detection System (Amersham Biosciences). pcDNA3.1 vector with myc-his-tag. The precise coordinates The primary antibodies used were c-myc (sc-40), Sin3A of the constructs will be supplied on request. The authenticity (sc-767), HDAC 1 (sc-7872), HDAC 2 (sc-7899) and actin of the constructs was confirmed by sequencing. (sc-8432) from Santa Cruz, and GFP (33–2600) from Zymed. Anti-rabbit and anti-mouse HRP-conjugated Cell culture, transfections secondary antibodies were from DAKO (p0217 and p0260, Human embryonal kidney epithelial cells (HEK293T) were respectively). Band intensities were quantified using Image- cultured in DMEM (Gibco) supplemented with penicillin– QuantTL –program (Amersham Biosciences) streptomycin antibiotics, 5% fetal bovine serum, 1 mM sodium pyruvate and 50 mg/ml of uridine. For mammalian Immunocytochemistry transfection experiments, 2 · 104 cells were seeded into 2 HEK293T cells were fixed with 4% paraformaldehyde in 1cm surface area of tissue culture dishes. DNA was trans- PBS [1· PBS (137 mM NaCl)] for 20 min and then washed fected with FuGENE 6 reagent (Roche) according to the with PBS and permeabilized for 10 min with 0.2% Triton manufacturer’s protocol for 18–30 h. Thereafter, cells were X-100 in PBS. Unspecific binding of the antibodies was either lysed in Laemmli solution/lysis buffer (see below) or blocked by 1% BSA in PBS for 60 min before incubation fixed with 4% paraformaldehyde for the immunostaining of the cells with primary antibody at 1:200 dilutions for experiments. 60 min at 37C. After washes with PBS, the cells were incub- ated with secondary antibody at 1:1000 [Alexa Fluorophor GST pull-downs conjugated anti-mouse (A11031) or anti-rabbit (A11034) GST-SAP30 and GST-SAP30L fusion proteins were pro- IgG], washed and mounted on a DAPI-mount duced in Escherichia coli (BL-21 strain) and purified with (VectaShield). The primary antibodies used were NPM Glutathione Sepharose 4B beads (Amersham Biosciences) (32-5200, Zymed), Flag (F-3165, Sigma), His (46-0693, according to manufacturer’s instructions. The gel profile Invitrogen), c-myc (sc-40; Santa Cruz), c-myc (sc-789; of the GST-fusion proteins is shown in Supplementary Santa Cruz) and Sin3A (sc-994; Santa Cruz). Slides were Figure 3. In vitro transcription and translation was carried analyzed and photographed with a confocal microscope 3290 Nucleic Acids Research, 2006, Vol. 34, No. 11

(Ultraview Confocal Imaging System, Perkin Elmer Life Figure 5d). We next created three deletion mutants of Sciences Inc., Boston, MA). Sin3A (amino acids 1–200, 1–400 and 1–855) in order to map SAP30L interaction domain(s) in Sin3A. Pull-down HDAC activity and repression analysis studies with in vitro transcribed and translated Sin3A proteins Histone deacetylase activity was measured using Fluor de revealed that the interaction with the GST-SAP30L requires Lys AK-500-kit (Biomol) according to manufacturer’s proto- PAH3/HID region of Sin3A protein (Figure 1e). This again cols. In order to explore the role of class III HDAC enzymes, resembles the interaction of SAP30 with Sin3A, which has NAD+ coenzyme (N1511, Sigma-Aldrich) was added at been reported to require the PAH3 region of Sin3A (6,10). 200 mM to reactions. HDAC inhibitor Trichostatin A (TSA) These results also suggest that the interaction between was added in control reactions at 1 mM concentration. SAP30L and Sin3A is direct. Fluorescence was measured at 460 nm with VICTOR2 1420 multilabel counter (Wallac, Perkin Elmer, Life Sciences). SAP30L is able to self-associate For the repression analysis, HEK293T cells were trans- We next investigated whether SAP proteins are able to inter- fected with Gal4DBD-SAP30, Gal4DBD-SAP30L or act with each other. As shown in Figure 2a, myc-his-tagged Gal4DBD-SAP30L mutants along with 5xGal4-TK-LUC or SAP30L associates with gfp-tagged SAP30L in vivo but not 5xGal4-14D-LUC luciferase reporter plasmids as indicated. with gfp-tagged SAP30. On the other hand, myc-his-tagged The results were normalized by the activity of the b-gal SAP30 could not associate with gfp-tagged SAP30 or produced (16) from the cotransfected pcDNA3.1-LacZ SAP30L. These results demonstrate that SAP30L is able to (Invitrogen). Twenty-fourhour post-transfection cells were self-associate. split into two dishes, and treated with either TSA at To characterize more closely the domain(s) in SAP30L 200 nM or DMSO for 24 h. Thereafter, cells were harvested needed for self-association, we carried out pull-down experi- and luciferase activity was measured using Luciferase Assay ments with in vitro translated SAP30L proteins. As Figure 2b System (Promega). Measurements were done in duplicate shows, SAP30L 1–120 was unable to associate with full- from two independent experiments and the range is reported. length SAP30L. Two other SAP30L C-terminal deletion mutants (1–140 and 1–160) showed markedly impaired self-association capability compared to full-length SAP30L, RESULTS suggesting that the entire C-terminus of SAP30L is needed SAP30L interacts with Sin3A for efficient binding. Consistent with this, we found that mutating the nucleolar localization signal, which is composed SAP30 is a well-characterized binding partner for Sin3A of amino acids 120–127 (SAP30L 8A-mutant, see Figure 5), (5,10). Owing to its similarity to SAP30L, we examined had no effect on self-association. Intriguingly, truncating if SAP30L interacts with Sin3A. As shown in Figure 1a, 60 residues from the N-terminus of SAP30L increased inter- myc-tagged mouse Sin3A co-immunoprecipitated with myc- action over 2-fold (Figure 2b). These results suggest that his-tagged SAP30L in transiently transfected HEK293T cells. the ability of SAP30L to self-associate is dependent on an In this experiment, we used myc-his-tagged SAP30 and an intact C-terminus and that deletion of the N-terminus empty myc-his vector as positive and negative controls, increases this capability, possibly through conformational respectively. Consistent with previous studies, Sin3A changes in the protein. Furthermore, nucleolar targeting of co-immunoprecipitated with SAP30, whereas the control SAP30L (see Figure 5) is independent of its self-association, experiment with the vector alone remained negative, con- since the mutant incapable of nucleolar localization can firming the specificity of the interactions. In a reciprocal self-associate with an affinity comparable with that of wild- co-immunoprecipitation experiment, green fluorescent type SAP30L. protein (GFP)-tagged SAP30L co-immunoprecipitated with the myc-tagged Sin3A while GFP alone was unable to co- SAP30L associates with histone deacetylases and immunoprecipitate with Sin3A (Figure 1b). These results represses transcription confirm that the interactions are independent of the tag used. In GST pull-down experiments with nuclear lysates SAP30 associates with HDAC activity and HDAC 1 and of HEK293T cells, SAP30L associated with endogenous HDAC 2 proteins (6,10,12). Therefore, we asked if human Sin3A similar to SAP30 (Figure 1c). SAP30L also associates with HDACs. First, we carried out Next, we mapped the domains responsible for the inter- pull-down experiments with GST, GST-SAP30L and GST- action between SAP30L and Sin3A. We used C-terminally SAP30 proteins from HEK293T nuclear lysates and measured truncated versions of SAP30L (constructs are shown in associating HDAC activity. GST-SAP30L pulled down Figure 2b) and cotransfected them with mouse Sin3A. HDAC activity comparable with GST-SAP30 (Figure 3a), SAP30L 1–140 truncation mutant co-immunoprecipitated and this activity was sensitive to TSA, an inhibitor of Sin3A whereas SAP30L 1–120 construct failed to associate HDACs. Addition of NAD+, which is an essential cofactor with Sin3A, suggesting that the region between residues for the activity of class III HDACs, did not increase HDAC 120 and 140 of SAP30L is critical for the interaction activity, suggesting that class III HDACs (17) do not contrib- (Figure 1d). This finding is similar to SAP30 interaction ute to HDAC activity associated with SAP proteins in this with Sin3A, where the interaction domain resides in the assay. An intact C-terminus of SAP30L was necessary for C-terminus between residues 130 and 167 of SAP30 (10), HDAC activity as shown by a series of mutants of SAP30L a region sharing eight analogous residues with the SAP30L (Figure 3b). Further pull-down experiments demonstrated 1–140 truncation mutant (see sequence alignment in that GST-SAP30 and GST-SAP30L interacted with class I Nucleic Acids Research, 2006, Vol. 34, No. 11 3291

AB

C D

E

Figure 1. SAP30L interacts with Sin3A. (A) Sin3A co-immunoprecipitates with SAP30L. Lysates from the transfected HEK293T cells were immunoprecipitated with agarose conjugated anti-his antibody, and the immunocomplexes were analyzed by western blotting with anti-myc antibody. (B) SAP30L co-immunoprecipitates with the Sin3A protein. Transfected HEK293T cells were lysed and immunoprecipitated with the agarose conjugated anti-myc antibody, and the immunocomplexes were analyzed by western blotting with anti-myc and anti-gfp antibodies. (C) GST-SAP30L and GST-SAP30 pulls down endogenous Sin3A protein from nuclear lysate of HEK293T cells, whereas GST alone does not. Western blotting was performed with Sin3A antibody. (D) Residues between 120 and 140 in the SAP30L protein are critical for the association with the Sin3A. Immunocomplexes in the upper panel and inputs in the lower panel were analyzed by western blotting with anti-myc antibody. (E) SAP30L binds directly to PAH3/HID domain in Sin3A. Sin3A constructs used are illustrated on the left side of the experimental panel. The Sin3A proteins were produced by coupled in vitro transcription/translation system and labeled with 35S-Methionine before subjecting to pull-down experiments with GST-fusion proteins as indicated. SDS–PAGE was subjected to autoradiography. (Asterisk, IgG heavy and light chains; GST, glutathione-S-transferase; PAH, paired amhipathic helix; HID, histone deacetylase interacting domain.)

HDACs 1–3 (Figure 3c), whereas they failed to interact with of SAP30L constructs and SAP30. SAP30 has previously a class II histone deacetylase, HDAC 4 (data not shown). been reported to be able to repress transcription (6,10). The association of SAP30L with a functional Sin3A Wild-type Gal4SAP30L repressed transcription of a complex was examined using Gal4DBD fusions with a series reporter containing 14D promoter and Gal4 binding sites 3292 Nucleic Acids Research, 2006, Vol. 34, No. 11

A A

B B

C

Figure 3. SAP30L associates with histone deacetylase activity. (A and B) GST-fusion pull-downs from the HEK293T nuclear extracts were performed and HDAC activities were measured with the Fluor de Lys kit. Schematic representation of SAP30L constructs used are shown in Figures 2b and 5. GST and GST-SAP30 were used as a negative and positive control, respectively. The basal level of fluorescence (blank) and sensitivity in the experiment were defined by measuring fluorescence from the assay buffer + Figure 2. SAP30L is able to self-associate. (A) Myc-his-tagged SAP30L and from the 1 mM deacetylated standard respectively (white bars). NAD associates with gfp-tagged SAP30L in vivo. Lysates from the transfected coenzyme and TSA were added in the cases indicated. Shown are the means HEK293T cells were immunoprecipitated with agarose conjugated anti-myc of two experiments performed in duplicate and the error bars represent the antibody and the immunocomplexes were analyzed with the antibodies range of measurements (AFU ¼ arbitrary fluorescence unit). (C) GST-SAP30 indicated. (B) GST-SAP30L can directly interact with in vitro translated and GST-SAP30L pulls down endogenous HDAC 1 and 2, and transfected SAP30L. SAP30L proteins were produced by coupled in vitro transcription/ myc-his tagged HDAC 3 from the HEK293T lysates, whereas GST alone translation and labeled with 35S-methionine and subjected to pull-down does not (negative control). Pull-down complexes were analyzed by western experiment with full-length GST-SAP30L as indicated. Pull-down complexes blotting with the antibodies indicated. were analyzed by autoradiography. The amount of bound protein was quantified and normalized to the input. Data are illustrated in the histogram. The mutants of SAP30L are shown below. 1–140 possessed moderate repression activity (Figure 4b). However, cotransfection of Gal4SAP30L or Gal4SAP30 with either myc-tagged SAP30L, SAP30 or Sin3A did not (5xGal14D-LUC) dramatically compared to Gal4 alone yield any additive repression effect (Figure 4c), suggesting (23-fold) and 1.6-fold compared to Gal4SAP30 that the amount of these binding partners was not rate- (Figure 4a). Again, an intact C-terminus of SAP30L was limiting. TSA treatment greatly diminished the repression needed for full repression capability although SAP30L activity of SAP proteins (except for the SAP30L 1–140 Nucleic Acids Research, 2006, Vol. 34, No. 11 3293

