Insight into estrogen action in breast cancer via the study of a … · Andrew Redfern, Lisa Stuart, Andrew Barker, Louisa MacDonald, Dominic Voon and Keith Giles, who between them

Insight into estrogen action in breast cancer via

the study of a novel nuclear receptor corepressor:

SLIRP.

Esme Claire Hatchell

BSc (Honours) Medical Genetics & Zoology, UWA

This thesis is presented for the degree of Doctorate of Philosophy (PhD)

University Department of Medicine and Pharmacology, Faculty of Science,

University of Western Australia

Date of Final Submission: June 2008

E. Hatchell Abstract

i

Declaration

This thesis “Insight into estrogen action in breast cancer via the study of a novelnuclear receptor corepressor: SLIRP” has been completed for the Doctorate ofPhilosophy degree in Medicine at the University of Western Australia. The work Ipresent is my own work and all advice and assistance received has been appropriatelyacknowledged. This experimental research was performed under the supervision ofProfessor Peter J Leedman. The data presented here is original, and has not beensubmitted in any other form to this or any other university.

Several aspects of the work presented here have together been published as aMolecular Cell Journal article on the 9th of June 2006. The author of this thesis beganthe project and identified SLIRP (Patent Number: WO/2007/009194), performed allbasic analysis of SLIRP, all REMSA and UVXL experimentation, made the plasmidconstructs, grew and purified protein, performed all the transient transfectionexperiments, assessed SLIRP’s role in HeLa and hormone-dependent cancer cell-lines,played a significant role in the intellectual and experimental design of the projectduring the period of her PhD and co-wrote the journal article; totalling 50 % of thework presented in the Molecular Cell article.

The contributions of the named authors to the journal article are as follows: ShaneColley performed the mitochondrial localisation, Northern and ChIP experiments andco-authored the journal article (totalling 25 % of the work towards the Molecular Cellarticle). Together, Dianne Beveridge and Mike Epis performed the ChIP and IP-RT-PCR analyses (7 %). Lisa Stuart performed the tissue microarray analysis of SLIRP inbreast cancer (1 %). Keith Giles performed westerns assessing for SLIRP antibodyleakage and SKIP expression (1 %). Andrew Redfern contributed to the design of thetransfection experiments (1 %). Lauren Miles contributed to supplementary data of theMolecular cell article, but it is not presented in this thesis. Andrew Barker helped withthe intellectual design of the STR7 hot probes (1 %). Louisa MacDonald, Peter Arthurand James Lui contributed to data not presented in this thesis. Jackie and MatthewWilce advised on where we should make mutations in order to reduce SLIRP bindingto SRA in REMSA (1 %). Rainer Lanz and Bert O’Malley are our internationalcollaborators who have had intellectual input to the project from its inception (3 %).Peter Leedman is my Supervisor and has had complete supervisory control over, andintellectual input into, this project, and he co-wrote the paper (10 %).

All appropriate acknowledgments have been made throughout this thesis during thedetailing of methods and experiments in each chapter. A copy of this journal article hasbeen inserted into Appendix III of this thesis with permission from Elsevier.

PhD Candidate Co-ordinating SupervisorEsme Hatchell Peter LeedmanJune 2008 June 2008


ii

Abstract

Breast cancer is the cause of significant suffering and death in our community. It is

now estimated that the risk of developing breast cancer for an Australian woman

before the age of 85 is 1 in 8, with this risk rising for unknown reasons. While

mortality rates from breast cancer are falling due to increased awareness and early

detection, few new treatments have been developed from an advanced understanding of

the molecular basis of the disease.

From decades of scientific research it is clear that estrogen (E2) has a large role to play

in breast cancer. However, the basic mechanism behind E2 action in breast cancer

remains unclear. E2 plays a fundamental role in breast cancer cell proliferation and is

highly expressed in breast cancers, thus, it is important to understand both E2 and its

receptor, the estrogen receptor (ER). The ER is a member of the nuclear receptor (NR)

superfamily. The NR superfamily consists of a large group of proteins which regulate a

large number of homeostatic proteins together with regulator proteins termed

coregulators and corepressors.

SRA (steroid receptor RNA activator) is the only known RNA coactivator and

augments transactivation by NRs. SRA has been demonstrated to play an importantrole in mediating E2 action (Lanz et al., 1999; Lanz et al., 2003) and its expression is

aberrant in many human breast tumors, suggesting a potential role in breasttumorigenesis (Murphy et al., 2000). Despite evidence that an alternative splice variant

of SRA exists as a protein (Chooniedass-Kothari et al., 2004), it has been conclusivelyshown that SRA can function as an RNA transcript to coactivate NR transcription

(Lanz et al., 1999; Lanz et al., 2002; Lanz et al., 2003). The precise mechanism by

which SRA augments ER activity remains unknown. However, it is currentlyhypothesized that SRA acts as an RNA scaffold for other coregulators at the

transcription initiation site.

Several SRA stem loops have been identified as important for SRA function, including

structure (STR) 1, 5 and 7 (Lanz et al., 2002; Zhao et al., 2007). Previously, I sought to

identify SRA-binding proteins using a specific stem-loop structure of SRA (STR7) that

was identified as both important for its coactivator function (Lanz et al., 2002) and also

as a target for proteins from breast cancer cell extracts (Hatchell, 2002). From a yeast


iii

III hybrid screen using STR7 as bait, I identified a novel protein which was named

SLIRP (Patent Number: WO/2007/009194): SRA stem-Loop Interacting RNA-binding

Protein (Hatchell, 2002; Hatchell et al., 2006). In this thesis I describe in detail the

characterization of SLIRP: a widely expressed small SRA-binding protein and arepressor of NR signaling in breast cancer.

SLIRP is a small protein that is expressed in normal and tumor tissues at the protein

and RNA level, contains an RNA recognition motif (RRM) and represses NR

transactivation in an SRA- and RRM-dependent manner. Importantly, SLIRP augments

the effect of tamoxifen, an anti-E2 therapeutic used in the management of patients with

ER+ breast cancer. SLIRP is recruited to endogenous promoters (pS2 and

metallothionein), the latter in an SRA-dependent manner.

SLIRP interacts with several other NR coregulators. It modulates association of both

SRC-1 (a potent NR coactivator) with SRA, and NCoR (a potent NR corepressor) with

the DNA. Detailed analysis of SLIRP via bioinformatics showed SLIRP colocalizes

with SKIP (on Chr14q24.3), another NR coregulator. Further transfection studies show

that SLIRP reduces SKIP-potentiated NR signaling. REMSA analysis of SHARP, a

RRM-containing corepressor recently found to interact with SRA, shows it also binds

STR7, and transfection data shows SHARP augments repression with SLIRP. This data

shows SLIRP plays a fundamental role in NR transcription control.

Immunohistochemistry analysis of human tumors found that SLIRP is readily detected

in primary human breast cancer. Assessing SLIRP’s function in hormone-dependent

cancer cells showed SLIRP is a potent repressor of E2 and androgen receptor (AR)

signaling in breast and prostate cancer. When SLIRP protein levels are reduced in

breast cancer cells, cell proliferation is increased, which is consistent with its action as

a NR corepressor.

The unexpected location of SLIRP predominantly to the mitochondria prompted our

examination of how SLIRP might regulate a mitochondrial gene linked with breastcancer, heat-shock protein 60 (HSP-60). Reduction of SLIRP leads to a reduction of

HSP60 levels in breast cancer whole cell lysates. This provided evidence that SLIRP

may have bifunctional roles: one in the nucleus and one in the mitochondria.


iv

This thesis demonstrates that SLIRP modulates NR transactivation, provides

mechanistic insight into interactions between SRA, SRC-1, HSP-60 and NCoR and

suggests that SLIRP may regulate mitochondrial function. These studies contribute

significantly to the growing field of NR biology, and contribute more specifically to the

elucidation of estrogen action in breast cancer. Furthermore, it lays a strong and

exciting foundation for further studies to evaluate SLIRP as a biomarker and potential

therapeutic target in hormone dependent cancers.

E. Hatchell Acknowledgments

v

Acknowledgments

Science is a collaborative effort; this thesis, and the published work that has arisen, isno exception. This research would not have been possible without the help of the

following people, their expertise, guidance, advice and support.

First and foremost, I would like to acknowledge and sincerely thank my supervisor

Professor Peter Leedman, to whom I am indebted and without whom this work wouldnever have happened. I look forward to many future collaborations when our research

paths meet again.

Special thanks must go to the Leedman Laboratory, especially Dianne Beveridge,

Andrew Redfern, Lisa Stuart, Andrew Barker, Louisa MacDonald, Dominic Voon andKeith Giles, who between them have taught and guided me throughout the years. Here

I would also like to acknowledge the efforts of all the lab members who were co-

authours of our Molecular Cell paper; together we have proven just how powerfulteam-work can be. Further to the names already mentioned above, these people

include: Shane Colley, Micheal Epis, Lauren Miles, Jemma Golding and RossMcCulloch.

Over the last 6 years we have held a long-standing and important collaboration withRainer Lanz and Bert O’Malley, from the Baylor College of Medicine in Texas, USA.

Their guidance and advice in many aspects of this work has been invaluable, and I amextremely grateful for their contributions.

Without the assistance of our many national and international collaborators much ofthis work would not have be possible. I would like to thank: Jennifer Byrne (Children’s

Hospital, Westmead, Sydney, Australia) for the human breast cancer cDNA library andrelevant clinical details, Xiatao Li (Baylor College of Medicine, USA) for technical

advice, Ronald M. Evans and Michael Downes (Salk Institute, USA) for the SHARP

clones, Gary Leong (University of Sydney, Australia) for the SKIP constructs, MarvinWickens (University of Wisconsin, USA) for the yeast three-hybrid system, George

Muscat (Institute for Molecular Bioscience, Queensland, Australia) for the PPARδ and

PPARE vectors and advice, Dennis Dowhan (Centre for Immunology and Cancer

E. Hatchell Acknowledgments

vi

Research, Queensland, Australia) for assistance with the ChIP assay and SKIP ab,Chris Glass (University of California, USA) for advice on the ChIP assay, Roger

Reddel (Westmead, Sydney, Australia) for the HMEC cells, Paul Rigby (University ofWA, Australia) for his assistance with the imaging studies and Evan Ingley (Western

Australian Institute for Medical Research (WAIMR), Australia) for his technical

assistance, Peter Aurthur and James Lui (UWA) for their help with mitochondrialproteins, Cecily Metcalf (Pathology, Royal Perth Hospital (RPH) Perth) for her

histological analysis of cancer tissues, and Jackie and Matthew Wilce (UWA) for theiradvice on structural biology.

This work was funded by National Health & Medical Research Council (NHMRC),National Breast Cancer Foundation (NCBF), Cancer Council of WA (CCWA),

WAIMR and Royal Perth Hospital Medical Research Foundation grants. I have beenextremely proud and grateful to be supported by a NHMRC Dora Lush Scholarship,

and am very thankful for the funding I received for national and international travel

from the following funding bodies: the Endocrine Society of Australia (ESA), theEndocrine Society (ENDO), the Australian Society for Medical Research (ASMR), the

Australian Society for Biochemistry and Molecular Biology (ASBMB), the Australiaand New Zealand Society for Cell and Developmental Biology (ANZSCDB), Royal

Perth Hospital (RPH), the Western Australian Institute for Medical Research

(WAIMR) and the University of Western Australia (UWA).

A special thanks must go to all my family and friends who have supported me over a

number of years, especially those friends in our laboratory: Viki Russell, MajhoobaSidiqi, Kerry Ramsay and Noel Chow. Here, a special acknowledgment is reserved for

Christin Down and Becky Webster; we travelled the PhD path together.

Lastly I would like to especially thank Helen and Liam for many years of love, support

and encouragement.

And, of course, Simon, my life-long partner. Thank you for believing in me.

E.Hatchell Abbreviations

vii

Abbreviations

a AdenosineA Alanineaa Amino acidAACR Australasian Association of Cancer Registriesab AntibodyACS American Cancer SocietyAF Activating functionAIB-1 Amplified in breast cancer-1AIHW Australian Institute for Health and WelfareANZSCDB Australia and New Zealand Society for Cell and

Developmental BiologyAP-1 Activating protein -1APS Ammonium persulphateAR Androgen receptorARE Androgen response elementARM Arginine rich motifASBMB Australian Society for Biochemistry and Molecular BiologyATCC American Type Culture CollectionATPase Adenosine triphosphotase catalysing enzyme3-AT 3-amino-triazolebp Base pairBSA Bovine serum albuminBRCA1 Breast cancer 1, early onsetC CytosineCASP-8 Caspase 8, apoptosis-related cytesine peptidaseCBP CREB-binding proteincDNA complimentary DNACEB Cytoplasmic extraction bufferChIP Chromatin immunopreciptationChr14ORF156 Chromosome 14 open reading frame 156cmv CytomegalovirusCoA CoactivatorCoAA Coactivator activatorCoAM Coactivator modulatorCoR Corepressorcpm Counts per minuteCREB cAMP response element binding-proteincRNA Complementary RNAC-terminal Carboxy-terminalDAKO DiaminobenxidineDBD DNA-binding domainDCIS Ductal carcinoma in situddH2O Distilled deionised waterDEPC Diethyl pyrocarbonateDex DexamethasoneDHT 4, 5 – dihydrotestosteroneDMSO Dimethyl sulphoxideDNA Deoxyribonucleic acidDNase Deoxyribonuclease


viii

dNTP deoxyribonucleotide triphosphateDP97 novel DEAD box helicase 97 kDaDSRM Double stranded RNA binding motifDRIP Vitamin D receptor-interacting proteinsDTT DithiothreitolE2 EstradiolE6AP E6-associated proteinECL Enhanced chemiluminescenceEDTA Ethylene diamine tetra-acetic acidEGF Epidermal growth factorEGFR Epidermal growth factor receptorEMSA Electromobility shift assayENDO the Endocrine SocietyER Estrogen receptorERE Estrogen response elementERRB2 Erythroblastic leukaemial viral oncogene homologue 2ESA Endocrine Society of AustraliaFCS Fetal calf serumFGFR2 Fibroblast growth factor receptor 2g GuanineGAPDH Glyceraldehyde-3-phosphate dehydrogenaseGAN Genbank accession numberGO Gene ontologyGFP Green fluorescent proteinGP General practitionerGST Glutathioine-S-transferaseGR Glucocorticoid receptorGrb-7 Growth factor receptor-bound protein-7GRIP2 Glucocorticoid receptor interacting protein-1HAT Histone acetylaseHDAC Histone deacetylaseHDAC-1 Histone deacetylase protein -1HEPES N-[2-hydroxyethyl]piperazine-N’-[2-ethane sulphonic acid]hER Human estrogen receptorHer2 Human epidermal growth factor receptor -2HG Human genomehnRNP Heterogeneous nuclear ribonucleoproteinhr hourHRE Hormone response elementHSP Heat-shock proteinHSP-60 Heat-shock protein 60H19 Housekeeping RNA 19HuR Hu protein RHuD Hu protein DIGF Insulin-like growth factorIGFR1 IGF type I receptorIL InterleukinIL-6 Interleukin-6IP ImmunoprecipitationIP-RT-PCR Immunoprecipitation reverse transcription PCRIPTG Isopropyl β-D-thio-galatopyrosanideIRE Iron response element


ix

kb KilobasekDa KilodaltonKH hnRNP K homologyL LeucineLARII Luciferase assay reagentLBD Ligand binding domainLCM Laboratory for Cancer MedicinelncRNA Long non coding RNALOH Loss of heterozygosityLuc LuciferaseMr Molecular weightMAPK Mitogen-activated protein kinaseMICoA MTA interacting coactivatormin MinutesmiRNA microRNAMMTV Mouse mammary tumour virusMOPS 3-[N-morpholino]propanesulfonic acidMR Mineralcorticoid receptormRNA Messenger ribonucleic acidmRNP Messenger ribonucloprotein complexMTA1 Metastasis-associated protein 1N Degenerate base symbol for any nucleotide in DNA or RNAns non-senseN-terminal(/us) Amino-terminalNAD+ Nicotinamide adenosine dinucleotide (non-reduced)NADH Nicotinamide adenosine dinucleotide (reduced)NCBI National Center for Biotechnology InformationNCoR Nuclear corepressor proteinncRNA Non-coding RNANFκB Nuclear factor kappa βNLS Nuclear localization signalNR Nuclear receptorsnt nucleotideORF Open reading framep300 Protein 300 (KDa)p53 Protein 53 (KDa)PABP Poly A binding proteinPACT PKR activating proteinPAGE Polyacrylamide gel electrophoresisPBS Phosphate-buffered salinepCAF p300/CREBBP-associated factorPCR Polymerase chain reactionPELP1 Proline-, glutamic acid-, leucine rich- protein -1PGC1α PPAR gamma coactivator 1 alphaPK Pyruvate kinasePKR Protein kinase RPLB Passive lysis bufferPMSF Phenylmethysulphonyl fluoridePol PolymerasePPAR Proliferated peroxisome activated receptorPR Progesterone receptorPUA Pseudouridine synthase and archaeosine transglycosylase


x

PUS Pseudouridine synthasePVDF Polyvinyl diflourideRAR Retinoic acid receptorRBP RNA-binding proteinRD Repression domainREMSA RNA electromobility shift assayRGG Arginine-glycine-glycineRNP RibonucleoproteinRNA Ribonucleic acidRNAse RibonucleaseRNAi RNA interferenceRPC RNA-protein complexRPH Royal Perth HospitalRPMI Roswell Park Memorial InstituteRRM RNA-recognition motifrRNA ribosomal RNART Reverse transcriptaseRTA Repressor of tamoxifen transcriptional activityRT-PCR Reverse transcription PCRRXR Retinoid X receptorS SerineSAGA Spt-Ada-Gch5-AcetyltransferaseSDM Site directed mutagenSDS Sodium dodecyl sulphatesec secondsSERM Selective estrogen receptor modifierSHARP SMRT/HDAC1 associated repressor proteinsiRNA Small interfering RNASKIP Ski-interacting proteinSL Stem loopSLIRP SRA stem-loop interacting RNA-binding proteinsmRNA Small ribonluceic acidSMRT Silencing mediator of retinoic acid and thyroid receptorSN supernatentSRA Steroid receptor RNA activatorSRA-Del SRA deletion splice variantSRC-1 Steroid receptor coactivator-1SRC-3 Steroid receptor coactivator-3STR StructureSV40 Simian virus 40Sxl Sex lethal protein genet ThymidineT3 Thyroid hormoneTAE Tris acetate EDTATAF TBP-associated factorsTam TamoxifenTATA a conserved promoter sequence: tataTBP TATA binding proteinTBS Tris buffered salineTBS-T Tris buffered saline – Tween 20TEMED TetramethylethyldiamineTFIID Transcription factor for RNA polymerase II D


xi

TR Thyroid hormone receptorTRAP Thyroid receptor associated proteintRNA Transfer RNAu Uracil/uradineUP1 Uuclear ribonuclear proteinUSA Upstream stimulatory activityUTP Uradine triphosphateUTR Untranslated regionUVC Ultraviolent CUVXL UV cross-linkingUWA University of Western AustraliaVit D(3) 1(α),25-dihydroxyvitamin D3, (1,25 D3)VDR Vitamin D receptorVDRE Vitamin D response elementWAIMR Western Australian Institute for Medical ResearchWT Wild-typeWTB Western Transfer Buffer


xii

E. Hatchell Table of contents

xiii

Table of contents

Declaration .................................................................................................................................... iAbstract ......................................................................................................................................... iiAcknowledgements.......................................................................................................................vAbbreviations ................................................................................................................................vii

Table of contents...........................................................................................................................xiii

Chapter 1: Introduction ..............................................................................................................11.1 Introduction ................................................................................................21.1.1 An introduction to cancer ...........................................................................21.1.2 Breast cancer ...............................................................................................31.2 The progression towards breast cancer ......................................................41.2.1 The genetic biology of cancer ....................................................................41.2.2 The biology of breast cancer ......................................................................51.2.3 Piecing together the evidence.....................................................................61.3 The nuclear receptor superfamily...............................................................71.3.1 Nuclear receptors ........................................................................................71.3.2 Nuclear receptor classes..............................................................................81.3.3 The steroid hormone receptor domains......................................................91.3.4 Nuclear receptor cross-talk within the mitochondrion ..............................101.4 Mechanisms of estrogen action ..................................................................101.4.1 Estrogen action............................................................................................101.4.2 Estrogens and receptors ..............................................................................111.5 The estrogen receptor..................................................................................121.5.1 The human estrogen receptors....................................................................121.5.2 Functional estrogen receptor domains........................................................131.5.3 Selective estrogen receptor modifiers (SERMs)........................................141.6 Coregulators ................................................................................................151.6.1 Controlling gene expression .......................................................................151.6.2 Nuclear receptor coactivators .....................................................................151.6.3 The TFIID complex: mechanisms of basal transcription ..........................161.6.4 Coregulator action on chromatin................................................................171.6.5 Nuclear receptor corepressors ....................................................................171.6.6 N-CoR..........................................................................................................181.6.7 ER coregulators...........................................................................................181.6.8 SKIP: a nuclear receptor coregulator .........................................................191.7 SRA .............................................................................................................211.7.1 SRA: an RNA coactivator ..........................................................................211.7.2 SRA in breast cancer...................................................................................231.7.3 SRA structure ..............................................................................................251.7.4 SRA stem loop 7 .........................................................................................241.8 RNA biology: a growing field....................................................................251.8.1 The changing role of RNA in biology........................................................251.9 RNA-binding proteins.................................................................................271.9.1 General ........................................................................................................271.9.2 Nuclear receptors and RNA-binding proteins............................................281.10 SRA RNA-binding proteins........................................................................291.10.1 SHARP ........................................................................................................291.10.2 SRA-interacting RNA-binding DEAD-box proteins.................................301.10.3 Pseudouridine synthase and SRA...............................................................311.11 The SRA project..........................................................................................311.11.1 My previous work on SRA.........................................................................311.11.2 Project hypotheses and aims.......................................................................321.11.3 Significance and outcomes of this study....................................................33Chapter 1 Figures ............................................................................................................35


xiv

Chapter 2: Materials and methods ............................................................................................ 492.1 Materials...................................................................................................... 502.1.1 Reagents ...................................................................................................... 502.1.2 General reagents ......................................................................................... 542.1.3 Consumables ............................................................................................... 552.1.4 Commercial kits.......................................................................................... 552.1.5 Equipment ................................................................................................... 552.1.6 Buffers and solutions .................................................................................. 562.2 Methods....................................................................................................... 572.2.1 Plasmid constructs ...................................................................................... 572.2.2 Bioinformatics ............................................................................................ 572.2.3 Polymerase chain reaction assays and primer design................................ 572.2.4 Cloning........................................................................................................ 582.2.5 Bacterial cell culture................................................................................... 582.2.6 GST fusion protein production................................................................... 592.2.7 RNA analysis .............................................................................................. 602.2.8 REMSA and UVXL experiments............................................................... 612.2.9 Tissue culture.............................................................................................. 632.2.10 Protein analysis ........................................................................................... 652.2.11 In vivo analysis of proteins......................................................................... 66Chapter 2 Figures ............................................................................................................ 68

Chapter 3: Basic characterisation of SLIRP ............................................................................ 713.1 Introduction................................................................................................. 723.1.1 Preface......................................................................................................... 723.1.2 SRA and the growing ncRNA field ........................................................... 723.1.3 RNA secondary structures.......................................................................... 733.1.4 RNA-binding proteins and RNA-binding domains................................... 743.1.5 RNA recognition motifs (RRMs)............................................................... 753.1.6 The identification of SLIRP ....................................................................... 763.1.7 Hypotheses and Aims ................................................................................. 773.2 Methods....................................................................................................... 783.2.1 Bioinformatics and sequence analysis ....................................................... 783.3.2 Tissue culture.............................................................................................. 783.2.3 Cytoplasmic and nuclear extracts............................................................... 783.2.4 Protein purification ..................................................................................... 793.2.5 Western blot ................................................................................................ 793.2.6 Radioisotope labelling of STR7................................................................. 793.2.7 REMSA and UVXL.................................................................................... 793.2.8 RT-PCR....................................................................................................... 803.2.9 IP-RT-PCR.................................................................................................. 803.2.10 Northerns..................................................................................................... 803.2.11 Immunohistochemistry ............................................................................... 803.3 Results ......................................................................................................... 813.3.1 SRA STR7 is a target for proteins in human breast cancer cells .............. 813.3.2 SLIRP: a SRA-binding protein .................................................................. 813.3.3 Bioinformatic analysis of SLIRP ............................................................... 813.3.4 SLIRP is expressed widely in human tissues and cancer cells ................. 833.3.5 Characterization of SLIRP’s interaction with SRA .................................. 843.4 Discussion ................................................................................................... 85Chapter 3 Figures ............................................................................................................ 88

Chapter 4: SLIRP: a predominantly mitochondrial nuclear receptor corepressor ........... 1014.1 Introduction................................................................................................. 1024.1.1 Preface......................................................................................................... 1024.1.2 Nuclear receptors ........................................................................................ 1024.1.3 Coregulators in the ER pathway ................................................................ 1034.1.4 Nuclear receptors in the mitochondria....................................................... 1054.1.5 NR coregulators and the mitochondria ...................................................... 1064.1.6 ncRNA-protein interactions ....................................................................... 1074.1.7 Hypotheses and Aims ................................................................................. 1084.2 Methods....................................................................................................... 109


xv

4.2.1 Tissue culture and transient transfections ..................................................1094.2.2 siRNA & RNAi technology........................................................................1094.2.3 SLIRP mutants ............................................................................................1094.2.4 REMSA .......................................................................................................1104.2.5 Confocal microscopy ..................................................................................1104.2.6 ChIP.............................................................................................................1104.2.7 IP-RT-PCR..................................................................................................1104.2.8 Immunohistochemistry ...............................................................................1114.3 Results .........................................................................................................1114.3.1 SLIRP represses SRA-mediated nuclear receptor coactivation ................1114.3.2 SLIRP modulates SHARP- and SKIP-mediated coregulation

of NR activity..............................................................................................1124.3.3 SLIRP function requires an intact RRM domain.......................................1134.3.4 SLIRP is recruited to endogenous NR target promoters ...........................1144.3.5 SLIRP is predominantly mitochondrial .....................................................1144.4 Discussion ...................................................................................................116Chapter 4 Figures ............................................................................................................119

Chapter 5: SLIRP and estrogen signaling in breast cancer ...................................................1315.1 Introduction .................................................................................................1315.1.1 Preface .........................................................................................................1315.1.2 Hormone-dependent cancers ......................................................................1315.1.3 Estrogen and cell growth in breast cancer .................................................1325.1.4 Mitochondrial proteins in breast cancer.....................................................1335.1.5 HSP-60 ........................................................................................................1345.1.6 Hypotheses and Aims .................................................................................1355.2 Methods .......................................................................................................1365.2.1 Tissue culture and transient transfections ..................................................1365.2.2 RNAi transfections......................................................................................1365.2.3 Western blot ................................................................................................1365.2.4 Cell titre.......................................................................................................1375.3 Results .........................................................................................................1375.3.1 SLIRP represses SRA-mediated NR coactivation in breast cancer

cells..............................................................................................................1375.3.2 Effect of SLIRP RNAi on estrogen signaling............................................1385.3.3 Reduction of SLIRP increases cell proliferation .......................................1385.3.4 SLIRP affects HSP-60 protein levels in breast cancer cells......................1385.4 Discussion ...................................................................................................139Chapter 5 Figures ............................................................................................................142

Chapter 6: Discussion & Conclusions........................................................................................1496.1 Introduction .................................................................................................1506.2 The emerging role for RNA and the mitochondria in breast cancer .........1506.3 The discovery of SRA represents a paradigm shift in the perception

of NR coregulators......................................................................................1506.4 SLIRP: a novel SRA-binding nuclear receptor corepressor......................1516.5 Complex interactions exist between SLIRP and other NR

coregulators .................................................................................................1536.6 SLIRP is a predominantly mitochondrial NR corepressor ........................1546.7 SLIRP in breast cancer ...............................................................................1566.8 The SRA story unfolds ...............................................................................1576.9 Final conclusion ..........................................................................................158

References......................................................................................................................................161

Appendix I: Buffers and Solutions.............................................................................................179

Appendix II: Prizes and Awards ................................................................................................193

Appendix III: Publications and Patents ....................................................................................195


xvi

E. Hatchell Chapter 1: Introduction

1

Chapter 1

Introduction


2

1.1 Introduction

1.1.1 An introduction to cancer

Cancer is a significant cause of suffering and death in our community. The most recent

Australian Cancer Council statistics states that approximately one in four Australian

females, and one in three Australian males will develop a malignant cancer prior to the

age of 75 (AIHW & AACR, 2004). An estimated 260,000 years of life will be lost each

year due to cancer deaths before the age of 75 (AIHW & AACR, 2004). Cancer is now

a leading cause of death in Australia, with over 36,000 people dying from the condition

each year (AIHW & AACR, 2004). This number accounts for one third of all deaths in

our community (Figure 1.1 A) and yet, despite such high statistics, the phenomenon of

cancer remains poorly understood (Gupta et al., 2005).

The term ‘cancer’ encompasses a wide range of diseases whereby an abnormal cell can

proliferate and spread out of control. This leads to a mass of cells termed a neoplasm, a

Greek word for ‘new formation’ (Jorde et al., 1995) or tumor. Benign tumors do not

spread throughout the body although they can interfere with healthy organs. Malignant

tumors are those that can invade nearby tissues and metastasise (spread) to other parts

of the body (AIHW & AACR, 2004). Cancer is a result of both genetic and

environmental influences, and is thought to be a multistep process (Hanahan &

Weinburg, 2000). There are numerous different cancers, some have a common cause

but a greater number produce similar symptoms (Hanahan & Weinburg, 2000). An

understanding of the biology of an individual neoplasm is essential in predicting

outcome and in providing effective treatment.

Over the past three decades there have been significant improvements in our

understanding of cancer biology which have led to earlier diagnosis and treatments,

such as chemotherapy and receptor-targeted therapies which have, in turn, significantly

reduced cancer mortality (Hortobagyi et al., 1999; AIHW & AACR, 2004; Eneman et

al., 2004; Vlastos & Verkooijen, 2007). However, while improvements have been

made in early diagnosis, locoregional treatments and our understanding of the

pathogenesis of cancer development, few new therapies directly resulting from basic

scientific research have been introduced into practice (Hortobagyi et al., 1999; Russo

and Russo, 2006). One of remaining gaps in our knowledge is the identification of the


3

key molecular players involved in gene expression (Lanz et al., 2003), which suggests

that such understanding may be the key to developing new therapies in the future.

Ultimately, the goal of cancer researchers is to understand what goes ‘wrong’ at the

level of gene expression which leads to the development of neoplasms and to develop

better systems for early detection of neoplasms. Such an understanding would allow for

rapid intervention and for the development of curative therapies for individuals

diagnosed with cancer.

1.1.2 Breast cancer

Breast cancer is the most common cancer in women and a major cause of morbidity

and mortality (Figure 1.1 B) (Sarkar et al., 2001). Worldwide, approximately 350,000

women die from breast cancer each year (Holst et al., 2007). In Australia alone, over

13,000 new cases are diagnosed each year which account for approximately 30 % of all

new cancers in women (AIHW & AACR, 2004; AIHW & NBCC, 2006). Metastatic

breast cancer is currently incurable (Al-Hajj et al., 2003) although if detected early

enough then the disease is treatable (Stockler et al., 2000). The life expectancy of an

Australian woman born between 2002 and 2004 is currently 83 years (AIHW &

NBCC, 2006). However, at present, the life-time risk for an Australian woman of

developing breast cancer before the age of 75 is approximately 1 in 11, (AIHW &

NBCC, 2006) and of developing breast cancer before the age of 85 is approximately 1

in 8 (AIHW & NBCC, 2006). Several categories of women are at a higher risk of

developing breast cancer. Women living in either an urbanised or high socio-economic

status area have a significantly higher risk of developing breast cancer when compared

to women living in either a rural or low socio-economic status area. The reasons for

this are still unknown and heavily debated (AIHW & NBCC, 2006).

Over 70 % of women who develop breast cancer will initially have an estrogen

receptor (ER) positive cancer on immunohistochemistry (IHC) analysis, which

signifies the presence of ERs and a high likelihood of response to anti-estrogen (E2)

(Green & Chambon, 1991, Clarke et al., 1997; Murphy et al., 2000) and aromatase

inhibitor treatment (Hind et al., 2007). Male breast cancer makes up 0.5 % of all breast

cancers (0.7 % of all male cancers), and over 80 % of male breast cancers are also

initially ER positive (ER+) (Dixon & Morrow, 1999). Research has shown estrogens,

acting via ER signaling pathways are pro-proliferative and play a fundamental role in


4

the development of breast cancer (Clarke et al., 1997; Hortobagyi et al., 1999; Murphy

et al., 2000; Hayashi et al., 2003). Taken together, this information is consistent with

the notion that the ER plays an important role in breast cancer by mediating the

mitogenic action of estrogens, however, there remains no clear understanding of the

mechanisms of how E2 causes cancer (Russo & Russo, 2006).

1.2 The progression towards breast cancer

1.2.1 The genetic biology of cancer

Traditionally, there are three major classes of cancer genes: those that normally inhibit

cell proliferation, termed tumor suppressor genes; those that normally activate

proliferation, termed oncogenes; and those involved in DNA repair (Jorde et al., 1995).

Knudson’s ‘Two Hit’ Model is applicable to tumor suppressor genes. These are genes

which require a loss of function mutation in not just one, but both, normal genes in the

cell. Some examples of tumor suppressor genes include Retinoblastoma 1, BRCA1,

BRCA2, p53, and APC (Jorde et al., 1995). Oncogenes, on the other hand, are quite

different from tumor suppressor genes. Oncogenes originate from cellular pro-

oncogenes that regulate normal cell growth. The mutation and overexpression of a

proto-oncogene can result in a transformation from regulated to unregulated growth

and differentiation (Lodish et al., 1995). Oncogenes include growth factors, receptors

for growth factors and various intracellular transducers, including RET, erb-B (the

epidermal growth factor receptor; EGFR), erb-A, H-RAS, N-myc, c-fos (Lodish et al.,

1995) and FGFR2 (fibroblast growth factor receptor 2) (Easton et al., 2007). The ER

acts as a growth promoter through its capacity to activate a variety of proliferative

genes, including oncogenes (Osborne et al., 2001). DNA-repair genes act via yet a

further different mechanism to produce cancer. In the normal cell, DNA must undergo

accurate replication from one generation to the next and this is achieved with the help

of repair mechanisms. When these mechanisms are defective, genomic instability

results, leading to a predisposition to tumorigenesis, best illustrated by some rare

genetic disorders including Bloom’s syndrome, ataxia telangiectasia and Werner’s

syndrome (Jorde et al., 1995).