A localization signal (NLS) was identified (15). We decided to examine more closely the subcellular localization of SAP30L using tagged wt and mutant SAP30L proteins transiently transfected into a variety of cell lines. GFP- or myc-his-tagged wt SAP30L was found in the nucleus of studied cell lines (MCF-7, COS-7, IMR-90, T84, Daudi, HEK293T), and a prominent, patchy staining pattern resem- bling that of the nucleolus was observed in the nucleus of many cells. To confirm this, we performed colocalization B experiments with a nucleolar marker, nucleophosmin (NPM or B23) (18). Figure 5a demonstrates that there is a marked colocalization between the two proteins in HEK293 cells and that SAP30L partly localizes to the nucleolus. Many nuc- lear and nucleolar proteins like HSP70, EBNA-5 (19), p53 and MDM2 (20) are known to accumulate in the nucleolus under proteotoxic stress caused by proteasome inhibitor MG132. Therefore, we tested whether MG132 affects the subnuclear localization of SAP30L and found that MG132 caused further accumulation of SAP30L in the nucleolus (Figure 5a). We decided to take advantage of this effect in our later experiments mapping the NoLS in SAP30L (see C below). GFP-tagged SAP30L showed similar strong nucleolar accumulation under MG132 treatment (Supplementary Figure 2) whereas GFP alone did not relocalize (data not shown) indicating that tags do not contribute to the results. SAP30 behaved similarly, although more slowly (Supple- mentary Figure 2): SAP30L accumulated into the nucleolus within 4 h whereas for SAP30 the accumulation took 6 h (data not shown). Confocal microscopy with the C-terminally truncated ver- sions of SAP30L protein (SAP30L 1–160, SAP30L 1–140 and SAP30L 1–120) revealed that the largest C-terminal dele- tion mutant (1–120) caused significant mislocalization of the protein to the cytoplasm and complete disappearance of the nucleolar localization (Figure 5b). This suggested the presence of a NoLS in the region between residues 120 and 140 of SAP30L. Closer examination of the sequence of SAP30L showed that this region harbors a stretch of basic residues consistent with a proposed NoLS consensus Figure 4. SAP30L is able to repress transcription. (A–C) HEK293T cells sequence (R/K-R/K-x-R/K) [Figure 5d and Ref. (21)]. were cotransfected with 5xGal4-14D luciferase reporter vector, Gal4DBD In order to assess the role of this region in the nucleolar fusions and LacZ-vector as indicated. Twenty-four hour post-transfection targeting of SAP30L, we constructed SAP30L 1–121, cells were either treated with TSA or DMSO (vehicle) for another 24 h. Lysed SAP30L 1–127 and SAP30L 1–131 mutants (Figure 5c). In cells were analyzed for luciferase and b-gal activity. The histogram illustrates the average fold-repressions of the Gal4DBD-fusions compared with Gal4 contrast to 1–121 truncation, SAP30L 1–127 deletion mutant alone. The measurements were done in duplicate from two independent accumulated in the nucleolus under MG132 treatment, experiments and the error bars represent the range. showing that the critical region responsible for nucleolar loc- alization resides between the residues 120–127 of SAP30L (Figure 5c). Next, we created two mutants with either three truncation), suggesting that HDAC activity plays an import- or four basic residues mutated to alanines in this region ant role in mediating the repression capability (Figure 4a of SAP30L (120RRYKR124 ! AAYARorAAYAA). These and b). Another reporter vector with TK promoter produced mutations reduced nucleolar accumulation of SAP30L, but identical results (Supplementary Figure 1). Taken together, failed to abolish it under MG132 treatment (data not these findings suggest that SAP30L represses transcription, shown). However, a larger mutation in the region (SAP30L and that this repression involves the recruitment of Sin3A 8A-mutant: 120RRYKRHYK127 ! AAAAAAAA) com- and histone deacetylases. pletely abolished nucleolar localization of SAP30L, demon- strating that these eight residues are responsible for correct localization of SAP30L to the nucleolus (Figure 5c and d). SAP30L has a functional NoLS and localizes to Previously, GFP-tagged SAP30L was reported to contain the nucleolus a functional NLS between residues 87 and 91 (15). We By transfection experiments, SAP30L was previously shown recreated this NLS mutant in a wt SAP30L construct with to localize to the nucleus of cells and a functional nuclear myc-his-tag (87KRKRK91 ! KAAAK). This mutant partly 3294 Nucleic Acids Research, 2006, Vol. 34, No. 11

A

B

C

D

Figure 5. SAP30L localizes in the nucleolus. (A), SAP30L colocalizes with nucleophosmin (NPM) in the nucleolus. Ten hour treatment with proteasome inhibitor MG132 causes further accumulation of SAP30L to the nucleolus. (B) NoLS signal resides between residues 120–140 in SAP30L. Arrows indicate cytoplasmic accumulation of the SAP30L 1–120 and SAP30L-NLSmut mutants. Arrowhead indicates the nucleolus. (C) Residues 120–127 in the SAP30L protein are necessary for its nucleolar accumulation. In the above experiments, HEK293T cells were transfected with the indicated myc-his-tagged constructs, and double-stained with anti-myc (construct) and anti-nucleophosmin (nucleolar marker) antibodies. Cells were further stained with DAPI in order to visualize the nuclei. Cells in the right panel were treated for 10 h with 10 mM proteasome inhibitor MG132. All the pictures were taken by confocal fluorescence microscope. Line-diagrams illustrate the fluorescence intensity (green, SAP30L; red, NPM) along the white line shown in merged images. SAP30L-myc-his constructs used are illustrated on the left side of each panel. (D) The NoLS of SAP30L and SAP30 identified in this study was manually aligned with three other previously published nucleolus localizations sequences of the following human proteins: catalytic subunit of human telomerase [TERT (34)]; NOLP (35) and hLa (21). Nucleic Acids Research, 2006, Vol. 34, No. 11 3295 localized to the cytoplasm but still showed some nucleolar A localization (Figure 5b), suggesting that NLS signal in SAP30L is functional only in nuclear targeting of the protein. When both signals were mutated simultaneously, nuclear loc- alization of SAP30L was significantly impaired (data not shown), indicating that the NoLS signal also contributes to the nuclear localization of SAP30L. We found an additional signal in the N-terminus of SAP30L (58KKLK61), which is also consistent with the NoLS proposed by Horke et al. (21). However, N-terminally deleted SAP30L protein (SAP30L 61–183) showed strong nucleolar localization of SAP30L, indicating that it is not a functional nucleolar local- ization signal (Figure 5b). Next, we investigated whether MG132 treatment causes relocalization to the nucleolus of other members of the Sin3A corepressor complex. A fraction of the endogenous Sin3A pool responded to MG132 treatment in a manner sim- ilar to the SAP proteins by accumulating in the nucleolus (Supplementary Figure 2). Consistent with other studies, HDAC1 and HDAC2 enzymes were detectable in the nucle- B olus (22) although there was no marked relocalization after MG132 treatment. Importantly, MG132 did not alter the sub- cellular localization of RbAp46, RbAp48 and YY1 proteins (Supplementary Figure 2). These results suggest that separate Sin3A complexes are present in cells, and that SAP proteins together with Sin3A and HDAC 1/2 enzymes belong to one such subcomplex, possibly within the nucleolus. This is consistent with a study reporting at least three separate Sin3A complexes with unique protein compositions (23).

SAP proteins target SIN3A to the nucleolus In quiescent cells, Sin3A is known to localize in the peri- nucleolar sites where early DNA replication origins are situated (24). Since Sin3A does not have any apparent NoLS sequence, we asked whether SAP30 and SAP30L pro- teins are able to target Sin3A to the nucleolus. To examine Figure 6. SAP30 and SAP30L target Sin3A to the nucleolus. (A) Cotransfected SAP proteins target Sin3A to the nucleolus. Myc-Sin3A this, we cotransfected Sin3A with either SAP30L, SAP30L was transfected with either myc-his tagged SAP30L, SAP30L 1–120 or 1–120 or SAP30 in HEK293T cells. In confocal microscopy, SAP30 and cells were treated for 10 h with MG132. Sin3A protein was wt SAP30L and SAP30 proteins dramatically increased the visualized by the Sin3A antibody and SAP proteins with the His-tag antibody. number of Sin3A-positive nucleoli (Figure 6a and b). Import- Stained cells were analyzed with confocal microscopy. Arrows indicate the nucleoli. (B) Data in A (without MG132) were scored as percent of cells antly, C-terminally deleted SAP30L (1–120), which does not expressing Sin3A in the nucleoli (total 100 cells counted for each associate with Sin3A, failed to target Sin3A to the nucleoli. experiment) and illustrated as the histogram. The results were similar in the presence or absence of MG132, although in MG132-treated cells, there was a low but constant level of Sin3A visible in the nucleolus (see SAP30L protein (Figure 7). The expression levels of two Supplementary Figure 2). We also scored Sin3A-positive other C-terminally deleted SAP30L proteins (1–140 and nucleoli, and found that 42 and 7% of the SAP30L- and 1–160) declined progressively: SAP30L 1–140 was expressed SAP30-transfected cells, respectively, were positive whereas 20 and SAP30L 1–160 16 times more efficiently than none of the control vector-transfected cells were (Figure 6b). the wt SAP30L protein. Furthermore, N-terminally deleted The transfected SAP30L and SAP30 proteins were also able SAP30L (61–183), which localizes intensively within the to relocate endogenous Sin3A to the nucleolus (data not nucleolus (Figure 5b), was expressed at very low levels, i.e. shown). These results indicate that Sin3A can be targeted five times less than the wt SAP30L protein. In other words, to the nucleolus by SAP proteins. It is noteworthy that nucleolar localization correlated inversely with protein SAP30L is more efficient than SAP30 in nucleolar targeting, expression levels. In these experiments, transfection effi- consistent with its more prominent localization within the ciencies were normalized to cotransfected lacZ protein and nucleolus. endogenous actin was used as a loading control. HEK293T cells treated for 10 h with MG132 stabilized ectopically The turnover of SAP30L is regulated by its C-terminus expressed wt SAP30L by 10-fold compared to control cells Transfected C-terminally deleted SAP30L (1–120) protein (DMSO-treated cells). Stabilization caused by MG132 was was expressed over 35 times more efficiently than the wt dependent of the residues between 120 and 140 of SAP30L 3296 Nucleic Acids Research, 2006, Vol. 34, No. 11

DISCUSSION We report here a novel component of the Sin3A corepressor complex, SAP30L. It binds to the PAH3/HID region of mouse Sin3A with residues 120–140, a region harboring sev- eral residues that are also conserved in SAP30. We provide evidence that SAP30L induces transcriptional repression, possibly via the recruitment of Sin3A and histone deacety- lases. We have also identified a region in SAP30L with a stretch of basic residues representing a functional NoLS signal (21). Moreover, both SAP proteins are capable of targeting Sin3A to the nucleolus. SAP30L and SAP30 are both transcribed from independent genomic loci (5q33.2 for SAP30L and 4q34.1 for SAP30). These two chromosomes are known to share chromosome- duplication blocks (26) and, in fact, macroscale analysis of gene composition of distal arms of 4q and 5q chromosomes suggests that a gene duplication event may have occurred during evolution. Closer inspection of the genomic sequences and phylogenetic footprinting analysis (Consite website, data not shown) of SAP genes reveal that they have different sets of conserved transcription factor binding sites on the promoters. Thus, different promoter sequences could allow specific transcription factors to regulate their expression in a timely and tissue-specific manner. Accordingly, previous studies (10,15) and gene expression databases (14) show that they are both ubiquitously expressed but with differences in expression pattern in, for example, testis, placenta and kidney. In our luciferase reporter analysis, SAP30L had 1.6-fold higher repression capacity than SAP30. If also true in vivo, use of a specific SAP protein could be a way to fine-scale the repression efficiency of the Sin3A corepres- sor complex. Various SAP proteins could also be used in spe- cific Sin3A-subcomplexes or recruited in response to varying demands of repression activity during the progression of cell cycle or in specific cell types. Figure 7. Fast turnover of the SAP30L protein correlates with its nucleolar Our results show that both SAP proteins localize partly localization. HEK293T cells were cotransfected with the indicated, myc-his- within the nucleolus. The nucleolus is the most prominent tagged constructs and a control vector (LacZ-myc-his). Cells were treated 10 h with either DMSO (vehicle), MG132 (proteasome inhibitor) or specialized organelle inside the nucleus. Its principal func- cyclohexamide (translation inhibitor). The lysed cells were subjected to tion is the transcription and processing of rRNA and the SDS–PAGE and analyzed by Western blotting with anti-myc and anti-actin assembly of ribosomes, although other functions, such as (loading control) antibodies. Band intensities were quantified and the amount ribonucleoprotein (RNP) assembly, cell cycle control, mRNA of the protein in the sample was calculated after normalization to LacZ expression. Representative results from three independent experiments are maturation, stress response and protein sequestration, shown. have recently been attributed to the nucleolus (27,28). In S.cerevisiae, SIN3A, SAP30 and RPD3 have been shown to since mutant lacking these residues (compare SAP30L 1–120 affect the transcription of the mating-type, telomeric and with SAP30L 1–140 and 1–160 mutants) did not show any rDNA loci. Interestingly, deletion of any of these genes stabilization after MG132 treatment (Figure 7). Since enhances silencing of RNA polymerase II-transcribed MG132 inhibits proteasomes, and ubiquitination of proteins reporter genes inserted into the above-mentioned three loci often marks them for degradation (25), we reasoned that (29). Similarly, a genetic screen for genes involved in SAP30L could be ubiquitinated. However, we failed to detect rDNA silencing in S.cerevisiae identified mutations in any ubiquitination of SAP30L even after MG132 treatment SIN3A, SAP30 and RPD3 genes (30). An alternate function (data not shown). This may imply that the stabilization of for the Sin3A complex in yeast was suggested by Meskauskas SAP30L is secondary to the inhibition of degradation of et al. (31), who showed that the main components of this other protein(s). Treating cells with cycloheximide, which complex participate not in the transcription, but in the early ceases protein translation, further demonstrated that mutants processing of rRNA. In the light of these reports, it is not lacking the C-terminus of SAP30L, and particularly the surprising that we found also mammalian SAP proteins in mutant lacking residues 120–140, have extended turnover the nucleolus. Furthermore, we were able to identify a compared to other SAP30L proteins (Figure 7). Others have NoLS signal in both SAP30L and SAP30. NoLS is generally previously reported that the turnover of endogenous SAP30 regarded as more of a protein–protein interaction domain protein is normally short, only 2 h in HeLa cells (10). than as a specific localization or targeting signal (32). It is Nucleic Acids Research, 2006, Vol. 34, No. 11 3297