Recent research has shown that even when a tumor cell escapes dysregulation by tumor

suppressor genes, proto-oncogenes and DNA-repair genes, it must still overcome the


5

predisposition towards senescence by activating the gene encoding telomerase (Ide,

2006). Human chromosomes contain TTAGGG repeats at their ends that provide

genomic stability, and are termed telomeres (Herbert et al., 2001). Telomerase is an

enzyme that replaces telomeric segments normally lost during cell division and allows

the cell to continue to multiply (Herbert et al., 2001). Telomerase is present in 85 % to

90 % of all human tumors, but not in normal cells, making it a useful marker of

malignancy (Elenitoba-Johnson, 2001). Throughout the progression to breast cancer,

E2 has a pro-proliferative effect which amplifies the signals produced by abnormalities

in either tumor suppressor genes, oncogenes, DNA-repair genes, telomerase or a

combination of these.

1.2.2 The biology of breast cancer

Normal breast epithelial cells undergo dramatic changes during a woman’s life, from

puberty through to the follicular and luteal phases of the menstrual cycle then

throughout menopause (Santen & Harvey, 1999). A coordinated interaction between

steroid hormones and growth factors regulates the proliferation and function of these

epithelial cells. Breast cancers can be divided into two groups, those dependent on and

those independent of hormones (Santen et al., 1990; Dixon & Morrow, 1999). In the

hormone-dependent group, E2 appears to play a dominant role in inducing cell

proliferation and growth leading to breast cancer (Santen & Harvey, 1999). Some

human breast tumors express high levels of growth factors and their receptors

(Mendelsohn & Baselga, 2000; Holst et al., 2007), for these hormone-independent

tumors growth factors such as the EGFR together with the erythroblastic leukaemial

viral oncogene homolouge 2 (ErbB2) can drive cancer growth.

E2 promotes the proliferation of breast cancer cells by enhancing the rate of cell

division via the stimulation of growth factors and by reducing the time available for

DNA repair (Santen & Harvey, 1999). Genome profiling studies have shown breast

cancers undergo a remarkable spectrum of genomic changes as cancer progression

occurs (Gray, 2001). Generally, breast cancer tumors evolve slowly and genomic

changes occur early (Waldman et al., 2000). Often, metastatic tumors are genetically

similar to the original tumors from which they arise (Gray, 2001). Recent

comprehensive genome profiling has shown regions of recurrent abnormalities, where

genetic mutations, translocations and epigenetic changes play a role in the development


6

of neoplasms, and are associated with increased breast cancer risk. One such example

is the gene CASP8 (Cox et al., 2007). However, there is still much to be learned to fully

understand the underlying mechanism of action of E2 in breast cancer, and this area is

the focus of intense investigation from multiple groups around the world.

1.2.3 Piecing together the evidence

Thirty years of laboratory research has established that E2 regulates the growth and

development of both normal and neoplastic human mammary tissue (Leygue et al.,

1999), and can directly cause breast cancer cells to proliferate (Murphy et al., 2000). It

was Lacassagne in 1936, who first proposed that breast cancer could be prevented via

an antagonist to E2 action (Jordan & Morrow, 1999). However, as no models of E2

action existed at that time, there was no way to develop an antagonist to it.

Until the middle of the 19th century, when an individual developed a clinically obvious

breast tumor, the predicted outcome was death usually within a 10 year period

(Haagensen, 1986). The operation for breast carcinoma was a simple crude removal of

the obvious tumor and some surrounding breast tissue, most often accompanied by

severe loss of blood and secondary infection (Haagensen, 1986). Little was understood

of the basic mechanisms behind cancer, and there was no available evidence to suggest

hormones may play a role in the development of breast cancer. In the mid 1890’s breast

cancer was considered a disease that developed in the breast, and, if left untreated,

would quickly spread from lymph node to lymph node leading to a ‘contiguous’

development of metastasis, which eventually spread to other organs in the body

(Fisher, 1999). Dr W.S. Halsted first articulated this concept, termed the Halsted

theory, and the Halsted radical mastectomy, whereby the entire breast organ and

axillary lymph nodes were removed, quickly became the advocated life saving

treatment (Webster, 1986).

In 1896 George Beatson published his findings demonstrating that the removal of

ovaries from a woman with advanced metastatic breast cancer could result in the

regression of the cancer, and increase survival (Beatson, 1896). However, it was over

fifty years later in 1962 when research showed that E2 (specifically estradiol) could

bind target tissues such as the uterus (Jensen & Jacobson, 1962). Furthermore, it was

demonstrated that tissues that did not bind estradiol, also did not respond to E2 (Jensen


7

& Jacobson, 1962). Jensen & Jacobson concluded that the E2 sensitive tissues (ovary,

uterus, breast) must have E2 receptors (ERs). Their research led to a series of

experiments worldwide that established an ER assay as a predictor of response to

hormone-withdrawal in breast cancer (Jordan & Morrow, 1999). Finally, a mechanistic

link between breast cancer and E2 action had been established (Jordan & Morrow,

1999).

Interestingly, after careful clinical research in the 1980’s, the Halstedian principles of

radical mastectomy were proven to have “no biological or clinical justification” (Fisher

& Ore, 1993, p. 98). A number of randomised trials showed there was no survival

advantage in total mastectomy over breast conserving therapy (Fisher, 1999).

Importantly, the observation of an inverse correlation between stage of cancer

development and survival rates led to an increased emphasis on early diagnosis, which

in turn has contributed to a falling mortality rate of 30 % in the past three decades

(Figure 1.1 C) (Fisher, 1999; Hortobagyi et al., 1999; Eneman et al., 2004).

Today, anti-estrogens such as tamoxifen (Tam) and aromatase inhibitors such as

anastrozole (Hind et al., 2007) have been developed as important therapeutic agents

in the breast cancer armamentarium (Emmens et al., 1962; Jordan & Morrow, 1999).

Combined with chemotherapy and radiation treatment, technological advances such as

these have provided wider choice for the doctor and patient in selecting therapy.

Unfortunately, none of the currently available treatments offers a guaranteed cure, so

that although the mortality rates of breast cancer are decreasing, the rate of breast

cancer incidence is increasing each year (Figure 1.1 C) (Eneman et al., 2004). In part,

this may have been due to enhanced mammographic screening detecting earlier and

smaller tumors. However, it is now generally accepted that a detailed understanding of

the molecular basis of the process leading to breast cancer development will provide

the fundamental basis for the generation of novel and more effective therapeutics.

1.3 The nuclear receptor superfamily

1.3.1 Nuclear receptors

E2 plays a fundamental role in breast cancer cell proliferation and its receptor (ER) is

highly expressed in breast cancers, thus, it is important to understand both E2 and its


8

receptor, the ER. The ER is a member of the nuclear receptor (NR) superfamily. The

NR superfamily consists of a large group of proteins (comprising 48 members in the

human) that play critical roles in regulating a large number of homeostatic proteins

(Evans, 1988; Freedman, 1999; Savkur et al., 2004). In fact, NRs are critical for the

regulation of eukaryotic gene expression. The ER belongs to the class I, or classical,

family of steroid NRs (Stryer, 1995), which are ligand-inducible transcription factors

that regulate the expression of target genes involved in important cell functions such as

development, reproduction and metabolism (McKenna et al., 1999; Leo & Chen, 2000;

Chawla et al., 2001; McKenna & O’Malley, 2002b; Nishihara et al., 2004; McKenna &

O’Malley, 2005).

Activation of a NR is a “multifaceted cascade of events” (Lanz et al., 1999) resulting in

the binding of the receptor to a specific DNA (regulatory) sequence. Once activated,

NRs direct, promote and stabilise formation of a preinitiation complex at a target gene

promoter region, aiding and modulating target gene expression (Shibata et al., 1997).

The NR superfamily (encompassing all NRs) can be broadly grouped into three groups

based on their DNA–binding properties and known ligands (Figure 1.2 A), and are

discussed in the next section.

1.3.2 Nuclear receptor classes

The first of the three groups of the NR superfamily, class I (‘classical’, ‘steroid’ or

‘endocrine’) receptors, function as ligand induced homodimers (Figure 1.2 B) which

bind in a head-to-head (organised as inverted repeats) manner to DNA sites (Kumar &

Chambon, 1988; Mangelsdorf et al., 1995; Chawla et al., 2001; McKenna & O’Malley,

2005). Further, they undergo nuclear translocation when activated by their ligand

(McKenna & O’Malley, 2002b). The group includes the receptors for androgen (AR),

progesterone (PR), glucocorticoid (GR), mineralocorticoid (MR) and E2 (ER).

Class II (‘nuclear’ or the second ‘endocrine’ class) receptors bind as heterodimers to

direct repeats of DNA (Figure 1.2 A & B) and are often retained in the nucleus prior to

and after binding to their appropriate ligand (Chawla et al., 2001; McKenna &

O’Malley, 2002b). These include receptors for vitamin D3 (VDR), thyroid hormone

(TR), all-trans retinoic acid (RAR) and 9-cis retinoic acid (RXR).


9

The class III receptors (which can also be subdivided into ‘adopted orphan’ and

‘orphan’ receptors) (Figure 1.2 A) bind to direct repeats of DNA as homodimers or to

extended core DNA cognate response elements as homodimers (Mangelsdorf et al.,

1995; Chawla et al., 2001; McKenna & O’Malley, 2005). The ligands for the orphan

receptors have not yet been identified and also ligand predictions based on their ligand

binding domain (LBD) sequences have not proven possible (McKenna & O’Malley,

2002b). However, the ligands for the adopted orphan receptors have proven to be low-

affinity dietary lipids (Chawla et al., 2001).

1.3.3 Steroid hormone receptor domains

All classes of the steroid NRs share six common functional domains designated A – F

(Figure 1.2 C) (Westin et al., 2000). Briefly, the A/B domains are found in the N-

terminal and are weakly conserved between the different classes. The N-terminal is of

variable length and contains a hormone-independent transactivation function or

activating function (AF-1) domain (Shibata et al., 1997; Robyr et al., 2000). The C

domain is highly conserved between the groups and is the DNA binding domain, a

66–68 amino acid long region consisting of two type II putative zinc fingers (Tora et

al., 1989; Shibata et al., 1997; Robyr et al., 2000). Although essential for DNA

binding, these zinc fingers are not able to bind DNA alone (Green & Chambon, 1991).

The D domain is a variable hinge and contains a nuclear localisation signal for ER, GR,

PR and AR, and a transactivation domain in both GR and TR (Shibata et al., 1997;

Robyr et al., 2000). The E domain contains the LBD, the ligand-dependent AF-2

domain, a dimerization domain, heat shock protein association ability, and a region

involved in nuclear localisation among other class specific functions (Shibata et al.,

1997; Robyr et al., 2000). In many receptors such as the ER, this E region is large (250

amino acids in length) and is hydrophobic (Green et al., 1986). Agonist bound LBDs of

the NR superfamily have been shown to bind a group of transcriptional regulatory

factors termed coregulators (Darimont et al., 1998). Domain F is located in the most C-

terminal region and its function is as yet unknown (Robyr et al., 2000). Interestingly

this domain is absent in PR, PPAR, RAR and RXR (Robyr et al., 2000), which fall into

the second and third NR categories, the type II endocrine receptors and the adopted

orphan receptors, respectively.


10

1.3.4 Nuclear receptor cross-talk within the mitochondrion

While the role of NRs in the nucleus has been studied in great detail, the role of NRs in

the mitochondrion affecting cell survival and energy homeostasis has recently come

under close scrutiny and is emerging as a complex and new area of NR biology. In

humans, mitochondrial cytochrome P450 activity is involved with steroid hormone(including E2 and DHT) and Vitamin D3 biosynthesis (Seliskar & Rozman, 2007).

Further, several NRs have been localized to the mitochondria, including the TR (Xu &

Koenig, 2004), GR (Scheller et al., 2000) and recently ERβ (Chen et al., 2004; Yang et

al., 2004). The demonstration that thyroid hormone (TR) could induce transcription in

the absence of nuclear factors in an isolated mitochondrial system acting via a 43 kDa

mitochondrial TR (Scheller et al., 2000) provided increased support for a primary

effect of steroid hormones on mitochondrial transcription. Additionally, T3 can induce

transcription in the absence of nuclear factors acting via mitochondrial TR (Scheller &

Sekeris, 2003).

The mitochondrial GR can interact with NFkB subunits (Tao et al., 2001), which have

also been found in the mitochondria (Cogswell et al., 2003), and putative GREs exist in

some key components of the mitochondrial oxidative phosphorylation pathway

(Scheller et al., 2000; Psarra et al., 2006). Thus, these mitochondrial GR-NFκB and

GR-GRE interactions may function to regulate key processes in metabolism and cell

growth. Recently, studies in breast cancer tissues suggest the mitochondrial ER plays a

role in tumor cell survival (Pedram et al., 2006). Taken together, these observations

indicate that the area of NR signaling in the mitochondria is rapidly evolving and

provides a strong foundation for further studies of NR action in the mitochondria.

1.4 Mechanisms of estrogen action

1.4.1 Estrogen action

As outlined above, the biological action of E2 is mediated by the ER (Huang et al.,

2001). ER are found in both the cytoplasm and the nucleus. Steroid hormones enter the

cell via passive diffusion (Shibata et al., 1997). The historical model for the

mechanisms of E2 action via the ER is depicted in Figure 1.3 A. Once in the cell

cytoplasm, the hormone, which in this case is estradiol, either: interacts with its


11

specific receptor (ER) and a complex set of chaperone proteins, including HSP-90 and

cyclophilin 40, prior to moving into the nucleus (Ward et al., 1999), or moves straight

to the nucleus where it causes conformational changes at the DNA level. The receptor

bound by hormone targets cognate hormone response elements (HREs) in the DNA as

a homodimer, which facilitates formation of a preinitiation complex followed by

transcription of the appropriate gene (Kumar & Chambon, 1988; Wong et al., 2001).

The resulting change in mRNA levels and subsequent translation will alter protein

levels in a tissue, thus potentially modulating the functional behaviour of the target

tissue (Shibata et al., 1997).

1.4.2 Estrogens and receptors

Until 1996, it was assumed that there was only one ER (now termed ERα) and that it

was responsible for mediating all pharmacological and physiological effects of both

natural and synthetic E2. However, in 1996, a second ER, termed ERβ, was cloned

from a rat prostate cDNA library (Kuiper et al., 1996). In the same year, ERβ was

cloned from a screen of a human testis cDNA library by a different research group, and

its expression revealed to be high in the human thymus, testis and ovaries (Mosselman

et al., 1996).

While E2 can activate both ERα and ERβ, the two receptors are quite different.

Further, until recently it was thought that the only mechanism through which E2

affected transcription of E2-responsive genes was by direct binding of activated ER to

estrogen response elements (EREs) (Klinge, 2001; Nilsson et al., 2001). It is now

known that the ERs can modulate the expression of genes without directly binding to

DNA. One example is the interaction between ER and the c-rel subunit of the NF-κB

complex in the interleukin-6 cytokine (IL-6) pathway. IL-6 is a multifunctional

cytokine involved in pro-inflammatory responses and recently found to be involved in

post-menopausal osteoporosis (Galien & Garcia, 1997). The ER/c-rel interaction

prevents NF-κB from binding to and stimulating expression from the interleukin-6 (IL-

6) promoter, thus directly inhibiting the expression of the cytokine IL-6 (Gailen &

Garcia, 1997).

Evidence for ligand-independent (non-genomic) activation of these receptors through

alternative signaling pathways has also emerged with an underlying mechanism of


12

receptor phosphorylation. In the absence of E2, other signaling pathways can modulate

ER through phosphorylation. A good example of such pathways includes the epidermal

growth factor (EGF) that has been found to mimic the effects of E2 on ER in the mouse

reproductive tract, and which is believed to partially mediate E2 induced growth and

differentiation (Ignar-Trowbridge et al., 1992). Another example is activator protein 1

(AP-1), which is a group of transcription factors composed of Jun, Fos and ATF family

proteins (Bakiri et al., 2002). The ER has been shown to activate transcription at AP-1

sites, that bind Jun and Fos, via interaction through the AF-1 and AF-2 sites with the

p160 component of the recruited coactivator complex (Kushner et al., 2000).

1.5 The estrogen receptor

1.5.1 Human estrogen receptors

The human ER (hERα) was cloned using cDNA libraries prepared from the MCF-7

breast cancer cell line (Walter et al., 1985). The MCF-7 cell ERα mRNA is 6322

nucleotides long, and encodes a 66 kDa (595 amino acid) protein (Green et al., 1986).

While similar, the 54 kDa ERβ appears to have diverged early in evolution based on

differences in the A/B and E domains and in AF-1 activities (Simons, 1998; Lazennec

et al., 2001). Overall the two ERs share a 97 % nucleotide sequence similarity.

However, in the ligand-binding domain the amino acid identity is only 55 %

(Pettersson & Gustafsson, 2001).

E2 activates both receptors and plays a major role in the development of breast cancer

(Haslam et al., 1993). As mentioned earlier, conclusive evidence has shown ERα to

exist as a heterocomplex with chaperone and accessory proteins, many of which have

also been found to be overexpressed in breast cancer cell lines and carcinomas (Ward

et al., 1999). Remarkably, while ERα has been shown to play an important role in

carcinogenesis, recent evidence suggests ERβ plays a role in the inhibition of cell

proliferation in breast cancer cells (Lazennec et al., 2001).


13

1.5.2 Functional estrogen receptor domains

The two ERs (ERα and ERβ) belong to the steroid receptor family (Murphy et al.,

2000; Leygue et al., 1999; Mosselman et al., 1996). These receptors are nuclear

proteins (Gorski et al., 1984) that, in the presence of E2, bind to DNA and activate

genes involved in cell growth. The ER is made up of six functional domains as

explained above (Fuqua et al., 1993; Shibata et al., 1997), and include a steroid domain

which undergoes a conformational change to lock E2 into a hydrophobic pocket

(Jordan & Morrow, 1999).

Changes in the conformation of ER allows for coregulator proteins that facilitate

transcription (termed coactivators) to bind selectively to the ligand independent

activating function (AF-1) and the hormone-dependent (AF–2) domains allowing the

ER to bind DNA via the DNA-binding domain (Tora et al., 1989; Sadovsky et al.,

1995; Jordan & Morrow, 1999; Cavarretta et al., 2002, Warnmark et al., 2003). The N-

terminal region of NRs which contain the AF-1 domain are characteristically

unstructured and the sequence is not well conserved, thus making predictions of

interacting protein partners difficult (Warnmark at al, 2003). The AF-2 domain is

within the ligand binding domain (LBD) and is part of a highly conserved sequence

and structural region. Binding partners of AF-2 domains often conform to obvious

sequence and structural characteristics (Warnmark at al, 2003).

The importance of the AF-2 domain was first identified via selective deletions of ERα,

where it was observed that ligand binding was affected by mutations in the LBD (Lees

et al., 1990). However, it has since become apparent that maximal ER transactivation

requires the synergistic activity of both the AF-1 and AF–2 domains (Cavarretta et al.,

2002). In ER transcription, the activators of the AF-1 and AF-2 domains have been

sought after by many research groups. While it is clear that the AF-2 response increases

in a ligand-dependent manner, less is known about the AF-1 domain and its mediators.

Mediator proteins are able to direct the assembly and stabilisation of a preinitiation

complex and are termed coactivators. Essentially, coactivators along with corepressors

(together termed coregulators) are able to control gene transcription via complex

interactions with the histone acetylase and histone deacetlyase complexes involved in


14

chromatin remodelling (Shibata et al., 1997; Leygue et al., 1999; McKenna et al.,

1999; Lanz et al., 1999; Barlev et al., 2001).

1.5.3 Selective estrogen receptor modifiers (SERMs)

Antiestrogen therapy, in particular the development of the selective ER modulators

(SERMs), agents designed to selectively target the ER, has been a cornerstone of breast

cancer treatment for many years (Lien & Lonning, 2000). SERMs such as tamoxifen

are known to mimic E2 action in certain tissues and to oppose E2 action in others

(Katzenellenbogen & Katzenellenbogen, 2002). Tam is used in a range of clinical

applications in breast cancer (Shang & Brown, 2002) and is proven to significantly

increase long-term survival when administered in early breast cancer (Lien & Lonning,

2000). The underlying biological therapeutic concept of Tam and other SERMs is the

development of novel ER ligands that vary in binding characteristics to the ER and

provide a range of tissue specific effects which differ from E2 (Levenson & Jordan,

1999).

An important advance was recently made in understanding the tissue specific effects of

various SERMs. Tam acts as a transcriptional regulator, and has recently been shown to

recruit coregulator proteins including coactivators (CoA) and corepressors (CoR), but

does so differentially in selective tissues. Raloxifene, a SERM used in the treatment of

osteoporosis, also recruits coregulator proteins. In the breast, where Tam acts as an

antagonist, corepressors are recruited; where as in the endometrium where Tam is an

agonist and raloxifene is an antagonist, Tam recruits coactivators and raloxifene

recruits corepressors (Figure 1.3 B) (Shang & Brown, 2002). Additionally, steroid

receptor coactivator protein-1 (SRC-1), a potent ER coactivator, has been shown by

Shang and Brown (2002) to significantly contribute to Tam’s agonistic effects. These

insights raise numerous other questions concerning the mechanisms involved in

response to therapeutic treatments based on the impact of coregulators in the ER

pathway in breast cancer (Hall et al., 2002; Yi et al., 2002).


15

1.6 Coregulators

1.6.1 Controlling gene expression

The regulation of eukaryotic gene expression in response to environmental stimuli is a

complex, multi-step and multi-regulated phenomenon (McKenna et al., 1999). Outlined

in detail earlier is how the NR superfamily comprises a large family of transcription

factors which play a major role in the control of critical pathways ranging from cell

development to homeostasis (McKenna et al., 1999; Chawla et al., 2001; McKenna &

O’Malley, 2002b). Essentially, NRs exert their action as transcription factors that

directly bind to the promoters of target genes and regulate their rate of transcription.

However, in order to modulate transcription, NRs must recruit a number of accessory

coregulators known as corepressors and coactivators (Savkur et al., 2004). NR

coregulators are those proteins that are required by the NRs for efficient transcriptional

regulation (Burke & Baniahmad, 2000; McKenna & O’Malley, 2002a). NR

coregulators have the ability to repress (corepressor) or enhance (coactivator) the

activity of NR regulated genes (Robyr et al., 2000).

1.6.2 Nuclear receptor coactivators

Nuclear coactivators are critical factors that augment NR activation of a target gene

(Robyr & Wolffe, 1998; Lanz et al., 1999). Early observations by Meyer et al. (1989)

that squelching, or transcriptional interference, occurred between steroid NRs in

transient receptor and reporter co-transfection assays indicated the existence of limiting

transcriptional cofactors that mediate AF-2 function (Meyer et al., 1989). Results from

ER, PR and GR pathway transfection experiments suggested that common cofactors

are also important in determining transcriptional efficiency (Meyer et al., 1989;

McKenna et al., 1999). Termed coactivators (Figure 1.3 C), these factors are required

for efficient and effective transcription (Tora et al., 1989; McKenna & O’Malley,

2002a; Lonard & O’Malley, 2006).

Some general criteria apply to NR coactivators. First they must significantly increase

transactivation without altering basal transcription activity (Wong et al., 2001).

Second, when over-expressed they must reverse squelching (transcriptional

interference) (Wong et al., 2001; McKenna & O’Malley, 2002b). Third, they must


16

contain transferable and autonomous activation domains (Lanz et al., 1999; McKenna

& O’Malley, 2002b).

Two main classes of coactivators exist, excluding the USA (upstream stimulatory

activity) fraction and the positive coactivators PC2, PC4 and P52. There are those

coactivators that interact with sequence-specific transcription factors, and those which

interact with general transcriptional promoter elements (Di Croce et al., 1999). The

latter class include the TATA binding protein (TBP) - associated factors (TAFs). TAFs

are found in at least two different complexes including the TFIID and the SAGA (Spt-

Ada-Gch5-Acetyltransferase) complex (Struhl & Moqtaderi, 1998). In the TFIID

complex, TAFs associate with TBP; in the SAGA complex, TAFs interact with

polypeptides involved in chromatin remodelling (Di Croce et al., 1999).

1.6.3 The TFIID complex: mechanisms of basal transcription

In 1991, a cell-free transcription system was devised by O’Malley et al. (1991) which

explored the mechanism by which a DNA-bound NR modulates transcription initiation.

It is now known that the process by which a preinitiation complex assembles at the

TATA box is highly organized and sequential.

Briefly, at many promoters the TBP together with approximately 10 TAFs comprise a

component of the TFIID complex (Goodrich & Tjian, 1994; Sauer et al., 1995; Struhl

& Moqtaderi, 1998). TFIID first associates with the TATA box via an interaction

between the TBP subunits and the TATA elements, forming a complex (Maldonado et

al., 1990; Struhl & Moqtaderi, 1998). This complex is stabilised by TFIIA, forming a

DA-TATA complex that is recognized by TFIIB forming the DAB-TATA complex

(Maldonado et al., 1990; Horwitz et al., 1996). TFIIB recruits RNA polymerase II, the

initiation of which into the complex is mediated by TFIIF, the resulting complex is

recognised by TFIIE and TFIIH (Bagchi, 1998). The complex that results is

DABpolFEH, and is able to direct a low level of basal RNA synthesis when in the

presence of ribonucleoside triphosphates (Goodrich & Tjian, 1994). For ERα ,

Sadovsky et al. showed in 1995 the AF-1 and AF-2 domains bind to TBP. Further, the

ER has been shown to bind TBP, TFIIB and many TAFs in the presence of coactivators

(Bagchi, 1998).


17

1.6.4 Coregulator action on chromatin

Nucleosomes fold eukaryotic chromosomal DNA into structures termed chromatin; this

coiling of DNA around a histone octamer in the nucleosome is recognized as a

cornerstone of transcriptional control (Kornberg, 1999). The correlation between

histone acetylation and gene activity was established by Hebbes et al. (1988). The

eukaryotic chromatin is a very organized and densely packed structure that must be

activated at an early stage if transcription is to occur (Ogbourne & Antalis, 1998). If

not activated, the chromatin may physically inhibit the function of initiation factors and

the transcription of DNA. The unravelling, or ‘priming’, of the chromatin is regulated

by acetylation or methylation of the free amino groups of lysine residues at the N-

terminal end of the histone molecules. This results in an increase in the net positive

charge and decreases the affinity for the DNA phosphate back-bone (Ogbourne &

Antalis, 1998). In general, acetylation of the lysine residues of the core histone tails is

associated with transcriptionally active DNA (Figure 1.4), and the deacetylation of

these residues is associated with transcriptionally repressed DNA (Kornberg, 1999).

Chromatin modification, as outlined above, is aided by the activity of histone

deacetylases (HDACs) in gene-repression and histone acetylases (HATs) in gene-

activation (Burke & Baniahmad, 2000). It has been found that chromatin structure

plays a role in widespread gene regulation (Beato et al., 1995). Early experiments with

ER induced genes showed an overwhelming correlation between hormone-mediated

transcription and chromatin conformational changes (Weintraub, 1985). However, it

was not until 1998 that TAFs were understood to be integral components of histone

acetylase complexes (Struhl & Moqtaderi, 1998). This finding provided a mechanistic

link between the RNA polymerase II machinery and the chromatin modifying

activities, and also provided an explanation for why coactivators such as SRC-1,

protein 300 (p300) and CREB-binding protein (CBP) were also histone

acetyltransferases (Ogryzko et al., 1996; Spencer et al., 1997; Kim et al., 2001; Cheung

et al., 2003; Lee et al., 2003).

1.6.5 Nuclear receptor corepressors

In order to understand the actions of coregulators and the importance of coregulators

which do not appear to regulate via chromatin modification, it is necessary to


18

understand the simultaneous activity of coactivators and corepressors in vivo.

Corepressors harbour intrinsic silencing ability and actively repress transcription.

Corepressors do not bind DNA directly, but are recruited by a number of transcription

factors that are bound directly to regulatory regions of target genes.

To date, a large number of corepressors have been found to play a role in a myriad of

pathways. Corepressors have been shown as essential for a wide range of biological

pathways and homeostatic mechanisms, including cell development, proliferation and

apoptosis (McKenna et al., 1999; Jepsen et al., 2000; Look, 1997). Moreover,

corepressor-repressor interactions play an important role in the activity of tumor

suppressors, hormone action and cell cycle regulators via interactions with NRs.

It has been suggested that a corepressor must fulfil two criteria: it must interact directly

with the unligated receptor leading to enhancement of basal transcriptional repression,

and it must interact with other basal transcriptional components (Robyr et al., 2000).

Regulation of the corepressor-silencer interaction is essential for control of the normal

cell cycle.

1.6.6 N-CoR

One of the most intrinsic and abundant corepressors is the nuclear corepressor protein

(N-CoR), initially discovered in 1995 (Horlein et al, 1995). N-CoR’s importance in the

cell is exemplified by data showing an embryonic lethality phenotype in the N-CoR

homozygous knock-out mouse (Jepsen et al., 2000). N-CoR has been shown to play a

major role in the corepression of several NR pathways (Cohen, 2006) including the ER

pathway (Klinge et al., 2004). Importantly, in the absence of ligand, NRs are bound by

either N-CoR, or a different NR corepressor: SMRT (silencing mediator of retinoid and

thyroid hormone receptors) (Chen & Evans, 1995; Cohen, 2006). N-CoR has been

identified in ovarian cancer cell lines and in ovarian cancer (Hussein-Fikret & Fuller,

2005).

1.6.7 ER coregulators

For the ER, a NR that plays a key role in the proliferation of breast cancer cells

(Herynk & Fuqua, 2004), a large number of coregulators have been identified. In fact,


19

it now appears that the spectrum of ER coregulator function is much wider than

initially envisaged (Nishihara et al., 2004). Known ER coregulators include SRC-1

(Onate et al., 1995), CBP and p300 (Liu et al., 1999); all three of which play a major

role in ER transactivation. SRC-1, for example, has several isoforms, of which SRC-1e

is a potent coactivator for ER (Liu et al., 1999). Upon binding by E2, ER becomes

associated with several proteins including SRC-1 and p300. CBP interacts with SRC-1

and the resulting p300/CBP complex binds ER in a ligand dependent manner and acts

as a coactivator (Liu et al., 1999). CBP/p300, P/CAF and SRC-1 all possess intrinsic

histone HAT activity (Ogryzko et al., 1996; Spencer et al., 1997; Liu et al., 1999).

Other examples of ER coregulators include ERAP160 (estrogen receptor-associated

protein 160) (Halachmi et al., 1994), SRC-2/GRIP (glucocorticoid receptor-interacting

protein)-1 (Hong et al., 1996), SRC-3/AIB-1 (amplified in breast cancer-1) (Suen et al.,

1988; Anzick et al., 1997), members of the TRAP (thyroid receptor (TR)-associated

proteins)/DRIP (vitamin D receptor-interacting proteins) complex (Fondell et al., 1996;

Rachez et al., 1998), E6-AP (E6-associated protein) (Nawatz et al., 1999), MTA1

(metastasis-associated protein 1) (Toh et al., 1994), MICoA (MTA interacting

coactivator) (Mishra et al., 2001), PELP1 (proline-, glutamic acid-, leucine-rich protein

1) (Vadlamudi et al., 2001), SHARP (SMRT/HDAC1 associated repressor protein)

(Shi et al., 2001), DP97 (novel DEAD box helicase 97 kDa) (Rajendran et al., 2003)

and SRA (steroid receptor RNA activator) (Lanz et al., 1999) among others (Figure 1.5

A).

1.6.8 SKIP: a nuclear receptor coregulator

NCoA-62 (SKIP) was discovered in 1998 as a novel coactivator protein found to be

involved in Vitamin D Receptor (VDR) mediated transcription (Baudino et al., 1998).

The VDR is a member of the class II NRs with known interacting ER coactivators

including SRC-1, CBP and SHARP (Baudino et al., 1998; Oswald et al., 2002). While

the mechanisms by which, and the complex in which, SKIP interacts with NRs have

yet to be determined, the general hypothesis is that SKIP plays a role in linking the

receptor complex to basal transcription machinery and RNA polymerase II. This is

unlike CBP and SRC-1, which both possess HAT activity (Baudino et al., 1998). That

said, many SKIP-interacting proteins have been identified including positive


20

transcription elongation factor b (P-TEFb) (Bres et al., 2005), and it is hypothesised

that SKIP plays a role in RNA splicing (Zhang et al., 2003).

A yeast two-hybrid screen using VDR as bait first identified NCoA-62, which was

subsequently cloned by Baudino et al. (1998). Termed NCoA-62 because of its size

(Novel coactivator of 62 kDa), NCoA-62 encodes a protein highly homologous to a

Drosophlia melangogaster protein involved in ecdysone-stimulated gene expression:

BX42 (Baudino et al., 1998). NCoA-62 has been found to interact with the retinoid

receptors, and also increases E2, retinoic acid and glucocorticoid mediated gene

expression when transiently expressed in cells (Baudino et al. 1998). It was named

SKIP, for SKI (an oncogene) Interacting Protein, due to an interaction between SKIP

and the highly conserved region of SKI required for its transforming activity

(Prathapam et al., 2001).

The active form of the steroid hormone vitamin D, 1(α),25-dihydroxyvitamin D3, (1,25

D3) or calcitriol, is the major high affinity ligand for the VDR (Barry, 2001). This

ligand (1,25 D3) has a major role as a calcium regulating hormone. However this ligand

has also been found to stimulate differentiation and inhibit proliferation of both

malignant and normal cells (James et al., 1996; Liu et al., 1996), and breast cancer

cells are known to express the VDR (Barry, 2001). Further, SRC-1 has been found to

stimulate 1,25 D3 alkaline phosphatase activity (Yanagisawa et al., 1999; Gill & Bell,

2000), and SKIP has been found to interact with several coregulators (Zhou et al.,

2000). SKIP activates ER transactivation and is present in breast cancer tissues, and

has been shown to interact with SHARP (Oswald et al., 2002). These observations

taken together, suggests SKIP may play a role in the development and proliferation of

breast cancer (Barry, 2001). SKIP is also predicted to play a role in additional

mammalian coregulator complexes (Zhou & Hayward, 2001).

Of the ER coregulators listed above, SRA, remarkably, is the only coregulator that has

the capacity to coactivate as an RNA rather than as a protein. For this reason SRA

stands alone in its functional characteristics.