thought that NoLS mediates interactions and thereby retains 10. Laherty,C.D., Billin,A.N., Lavinsky,R.M., Yochum,G.S., Bush,A.C., proteins in the nucleolus (33). In addition, our results show Sun,J.M., Mullen,T.M., Davie,J.R., Rose,D.W., Glass,C.K. et al. (1998) that Sin3A can be targeted to the nucleolus by SAP proteins. SAP30, a component of the mSin3 corepressor complex involved in N-CoR-mediated repression by specific transcription factors. Therefore, it can be postulated that the SAP proteins interact Mol. Cell, 2, 33–42. with component(s) of the nucleolus and, by this means, 11. Hsieh,J.J., Zhou,S., Chen,L., Young,D.B. and Hayward,S.D. (1999) recruit the Sin3A corepressor complex into the nucleolus CIR, a corepressor linking the DNA binding factor CBF1 to the for the regulation of rDNA transcription and/or rRNA histone deacetylase complex. Proc. Natl Acad. Sci. USA, 96, 23–28. processing. In future experiments, it will be essential to 12. Huang,N.E., Lin,C.H., Lin,Y.S. and Yu,W.C. (2003) Modulation of examine further the functional consequences of this nucleolar YY1 activity by SAP30. Biochem. Biophys. Res. Commun., 306, localization and recruitment. 267–275. 13. Kadosh,D. and Struhl,K. (1998) Targeted recruitment of the Sin3-Rpd3 histone deacetylase complex generates a highly localized domain of repressed chromatin in vivo. Mol. Cell Biol., 18, 5121–5127. SUPPLEMENTARY DATA 14. Kent,W.J., Sugnet,C.W., Furey,T.S., Roskin,K.M., Pringle,T.H., Zahler,A.M. and Haussler,D. (2002) The human genome browser at Supplementary Data are available at NAR Online. UCSC. Genome Res., 12, 996–1006. 15. Lindfors,K., Viiri,K.M., Niittynen,M., Heinonen,T.Y., Maki,M. and Kainulainen,H. (2003) TGF-beta induces the expression of SAP30L, ACKNOWLEDGEMENTS a novel nuclear protein. BMC Genomics, 4, 53. 16. Sambrook,J. and Russel,D.W. (2001) Molecular Cloning: The authors thank Taisto Heinonen for advice, Jorma A Laboratory Manual, 3rd edn. Cold Spring Harbor Laboratory, Kulmala for technical assistance and Olli Silvennoinen and Cold Spring Harbor, NY. 17. Imai,S., Armstrong,C.M., Kaeberlein,M. and Guarente,L. (2000) Leena Viiri for comments on the manuscript. The authors Transcriptional silencing and longevity protein Sir2 is an are grateful to Drs D. Ayer, D. Reinberg, W. Yu and NAD-dependent histone deacetylase. Nature, 403, 795–800. U. Mahlknecht for plasmids used in this work. This work was 18. Chan,W.Y., Liu,Q.R., Borjigin,J., Busch,H., Rennert,O.M., Tease,L.A. supported by the Academy of Finland Research Council for and Chan,P.K. (1989) Characterization of the cDNA encoding human Health (funding decision number 201361), the Foundation nucleophosmin and studies of its role in normal and abnormal growth. Biochemistry, 28, 1033–1039. for Paediatric Research in Finland, the Medical Research 19. Pokrovskaja,K., Mattsson,K., Kashuba,E., Klein,G. and Szekely,L. Fund of Pirkanmaa Hospital District, Maud Kuistila (2001) Proteasome inhibitor induces nucleolar translocation of Memorial Foundation, and Nona and Kullervo Va¨re Epstein–Barr virus-encoded EBNA-5. J. Gen. Virol., Foundation. Funding to pay the Open Access publication 82, 345–358. 20. Klibanov,S.A., O’Hagan,H.M. and Ljungman,M. (2001) Accumulation charges for this article was provided by Medical Research of soluble and nucleolar-associated p53 proteins following cellular Fund of Pirkanmaa Hospital District. stress. J. Cell Sci., 114, 1867–1873. 21. Horke,S., Reumann,K., Schweizer,M., Will,H. and Heise,T. (2004) Conflict of interest statement. None declared. Nuclear trafficking of La protein depends on a newly identified nucleolar localization signal and the ability to bind RNA. J. Biol. Chem., 279, 26563–26570. REFERENCES 22. Andersen,J.S., Lam,Y.W., Leung,A.K., Ong,S.E., Lyon,C.E., Lamond,A.I. and Mann,M. (2005) Nucleolar proteome dynamics. 1. Ayer,D.E. (1999) Histone deacetylases: transcriptional repression with Nature, 433, 77–83. SINers and NuRDs. Trends. Cell Biol., 9, 193–198. 23. Kuzmichev,A., Zhang,Y., Erdjument-Bromage,H., Tempst,P. and 2. Narlikar,G.J., Fan,H.Y. and Kingston,R.E. (2002) Cooperation between Reinberg,D. (2002) Role of the Sin3-histone deacetylase complex in complexes that regulate chromatin structure and transcription. Cell, growth regulation by the candidate tumor suppressor p33(ING1). 108, 475–487. Mol. Cell Biol., 22, 835–848. 3. Santoro,R. and Grummt,I. (2005) Epigenetic mechanism of rRNA gene 24. Lai,A., Kennedy,B.K., Barbie,D.A., Bertos,N.R., Yang,X.J., silencing: temporal order of NoRC-mediated histone modification, Theberge,M.C., Tsai,S.C., Seto,E., Zhang,Y., Kuzmichev,A. et al. chromatin remodeling, and DNA methylation. Mol. Cell Biol., 25, (2001) RBP1 recruits the mSIN3-histone deacetylase complex to the 2539–2546. pocket of retinoblastoma tumor suppressor family proteins found in 4. Silverstein,R.A. and Ekwall,K. (2005) Sin3: a flexible regulator of limited discrete regions of the nucleus at growth arrest. Mol. Cell Biol., global gene expression and genome stability. Curr. Genet., 47, 1–17. 21, 2918–2932. 5. Zhang,Y., Iratni,R., Erdjument-Bromage,H., Tempst,P. and 25. Ciechanover,A. (2005) Proteolysis: from the lysosome to ubiquitin and Reinberg,D. (1997) Histone deacetylases and SAP18, a novel the proteasome. Nature Rev. Mol. Cell Biol., 6, 79–87. polypeptide, are components of a human Sin3 complex. Cell, 89, 26. Friedman,R. and Hughes,A.L. (2003) The temporal distribution of gene 357–364. duplication events in a set of highly conserved human gene families. 6. Zhang,Y., Sun,Z.W., Iratni,R., Erdjument-Bromage,H., Tempst,P., Mol. Biol. Evol., 20, 154–161. Hampsey,M. and Reinberg,D. (1998) SAP30, a novel protein conserved 27. Shaw,P.J. and Brown,J.W. (2004) Plant nuclear bodies. Curr. Opin. between human and yeast, is a component of a histone deacetylase Plant Biol., 7, 614–620. complex. Mol. Cell, 1, 1021–1031. 28. Bernardi,R., Scaglioni,P.P., Bergmann,S., Horn,H.F., Vousden,K.H. 7. Alland,L., David,G., Shen-Li,H., Potes,J., Muhle,R., Lee,H.C., and Pandolfi,P.P. (2004) PML regulates p53 stability by sequestering Hou,H.,Jr, Chen,K. and DePinho,R.A. (2002) Identification of Mdm2 to the nucleolus. Nature Cell Biol., 6, 665–672. mammalian Sds3 as an integral component of the Sin3/histone 29. Sun,Z.W. and Hampsey,M. (1999) A general requirement for the deacetylase corepressor complex. Mol. Cell Biol., 22, 2743–2750. Sin3-Rpd3 histone deacetylase complex in regulating silencing in 8. Fleischer,T.C., Yun,U.J. and Ayer,D.E. (2003) Identification and Saccharomyces cerevisiae. Genetics, 152, 921–932. characterization of three new components of the mSin3A corepressor 30. Smith,J.S., Caputo,E. and Boeke,J.D. (1999) A genetic screen complex. Mol. Cell Biol., 23, 3456–3467. for ribosomal DNA silencing defects identifies multiple DNA 9. Zhang,Y., Ng,H.H., Erdjument-Bromage,H., Tempst,P., Bird,A. and replication and chromatin-modulating factors. Mol. Cell Biol., 19, Reinberg,D. (1999) Analysis of the NuRD subunits reveals a histone 3184–3197. deacetylase core complex and a connection with DNA methylation. 31. Meskauskas,A., Baxter,J.L., Carr,E.A., Yasenchak,J., Gallagher,J.E., Genes. Dev., 13, 1924–1935. Baserga,S.J. and Dinman,J.D. (2003) Delayed rRNA processing results 3298 Nucleic Acids Research, 2006, Vol. 34, No. 11

in significant ribosome biogenesis and functional defects. Mol. Cell 34. Etheridge,K.T., Banik,S.S., Armbruster,B.N., Zhu,Y., Terns,R.M., Biol., 23, 1602–1613. Terns,M.P. and Counter,C.M. (2002) The nucleolar localization 32. Olson,M.O. and Dundr,M. (2005) The moving parts of the nucleolus. domain of the catalytic subunit of human telomerase. J. Biol. Chem., Histochem. Cell Biol., 123, 203–216. 277, 24764–24770. 33. Nagahama,M., Hara,Y., Seki,A., Yamazoe,T., Kawate,Y., 35. Ueki,N., Kondo,M., Seki,N., Yano,K., Oda,T., Masuho,Y. and Shinohara,T., Hatsuzawa,K., Tani,K. and Tagaya,M. (2004) NVL2 is a Muramatsu,M. (1998) NOLP: identification of a novel human nucleolar AAA-ATPase that interacts with ribosomal protein L5 nucleolar protein and determination of sequence requirements for its through its nucleolar localization sequence. Mol. Biol. Cell, 15, nucleolar localization. Biochem. Biophys. Res. Commun., 252, 5712–5723. 97–102. Supplementary Figure 1

Supplementary Figure 2

Supplementary Figure 3 MOLECULAR AND CELLULAR BIOLOGY, Jan. 2009, p. 342–356 Vol. 29, No. 2 0270-7306/09/$08.00ϩ0 doi:10.1128/MCB.01213-08 Copyright © 2009, American Society for Microbiology. All Rights Reserved.

DNA-Binding and -Bending Activities of SAP30L and SAP30 Are Mediated by a Zinc-Dependent Module and Monophosphoinositidesᰔ† Keijo M. Viiri,1 Janne Ja¨nis,2 Trevor Siggers,3 Taisto Y. K. Heinonen,1 Jarkko Valjakka,5 Martha L. Bulyk,3,4,6 Markku Ma¨ki,1 and Olli Lohi1* Paediatric Research Centre, University of Tampere Medical School and Tampere University Hospital, 33520 Tampere, Finland1; Department of Chemistry, University of Joensuu, FI-80101 Joensuu, Finland2; Division of Genetics, Department of Medicine, Brigham & Women’s Hospital and Harvard Medical School, Boston, Massachusetts 021153; Department of Pathology, Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts 021154; Institute of Medical Technology and Tampere University Hospital, University of Tampere, FI-33014 Tampere, Finland5; and Harvard/MIT Division of Health Sciences and Technology, Harvard Medical School, Boston, Massachusetts 021156

Received 1 August 2008/Returned for modification 17 September 2008/Accepted 6 November 2008

Deacetylation of histones is carried out by a corepressor complex in which Sin3A is an essential scaffold protein. Two proteins in this complex, the Sin3A-associated proteins SAP30L and SAP30, have previously been suggested to function as linker molecules between various corepressors. In this report, we demonstrate new functions for human SAP30L and SAP30 by showing that they can associate directly with core histones as well as naked DNA. A zinc-coordinating structure is necessary for DNA binding, one consequence of which is bending of the DNA. We provide evidence that a sequence motif previously shown to be a nuclear localization signal is also a phosphatidylinositol (PI)-binding element and that binding of specific nuclear monophosphoi- nositides regulates DNA binding and chromatin association of SAP30L. PI binding also decreases the repres- sion activity of SAP30L and affects its translocation from the nucleus to the cytoplasm. Our results suggest that SAP30L and SAP30 play active roles in recruitment of deacetylating enzymes to nucleosomes, and mediate key protein-protein and protein-DNA interactions involved in chromatin remodeling and transcription.

A basic unit of chromatin is the nucleosome, in which 147 bp actions and thereby forms a platform for several enzymes (e.g., of DNA is wrapped around a histone octamer core composed HDACs and methyltransferases), DNA-binding transcription of the four histones H2A, H2B, H3, and H4. The N-terminal factors (e.g., Mad family repressors, MeCP2, and Pf1), and “tail” domains of these histones project out of the nucleosome other “bridging” proteins (e.g., SDS3) (52). The Sin3A-HDAC core and are the main sites of posttranslational modifications, corepressor complex contains HDAC1 and HDAC2, the his- such as acetylation, methylation, and phosphorylation. These tone-binding proteins RbAp46 and RbAp48, Sin3A-associated covalent modifications have been proposed to play important protein 18 (SAP18), SAP30, and SDS3 (52). roles in regulation of gene expression. According to the “his- Mammalian HDAC1 and HDAC2 are almost identical, each tone code” hypothesis, the modifications function as “marks” containing an N-terminal catalytic domain, which removes which are recognized by various proteins required for the dy- acetyl moieties from the ε-amino groups of lysine residues, and namic alterations in chromatin structure that are needed to a C-terminal tail. They are class I HDACs, which share se- make a gene accessible to the components of the transcription quence similarity with the Rpd3 (reduced potassium depen- machinery. dency-3) protein in Saccharomyces cerevisiae (59). The core- The recruitment of histone deacetylases (HDACs) to chro- pressor complex components RbAp46 and RbAp48 share 90% matin is a common mechanism of transcriptional repression sequence identity and belong to the WD repeat family. These (59). While certain repressors, such as Rb and YY1, are able to proteins have been shown to bind core histones H3 and H4 and recruit HDACs directly (5, 58), others, such as Mad-Max and are therefore thought to target HDAC-containing complexes the nuclear hormone receptor, require association with core- to their histone substrates (56). Another member of the com- pressors (1, 17, 18, 29, 42). Another example of the latter is the plex, SAP18, interacts directly with Sin3A and has multiple Sin3A corepressor complex, in which the Sin3A protein itself functions (60). In addition to regulating transcription, it also does not bind DNA or possess any enzymatic activity. Instead, participates in mRNA processing (49). SDS3 has been sug- it is composed of domains that mediate protein-protein inter- gested to play a role in stabilizing the Sin3A complex, and a yeast strain lacking SDS3 possesses only residual Sin3A-asso- ciated HDAC activity (31). SDS3 also participates in pericen- * Corresponding author. Mailing address: Paediatric Research Cen- tric heterochromatin formation and chromosome segregation tre, University of Tampere Medical School and Tampere University (9). Other components, such as SAP25, SAP130, and SAP180, Hospital, 33520 Tampere, Finland. Phone: 358 3 35518405. Fax: 358 3 have been reported to associate with the Sin3A-HDAC com- 3551 8402. E-mail: olli.lohi@uta.fi. † Supplemental material for this article may be found at http://mcb plex, but their roles in the complex have remained elusive .asm.org/. (13, 51). ᰔ Published ahead of print on 17 November 2008. SAP30 (Sin3A-associated protein 30) was originally identi-

342 VOL. 29, 2009 SAP30L AND SAP30 BIND AND BEND DNA 343

fied as a conserved member of the Sin3A corepressor complex tively. The H2B-GFP construct was a generous gift from P. Peterson (Tartu, (28, 61). It was shown to be required for N-CoR-mediated Estonia) (46). Point mutations were created using a QuikChange site-directed repression by the antagonist-bound estrogen receptor and the mutagenesis kit (Stratagene) according to the manufacturer’s instructions. The luciferase reporter vector, under the control of 14D promoters harboring 5ϫ POU domain protein Pit-1 but not by the unliganded retinoic Gal4 sites, was generously provided by D. Ayer (Salt Lake City, UT). The precise acid receptor or thyroid hormone receptor complexes (28). In coordinates of the constructs will be supplied on request. The integrity of the Saccharomyces cerevisiae, SAP30 was demonstrated to be im- constructs was confirmed by sequencing. portant for cell growth and to affect gene expression in a Cell culture, H2O2 treatment, and transfections. Human embryonic kidney epithelial cells (HEK293T) were cultured in Dulbecco’s modified Eagle’s me- promoter-dependent manner (61). A SAP30-deficient mutant dium (Gibco) supplemented with penicillin and streptomycin, 5% fetal bovine strain exhibited enhanced silencing of the ribosomal DNA, serum, 1 mM sodium pyruvate, and 50 ␮g/ml uridine. HeLa cells were cultured HMR, and telomeric loci, which suggested an antisilencing in RPMI 1640 (Gibco) supplemented with penicillin and streptomycin, 10% fetal function for SAP30 (36, 54, 55). Meskauskas et al. observed bovine serum, and L-glutamine. Cells were treated for 15 min with 0.5 mM H2O2, that mutations in Rpd3p, Sin3p, and Sap30p resulted in a washed, and allowed to grow for 4 h before being harvested for analysis. DNA was transfected with FuGENE 6 (for HEK293T cells) and FuGENE HD (for defect in rRNA processing rather than ribosomal DNA tran- HeLa cells) reagents (Roche) according to the manufacturer’s protocol. scription (41). Human SAP30, in addition to its associations in GST pulldown experiments. GST fusion proteins were produced in Escherichia the Sin3A-HDAC complex, has been reported to interact with coli (BL-21 strain) and purified with glutathione-Sepharose 4B beads (Amer- a number of other proteins, such as retinoblastoma-binding sham Biosciences) according to the manufacturer’s instructions. The fusion pro- teins were eluted from the beads with reduced glutathione, if necessary. The protein 1 (RBP1) (30), the CBF1-interacting corepressor quantity and integrity of the GST fusion proteins were checked using Coomassie- (CIR) (20), the YY1 transcription factor (21), and the inhibitor stained sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) of growth 1b (ING1b) tumor suppressor protein (27, 53). Re- gel. cently, SAP30-mediated transcriptional repression was shown GST pulldown experiments with nucleosomes and histones, with the Sin3a to play a role in the transmission and propagation of certain 1-855 protein transcribed and translated in vitro, were carried out as described previously (57), except that PI or PI 5-phosphate [PI(5)P] (Echelon, Inc.) was viruses (26, 32). The above observations indicate that SAP30 added to the reaction mixtures where indicated. plays a vital role in transcriptional regulation, which can be Electrophoretic mobility shift assays (EMSA) and a novel DNA ladder EMSA either negative or positive, depending on local factors, such (L-EMSA). A 150-bp DNA probe comprising the mitochondrial tRNA- 32 as chromatin state and the presence of various interacting Leu(UUR) sequence was end labeled with [␥- P]ATP as described in reference ␮ partners. 22. The probe was incubated with 0.5 g of a GST fusion protein on ice for 30 min in a buffer containing 50 mM Tris-HCl, pH 7.5, 125 mM NaCl, 2.5 mM Human SAP30L (SAP30-like), thus named because it shares dithiothreitol, 0.5 mM EDTA, 1 mM MgCl2, and 4% glycerol. The reaction 70% sequence identity with SAP30, is the “newest” member of products were analyzed on a 6% nondenaturing polyacrylamide gel, dried, and the Sin3A corepressor complex. It was originally discovered as autoradiographed. an expressed transcript in cultured T84 cells induced to differ- Because of the non-sequence-specific nature of the binding by the SAP30L ␤ and SAP30 proteins, a faster and simpler assay, L-EMSA, was designed for the entiate in response to transforming growth factor (35). Since protein-DNA interaction studies. Five micrograms of the fusion protein was then, SAP30L has been shown to associate with the Sin3A- incubated with 0.25 ␮g of a 1-kb DNA ladder (GeneRuler; Fermentas) in HDAC complex and to induce transcriptional repression in a phosphate-buffered saline for 10 min at room temperature. The reactions were Sin3A- and HDAC-dependent manner (57). Both SAP30 and run on ethidium bromide-containing 1% agarose gel with standard DNA gel SAP30L are able to localize to the nucleus or the nucleolus, loading buffer. Prior to use, this method was validated by comparing the DNA band shifts in L-EMSA to shifts in conventional EMSA. The GST fusion proteins and we have demonstrated that they can direct Sin3A to the used in Fig. 1A generated identical shifts in both assays (data not shown). Where nucleolus (57). indicated, PI and PI(5)P were added after the protein-DNA complex formation. Previous work has led to the view that SAP30 and SAP30L PBM experiments and data analysis. Protein binding microarray (PBM) ex- serve mainly a bridging role in various corepressor complexes. periments and analyses were performed as described in reference 3 for four different GST-tagged protein constructs (full-length SAP30, full-length SAP30L, In this study, we set out to investigate the functions and the SAP30 residues 1 to 131 [SAP30 1-131], and SAP30L residues 1 to 92 [SAP30L domain structures of these proteins in more detail. We show 1-92]) and a GST control (the protein concentration was ϳ1uM). An Alexa that both proteins directly bind and bend DNA and interact 488-conjugated anti-GST antibody was applied to the protein-bound microarray with core histones 2A/2B. Interestingly, our results suggest that to detect bound protein. The feature set of ϳ44,000 oligonucleotides present on both DNA binding and chromatin association are regulated by the custom-designed Agilent microarrays followed the design described in ref- erence 3 but was incorporated on the Agilent 4x44K array platform, allowing nuclear monophosphorylated phosphoinositides (PIs). four independent PBM experiments to be performed simultaneously on the same microarray. To identify DNA-binding-site motifs, two approaches were used: (i) the approach described in reference 3, based on perturbations of the highest- MATERIALS AND METHODS ranked 8-bp sequence; and (ii) a de novo motif search of the sequences from the Antibodies, immunoblotting, and immunofluorescence. The primary antibod- 20, 30, and 50 brightest microarray probes, using the program MEME (2). ies used were those against histone 2B (sc-8650), c-myc (sc-40), Sin3A (sc-767 Mass spectrometry. Wild-type and mutant SAP30L 1-92 peptides were cleaved and sc-5299), HDAC 1 (sc-7872), calregulin (sc-11398), and histone H1 (sc-8030) from GST by prothrombin, yielding SAP30L 1-94 peptides with two additional from Santa Cruz and glutathione S-transferase (GST) (27-4577-01) from GE amino acids from the GST vector. The samples were desalted on PD-10 columns Healthcare. Immunoblot analyses were done according to standard protocols, (Amersham Biosciences, Uppsala, Sweden), and concentrations were estimated ε ϭ Ϫ1 Ϫ1 and anti-rabbit, anti-mouse, or anti-goat horseradish peroxidase-conjugated sec- from absorbance at 280 nm by using 280 2,560 cm M . Prior to measure-

ondary antibodies (p0217, p0260, or p0449, respectively) were obtained from ments, the samples were further diluted with the appropriate solvents: CH3CN-

Dako. Immunofluorescence was performed as described previously (57), and H2O-acetic acid (49.5:49.5:1.0, vol/vol, pH 3.2) for denaturing solution condi- Alexa fluorophore-conjugated anti-mouse antibody (A11031) was used for de- tions and 10 mM ammonium acetate buffer (pH 6.8) for nondenaturing solution tection of immunocomplexes. conditions. All experiments were performed with a 4.7-T hybrid quadrupole Cloning and plasmid constructs. Full-length SAP30L and deletion mutants Fourier transform ion cyclotron resonance (Q-FT-ICR) instrument (APEX-Qe; were cloned into pcDNA 3.1-myc-his (Invitrogen), pGEX-4T1, and GAL4-DBD Bruker Daltonics, Billerica, MA) interfaced with an external electrospray ion- (Stratagene) vectors, and some of the constructs have already been described in ization (ESI) source (Apollo-II). The samples were infused directly at a flow rate ␮ Ϫ1 2 references 25 and 57. The SAP30L-green fluorescent protein (GFP) and of 1.5 l min , with dry N2 serving as the drying (10 lb/in , 200°C) and nebu- pcDNA3.1-Sin3A1-855 constructs are described in references 35 and 57, respec- lizing gas. ESI-generated ions were externally accumulated in a hexapole ion trap 344 VIIRI ET AL. MOL.CELL.BIOL.

FIG. 1. The N-terminal domains of SAP30L and SAP30 bind DNA in a sequence-independent manner. (A) A 32P-labeled DNA probe was incubated with GST fusion proteins, and the DNA-protein complexes were analyzed by an EMSA. (B) Median fluorescence intensities of all microarray probe oligonucleotides containing a particular 8-mer sequence, from PBM experiments performed with GST, GST-SAP30L, GST- SAP30, and GST-Cbf1.

for 0.5 to 1.0 s and transferred to an Infinity ICR cell for Sidekick trapping, RESULTS conventional “RF-chirp” excitation, and broadband detection. A total of up to 256 coadded (1-megaword) time domain transients were fast Fourier trans- SAP30L and SAP30 bind DNA. It is generally believed that formed prior to magnitude calculation and external frequency-to-m/z calibration specific repressor proteins are needed to direct the Sin3A- with respect to the ions of an ES tuning mix (Agilent Technologies, Santa Clara, HDAC complex to its target sequence, because Sin3A itself CA). All data were acquired and processed with the use of Bruker XMASS 7.0.8 software. does not bind DNA. Since SAP30L is able to repress transcrip- DNA bending/ligation-mediated circularization assay. The ligation-mediated tion in a Sin3A- and HDAC-dependent manner (25, 57), we circularization assay was essentially performed as described previously (45). See examined if it could bind DNA directly. An EMSA was carried also the legend to Fig. 3. out using a GST-SAP30L fusion protein incubated in the pres- Protein-lipid blot assays. PI phosphate (PIP) strips and arrays were purchased from Echelon Biosciences. Protein-lipid blot assays were performed by adding ence of mitochondrial DNA. As shown in Fig. 1A, a marked 0.5 ␮g/ml of GST fusion proteins and were further processed as described in the shift was detected in the mobility of the DNA after incubation manufacturer’s protocol. Each protein-lipid blot experiment was repeated at with GST-SAP30L, indicating a direct interaction with DNA. least once. A GST-SAP30 fusion protein also bound DNA, whereas GST Nucleosome preparations. Intact nucleosomes and tailless nucleosomes were prepared as described previously (37). The presence of solubilized nucleosomes alone did not. SAP30L proteins with a C-terminal deletion was confirmed by DNA agarose gel electrophoresis and SDS-PAGE, followed by (GST-SAP30L 1-120) or a mutated nucleolar localization sig- Coomassie staining (see Fig. S4A in the supplemental material). Calf thymus nal (NoLS; 8A mutant) were also able to interact with DNA. histones were purchased from Roche. For the GST fusion pulldown experiments, Deletion of 25 amino acid residues from the N terminus did 30 ␮g of nucleosomes, tailless nucleosomes, or histone proteins was used. For the initial screening experiment (Fig. 4B), 100 ␮g of histones was used. not affect DNA binding, whereas a larger deletion of 60 N- Interphase chromatin spreads, chromatin isolation, and subcellular fraction- terminal residues (GST-SAP30L 61-183) abolished the shift. ation. Chromatin spread preparation for the SAP30L-GFP-transfected Similarly, a deletion mutant lacking 40 residues from the N HEK293T cells was performed as described previously (38), with the following terminus failed to interact with DNA. These results show that modifications. Nocodazole was not added, and the collected, phosphate-buff- ered-saline-washed, hypotonically swollen cells were dropped to a tilted micro- both SAP30L and SAP30 are able to bind DNA and that the scope glass. The unfixed, dried drop was counterstained with DAPI (4Ј,6- region between residues 25 to 120 of SAP30L contains the diamidino-2-phenylindole) and photographed under a confocal microscope. DNA-binding determinants. Chromatin isolation and subcellular fractionation were performed as described DNA binding by SAP30L and SAP30 is not sequence spe- previously (39). Repression analysis. GAL4DBD-based repression analyses were performed as cific. In order to investigate whether SAP30L and SAP30 can described previously (57). target Sin3A to specific DNA sequences, we took advantage of VOL. 29, 2009 SAP30L AND SAP30 BIND AND BEND DNA 345

FIG. 2. The N-terminal domains of SAP30L and SAP30 contain an evolutionarily conserved zinc-binding module which is needed for DNA binding. (A) Clustal V alignment of the N-terminal amino acid sequences of SAP30L and SAP30 from various (selected) animals. Conserved cysteine (positions 29, 30, 38, and 74) and histidine (positions 70 and 77) residues in SAP30L are boxed, and the previously identified NLS is shaded in gray. Identical amino acids are marked by asterisks, and conservative substitutions are indicated by punctuation marks (colons and dots). (B) Charge-deconvoluted ESI Q-FT-ICR mass spectra of wild-type SAP30L in denaturing (upper) and nondenaturing (lower) solutions. The most abundant isotopic masses for the detected peptide variants are indicated. The insets show an expanded view of the isotopic distributions for the variant comprising residues 1 to 94 (this construct contains aa 1 to 92 of SAP30L and two amino acids derived from the thrombin cleavage site of the vector). The theoretical isotopic distributions were calculated from the sequence-derived elemental compositions (Table 1). The small arrow indicates the most abundant isotopic peak. (C) The SAP30L C29S, C38S, C74S, and H77A mutant constructs are degraded within 16 h after cleavage from GST, as analyzed by SDS-PAGE and Coomassie blue staining. wt, wild type. (D) L-EMSA with the GST-SAP30/SAP30L fusion proteins. (E) L-EMSA with the GST-SAP30L 1-92 fusion protein in the presence of 50 mM 1, 10-o-phenanthroline, a zinc-chelating agent. a recently developed method, the PBM (6a, 41a). GST-tagged sequence variants. Utilizing the universal microarray design proteins were applied to a microarray synthesized with double- and binding protocol recently described (3) (see Materials and stranded DNA oligonucleotides in order to assess the binding Methods), we performed PBM experiments using four con- preference for all possible contiguous and gapped 8-bp DNA structs: full-length fusion proteins GST-SAP30L and GST- 346 VIIRI ET AL. MOL.CELL.BIOL.