21

1.7 SRA

1.7.1 SRA: an RNA coactivator

In 1999, Lanz et al. reported the cloning and characterisation of SRA in a seminal

paper in Cell. In stark contrast to all other coactivators which act as proteins,

conclusive evidence shows SRA is a coactivator which acts as an RNA transcript. SRA

has a cDNA sequence of approximately 870 nucleotides (nt) and Northern analysis

shows it is ubiquitously expressed at the mRNA level in all human tissues (Lanz et al.,

1999). SRA mRNA levels are enriched in the liver and skeletal muscle and decreased

in the brain (Lanz et al., 1999). The gene for SRA is located on chromosome 5q31.3-32

(Leygue et al., 1999), is highly conserved across species, and is composed of 5 exons

in human, rat and mouse genomes. Sequence homology searches have shown mouse

SRA to share sequences of 75 % nucleotide (nt) similarity to that of human SRA

cDNA, which indicates that SRA is not species specific for Homo sapiens and may

play an important role in both cellular processes and mammalian evolution.

Of interest, SRA was initially identified via a yeast-II hybrid screen using a human

progesterone receptor (PR) AF-1 domain as a probe. This identification of SRA was

subsequently confirmed with a number of different screens in several different human

and mouse libraries including those made from heart and skeletal muscle (Lanz et al.,

1999). As mentioned previously, while the AF-2 domains of NRs are highly conserved

in both sequence and structure, much less is known of the binding partners to the more

poorly conserved AF-1 domain (Warnmark et al., 2003). Basing their work on a

hypothesis that “specificity in steroid receptor-mediated transactivation might be

provided by factors that associate with the poorly conserved AF-1” (Lanz et al., 1999,

p. 17), Lanz et al. had demonstrated that the AF-1 domain has specific and crucial

importance to NR transactivation and in addition that SRA was a remarkable and novel

gene.

Initially, Lanz et al. assayed the effect of SRA on PR mediated transactivation. SRA

enhanced PR transactivation but it did not increase basal transcription, which is

consistent with its functioning as a coactivator (Lanz et al., 1999). Further transfection

studies conclusively showed SRA only enhanced GR, AR, MR, PR, and ER


22

transcription. SRA selectively enhanced steroid receptor transcription via the NR AF-1

domain (Lanz et al., 1999).

SRA has several interesting characteristics with the most remarkable being that its

coactivation is not a function of SRA protein (Lanz et al., 1999). To support this

notion, the addition of cyclohexamide (a protein synthesis inhibitor) did not decrease

SRA transactivation activity (Lanz et al., 1999). Furthermore, in standard

transactivation assays, conservative mutations that did not change the amino acid

protein sequence did affect SRA's coactivator ability. When Lanz et al. attempted to

detect whole cell translated SRA proteins with antipeptide SRA antibodies in vitro they

had no success. However, SRA RNA could be detected in a 600-700 kDa protein

complex and coimmunoprecipitated with SRC-1 (Lanz et al., 1999). Thus, extensive

studies performed by Lanz et al. showed that SRA-mediated NR coactivation did not

require the expression of SRA protein. Lanz et al. concluded that while SRA may

encode a protein, SRA certainly acts as an RNA coactivator and that its existence in a

protein complex indicates it may act as part of a ribonucleoprotein scaffold in specific

coactivator complexes.

The early description of SRA indicated the existence of multiple isoforms. SRA was

initially found to exist in four distinct splice isoforms, SRA1, SRA2, SRA3 and SRA-

Del (Lanz et al., 1999; Leygue et al., 1999), recently there have been several more

human SRA isoforms discovered (Emberley et al., 2003). The SRA1, SRA2 and SRA3

sequences are similar in the core central region, but differ in their 5’ and 3’ regions.

SRA-Del has a deletion of exon 3 (Leygue et al., 1999). The recent discovery of more

SRA isoforms includes isoforms with larger open reading frames, including an SRA1

isoform encoding a 236 aa protein (theoretically coded for with SRA but with the

inclusion of 73 aas) known as SRA protein (SRAP).

While it has been conclusively shown that SRA can function as an RNA, it has also

become clear that SRAP may also have a biologically functional role. SRAP was first

found in prostate cancer cells by several groups (Kawashima et al., 2003; Chooniedass-

Kothari et al., 2004; Kurisu et al., 2006). However, SRAP has recently been found in

breast cancers (Chooniedass-Kothari et al., 2006) and an alternative splice variant of

SRA (SRA1) has been found more recently in breast cancer cell lines where two start

codons have been inserted at the 5’ end of the mRNA leading to the production of both


23

SRA mRNA and SRA protein, which suggests the possibility that SRA RNA and SRA

protein (SRAP) regulate each other (Hube et al., 2006). SRA appears to function in an

isoform- and cell-specific manner and to be present in an RNA-protein complex

(possibly together with SRAP).

1.7.2 SRA in breast cancer

Of importance to this thesis, SRA is overexpressed in breast cancer and breast cancer

cell-lines suggesting it may play an important role in breast cancer tumorigenesis (Lanz

et al., 1999; Murphy et al., 2000; Hussein-Firket & Fuller, 2005). In transient

transfection experiments, when SRA is overexpressed in HeLa and T47D cells,

transactivation activity is increased by five to ten fold (Lanz et al., 1999; Lanz et al.,

2002).

There is mounting evidence to implicate SRA in human breast tumorigeneis. SRA is

overexpressed in breast tumors correlating with ERα+ levels (Leygue et al., 1999;

Murphy et al., 2000; Hussein-Firket & Fuller, 2005). Such data suggest a role for SRA

in modulating E2 action via the ER in ERα+ breast cancer. Furthermore, a perfect

deletion of SRA exon 3 exists in the SRA-Del splice variant which is associated with a

more aggressive tumor phenotype suggesting such a deletion could result in

modification of the ER signaling pathway (Leygue et al., 1999). Leygue et al. further

suggested this deletion may lead to a short SRA transcript and the premature

termination of putative SRA-protein interactions, thus altering the activity of SRA.

Cavarratta et al. (2002) reconfirmed the importance of SRA in ERα+ breast cancer by

showing that SRA functions in a coactivator complex together with SRC-1. Further,

they detailed decreased ERα transcriptional activation caused directly by the inhibition

of three transiently transfected coactivators, SRA, TIF2 and SRC-1 (Cavarratta et al.,

2002).

An SRA overexpressing transgenic mouse model was developed (Lanz et al., 2003).

The mouse showed signs of increased mitosis and cell death of the mammary

epithelium, but interestingly there was no increase in tumor incidence (Lanz et al.,

2003). Furthermore, when the SRA transgenic mice were crossed with MMTV-ras

overexpressing mice, mammary tumor development was reduced (Lanz et al., 2003).


24

More recent evidence indicates that SRA may play a role in Tam resistance (Lanz et

al., 2002; Coleman et al., 2004), SRA can also be activated by mitogen-activated

protein kinase (MAPK) (Deblois et al., 2003), it has recently been found to coactivate

muscle cell differentiation factor MyoD (Caretti et al., 2006), and localises to the

cytoplasm of the cell for the majority of the time (Zhao et al., 2007). This suggests

SRA has a broad biological function both inside and outside the nucleus. Recently,

SRAP has also been implicated in breast cancer, with recent data suggesting patients

with SRAP positive tumors have a lower rate of tumor recurrence and improved

outcomes (Chooniedass-Kothari et al., 2006). These data support an important role for

SRA in breast cancer cells and also indicate that the mechanisms of SRA action are

likely to be complex (Cavarratta et al., 2002).

Taken together, the data mentioned above suggest that SRA plays an important role in

both normal mammary development and breast cancer growth. In summary, SRA is anovel RNA-coactivator that plays a central role in ER-mediated transactivation. It is

overexpressed in ERα+ breast cancer and is a target for mutation in a subset of women

with breast cancer. The discovery of SRA opens up an entirely new concept for how

eukaryotic transcriptional activation is regulated (Lanz et al., 1999). The precisemechanisms by which SRA facilitates transcription have yet to be determined, and in

particular, in respect of any RNA-protein interactions that involve SRA, of which littleis known.

1.7.3 SRA structure

In order to understand how SRA functions as a molecule, it is essential to appreciate its

structure. SRA is a complex RNA molecule with several predicted secondary stem-

loops (Figure 1.6 A). Site directed mutants (SDMs) of SRA stem-loops were originally

produced by Lanz et al. (1999) in an attempt to determine which region(s) contribute to

transactivation. The most important three regions of SRA for transactivation were

originally shown to be: the 5’ region (stem loop structure 1; STR1); the middle region

(STR7); the 3’ region (STR11). Important co-operativity has been demonstrated

between these portions of SRA, such that each contributes to mediate basal and ligand

stimulated transcription (Lanz et al., 2002). Intriguingly, STR5 has been demonstrated

as capable of allowing SRA to switch from a coactivator to a corepressor via Pus


25

(pseudouridine synthase)-1p and Pus-3p pseudouridylation (Zhao et al., 2004; Zhao et

al., 2007).

1.7.4 SRA stem loop 7

Stem loop or structure (STR) 7 is the largest and one of the most stable stem loop

structure in the SRA molecule. In a series of RNA secondary structure plots, the

conformation of STR7 remains constant. STR7 is 89 nt in length. In transient

transfection transactivation experimental mutations of STR7 (termed SDM7 for site

directed mutagenesis of STR7) significantly decreased transcriptional activity of a

heterologous luciferase reporter (Lanz et al., 2002): SRA transactivation efficiency was

decreased to ~30 % of basal activity (Figure 1.6 B), the largest decrease in activity

measured. Furthermore, STR7 interacts with several other stem-loops of SRA to

modify transactivation capacity.

To date, available evidence suggests that SRA STR7 plays an important role in the

coactivator function of the SRA molecule. In order to further understand the molecular

mechanism(s) behind this coactivator function, it is essential to determine what

interacting proteins may be enabling STR7 to contribute to the coactivation ability of

SRA.

1.8 RNA biology: A growing field

1.8.1 The changing role of RNA in biology

In 1978, Francis Crick stated that:

Biologists should not deceive themselves with the thought that some new class of biological

molecules, of comparable importance to proteins, remains to be discovered. This seems highly

unlikely.

(Francis Crick in 1978, cited by Eddy, 2002, p.137)

At this point, before continuing the discussion on SRA, it appropriate to delve a little

more deeply into the recent explosion of literature on the roles of RNA in the cell. The

“Central Dogma” of molecular biology (Crick, 1970) has long held the belief that DNA


26

(regarded as the information) is transcribed into RNA (the messenger) and is translated

into proteins (the product and “active” or “doing” molecule). However, the role of

RNA is now proving to be substantially more complicated.

RNA can be split into two major categories: messenger RNA (mRNA) and non-coding

RNA (ncRNA) (Huttenhofer & Vogel, 2006). Generally speaking, mRNA is the RNA

most referred to, the product between (and essential to) transcription and translation. In

contrast to mRNA, ncRNA functions at the RNA level and is not translated into

protein. We have recently witnessed an explosion in the number of ncRNA families

(Huttenhofer & Vogel, 2006) and ncRNA now includes the traditional RNA families

such as transfer RNA (tRNA), ribosomal RNA (rRNA) and splicosomal RNA (sRNA)

as well as: small nucleolar RNA (snoRNA), micro RNA (miRNA) and small

interfering RNA (siRNA).

In 2002, the FANTOM (Functional Annotation Of Mouse) consortium found several

mRNA transcripts that were not translated into protein and yet seemed to occur with agreater than random frequency (Okazaki et al., 2002 Tomaru & Hayashizaki, 2006). In

2005, FANTOM published that they had found over 100,000 sequences that theyclassified as ncRNA (Carninci et al., 2005). This number of ncRNA transcripts in the

cell indicates that RNA plays a much more complicated role in biology than originally

suspected.

The importance of RNA in biology is exemplified by the discovery of SRA inoncogenesis, Evf-2 in organogenesis (Feng et al., 2006; Shamovsky & Nudler, 2006),

and H19, which has recently been associated with increased risk in breast cancer

(Easton et al., 2007). ncRNAs have been shown to regulate gene expression by novel

mechanisms including as RNA interference, gene suppression, gene silencing, DNA

demethylation, RNA splicing, chromatin remodelling and imprinting. There is also

mounting evidence to suggest RNA involvement in neurological disease and cancer

(Costa, 2005; Mattick & Makunin, 2006). It is now becoming clear that RNA performs

critical functions in development and cell differentiation (Costa, 2005; Shamovsky &

Nudler, 2006). It is also of interest to note that in 2006, the Nobel Prize for Medicine

was awarded to Andrew Fire and Craig Mellow for “their discovery of RNA

interference – gene silencing by double stranded RNA” (Nobel Web, 2007, p. 1).


27

1.9 RNA-binding proteins

1.9.1 General

Given the growing importance of RNA in biology, it is easy to understand that RNA-

binding proteins (RBP) play a major role in the various components of the regulation of

the gene expression pathway (McCarthy & Kollmus, 1995; Mattick & Makunin, 2006).

The discovery of SRA has transformed our understanding of coactivators and the role

of RNA-protein interactions in the nucleus. In order to understand the implications of

novel RNA-protein interactions it is imperative to explain the basic mechanisms

involved.

Messenger RNAs (mRNAs) are produced from pre-mRNAs or heterogenous nuclear

RNAs (hnRNAs) in the nucleus via a series of processes including 5’ capping, splicing

and polyadenylation (Burd & Dreyfuss, 1994). The mRNAs are then transported from

the nucleus into the cytoplasm where translation and further processing takes place

(Burd & Dreyfuss, 1994). mRNA processing is mediated by clusters of small RNAs

and RNA-binding proteins called ribonucleoprotein (RNP) complexes (Burd &

Dreyfuss, 1994). Most of the RNA in the cell is not mRNA, but RNA in the form

complexed to RNPs (Dreyfuss et al., 1993), many of which exist in different splice

forms (Iwasaki et al., 2001). Until recently, RNA has been considered to be only the

intermediate message, prior to production of protein. In 1988, however, Dreyfuss et al.

pointed out that this inferred “protein-centric point of view” will not always be a

correct interpretation (Dreyfuss et al., 1988, p. 1419).

Proteins that bind RNA contain distinct conserved motifs which facilitate interactions

with RNA (Burd & Dreyfuss, 1994). The presence of one or more of these domains is

highly predictive of the capacity to bind RNA. One of the best characterised RNA-

binding domains is the RNA recognition motif (RRM) (Maris et al., 2005). It is 90 to

100 amino acids (aa) in length and is often present in more than one copy within

proteins (Burd & Dreyfuss, 1994; Siomi & Dreyfuss, 1997). Two internal consensus

sequences within the RRM have been characterised, RNP1 and RNP2, which contain

highly conserved aa at specific locations (Burd & Dreyfuss, 1994). Proteins with an

RNP domain are involved in most, if not all, pathways of posttranscriptional regulation


28

(Maris et al., 2005) of gene expression (examples include HuR and HuD) (Varani &

Nagai, 1998).

There are several other RNA-binding motifs, including the RGG Box, which comprises

20-25 aa and contains repeated arginine-glycine-glycine sequences (Burd & Dreyfuss,

1994). The KH (hnRNP K homology) motif (~50 aa) and the double-stranded (ds)

RNA-binding motif (65-68 aa) are also found in single or multiple copies within the

protein sequence (Burd & Dreyfuss, 1994). dsRNA-binding proteins demonstrate less

obvious sequence specificity, making them harder to characterise. The arginine-rich

motif (ARM) is a short 10-20 aa sequence found in a variety of different proteins (Burd

& Dreyfuss, 1994; Draper, 1999). Studies have shown that ARMs may increase non-

specific binding of RNA via positively charged arginines within the domain (Burd &

Dreyfuss, 1994). Other RNA-binding domains include the zinc finger/knuckle domain,

the cold shock domain, and a number of rarer domains for which there are only a few

examples (Burd & Dreyfuss, 1994).

1.9.2 Nuclear Receptors and RNA-binding proteins

The importance of RNA-binding proteins in diverse biological pathways is now well

established. There is vast evidence in the literature detailing interactions between

RNA-binding proteins and NRs. For example the binding of, and modification by, HuR

and HuD to the AR mRNA (Yeap et al., 2004). Relevant to this thesis, several recent

observations have elucidated significant interaction between RNA, RNA-binding

proteins and NRs. For example, Xu and Koeing (2004) found a novel RNA-binding

domain (via a series of deletion mutations) that localized to a 41-aa segment of the

TRα residing between the second zinc finger and the ligand-binding domain.

Furthermore, transient transfections showed this RNA-binding domain allowed the TR

to bind to SRA (Xu & Koeing, 2004).

Recently, RTA, repressor of tamoxifen transcriptional activity, was described (Norris et

al., 2002). Expression of RTA inhibits Tam-mediated agonist activity and has been

found to partially inhibit E2-mediated transactivation (Norris et al., 2002).

Interestingly, RTA is a coregulator which has been found to interact with RNA in vitro

via an RRM. When this RRM is mutated a dominant negative form of RTA results


29

which increases ERα transcriptional activity (Norris et al., 2002). However, no work

has yet been undertaken regarding potential RTA and SRA interactions.

The human coactivator activator (CoAA) was discovered by Iwasaki et al. (2001) as a

coactivator for the thyroid hormone receptor-binding protein (TRBP) and p300. It has

been suggested that CoAA acts synergistically with TRBP and CBP, coactivating

transcription which is mediated by multiple hormone-response elements (Iwaskai et al.,

2001). CoAA has two RRMs in its 669 aa coding sequence (Iwaskai et al., 2001). It is

not known to which RNA CoAA binds although a splice variant termed coactivator

modulator (CoAM) has also been discovered. CoAM only contains the RRM region of

CoAA and has been found to strongly repress both TRBP and CBP action, suggesting

it may modulate CoAA function.

The discovery of these coregulators, which contain RNA-binding domains and have the

capacity to modulate NR signaling, illustrates the complexity of cofactor involvement

in transcription and highlights the potential importance of these RNA-protein

interactions. The discovery of SRA and the aforementioned coregulators, together with

the RRM-containing coregulators with roles in metabolism, RNA splicing and as an

RNA-binding Dead-box proteins: PGC-1, CAPERα, CAPERβ and p72 (Wantanabe et

al., 2001; McKenna & O’Malley, 2002; Dowhan et al., 2005), has generated significant

interest in understanding RNA-protein interactions, especially in NR signaling.

1.10 SRA RNA-binding proteins

1.10.1 SHARP

Given the importance of SRA as a modulator of transcriptional regulation by NRs,

identification of proteins that target it is an important goal. The following sections

describe the identification of some RBPs that bind SRA. The first is SHARP

(SMRT/HDAC1 Associated Repressor Protein), which has been described as a ‘potent

transcriptional repressor’ and interacts with SMRT, HDAC1 and HDAC2 directly (Shi

et al., 2001, p. 1140). SHARP is a large protein (~220 kDa) which contains three RNA

Recognition Motifs (RRMs) and binds to SRA. Furthermore, mutations to the RRMs

abolish binding between SHARP and SRA, implicating the RRMs as essential for the

interaction (Shi et al., 2001).


30

A truncated recombinant SHARP (a ~65 kDa protein) was shown to bind SRA in GST

pull-down experiments, and immunoprecipitation real time PCR (IP-RT-PCR) showed

coassociation of FLAG-SHARP with SRA, an association which was further prevented

with the addition of RNase treatment (Shi et al., 2001). In contrast, SHARP had no

effect on potential SRC-1 transcriptional activity, indicating a novel mechanism in

which a repressor can attenuate hormone action (Shi et al., 2001).

SHARP is different from other corepressors in that it does not repress SRA’s potential

transactivation via HDAC activity. However, it does compromise SRA-potentiated

transactivation when overexpressed with ER and GR (Shi et al., 2001). SHARP binds

to both SRA, a coactivator, and to SMRT, a corepressor, and has been shown to

interact with SKIP in the Notch pathway (Oswald et al., 2002). Interestingly, at the

time of commencing this thesis, the structural motif within SRA that is bound by

SHARP had not been identified.

1.10.2 SRA interacting RNA-binding DEAD-box proteins

Wantanabe et al. (2001) recently described a new subfamily of RNA-binding DEAD-

box proteins p72/p68 that coactivate ERα but not ERβ. p68 is an RNA helicase

described by Endoh et al. (1999) which had previously been identified as a hERα

coactivator. p72/68 can interact with the AF-2 of any SRC-1/TIF2 protein family and

was shown to interact with SRA (Wantanabe et al., 2001). In transient transfection

assays the combination of p72/p68 and SRA induced a significant increase of hERα-

mediated transactivation. A current hypothesis predicts that p72/p68 in some way

bridges the N-terminal hERα AF-1 domain and SRA. It should be noted that there was

little definitive binding of SRA in the report of p72/68, and there was no identification

of the region within SRA targeted by p72/68. The identification of the p72/p68

subfamily of RNA-binding DEAD-box proteins interacting with SRA and modulating

ER action is consistent with the thesis that SRA may play a major role in the chain of

events leading to the increased transactivation of hERα.


31

1.10.3 Pseudouridine synthase and SRA

It has very recently been discovered that SRA transcripts can be bound and

pseudouridlated by two pseudouridine synthases (PUS), mPus1p (Zhao et al., 2004)

and mPusP3 (Zhao et al., 2007). PUSs are evolutionarily conserved enzymes which areable to catalyze isomerization of uridine to pseudouridine in RNA (Zhao et al., 2004),

and deregulated PUS expression has been associated with cancer (Zhao et al., 2007).

Zhao et al. (2004 & 2007) found that both mPusP1 and mPusP3 can act as NRcoactivators, but each through a slightly different mechanism.

mPus1p and mPus3p have been found to modify SRA at several different positions

(including some common targets) and the order of modification of SRA by these two

enzymes determines how many positions are pseudouridylated (Zhao et al., 2004; Zhaoet al., 2007). Interestingly, hyperpseudouridylation of SRA can occur, through a

mutation of a common target of both mPusP1 and mPusP3, leading to a dramaticchange in SRA function where SRA is switched from a coactivator to dominant-

negative molecule (Zhao et al., 2007). PUS may stabilize base stacking and hydrogen

bonding within SRA and its binding partners, as has been shown to be the case withother ncRNAs.

Such data suggests that a complex set of post-translational NR-coactivator complex

modifications exist (Zhao et al., 2004) which play a role in both SRA activation and

repression in NR coregulator complexes (Zhao et al., 2007). Such activity may helpexplain how SRA can augment NR activity, and this exciting revelation of further

complexity of SRA biology highlights the importance of understanding the functionalrole of SRA activity in NR signaling.

1.11 The SRA project

1.11.1 My previous work on SRA

In order to understand how SRA functions as a molecule, it is essential to appreciate its

structure. SRA is a complex RNA molecule with several predicted secondary stem-

loops. Historically, site directed mutants have been made by Lanz et al., who mutated

several parts of SRA in an attempt to determine which region(s) contribute to


32

transactivation. Stem loop, or Structure, 7 (STR7) is the largest and most stable stem

loop structure in the SRA molecule. In a series of RNA secondary structure plots, the

conformation of STR7 is always constant (Figure 1.6 A). The STR7 is 89 nt in length.

In transient transfection transactivation experiments, mutation of STR7 (termed SDM7

for site directed mutagenesis of STR7) significantly decreased transcriptional activity

of a heterologous luciferase reporter (Lanz et al., 2002). SRA transactivation efficiency

was decreased to ~30 % of basal activity, the largest decrease in activity measured

(Figure 1.6 B). Furthermore, STR7 interacts with several other stem-loops of SRA to

modify transactivation capacity.

My previous work (during my Honours year) on this project and in this laboratory, has

shown that a number of breast cancer proteins bind to STR7 (Figure 1.7 A, B, C).

Based on this observations I performed a yeast three hybrid screen (SenGupta et al.,

1996) of a human primary breast cancer cDNA library (Byrne et al., 1998) with STR7

as bait (Figure 1.7 D), and this led the identification of a novel RNA-binding protein

that is now termed SLIRP (SRA stem loop interacting RNA-binding protein).

SLIRP is a small protein, with a RRM spanning nearly the entire protein sequence, this

RRM is highly homologous to the three RRMs in SHARP’s N-terminus (Shi et al.,

2001). When I began work on my PhD thesis, nothing was known about SLIRP and its

function in the cell. My PhD has aimed to characterise SLIRP, and to determine its

functional role in normal and breast cancer cells.

1.11.2 Project hypotheses and aims

Given the pressing need to better understand the mechanisms behind E2 action in

breast cancer cells in order to devise more specific and effective therapies, we need a

detailed knowledge of the molecular action of E2 in the cell. SRA is a novel RNA

coactivator involved in E2 action in breast cancer, yet little is known of how SRA

mediates its effects. SLIRP is a novel SRA-binding protein, but little is known of its

function in the cell. SRA STR7, a highly stable stem loop, appears to play a critical role

in the transactivation activity of SRA as a whole, and furthermore, SLIRP binds to

STR7. Thus the hypotheses for this project were:

Hypothesis 1: SLIRP specifically binds SRA STR7.


33

Hypothesis 2: SLIRP modifies ER transactivation, interacts with other NR

coregulators and has an important functional role in other NR pathways.

Hypothesis 3: SLIRP is expressed in breast cancer cells and regulates breast cancer

cell growth.

In order to address these hypotheses, the overall aims of my PhD have been:

Aim 1: To perform basic characterisation of SLIRP via bioinformatic and sequence

analysis.

Aim 2: To assess SLIRP’s functional role in different NR signaling pathways in the

model human cancer cell line, HeLa.

Aim 3: To assess SLIRP’s expression in human cancer cell lines, and its intracellular

localization.

Aim 4: To assess SLIRP’s interactions with other NR coregulators.

Aim 5: To assess SLIRP’s functional role in breast cancer cells.

1.11.3 Significance and outcomes of this study

Nuclear coregulators play an essential role in E2-mediated transcription and E2

mediated transcription is an important component of the growth of human breast

cancer. RNA coactivators are novel in the transcriptional context and the discovery of

SRA has changed our understanding of transcriptional regulation. To date, SRA is the

only RNA NR coactivator described in the literature and it appears to play an important

role in the transactivation of the NRs (Lanz et al., 1999). Its discovery has generated

many important questions, some of which I will address in this thesis. These include,

how does SRA function as an RNA coactivator? Second, what are the proteins that

bind to different portions of SRA? Third, do SRA-binding protein family members

function cooperatively to regulate transactivation? Fourth, how many proteins bind to

SRA and with what affinity? Fifth, are these proteins involved in cancer or normal cell

growth regulation? Sixth, is any effect on transactivation related to the stage of cancer

cell development? Seventh, is there an advantage to having an RNA as a backbone for

a coactivation complex? Eighth, are there other NR RNA coregulators?

The research I will describe in the following Chapters provides considerable insight

into the mechanisms of action of SRA as an RNA coactivator in breast cancer via its


34

interaction with other NR coregulators, including the novel NR corepressor SLIRP.

Such research is critical in order to generate a more complete picture of the molecular

mechanisms involved in E2 action in breast cancer. A comprehensive understanding of

SRA-protein interactions in breast cancer cells and their impact on gene expression and

cell growth is an essential step towards the future development of new breast cancer

therapies that may specifically target such RNA-protein interactions.

In Chapter I have introduced SLIRP and the importance of RNA biology to cancer

research. In Chapter 2 I will detail the Materials and Methods used for all

experimentation undertaken in this thesis. In Chapter 3 I describe the results for the

basic characterisation of SLIRP, including all bioinformatic analysis, expression

studies and SRA-SLIRP interaction studies. I discuss some specific aspects of its

identification in order to establish context for the rest of this thesis. In Chapter 4 I

detail the discovery of three fascinating aspects of SLIRP biology: that SLIRP is a NR

corepressor, that SLIRP is predominantly mitochondrial and that SLIRP interacts with

other NR corepressors. Chapter 5 presents data that highlights the complex role for

SLIRP in the ER pathway of E2-dependent breast cancer cells. The final Chapter,

Chapter 6, provides discussion and conclusions which highlight the importance of both

SLIRP and SRA in ER biology and emphasizes how this thesis provides a new

foundation of NR coregulator knowledge that will act as a firm basis for further study.

Figure 1.1 A, B: Cancer affects a large proportion of our population. Panel A. Twographs taken from the Victoria Cancer Council (CanStat, 2006), showing the proportion(in a percentage of the total) of all deaths that can be attributed to cancer and otherdiseases, disorders and accidents in Victoria, Australia. While these statistics are not forthe whole of Australia, the relative proportions are similar. The second graph shows theyears of potential life lost (measured in the thousands) due to these various forms ofdeath. Panel B again shows two graphs, taken from the AIHW & AACR (2004), thegraph on the left depicts incidence of cancers that specifically affect females in Australia,and on the right, the mortality rate of these cancers.

A

B

Proportion of total deaths (%) Years of potential life lost to age 78years (measured in thousands)

Cancer

IHD

CVD

CLRD

Diabetes

Suicide

Transport

accidents

Other

causes

29.6%

17.8%

8.5%

4.7%

3.3%

1.5%

1.2%

33.5%

35

C

Figure 1.1 C: Breast cancer incidence continues to rise in our community. Panel C is agraph taken from the AIHW & AACR (2004) depicting the incidence and mortality ratesof breast cancer in Australian females. While the mortality rate of breast cancer isdecreasing due to early intervention and chemotherapy treatments, the incidence rate isincreasing for unknown reasons. The risk for an Australian woman living in a urbanenvironment of developing breast cancer is expected to continue to rise (AIHW &AACR, 2004).

36

ER α, β

PRARGRMR

RAR α, β, γ

TR α, β

VDREcR

Endocrine / NuclearReceptors

Ligands:High affinity,

hormonal receptors

RXR α,β

PPAR α,β, γ, δ

LXR α,β

FXRPXR / SXRCAR

Adopted OrphanReceptors

Ligands:Low affinity,dietary lipids

SF-1LRH-1DAX-1SHPTLXPNRNGFI-B α,β, γ

ROR α,β, γ

ERR α,β, γ

RVR α,β, γ

GCNFTR 2, 4HNF-4COUP-TE α,β, γ

Orphan Receptors

Ligands:unknown

Type I

Type II

Figure 1.2 A, B: The nuclear receptor superfamily. Panel A depicts the acceptedgrouping of steroid hormones and their receptors into 3 (or 4) categories. First are theendocrine receptors, including both type I and type II endocrine receptors: these two sub-groups simply refer to the way the receptor binds the DNA in either a homo- or hetero-dimer and the direction in which they bind the DNA (refer to panel B). The second groupare the adopted orphan receptors which have ligands with a low affinity, and are oftendietary lipids. These receptors were previously categorized as orphan receptors. The thirdgroup are the orphan receptors whose ligands currently remain unknown. This figure isadapted from Chawla et al (2001). All the receptors listed in a known ligand category arevertebrate receptors except EcR, the only non-vertebrate receptor for which the ligand isknown. Panel B is taken from McKenna & O’Malley (2005). The two figures describe thetwo types of endocrine receptors, and detail how they either homodimerise and bind theDNA in inverted repeats (type I) or heterodimerise and bind the DNA in forward repeats(type II). Examples of each type are listed below the diagram.

A

B

37

A/B C D FE

Ligand Binding Domain

DNABindingDomain

HingeDomain

AF-2AF-1

Figure 1.2 C: Schematic showing the common nuclear receptor domains. Thisschematic adapted from Simons (1998) shows the sequence domains of the endocrine,or steroid, nuclear receptors. Briefly, the A/B domain is weakly conserved betweengroups and contains the activating function (AF)-1 domain. The C domain is highlyconserved between groups which is the location of two zinc fingers and contains theDNA binding domain. Domain D is the variable hinge domain and contains nuclearlocalization signals for the receptor. The ligand binding domain E contains the AF-2.Domain F is not conserved between receptors. Some receptors appear not to havedomain F.

C

38

AAARibosomes

Gene

AAA

Nucleus

Cytoplasm

CELL

Estrogen Receptor

Chaperone proteins

EGF-R erbB-2

receptortyrosinekinases

A

Figure 1.3 A: Schematic detailing our basic understanding of the ER pathway. Panel Ais a simplified schematic depicting our current understanding of estrogen action in thecell. Essentially, estradiol (E2) enters the cell membrane via passive diffusion whereeither in the cytoplasm, or the nucleus (more often) it is bound by its receptor, theestrogen receptor (ER). In the nucleus this complex can initiate transcription. An mRNAis produced, exported to the cytoplasm and eventually a protein product is produced. Thisby no means is a static event, rather it is extremely dynamic with any one cell potentiallyundergoing activation of several pathways simultaneously, including non-classical pathsof activation such as by tyrosine receptor kinases, which are depicted on this schematic.

Estradiol (E2)

39

Figure 1.3 A: The important role of coregulators in cancer biology is becoming moreapparent. This is a schematic taken from Katzenellenbogen & Katzenellenbogen (2002,p. 2381) showing the involvement of coactivators in SERM (selective estrogen receptormodulator) action. This picture illustrates how tamoxifen and raloxifene recruit bothcoactivators and corepressors under different conditions. For example, in the breast,where both tamoxifen and raloxifene are antagonists, corepressors are recruited. In theendometrium, tamoxifen is an agonist, and recruits coactivators. The different receptorconformations which form can directly affect the transcription of a gene (Shang &Brown, 2002). Coregulators are involved in many processes in the breast cancerpathway and in the treatment of some breast cancers, their importance is becomingparamount and yet our understanding of the mechanisms involved in coregulator actionin breast cancer remains poor.

A

40

GeneERE

RNAPolymerase II

Transcription

ER

AF-2AF-2

AF-1 AF-1

E2

AF-2AF-2

AF-1 AF-1

E2

Coregulators(coactivators & corepressors)

SRC-1

SHARP

pCAF

N-CoR

SRAB

Figure 1.3 B: Our current understanding of transactivation in the ER pathway. Panel B isa schematic explaining our current understanding of E2 action in the nucleus. Essentially,the dimerized ER bound with E2 enters the nucleus, binds the hormone response element(HRE), in this case the estrogen response element (ERE), and turns on transcription.Although this pathway has been well established for many decades, it has only beenrecently that the actual mechanistics of how the ER pathway is controlled has come tolight. Essentially, this pathway is controlled by mediators or coregulators. Coregulatorsare either coactivators (which increase the rate of transcription) or corepressors (whichdecrease the rate of transcription).

41

Figure 1.4: Coactivators and corepressors. Diagram taken from Robyr et al (2000, p.339) showing the ligand dependent switch between a NR and a co-regulator. In the firstdiagram the NR associates with the co-repressors N-CoR and SIN3, a complex which inturn recruits a HDAC. Deacetylation of the histone then leads to transcriptionalrepression. The co-activator association with histone is shown for comparison. Here aco-activator such as SRC-1 and p/CAF exert HAT activity leading to acetylation of thehistone and transcriptional repression. (The dashed lines indicate those co-regulatorcomplexes whose exact composition remains undetermined.)

42

Figure 1.5: ER coregulators. Coregulators which interact in various parts of the ERpathway are presented in a schematic above, as taken from McDonnell and Norris(2002). Red indicates corepressors and green indicates coactivators. The proteins whichact as coregulators of the ER pathway often also have other functions in the cell. This isa simple schematic which depicts several of the coregulators which will be discussed indetail in this thesis.

43

SRA

SRA STR7A

Figure 1.6 A: Mfold secondary structure plot of SRA . SRA is predicted to be acomplicated structure, of which there are several obvious and conserved stem loop sub-structures. rE value for the full-length SRA structure was –243.1 kJ mol-1. STR7 (labeled)is one of the largest and most stable sub-structures of SRA and is depicted above.

44

Mutations of SRA: Effect on SRACoactivation

B

Figure 1.6 B: The contribution of different SRA stem loops to total SRA coactivation.Panel B is adapted from Lanz et al (2002) and shows that several stem loops of SRA areimportant for total SRA coactivation. The box highlights SDM7 (site directed mutagen 7).When STR7 (structure 7) was mutated (with 3 point mutations expected to only affectSRA sequence and not structure) a large decrease in SRA coactivation was noted. Thisled us to hypothesize that there would be important RNA-protein interactions at STR7,affecting SRA coactivation function.

45

Relative coactivationSRA

SDM1

SDM2

SDM3

SDM9

SDM4

SDM5

SDM10

SDM7

SDM7a

SDM11

0 20 40 60 80 100

STR7 ProbeMCF-7 nuclear

MCF-7 cytoplasmicMDA-MB-468 nuclear

MDA-MB-468 cytoplasmicHeLa nuclear

HeLa cytoplasmic

+------

+-+----

+---+--

+--+---

+-----+

+----+-

++-----

1 4 5 6 72 3

RPC

FreeProbe

STR7 ProbeHeLa nuclear

Cold STR7 ProbeCold pBlue

tRNA

1 2 3 4 5 6 7

++---

+++--

++-

++-

++--+

++--

++

++--

+++

++

++--

FreeProbe

RPC

A

Figure 1.7 A, B: Breast cancer cells contain proteins that specifically target SRA STR7.Panel A shows a comparison of nuclear and cytoplasmic MCF-7, HeLa and MDA-MB-468 extract binding to SRA STR7 via REMSA. Panel B shows REMSA with SRA STR7showing reduced complex formation with unlabeled “cold” competitor STR7, but notexcess cold pBlue vector or tRNA. RPC, RNA-protein complex. +-+++, increasing coldcompetitor RNA. Both figures are taken from Hatchell et al. (2006, p. 659).

B

46

FreeProbe

++--

+-+-

+--+

+---

STR7 Probe MCF-7 nuclear

MDA-MB-468 nuclearHeLa nuclear

220

97.4

66

46

30

21.5

*

2 3 41 Mr (kDa)C

Figure 1.7 C, D: Identifying SRA RNA-binding proteins. Panel C: UV cross-link assaywith nuclear extracts from cell lines in Panel A and B. Cell extract (30 µg) was incubatedwith labeled STR7, UV irradiated, RNase A digested, resolved by SDS-PAGE anddetected by PhosphorImager after transfer to PVDF membrane. [14C]-molecular weightmarkers used as size standards. Arrows highlight 39 & 40 kDa RPCs in MDA-MB-468cells. (Taken from Hatchell et al., 2006, p. 659). Panel D: After determining that SRASTR7 was a target for RNA-binding proteins, a yeast III hybrid screen of a humanprimary breast cancer cell library was performed, and is depicted in D, a schematicdiagram adapted from Yeap (2000). There are three plasmids involved: the bait RNA(SRA STR7), the human breast cancer library, and the MS2 coat. The hybrid RNA bindswith high affinity ot the MS2 bacteriophage of the hybrid protein 1. When ‘prey’ from thebreast cancer cDNA library bind (via RNA-binding domains) to STR7, the activationdomain is brought near enough o the promoter resulting in transcription of the reportergene. From this screen we identified SLIRP.

Lex A op HIS3 reporter

Lex A

Human breastCancer library

Gal4 activation domain

Reporter genetranscription

MS2

Bait RNASTR7

MS2Coat

D

47

48

E. Hatchell Chapter 2: Materials and Methods

49

Chapter 2

Materials & methods


50

2.1 MATERIALS

2.1.1 Reagents

Cell culture

22RV1 cell line American Type Culture Collection, Rockville, MD, USAC2C12 cell line American Type Culture Collection,

Rockville, MD, USACO57 cell line American Type Culture Collection,

Rockville, MD, USACalu-6 cell line American Type Culture Collection,

Rockville, MD, USAHeLa cell line American Type Culture Collection, Rockville, MD, USAHT1080 cell line American Type Culture Collection,

Rockville, MD, USAHMEC cell line Dr. Roger Reddel, Westmead, Sydney, AustraliaJ2E cell line American Type Culture Collection,

Rockville, MD, USALNCaP cell line American Type Culture Collection, Rockville, MD, USAMCF-7 cell line American Type Culture Collection,

Rockville, MD, USAMDA-MB-468 cell line American Type Culture Collection,

Rockville, MD, USANIH-3T3 cell line American Type Culture Collection,

Rockville, MD, USAOD9DL cell line American Type Culture Collection,

Rockville, MD, USAPC3 American Type Culture Collection,

Rockville, MD, USASKBR-3 cell line American Type Culture Collection,

Rockville, MD, USAT47-D cell line American Type Culture Collection,

Rockville, MD, USA17β-estradiol Sigma, AustraliaCellTitre Aqueous One Solution cell proliferation assay reagent Promega, AustraliaDexamethasone Sigma, AustraliaDimethylsulphoxide (DMSO) Sigma, AustraliaDulbecco's modified eagle medium/F12 Invitrogen, Australia (DMEM/F12) Fetal calf serum Invitrogen, AustraliaGW501516 Axora, USAICI182780 Tocris, USAInsulin Sigma, AustraliaOptiMem Media Invitrogen, AustraliaPhosphate Buffered Saline Invitrogen, Australia


51

Penicillin Gibco, AustraliaRPMI Media Invitrogen, AustraliaStreptomycin Invitrogen, AustraliaTamoxifen Sigma, AustraliaThyroid hormone Sigma, AustraliaTrypsin Invitrogen, AustraliaVitamin D Sigma, Australia

Enzymes and appropriate buffers

BamHI Promega, AustraliaBuffer C Promega, AustraliaBuffer H Promega, AustraliaCalf intestinal alkaline phosphatase Promega, AustraliaCalf intestinal alkaline phosphatase buffer Promega, AustraliaDNA ligase Promega, AustraliaDNA ligase buffer Promega, AustraliaEcoRI Promega, Australia

Ribonuclear probes and cold probes

32P Amersham AustraliaAcrylamide 19:1 BioRad, AustraliaAmmonium persulphate BioRad, AustraliaBromophenol blue BioRad, AustraliaDNase RQ1 Promega, AustraliaDNase-1 Invitrogen, AustraliadNTPs (dATP, dTT, dGTP, dCTP) Promega, AustraliaDithiothrietol Invitrogen, AustraliaFormamide BioRad, AustraliaNICK column Pharmacia Biotechnology, AustraliaRNasin Invitrogen, AustraliaT3 RNA polymerase Invitrogen, AustraliaT7 RNA polymerase Invitrogen, AustraliaTEMED BioRad, AustraliaTranscription buffer x 5 Invitrogen, AustraliaXylene cyanol Sigma, Australia

REMSA & UV cross-linking

β-mercaptoethanol BHD Biochemicals, England14C Rainbow molecular weight marker Amersham, AustraliaAcrylamide 30 % 19:1 BioRad, AustraliaAcrylamide 30 % 29:1 BioRad, AustraliaHeparin Sigma, USAHepes BHD Biochemicals, EnglandMgCl2 BHD Biochemicals, EnglandNP-40 BHD Biochemicals, EnglandRNase A Roche, Australia


52

RNase T1 Roche, USAtRNA Sigma, Australia

RNA interference

GFP control siRNA Dharmacon, USALipofectamine 2000 Invitrogen, AustraliaNon-targeting control SMARTPOOL siRNA Dharmacon, USASLIRP individual siRNAs Dharmacon,USASLIRP SMARTPOOL siRNA Dharmacon, USA

Antibodies (and concentrations)

Anti-mouse Ig Chemicon, Australia 1/10,000Anti-rabbit Ig Amersham, Australia 1/10,000β-actin antibody Abcam, Australia 1/10,000Cytochrome c antibody Santa Cruz, USA 1/10,000ERα antibody Santa Cruz, USA 1/10,000FLAG antibody Santa Cruz, USA 1/5,000HSP-60 antibody Santa Cruz, USA 1/5,000HuD antibody Santa Cruz, USA 1/5,000HuR antibody Santa Cruz, USA 1/5,000GR antibody Santa Cruz, USA 1/10,000N-CoR antibody Santa Cruz, USA 1/5,000SKIP antibody Gift from G. Leong 1/1,000SLIRP antibody Developed in-house 1/250SRC-1 antibody Gift from B. O’Malley 1/1,000

Transient transfection assays

Fugene9 Invitrogen, AustraliaMitotracker Molecular Probes, USA

Antibiotics

Ampicillin Roche, AustraliaChloramphenicol Roche, AustraliaTetracycline Roche, Australia

Bacterial culture & cloning

Bacto yeast extract Oxiod, EnglandBactotryptone Oxiod, Englandglutathoine-sepharose beads Sigma, AustraliaIsopropyl-β-D-thiogalactoside (ITPG) Roche, AustraliaPreScission protease Amersham, AustraliaReduced glutathione Sigma, Australia


53

PCR reagents

1 Kb Plus DNA ladder Invitrogen, AustraliaBuffer 10 x PCR polymerase Invitrogen, AustraliadNTPS Invitrogen, AustraliaMgCl2 Invitrogen, AustraliaRandom primers Invitrogen, AustraliaTaq polymerase Invitrogen, Australia

Transfection plasmids and vectors

3xFLAG-CMV9 Sigma, AustraliaAR E. Wilson, North Carolina UniversityARE-Luc E. Wilson, North Carolina UniversityER R. Lanz, Baylor CollegeERE-Luc R. Lanz, Baylor CollegeGR R. Lanz, Baylor CollegeGRE-Luc R. Lanz, Baylor CollegePPARδ G. Muscat, University of QueenslandPPAREδ-Luc G. Muscat, University of QueenslandpSCT R. Lanz, Baylor CollegepSCT-SRA R. Lanz, Baylor CollegeTR E. Ingley, WAIMRTRE-Luc E. Ingley, WAIMRVR G. Muscat, University of QueenslandVDRE-Luc G. Muscat, University of Queensland

DNA primers

(Refer to table 2.1A for sequences)

STR7 sense GeneWorks, AustraliaSTR7 reverse GeneWorks, AustraliaSLIRP sense GeneWorks, AustraliaSLIRP reverse GeneWorks, AustraliaSDM7 sense GeneWorks, AustraliaSDM7 reverse GeneWorks, AustraliaSKIP sense GeneWorks, AustraliaSKIP reverse GeneWorks, AustraliaFLAG sense GeneWorks, AustraliaFLAG reverse GeneWorks, AustraliapGEX sense GeneWorks, AustraliapGEX reverse GeneWorks, AustraliaSRAIEV sense GeneWorks, AustraliaSRAIEV reverse GeneWorks, AustraliapS2 sense GeneWorks, AustraliapS2 reverse GeneWorks, AustraliaMetallothione sense GeneWorks, Australia


54

Metallothione reverse GeneWorks, Australia

Protease inhibitors

Aprotinin Roche, AustraliaBenzamadine Sigma, USADithiothreitol (DTT) Roche, AustraliaLeupeptin Roche, AustraliaPhenylmethylsulfonyl fluoride (PMSF) Roche, Australia

2.1.2 General reagents

Acetic Acid (glacial) BDH Biochemicals, England, UKAgarose Scientifix Pty Ltd AustraliaAlbumin (Bovine) Sigma Chemical Company, UKAmmonium sulphate Fluka Chemika, SwitzerlandAntioxidant Invitrogen, AustraliaBaxter Water Baxter Healthcare Pty Ltd AustraliaBis-Tris Precast Gels (10-12 %) Invitrogen, AustraliaBovine serum albumin Sigma Chemical Company, USAChloroform BDH Biochemicals, England, UKDiethyl pyrocarbonate Sigma Chemical Company, USAEthanol (absolute) BDH Biochemicals, England, UKEthidium bromide Sigma Chemical Company, UKGlass beads Sigma Chemical Company, UKGlucose Sigma Chemical Company, UKGlycerol Sigma Chemical Company, UKHydrochloric acid BDH Biochemicals, England, UKIsoamylalcohol BDH Biochemicals, England, UKIsopropanol BDH Biochemicals, England, UKMagnesium chloride Sigma Chemical Company, UKMES buffer Invitrogen, AustraliaMethanol BDH Biochemicals, England, UKMOPS buffer Invitrogen, AustraliaMulti Colour Marker Invitrogen, AustraliaNonidet P40 Shell, AustraliaNuPage antioxidant Invitrogen, AustraliaNuPage LDS sample buffer Invitrogen, AustraliaNuPage MES SDS running buffer Invitrogen. AustraliaNuPage MOPS SDS running buffer Invitrogen, AustraliaNuPage transfer buffer Invitrogen, AustraliaPEG 3350 (40 %) Fluka Chemika, SwitzerlandPhenol Wako Pure Chemical Industries, JapanPotassium chloride Sigma Chemical Company, USAProtein A beads Pharmica, AustraliaProtein assay dye BioRad, AustraliaProtein G beads Sigma, AustraliaRainbow Markers Amersham, AustraliaRNAse A Molecular Probes, USARNAse T1 Molecular Probes, USA


55

Sheared salmon sperm Roche, AustraliaSodium acetate BDH Biochemicals, England, UKSodium bicarbonate Sigma Chemical Company, UKSodium chloride Sigma Chemical Company, UKTEMED Biorad, AustraliaTris base Sigma Chemical Company, USATriton X-100 BioRad, AustraliaTrizol Invitrogen, AustraliaUrea Sigma Chemical Company, USAVectasheild Vector Laboratories, USA

2.1.3 Consumables

6 well plate Greiner Bio-One, Germany96 well plate Greiner Bio-One, Germany175 cm2 Tissue Culture Flask Greiner Bio-One, GermanyCell Scraper Startedt, FranceCryotubes Nunc, DenmarkGloves Kimberly Citerile, USAMedia Filter Sartolab, USANick Columns Amersham, AustraliaNuPage 10 % Bis Tris Precast Gel Invitrogen, AustraliaPipettes (1, 5, 10 & 25 mL) Starstedt, France

2.1.4 Commercial kits

ECL Plus Kit Amersham, AustraliaEcoRI and BamHI Restriction Enzymes Promega AustraliaInvitrogen MidiPrep Kit Invitrogen, AustraliaQIAGEN PCR Purification System Kit Qiagen, AustraliaQIAGEN Plasmid Maxi Preparation Kit Qiagen, AustraliaSilver Stain Plus Kit Biorad, AustraliaRNeasy Kit Qiagen, AustraliaTaq DNA Polymerase Invitrogen, AustraliaWizard SV Gel and PCR Clean Up Kit Promega, Australia

2.1.5 Equipment

Agarose gel tanks Biorad, AustraliaBiological safety cabinet class II Clyde-AlacBeckman avatiTM J-301 centrifuge Beckman Coulter, USABeckman avanti™ J-251 centrifuge Beckman Coulter, USAChemidoc XRS Documentation System Biorad, AustraliaCO2 incubator Sanyo, JapanCR312 centrifuge Jouan, FranceCV2 microscope Olympus, JapanDigital Camera Nikon, JapanDeveloper CP1000 Agfa, USADry block heater Thermolyne Scientific Equipment,


56

AustraliaEppendorf centrifuge 5415C Eppendorf, GermanyFalcon tissue culture flasks Becton Dickinson, USAFLUOstar OPTIMA BMG laboratories, AustraliaFume hood BH2000 Cyde-Apac, AustraliaGallenhamp 37oC shaker Rowe Scientific, AustraliaGel dryer model 583 Biorad, AustraliaGel electrophoresis tank BioRad AustraliaGene Pulser BioRad, AustraliaHeidolph MR 1000 magnetic stirrer John Morris Scientific, AustraliaHot water bath Grant Instruments, EnglandIncubator shaker Bioline, AustraliaLiquid scintillation counter Pharmacia, USAMini-protean electrophoresis and transfer BioRad, AustraliaMolecular Imager FX Biorad, AustraliaNanoDrop BioLab, AustraliaNuPage PreCast gel electrophoresis system Invitrogen, AustraliaPerkin Elmer DNA thermo cycler Perkin Elmer, EnglandPhosphorImager 445 Si Molecular Dynamics, USAPhosphorImager (new) Biorad, AustraliaPlatform rocker Bioline, AustraliaPower Pac™ 1000 BioRad, AustraliaQuantity One 1-D analysis software BioRad, AustraliaRatek water bath Scott Scientific, AustraliaSanyo –85oC Freezer Qantam Scientific, AustraliaScintillation Counter Beckman, AustraliaSonicator 250 Branson, USASpectrophotometer UV / Vis Perkin Elmer, USASpeedvac system ISSIIO Thermo SavantStratagene ®UV crosslinker Stratagene, USAThermocycler MJ PTC-200 Peltier, USAUV transilluminator Hoefer, Australia

2.1.6 Buffers and solutions

Please see alphabetical listing in Appendix I.


57

2.2 Methods

Throughout the methods section of this Chapter, buffers and solutions are numericallysuperscripted; these numbers correspond to the buffers and solutions listed in Appendix1.

2.2.1 Plasmid constructs

SLIRP plasmid constructs, sequences and vectors

SLIRP coding domain was amplified from the original yeast three hybrid clone(Hatchell, 2002) with SLIRP (sense) 5’ cgc gga tcc gcg gcc tca gca gca 3’, SLIRP(reverse) 5’ gcg cgg atc cta ggc tgc agt ctca 3’ primers and subcloned into BamHI cutpCMV-FLAG 7.1 for transfection assays and pGEX-6P2 for REMSA. STR7oligonucleotides (sense) 5’ agg agg cag gta tgt gat gac atc agc cga cgc ctg gca ctg ctgcag gaa cag tgg gct gga gga aag ttg tca ata cct gta aag aa 3’ and (reverse) 5’ ttc ttt acaggt att gac aac ttt cct cca gcc cac tgt tcc tgc agc agt gcc agg cgt cgg ctg atg tca tca catacc tgc ttc ct 3’ were cloned into EcoRV-digested pBluescript II KS+ (Stratagene) togenerate labeled riboprobes. To generate SRA SDM7, underlined residues above weremutated to tcc, ctc and ctc respectively as previously outlined (Lanz et al., 2002).pBluescript vectors were linearized and riboprobes generated as described (Thomson etal., 1999). SLIRP mutants (R24,25A and L62A) were prepared by Ross McCulloch(Royal Perth Hospital, Perth, Australia) by PCR-based mutagenesis and subcloned intoBamH1-cut pCMV-FLAG 7.1 and pGEX-6P2 as above.

2.2.2 Bioinformatics

Searches were performed for potential protein matches via the NCBI interactive webhomepage (http://www.ncbi.nlm.nih.gov/), using several different BLAST programs. Allsequences were translated at the ExPasy website: (http://www.expasy.org/tools/dna.html). ThemFold program can be found at Michael Zuker’s mFold server web page:(http://bioinfo.math.rpi.edu/~mfold/rna/form1.cgi). All searching via the Celera database wasperformed via this website: (http://www.celera.com). All plasmid sequences were alignedusing “BLAST 2 Sequences” at the NCBI website. All sequence alignments of SLIRPand SHARP and SLIRP inter-species comparisons were performed using algorithmsdescribed by Stothard (2000) and Chenna et al. (2003).

2.2.3 Polymerase chain reaction assays and primer design

Polymerase chain reactions

Polymerase chain reactions (PCR) were performed on a Perkin Elmer DNA ThermoCycler. A 20 µL reaction contained: 1 µL (1 µg/µL) DNA; 1 µL (50 ng/µL) senseprimer; 1 µL (50 ng/µL) anti-sense primer; 2 µL 10 x buffer42; 0.5 µL 10 mM dNTPs17;0.2 µL Taq (1 U); 0.6 µL 50 mM MgCl2; 11.4 µL sterile Baxter water. The sameprocedure was followed for PCR on bacterial DNA except that instead of adding of 1µL of DNA, one colony was added directly to the mix. PCR was performed with a hotstart at 95°C, 30 cycles of 95°C 1’, 55°C 1’, 72°C 2’ and a long elongation period of


58

72°C 10’ on the PCR machine. The samples were then run on 0.8 %1 or 1.2 %2 agarosegels, in 1 x TAE76 with DNA loading dye15, to confirm the PCR results.

Primers

All primers were ordered via Geneworks, and reconstituted in sterile Baxter water to 1µg/µL. All primers and primer dilutions were stored at –20°C. A list of primers usedcan be seen in Table 3.1.

2.2.4 Cloning

Cloning into expression vectors

One cDNA isolated from the yeast III hybrid screen was selected for further analysis. Itwas sub-cloned into the pGEX-6P2 vector to generate recombinant fusion protein forprotein studies, and the pCMV-FLAG 7.1 vector for eukaryotic cell transfection (Table3.2). Primers were designed for SLIRP in which a BamHI restriction digest site addedat the 5’ and 3’ ends. After a hot start at 95°C 5’, 30 cycles of 95°C 1’, 55°C 1’, 72°C2’ were performed followed by an elongation period of 72°C 10’. The resultant DNAwas visible on an agarose (0.8 %) gel1. The DNA was purified via the PCR PurificationSystem Kit. Both vectors and the SLIRP PCR product were cut with BamHI in anovernight digest consisting of 10 µg DNA, 4 µL restriction enzyme, 10 µL buffer9,made up to 100 µL of sterile water at 37°C. The DNA was again purified via the PCRPurification System Kit. Both vectors were dephosphorylated with 0.5 U calf intestinalalkaline phosphatase (CIAP) at 37°C for 30 min. A further 0.5 U CIAP was added toeach and left another 30 min at 37°C followed by the addition of EDTA22 to 5 mM at atemperature of 75°C for 10 min to inactivate the CIAP. The DNA was purified with thePCR Purification System Kit as described previously and resuspended in 20 µL TE73.SLIRP was then ligated overnight at room temperature to each of the digested vectors(7 µL SLIRP DNA, 1 µL vector, 1 µL DNA ligase, 1 µL ligase buffer42), andtransformed into E.coli XL2-Blue cells. Bacterial PCR was then performed to confirmpresence of the SLIRP insert (of the correct size) of several transformants. Maxi-preparations were then generated of two positive clones from the pGEX-6P2 ligationsand five of the p3XFLAG-CMV9 ligations to validate orientation and sequence of theinsert.

2.2.5 Bacterial cell culture

Generating electro-competent cells

A 2 mL low salt LB44 starter culture containing tetracycline74 (12.5 µg/mL) and onesterile loop of appropriate bacteria cells (E.coli XL2-Blue, E.coli BL21 Codon +) weregrown at 37°C shaking (170 rpm) for 8 hours. A 10 mL LB44 (with 12.5 µg/mLtetracycline74) culture was inoculated with the 2 mL culture and grown under the sameconditions overnight. The next day, 1 L low salt prewarmed LB44 was inoculated withthe overnight culture and grown to an OD600 of 0.5 ~ 0.6. The flask was chilled on ice at4°C for 15 - 30 min to inhibit further growth. The cells were then pelleted by spinningat 4000 x g for 15 min at 4°C, and washed repeatedly (four times) in decreasingamounts of sterile ice cold 10 % glycerol29. The last sample was resuspended in 3 mL


59

ice cold 10 % glycerol29 and the bacteria cells were aliquoted into 50 µL samples, snapfrozen in liquid nitrogen and stored at –80 °C.

Transformation into bacterial cells

BioRad GenePulser 0.2 cm cuvettes were cooled on ice for 15 min and vials (50 µL) ofcompetent cells defrosted on ice. Sample DNA (1 - 2 µL) was added to the competentcells and mixed gently. The mix was then added to the bottom of the pre-cooled cuvettewith a sterile pipette, ensuring that the liquid was evenly spread between the twoelectrodes, and returned to ice. The BioRad GenePulser machine was set (200 mAresistance, 25 capacitance, and 2.50 Kvolts), following which the cells wereelectroshocked and then 1 mL SOC65 media was added to the cuvette immediately. Thebacterial cells were transported to a sterile 10 mL tube and placed in a 37°C water bathfor 1 hour. Following incubation the cells were plated onto appropriate antibiotic platesand incubated over night at 37°C. E. coli XL2-Blue cells were plated onto (50 µg/mL)ampicillin plates, and E. coli BL21 Codon + cells were plated on to (12.5 µg/mL)tetracycline74, (100 µg/mL) chloramphenicol11 and (50 µg/mL) ampicillin5 plates40.

Maxi-preparations of plasmid DNA

Large scale plasmid DNA preparations were made using QIAGEN Plasmid MaxiPreparation Kits as per the manufacturers instructions, whereby the bacterial cellculture containing plasmids were pelleted, lysed, the plasmid DNA bound to a columnand washed, and then finally eluted into sterile Baxter water. The DNA concentrationwas always checked on an 0.8 % agarose gel1 with DNA loading dye15 prior tosequencing. For sequencing, 10 µL of 100 ng/µL DNA and 5 µ L of 1 pmol ofappropriate primer was sent to an automatic sequencer at the DNA Sequence Service,Department of Clinical Immunology at Royal Perth Hospital. Using Big DyeTerminator 3.1 Chromatgrams were checked by eye and corrected when possible. Allplasmid sequences were aligned using “BLAST 2 Sequences” at the NCBI website.

2.2.6 GST fusion protein production

Large scale protein expression

The pGEX-6P2-SLIRP hybrid was transformed into the electrocompetent E.coli BL21Codon + cells and grown on antibiotic selective plates40. Large scale protein expressionwas performed at temperatures of 30°C and/or 37°C. LB39 (500 mL) with (12.5 µg/mL)tetracycline74, (100 µg/mL) chloroamphenicol11 and (50 µg/mL) ampicillin5 wasinoculated with an overnight 50 mL culture of either pGEX-6P2 (GST alone), or thehybrid GST-DC50-pGEX-6P2. The cultures were grown at 37°C shaking at 170 rpm toan OD600 ranging between 0.6 – 1. The cultures were cooled for 5 min on ice to inhibitgrowth, then induced with 0.1 or 0.4 mM isopropyl-β-d-thiogalactoside (IPTG37) andplaced at the desired temperature for 2 - 3 hours while shaking. Samples of 500 µLwere taken at various time points after induction, centrifuged, resuspended in 25 µLBaxter water and stored at –20 °C. After the induction period the bacterial cells werecentrifuged at 3000 x g for 15 min at 4 °C. The pellet was resuspended in 10 mLNETN48 (always with final concentrations of protease inhibitors as follows 1 mMDTT14, 0.5 mM PMSF50, 10 µg/µL leupeptin41, 2 µg/µL aprotinin6, and 5 mMbenzamadine7 and antibiotics as previously). Following resuspension, 20 mL 0.2 %


60

Triton X-10080 was added to lyse the cells, and the samples sonicated 3 x 30 seconds onice. The sonicated supernatant from the large scale protein preparations was incubatedwith 30-50 mg of pre-swelled Glutathione-Sepharose beads including proteaseinhibitors for 30 min at 4°C on a rotating wheel. The beads were then washed six timesat 3000 rpm (for 5 min at 4°C) with 10 ml NETN48 including protease inhibitors. Afurther 2 mL NETN48 (with protease inhibitors) wash was performed and thesupernatant was removed. For elution, 500 µL GST elution buffer31 was added to theglutathione beads and the samples were incubated at 4°C for 30 min on a rotatingwheel. The samples were centrifuged at 14,000 x g for 5 min at 4°C and the elutedprotein was collected in the supernatant. Where the proteins were cleaved, the beadswere washed with 10 times the bed volume with GST cleavage buffer30, this wash wasrepeated six times at 3000 x g, for 5 min at 4°C. The bead slurry was pelleted, and anequal amount of PreScission Protease cleavage buffer57 added and the bed slurryresuspended. The mix was placed on the rotating wheel for 4 hours at 4°C, thencentrifuged at 12,000 x g for 5 min at 4°C. The supernatant containing the cleavedprotein was stored in a screw cap eppendorf at 4°C. All samples were run on a SDS-PAGE (8-15 %) gel53-56, at 30 – 40 mA for 2 - 4 hours. All gels were silver stainedimmediately using the Silver Stain Plus Kit.

2.2.7 RNA analysis

RNA extraction

Trizol reagent was used to extract RNA from cells as described in the manufacturer’sprotocol. Medium was aspirated from confluent cells in 9 cm2 (6-well) tissue cultureplates, 1 mL of Trizol was added per well and lysis performed for 5 min at roomtemperature. The plates were scraped and the contents collected in sterilemicrocentrifuge tubes. Samples were processed to completion immediately or stored at-80°C. To precipitate protein contaminants, 200 µL of chloroform was added to eachsample which was then shaken vigorously by hand for 15 sec before being centrifugedat 12,000 x g for 15 min at 4°C. The aqueous phase containing the RNA was thenremoved to a new tube, 500 µL of isopropanol then added and the RNA precipitated for10 min at room temperature. The RNA was pelleted by centrifugation at 12,000 x g for10 min at 4°C, the supernatant was removed and discarded and the pellet washed with 1mL of 75 % ethanol26 in diethyl pyrocarbonate (DEPC)-treated ddH2018. The tubes wereflicked to mix and re-centrifuged at 7,500 x g for 5 min at 4°C. The ethanol wasremoved, the pellet air-dried and resuspended in 20 µL of DEPC-treated ddH2018.

Northern analysis

Human tissue blot obtained from BD Biosciences. Total RNA isolated using Trizol,Poly A RNA using Poly ATtractTM. Samples (2 µg) were electrophoresed andtransferred to Hybond-N membrane. Human SLIRP, SKIP or β−actin cDNAs wererandom primer labeled ([32P]dCTP) and hybridized with membranes for 2 - 5 h, washedin 0.1 x SSC / 0.1% SDS at 65°C and visualized by PhosphoImager. Northerns wereperformed by Shane Colley (Western Australian Institute for Medical Research, Perth,Australia).


61

Bioinformatic Analysis of STR7

Substantial analysis of the predicted structure of SRA STR7 within the entire SRARNA was performed. This required considerable structural analysis to determine theoptimal orientation (with the lowest free energy [–kcal/mol]).

2.2.8 REMSA and UVXL experiments

Linearizing SL7 for riboprobe synthesis

This reaction was performed on several occasions, exactly as described below.Maxiprep QIAGEN SL7 pBluescript (22.75 µg/µL) DNA (175 µL) was incubatedovernight at 37°C with 20 µL Buffer H9 and 5 µL EcoRI. This mixture was purifiedwith the PCR Purification Kit and re-suspended in 20 µL TE73, resulting in a finalconcentration of 1 µg/µL. Confirmation of the digestion of SL7 RNA was provided byassessment of the reaction via 0.8 % agarose gel1 electrophoresis.

Generation of 32P labelled riboprobes

REMSA and UVXL were performed as described (Thomson et al., 1999). Incompetition REMSA, up to 100 fold excess unlabeled pBlue, STR7 or tRNA was used.Large scale recombinant GST protein expression was performed as described (Giles etal., 2003). Recombinant proteins were digested with PreScission Protease. Thisprotocol is essentially as described in Thomson et al. (1999). Initially, 10 µL ofradioactive [α-32P] UTP (37 Ci/mmol) was vacuum dried for 15 min in a screw capeppendorf tube. The following was then added to the dried 32P pellet: 1 µL (1 µg/µL)linearized SL7; 2 µL x 542 transcription buffer; 2 µL rNTPs (2.5 mM; without UTP62); 1µL DTT14 (0.1 M); 2 µL DEPC treated water18; 1 µL RNAse In; and 1 µL appropriateRNA polymerase (T3 for SL7, T7 for pBlue). The mix was briefly centrifuged at12,600 x g, then incubated at 37°C for 60 min. DNAse-1 (1 µL) was added and thesample incubated at 37°C for 10 min, immediately followed by 65°C for 5 min. Stopsolution70 (12 µL) was added, followed by incubation at 80°C for 3 min. A 6 % (5M)urea gel82 was preelectrophoresed at 200 V for 30 min at room temperature in 1 xTBE76 and the labelled transcripts were loaded and separated on the gel (20 min at 200V). The wet gel was exposed to an x-ray film, the appropriate transcript was cut out ofthe gel, eluted with 500 µL riboprobe elution buffer61, precipitated, washed (with cold70 % ethanol25) and resuspended in sterile DEPC water18. The riboprobe was stored at–80°C for up to a week and had a specific activity of ~ 2x108 cpm/µg RNA.

Generating ribonucleotide cold probes

Unlabelled (cold) RNA probes were prepared as described above except the [α-32P]UTP was replaced with unlabelled 2.5 mM uridine 5’ triphophatase (UTP). Cold probeswere always made as follows: 2 µg linear DNA, 20 µL 5 x transcription buffer42, 20 µLrNTPs (2.5 mM; with UTP63), 10 µL DTT14 (0.1 M), 43.5 µL DEPC18 water, 2.5 µLRNAsin, 2 µL RNA polymerase (T3 for SL7 as it is in the antisense orientation and T7for pBlue). The mix was incubated at 37°C for 60 min. DNAse RQ1 (2 µL) was thenadded prior to a further 10 minute incubation at 37°C, followed by a 5 minute 65°Cincubation. The samples were stored on ice and then either subjected to NICK column


62

purification (where the sample is run through a filter following manual directions) or aphenol isoamylalcohol chloroform (PIC52) extraction followed.

PIC extraction

In a PIC extraction, equal volume of PIC52 was added to the sample, mixed, spun andthe supernatant removed. The same volume of chloroform was then added to thesample, the solution mixed and spun and the resultant supernatant removed to a freshtube. For both NICK column purification or PIC extracts the DNA was precipitatedwith the addition of 21/2 times the volume of ethanol (100 %) and 1/10 the volume of 5 Msodium acetate (pH 5.5)66 at –80°C for 2-16 hours, washed and resuspended in DEPCwater. Samples were run on a 6 % (5M) urea gel82 for 20 min, then stained withethidium bromide28 for 20 min. Cold probes generated in this way were quantifiedagainst known standards.