SAP30 (Fig. 1B) and N-terminally truncated fusion proteins TABLE 1. Calculated and experimentally determined masses for GST-SAP30L (amino acids [aa] 1 to 92) and SAP30 (aa 1 to wild-type SAP30L and the C30S and H70A mutants 131) (see Fig. S1 in the supplemental material). Using two m m m Ϫ m Elemental Peptidea exp calc exp calc different approaches (see Materials and Methods), we were (Da)b (Da)c (Da) compositiond unable to derive any binding motifs that would demonstrate apo wild type 10,413.33 10,413.28 0.05 C447H726N140O137S5 specific binding behavior. We further examined the median holo wild type 10,476.25 10,476.20 0.05 C447H724N140O137S5Zn1 apo(C30S) 10,397.31 10,397.35 Ϫ0.04 C H N O S signal intensity distribution of probe sequences containing 447 726 140 138 4 holo(C30S) 10,460.24 10,460.22 0.02 C447H724N140O138S4Zn1 Ϫ each 8-bp sequence variant and compared it to the distribution apo(H70A) 10,347.26 10,347.30 0.04 C444H724N138O137S5 Ϫ from a PBM experiment in which the sequence-specific tran- holo(H70A) 10,410.18 10,410.32 0.14 C444H722N138O137S5Zn1 scription factor Cbf1 from S. cerevisiae was used as a positive a The data are presented only for the peptide variants comprising residues control (3). The Z-transformed distributions of results from all 1to94. b Most abundant isotopic mass, experimentally determined. four experiments with SAP30L were very similar to that for the c Most abundant isotopic mass, calculated on the basis of the sequence-derived negative-control experiment (GST only), while the distribution elemental composition. d of results from the Cbf1 control experiment showed a narrower In the case of zinc binding, a loss of two protons was considered. distribution and a longer tail at high scores, indicative of spe- cifically bound probes (Fig. 1B). This comparison further sug- gested a lack of sequence-specific DNA binding of SAP30L and SAP30. theoretical 62.92-Da increase in mass, as discussed in detail SAP30L and SAP30 contain an N-terminal zinc-coordinat- elsewhere (12). The precision of the mass measurements was ing signature. A well-characterized DNA-binding element is approximately 0.01 Da, and such high accuracy provides an 2ϩ the zinc finger, which resembles a finger with a base of four unequivocal identification of the incorporated Zn cation in cysteine and/or histidine residues that coordinate a zinc ion SAP30L. No apo peptides or peptides with higher zinc-binding through the thiol groups of cysteine residues and/or the im- stoichiometries were detected. Similar results were obtained idazole nitrogens of histidine residues (33, 48). In a stretch of for the SAP30L C30S and H70A mutants (see Fig. S2A in the 49 residues (aa 29 to 77 in human SAP30L), we identified four supplemental material). The calculated and determined cysteine and two histidine residues, which suggested the pos- masses for the peptides SAP30L 1-94, C30S, and H70A are sibility of a zinc-coordinating motif. These residues are com- listed in Table 1. The SAP30L C29S, C38S, C74S, and H77A pletely conserved in a phylogenetic comparison of SAP30/ mutants had completely degraded into small peptide fragments SAP30L sequences from several species, including the fruit fly when expressed in E. coli (Fig. 2C; also see Fig. S2B in the and the human (Fig. 2A). To investigate whether SAP30L supplemental material) and when transiently transfected into binds zinc, we determined ESI Q-FT-ICR mass spectra for an mammalian cells (data not shown). Many of the peptide frag- N-terminal peptide of SAP30L (aa 1 to 92) and for mutants in ments that could be identified seemed to contain disulfide which the putative zinc-coordinating cysteine residues were bridges. In summary, these results indicate that the C2CH replaced by serines (C29S, C30S, C38S, and C74S) and histi- motif (C-X8-C-X35-C-X2-H) forms the zinc-binding module in dines by alanines (H70A and H77A), one at a time. Figure 2B SAP30L (see Fig. 7A) and that disruption of this module presents the spectrum measured for SAP30L 1-94 (this peptide destabilizes the protein and leads to its rapid degradation. contains two additional residues from the GST vector at its N Similar instability has been reported to occur in the zinc-defi- terminus) under denaturing conditions. To aid in the interpre- cient mutant of the Spt10p zinc finger protein in yeast (40). tation, the mass spectra were subjected to charge deconvolu- The C2CH module is needed for DNA binding. To further tion (i.e., conversion of m/z to Da). Four major peptide vari- explore the DNA-binding domains of SAP30L, we used a sim- ants, with relative abundance ratios of ϳ4:31:7:58, were plified EMSA which utilizes a commercial DNA ladder (L- detected, a result that is consistent with the SDS-PAGE anal- EMSA). To validate the assay, we incubated the full-length ysis (Fig. 2C). The mass of the second lightest variant GST-SAP30L and GST-SAP30 fusion proteins with the DNA (10,413.33 Da) agrees well with the mass calculated for the ladder and observed a marked shift in the mobility of the SAP30L 1-94 construct (Table 1). The species with the smallest DNA. The GST-SAP30L 1-92 protein was able to bind DNA mass (10,028.09 Da) is consistent with cleavage of three resi- and therefore contains all determinants for interaction with dues from the C terminus and is presumed to comprise resi- DNA. Two polybasic regions (PBRs) were shown to be neces- dues 1 to 89 of SAP30L. However, two heavier variants sary for DNA binding (Fig. 2D). One of these is in the loop of (11,131.58 and 11,766.95 Da) could not be assigned to any the zinc-coordinating structure (aa 50 to 69), and the other peptide sequence, even when all possible additional residues region (aa 84 to 92) has previously been identified as a nuclear from the expression vector were considered. All peptides ap- localization signal (NLS) (35). A hydrophobic region (aa 78 to peared as apo peptides, i.e., no zinc binding was detected 84) between the zinc-coordinating structure and the NLS is under denaturing conditions. also required for DNA binding, since a construct containing To detect zinc binding, the SAP30L 1-94 peptide was ana- both the zinc-coordinating structure and the hydrophobic re- lyzed under nondenaturing conditions (Fig. 2B). A 62.94-Da gion (aa 1 to 84) was able to bind DNA, whereas the zinc- increase in mass was detected for each of the four peptide coordinating structure alone (aa 1 to 77) was not. Similarly, a variants, consistent with the binding of one Zn2ϩ cation (Table construct which includes both the NLS motif (aa 78 to 92) and 1). The binding of Zn2ϩ (average mass ϭ 65 Da) by zinc finger the hydrophobic region is able to bind DNA, but the NLS domains is always accompanied by the loss of two protons alone (aa 84 to 92) is not. The role of the NLS is further (deprotonation of two coordinating cysteines), which gives a demonstrated by the full-length GST-SAP30L-KAAAK con- VOL. 29, 2009 SAP30L AND SAP30 BIND AND BEND DNA 347

FIG. 3. SAP30L bends DNA. (A) Probes of various lengths (170, 150, 130, 110, 90, and 70 bp), labeled internally with 32P and containing EcoRI “sticky ends,” were incubated with GST only (lanes 4, 9, 14, 19, 24, and 29) or with GST-SAP30L (lanes 5, 10, 15, 20, 25, and 30). The reaction mixtures for lanes 2 to 5, 7 to 10, 12 to 15, 17 to 20, 22 to 25, and 27 to 30 were incubated in the presence of T4 DNA ligase at 30°C for 20 min. The reaction mixtures for lanes 3 to 5, 8 to 10, 13 to 15, 18 to 20, 23 to 25, and 28 to 30 were subsequently treated with exonuclease (exo) III to remove any linear ligation products. The reaction products were electrophoresed on a 7% polyacrylamide gel, which was dried and subjected to autoradiography. Mono-, di-, and tricircular DNA ligation products are indicated by the numbered circles. nt, nucleotides or base pairs. (B) Increasing amounts of GST-SAP30L 1-92 were used in the ligation reactions, which were performed as described above. struct, which has markedly reduced affinity for DNA compared following the zinc-binding module and the intervening hydro- to wild-type GST-SAP30L. The importance of the zinc-coor- phobic pocket are also critical for DNA binding. dinating structure is best demonstrated in Fig. 1A, where it SAP30L has DNA-bending activity. The high-mobility-group can be seen that disruption of this structure (constructs (HMG) proteins provide a well-known example of a protein GST-SAP30L 61-183 and 40-183) completely abolishes family in which there can exist sequence-independent DNA DNA binding. binding, accompanied by bending of the DNA (19). To test the The mapping experiments described above suggested that ability of SAP30L to mediate bending of double-stranded the zinc-coordinating structure is necessary for DNA binding. DNA, we examined its effect on the T4 DNA ligase-dependent To explore whether DNA binding is dependent on zinc, the cyclization of short DNA fragments. This ligation-mediated GST-SAP30L 1-92 fusion protein was incubated with 1,10-o- circularization assay works on the principle that any double- phenanthroline, a zinc-chelating agent, and the L-EMSA was stranded DNA of less than 150 bp in length will not self- performed. As shown in Fig. 2E, 1, 10-o-phenanthroline at a 50 circularize in the presence of ligase, because of inherent limi- mM concentration abolished the DNA binding, as evidenced tations in the flexibility of DNA. Therefore, circularization is by a lack of shift in the mobility of DNA. Neither the GST seen only in the presence of a DNA-bending protein. 32P- peptide alone nor the methanol solvent elicited any mobility labeled PCR fragments (see Fig. S3 in the supplemental ma- changes. Addition of different divalent cations showed that terial) were incubated with the GST-SAP30L fusion protein or only Zn2ϩ could partially restore the DNA-binding activity of the GST-only control as detailed in Materials and Methods. As the N-terminal domain of SAP30L (see Fig. S2C in the sup- shown in Fig. 3A, 170-bp and 150-bp DNA fragments were plemental material). With DNA fragments less than 1,500 bp able to self-circularize, whereas shorter fragments were not. In in length, we saw reproducible mobility shifts, which evidenced the cases of shorter fragments (110 bp, 90 bp, and 70 bp), the zinc dependence and specificity of DNA binding by addition of GST-SAP30L resulted in the formation of circular SAP30L. We were unable to test the DNA-binding activity of monomers, whereas GST alone did not form any circular the specific zinc-coordinating mutants described above, be- monomer molecules. The lack of monomer formation in the cause of extensive degradation of these proteins. case of 130-bp fragments suggests that SAP30L introduces In summary, we have identified a putative zinc-binding such a strong bend to the DNA that cohesive ends are unable C2CH module in SAP30L and SAP30 and shown that DNA to line up. Such a lack of monomer formation in the cases of binding is dependent on this module. This zinc-dependent DNA molecules of certain lengths has also been reported for structure appears to be important for the stability of the entire HMG proteins (45). The GST-SAP30L 1-92 region was able to protein, and its disruption destabilizes the protein. Our results bend DNA in a dose-dependent manner (Fig. 3B). It is rea- also show that the loop region in the C2CH module is required sonable to infer that SAP30 similarly bends DNA, given its for DNA binding. Furthermore, the polybasic motif (NLS) identical zinc-binding module and DNA-binding properties. 348 VIIRI ET AL. MOL.CELL.BIOL.

FIG. 4. SAP30L and SAP30 bind histones and nucleosomes. (A) Schematic representation of the domain architecture of the HMG proteins, SAP30L, and SAP30. The charge average of SAP30L is presented as a sliding window of 10 aa. Surface probability predictions were performed using the approach of Emini (11), and values greater than 2 are shown as black lines. Zn, zinc-binding module; DNAbd & bending, DNA-binding and -bending domain. (B) GST fusion protein pulldowns of calf thymus histones (Roche) were analyzed by SDS-PAGE and stained with Coomassie blue. The asterisks mark the GST fusion proteins, and the arrows indicate interaction with histones 2A/2B. (C) GST fusion protein pulldowns of intact nucleosomes, of nucleosomes from which the tails had been removed with trypsin (see Fig. S4 in the supplemental material), and of calf thymus histones. WB, Western blot. (D) HEK293T cells were cotransfected with an H2B-GFP fusion protein and either wild-type SAP30L or SAP30Ldel109-113 containing a myc-His tag. The cells were stained with the anti-myc antibody, and nuclei positive for both H2B and GFP were scored from 50 cells. The results are illustrated in the histogram. (E) GST fusion pulldowns of intact nucleosomes were analyzed using a Western blot probed with the anti-H2B antibody.

SAP30L and SAP30 bind core histones and nucleosomes. end, we used purified histones, isolated nucleosomes, and tryp- We noticed a similarity in the domain architecture of SAP30L sin-cleaved nucleosomes (see Fig. S4A in the supplemental and SAP30 to that of the HMG proteins, both having an material). In a pulldown experiment, GST-fused SAP30L was N-terminal DNA-binding/bending domain followed by an able to associate with purified core histones 2A and 2B (Fig. acidic domain (Fig. 4A). As some reports indicate that the 4B). GST-SAP30L 1-92, which lacks the central acidic region, acidic region of HMG proteins mediates interactions with H1 interacted with histone 2A/2B slightly less than full-length or core histones (4, 7), we set out to examine whether SAP30L GST-SAP30L (Fig. 4B and C, lower panels). This comparison may also associate with core histones and nucleosomes. To this is made difficult by the degradation of the full-length GST- VOL. 29, 2009 SAP30L AND SAP30 BIND AND BEND DNA 349

SAP30L fusion protein and the possible loss of some of its followed by a PBR (85RNKRKRK91) (Fig. 5A). In SAP30L, binding determinants, which may lead to an underestimate of the PBR motif has previously been shown to act as an NLS the difference. When the GST-SAP30L 1-120 construct, which (35). To investigate the PI binding of SAP30L and SAP30, the contains the central acidic region, was used, histone binding GST fusion proteins were tested for binding to a variety of was substantially increased (Fig. 4B and C, lower panels), sug- immobilized lipids, as depicted in Fig. 5B. Both GST-SAP30L gesting that the central acidic region makes a significant con- and GST-SAP30 bound the monophosphorylated PIs PI(3)P, tribution to histone binding. The interaction of SAP30L with PI(4)P, and PI(5)P (Fig. 5C). No lipid binding was detected for histones was confirmed for histone 2B by using a specific an- GST alone. As a positive control, we used the PH domain of tibody (Fig. 4C), whereas in the case of histone 2A, the anti- phospholipase C-delta1, which interacted specifically with body cross-reacted with SAP30L (data not shown), and this PI(4,5)P2 in this assay. hampered the interpretation of the result. GST-SAP30 also To quantify the relative affinities of SAP30L and SAP30 for interacted with histone 2B (data not shown). various PIs, we used their fusion proteins to probe a lipid blot Full-length GST-SAP30L and GST-SAP30L 1-92 were able that contained serial dilutions of eight different PIs (Fig. 5D). to interact with DNA-containing nucleosomes. Even though Full-length GST-SAP30L bound most tightly to PI(5)P, fol- the SAP30L 1-120 construct showed a high affinity for histones lowed by PI(3)P and PI(4)P. The level of PI(5)P binding to without DNA, it did not interact with nucleosomes (Fig. 4C, GST-SAP30L was fourfold higher than that for PI(3)P and upper). In the 1-120 construct, the acidic region is likely to be eightfold higher than that for PI(4)P. GST-SAP30 bound to artificially exposed to acidic DNA, and consequently, its inter- immobilized PIs in an identical manner, though with slightly action with nucleosomes is prevented. Interestingly, the inter- lower affinities (Fig. 5D). action of SAP30L with nucleosomes was not dependent on the The determinants for PI binding were analyzed using trun- protruding N-terminal tails of histones, as demonstrated by a cated GST-SAP30L fusion proteins. Strikingly, deletion of 60 lack of effect after trypsin cleavage of the tails (Fig. 4C, mid- residues from the N terminus (SAP30L 61-183) resulted in dle). GST-SAP30 was similarly able to interact with nucleo- complete loss of PI binding (Fig. 5E), which was rescued only somes (Fig. 4E). Zhang et al. (61) have previously reported by inclusion of the entire zinc-binding structure, indicating that that SAP30 is unable to bind nucleosomes and core histones 3 the 25 residues in the N terminus are dispensable for this and 4. This is partially in line with our results in that we interaction. On the other hand, a construct containing the N detected only nonspecific interactions with histones 3 and 4, as terminus and the intact zinc-coordinating structure (aa 1 to 77) the GST moiety alone also bound them (Fig. 4B). was unable to bind PIs. Addition of the hydrophobic region (aa We have previously identified several mRNA isoforms of 1 to 84) resulted in weak PI binding, whereas inclusion of the SAP30L, including an isoform which lacks the entire exon 2 PBR motif following (aa 1 to 92) fully restored the interaction and five residues of exon 3. Specific deletion of these five (this construct also exhibited some nonspecific binding, but the residues (del109-113) markedly reduced the repression activity specificity was restored in the 1-120 construct). The PBR motif of SAP30L, whereas most of the HDAC activity was retained (aa 84 to 92) by itself was not sufficient for PI binding, but a (25). Intriguingly, these residues reside in the central acidic construct which includes both the PBR motif and the preced- domain and include a single aspartate residue. As shown in ing hydrophobic region (aa 78 to 92) was sufficient for the Fig. 4C, the del109-113 mutant was still able to interact with interaction. In the case of full-length SAP30L, mutating three purified histone 2B but not with purified nucleosomes contain- basic residues in the PBR motif to alanines markedly reduced ing DNA, a result that may explain the reduced repression its binding activity (KAAAK mutant). It is noteworthy that capability of this mutant. The lack of nucleosome binding by disruption of the loop in the zinc-coordinating structure by the del109-113 mutant can be explained only by this mutant’s deletion of residues 50 to 69 from otherwise intact SAP30L inability to interact with a histone-DNA complex, since it binds completely abolished the PI interaction (Fig. 5E). Finally, naked DNA with the same affinity as does the wild-type protein swapping of basic residues in the PBR motif with the polybasic (see Fig. S4C in the supplemental material). NoLS region (aa 1 to 84 plus the NoLS) changed the specificity In confocal microscopy, colocalization of histone 2B and of PI binding (Fig. 5E). SAP30L was detected, with simultaneous relocalization of hi- The above-mentioned results give rise to three conclusions. stone 2B around the nucleolus in response to overexpression of First, SAP30L and SAP30 interact specifically with monophos- SAP30L (Fig. 4D). Coexpression with wild-type SAP30L, but phorylated PIs. Second, interaction of SAP30L with PIs is not with the SAP30Ldel109-113 mutant, increased the pe- mediated by the PBR motif and supported by the preceding rinucleolar localization of H2B from 10% to over 80% (Fig. hydrophobic region and the zinc-coordinating structure. Third, 4D). Intriguingly, overexpressed histone 2B was able to direct the specificity of PI binding is partially determined by the NoLS-mutated SAP30L (8A) to the nucleolar region and thus composition of the basic sequence of the PBR. It is also note- overcome the lack of NoLS (data not shown). As a control, worthy that the same region in SAP30L interacts with both endogenous histone 1 showed no changes upon transient over- DNA and PIs, as summarized in Table 2. expression of SAP30L (data not shown). Monophosphoinositides regulate chromatin association of The polybasic NLS binds monophosphoinositides. Pf1, a SAP30L. GFP-tagged SAP30L associated with chromatin in recently identified Sin3A-binding protein, has a PI-binding vivo when hypotonically swollen HEK293 cells were splashed PBR following the first PHD zinc finger (24). A similar orga- on a microscope slide and counterstained with DAPI (Fig. 6A). nization is found in ING2, in which the PHD zinc finger is also It should be noted that SAP30L is not a component of chro- followed by a PBR. We identified a similar modular organiza- matin in the same way as histones, since it is not present in tion in SAP30L, which also contains a zinc-binding element mitotic chromosomes (data not shown). We next explored the 350 VIIRI ET AL. MOL.CELL.BIOL.

FIG. 5. The zinc-binding structure and the PBR in SAP30L and SAP30 bind monophosphoinositides. (A) Manual alignment of the sequences of the PBRs following the zinc-binding modules in SAP30L, SAP30, and Pf1. The last zinc-coordinating residue is boxed, and the basic residues are indicated with bold letters. (B) Schematic diagram of a lipid blot membrane (PIP strip) containing 20-pmol spots from samples of the following: lysophosphatidic acid (LPA), lysophosphocholine (LPC), PI (PtdIns), PtdIns(3)P, PtdIns(4)P, PtdIns(5)P, phosphatidylethanolamine (PE), phosphatidylcholine (PC), sphingosine 1-phosphate (S1P), PtdIns(3,4)P2, PtdIns(3,5)P2, PtdIns(4,5P)2, PtdIns(3,4,5)P3, phosphatidic acid (PA), phosphatidylserine (PS), and blank. (C, D and E) The indicated GST fusion proteins (0.5 ␮g/ml) were incubated with PIP strips or with the PIP array as described in Materials and Methods. The lipids which bound most strongly are indicated. Arrows indicate the specificity differences from the 1-92 construct. VOL. 29, 2009 SAP30L AND SAP30 BIND AND BEND DNA 351

TABLE 2. Summary of mapping studies of DNA and DISCUSSION PIP interactions Although the Sin3A-HDAC corepressor complex has been Resulta for: Construct studied extensively, the roles of the various members of this DNA interaction PIP interaction complex are poorly understood. In this study, we have explored -Wild-type 1-183 ؉؉؉ ؉؉؉ the functions of two members of this complex, the Sin3A 61-183 ϪϪassociated proteins SAP30L and SAP30, which share 70% se- Ϫ 40-183 NA quence identity. We have discovered three types of interactions 35-183 NA Ϫ 25-183 ϩϩϩ ϩϩϩ that illuminate the functional roles of these proteins. First, 1-77 ϪϪboth SAP30L and SAP30 interact directly with the core his- 1-84 ϩϩ ϩϩ tones 2A/2B. Second, we demonstrate that both proteins have 1-92 ϩϩϩ ϩϩϩ intrinsic DNA-binding activity which is partly mediated ϩϩϩ ϩϩϩ 1-120 through a novel N-terminal zinc-containing structure consist- 84-92 ϪϪ 78-92 ϩϩϩ ϩϩϩ ing of a C2CH module and a coordinated zinc ion. Binding to del50-69 ϪϪDNA is sequence independent and induces strong bending of 87KAAAK91 ؉؉the DNA. Third, we have identified a PI-binding site, a basic a NA, not available; Ϫ, no interaction; ϩ, weak interaction; ϩϩ, moderate domain which binds monophosphorylated PIs specifically, ad- interaction; ϩϩϩ, strong interaction. jacent to the zinc-binding element. Intriguingly, we have found that PI binding has a strong influence on the proteins’ affinity for DNA in vitro, which leads us to suggest that the DNA domains that regulate the chromatin association of SAP30L. binding is actually regulated by PIs. An increase in the con- As shown in Fig. 6B, KAAAK and del50-69 mutants of centration of PIs in the nucleus caused by hydrogen peroxide SAP30L associated significantly less than wild-type SAP30L leads to reduced repression activity and cytoplasmic relocal- with the chromatin-enriched fraction, as assayed by subcellular ization of SAP30L. fractionation. Also, an intact C-terminal domain was needed, Previously, SAP30 has been assigned the role of a linker presumably reflecting the importance of protein-protein inter- protein that mediates interactions of the Sin3-HDAC complex actions mediated by the C-terminal region. with various transcriptional repressors (e.g., YY1) or corepres- As the KAAAK and del50-69 mutants are deficient in both sors (e.g., N-CoR, CIR, and RBP1). Specifically, the interac- DNA/chromatin and PI binding (Fig. 2D and 5E), we asked if tion of SAP30 with N-CoR was demonstrated to occur through the association of SAP30L with chromatin is regulated by PIs, the N terminus, whereas the C terminus bound mSin3a (28). which could compete for the same binding sites and thus de- Our results suggest a second function for the N terminus, tach SAP30L from chromatin. Two lines of evidence indicate which we found to bind DNA. Our results do not necessarily that they do. First, the mobility shift generated by binding of contradict the previous reports, because one can imagine that SAP30L to DNA in the L-EMSA was greatly diminished after SAP30 can either bridge different multiprotein complexes or addition of equivalent molar amounts of monophosphorylated anchor a specific complex to nucleosomes, depending on the

PIs but not other PIs (Fig. 6C). Second, H2O2 treatment, which circumstances. Functional diversity of this kind is not unprec- has previously been shown to increase the amount of intranu- edented in Sin3A-associated proteins, since Fleischer et al. clear monophosphorylated PIs (23), led to a significant relo- (13) have observed at least three separate Sin3A-containing calization of myc-tagged SAP30L as assayed by the chromatin- complexes. Furthermore, the C terminus also seems to carry enriched fraction in HEK293 cells (Fig. 6D). In confocal multiple functions. Huang et al. (21) identified SAP30 as a microscopy of HeLa cells, 9% of nontreated and 41% of H2O2- binding partner for the transcription factor YY1 and showed treated cells expressed cytoplasmic GFP-SAP30L (Fig. 6E). that it is able to enhance YY1-mediated repression in a dose- The interaction of SAP30L with nucleosomes and Sin3A re- dependent manner. This interaction was mapped to the C- mained unchanged after addition of monophosphorylated PIs terminal region, i.e., the same region which also binds Sin3A, (Fig. 6F). These results suggest that intranuclear monophos- prompting the authors to suggest that the interactions of phorylated PIs associate with the PI-binding domain of SAP30 with YY1 and Sin3A are mutually exclusive. Huang et SAP30L and thereby regulate its association with chromatin. al. (21) suggested that HDAC activity could be brought to the PI binding decreases the repression activity of SAP30L. YY1-SAP30 complex through a direct interaction of SAP30 Finally, we tested if association of SAP30L with PIs influenced with HDAC1 (61). If SAP30L and SAP30 were to bind DNA its repression activity by utilizing a Gal4 fusion system with a and histones independently of Sin3A, it easy to envision that luciferase reporter vector as described previously (57). As their N-terminal domains could participate in anchoring the shown in Fig. 6G, reduced repression activity was observed YY1-SAP30-HDAC complex to chromatin to induce repres- both in the PBR mutant (SAP30L KAAAK), which lacks DNA sion of transcription. binding and mimics PI binding, and after H2O2 treatment, Sin3A by itself does not bind DNA or repress transcription which increases nuclear monophosphorylated PIs (23). Com- but instead mediates gene silencing through the enzymes that bined, these results suggest that association of SAP30L with it associates with (52). Targeting of the Sin3A complex is chromatin is dependent on intact C-terminal and PI-/DNA- carried out by DNA sequence-specific repressor proteins. binding domains and that monophosphorylated PIs disrupt this Here, we show that SAP30L and SAP30 are able to bind DNA association, leading to decreased transcriptional repression without any sequence specificity. This binding is dependent on through SAP30L. an intact N terminus that contains a C2CH-type zinc module, 352 VIIRI ET AL. MOL.CELL.BIOL.