RNA ElectroMobility Shift Assay (REMSA)

The protocol used was essentially as described in Thomson et al. (1999). The riboprobewas thawed and incubated at 70°C for 10 min, followed by a 10 min incubation at roomtemperature (22ºC). Cytoplasmic extracts, or fusion proteins (Figure 2.1 A), werethawed on ice, 5 – 10 µg of protein mixed with CEB12 to a volume of 8 µL, wasaliquoted to each screw cap eppendorf. Following thawing, denaturation andrenaturation, 50,000 cpm of 32P labelled riboprobe was added to each reaction and themix incubated at room temperature (22ºC) for 30 min. In RNA competition REMSAassays, excess (50-150 fold) unlabelled cold sense RNA transcript (SL7 or pBlue) wasfirst denatured at 70°C then cooled to room temperature as with the hot probe. The coldprobe was then incubated with the extract for 30 min, at 22°C prior to incubation forthe 32P labelled probe. RNase T1 (0.3 U) was added and incubated for 10 min at roomtemperature. Heparin (50 µg)35 was added and the solution further incubated for 10 minat 22ºC. Following the addition of 1 µL of RNA loading dye60, the samples were loadedonto a 5 % non-denaturing polyacrylamide mini-gel49 and electrophoresed at 200 V for20 min at 4ºC. Following electrophoresis REMSA gels were fixed in REMSA fixingsolution59, dried and exposed to phosphorImage screen overnight and visualised byPhosphorImager.

UV cross-linking

The protocol used was as described in Thomson et al. (1999). The riboprobe wasthawed and then denatured at 70 °C for 10 min followed by a 10 min incubation atroom temperature. Concurrently, 15 – 20 µg of protein was thawed and made up withCEB12 to 15 µL in screw cap eppendorfs. Following thawing, denaturation andrenaturation of the 32P probe, 200,000 cpm of riboprobe was added to each reaction andthe mix incubated for 30 min at room temperature. RNase T1 and heparin35 were addedto the same final concentration as outlined in previously. The samples are transferred toa 96 well plate and UV-irradiated 1 cm from the UV source for 15 min on ice using aStratagene ®UV Crosslinker (3 x 105 µJ, 254 nm bulbs). RNase A (10 µg) was added toeach and the mix incubated at 37°C for 15 min. Equal volumes of 2 x SDS buffer67 wasadded (including 4 µL/100 µL of fresh β-mercaptoethanol) and the samples boiled for 3min prior to loading onto an 8, 10, 12.5 or 15 % SDS-PAGE gel53-56. A 15 cm x 15 cmgel was run for 16 hours at 8 mA; 14C Rainbow molecular weight markers were used as


63

size standards. The gel was fixed for 10 min in UV-cross linking fixer solution53, dried,and exposed to a phosphorImage screen for 2 days and visualised by PhosphorImager.

2.2.9 Tissue culture

Cell culture

MCF-7 (ATCC HTB 22), MDA-MB-468 (ATCC HTB 132), SK-BR-3 (ATCC HTB30), T47D (ATCC HTB 133), LNCaP (ATCC CRL 1740), PC-3 (ATCC CRL 1435),HepG2 (ATCC HB 8065), Calu-6 (ATCC HTB 156), HT1080 (ATCC CCL 121),COS-7 (ATCC CRL 1651), HeLa (ATCC CCL 2) cell lines were obtained from theAmerican Type Culture Collection and HMEC cells were a gift from Dr. Roger Reddel(Westmead, Sydney, Australia). Cells were grown as per supplier’s recommendationsor as previously described (Giles et al., 2003) and used within 20 passages of originalstock for all experiments. MCF-7 cells originate from a human breast cancer andrepresent an ER+ and EGFR- phenotype; MDA-MB-468 cells originate from a humanbreast cancer and have an ER- and EGFR+ phenotype; and HeLa cells originate from ahuman andenocarcinoma cancer an ER-, EGFR- phenotype (Brooks et al., 1973;Lippman & Bolan, 1975; Maminta et al., 1991; Levenson & Jordan, 1997). Cells weregrown at 37°C and 5 % CO2 in Dulbecco's modified Eagle medium/F12 (DMEM-F12)21 or RPMI64 medium. The medium contained (2.4 g/L) sodium bicarbonate, 10 %heat-inactivated fetal calf serum, 100 U/mL penicillin and 100 mg/mL streptomycin, ata pH ~7.35. MCF-7 cells were grown in 10 nM insulin36 to retain their estrogenresponsiveness (as per ATCC recommendations). Cells were grown to a confluency of80 % in 10 cm2 tissue culture dishes, prior to being trypsinized and split at a ratio of1:4. Cells to be stored for future use were pelleted and resuspended in 1 mL of 90 %heat-inactivated fetal calf serum and 10 % dimethylsulphoxide (DMSO) and stored incryotubes frozen under liquid nitrogen until needed. For experiments where low or nobackground levels of steroid hormones was required, cells were fed media treated with10 % charcoal stripped FCS20 24 hr prior to experiments.

Charcoal-stripped FCS

FCS (500 mL) was thawed overnight at 4oC and 2 mL were set aside for a biochemicalassay. The remainder FCS was incubated with dextran-coated charcoal whilst stirringfor 3 hr at room temperature. FCS was recovered by centrifugation 1,800 x g for 20min at 4oC. The supernatant was decanted and centrifuged again, FCS was then filtered.Charcoal-stripped FCS and FCS were assayed by Mario Taranto (Core and ClinicalBiochemistry, Royal Perth Hospital, Perth, Australia) for steroid hormones to ensureminimum hormone levels after stripping.

Cell extracts: cytoplasmic extracts

For each 10 cm dish of cells, the media was removed, the cells washed in ice coldphosphate buffered saline (PBS51) and then scraped off the dish and transferred to afalcon tube at a concentration of 15 plates per falcon tube. The cells were centrifuged at1800 x g for 3 min at 4°C and the PBS51 subsequently removed. Cytoplasmic extractionbuffer (CEB13) (150 µL) including protease inhibitors (see GST fusion protein protocolfor concentrations of protease inhibitors) was added to each tube and the solutionmixed well and stored on ice for 20 min. The samples were spun down for 5 min at12,000 x g at 4°C and the supernatant containing the cytoplasmic extract was collected


64

and stored at –80°C in aliquots of 10 and 100 µL. The remaining pellet was stored onice and used immediately for nuclear extracts. Each cell line was performed as above.If a whole cell protein extraction was necessary, mid-RIPA45 or passive lysis buffer (50µL per well of a 6 well plate) was used.

Cell extracts: nuclear extracts

Each pellet (produced from the 15 petri dishes of cells from above) from thecytoplasmic extraction was resuspended in 300 µL low salt buffer43 and 300 µL highsalt (added dropwise) buffer24. The mix was placed on a rotating wheel at 4°C for 30min and then centrifuged at 4°C, at 12000 x g for 30 min. Following the spin, thesupernatant was collected, aliquoted and stored at –80°C. All cell lines were processedthe same way.

Transfection and luciferase assays

HeLa and MCF-7 cells (150,000 cells / well) were transfected (Figure 2.1 B) in RPMI,containing 5 % stripped serum and with no additional hormone, using FuGENE6 withequal molar ratios of control empty and/or cDNA expression plasmids plus: 50 ng/wellERα, TRβ, AR, PPARδ; 0.6 µg/well ERE-Luc, GRE-Luc, TRE-Luc, ARE-Luc,VDRE-Luc, PPARE-Luc; 0.0625 µg/well pSCT or 0.08 µg/well pSCT-SRA, pSCT-SRA-SDM7a; 0.049 µg/well pCMV-FLAG-7.1 or 0.05-0.5 µg/well FLAG-SLIRP,FLAG-SLIRP-R24,25A, FLAG-SLIRP-L62A, FLAG-SLIRP-DM; 0.3 µg/well pCMXor 0.5 µg/well pCMX-SHARP; 100 ng/well pCGN or pCGN-SKIP. After transfection,cells were cultured for 24 h, prior to addition of E224, Dex19, DHT13 (10 nM), T371 (1nM), VitD84 (100 nM), GW50151632 (500 nM), Tam37 (1 µM), ICI35 (1 µM) or equalvolume of absolute ethanol or DMSO. After 8 h, lysates assayed for luciferase activityrelative to protein levels using Promega Luciferase and BioRad Protein Assays on aFluorStar Optima. Briefly, for measurement of luciferase activation, 20 µL sample and50 µL Luciferase Assay Reagent (LARII) were read on black 96 well plates. LARIIwas added by injection, and luminescence was read over 10 seconds by the FluorStarOptima. All transfection studies were performed in triplicate at least 3 times. (Of note,the core transfection - where SRA coactivation and SLIRP repression is assessed - wasperformed over 100 times.)

RNA interference

Cells were transfected with siRNA complexes directed against SRA, SLIRP, SKIP orcontrol (nonsense) (final concentration 20-200 nM) using Lipofectamine 2000. Toassess SLIRP knock down on GRE-luc activity, reporter plus pSCT or pSCT-SRA andsiRNA conjugates were added simultaneously. After 48 h, cells were induced withDex13 (10 nM) and lysates processed as described above. For SLIRP and SKIP siRNAstudies, cells were transfected with siRNA for 72 h, ERα and ERE-Luc co-transfectedat 48 h and E224 (10 nM) added 8 h prior to harvest. SKIP (sense) 5’-cat tca act ctg gagcta aac aga-3’ and (reverse) 5’-tcc gat cag caa tgt aga ggg ct-3’; SLIRP (sense) 5’-gcgctg cgt aga agt atc aa-3’ and SLIRP (reverse) 5’-tcg att ccg aag tcc ttc tt-3’; β−actin(sense) 5’-gcc aac aca gtg ctg tct gg-3’ and actin (reverse) 5’-tac tcc tgc ttg ctg atc ca-3’primers used to quantitated targets by RT-PCR. All RNAi transfections were performedin triplicate, and were performed at least three times.


65

Cell titration assay

The CellTiter 96 Aqueous One Solution cell proliferation assay kit was used to assesscell proliferation of MCF-7 cells after treatment with siRNA. Cells were plated at adensity of 2.5 x 104 cells/cm2 in 100 µL in 96-well plates, in triplicate, and transfectedat time of plating as described previously. To each well 50 µL of transfection mix wasadded, containing SLIRP or GFP control siRNA, to a final concentration of 50 nM. Toreduce handling and variation, media was not changed throughout the course of anexperiment. Cell proliferation was measured at 24 h intervals for 5 days by adding 20µL of CellTiter reagent, incubating under normal growth conditions for 1 h and readingthe absorbance at 490 nM using the Fluostar Optima platereader.

Western analysis

Cell lysates were resolved by SDS-PAGE and transferred onto PVDF Membrane.SLIRP was detected with an in-house rabbit polyclonal antisera raised against GST-SLIRP protein (1:250). SRC-1, SKIP, ERα and β-actin abs were used (all 1 in 10,000)with HRP conjugated anti-rabbit or anti-mouse secondary (1 in 10,000) abs and ECLPlus. All western blots were performed by the author except Figure 3.4 C and Figure3.5 where the SKIP and cross-species westerns were performed by Keith Giles(Western Australian Institute for Medical Research, Perth, Australia), and the SLIRPwestern was performed by Dianne Beveridge (Western Australian Institute for MedicalResearch, Perth, Australia).

2.2.10 Protein analysis

Protein concentration assay

Protein concentrations of cell lysates were determined using the Bio-Rad protein assay.A protein standard curve was generated each time samples were analysed, by dilutingbovine serum albumin (BSA)8 (2 µg/µL) to 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0,1.5, 2.0 and 5.0 mg/mL and adding 1 µL of each solution, in duplicate, to a 96-wellmicrotitre plate. In the same plate, 1 or 2 µL of each unknown sample was added induplicate. Bio-Rad Protein Assay Reagent was diluted 5-fold with ddH20 and 250 µL ofthis was added to each sample. The plate was then incubated for 5-10 min at roomtemperature before the absorbance at 590 nM was measured in the Fluostar Optimaplatereader. Concentrations of unknown samples were automatically calculated fromthe standard curve by the Fluostar Optima and Microsoft Excel software.

Polyacrylamide gel electrophoresis of proteins

Protein expression was studied using denaturing sodium dodecyl sulphatepolyacrylamide gel electrophoresis (SDS-PAGE). Polyacrylamide gels54 (10 % w/v)were prepared or NuPage 10 % Bis-Tris polyacrylamide were purchased. Equalmicrogram amounts (10-25 µg/lane) of extracts to be analysed were made up to anequal final volume with the buffer in which they had originally been prepared. Thesesamples were mixed in a ratio of four parts sample to one part Invitrogen LDS SampleBuffer and heated to 95°C for 4 min. After denaturing, the samples were centrifugedbriefly and cooled on ice. Samples were loaded into the gel, one lane included 5 µL ofKaleidoscope molecular weight marker, and electrophoresed at 35 mA per gel for 3


66

hours or until the dye front reached the bottom of the gel. Pre-cast gels wereelectrophoresed in 1 x MOPS-SDS or 1 x MES-SDS buffer and prepared gels were runin 1 x electrode buffer23.

Western blotting

After electrophoresis as described above, proteins were transferred onto PVDF-pluswestern blotting membrane. The membrane was dipped in methanol, rinsed in ddH20and pre-equilibrated in western transfer buffer85 (WTB). The gel, two pieces of filterpaper and two sponges were also pre-equilibrated in WTB85. The blotting apparatuswas assembled as follows: one sponge then one piece of filter paper were stacked onthe black side of the blotting cassette, then the gel was placed on the filter paper,followed by the PVDF-plus membrane, the other piece of filter paper and lastly thesecond sponge. The cassette was closed and proteins transferred to the membrane byelectroblotting overnight in WTB85 at 30 V at 4°C with a stirring bar.

Chemiluminescent detection of proteins

All incubations were performed at room temperature for at least 1 h, with rocking. ThePVDF-plus membrane with electroblotted proteins was blocked in 5 % skim milkpowder dissolved in Tris Buffered Saline (TBS)77. Following blocking, the membranewas incubated with primary antibody (SLIRP, HuR, Actin, SKIP), diluted in 1 % skimmilk powder dissolved in Tris Buffered Saline-Tween (TBS-T)78. Excess primaryantibody was removed by three 10 min washes in TBS-T78 before incubation withsecondary antibody diluted in 1 % skim milk powder dissolved in TBS-T78. Excesssecondary antibody was removed by three 10 min washes in TBS-T78. Enhancedchemiluminescence detection reagent was prepared in accordance with themanufacturer's instructions and gently poured onto the membrane. After five minutes atroom temperature, excess detection reagent was blotted off with filter paper and themembrane exposed to light-sensitive film or the image recorded by a charge-coupleddevice camera using a ChemiDoc XRS. Membranes were rinsed in TBS78, dried andstored at 4°C. If pictures were taken using the ChemiDoc XRS, western blots wereanalysed by densitometry, using Quantity One 1-D analysis software (Version 4.5.2,build 070).

2.2.11 In vivo analysis of proteins

Immunoprecipitation RT-PCR assay

Method is as described previously in the literature (Giles et al., 2003). Using MCF-7,MDA-MB-468 or HeLa cells with either SLIRP ab, SRC-1 ab, β-actin ab or no ab, co-immunopurifying. MCF-7, MDA468 or HeLa cells were grown to ~70 % confluence in175 cm2 flasks, washed with PBS, trypsinized, washed again with PBS and then lysedin 1 ml of CEB (see REMSA methods) on ice for 20 min. Lysates were centrifuged at11,500 rpm for 10 min and supernatant removed to fresh tubes. SLIRP antibody (20µl), PACT antibody (20 µl), Actin (3 µg) or no antibody were added to lysates andincubated for 60 min, at 4°C, on a rotating wheel. Fifty µl of a 50 % slurry of a mixtureof Protein A beads and Protein G beads (preswollen and equilibrated in CEB) wereadded to each sample and incubated for a further 60 min. After 4500 rpm for 2 minsupernatants were removed, pelleted beads were washed with cold CEB (5 times) andRNA was extracted using Trizol reagent. RNA was treated with RQ1-RNase-free


67

DNaseI to eliminate any genomic DNA and RT performed using random hexamers,Superscript II and standard procedures. PCR was performed for 40 cycles, withannealing at 55°C, with the primers: SRAEIV(F): SRAEIV(sense): 5’-tga tga cat cagccg acg cct-3’ and SRAEIV(reverse) 5’-gct gca gat ttc tct tca ttg-3’ flanking an intronsite within SRA cDNA to avoid any genomic DNA effect. PCR products were resolvedon ethidium bromide-stained 1.5 % agarose gels. Dianne Beveridge and Micheal Episperformed the SLIRP IP-RT-PCRs presented in this thesis (Western Australian Institutefor Medical Research, Perth, Australia).

Chromatin immunoprecipitation (ChIP) assay

Assay was performed as previously described (Dowhan et al., 2005). MCF-7 and HeLacells were treated with 100 nM E224 for 0 - 120 min or Dex19 100 nM for 15 min priorto fixation. Soluble chromatin incubated with 4 µg ERα, SLIRP polyclonal, HuD,SRC-1, N-CoR or GR abs. Recovered DNA fragments amplified with either pS2(sense) 5’-ggc cat ctc tca cta tga atc act tct gc-3’ and (reverse) 5’-ggc agg ctc tgt ttg cttaaa gag cg-3’ or metallothionein (MTA2) (sense) 5’-act cgt ccc ggc tct ttc ta-3’ and(reverse) 5’-agg agc agt tgg gat cca t-3’primers. Micheal Epis performed the ChIPassays presented in this thesis (Western Australian Institute for Medical Research,Perth, Australia).

Immunohistochemistry

Immunohistochemistry was performed in the Department of Anatomical Pathology atRoyal Perth Hospital by Lisa Stuart (Western Australian Institute for MedicalResearch, Perth, Australia), and analysed by Dr Cecily Metcalf (Department ofAnatomical Pathology, Royal Perth Hospital, Perth, Australia) and Lisa Stuart. Full-face sections of primary breast cancer tissue were immunostained for SLIRP. Sectionswere deparaffinised through 3 changes of xylene (3 min), rehydrated through gradedalcohols to distilled water and subjected to antigen retrieval in 500 mM EDTA22 (pH8.0) under pressure. After blocking endogenous peroxidase activity with hydrogenperoxide the SLIRP polyclonal ab or HSP-60 ab were used at 1:2500 for 60 min.Immunoreactivity was detected by incubating slides with biotinylated goat anti-rabbitand anti-mouse ab followed by streptavidin-horseradish peroxidase as secondary andtertiary reagents. Sections were vizualised with DAKO followed by a light counterstainwith haematoxylin.

Imaging studies

HeLa cells cultured on glass cover slips were incubated with MitoTracker, SLIRP,HSP-60 or cytochrome c ab and Alexa Fluor 488 goat-anti-rabbit secondary ab added.Cover slips bathed in 100 ng/ml Hoechst 33258/PBS, washed and mounted inVectashield. For FLAG imaging, cells were transfected 24 hours prior to staining asabove and visualised using a BioRad MRC1000 confocal microscope. Microscopepictures were taken by Shane Colley (Western Australian Institute for MedicalResearch, Perth, Australia).

GST-SLIRP-DM (L24A/A62L)

GST-SLIRP

GST-SLIRP-A62L

GST

GST-SHARP-RRM

GST-SHARP-RD

GST-SLIRP-L24A

REMSA plasmids

1 109

24

24

62

62

GST

GST

GST

GST

GST

109

109

109

1

1

1

GST

GST

22 52 60 95RRM RRM

RRM RRMRRM1

1

RD

SLIRP-FLAG

SLIRP-FLAG-R62L

SLIRP-FLAG-DM (R24/L62)

SLIRP-FLAG-L24A

FLAG1 109

FLAG1 10924

FLAG1 109

FLAG1 10924 62

62

RRM RRM22 52 60 95

1pCMX-SHARP

pSCT-SRA

pCMX RRM RRMRRM

1

3651

pSCT-SRA-SDM8

HA-SKIP

pSCT

pSCT

HA536

1

1

870 nt

870 nt

Transfection Plasmids

FLAGFLAG

pCMX

pSCT pSCT

HAHA

A

B

Figure 2.1 A, B: Schematic illustrating plasmid constructs. Panel A includes stickfigures to represent all plasmids used in REMSA. Panel B includes stick figures torepresent all plasmids used in transfections.

68

Table 2.1: Primers. This table details all the primer names and sequences that were used in this thesis forsequencing, PCR, RT-PCR and IP-RT-PCR experiments.

SLIRP sense cgc gga tcc gcg gcc tca gca gcaSLIRP antisense gcg cgg atc cta ggc tgc agt ctcaSLIRP sense (RT) gcg ctg cgt aga agt atc aaSLIRP antisense (RT) tcg att ccg aag tcc ttc ttSTR7 sense agg agg cag gta tgt gat gac atc agc cga cgc ctg gca ctg ctg cag gaa cag

tgg gct gga gga aag ttg tca ata cct gta aag aaSTR7 antisense ttc ttt aca ggt att gac aac ttt cct cca gcc cac tgt tcc tgc agc agt gcc agg

cgt cgg ctg atg tca tca cat acc tgc ttc ctSDM7 sense agg agg cag gta tgt gat gac atc tcc cga cgc ctc gca ctg ctc cag gaa cag

tgg gct gga gga aag ttg tca ata cct gta aag aaSDM7 antisense ttc ttt aca ggt att gac aac ttt cct cga gcc cac tgt tcc tgc agc agt gcc agg

cgt cgg gag atg tca tca cat acc tgc ttc ctSKIP sense cat tca act ctg gag cta aac agaSKIP antisense tcc gat cag caa tgt aga ggg ctβ Actin sense gcc aac aca gtg ctg tct ggβ Actin antisense tac tcc tgc ttg ctg atc caGST sense gta ctt gaa atc cag caa gGST antisense cag atc gtc agt cag tcaFLAG sense ggg cgt gga tac ggt ttg act cac ggg gaFLAG antisense tat tag gac aag gct ggt cacSRA IEV sense tga tga cat cag ccg acg cctSRA IEVantisense gct gca gat ttc tct tca ttgpS2 sense ggc cat ctc tca cta tga atc act tctpS2antisense ggc agg ctc tgt ttg ctt aaa gag cgMetallothionein (MT2) sense act cgt ccc ggc tct ttc taMetallothionein (MT2) sense agg agc agt tgg gat cca

69

70

E. Hatchell Chapter 3: Basic Characterisation of SLIRP

71

Chapter 3

Basic characterisation of SLIRP


72

3.1 Introduction

3.1.1 Preface

SRA is an RNA coactivator, which, as discussed in detail in Chapter 1, plays animportant role in the ER pathway in breast cancer and breast cancer cell lines. Lanz and

O’Malley (1999) published their finding of the novel RNA coactivator in Cell, andhave since detailed the importance of several of SRA's stem loops. Before I embarked

on my PhD, I performed a yeast III hybrid screen with a human primary breast cancer

cell library using STR7, a functionally relevant stem loop required for SRA-mediatedcoactivation. From this screen a previously hypothetical and novel SRA binding protein

(Hatchell, 2002) was identified, which was termed SLIRP: SRA stem-loop interactingRNA-binding protein (Hatchell et al. , 2006 & Patent Number: WO/2007/009194). In

this Chapter, I describe the basic characterisation of SLIRP, including the bioinformatic

analysis, mRNA and protein expression and evaluation of the SRA-SLIRP interaction.Greater in depth discussion of the identification of SLIRP can be found in two other

documents (Hatchell, 2002 & Hatchell et al., 2006). However this Chapter will discussspecific aspects of its identification throughout this Chapter in order to establish

context for the remainder of this thesis.

3.1.2 SRA and the growing ncRNA field

The recent discovery of a large family of ncRNAs that play critical roles in the cell(Pang et al., 2005) has led to a significant shift in our understanding of the basic

biology involved in gene regulation. Approximately 98 % of all transcriptional outputis in the form of non-coding RNA (Mattick, 2001) and 30 % of the human genome is in

the form of introns, which are now predicted to act as a source of ncRNA (Mattick &

Makunin, 2006). It is of particular interest to note that there is an inverse relationshipbetween the complexity of an organism and the percentage of its genome that is

comprised of “junk” DNA (Szymanski et al., 2005). Junk DNA accounts for only 10-20 % of single unicellular organism genomes with this number increasing to 98 % in

complex eukaryotes (Szymanski et al., 2005). Such observations have led several

research laboratories to hypothesise that the non-coding regions of the genome mayplay crucial regulatory roles in the cell and may account for the differences between


73

organisms that have not been accountable by the differences in protein encoding genes

in the sequenced genomes (Mattick, 2001; Szymanski et al., 2005). A very recentpublication has shown that through the use of whole-genome tiling arrays that there

appears to be a subset of ncRNA that are evolutionarily conserved (Perez et al., 2007),

further suggesting ncRNA do play an important role throughout biology.

There are now a great number of functional ncRNAs described in the literature which

have roles in various biological pathways, including cancer. Some examples ofncRNAs whose functions are well studied include XIST, an essential molecule in the x-

inactivation process (Perez et al., 2007), and B200, a regulator of decentralizedtranslation in dendrites (Duning et al., 2007). Some examples of ncRNA involved in

cancer include DD3 (Bussemakers et al., 1999) and PCGEM1 (Srikantan et al., 2000)

which are both overexpressed in prostate cancer, MALAT-1 which is expressed in lungcancer (Muller-Tidow et al., 2004), H19 which was first found to be highly expressed

in many varied and different tumors (prostate, bladder, pancreatic and uterine to name afew) including breast tumors (Looijenga et al., 1997), and which recently has been

found to be strongly associated with increased breast cancer risk (Easton et al., 2007),

and SRA (Lanz et al., 1999; Szymanski et al., 2003; Szymanski et al., 2005).

3.1.3 RNA secondary structures

The secondary structure of RNA plays a critical role in determining protein-binding

interactions (Matthews et al., 1999). The secondary structure of RNA is usuallydetermined using free energy minimisation techniques (Jaeger et al., 1993) there are

publicly available programs that minimise the thermodynamic free energies of RNAsecondary structure formation. The most recent secondary structure prediction

programs essentially use a combination of techniques which have been used in this

field for a number of years, including polymer, dumbbell and oligonucleotide nearest-neighbour (NN) thermodynamics (SantaLucia, 1998). The high quality mFold program

which was employed to determine RNA secondary structures in this thesis was written

by Michael Zuker (2003).


74

3.1.4 RNA-binding proteins and RNA-binding domains

As outlined in Chapter 1, RBPs play a major role in many fundamental cellular

processes including early development, infection by RNA viruses and regulation of a

wide array of gene expression pathways (SenGupta et al., 1996; Mattick, 2001). RNA-binding proteins play a major role in the post-transcriptional regulation of gene

expression (Gamberi et al., 2006). The essential role of RNA–protein interactions for

normal cell function is highlighted in cases where RNA-binding protein dysfunctionleads to severe clinical consequences. For example, mutations in the FMR1 gene,

which encodes a cytoplasmic RNA binding protein, leads to the development of theFragile X syndrome (Verkerk et al., 1991). Another example includes the altered

interaction between iron-regulatory proteins (IRPs) and cognate response elements

(iron responsive elements) which is associated with abnormal iron homeostasis (Statonet al., 2000).

Several classes of RBPs have been described based on the presence of characteristic

conserved binding motifs (Burd & Dreyfuss, 1994). One of the most well-characterised

domains is the RNA recognition motif (RRM). The RRM is 90 to 100 amino acids (aa)and is typically present in one or more copies in the sequence (Burd & Dreyfuss, 1994;

Shamoo et al., 1995; Siomi & Dreyfuss, 1997). Two consensus sequences are usuallypresent (RNP1 and RNP2) as well as a number of conserved single aa (Siomi &

Dreyfuss, 1997).

Recently, the field has expanded with the characterisation of novel domains within

RNA modifying enzymes and proteins involved in translational regulation (Aravind &Koonin, 1998; Zhao et al., 2007). For example, a significant impact in the area of

RNA-protein interactions by the discovery of the PUA (PseudoUridine synthase and

Archaeosine transglycosylase) domain, the S4 domain (Aravind & Koonin, 1998) andthe NAD+ binding fold of NAD+ dependent dehydrogenases (Nagy et al., 2000).

Motifs such as the arginine rich motif (ARM) are becoming recognised. The ARMconsists of a short sequence (10–20 aa) that is arginine rich (Burd & Dreyfuss, 1994).

ARMs are generally positively charged, which increases their non-specific affinity forRNA, thus allowing proteins to search for high-affinity binding sites (Burd & Dreyfuss,

1994).


75

When RBPs bind RNA they may alter the RNA structure and either hinder or facilitateinteractions with other trans-factors (Siomi & Dreyfuss, 1997). Such interactions may

also provide localisation and target signals for the transport of RNA to different cellular

compartments (Siomi & Dreyfuss, 1997). Furthermore, a potential role of RNA-proteininteractions which Cech & Bass (1986) first suggested and Dreyfuss et al. (1988)

reiterated, is that RNA-protein units, or perhaps even the RNA itself, may act as a

catalyst in transcription. Due to the increased importance of RNA-protein complexesapparent almost 20 years ago, Dreyfuss et al. (1988) hypothesised that some RNAs

may in this way act as a scaffold upon which proteins can assemble to form acomplexes utilised in the transcription process.

Over 300 different known RNP motif proteins were published by 1997 (Siomi &Dreyfuss, 1997) and this number has continued to increase. Most of these are known to

have both specific and non-specific components of binding. In fact, most RNA-bindingproteins have some level of basal non-specific binding (Wang & Schimmel, 1999). In

terms of functional relevance, it has been historically hard to evaluate the difference

between specific and non-specific binding for some proteins (Wang & Schimmel,1999). While less is known of the role of non-specific interactions it is important not to

discard these interactions as worthless. Evidence suggests that some non-specific RNA-binding domains may have important functional roles, for example, increasing the

functional activity of a highly specific enzyme in vivo (Wang & Schimmel, 1999).

Together with the fact that hundreds, possibly thousands, of yet to be delineated smallRNAs (sRNA) are thought to exist in the cell (Stormo & Ji, 2001; Lee & Ambros,

2001; Lau et al., 2001; Ruvkun, 2001; Lagos-Quintana et al., 2001), such non-specificinteractions may play an important role in determining functionality of the cell.

3.1.5 RNA recognition motifs (RRMs)

In depth discussion of the importance of RBPs in normal cellular processes and the

existence of different RNA-binding domains has already been discussed in Chapter 1.However, greater discussion of the RNA recognition motif (RRM) is necessary at this

point because of its significance to the results presented in this Chapter.


76

The RRM is one of the most well conserved RNA-binding domains with over 300

known RRM proteins (Dreyfuss et al., 1993; Siomi & Dreyfuss, 1997). The RRM isknown to contain two RNP sequences (Burd & Dreyfuss, 1994; Siomi & Dreyfuss,

1997). These sequences, RNP1 and RNP2, are part of an RNA-binding surface of the

protein that form a β-α-β-β-α-β pattern (Siomi & Dreyfuss, 1997; Wang & Hall,

2001). The two sequences generally fall 30 aa apart in a 90 aa total RRM sequence(Dreyfuss et al., 1993). Crosslinking assays performed on RNP1 and RNP2 sequences

have historically (Merrill et al., 1988) identified each of them is involved in RNA-

binding. More recent data, however, has determined the actual structural biology of theRRM in proteins such as HuD which binds to androgen response elements (AREs) in

the androgen receptor mRNA (Wang & Hall, 2001). A comparison of the RRMdomains in HuD, sex lethal protein (Sxl), poly(A)+ binding protein (PABP), nuclear

ribonuclear protein (UP1), and the splicosomal proteins UA1 and U2B, has definitively

shown the site of RNA interaction for all of these proteins is through the RNP domains(Wang & Hall, 2001).

The RNP1 and RNP2 sequences occur juxtaposed on the two central β strands (Figure

3.1) on what is termed a β sheet (Siomi & Dreyfuss, 1997; Wang & Hall, 2001). In

fact, these β sheets are widely used by RNA-binding proteins as a surface for

interactions for RNA-binding (Siomi & Dreyfuss, 1997). Protein inducedconformational changes of RNA via the unstacking of nucleotide bases at this non-

sequence specific RNA-binding platform (the β sheet) appears to be a common

although not universal theme in RNA-binding proteins aiming to achieve specificrecognition of RNA and the stabilisation of the resulting complex (Siomi & Dreyfuss,

1997).

It is of interest to note that several RNA-binding proteins have been found to have more

than one RRM domain (for example PABP, U1A, HuR, HuD and SHARP). Evidence

suggests such proteins may be able to bind to more than one segment of RNAsimultaneously (Dreyfuss et al., 1988; Lutz & Alwine, 1994, Wang & Hall, 2001), thus

allowing the formation of a complex of simultaneous interactions between severalRNAs and a single protein.


77

3.1.6 The identification of SLIRP

SLIRP was identified as part of my Honours project (Hatchell, 2002). However, all

characterisation of SLIRP and investigation of its biological function in the cell has

been undertaken during my PhD. SLIRP was first identified from a yeast III screen of ahuman primary breast cancer cell library. The sequence identified was that of a

hypothetical protein previously termed DC50 (Genebank Accession Number [GAN]:

AF271779), then Chr14 Open Reading Frame 156 (Chr14ORF156) (GAN: BC017895).Both of these mRNAs were obtained previously via theoretical translation and entered

into GenBank.

The SLIRP sequence predicted a protein of 109 aa in length, and the entire coding

region (except for the start codon) and 3’ UTR (including the poly A tail) was isolatedvia the screen. When searching the protein sequence it was obvious that the sequence

consisted almost entirely of an RRM RNA-binding domain containing two RNP-consensus sequences RNP2 and RNP1, consistent with its capacity to bind SRA.

3.1.7 Hypotheses and Aims

After the discovery of a novel and uncharacterised gene, SLIRP, my initial goal was toperform basic characterisation of the SLIRP protein. Based on the observations in the

introduction of this Chapter, the following specific hypotheses were developed:

Hypothesis 1: SLIRP exists in cancer cell lines and primary breast cancers.

Hypothesis 2: SLIRP binds SRA specifically.Hypothesis 3: The RRM domains of SLIRP and SHARP are critical for binding to

SRA.

Hypothesis 4: The interaction of SLIRP with SRA is biologically relevant.