FIG. 6. Chromatin association, subcellular localization, and transcriptional repression activities of SAP30L are regulated by its interactions with DNA and monophosphoinositides. (A) The SAP30L-GFP fusion protein colocalizes with interphase chromatin. (B) HEK293T cells were transfected with the indicated constructs, fractionated into subcellular fractions, and immunoblotted as indicated. Data from three independent experiments are illustrated as histograms, in which the bars represent the ranges of band intensities as measured by a densitometer. wt, wild type. (C) L-EMSA with GST-SAP30L

1-92 in the presence of equivalent molar quantities of PI (PtdIns) or PtdIns(5)P. (D) Cells were transfected with wild-type SAP30L, treated with H2O2, and subjected to subcellular fractionation as described for panel B. (E) Confocal images of cells transfected with SAP30L-GFP and treated with H2O2. (F) (Upper) In vitro-translated, 35S-methionine-labeled Sin3A1-855 was subjected to a pulldown experiment with GST-SAP30L and a GST-only control. PtdIns(5)P was added as indicated, and the results from the experiment were analyzed by SDS-PAGE and autoradiography. (Lower) Pulldown of nucleosomes with GST-SAP30L in the presence of PIs. (G) HEK293T cells were cotransfected with a 5ϫ Gal4-14D luciferase reporter vector, GAL4DBD ␮ fusions, and a LacZ vector, as indicated. At 24 h posttransfection, the cells were treated with 500 MH2O2 for 15 min, washed, and lysed after4hofincubation. Luciferase and ␤-galactosidase activities were measured, and the histograms illustrate the average repressions of the GAL4DBD fusions relative to the level for GAL4 alone. The measurements were done in duplicate in two independent experiments, and the bars represent the ranges of observed values. VOL. 29, 2009 SAP30L AND SAP30 BIND AND BEND DNA 353 whose disruption abolishes the DNA binding. Zinc fingers also modulated by specific inositol polyphosphates, the cleav- were originally identified as DNA-binding motifs, but they are age products generated by PI-specific phospholipase C (50). now known to bind RNA, protein, and lipid substrates as well Additionally, another SWI/SNF-like chromatin remodeling (6, 14, 16, 24). A zinc finger consists of two antiparallel ␤ complex, BAF, is targeted to chromatin and the nuclear matrix strands and an ␣ helix, and the zinc ion is crucial for its specifically by a PIP2-dependent mechanism upon lymphocyte stability. Usually, a single zinc finger does not bind DNA with activation (62). Pf1, a recently identified nuclear binding part- very high affinity and can recognize only two or three base ner for the corepressors mSin3A and TLE, has a PBR which pairs, but when several, up to 60, zinc fingers are strung to- binds specific monophosphoinositides (24). The Sin3A-binding gether, the group binds more tightly and can recognize longer tumor suppressor ING2 binds PI(3), PI(4)P, and PI(5)P and DNA sequences. In the cases of both SAP30L and SAP30, only shows PI(5)P-dependent association with chromatin and in- a single zinc-coordinating element was identified. Moreover, duction of p53-dependent apoptosis (16, 24). In response to the stability of SAP30L was dependent on the zinc module cellular stress by UV irradiation or hydrogen peroxide, ING2 since mutations in its zinc-coordinating residues led to rapid associates with chromatin through a PI(5)P-mediated mecha- degradation of the protein. We have previously observed that nism (23). Initially, the PHD domain of ING2 was reported to N-terminally truncated SAP30L is poorly expressed in tran- be sufficient for PI binding, but later, the PBR motif was sient transfections, but this could be overcome by using demonstrated to be both necessary and sufficient on its own MG132, a proteasome inhibitor, and now this can be explained (16, 24). Even though the PBRs of Pf1 and ING2 were deemed by the loss of the stabilizing zinc-dependent module in the N critical for the binding activity and specificity, the preceding terminus (57). Zinc-binding domains are usually relatively zinc-binding PHD domain contributed some specificity to the short, i.e., 20 to 30 residues, and the spacing of 35 residues interaction with PIs (24). SAP30 and SAP30L have a number between the C2 and CH coordinating residues in SAP30L is of similarities with the PI-binding proteins Pf1 and ING. First, unusually long. There is, however, a precedent for a large the domain architecture of SAP30/SAP30L resembles that of zinc-binding module, since THAP domains, which are con- Pf1, with a zinc-binding element followed by a basic PI-binding served zinc-dependent modules capable of sequence-specific module in both cases. Second, like SAP30/SAP30L, Pf1 and DNA binding, have a loop of 35 to 53 residues in the middle of ING2 are part of the Sin3A complex. Third, all three proteins the zinc-binding motif (8). The THAP domain, however, con- are nuclear and bind monophosphorylated PIs, albeit with tains other conserved elements in addition to the C2CH mod- different preferences. Fourth, in the cases of SAP30/SAP30L ule, making it distinct from the zinc-binding motif in SAP30L. and ING2, the subcellular localization and chromatin associa- The sequence-independent nature of the DNA binding rules tion are modified by PI binding. However, in the cases of out a sequence-specific targeting role for SAP30 and SAP30L SAP30/SAP30L (but not in ING2), PI binding competes with and suggests a more general role in anchoring to nucleosomal/ DNA binding in vitro so that an increase in the concentration linker DNA. In addition, we demonstrated that this DNA of monophosphorylated PIs causes SAP30L to detach from binding results in strong bending of the DNA. Classical exam- DNA. Furthermore, an increase in the concentration of nu- ples of proteins that bind and bend DNA in a sequence-inde- clear PIs elicited with hydrogen peroxide leads to reduced pendent manner are the HMG proteins (45), which interact repression activity and cytoplasmic relocalization of SAP30L. transiently with DNA. They are thought to antagonize histone Although we note that these results are preliminary and mostly H1 binding by competing for the same chromatin sites. Gen- based on in vitro experiments, they suggest the intriguing pos- erally, they are thought to open up chromatin, although some sibility that changes in the concentration of nuclear monophos- HMG proteins may also compact chromatin (43). We find phorylated PIs may regulate transcriptional repression through interesting parallels between the HMG proteins and SAP30/ SAP30/SAP30L in vivo. The site of PI binding was mapped to SAP30L. Both are small and localized in the nucleus. Their a region containing a motif previously shown to act as an NLS, domain structures are also similar, as both contain an N-ter- and this motif is necessary for the PI-binding activity. However, minal DNA-binding domain followed by an acidic region which the adjacent zinc-coordinating module also contributes to this contributes to histone interactions. This could imply functional interaction, a result that is in agreement with studies of other similarity as well, and it seems likely that SAP30/SAP30L have proteins (47). The binding interface may reside on one side of roles in stabilizing the multiprotein complex on its target, in- the loop region of the zinc module, in a region which contains creasing the availability of enzymatic targets to the complex, or a stretch of basic residues. The specificity of PI binding is partly promoting the recruitment of interacting proteins, the cumu- determined by the amino acid residue composition of the bind- lative effect being increased repression activity. ing motif, since replacing the NLS motif with another basic Perhaps one of the most intriguing features of SAP30L and motif (NoLS motif) in SAP30L led to changes in PI binding. SAP30 is the presence of a PI-binding site. PIs are known to Differences in binding specificity between different proteins function in nuclear signaling, and local changes in PI concen- are also evident. SAP30/SAP30L prefer PI(5)P over PI(3)P/ trations are sensed by proteins with specific PI-binding do- PI(4)P, whereas Pf1 prefers PI(3)P, with some binding activity mains, such as PH, ENTH, FYVE, and PHOX domains and toward PI(3,5)P species (24). lysine/arginine-rich patches (34, 44). A number of PI kinases We propose a model in which SAP30L/SAP30 are actively and phosphatases translocate to the nucleus upon activation, involved in multiple protein-protein and protein-DNA inter- and many PI species have been shown to be intranuclear (10, actions that modulate transcriptional repression. The domain 15). Intervention of chromatin biology by signaling lipids is not structures of SAP30L/30 and the proposed model are depicted unprecedented, since ATP-dependent chromatin-remodeling in Fig. 7B and C, respectively. Briefly, we suggest that the complexes, such as NURF, ISW2, INO80, and SWI/SNF, are DNA-binding activity plays a role in anchoring the Sin3A com- 354 VIIRI ET AL. MOL.CELL.BIOL.

FIG. 7. Domain structure and a proposed mode of action for SAP30L and SAP30. (A) Schematic representation of the N-terminal zinc- coordinating motif of SAP30L. (B) Various domains of SAP30L identified in this and other studies (35, 57). Zn, zinc-coordinating motif; DNAbd, DNA-binding domain; PIPbd, PIP binding domain; protein bd, protein binding domain; acidic region, a central region contributing to histone binding. (C) Proposed model. (Panel 1) When the histones are acetylated, the DNA is loosely packed and therefore accessible to RNA polymerase

II. (Panel 2) A sequence-specific transcriptional repressor (TF) recruits the Sin3A complex to its target promoter. SAP30 or SAP30L stabilizes the complex through interactions with DNA and histones 2A/2B. The interaction of SAP30/SAP30L with DNA induces bending of the DNA, as a result of which the nucleosomes are more accessible to HDAC enzymes, and the repressome is fully formed. (Panel 3) Nuclear PIPs (PtdInsPs) interact with the N-terminal domain of SAP30/SAP30L, displacing the DNA, which leads to relocalization of SAP30/SAP30L to the cytoplasm.

plex to nucleosomal and/or linker DNA in chromatin and that REFERENCES this binding is further strengthened by the interaction with core 1. Alland, L., R. Muhle, H. Hou, Jr., J. Potes, L. Chin, N. Schreiber-Agus, and histone 2A/2B dimers. One consequence of DNA binding is R. A. DePinho. 1997. Role for N-CoR and histone deacetylase in Sin3- mediated transcriptional repression. Nature 387:49–55. bending of the DNA, and we envision that this leads to en- 2. Bailey, T. L., N. Williams, C. Misleh, and W. W. Li. 2006. MEME: discov- hanced accessibility of nucleosomes and histone tails to ering and analyzing DNA and protein sequence motifs. Nucleic Acids Res. deacetylating enzymes. Moreover, our results provide new ev- 34:W369–W373. 3. Berger, M. F., A. A. Philippakis, A. M. Qureshi, F. S. He, P. W. Estep III, and idence for the regulatory role played by nuclear PIs in tran- M. L. Bulyk. 2006. Compact, universal DNA microarrays to comprehensively scriptional repression and relocalization of nuclear proteins. determine transcription-factor binding site specificities. Nat. Biotechnol. 24: 1429–1435. 4. Bernues, J., E. Espel, and E. Querol. 1986. Identification of the core-histone- binding domains of HMG1 and HMG2. Biochim. Biophys. Acta 866:242– ACKNOWLEDGMENTS 251. 5. Brehm, A., E. A. Miska, D. J. McCance, J. L. Reid, A. J. Bannister, and T. We thank Jorma Kulmala and Ritva Romppanen for technical as- Kouzarides. 1998. Retinoblastoma protein recruits histone deacetylase to sistance, Mike Berger for advice and assistance with PBM experiments repress transcription. Nature 391:597–601. and data analysis, and Olli Silvennoinen for comments on the manu- 6. Brown, R. S. 2005. Zinc finger proteins: getting a grip on RNA. Curr. Opin. Struct. Biol. 15:94–98. script. We are grateful to P. Peterson (Tartu, Estonia) and D. Ayer 6a.Bulyk, M. L., X. H. Huang, Y. Choo, and G. M. Church. 2001. Exploring the (Salt Lake City, UT) for H2B-GFP and 14D promoter plasmids, re- DNA-binding specificities of zinc fingers with DNA microarrays. Proc. Natl. spectively, and to Anne Hyva¨rinen (Tampere, Finland) for the EMSA Acad. Sci. USA 98:7158–7163. probe used in this work. 7. Carballo, M., P. Puigdomenech, and J. Palau. 1983. DNA and histone H1 This work was supported by the Academy of Finland Research interact with different domains of HMG 1 and 2 proteins. EMBO J. 2:1759– Council for Health (funding decision numbers 115260 and 201361) and 1764. for Natural Sciences and Engineering (108533), the Foundation for 8. Clouaire, T., M. Roussigne, V. Ecochard, C. Mathe, F. Amalric, and J. P. Paediatric Research in Finland, the Competitive Research Funding of Girard. 2005. The THAP domain of THAP1 is a large C2CH module with the Pirkanmaa Hospital District (EVO), the Nona and Kullervo Va¨re zinc-dependent sequence-specific DNA-binding activity. Proc. Natl. Acad. Sci. USA 102:6907–6912. Foundation, the Pa¨ivikki and Sakari Sohlberg Foundation, and grant 9. David, G., G. M. Turner, Y. Yao, A. Protopopov, and R. A. DePinho. 2003. R01 HG003985 from NIH/NHGRI to M.L.B. T.S. was supported in mSin3-associated protein, mSds3, is essential for pericentric heterochroma- part by a U.S. National Science Foundation Postdoctoral Research tin formation and chromosome segregation in mammalian cells. Genes Dev. Fellowship in Biological Informatics. 17:2396–2405. VOL. 29, 2009 SAP30L AND SAP30 BIND AND BEND DNA 355