These hypotheses were pursed via the aims outlined below:

Aim 1: To assess SLIRP expression in cancer cell lines, normal and cancer tissues, and

primary breast cancer.Aim 2: To assess SLIRP interaction with SRA via REMSA analysis.


78

Aim 3: To compare SLIRP and SHARP binding to STR-7, and to assess SLIRP

binding to a mutant STR-7 (SDM-7) in REMSA.Aim 4: To perform IP-RT-PCR experiments in cancer cell lines looking for a close

association of SRA and SLIRP in vivo.

3.2 Methods

3.2.1 Bioinformatics and sequence analysis

All bioinformatics was performed via NCBI or Celera web databases. The mFoldprogram (Zucker, 2003) was accessed via either the NCBI web site, or Michael Zuker’s

mFold server web page: (http://bioinfo.math.rpi.edu/~mfold/rna/form1.cgi). All sequencealignments were performed using algorithms described by Stothard (2000).

3.2.2 Tissue culture

The MCF-7 breast cancer cell line represents an ER+ and EGFR- phenotype; MDA-MB-468, also a breast cancer cell line, are an ER- and EGFR+ phenotype; and HeLa, a

cervical cancer cell line, have an ER-, EGFR- phenotype (Brooks et al., 1973; Lippman

& Bolan, 1975; Marminta et al., 1991). SRA is expressed in MCF-7 cells (Lanz et al.,1999). MDA-MB-468 cells have an ER- phenotype so that using both cell lines has

enabled a direct comparison of breast cancer cell lines with opposite phenotypes. HeLa

cells were chosen as a non-breast cancer cell line which is the model cell line that Lanzet al. (1999) originally performed all SRA experimentation in. These cells show a very

different ER- and EGFR- phenotype.

3.2.3 Cytoplasmic and nuclear extracts

Generation of cytoplasmic and nuclear extracts was achieved as detailed in the

Materials and Methods Chapter. Bovine serum albumin (BSA) controls served toprovide readings at OD595 for a standard curve. These controls were always run

simultaneously with the data while using fresh BSA concentrations of 0.1, 0.2, 0.3, 0.4,

0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 1.25, 1.5, 1.75, 2 and 5 µg/µL.


79

3.2.4 Protein production

SLIRP protein for use in REMSA experiments was generated as described in the

Materials and Methods Chapter and previously by Giles et al. (2003). GST-SLIRP was

used for most experiments and GST alone was used as a control.

3.2.5 Western blot

Whole cell lysates were collected with 150 µL passive lysis buffer per well of a 6 well

plate. Further detail can be found in the Materials and Methods Chapter.

3.2.6 Radioisotope labelling of STR7

SRA STR7 pBlue was successfully linearized in an overnight digestion as described inthe Materials and Methods Chapter. All 32P labelled riboprobes were successfully

generated with a specific activity of ~2x108 cpm/µg RNA. RNA probes (~5 x 104 cpm

of 32P labelled probe) were utilized in REMSA. Specific or non-specific (pBluescript)competitor unlabelled RNA cold probes were produced (STR7) and together with

tRNA used at increasing concentrations for cold competition assays.

3.2.7 REMSA and UVXL

REMSA and UVXL were performed as described in the Materials and MethodsChapter. RNase T1 was added to the incubation mixture of riboprobe and selected

proteins to reduce the background. RNase T1 digests 3’ of guanosine residues andcleaves 5’ phosphate linkages of RNA molecules not protected by a protein (Thomson

et al., 1999). Heparin (a negatively charged polymer which mimics the RNA phosphate

backbone) was also added to reduce non-specific protein binding.

Supershift analysis was performed with specific REMSA to detect specific RNA-protein interaction with the ab of interest.


80

3.2.8 RT-PCR

RT-PCR was performed as described (Materials and Methods Chapter) to assess if

SLIRP RNA was expressed in a variety of cell lines. All primers were designed to

cover exon-exon boundaries to remove the chance of amplification from genomicDNA.

3.2.9 IP-RT-PCR

IP-RT-PCR was performed as described (Materials and Methods Chapter) to identifyclose association between a specific RNA and protein in cell extracts using primers

designed to cover exon boundaries. The IP-RT-PCR experiments presented in this

Chapter were performed by Dianne Beveridge (WAIMR).

3.2.10 Northerns

Northern blot assays were performed to assess SLIRP mRNA expression in cancer cell

lines and normal tissues. The same RNA membranes that Lanz et al. (1999) used todescribe SRA expression were purchased. These experiments were performed by Shane

Colley (WAIMR).

3.2.11 Immunoshisotchemistry

Immunohistochemical expression analysis of SLIRP was performed as described in

Chapter 2. Lisa Stuart (WAIMR) performed the analysis of SLIRP in several primaryhuman breast cancers in a collaboration with Cecily Metcalf (Department of

Anatomical Pathology, Royal Perth Hospital, Perth, Australia).


81

3.3 Results

3.3.1 SRA STR7 is a target for proteins in human breast cancer cells

SRA is a complex RNA molecule (Lanz et al., 2002) predicted to contain multiplestable stem-loop structures (Figure 1.5 A). STR7, an 89-nucleotide (nt) sequence, is the

largest and one of the most stable stem-loop SRA structures and functions in acooperative manner with other stem-loops to augment ER transactivation (Lanz et al.,

2002). Based on these observations, my previous investigation evaluated if SRA STR7

was a target for RNA-binding proteins REMSA and UVXL studies (Hatchell, 2002;Hatchell et al., 2006). These studies demonstrated a significant interaction between

SRA STR7 and nuclear proteins from human cancer cells (see Introduction Chapter andFigure 1.8 A-C). Confident that STR7 was a target for SRA binding proteins, a yeast III

hybrid screen (Sengupta et al., 1996) was performed to further identify these proteins

(Figure 1.8 D). The yeast III screen was undertaken using a primary human breastcancer library that originated from a woman diagnosed in 1991 with an ER+ metastatic

breast cancer at the age of 60 (Byrne et al., 1998).

3.3.2 SLIRP: a SRA-binding protein

From the yeast III hybrid screen a cDNA clone that contained an open reading frame

with a 3’ untranslated region (UTR) and polyadenylated (poly A) tail (Figure 3.2 A)

was isolated. The cDNA sequence predicted a protein of 109 aa with a Mr of 12.7 kDa.Database analysis revealed that this clone, which was named SLIRP: SRA stem loop

interacting RNA-binding protein (Hatchell et al., 2006 & Patent Number: WO/2007/009194), was identical to human sequences Hypothetical Protein DC50 and Chr 14

Open Reading Frame 156 (C14orf156) (Hatchell et al., 2006).

3.3.3 Bioinformatic Analysis of SLIRP

The predicted protein sequence of SLIRP (GAN: AY860853) is composed almostentirely of an RRM (Figure 3.2 A) containing RNP1 and RNP2 submotifs (Burd and

Dreyfuss, 1994). Bioinformatic analysis demonstrated that the SLIRP gene contains 4exons on chromosome 14. The RRM domain in SLIRP shares substantial aa homology

with SHARP (Figure 3.2 B), a SRA corepressor (Shi et al., 2001), and nucleolin, a


82

transcriptional repressor (Yang et al., 1994) and canonical RRM-containing protein

(Ghisolfi-Nieto et al., 1996; Bouvet et al., 1997). Significantly, homology betweenSLIRP and SHARP, particularly within key predicted interacting residues of their

RRMs (Figure 3.2 C), suggests these molecules may bind the same RNA targets, in

particular SRA STR7.

The aa sequence of SLIRP is highly conserved across human, rat and mouse species

(Figure 3.2 D). Curiously, human, rat and mouse SLIRP homologs have sharedsynteny. In humans, SLIRP is located on chromosome 14, mouse chromosome 12 and

rat chromosome 6. This striking sequence conservation suggests an important functionfor SLIRP in multiple species.

Of interest, human SLIRP is positioned within 1750 nt of SKIP, on Chr 14q24.3 withno intervening genes. SKIP is a vitamin D receptor (VDR) corepressor (as described in

detail in Chapter 1) implicated in oncogenesis and vitamin D metabolism (Barry et al.,2003; MacDonald et al., 2004). SKIP is expressed in breast cancer tissue and regulates

ER transactivation (Barry et al., 2003). Colocalization of these genes initially suggested

that SLIRP and SKIP may participate in the same NR pathways and could be co-ordinately expressed (Figure 3.3 A, B). There are some examples in the literature where

colocalisation and coexpression of proteins occurs. One such example is expression ofGrb-7 and HER2 which lie on Chromomsome 17, where the protein levels of each are

coregulated (Daly, 1998).

Loss of heterozygosity (LOH), tumor suppressor genes and LOH loci are critical

components of cancer biology (O’Connell et al., 1999). With the exception of therecently described BRACA1 and BRACA2 genes (which account for the vast majority of

hereditary breast cancers in Australia), and more recently FGFR2 (Easton et al., 2007),

no single genetic event is known to be necessary and sufficient to lead to breast cancer(O’Connell et al., 1999). Previous studies have shown chromosome 16 (q) and 18 to

play an important role in breast cancer metastasis (Adeyinka et al., 1999). Recently,

chromosome 14q, the chromosomal location of the SLIRP gene, has been implicated inthe metastasis of some breast cancers via LOH gene profiles (O’Connell et al., 1999;

Martin et al., 2001) and microsatellite marker analysis (Bjorkqvist et al., 1999; DeRienzo et al., 2000). It is also of interest to note that very recently the chromosomal


83

location of the SRA gene, chromosome 5, has been implicated in LOH in breast cancer

(Georgia Chenevix-Trench, personal communication, 28th March 2007).

O’Connell et al. (1999) found that the rate of LOH of chromosome 14q31.2 was much

higher in some cancers (axillary lymph node negative breast tumors) than in others(axillary lymph node positive breast tumors). Such information suggested that gene(s)

mapping to this region may be rate-limiting for breast cancer metastasis (O’Connell et

al., 1999). Martin et al. (2001) followed up this study by mapping chromosome 14qand found a number of candidate metastasis genes in the 14q31-32 region, including

MTA1. While SLIRP is not located in this region, this information does implicatechromosome 14q as both a region of importance in breast cancer and suggesting a

closer examination of SLIRP polymorphisms may be warranted.

3.3.4 SLIRP is expressed widely in human tissues and cancer cells

Expression of SLIRP was assessed using a range of techniques, including Northern and

Western blotting. RT-PCR analysis showed SLIRP is expressed at the RNA level in

HeLa and MCF-7 cancer cell lines (Figure 3.4 A). Northern analysis indicated that innormal human tissue SLIRP mRNA is ubiquitously expressed, but in varying amounts

with the highest levels in heart, liver, skeletal muscle and testis (Figure 3.4 B). SLIRPwas readily detected in a variety of cell lines, including SK-BR-3 (breast cancer),

MCF-7, HMEC (normal mammary epithelial cells), MDA-MB-468, LNCaP (prostate

cancer) and COS-7 (monkey kidney cancer) (Figure 3.4 C) and increased in HeLa,Calu-6 and HepG2 cells. Notably, SLIRP expression across multiple tissues (Figure 3.4

B) and cell lines (Figure 3.4 C) was similar to that of SRA (refer to Lanz et al., 1999,Figure 3.4 D).

A polyclonal SLIRP ab was generated and demonstrated SLIRP protein (Mr ~12.7 kDa)expression in multiple human cell lines, including those derived from breast, prostate

and lung carcinomas (Figure 3.5 A). The ab was highly specific for human SLIRP with

virtually no cross reactivity with a number of other species (Figure 3.5 B). Expressionof SLIRP protein varied across different breast cancer cell lines and in some cells

discordant levels of SLIRP mRNA and protein were observed (eg. HeLa cells). Whileexpression of SLIRP and SKIP was similar, for some cell lines there was little evidence

that their expression at the mRNA or protein level was co-ordinately regulated.


84

Immunohistochemistry (IHC) with SLIRP ab on human primary breast cancer tissueshowed SLIRP staining in normal ductal tissue, but little in the surrounding stroma

(Figure 3.5 C, a & b). Intense SLIRP staining was noted in carcinoma tissue (Figure 3.5

C, d) compared to control (Figure 3.5 C, c). Staining was evident throughout the cellbut predominantly with punctate, cytoplasmic distribution (Figure 3.5 C, b & d).

3.3.5 Characterization of SLIRP’s interaction with SRA

I next addressed the question of SLIRP’s capacity to bind SRA. Using recombinantSLIRP proteins (Figure 3.6 A), both GST-SLIRP (Figure 3.6 B, lane 2) and cleaved

SLIRP (Figure 3.6 C) bound STR7 avidly, while GST alone did not (Figure 3.7 B lane

3). A common finding during the production of SLIRP protein was that when SLIRPprotein was cleaved off GST it became insoluble and fell out of solution, thus making it

extraordinarily hard to work with. It was, therefore, important to keep cleaved SLIRPin non-concentrated state in solution. Such a low concentration explains the weaker

binding results seen in Figure 3.6 C.

The supershift experiments described in this Chapter were attempted with cleaved and

eluted protein, and produced the same effect. Addition of increasing amounts ofunlabeled (cold) STR7 probe efficiently competed out the complex (Figure 3.6 B, lanes

3 & 4). In contrast, neither addition of excess unlabeled pBlue (Figure 3.6 B, lanes 5 &

6) or high amounts of tRNA significantly affected SLIRP-STR7 complex formation(Figure 3.6 D). Taken together, these results indicate that SLIRP binds STR7 in vitro

with a high degree of specificity.

Results showed that binding of GST-SLIRP to the SRA SDM7 probe (a SRA STR7

mutant containing several stem-loop point mutations and having reducedtransactivation activity compared to wildtype; Lanz et al., 2002) was consistently

reduced compared to the STR7 probe (up to 2.9-fold) (Figure 3.6 E, lane 6). SLIRP

binding to the SDM7 probe could also be overcome with excess unlabeled SRA STR7(lanes 7 & 8).

Given the homology between SLIRP and SHARP within their RRM domains, the next

study was to see if SHARP could also bind SRA STR7 (Figure 3.7 A). A GST-SHARP


85

fusion protein (Shi et al., 2001) containing SHARP’s three RRM domains (GST-

SHARP-RRM) bound STR7 avidly (Figure 3.7 B, lane 1), whilst GST-SHARP-RD(containing only its repression domain) did not (Figure 3.7 B, lane 2). These data

indicate that the SHARP RRMs may compete with SLIRP for binding to SRA STR7.

In order to further assess if competition between SLIRP and SHARP exists in whole

cell extracts, supershifts were performed using SLIRP protein and a SLIRP polyclonal

ab in REMSA. While it appeared that with increasing amounts of GST-SHARP-RRM,SLIRP binding to the STR7 RNA probe was decreased, unfortunately, the SLIRP ab

had a degree of affinity for the probe, rendering the significance of this resultimpossible to interpret. This result, therefore, is not presented or discussed in detail

here.

To confirm SLIRP’s interaction with SRA in vivo, immunoprecipitation-RT-PCR (IP-

RT-PCR) assays with SLIRP ab was performed. Using HeLa (Figure 3.8 A), MDA-MB-468 and MCF-7 cells (Figure 3.8 B, C), SRA co-immunopurified with SLIRP

(Figure 3.8 A, lane 5) but not β-actin (Figure 3.8 A, lane 6). This data suggests that

SLIRP closely associates with SRA in whole cells in several cancer cell lines.

3.4 Discussion

In this Chapter I have provided a characterisation of SLIRP from a bioinformatic and

RNA-binding perspective. SLIRP is a small protein comprising of only 109 aa. The

calculated size of SLIRP is 12.7 kDa. The genomic organization of SLIRP indicatesthat it contains 4 exons on chromosome 14, is highly conserved between species,

suggesting a potential important role through evolution. The SLIRP protein iscomposed almost entirely of an RRM with minimum other domains, indicating that it

may act as a connecting or adaptor molecule between RNA and protein or may have a

functionally important role in binding RNA.

It was intriguing to observe that SKIP and SLIRP colocalize to human Chr 14q24.3 andthat this genomic colocalisation is conserved across species. This observation raised the

possibility that they may be co-ordinately regulated, as is the case for Grb7 and HER2

that lie adjacent to each other on human Chr 17 (Daly, 1998). However, the expression


86

data did not support this idea. Although SKIP and SLIRP are not regulated at the

protein level, their RNA expression levels are similar which suggests that they may stillbe regulated together and differentially post-translationally modified. Although LOH

has been described at Chr 14q31.2 in breast tumors (Martin et al., 2001; O'Connell et

al., 1999), LOH in the genomic area of SLIRP/SKIP in breast cancer has not beenpreviously described, providing an impetus to examine this in primary breast cancers.

As SLIRP was a hypothetical protein when I identified it, it was important to verify itsexpression in vivo. This was done via RT-PCR, Northern and Western analysis. It was

found that SLIRP was expressed at the RNA level in both normal and human cancercells lines, and highly expressed in skeletal muscle, heart, testes and liver. When SRA

and SLIRP expression was compared on the same blots, their relative expression levels

were very similar, leading to the hypothesis that the two molecules may be similarlyregulated in cancer cells.

The next experiment reconfirmed SLIRP’s ability to bind SRA STR7 via REMSA.

SLIRP binding to STR7 was avid. However, binding was significantly reduced when a

mutant probe, SDM7, was used in REMSA indicating sequence specificity of theSLIRP–SRA interaction. IP-RT-PCR confirmed SLIRP’s close association with SRA

in whole cells in the model cell line HeLa and in two breast cancer cell lines, MCF-7and MDA-MB-468. This data indicates that these studies in HeLa are a reasonable

approximation to the molecular arrangement of these proteins in breast cancer cells and

this data has provided a foundation for further studies in breast cancer cells which aredescribed in Chapter 5.

SHARP was discovered as binding to SRA (Shi et al., 2001). However, the general

nature of the pull down assay did not allow the identification of the exact location of

SRA to which SHARP bound. Based on a very high homology between the RRMdomain of SLIRP and each of the three RRM domains in SHARP’s N terminus, it was

then hypothesised that SHARP would also bind SRA STR7. REMSA clearly showed

that this was the case: the RRM domains of SHARP bound STR7 while the controls didnot.

As a result of these experiments, it was found that both SLIRP and SHARP can bind

SRA STR7 so therefore it was then hypothesised that they may also compete for


87

binding of STR7. Supershift assays were explored in an attempt to identify if SLIRP

and SHARP could compete for binding of STR7. Unfortunately due to the antibodyaffinity for the STR7 probe, the work remains preliminary. Data provided in this thesis

suggests there is competition due to a shift from SLIRP to SHARP. However, in order

to perfect the supershift analysis of this competition, monoclonal antibodies will needto be developed to remove all non-specific binding. Further assessment of SLIRP and

SHARP binding to each of SRA’s stem loops will be of extreme interest and necessary

to conclude the degree of specificity of these RNA-protein interactions.

Conclusions

In this Chapter I have provided results which detail the basic biology of a novel protein,

SLIRP, that was identified from a yeast III hybrid screen of a human primary breastcancer cell library. SLIRP is a small RNA-binding protein with an RRM spanning

almost the entirety of the molecule. SLIRP is highly expressed in normal tissues andcancer cell lines at the RNA level and, remarkably, has a similar expression pattern to

SRA RNA. SLIRP is ubiquitously expressed at the protein level in cancer cell lines and

importantly is highly expressed in primary human breast cancers. SLIRP associateswith SRA in whole cells and specifically binds STR7 in vitro. This Chapter has also

shown here that another SRA-binding protein, SHARP, can specifically target STR7via its RRM. Such observations suggest SLIRP may interact with other NR

coregulators involved with the ER signaling pathway, which will be the subject of

further investigation in the next Chapter.

Figure 3.1 A: The RRM RNP domain. Schematic taken from Varani & Nagai (1998, p.422) showing the three dimensional structure of the most common RNA-binding domainknown, the RRM domain. The RRM domain consists of a repeated β-α-β topology.The two RNP sequences (1 & 2) are juxtaposed on the two central β strands, the β sheet.Two other common RNA-binding domains, the double stranded RNA-binding domainand the KH domain, also have a similar α β architecture.

A

88

A

B

1 g g c cattatggccggggaaggtgctttagtctgaagATGgcggcctcagcagcgcgaggt M A A S A A R G 61 gctgcggcgctgcgtagaagtatcaatcagccggttgcttttgtgagaagaattccttgg A A A L R R S I N Q P V A F V R R I P W 121 actgcggcgtcgagtcagctgaaagaacactttgcacagttcggccatgtcagaaggtgc T A A S S Q L K E H F A Q F G H V R R C 181 attttaccttttgacaaggagactggctttcacagaggtttgggttgggttcagttttct I L P F D K E T G F H R G L G W V Q F S - 241 tcagaagaaggacttcggaatgcactacaacaggaaaatcatattatagatggagtaaag S E E G L R N A L Q Q E N H I I D G V K - - - * * * * * * 301 gtccaggttcacactagaaggccaaaacttccgcaaacatctgatgatgaaaagaaagat V Q V H T R R P K L P Q T S D D E K K D ̂ ^ ^ - - - - - 361 tttTGAgactgcagcctattaataaagttaacataactg F

RNP2. RNP1 .SHARP_RRM_1 336 --GIKVQNLPVRSTDTSLKDGLFHEFKKFGKVTSVQIHGTSE----ERYGLVFFRQQEDQEKSHARP_RRM_2 437 TRTLFIGNLEKTTTYHDLRN----IFQRFGEIVDIDIKKV------NGVPQYAFLQYCDIASVCKSHARP_RRM_3 517 -NCVWLDGLSSNVSDQYLTR----HFCRYGPVVKVVFDRL------KGMALVLYNEIEYAQAnucleolin 573 TLFVKGLSE--DTTEETLKE----SFD--GSVR-ARIVTDRETGSSKGFGFVDFNSEEDAKASLIRP 21 ---AFVRRIPWTAASSQLKE----HFAQFGHVRRCILPFDKETGFHRGLGWVQFSSEEGLRN

Consensus LFVGNL E L F FG I I KGFGFVXF IYIKGM D Y V V R YA

SHARP_RRM_3 AVKETKGRKIGGNKIKVDFANRSHARP_RRM_2 AIKKMDGEYLGNNRLKLGFGKSnucleolin AKEAMEDGEIDGNKVTLD----SHARP_RRM_1 ALTASKGKLFFGMQIEV-----SLIRP ALQQ-ENHIIDGVKVQV-----

Consensus A L G I V

Figure 3.2 A, B: SLIRP is an RRM-containing SRA-binding protein. Panel A:Nucleotide and amino acid (aa) sequence of SLIRP. The entire mRNA is shown: arrowdenotes sequence isolated via yeast three-hybrid screen; start and stop codons in capitals;italics denote poly A signal; SLIRP contains a highly conserved RRM (underlined) withconsensus RNP2 & RNP1 sub-motifs (highlighted). *, ∧ and – denote putative N-myristoylation, protein kinase C phosphorylation and casein kinase II phosphorylationsites, respectively. Panel B: Sequence alignment comparing SLIRP, SHARP andnucleolin RRMs. Black boxes indicate aas conserved with consensus RRM sequencedescribed by Burd & Dreyfuss (1994).

89

C

DHomo sapiens MAASAARGAAALRRSINQPVAFVRRIPWTAASSQLKEHFA 40 Mus musculus MAASAIKGLSALRSSTGRPIAFVRKIPWTAAASELREHFA 40 Rattus norvegicus MAASAVKGLTALRSRTGRPIAFIRKIPWTAASSELREHFA 40 Homo sapiens QFGHVRRCILPFDKETGFHRGLGWVQFSSEEGLRNALQQE 80 Mus musculus QFGHVRRCTVHFDKETGFHRGMGWVQFSSQEELQNALQQE 80 Rattus norvegicus QFGHVRRCTVPFDKETGFHRGMGWVQFSSQEELQNALQEE 80 Homo sapiens NHIIDGVKVQVHTRRPKL---PQTSDDEKKDF 109 Mus musculus HHIIDGVKIHVQAQRAKALHGAQTSDEERFLR 112 Rattus norvegicus HHIIDGVEIHVQPQKPKVLHGAQTSDEEKDF 111

1 10991 SLIRP

Putative mitochondriallocalization sequence

RRM

RNP2

21

RNP1

60 6726

1SHARP

RRMs

RID SID/RD

3417336 2201 3651585 2707

RRM

Figure 3.2 C, D: SLIRP is an RRM-containing SRA-binding protein. Panel C showsSLIRP and SHARP functional domains. RRM, RNA recognition motif. RID, receptorinteraction domain. SID/RD, repression domain. Numbers denote aa sequence position.Panel D is an alignment of human, mouse and rat SLIRP aa sequences illustrates highdegree of homology between species. Black, aa identity; grey, similarity; white, nohomology.

90

Chr 14

14q24.3

SLIRP(76,170K)

SKIP (76,190K)

ERRβ

SLIRP

SHARP

RTA

CoAA

SKIP

1 22 95 109

RRM RNP 2 RRM RNP 1

RRMs RID SID/RD

NR-Inter SH2 Trans

RRM1 RRM2 TRBP ID

1 3417336 2201 3651

1 390

1 669

1 536

RGG RGG RGGRRM RNP 2 RRM RNP 1

123 167 250 286 381

68 81 144 307 584

A

B

Figure 3.3 A, B: SLIRP and other NR coregulators. Panel A shows the chromosomallocalization of SLIRP and SKIP on human chromosome 14. They are only 1750 bp apartfrom each other leading to the hypothesis that they may be coordinately regulated. PanelB: Schematic diagram illustrating SLIRP and three other RRM containing coregulators(SHARP, RTA and CoAA) each of which has been shown to bind RNA. SKIP is alsopictured to illustrate a coregulator whose genomic localisation is adjacent to SLIRP.

91

CSLIRP

GAPDH

SK

-BR

-3M

CF-

7

MD

A-M

B-4

68

HM

EC

HeL

aLN

CaP

Cal

u-6

Hep

G2

CO

S-7

SKIP

Pan

crea

s

SLIRP

β-actin

B

Hea

rtB

rain

Pla

cent

aLu

ng

Live

rS

kele

tal m

uscl

eK

idne

y

Sp

leen

Thym

usP

rost

ate

Tes

tis

Figure 3.4 A, B, C: SLIRP is widely distributed in normal human tissues and cancercell lines. Panel A: RT-PCR for SLIRP from RNA from MCF-7 and HeLa cancer celllines. Lane 6 is the water control for the oligo DT although there is some spillage fromthe previous lane. This was not seen in other gels. Lane 8 is the positive SLIRP control,lane 9 is the no RT control using HeLa RNA. Panel B: Northern analysis of SLIRPcompared with β-actin in normal human tissues. Note, these two Northern blots are thesame blots that Lanz et al. (1999) used when SRA RNA expression was analysed.Panel C. Northern analysis of SLIRP in cancer cell lines. mRNA from human breast(SK-BR-3, MCF-7, MDA-MB-468); prostate (LNCaP); lung (Calu-6); cervical(HeLa); liver (HepG2) cell lines, normal mammary (HMEC) and monkey kidney(COS-7) cells probed with SLIRP, SKIP and GAPDH probes.

A1 Kb LadderMCF-7 RNA

HeLa RNAH2O

Oligo DTRandom Primers

SLIRP control

+------

---++--

---+-+-

------+

----+--

+------

-+--+--

-+---+-

--+-+--

--+--+-

1 2 3 4 5 6 7 8 9 10

92

Hea

rtB

rain

Pla

cent

aLu

ngLi

ver

Ske

leta

l mus

cle

Kid

ney

Pan

crea

sSLIRP

β-actin

SRA

β-actin

Lanz et al(1999) Cell

Hatchell et al(2006) Mol Cell

NorthernsD

Figure 3.4 D: SLIRP and SRA RNA are expressed in similar normal tissues. Twopanels of Northern blots showing a comparison of SLIRP mRNA expression (Hatchellet al., 2006), to SRA RNA expression (Lanz et al., 1999, p. 18). These experimentswere both done on blots bought from the same company, and show a remarkablesimilarity in the expression patter of SLIRP and SRA at the RNA level.

93

SLIRP

β-actin

LNC

aP

HeL

a

SK

-BR

-3

MD

A-M

B-4

68

MC

F-7

Cal

u-6

HT1

080

PC

3

A

SKIP

T47D

B

Figure 3.5 A, B: Analysis of SLIRP protein expression. Panel A: Immunoblot ofprotein lysates from breast (SK-BR-3, MCF-7, MDA-MB-468, T47D), cervical(HeLa), prostate (LNCaP, PC3), lung (Calu-6) and fibrosarcoma (HT1080) cells usingSLIRP, SKIP or β-actin abs. Panel B: Immunoblot of lysates from several species,including human (HeLa, MCF-7), murine (NIH-3T3, J2E, OD9DL, C2C12), monkey(COS-7) and rat brain using the SLIRP polyclonal ab or a β-actin ab. All non-humanderived samples are negative for SLIRP except for marginal reactivity in the J2E celllysate, suggesting that the ab is specific for human SLIRP.

HeL

a

NIH

-3T3

OD

9DL

J2E

CO

S-7

S26 MC

F-7

C2C

12R

at b

rain

SLIRP

β-Actin

94

Da b

c d

Stroma

Ducts

Tumor

40x20x

Immunohistochemistry of aPrimary Human Breast Cancer

No Ab

Ducts

Carcinomatissue

Figure 3.5 C: SLIRP in expressed in primary human breast cancer tissue. Sections(20 x a,c; 40 x b,d) from a human breast ductal cancer were probed with SLIRP ab (a,b and d) and compared with sections from the same tumor with no ab (c). Arrowsdenote stroma, ducts and tumor tissue. Box in Panel a denotes region magnified inPanel b (40 x).

95

GST-SLIRP

1091GST

21RRM

91

STR7 ProbeSDM7 Probe

GST-SLIRPCold STR7

+---

-++-

-++

-++

+-+-

+-+

+-+

-+--

1 4 5 6 72 3 8

RPC

FreeProbe

CB

RPC

FreeProbe

1 4 5 62 3

+---

++--

++

-

++-

++

-

++-

STR7 ProbeGST-SLIRPCold STR7Cold pBlue

GST GSTA

Figure 3.6 A, B, C, D, E: SLIRP REMSA studies. Panel A: Schematic of plasmidsused in REMSA studies: GST-alone, GST-SLIRP (wild-type), GST-SHARP-RRM(SHARP aas 1-608) and GST-SHARP-RD (SHARP aas 3420-3651). Panel B:Binding of recombinant GST-SLIRP to SRA STR7. Specific binding of STR7 byGST-SLIRP (lane 2) is reduced with unlabeled “cold” STR7 (lanes 3-4, up to 100-fold excess) but not excess cold pBlue (lanes 5-6). Panel C: REMSA showing thatcleaved SLIRP binds SRA STR7. Panel D: REMSA showing that increasing amountsof tRNA has little effect on wild-type SLIRP binding to SRT STR7. Panel E: REMSAdemonstrating GST-SLIRP binds the wild-type SRA STR7 probe more avidly thanSDM7 mutant probe. Binding to both probes was reduced following addition of up to100 fold excess “cold” STR7 competitor (lanes 3-4 and 7-8).

E

STR7 ProbeGST-SLIRP

tRNA

++-

++-

D

STR7 ProbeSLIRP

++

96

A

GST-SHARP-RRM

GST GST

GST-SHARP-RD

1 608GST RRM RRM RRM

3420 3651SID/RDGST

B

STR7 ProbeGST-SHARP-RRM

GST-SHARP-RDGST

RPCFreeProbe

++--

+-+-

+--+

1 2 3

Figure 3.7 A, B: SHARP binds SRA STR7. Panel A: Schematic showing SHARPconstructs used in REMSA. Panel B: REMSA showing that STR7 is bound by GST-SHARP-RRM (lane 1; this protein contains all 3 SHARP RRM domains) but not witheither GST-SHARP-RD (lane 2; this protein contains no RRM domains) or GST alone(lane 3).

97

A

No abSLIRP abβ-actin ab

RT

+--+

-+-+

--++

+--+

-+-+

--++

+---

-+--

--+-

+---

-+--

--+-

----

----

S/N P WBeads S/N Beads131 4 5 6 72 3 8 11 12109 14

SRA

Figure 3.8 A: SLIRP associates with SRA in vivo. Panel A shows SLIRP associateswith SRA in vivo. SRA was detected by RT-PCR in supernatant samples or followingimmunoprecipitation with beads plus SLIRP but not β-actin ab. No product generatedfrom RT- samples. (P) SRA expression plasmid; (W) no template. Arrow denotes 260bp SRA-specific PCR product. This IP-RT-PCR was performed in HeLa cells.

98

S/N Beads -RT

+- - - - + - - - + - + - -

-- - + - - - + - - - - - -

-- + - - - + - - - - - - -

-+ - - + - - - + - + - - -

No ab

SLIRP ab

PACT abβ- actin ab

RT

+ - - + - - - + - - - + - - - - --

- - + - - - + - - - + - - - + - - -

- + - - - + - - - + - - - + - - - -

+ + + + + + + + - - - - - - - - - -

- - - + - - - + - - - + - - - + - -

S/N Beads S/N Beads

Figure 3.8 B, C: SLIRP associates with SRA in vivo. Panel B, C shows SLIRPassociates with SRA in vivo. Panel B is in MDA-MB-468 cells. Panel C is in MCF-7cells. The positive control (PACT) ab in C is for another SRA-binding protein in thelaboratory where these experiments were conducted (Redfern et al., in review). SRAwas detected by RT-PCR in supernatant samples or following immunoprecipitationwith beads plus SLIRP but not β-actin ab. No product generated from RT- samples.(P) SRA expression plasmid; (W) no template. Arrow denotes 260 bp SRA-specificPCR product.

B

C

99

No abSLIRP ab

β-actin ab

RT

SRA

SRA

WP

W P

100

E. Hatchell Chapter 4: SLIRP: a predominantly mitochondrial nuclear receptor corepressor

101

Chapter 4

SLIRP: a predominantly mitochondrial

nuclear receptor corepressor


102

4.1 Introduction

4.1.1 Preface

The basic biology and predicted structure of SLIRP and the specificity of its interactionwith SRA was detailed in the previous Chapter. In this Chapter I detail the discovery of

three fascinating aspects of SLIRP biology: first, that SLIRP is a NR corepressor;

second, that SLIRP is predominantly mitochondrial; and third, that SLIRP interactswith other NR coregulators.

4.1.2 Nuclear receptors

The regulation of eukaryotic gene expression in response to environmental stimuli is a

complex, multi-step and multi-regulated phenomenon (McKenna et al., 1999). Outlined

in detail in the Introduction is how the NR superfamily comprises a large family of

transcription factors which play a major role in the control of critical pathways ranging

from cell development to homeostasis (McKenna et al., 1999; Chawla et al., 2001;

McKenna & O’Malley, 2002b). NRs are transcriptional regulators and can regulate

gene expression when bound by their appropriate ligands (Jenster, 1998). Essentially,

NRs exert their action as transcription factors that directly bind to the promoters of

target genes and regulate their rate of transcription. In order to modulate transcription,

NRs must recruit a number of accessory coregulators known as corepressors and

coactivators (Savkur et al., 2004).

The ERα belongs to the steroid receptor family (Murphy et al., 2000; Leygue et al.,

1999; Mosselman et al., 1996). These receptors are nuclear proteins (Gorski et al.,

1984) that in the presence of E2 bind to DNA and activate genes involved in cell

growth. The ER is made up of six functional domains as explained previously (Fuqua

et al., 1993; Shibata et al., 1997) including a steroid domain which undergoes a

conformational change to lock E2 into a hydrophobic pocket (Jordan & Morrow, 1999).

Changes in the conformation of ER allows for coactivator proteins that facilitate to

bind selectively to the ligand independent AF-1, and the hormone-dependent AF–2

domains allowing the ER to bind DNA via the DNA-binding domain (Tora et al., 1989;


103

Sadovsky et al., 1995; Jordan & Morrow, 1999; Cavarretta et al., 2002, Warnmark et

al., 2003). Mediator proteins are able to direct the assembly and stabilisation of a

preinitiation complex and are termed coactivators. Essentially, coactivators along with

corepressors (together termed coregulators) are able to control gene transcription

primarily via complex interactions with the histone acetylase and histone deacetlyase

complexes involved in chromatin remodelling (Shibata et al., 1997; Leygue et al.,

1999; McKenna et al., 1999; Lanz et al., 1999).

4.1.3 Coregulators in the ER pathway

Although the estrogen receptor is required for a cell to respond to an estrogenic stimulus, the

nature and extent of that response are determined by the proteins, pathways, and processes with

which the receptor interacts.

-McDonnell & Norris, 2002, p. 1642

E2 can regulate a number of different pathways in the cell and recent progress in the

field has determined the many different ways in which this occurs. Essentially, the

transactivation of the ER pathway requires it to engage with transcriptionalcoregulators (O’Malley, 2006), although these coregulators are not expressed equally

(McDonnell & Norris, 2002). The importance of coregulator interactions has recentlycome under close scrutiny. It has been conclusively shown by several groups that ERE-

bound ER recruits coactivators and corepressors (Klinge et al., 2004) and that the order

of this recruitment can have substantive impact on transactivation both in vitro and invivo through structural modification of the transactivation complex (Klinge et al., 2004;

Fleming et al., 2004).

As mentioned in detail in the Introduction, a number of ER coregulators have been

discovered. These include, SCR-1, SRC-2, SRC-3, CBP, NCoR, TRAP, E6-AP, RTA,CoAA, CoAM, SHARP, SRA and SLIRP. The mechanism of recruitment of

recruitment is complicated and depends on the state of the cell and the number ofcoregulators free for recruitment. The intrinsic properties of the coregulator, and its

interaction with other coregulators, combine to stabilize the coregulator complex in a

either a state of activation or repression.


104

A coactivator must significantly increase transactivation without altering basaltranscription, reverse squelching when overexpressed and must contain transferable

autonomous activation domains (Wong et al., 2001; McKenna & O’Malley, 2002b;

Lanz et al., 1999). One of the most important ER coactivators is SRC-1. Upon binding

by E2, ER becomes associated with several proteins including SRC-1 in a complex.

This complex binds ER in a ligand-dependent manner and acts as a coactivator (Liu et

al., 1999). SRC-1 expression is currently being investigated for use as a biomarker of

breast cancer endocrine treatment resistance and may be an important predictiveindicator and therapeutic target in breast cancer (Fleming et al., 2004).

Corepressors in general must have direct interaction with the unligated receptor, and

interact with other basal transcription components (Robyr et al., 2000). One of the most

well characterised and abundant corepressors is NCoR. NCoR has been shown to play a

major role in the corepression of several NR pathways (Cohen, 2006), including the ER

pathway (Klinge et al., 2004). NCoR was first found to irreversibly silence ER function

in 1999 (Chien et al., 1999). Since then NCoR has been identified and extensively

studied in ovarian cancer cell lines and in ovarian cancer (Hussein-Fikret & Fuller,

2005). It has also been shown that E2 can downregulate NCoR levels in breast cancer

cells, thus causing a relief of NCoR repression and potentially having a broad impact

on the activation of transcription in tumors (Frasor et al., 2005). NCoR contains three

repression domains in its N-terminus and two receptor-interacting domains, crucial for

NR interaction (Seol et al., 1996). NCoR can interact with NRs in the absence of

ligand, causing ‘transrepression’, a state of repressive interaction of unliganded

receptors with components of the basal transcriptional machinery (McKenna et al.,

1999).

Importantly, in the absence of ligand, NRs are bound by NCoR, and another NR

corepressor: SMRT (silencing mediator of retinoid and thyroid hormone receptors)

(Chen & Evans, 1995; Cohen, 2006). SMRT was initially isolated by a yeast II-hybrid

screen of a human lymphoctye cDNA library with RXR as bait (Chen & Evans, 1998).

Significant similarity exists between both termini of NCoR and SMRT (SMRT has 2

fewer repression domains), however, they recruit significantly different complexes,

differentially mediate several orphan receptors and are differentially degraded;

suggesting several mechanisms through which these repressors can work together to

selectively abolish certain NR activity (McKenna et al., 1999).


105

4.1.4 Nuclear receptors in the mitochondria

Accumulating data suggests that there is significant cross-talk between NR in the

nucleus and the mitochondria. Recently the mitochondrion has been shown to act as a

primary site of action of steroid and thyroid hormones (Psarra et al., 2006) and severalinteractions exist between ER, AR, GR, PR and TR with the mitochondria (Gavrilova-

Jordan & Price, 2007). Such NR-mitochondrial interactions include transcriptionalregulation of nuclear DNA-encoded mitochondrial proteins, transcriptional regulation

of mitochondrial DNA-encoded proteins and indirect effects on mitochondria through

cytoplasmic signaling peptides (Gavrilova-Jordan & Price, 2007). It is currentlyhypothesized that NR-dependent interactions may play key roles in the mitochondrial-

dependent processes of both apoptosis and oxidative phosphorylation.

The GR was the first NR found to localise to the mitochondrion via a number of

different techniques, including immunofluorescence labeling and confocal microscopy(Scheller et al., 2000; Psarra et al., 2005), Western blotting of protein extracts (Scheller

et al., 2000; Psarra et al., 2005) and electron microscopy (Scheller et al., 2000).Putative GREs have also been found in the mitochondrial genome which can be

activated by nuclear GR (Psarra et al., 2006).

Recently, TRα has also been found to localise to cardiac mitochondria, bind putative

mitochondrial TREs (Morrish et al., 2006) and to promote ligand-dependent

transcription (Casas et al., 1999; Casas et al., 2003). Further, T3 has been shown to

coordinate cardiac mitochondrial and nuclear transcription (Goldenthal et al., 2004;Goldenthal et al., 2005). Interestingly, in whole mitochondria derived from

hypothyroid rats, RNA production is significantly reduced, and this is reversed with theaddition of T3 (Casas et al., 1999).

While the majority of work performed in this thesis focuses on ERα, importantly, both

the ERα and ERβ have been shown to localize to the mitochondria (including in breast

cancer cells) via immunofluorescence labelling and confocal microscopy (Solakidi et

al., 2005; Pedram et al., 2006), Western blotting of protein extracts (Solakidi et al.,

2005) and electron microscopy (Yang et al., 2004). Several putative EREs have been

identified in the mitochondrial genome (Ioannou et al., 1988; Sekeris, 1990) and ERβ


106

can bind mitochondrial ERE sequences from MCF-7 cells (Chen et al., 2004). Ofinterest to the work presented in this Chapter is evidence that E2 has vasoprotective

activity which may relate to stimulation of mitochondrial energy production andinhibition of reactive oxygen species production, suggesting that the ER may

coordinate gene expression in both nuclear and mitochondrial genomes (Burris &

Krishnan, 2005). Of further significance to this Chapter is the observations thatmitochondrial impairment in breast cancer cells results in altered expression of nuclear

genes involved in signaling, metabolism, cell growth, differentiation, apoptosis andcellular architecture which further implicates significant links between the nucleus and

the mitochondria (Delsite et al., 2002).

The PPARs are not members of the classical endocrine receptor subgroup, but instead

belong to the adopted orphan NR subgroup. There are three members of the PPARfamily: PPARα, PPARγ and PPARδ (Johnson et al., 2000), and they are highly

expressed in the liver, kidney, heart, skeletal muscle and brown adipose tissue. PPARs

play a fundamental role in mitochondria biology (Tobin & Freedman, 2006). PPARs

have been found in the mitochondria of several cancer cell lines and in vivo models(Psarra et al., 2006). Several nuclear transcription factors NF-κB, AP-1, CREB and p53

have also been found in the mitochondria (Tao et al., 2001; Cogswell et al., 2003;

Christian et al., 2006; Psarra et al., 2006).

Taken together, the current literature suggests a significant role for NR in the

mitochondria. The mode of action of NRs on mitochondrial transcription processes,and further detailing of other NR action in the mitochondria, is now under investigation

in many laboratories around the world. The mitochondria is already seen as a potential

cancer drug target (Don & Hogg, 2004; Nilsen & Brinton, 2004; Tobin & Freedman,2006) via the PPAR pathways. Thus, NR mitochondria biology is of considerable

significant importance to the cancer field.

4.1.5 NR coregulators and the mitochondria

RIP140 is a large NR corepressor which was first found to be recruited to the ER AF-2

domain (Augereau et al., 2006) and plays an important role in female fertility (White et

al., 2000), fat accumulation and energy homeostasis (Leonardsson et al., 2004). It is

expressed in tissues similar to SLIRP and is recruited by the ER, AR, GR, VDR, TR,


107

RAR, RXR and PPARα (Augereau et al., 2006). It is also a negative regulator of

insulin-responsive hexose uptake and oxidative metabolism (Powelka et al., 2006). Ithas been found that significant decrease of RIP140 in adipocyte cells significantly

upregulates genes involved in glycolysis, mitochondrial biogenesis and oxidativephosphorylation (Powelka et al., 2006). Essentially, decreasing RIP140 expression

causes an increase in mitochondrial oxygen consumption that has a profound impact on

energy homeostasis (Powelka et al., 2006). Taken together, the data in the literatureregarding RIP140 suggests it is a key player in the cross-talk and interaction between

NR coregulators and the mitochondria.

Another coregulator with important links to the mitochondria is peroxisome

proliferator-activated receptors gamma coactivator 1, PGC1α . PGC1α is a key

regulator of mammalian metabolism including skeletal muscle fibre type switching,adaptive thermogenesis in brown adipose tissue, glucose uptake, insulin secretion and

mitochondrial biogenesis (Tiraby & Langin, 2005). It is activated by signals that

control energy homeostasis, then induces and coordinates gene expression whichstimulates mitochondrial biogenisis essentially regulating cell fate decisions

(Puigserver, 2005; Puigserver & Spiegleman, 2003). PGC1α activates gene expression

through interaction with transcription factors which bind to the promotors of, and thuscan regulate, genes involved in metabolic pathways (Puigserver, 2005).

Both RIP140 and PGC1α have emerged as key players in the cross-talk between the

mitochondria and the nucleus. They both have crucial roles at the nuclear level, toregulate mitochondrial gene expression and energy homeostasis, and more complicated

levels of regulation may exist between the two organelles. The discovery of proteinsthat are involved in both nuclear transcription and mitochondrial biogenesis would be

of great interest in this emerging area.

4.1.6 ncRNA-protein interactions

The role of ncRNA in biology has recently received much attention. The importance oflong ncRNAs has been exemplified by the discovery of SRA in oncogenesis, Evf-2 in

organogenesis (Feng et al., 2006; Shamovsky & Nudler, 2006) and H19 in breast

cancer (Easton et al., 2007). ncRNAs have been shown to regulate gene expression in

many ways, including through RNA interference, gene suppression, gene silencing,


108

DNA demethylation, RNA splicing, chromatin remodeling, imprinting and

transcriptional activation. There is now mounting evidence to suggest ncRNA

involvement in cancer (Costa, 2005; Mattick & Makunin, 2006), and more specifically

in breast cancer (Lanz et al., 1999; Easton et al., 2007). Thus any new data detailing

the regulation of gene expression by important ncRNA-protein interactions will

contribute significantly to our understanding of the physiological and

pathophysiological pathways in breast cancer. The discovery of SLIRP, an SRA-

binding protein, thus provides us with the opportunity to elucidate the mechanisms of

its interactions, its functional role and provide novel insight into coregulator biology.


After a thorough basic bioinformatic and RNA-binding assessment of SLIRP in

Chapter 3, my next aim was to assess if SLIRP could modify ER transactivation,

interact with other NR coregulators and assess if it has an important functional role in

other NR pathways. Thus, the hypotheses and aims which guided my research in this

Chapter were as follows:

Hypothesis 1: SLIRP is a NR coregulator that modifies ER transactivation.

Hypothesis 2: SLIRP interacts with other NR coregulators.

Hypothesis 3: SLIRP is a nuclear protein.

These hypotheses were pursed via the aims outlined below:

Aim 1: To assess SLIRP’s role in the ER and other NR pathways with SRA, via

transient transfections and siRNA.Aim 2: To determine, using site directed mutagenesis of SLIRP’s RRMs, the

importance of the RNA binding site for a functional activity.Aim 3: To perform IP-RT-PCR and ChIP analysis of SLIRP in cancer cell lines to

assess for interaction with other NR coregulators.

Aim 4: To assess wild-type SLIRP intra-cellular localisation via confocal studies.


109

4.2 Methods

4.2.1 Tissue culture and transient transfections

Tissue culture techniques were performed as previously described in the Materials and

Methods Chapter. Essentially, in order to first repeat and then to further elaborate on

work that was performed by Lanz et al. (1999), all transient transfections wereperformed in the HeLa cell-line. All hormone treatment was at 10 nM (E2, Dex, DHT,

T3, G501515, Vit D, Tam, ICI) and experimental cells were fed hormone free media 24hr prior to hormone treatment, and samples were analysed 8 hr after hormone

treatment. Constructs were used at the concentrations detailed in the Materials and

Methods Chapter, and for each hormone treatment an appropriate luciferase reporterwas also transfected (ERE-Luc, GRE-Luc, ARE-Luc, PPAREδ-Luc, TRE-Luc, VDRE-

Luc).

4.2.2 siRNA & RNAi technology

RNAi was delivered via siRNA against SLIRP (or SKIP or SHARP) as described indetail in the Materials and Methods Chapter (Chapter 2). For these experiments, a

SMARTPOOL siRNA was employed, consisting of 4 individual siRNAs targetedagainst the same gene. Such RNAi treatment significantly enhanced target gene knock-

down. SMARTPOOL siRNA targeted against GFP was used as a control.

4.2.3 SLIRP mutants

Mutations of SLIRP were made using PCR-mutagenesis. Mutations (as described inChapter 2), based on predictions by Shane Colley and Matthew Wilce (School of

Biomedical and Chemical Sciences, University of Western Australia, Perth, Australia)were generated in order to assess the importance of SLIRP’s RRM for binding to SRA

STR7 in REMSA and transfection studies. The mutant forms of SLIRP were made by

Ross McCulloch (Royal Perth Hospital, Perth, Australia).


110

4.2.4 REMSA

REMSA was performed as has been previously described in detail in Chapters 2 and 3,and was utilized to assess the importance of SLIRP’s RRM for binding to SRA in vitro.

4.2.5 Confocal microscopy

Confocal images were taken of fluorescently tagged SLIRP and its mutants asdescribed in Chapter 2. Mitotracker was used as a mitochondrial control. SLIRP

localisation was assessed under E2 depleted and E2 saturated conditions. Microscope

images were taken by Shane Colley (Western Australia Institute for Medical Research,Perth, Australia).

4.2.6 ChIP

In 1999, Kuo & Allis published the first chromatin immunoprecipitation (ChIP)method, a versatile and high-resolution DNA-protein interaction assay with two basic

steps. The first step is in vivo formaldehyde cross-linking of intact cells, followed bythe second step, a selective immunoprecipitation of protein-DNA complexes with

specific antibodies (Kuo & Allis, 1999). In this project ChIP was performed as

optimised by Dowhan et al. (2005). ChIP was performed to assess SLIRP’s interactionwith the DNA, as well as SLIRP’s affect on other coregulator interactions at hormone-

regulated promoters. These experimental procedures are described in detail in Chapter

2. ChIP was performed by Mike Epis (WAIMR).

4.2.7 IP-RT-PCR

IP-RT-PCR was performed as explained in detail in Chapters 2 and 3. Essentially IP-

RT-PCR was used to assess for association of SRA with SLIRP and SRC-1. IP-RT-PCRs presented in this Chapter were performed by Mike Epis (WAIMR).


111

4.2.8 Immunohistochemsitry

Immunohistochemistry analysis of SLIRP and HSP-60 was carried out as described indetail in the Materials and Methods Chapter. All immunohistochemistry analysis was

carried out by Lisa Stuart (WAIMR).

4.3 Results

4.3.1 SLIRP represses SRA-mediated nuclear receptor coactivation

As SLIRP was identified from a human breast cancer library as an SRA-binding

protein I was particularly interested in assessing SLIRP’s potential role as a modulatorof E2 action. In transient transfection assays in HeLa cells using an E2-responsive

reporter, I found that SRA coactivated reporter activity approximately 3-4-fold (Figure4.1 A), as previously reported (Lanz et al., 1999). When cotransfected with SRA,

SLIRP repressed SRA-augmented coactivation of the E2-responsive reporter by up to

3-fold in a dose-dependent manner (Figure 4.1 A, B). When assessed for statisticalsignificance via a Student’s t-test, the repression is a statistically significant change

with 99 % confidence intervals, p = 0.01. I found that in HeLa cells SLIRP repressionwas SRA dependent: when SLIRP was transfected without SRA, repression was not

seen (data not shown). Interestingly, addition of SLIRP to cells cotransfected with SRA

and treated with Tam or ICI (E2 antagonists and repressors of ER signaling) furtherenhanced the E2-antagonistic activities of each of these compounds (Figure 4.1 B).

This repression was statistically significant with 99 % confidence intervals, p < 0.01 ineach case, using a Students t-test. Thus, SLIRP augmented the anti-E2 effects of both

these agents. These data convincingly define SLIRP as an ER corepressor.

Unexpectedly, when transiently transfected HeLa cells were analysed by Western blot

it was noted that when Flag-tagged SLIRP was overexpressed in the presence of SRA,

endogenous levels of SLIRP were significantly reduced (Figure 4.1 C, lanes 7 - 10).Further, when Flag-SLIRP was overexpressed in the presence of SRA and E2,

endogenous SLIRP levels were further reduced (Figure 4.1 C, lanes 8 and 10). It was ofinterest to note, when E2 is added to the transfection, FLAG-SLIRP expression is

slightly decreased, consistent with its repressor role in transfection (Figure 4.1 C, lanes


112

8 and 10). This data suggests that expression of SLIRP protein maybe tightly regulatedin the cell.

The next set of studies aimed to define whether SLIRP repressed other NRs.

Cotransfection of SLIRP with NR reporters and SRA into HeLa cells resulted in strong

repression of GR, AR, TR and VDR-mediated transactivation (Figure 4.2 A),indicating that both SRA and SLIRP can modulate several different NR signaling

pathways. To determine if SLIRP could act on orphan NRs, a peroxisome proliferator-activated receptor δ (PPARδ) agonist GW501516 (Oliver et al., 2001) was used

together with a PPAR-Luc reporter (PPARE) (Dressel et al., 2003). SRA augmented

the activation by the agonist ~2-fold, which was repressed (up to 2.9-fold) by SLIRP

(Figure 4.2 A). Statistical significance testing showed SLIRP repression to besignificant in all NR pathways, p = / < 0.01, that is, significant with 99 % confidence

intervals. These data suggest that SLIRP has broad corepressor activity within the NRsuperfamily.

To complement our SLIRP overexpression studies, the effects of SLIRP siRNA onDex-responsive reporter activity was investigated in HeLa cells. In cells with reduced

endogenous SLIRP expression, a 10-fold increase in GRE-luc activity was found,

further confirming that SLIRP acts as a NR corepressor (Figure 4.2 B).

4.3.2 SLIRP modulates SHARP- and SKIP-mediated coregulation of NR activity

The high aa sequence homology between SHARP and SLIRP and their avid binding to

SRA STR7 in vitro suggested a functional interaction may exist between thesemolecules in vivo. When cotransfected with SRA, SHARP repressed SRA-mediated

coactivation of the E2-responsive reporter (Figure 4.2 C, lane 4), as previously reported(Shi et al., 2001). When SLIRP was cotransfected with SHARP and SRA, an additional

2-fold repression of SRA-augmented coactivation was observed (Figure 4.2 C, lane 5),

the repression was statistically significant, p < 0.05. Thus, SHARP and SLIRPappeared to act in an additive fashion to enhance repression of the E2-responsive

reporter.

The effects of SKIP on SLIRP repression in further transfections was also investigated

(Figure 4.2 C, lanes 6-12). In the presence of transfected SKIP alone, reporter activity


113

was increased ~2-fold, consistent with SKIP functioning as a coactivator of ERtransactivation. In the presence of cotransfected SRA, an additive effect was observed

with a total increase in activity of ~6-fold. When SLIRP was added, reporter activitywas reduced by more than 5-fold, the change between each individual transfected

constructs and the combination of SLIRP and SKIP is statistically significant (p <

0.0003). Thus, in the presence of SRA, SLIRP is a potent repressor of SKIP-mediatedcoactivation. When endogenous SLIRP expression was reduced using siRNA, SLIRP

repression was abrogated. Reduction of SKIP expression with siRNA reversed SKIP’scoactivation effect. When expression of SLIRP and SKIP were both reduced, an

intermediate reporter activity resulted. Taken together, these data validate the

functional role of each of these coregulators on ER transactivation and suggest acompetitive interaction exists between SLIRP and SKIP in NR signaling.

4.3.3 SLIRP function requires an intact RRM domain

To investigate the structural and functional significance of the RRM domain withinSLIRP, the properties of proteins with mutations to this motif was assessed (Figure 4.3

A). Based on binding predictions from other RRM-containing proteins, arginine 24 and25 were mutated to alanines (R24, 25A) in the RNP2 submotif and within the RNP1

domain leucine 62 was mutated to alanine (L62A). A double mutant (DM) containing

both the R24,25A and L62A substitutions was also prepared. In REMSA studies, eachof the mutations markedly reduced binding to the SRA STR7 probe (Figure 4.3 B,

lanes 4-9 and Figure 4.3 C). In transfections, each mutant partially relieved the SLIRP-

mediated repression (Figure 4.3 D, lanes 4 [p = 0.07], 5 [p < 0.05] & 6 [p < 0.05]),indicating the requirement of an intact RRM domain for SLIRP to function as a

repressor of E2-induced SRA coactivation.

To examine the functional specificity of the SLIRP-STR7 interaction in vivo, the SRA-

SDM7 mutant (Lanz et al., 2002) was utilized, in which the stem-loop structure ismutated, but preserved (Figure 4.3 D). This mutation decreased SRA-mediated

coactivation to ~70 % of wild-type levels. Furthermore, when SLIRP was cotransfectedwith SRA-SDM7, SLIRP was unable to function as a repressor. This lack of repression

was statistically significant, p = 0.014. These data suggested that a direct interaction

between SLIRP and STR7 is critical for SLIRP’s repressive activity.


114

4.3.4 SLIRP is recruited to endogenous NR target promoters

To determine if SLIRP is recruited to E2- and Dex-responsive promoters, chromatinimmunoprecipitation (ChIP) assays were performed. It was found that SLIRP was

recruited to the E2-responsive pS2 promoter within 60 min of ligand treatment (Figure

4.4 A), but this association had returned to undetectable levels by 120 min. ERα

binding increased in response to ligand returning to basal levels within 120 min. Incontrast HuD, another well-characterized RRM-containing RNA-binding protein

(Chung et al., 1996), was not recruited to the DNA. These data confirmed that SLIRPcan closely associate with the response element of an E2-regulated gene.

To investigate the mechanism by which SLIRP might mediate its effect at thetranscriptional level, ChIP assays were performed in HeLa cells treated with SRA

siRNA. Interestingly, in cells with reduced SRA expression, ~50 % less SLIRP wasrecruited to the Dex-responsive metallothionein promoter (Figure 4.4 B, lane 10). This

suggests that the presence of SRA is critical for recruiting SLIRP to the promoter and

consequently mediating SLIRP’s repressive effects.

To investigate interactions of SLIRP with other corepressors, ChIP studies were

performed in cells treated with SLIRP siRNA. In the absence of E2, NCoR is recruitedto the pS2 promoter together with a small amount of ER (Figure 4.4 C). However, in

cells with reduced SLIRP, NCoR could no longer be detected on the promoter and ERrecruitment was significantly higher. This suggests a key role for SLIRP in facilitating

recruitment of NCoR to the promoter.

As SRA also co-purifies with SRC-1 associated with AR (Lanz et al., 1999), the effects

of reducing intracellular SLIRP levels with siRNA were examined and it was foundthat there was a corresponding increase in SRC-1 associated with SRA (Figure 4.4 D,

E). This suggests competition may exist between SRC-1 and SLIRP for association

with SRA in vivo, which could directly impact on their coregulator effects.

4.3.5 SLIRP is predominantly mitochondrial

Based on the transfection and ChIP data presented so far, it was envisaged that SLIRP

would be a predominantly nuclear protein. However, imaging studies using the SLIRP


115

ab revealed endogenous SLIRP to have a filamentous cytoplasmic distribution confinedpredominantly to the mitochondria (Figure 4.5 A, top panel). A similar pattern was

observed using ab to HSP-60, a mitochondrial-specific protein (Figure 4.5 A, secondrow) (Gupta & Knowlton, 2005). In cells transfected with FLAG-tagged SLIRP, it was

found that SLIRP colocalized with another mitochondrial-specific protein cytochrome

c oxidase (Figure 4.5 A, third row). Sequence analysis of SLIRP revealed a N-terminal26 aa domain highly predictive of an amphipathic α-helical mitochondrial targeting

sequence conserved between the mouse, rat and human genomes (see Figure 3.2 C).

This mitochondrial signal sequence is evident, in the 3-D predicted structure of SLIRP,as an independent helix linked to the RRM (Figure 4.5 B). To evaluate the importance

of the N-terminal signal sequence, the intracellular localization of SLIRP-FLAG versus

FLAG-SLIRP constructs were compared. Interestingly, SLIRP-FLAG localized to themitochondria, whereas FLAG-SLIRP was pan-cellular (Figure 4.5A, bottom two rows).

These findings are consistent with the notion that the N-terminal mitochondrial signalsequence is critical for targeting SLIRP to the mitochondria.

Primary human breast tissues were examined with SLIRP and HSP-60 abs to furtherinvestigate the mitochondrial location of endogenous SLIRP. A punctate cytoplasmic

staining pattern, characteristic of mitochondria, was observed with both SLIRP and

HSP-60 abs (Figure 4.5 C). Taken together, these data confirm that SLIRP residespredominantly in the mitochondria and that interference with the N-terminal signal

sequence can substantially alter the intracellular distribution of the protein.

A triple mutant of the first three arginines in SLIRP (R7, 13, 14A) was generated to

evaluate the functional importance of the mitochondrial signal sequence (see Figure 4.5B). When cotransfected into HeLa cells with SRA, this SLIRP mutant had statistically

significant (p < 0.01) reduced ability to function as a repressor (Figure 4.5 D). This datasuggested that the mitochondrial signal sequence is required for maintaining

corepressor function, raising the possibility that SLIRP has bifunctional capacity as a

NR corepressor in the nucleus and mitochondria.

SRA’s cellular localisation was questioned after determining that SLIRP could exist inboth the nucleus and the mitochondria of the cell. It is of interest to note that earlier this

year (2007), SRA was found to localise 90 % of the time to the cytoplasm (Zhao et al.,


116

2007) (Figure 4.5E), which raises the possibility of dual roles for both SRA and SLIRPin the nucleus and also outside of the nucleus.

4.4 Discussion

The studies in this Chapter have shown that SLIRP can act as a potent repressor of E2,

glucocorticoid, androgen, thyroid hormone (T3) and VitD action. In addition, SLIRP

represses orphan NR activity as shown by its effects on PPARδ–mediated

transactivation. SLIRP also interacts with other NR coregulators (NCoR and SRC-1).This data defines SLIRP as a NR corepressor.

Shi et al., (2001) showed the coregulator activity of SHARP requires its RRM. Thisproject observed similar findings with SLIRP in that discrete single and double aa

substitutions of the RRM domain significantly reduced its SRA-binding andcorepression activities. When the L62A and the DM SLIRP mutants were transfected, a

statistically significant change was seen. While a statistically significant change was

not seen with the R24, 25A mutant, the trend was the same, and the p value was closeto significance (p = 0.07).

The SRA STR7 stem-loop is the longest and one of the most stable identified by 20

structure predictions and accounts for a substantial proportion of SRA’s overall

coactivator activity (Lanz et al., 2002). The SRA mutation data presented in thisChapter shows that STR7 is required for SLIRP to act as a corepressor, which further

strengthens the case for a direct interaction between STR7 and SLIRP in vivo.Reduction of endogenous SLIRP expression increases SRA’s coactivation ability,

which suggests not only that this interaction is functionally relevant but also that

SLIRP could play an important tumor-suppressor role in SRA-activated NR pathways.The additive repressive effect of SHARP and SLIRP, both of which bind to SRA

STR7, suggests these proteins could function to significantly down-regulate ERsignaling in breast cancer cells.

These studies provide new insight into the mechanism of interaction between SRA,SLIRP and SRC-1. In particular, it was found that recruitment of SLIRP to an

endogenous Dex-responsive promoter is regulated by the amount of SRA in the cell.


117

Furthermore, this Chapter has shown that SRC-1 and SLIRP appear to compete forassociation with SRA. Specifically, when SLIRP levels are reduced, SRC-1 association

with SRA increases. This is consistent with the opposing function of these twocoregulators: the association of SRC-1 with SRA results in coactivation whilst

association of SLIRP with SRA results in corepression.

The data resulting from ChIP experiments performed in cells with reduced SLIRP

expression provide additional mechanistic insights. The results suggest SLIRP isessential for mediating NCoR’s association with the promoter in the absence of E2.

Most interestingly, in SLIRP siRNA treated cells not only is binding of NCoR

abrogated but the ER is strongly recruited, which suggests that removal of SLIRP fromthe cell alters the promoter state from one of repression to activation.

This project has provided intriguing observations that SKIP and SLIRP colocalize to

human Chr 14q24.3 and that this genomic distribution is conserved across species. This

raised the possibility that they may be co-ordinately regulated. However, the expressiondata (see Chapter 3) did not support this theory. The transfection and siRNA studies in

this project suggest that they have opposing and possibly competitive effects on E2signaling, rather than working in concert. Given the interaction between SKIP and

NCoR in VDR transactivation (Leong et al., 2004), these studies with SKIP and NCoR

suggest a complex role for SLIRP in modulating VDR signaling.

The increased expression of SLIRP in high-energy demand and mitochondria-rich

tissues, such as skeletal muscle, heart and liver, is consistent with its predominantlymitochondrial location. Multiple imaging studies suggest that more than 90 % of

SLIRP is located in the mitochondria, which raises the possibility that it may functionboth in the nucleus and mitochondria to regulate NR activity. Whether this is via

interactions with SRA or other mitochondrial RNA targets is unknown. In addition,

SLIRP’s capacity to represses PPARδ-mediated-signaling suggests a potential role in

regulating lipid homeostasis in energy-rich tissues. SLIRP does not have a nuclearlocalisation sequence, but as it is significantly smaller than 40 KDa, it is potentially

capable of traversing the nuclear membrane without one (Koolman & Roehm, 2005).

The role of NRs in the mitochondrion, affecting cell survival and energy homeostasis,

has recently come under close scrutiny. Studies in breast cancer tissues suggest


118

mitochondrial ER plays a role in tumor cell survival (Pedram et al., 2006).Additionally, T3 can induce transcription in the absence of nuclear factors acting via

mitochondrial TR (Scheller & Sekeris, 2003). GR interacts with NFκB subunits (Tao et

al., 2001) present in the mitochondria (Cogswell et al., 2003) and putative GREs existin some key components of the mitochondrial genome-encoded oxidative

phosphorylation pathway (Psarra et al., 2006). Taken together, these data provide a

foundation for NR and coregulator action in the mitochondria and a rationale forSLIRP’s presence there.

The data presented in this Chapter suggests a complex role for SLIRP both in NR and

mitochondrial biology. However, while the over- and under-expression studies

presented here have been essential in quantifying SLIRP’s role in transcription in vitro,a project analysing a mouse knock-out of SLIRP has already commenced in the

laboratory of my Supervisor and will be a critical whole animal assessment of SLIRP’sbiological function.

Conclusions

In this Chapter I have detailed SLIRP as a potent and wide-ranging NR corepressor that

can interact with other NR coregulators. The data suggests SLIRP may have dualfunction in the nucleus and in the mitochondria. Taken together with SLIRP’s ability to

bind SRA and the recent discovery that SRA localises to the cytoplasm, these data areemphasize the important and emerging links between nuclear and mitochondrial

signaling and reinforce the increasing role of ncRNAs (such as SRA) in human cellular

biology.

A

E2SRA

SLIRPTam

ICI

+----

++---

++

--

++

--

++

--

00.5

11.5

22.5

33.5

44.5

5ERE

E2

Luc

++++-

+++-+

++--+

++-+-

Fold

tran

sact

ivat

ion

E2SRASLIRP

+--

++-

+++

B

Figure 4.1 A, B: SLIRP is an SRA-dependent repressor of ER transactivation. PanelA: SLIRP represses ER transactivation. Using a Student’s t-test to assess for statisticalsignificance the repression seen with SLIRP is statistically significant with 99 %confidence intervals, p = 0.01, indicated with the *. Panel B: SLIRP repression isaugmented by Tam and ICI. HeLa cells were cotransfected with ERE-luciferase(Luc), expression vectors for ERα ± SRA and increasing amounts of SLIRP. After 24hr, cells were treated for 8 hr with E2 prior to assessment of Luc activity (normalizedto protein). Tam or ICI were added where indicated at same time as E2. The *indicates statistical significance at 99 % confidence intervals, p < 0.01. All results arerepresentative of triplicate experiments; error bars represent standard deviation.