10. Deleris, P., D. Bacqueville, S. Gayral, L. Carrez, J. P. Salles, B. Perret, and 32. Le May, N., Z. Mansuroglu, P. Leger, T. Josse, G. Blot, A. Billecocq, R. Flick, M. Breton-Douillon. 2003. SHIP-2 and PTEN are expressed and active in Y. Jacob, E. Bonnefoy, and M. Bouloy. 2008. A SAP30 complex inhibits vascular smooth muscle cell nuclei, but only SHIP-2 is associated with nu- IFN-beta expression in Rift Valley fever virus infected cells. PLoS Pathog. clear speckles. J. Biol. Chem. 278:38884–38891. 4:e13. 11. Emini, E. A., J. V. Hughes, D. S. Perlow, and J. Boger. 1985. Induction of 33. Lee, M. S., G. P. Gippert, K. V. Soman, D. A. Case, and P. E. Wright. 1989. hepatitis A virus-neutralizing antibody by a virus-specific synthetic peptide. Three-dimensional solution structure of a single zinc finger DNA-binding J. Virol. 55:836–839. domain. Science 245:635–637. 12. Fabris, D., Y. Hathout, and C. Fenselau. 1999. Investigation of zinc chelation 34. Lemmon, M. A. 2003. Phosphoinositide recognition domains. Traffic 4:201– in zinc-finger arrays by electrospray mass spectrometry. Inorg. Chem. 38: 213. 1322–1325. 35. Lindfors, K., K. M. Viiri, M. Niittynen, T. Y. Heinonen, M. Maki, and H. 13. Fleischer, T. C., U. J. Yun, and D. E. Ayer. 2003. Identification and charac- Kainulainen. 2003. TGF-beta induces the expression of SAP30L, a novel terization of three new components of the mSin3A corepressor complex. nuclear protein. BMC Genomics 4:53. Mol. Cell. Biol. 23:3456–3467. 36. Loewith, R., J. S. Smith, M. Meijer, T. J. Williams, N. Bachman, J. D. Boeke, 14. Gamsjaeger, R., C. K. Liew, F. E. Loughlin, M. Crossley, and J. P. Mackay. and D. Young. 2001. Pho23 is associated with the Rpd3 histone deacetylase 2007. Sticky fingers: zinc-fingers as protein-recognition motifs. Trends Bio- and is required for its normal function in regulation of gene expression and chem. Sci. 32:63–70. silencing in Saccharomyces cerevisiae. J. Biol. Chem. 276:24068–24074. 15. Gozani, O., S. J. Field, C. G. Ferguson, M. Ewalt, C. Mahlke, L. C. Cantley, 37. Macfarlan, T., S. Kutney, B. Altman, R. Montross, J. Yu, and D. Chakra- G. D. Prestwich, and J. Yuan. 2005. Modification of protein sub-nuclear varti. 2005. Human THAP7 is a chromatin-associated, histone tail-binding localization by synthetic phosphoinositides: evidence for nuclear phospho- protein that represses transcription via recruitment of HDAC3 and nuclear inositide signaling mechanisms. Adv. Enzyme Regul. 45:171–185. hormone receptor corepressor. J. Biol. Chem. 280:7346–7358. 16. Gozani, O., P. Karuman, D. R. Jones, D. Ivanov, J. Cha, A. A. Lugovskoy, 38. McGuinness, B. E., T. Hirota, N. R. Kudo, J. M. Peters, and K. Nasmyth. C. L. Baird, H. Zhu, S. J. Field, S. L. Lessnick, J. Villasenor, B. Mehrotra, 2005. Shugoshin prevents dissociation of cohesin from centromeres during J. Chen, V. R. Rao, J. S. Brugge, C. G. Ferguson, B. Payrastre, D. G. Myszka, mitosis in vertebrate cells. PLoS Biol. 3:e86. L. C. Cantley, G. Wagner, N. Divecha, G. D. Prestwich, and J. Yuan. 2003. 39. Me´ndez, J., and B. Stillman. 2000. Chromatin association of human origin The PHD finger of the chromatin-associated protein ING2 functions as a recognition complex, Cdc6, and minichromosome maintenance proteins dur- nuclear phosphoinositide receptor. Cell 114:99–111. ing the cell cycle: assembly of prereplication complexes in late mitosis. Mol. 17. Hassig, C. A., T. C. Fleischer, A. N. Billin, S. L. Schreiber, and D. E. Ayer. Cell. Biol. 20:8602–8612. 1997. Histone deacetylase activity is required for full transcriptional repres- 40. Mendiratta, G., P. R. Eriksson, C. H. Shen, and D. J. Clark. 2006. The sion by mSin3A. Cell 89:341–347. DNA-binding domain of the yeast Spt10p activator includes a zinc finger that 18. Heinzel, T., R. M. Lavinsky, T. M. Mullen, M. Soderstrom, C. D. Laherty, J. is homologous to foamy virus integrase. J. Biol. Chem. 281:7040–7048. Torchia, W. M. Yang, G. Brard, S. D. Ngo, J. R. Davie, E. Seto, R. N. 41. Meskauskas, A., J. L. Baxter, E. A. Carr, J. Yasenchak, J. E. Gallagher, S. J. Eisenman, D. W. Rose, C. K. Glass, and M. G. Rosenfeld. 1997. A complex Baserga, and J. D. Dinman. 2003. Delayed rRNA processing results in containing N-CoR, mSin3 and histone deacetylase mediates transcriptional significant ribosome biogenesis and functional defects. Mol. Cell. Biol. 23: repression. Nature 387:43–48. 1602–1613. 19. Hock, R., T. Furusawa, T. Ueda, and M. Bustin. 2007. HMG chromosomal 41a.Mukherjee, S., M. F. Berger, G. Jona, X. S. Wang, D. Muzzey, M. Snyder, proteins in development and disease. Trends Cell Biol. 17:72–79. R. A. Young, and M. L. Bulyk. 2004. Rapid analysis of the DNA binding 20. Hsieh, J. J., S. Zhou, L. Chen, D. B. Young, and S. D. Hayward. 1999. CIR, specificities of transcription factors with DNA microarrays. Nat. Genet. a corepressor linking the DNA binding factor CBF1 to the histone deacety- 36:1331–1339. lase complex. Proc. Natl. Acad. Sci. USA 96:23–28. 42. Nagy, L., H. Y. Kao, D. Chakravarti, R. J. Lin, C. A. Hassig, D. E. Ayer, S. L. 21. Huang, N. E., C. H. Lin, Y. S. Lin, and W. C. Yu. 2003. Modulation of YY1 Schreiber, and R. M. Evans. 1997. Nuclear receptor repression mediated by activity by SAP30. Biochem. Biophys. Res. Commun. 306:267–275. a complex containing SMRT, mSin3A, and histone deacetylase. Cell 89:373– Hyvarinen, A. K., J. L. Pohjoismaki, A. Reyes, S. Wanrooij, T. Yasukawa, 22. 380. P. J. Karhunen, J. N. Spelbrink, I. J. Holt, and H. T. Jacobs. 2007. The 43. Narita, M., V. Krizhanovsky, S. Nunez, A. Chicas, S. A. Hearn, M. P. Myers, mitochondrial transcription termination factor mTERF modulates replica- and S. W. Lowe. 2006. A novel role for high-mobility group a proteins in tion pausing in human mitochondrial DNA. Nucleic Acids Res. 35:6458– cellular senescence and heterochromatin formation. Cell 126:503–514. 6474. 44. Overduin, M., M. L. Cheever, and T. G. Kutateladze. 2001. Signaling with 23. Jones, D. R., Y. Bultsma, W. J. Keune, J. R. Halstead, D. Elouarrat, S. phosphoinositides: better than binary. Mol. Interv. 1:150–159. Mohammed, A. J. Heck, C. S. D’Santos, and N. Divecha. 2006. Nuclear Paull, T. T., M. J. Haykinson, and R. C. Johnson. PtdIns5P as a transducer of stress signaling: an in vivo role for PIP4Kbeta. 45. 1993. The nonspecific Mol. Cell 23:685–695. DNA-binding and -bending proteins HMG1 and HMG2 promote the as- sembly of complex nucleoprotein structures. Genes Dev. 7:1521–1534. 24. Kaadige, M. R., and D. E. Ayer. 2006. The polybasic region that follows the plant homeodomain zinc finger 1 of Pf1 is necessary and sufficient for specific 46. Pitkanen, J., A. Rebane, J. Rowell, A. Murumagi, P. Strobel, K. Moll, M. phosphoinositide binding. J. Biol. Chem. 281:28831–28836. Saare, J. Heikkila, V. Doucas, A. Marx, and P. Peterson. 2005. Cooperative 25. Korkeamaki, H., K. Viiri, M. K. Kukkonen, M. Maki, and O. Lohi. 2008. activation of transcription by autoimmune regulator AIRE and CBP. Bio- Alternative mRNA splicing of SAP30L regulates its transcriptional repres- chem. Biophys. Res. Commun. 333:944–953. sion activity. FEBS Lett. 582:379–384. 47. Sankaran, V. G., D. E. Klein, M. M. Sachdeva, and M. A. Lemmon. 2001. 26. Krithivas, A., D. B. Young, G. Liao, D. Greene, and S. D. Hayward. 2000. High-affinity binding of a FYVE domain to phosphatidylinositol 3-phosphate Human herpesvirus 8 LANA interacts with proteins of the mSin3 corepres- requires intact phospholipid but not FYVE domain oligomerization. Bio- sor complex and negatively regulates Epstein-Barr virus gene expression in chemistry 40:8581–8587. dually infected PEL cells. J. Virol. 74:9637–9645. 48. Schwabe, J. W., and A. Klug. 1994. Zinc mining for protein domains. Nat. 27. Kuzmichev, A., Y. Zhang, H. Erdjument-Bromage, P. Tempst, and D. Rein- Struct. Biol. 1:345–349. berg. 2002. Role of the Sin3-histone deacetylase complex in growth regula- 49. Schwerk, C., J. Prasad, K. Degenhardt, H. Erdjument-Bromage, E. White, P. tion by the candidate tumor suppressor p33ING1. Mol. Cell. Biol. 22:835–848. Tempst, V. J. Kidd, J. L. Manley, J. M. Lahti, and D. Reinberg. 2003. ASAP, 28. Laherty, C. D., A. N. Billin, R. M. Lavinsky, G. S. Yochum, A. C. Bush, J. M. a novel protein complex involved in RNA processing and apoptosis. Mol. Sun, T. M. Mullen, J. R. Davie, D. W. Rose, C. K. Glass, M. G. Rosenfeld, Cell. Biol. 23:2981–2990. D. E. Ayer, and R. N. Eisenman. 1998. SAP30, a component of the mSin3 50. Shen, X., H. Xiao, R. Ranallo, W. H. Wu, and C. Wu. 2003. Modulation of corepressor complex involved in N-CoR-mediated repression by specific ATP-dependent chromatin-remodeling complexes by inositol polyphos- transcription factors. Mol. Cell 2:33–42. phates. Science 299:112–114. 29. Laherty, C. D., W. M. Yang, J. M. Sun, J. R. Davie, E. Seto, and R. N. 51. Shiio, Y., D. W. Rose, R. Aur, S. Donohoe, R. Aebersold, and R. N. Eisenman. Eisenman. 1997. Histone deacetylases associated with the mSin3 corepressor 2006. Identification and characterization of SAP25, a novel component of mediate mad transcriptional repression. Cell 89:349–356. the mSin3 corepressor complex. Mol. Cell. Biol. 26:1386–1397. 30. Lai, A., B. K. Kennedy, D. A. Barbie, N. R. Bertos, X. J. Yang, M. C. 52. Silverstein, R. A., and K. Ekwall. 2005. Sin3: a flexible regulator of global Theberge, S. C. Tsai, E. Seto, Y. Zhang, A. Kuzmichev, W. S. Lane, D. gene expression and genome stability. Curr. Genet. 47:1–17. Reinberg, E. Harlow, and P. E. Branton. 2001. RBP1 recruits the mSIN3- 53. Skowyra, D., M. Zeremski, N. Neznanov, M. Li, Y. Choi, M. Uesugi, C. A. histone deacetylase complex to the pocket of retinoblastoma tumor suppres- Hauser, W. Gu, A. V. Gudkov, and J. Qin. 2001. Differential association of sor family proteins found in limited discrete regions of the nucleus at growth products of alternative transcripts of the candidate tumor suppressor ING1 arrest. Mol. Cell. Biol. 21:2918–2932. with the mSin3/HDAC1 transcriptional corepressor complex. J. Biol. Chem. 31. Lechner, T., M. J. Carrozza, Y. Yu, P. A. Grant, A. Eberharter, D. Vannier, 276:8734–8739. G. Brosch, D. J. Stillman, D. Shore, and J. L. Workman. 2000. Sds3 (sup- 54. Smith, J. S., E. Caputo, and J. D. Boeke. 1999. A genetic screen for ribo- pressor of defective silencing 3) is an integral component of the yeast somal DNA silencing defects identifies multiple DNA replication and chro- Sin3[middle dot]Rpd3 histone deacetylase complex and is required for his- matin-modulating factors. Mol. Cell. Biol. 19:3184–3197. tone deacetylase activity. J. Biol. Chem. 275:40961–40966. 55. Sun, Z. W., and M. Hampsey. 1999. A general requirement for the Sin3- 356 VIIRI ET AL. MOL.CELL.BIOL.

Rpd3 histone deacetylase complex in regulating silencing in Saccharomyces from bacteria and yeast to mice and men. Nat. Rev. Mol. Cell Biol. 9:206– cerevisiae. Genetics 152:921–932. 218. 56. Verreault, A., P. D. Kaufman, R. Kobayashi, and B. Stillman. 1998. Nucleo- 60. Zhang, Y., R. Iratni, H. Erdjument-Bromage, P. Tempst, and D. Reinberg. somal DNA regulates the core-histone-binding subunit of the human Hat1 1997. Histone deacetylases and SAP18, a novel polypeptide, are components acetyltransferase. Curr. Biol. 8:96–108. of a human Sin3 complex. Cell 89:357–364. 57. Viiri, K. M., H. Korkeamaki, M. K. Kukkonen, L. K. Nieminen, K. Lindfors, 61. Zhang, Y., Z. W. Sun, R. Iratni, H. Erdjument-Bromage, P. Tempst, M. P. Peterson, M. Maki, H. Kainulainen, and O. Lohi. 2006. SAP30L interacts Hampsey, and D. Reinberg. 1998. SAP30, a novel protein conserved between with members of the Sin3A corepressor complex and targets Sin3A to the nucleolus. Nucleic Acids Res. 34:3288–3298. human and yeast, is a component of a histone deacetylase complex. Mol. Cell 58. Yang, W. M., C. Inouye, Y. Zeng, D. Bearss, and E. Seto. 1996. Transcrip- 1:1021–1031. tional repression by YY1 is mediated by interaction with a mammalian 62. Zhao, K., W. Wang, O. J. Rando, Y. Xue, K. Swiderek, A. Kuo, and G. R. homolog of the yeast global regulator RPD3. Proc. Natl. Acad. Sci. USA Crabtree. 1998. Rapid and phosphoinositol-dependent binding of the SWI/ 93:12845–12850. SNF-like BAF complex to chromatin after T lymphocyte receptor signaling. 59. Yang, X. J., and E. Seto. 2008. The Rpd3/Hda1 family of lysine deacetylases: Cell 95:625–636. GST-SAP30L 1-92 GST-SAP30 1-131 GST (negative control) GST-Cbf1 (positive control)

Z-transformed 8-mer Median intensities

Supplemental figure 1 A

Supplemental figure 2A B

Supplemental figure 2B C 20mM Ca2+ Mg2+ Zn2+ 1,10-Phenanthroline 60mM: GST-SAP30L 1-92:

bp

1500

1000 Zinc-dependent and -specific 750 protein-DNA complexes

500

250

Supplemental figure 2C bp

200 200 150 100

50

Supplemental figure 3 A

isolated nucleosomes histones B Trypsin:

H3 H2A H2B anti-H2B H4 kDa

Coomassie WB DNA precipitation bp 75 800 600 50 tri- 37 400 di- 25

200 mono- nucleosomes

C

Supplemental figure 4