ERE

E2

Luc

00.5

11.5

22.5

33.5

44.5

5

Fold

tran

sact

ivat

ion

119

*

**

C

66

45

30

20

14

KDa 1 2 3 4 5 6 7 8 9 10

Figure 4.1 C: Endogenous SLIRP is tightly regulated. Panel C: HeLa cells werecotransfected with ERE-Luc and expression vectors for ERα, pSCT or pSCT-SRA,FLAG or FLAG-SLIRP. Cell protein lysates were then generated and then analysed ona Western gel and exposed to ECL. It is of interest to note that when FLAG taggedSLIRP is overexpressed, endogenous SLIRP is significantly decreased. Further, whenE2 is added to the transfection, FLAG-SLIRP expression is decreased. The + indicatesaddition of the named construct or ligand, ++ indicates double the amount of the + (ie.,for SLIRP amounts are 1 µg and 2 µg).

ERERE-Luc

E2pSCT

pSCT-SRAFLAG

FLAG-SLIRP

++-----

+++----

++++-+-

++++--+

+++-++-

++--+-+

+++-+-+

++--+-++

+++-+-++

β−actin

SLIRP

FLAG-SLIRP

120

Figure 4.2 A, B: SLIRP represses signaling of multiple NR pathways. Panel A: HeLacells were cotransfected with either a GRE-Luc, ARE-Luc, TRE-Luc, VitD-Luc orPPARE-Luc reporter plus corresponding AR, TR, VDR, PPARδ, SRA and SLIRPexpression vectors, incubated with ligand (Dex, DHT, T3, VitD, GW501516) for 8 hand Luc activity determined as above. The * indicates statstical significance assessedvia a Student’s T-test, 99 % confidence intervals, p < 0.01 for each NR pathway. PanelB: Targeted reduction of SLIRP expression potentiates GR transcription. HeLa cellswere cotransfected with GRE-Luc and siRNA directed against either SLIRP or anonsense target. After 48 h, cells were treated with Dex (8 h) prior to assessment of Lucactivity. Immunoblot confirmed reduced endogenous SLIRP expression in SLIRPsiRNA treated cells (lane 1) compared with nonsense (lane 2) relative to β-actin.

DexDHT

T3VitD

GW501516SRA

SLIRP

00.5

11.5

22.5

33.5

44.5

5

Fold

tran

sact

ivat

ion

-+-----

-+---+-

-+---++

--+----

--+--+-

--+--++

---+---

---+-+-

---+-++

+------

+----+-

+----++

GRE

Dex

Luc LucARE

DHT

Luc TRE

T3

Luc VDRE

VitD

Luc LucPPARE

GW501516

Luc

----+--

----++-

----+++

A

B

DexSRA

SLIRP siRNANonsense siRNA

+++-

++-+

GRE Luc

0

2

4

6

8

10

12

Fold

tran

sact

ivat

ion

SLIRP

β-actinSLIRP siRNA

Nonsense siRNA+-

-+

1 2

Dex

121

*

** * *

C

E2SRA

SLIRPSHARP

SKIP

Figure 4.2 C: SLIRP augments SHARP’s repression and antagonizes SKIP’scoactivation of ER. Panel C: HeLa cells were cotransfected with ERE-Luc andexpression vectors for ERα alone, and/or empty, SHARP, SLIRP or SKIP vectors, +/-siRNA (nonsense, SLIRP or SKIP). The * indicates statistical significance with 95 %confidence intervals, p < 0.05. RT-PCR confirmed reduced SLIRP and SKIPexpression in siRNA treated cells (lower panel).

0

1

2

3

4

5

6

7

++---

++++-

+---+

++-+-

+----

++--+

+++-+

ERE

E2

Luc

Fold

tran

sact

ivat

ion

+++--

SLIRP

β-actin

Nonsense siRNASLIRP siRNA

SKIP siRNA

+--

-+-

SKIP

--+

-++

1 2 3 4

122

*

**

*

SLIRP-FLAG

SLIRP (L62A)-FLAG

SLIRP (R24,25A)-FLAG

FLAG

FLAG1 10924

RNP2 RNP1

FLAG1 10962

SLIRP DM (R24,25A,L62A)-FLAG FLAG1 10924 62

1 10921 26 54 61

SLIRP (R7,13,14,A)-FLAG FLAG1 10915

A

B1 4 5 62 3

+----

+

---

+

---

+-

--

+-

--

+--

-

7 8 9

+--

-

+---

+---

STR7-SLIRP

STR7 ProbeGST-SLIRP

GST-SLIRP-R24,25AGST-SLIRP-L62A

GST-SLIRP-R24,25A,L62A

C

Figure 4.3 A, B, C: SLIRP protein RRM mutant studies. Panel A: Plasmids forexpression of wild-type and mutant SLIRP with carboxy terminal FLAG epitope.Panel B: Mutation of the RRM domain abrogates binding of SLIRP to SRA. REMSAusing labeled SRA STR7 probe and increasing amounts of GST-SLIRP fusionproteins (wild-type, mutants R24,25A, L62A or double mutant R24,25A,L62A).Panel C: Immunoblot of GST-fusion proteins used in mutant REMSA binding studies.

1 2 3 4 5

GST-SLIRPGST-SLIRP-R24,25A

GST-SLIRP-L62AGST-SLIRP-R24A,25/L62A

GST

+----

-+---

--+--

---+-

----+

123

D

E2 SRA SLIRP

R24,25A L62A DM

SRA SDM7

00.5

11.5

22.5

33.5

44.5

5

+------

+++----

++--+--

++---+-

++-+---

++-----

+-----+

+-+---+

ERE Luc

E2

Fold

tran

sact

ivat

ion

Figure 4.3 D: SLIRP mutants have reduced ability to repress ER activity. Panel D:HeLa cells were transfected with ERα, ERE-Luc and either wild-type (SLIRP-FLAG)or mutated SLIRP-FLAG (R24,25A, L62A, DM) expression vectors together withSRA or SRA SDM7 (stem-loop mutant) and reporter activity assessed as in A above.The * indicates statistical significant change p < 0.05, with 95 % confidenceintervals.

124

**

B SRA siRNA

NS siRNA

SRAβ-actin

Dex - + - +

A

- Dex

+ Dex

InputNo AbGR Ab

SRC-1 AbSLIRP Ab

1 2 3 4 5 6 7 8 9 10

NS siRNA SRA siRNA

+----

-+---

--+--

---+-

----+

+----

-+---

--+--

---+-

----+

C

+----

-+---

--+--

---+-

----+

InputNo AbER Ab

SLIRP AbHuD Ab

0 min60 min

120 min

1 2 3 4 5

InputNo AbER Ab

NCoR Ab

+---

-+--

--+-

---+

-+--

--+-

---+

+---

1 2 3 4 5 6 7 8

- E2

NS siRNA SLIRP siRNA

+ E2

SLIRP

NS siRNASLIRP siRNA

+-

-+

1 2

β-actin

Figure 4.4 A, B, C: SLIRP is recruited to endogenous promoters and modulates NCoRrecruitment. Panel A: ChIP assay demonstrating recruitment of SLIRP and ER, but notHuD, to the pS2 promoter of MCF-7 cells in response to E2. Sheared, genomic, MCF-7DNA used as input control. Panel B: Recruitment of SLIRP to the metallothioneinpromoter is regulated by SRA. HeLa cells treated with SRA siRNA or non-sense siRNA(NS siRNA) for 3 days, were incubated with Dex and then ChIP assays performed witheither GR, SRC-1 or SLIRP ab (left panel). RT-PCR shows knockdown of SRA withoutaffecting β−actin (right panel). Panel C: SLIRP regulates NCoR association with the pS2promoter. MCF-7 cells were treated with either SLIRP siRNA or NS siRNA (3 days)followed by E2 for 45 min before ChIP assay using ER or NCoR ab as above (leftpanel). RT-PCR demonstrating SLIRP knockdown (right panel).

125

Supernatants

Beadsβ-actin

SLIRP

SRA

+-+-+-+-+-+-+-++--++-

-+-

+-+--+-+++--

SLIRP siRNANo ab

SRC-1 ab

SLIRP siRNANo ab

SRC-1 ab

β-actin

SLIRP SRC-1

+++-SLIRP siRNA

SRC-1 ab

D

Nonsense siRNA --++----

Nonsense siRNA --++

Nonsense siRNA -+

Figure 4.4 D: SLIRP knockdown augments SRC-1 association with SRA. Panel D:Lysates of MCF-7 cells treated with SLIRP siRNA were incubated with no ab orSRC-1 ab and SRA detected as above. SLIRP was significantly knocked downwithout affecting β-actin or SRA. Substantially more SRA co-purified with SRC-1 inSLIRP siRNA treated cells than nonsense controls. Below is an immunoblotconfirming SLIRP protein knockdown without affecting SRC-1.

126

A

HeLa

HeLa

HeLa + SLIRP-Flag

HeLa + Flag-SLIRP

HeLa + SLIRP-Flag

Wt-SLIRP MitoTracker Overlay

Cyto C

HSP-60

MitoTracker

MitoTracker

Flag

Flag

Flag

Wt-SLIRP

Nuclear

Overlay

Overlay

Overlay

Overlay

Nuclear

Nuclear

Nuclear

Nuclear

Figure 4.5 A: SLIRP localizes predominantly to the mitochondria Panel A:Simultaneous mitochondrial (red, Mitotracker), nuclear (blue, Hoescht 33256) andendogenous SLIRP protein staining (green, rabbit polyclonal sera, AlexaFluor 488secondary ab) of HeLa cells. Overlaying of confocal images reveals colocalization ofSLIRP and the mitochondria (yellow) (top row). Endogenous SLIRP also colocalizeswith mitochondria specific HSP-60 (second row). Transfected SLIRP-FLAGcolocalized with cytochrome c (middle row) and Mitotracker stain (fourth row).Transfected FLAG-SLIRP was pan-cellular and did not colocalize with theMitotracker stain (bottom row).

127

R13 Mitochondrial Signal Sequence(α-helix)

R7

R14

L62R24

RNA Binding Surface (β-sheet)

C

B

SLIRP HSP-60

Figure 4.5 B, C: SLIRP is predominantly mitochondrial. Panel B: Three-dimensionalmodelling of the SLIRP protein predicts a mitochondrial localization signal in theamino terminal 26 aas. Residues subjected to point mutation are indicated. Panel C:SLIRP and HSP-60 stain similarly in human breast cancer tissue. IHC of primaryhuman breast cancer tissue using either SLIRP or HSP-60 abs. Ducts stained readilywith both abs in a punctate cytoplasmic pattern, consistent with a mitochondriallocation for HSP-60 and SLIRP.

128

Imaging Studies

E

D

+++-

E2 SRA

SLIRP R7,13,14A

+---

++--

++-+

00.5

11.5

22.5

33.5

44.5

5

ERE Luc

E2

Fold

tran

sact

ivat

ion

Figure 4.5 D, E: Mutations in the SLIRP mitochondrial sequence relieve its repressiveactivity. Panel D: HeLa cells were transfected as previously with ERE-luc, ERα, wild-type SRA, and either wild-type SLIRP or the R7,13,14A mutant. The * indicatesstatistical significance with 99 % confidence intervals, p < 0.01. Results arerepresentative of triplicate experiments; error bars represent standard deviation. SRA iscytoplamsic Panel E: Image taken from Zhao et al. (2007), p. 692, where in MCF-7 cellstransfected with tagged SRA, and E2, SRA was found to be cytoplasmic approximately90 % of the time (these cells are indicated by black arrows), the other 10 % of the timethey reported SRA to be localized to distinct nuclear subcellular compartments (thesecells are indicated by white arrows).

129

*

130

E. Hatchell Chapter 5: SLIRP and estrogen signaling in breast cancer

131

Chapter 5

SLIRP and estrogen signaling

in breast cancer


132

5.1 Introduction

5.1.1 Preface

As SLIRP is a NR corepressor of the ER pathway, and is detected in primary humanbreast cancers, it was considered particularly important to assess SLIRP’s biological

function in breast cancer cells. In this Chapter I detail experiments performed in a

human breast cancer E2-dependent cell-line. The data from these studies revealed thatSLIRP functions as an ER corepressor in breast cancer cells in an SRA- and E2-

dependent manner. Interestingly, targeted knock-down of SLIRP affects endogenousHSP-60 levels, a mitochondrial protein whose high expression in human breast cancers

is related to a poor prognosis. Knock-down of SLIRP also increases cell proliferation,

which is consistent with its role as a corepressor of ER transactivation. Thus, the datapresented in this Chapter illustrates a complex role for SLIRP in the ER signaling

pathway of E2-dependent breast cancer cells.

5.1.2 Hormone-dependent cancers

Breast and prostate cancer, the two major hormone-dependent cancers, cause

significant morbidity and mortality. Estrogens and androgens, unlike other steroid

hormones which induce cell differentiation, induce cell proliferation and play afundamental role in the growth and development of breast and prostate cancer (Cheng

& Balk, 2003; Ko & Balk, 2004).

There is significant overlap of the molecular mechanisms underlying breast and

prostate cancer. In prostate caner, E2 has a significant influence over tumor severityand can increase or decrease the proliferation of cancer cells depending on the receptor

status of the tumor (Roger et al., 2005; Carruba, 2006). In breast cancer, E2 issynthesised from androgens (di-hydroxy-testosterone; DHT) by the cytochrome P450

complex known as aromatase (Brueggemeier et al., 2005). Aromatase inhibitors are

currently used as part of the armamentarium of endocrine treatment for E2-dependentcancer (Brueggermeier, 2001). Interestingly, there is increased survival after diseaserecurrence with AR overexpressing breast cancer (Schippinger et al., 2006) and there is

significant evidence to suggest that E2 in synergy with an androgen (DHT) has a major


133

role in prostate tumor pathogenesis (Ho et al., 2006). Thus, for each of breast andprostate cancer, a better understanding of steroid receptors and the mechanisms through

which they exert their effects in normal and tumor tissue will be key to developingnovel treatments targeted at the hormone signaling pathways (Cordera & Jordan, 2006).

5.1.3 Estrogen and cell growth in breast cancer

E2 is the most biologically active hormone in the developing breast (Russo et al., 1999)and is considered to play a major role in growth of normal and neoplastic breast cells

(Russo & Russo, 2006). Epidemiological data indicates that long term exposure to E2

increases breast cancer risk (Dorgan et al., 1996) and unnatural E2 deprivation (e.g.premature menopause) is associated with a reduced incidence of breast cancer (Howell

et al., 2007). E2 is hypothesised to act as a carcinogen in three ways: (i) as a stimulatorof cell growth through its receptor activity; (ii) as an inducer of aneuploidy; and (iii) as

an inducer of mutations through cytochrome P450-mediated metabolic activity (Chen

& Yager, 2004; Russo & Russo, 2006).

Most mammalian cytochrome p450 exists as membrane bound proteins (Figure 5.1) ineither the endoplasmic reticulum or the inner mitochondrial membrane (Seliskar &

Rozman, 2006). In humans, mitochondrial cytochrome P450 is involved with steroid

hormone (including E2 and DHT) and vitamin D3 biosynthesis (Seliskar & Rozman,2007). Furthermore, mutations in cytochrome P450 components have been shown to

lead to a significant variety of cancers, including ovarian (Leung et al., 2005) and

breast cancer (Segersten et al., 2005).

5.1.4 Mitochondrial proteins in breast cancer

Nowadays we are facing a renaissance of the role of mitochondria in cancer biology.

- Isidoro et al., 2005 (p. 2095)

Recent literature highlights the elegant cross-talk and dependency between the

mitochondria and the nucleus in normal and cancerous cells. Several mitochondrialproteins (as well as cytochrome P450 components) have been significantly associated

with cancer development (Carew & Huang, 2002; Cuezva et al., 2002). In breastcancer, a high frequency of somatic mitochondrial DNA mutations has been associated


134

with tumor development (Parrella et al., 2001; Tan et al., 2002), and abnormalmetabolic signatures have been associated with more advanced breast tumors (Isidoro

et al., 2005). With the hope of one day contributing to our current breast cancerdiagnosis and treatment regimes, recent attempts to analyse the genes and proteins

associated with a metabolic phenotype of breast cancer tumors have led to a greater

understanding of metabolic protein action in breast cancer. The dysregulation ofseveral mitochondrial proteins critical for cellular homeostasis has been implicated in

breast cancer (Isidoro et al., 2004). These proteins include pyruvate kinase (PK), β-F1-

ATPase, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and heat shock protein60 (HSP-60) (Figure 5.2 A, B) (Isidoro et al., 2004; Li et al., 2006).

5.1.5 HSP-60

HSP-60 is a member of the heat shock (HS) family of proteins. HS proteins are inducedin order to limit damage to the cell caused by stress; they facilitate cellular recovery

(Arya et al., 2007). Each heat shock protein acts differently under stress conditions: for

example, HSP-90 is involved in activating apoptotic pathways, HSP-27 and HSP-70are anti-apoptotic (Arya et al., 2007), and HSP-60 is both apoptotic and anti-apoptotic

(Chang et al., 2006; Arya et al., 2007). HSP60 is primarily known as a mitochondrial

protein important for folding proteins into their mature conformation after import intothe mitochondria (Hood et al., 2003; Di Felice et al., 2005; Gupta & Knowlton, 2005).

Highlighting the cross-talk between the nucleus and the mitochondria is the discoverythat T3 increases the rate of mitochondrial protein import thus altering active HSP

levels in the mitochondria (Hood et al., 2003).

Mitochondrial HSP-60 protein levels are perturbed in several diseases and disorders.

When mitochondrial HSP-60 is redistributed to the cytosol in rats with cardiovasculardepression during endotoxaemia, fatality is significantly decreased due to the loss of

apoptotic signaling through mitochondrial Bax and Bcl-2 (Chang et al., 2006).

Increased levels of HSP-60 have been associated with increasing severity ofatherosclerosis (Mandal et al., 2005). Whilst in the oncology clinic, perturbation of

HSP-60 protein has been implicated in ovarian, cervical (Di Felice et al., 2005),colorectal, prostate (Cappello et al., 2007) and breast cancer (Isidoro et al., 2005).


135

Increased HSP-60 levels have been observed in advanced breast tumors whencompared to earlier stage carcinomas, and women with higher levels of HSP-60 have

significantly less disease free months than women with low expression levels of HSP-60 (Figure 5.2 B) (Isidoro et al., 2005; Li et al., 2006). HSP-60 is a molecular modifier

of alpha 3 beta 1 integrin, which enables breast cancer cells to attach to lymph nodes,

enhancing the metastatic potential of cancerous cells (Barazi et al., 2002). In thelaboratory, HSP-60 protein increases in response to E2 treatment of the E2-dependent

MCF-7 cell-line (Kim et al., 2005). Given the fundamental role of HSP-60 inmitochondrial apoptosis pathways and its expression in many cancers including breast

cancer, it is not surprising that HSP-60 is currently under intense investigation as both a

biomarker (Czarnecka et al., 2006) and an anti-tumor molecular agent (Cappello et al.,2007).


This Chapter will investigate the functional role of SLIRP in breast cancer cells, and apossible relationship between SLIRP and HSP-60 expression. The existence of such a

relationship could illustrate a novel mechanism of cross-talk between nuclear andmitochondrial signaling in breast cancer. The following list of hypotheses guided my

research for this Chapter.

Hypothesis 1: SLIRP is an ER corepressor in breast cancer cells.

Hypothesis 2: SLIRP is an AR corepressor in prostate cancer cells.

Hypothesis 3: SRA transactivation, and thus cell growth, is increased when SLIRPlevels are reduced.

Hypothesis 4: Alteration of SLIRP expression affects mitochondrial proteinexpression.

To this end, Chapter 5 had the following aims:

Aim1: To assess the functional role of SLIRP as a coregulator in two hormone-dependent breast cancer cell lines (MCF-7 and MD-MB-468) and a prostate cancer cell

line (22RV1).

Aim 2: To determine the functional effects on NR-mediated transactivation withSLIRP levels are decreased in cells..


136

Aim 3: To analyse the effects of reduced SLIRP in breast cancer cell proliferation.Aim 4: To examine the expression of SLIRP and HSP-60 in breast cancer cells.

5.2 Methods

5.2.1 Tissue culture and transient transfections

Transient transfections were performed with the concentrations of DNA as described indetail in the Materials and Methods Chapter. Transfections were predominantly

performed in MCF-7 cells, with the addition of a growth medium containing insulinthat optimises E2 sensitivity. It is known that there is significant cross over between

insulin and E2 pathways: insulin is able to control MCF-7 cell cycle progression

(Chappell et al., 2001; Felty, 2006). MDA-MB-468 (ER- breast cancer cells) and22RV1 (androgen dependent prostate cancer cells) were transfected to compare

SLIRP's role under different hormone-dependent conditions. Cells were not transfected

with a receptor plasmid, except when using MDA-MB-468 cells where the ERα was

cotransfected to allow analysis of another hormone-dependent cell system.

5.2.2 RNAi transfections

Cells were treated with either a SMARTPOOL (a group of 4) siRNA against SLIRP or

a single SLIRP siRNA, as detailed in the Materials and Methods Chapter. The SLIRPsiRNA sense sequences are as follows: SLIRP siRNA # 1: cga guc agc uga aag aac

auu; SLIRP siRNA # 2: uca auc agc cgg uug cuu uuu; SLIRP siRNA # 3: gca cag uucggc cau guc auu; SLIRP siRNA # 4: guu cac acu aga agg cca auu. SLIRP

SMARTPOOL siRNA included a mix of all four sequences. RNAi treatments were

first optimised and for all experiments cells were treated with a standard dose of 50 nMsiRNA and left for 3 days before knock-down was assessed.

5.2.3 Western blot

Whole cell lysates were collected with 150 µL passive lysis buffer (Promega) per well

of a 6 well plate. Cells were treated with siRNA against SLIRP (or GFP for a control),


137

and for some experiments this was followed by a transient transfection. Further detailcan be found in the Materials and Methods Chapter.

5.2.4 Cell titre

MCF-7 cells were grown in 96 well plates and treated in the presence or absence of E2.Cell titre analysis of plates occurred every 24 hours over a period of 5 days. Each

experiment was repeated in triplicate, and knock-down was detected via Western blotanalysis. Further details can be found in the Materials and Methods Chapter.

5.3 Results

5.3.1 SLIRP represses SRA-mediated NR coactivation in breast cancer cells

As SLIRP was identified from a human breast cancer library and SRA is an activator of

ER signaling in breast cancer cells (Lanz et al., 1999), I planned to initially assessSLIRP’s potential role as a modulator of E2 action in breast cancer cells. In transient

transfection studies in the ER+ breast cancer cell line MCF-7 using an E2-responsiveluciferase reporter, SLIRP repressed E2-mediated signaling as had previously been

found to be the case in HeLa cells. SLIRP repressed ER transactivation 2–3 fold

(Figure 5.3 A). In MCF-7s, an E2 dependent cell line, SLIRP repression was E2- andSRA-dependent.

The same experiment was repeated in 2 other cell types. First, in MDA-MB-468s, an

ER- breast cancer cell line where SLIRP, SRA, ERE-luc and an ERα plasmid were

cotransfected (Figure 5.3 B). Second, in 22RV1s, a DHT-dependent prostate cancer

cell line, enabling assessment of SLIRP function in the AR pathway (Figure 5.3 C). Ineach case similar results were found: SLIRP repressed SRA transactivation. In MDA-

MB-468 cells this repression was ~ 2-fold, whereas in 22RVI cells repression wassignificantly greater and up to 10-fold.


138

5.3.2 Effect of SLIRP RNAi on estrogen signaling

After undertaking SLIRP overexpression assays in MCF-7 cells, complementarySLIRP RNAi studies were employed to assess the impact of reducing endogenous

SLIRP on SRA mediated ER transactivation. Using either a SMARTPOOL siRNA or

one single siRNA targeted against SLIRP (siRNA #4), it was found that reduction ofendogenous SLIRP from MCF-7 cells significantly increased SRA mediated ER

transactivation (Figure 5.4 A). Furthermore, when SLIRP was cotransfected back intocells that had been treated with SLIRP siRNA repression of E2 signaling was observed

(Figure 5.4 B).

5.3.3 Reduction of SLIRP increases cell proliferation

Given the fundamental role of E2 in the growth of breast cancer cells and SLIRP’s role

as a NR corepressor of gene transactivation in the ER pathway, it was hypothesised that

when SLIRP levels are reduced in MCF-7 cells total cell proliferation would increase.This hypothesis was tested in normal dividing MCF-7 cells and found that cell

proliferation did not change (Figure 5.5 A, C) even though SLIRP protein was reducedfrom day 2 onwards. However, when the cells were stimulated with E2, a noticeable

and statistically significant (p < 0.05) increase in cell proliferation was observed from

day 2 (the first time point after the addition of E2) until day 4 (Figure 5.5 B). At day 5,the SLIRP and GFP siRNA treated cells had both reached the same plateau-level, as is

to be expected once the cells reach a density preventing further growth. This data is

consistent with SLIRP acting as a corepressor of E2 signaling. It seems when SLIRPexpression is decreased, there is a release on a brake controlling cell proliferation.

5.3.4 SLIRP affects HSP-60 protein levels in breast cancer cells

Treatment of MCF-7 cells with siRNA targeted against SLIRP resulted in a significantknock-down of SLIRP protein in the cell. Based on observations in the literature of

HSP-60 as a mitochondrial protein affected by E2, highly expressed in primary breastcancers and currently under assessment as a biomarker of breast cancer (Czarnecka et

al., 2006), I was particularly interested in assessing if reduction of SLIRP via siRNA

had any effect on HSP-60 levels in breast cancer cells. Using a SMARTPOOL siRNAagainst SLIRP, when SLIRP protein levels were reduced to approximately 30 % of


139

endogenous levels, HSP-60 protein level was reduced (Figure 5.5 D). This knock-downsuggests HSP-60 levels may be regulated by SLIRP as well as regulated by E2 as has

been previously discovered.

5.4 Discussion

As SLIRP was originally identified from a yeast III screen of a primary human breast

cancer cell library and is a potent ER corepressor, there is much interest in assessingSLIRP’s role in hormone-dependent breast cancer cells. The MCF-7, E2-dependent,

cell-line was utilized and it was found that SLIRP was an ER corepressor in the contextof breast cancer. These data were consistent with my previous results of SLIRP activity

in HeLa cells (see Chapter 4). However, SLIRP’s function in prostate cancer cell-line

was also assessed, and it has been shown that SLIRP also acts as a corepressor in theAR pathway. This data confirms SLIRP’s role as a NR corepressor that plays a

fundamental role in the regulation of a range of hormone action in breast and prostate

cancer. Future work in the laboratory of my Supervisor is examining in closer detailSLIRP’s role in prostate cancer.

In order to further validate the capacity of SLIRP, a complex study was performed in

which SLIRP was removed from- and then transfected back into- breast cancer cells.

Repression of SRA-mediated transactivation of the ER pathway was reduced whencells were treated with siRNA against SLIRP. Interestingly the repression could be re-

introduced when SLIRP was transfected back into the same cells. This data providesadditional confirmation of the repressive activity of SLIRP.

Given the fundamental role of E2 in the growth of breast cancer cells and SLIRP’s roleas a NR corepressor of gene transactivation in the ER pathway, it was hypothesised that

when SLIRP levels are reduced in MCF-7 cells, cell proliferation would increase.Interestingly, when these SLIRP knock-out experiments were performed under normal

growth conditions, no effect on cell proliferation was observed. However, if cells were

treated with E2 24 hours after siRNA treatment, cell proliferation increased. These dataare consistent with SLIRP acting as a potent E2-signaling pathway corepressor. The

change in proliferation was noted early, suggesting reasonably rapid changes resultingfrom SLIRP’s knock-down. Future work in the project will focus in more detail on


140

SLIRP’s effect on cell proliferation and cell cycle in a range of breast and prostatecancer cells, both in vitro and in vivo.

Because HSP-60, a mitochondrial protein, is potentially regulated by E2, highly

expressed in primary breast cancers and currently under assessment as a biomarker of

breast cancer, I undertook steps to assess if reduction of SLIRP via siRNA couldregulate HSP-60 levels in breast cancer cells. I was intrigued to find that a reduction of

SLIRP led to a reduction of HSP-60 levels in whole cell lysates. This suggests thatSLIRP may have important roles to regulate key mitochondrial genes, above and

beyond the E2-signaling pathway in breast cancer cells. Thus, HSP-60 and SLIRP may

each be important biomarkers for breast cancer and could even potentially representnovel targets for breast cancer therapeutics.

The potential importance of the SLIRP-HSP-60 relationship in terms of cell

proliferation is important to note here. Women with breast cancers with high HSP-60

expression have a poor prognosis. Unpublished data from our laboratory also suggestswomen with high levels of SLIRP have a poor prognosis (data not shown). Thus we

hypothesise that the SLIRP-HSP-60 relationship could be key for driving certainaspects of mitochondrial biology and energy homeostasis. However, it is clear that

there is a substantial difference between what we observe in vitro in cell culture and in

primary human breast cancer. My studies indicate that removal of SLIRP from cellscan enhance cell growth. However, data from the lab of my Supervisor indicates that

excess SLIRP in primary breast cancer carries a poor prognosis, and presumably

augments growth. This may reflect the basic difference between the two phases ofclinical disease being studied (primary [tissue] versus secondary [cell lines]) and

emphasizes the need to carefully investigate the role of SLIRP in each clinicalsituation. Future work will be necessary to assess the extent of the SLIRP-HSP-60

relationship in detail in a range of hormone dependent cancer paradigms.

It is clearly going to be very important to understand the importance of SLIRP’s role in

both the nucleus and mitochondria. In order to address these issues, in the near futurethe laboratory of my Supervisor will be conducting cDNA microarray analysis of

SLIRP pathways using under- and over-expression techniques. Proteins that interact

with SLIRP will also be identified, and a SLIRP knock-out mouse is being generated toassess the functional biology of SLIRP in vivo.


141

Conclusion

The studies in this Chapter have established SLIRP as a NR corepressor in hormone

dependent cancer cell systems and, more specifically, as an ER corepressor in the

context of breast cancer. Furthermore, the data demonstrating SLIRP’s regulation ofHSP-60 levels and cell proliferation suggest key roles for SLIRP in mitochondrial

biology in these cells.

Figure 5.1 A: The breadth of mitochondrial cytochrome P450 proteins and theirinteractions with steroid hormones. Panel A: This schematic is taken from Seliskar &Rozman (2007, p. 460). Mitochondrial cytochromes P450 are highlighted in orange:this figure highlights how they interact with and facilitate steroid hormonebiosynthesis.

A

142

B

Figure 5.2 A, B: HSP-60 expression in breast cancer. Panel A: Two figures takenfrom Isidoro et al. (2004, p. 18), where HSP-60 was found to be highly expressed inhuman breast cancers. (a) is a Western blot showing two patients, 9 (control) and 10(breast cancer patient) and their relative HSP-60 expression levels. (b) is the real-timePCR quantification of this increase. HSP-60 levels are increased in human breastcancers. Panel B: A Kaplan-myer survival graph taken from Isidoro et al. (2005, p.2100), where HSP-60 was also found to be overexpressed in a number of primaryhuman breast cancers. One hundred and one breast caners were observed andclassified into low, medium and high based on their expression of tumourigenicglycolitic markers (Isidoro et al., 2005). High level HSP-60 expression was correlatedwith an increase in aggressive tumour types, fewer disease free months and asignificantly decreased survival.

Aa b

9 10

Rela

tive

fold

incr

ease

143

ERE

E2

Luc

A

ERE

E2

Luc

0

5

10

15

20

25

ReporterE2

pSCTpSCT-SRA

FLAGFLAG-SLIRP

+-----

++----

+++-+-

+++--+

++-++-

++-+-+

Fold

tran

sact

ivat

ion

B

Figure 5.3 A, B: SLIRP is an ER corepressor in two breast cancer cell-lines. Panel A:a transient transfection in the E2-dependent MCF-7 cell-line. All the controls areincluded in this figure to illustrate the degree to which SLIRP repression is E2- andSRA-dependent in this cell-line. The last lane includes SLIRP and SRA, definingSLIRP as an ER corepressor. Panel B: is the same experiment performed in MDA-MB-468, an ER- breast cancer cell line, in which, when transfected with ERα, SLIRPis a E2- and SRA- dependent ER corepressor.

144

0

2

4

6

8

10

12

14

ReporterE2

pSCTpSCT-SRA

FLAGFLAG-SLIRP

+-----

++----

+++-+-

+++--+

++-++-

++-+-+

Fold

tran

sact

ivat

ion

ARE

DHT

Luc

0

5

10

15

20

25

ReporterDHT

pSCTpSCT-SRA

FLAGFLAG-SLIRP

+-----

++----

+++-+-

+++--+

++-++-

++-+-+

Fold

tran

sact

ivat

ion

C

Figure 5.3 C: SLIRP is an ER corepressor in two breast cancer cell-lines. Panel C:Transient transfections using an androgen-responsive luciferase reporter in theandrogen-dependent prostate cancer cell-line, 22RV1, showed SLIRP is acorepressor. This repression, however, is not entirely dependent on the addition ofexogenous SRA as seen in lane 4 above.

145

146

0

1

2

3

4

5

6

7

8

9

10

SLIRP siRNAGFP siRNApSCT-SRA

+-+

-++

ERE

E2

LucFo

ld tr

ansa

ctiv

atio

n

A

0

2

4

6

8

10

12

14

SLIRP siRNAGFP siRNApSCT-SRA

FLAGFLAG-SLIRP

+-++-

-+--+

-+++-

-++-+

Fold

tran

sact

ivat

ion

ERE

E2

Luc

B

Figure 5.4 A, B: Complementary SLIRP siRNA experiments. Panel A: Transienttransfection analysis in MCF-7 cells. When SLIRP is significantly reduced withsiRNA (confirmed by Western blot in the right hand panel), SRA transactivation ofthe ER pathway is increased when compared to GFP siRNA treated cells. Panel B: acomplex reversal experiment, where MCF-7 cells were treated with siRNA againsteither SLIRP or GFP for 3 days, then SLIRP was transfected back in to the cells, andthe cells stimulated with E2 for 24 hours. When SLIRP is reintroduced into the cells itinduces repression (compare lane 1 and 4). Western blot confirming knock-down ofSLIRP is shown on the right hand panel, each experiment was performed in triplicate,results shown are representative over all experiments.

β-Actin

SLIRP

β-Actin

SLIRP

0

0.5

1

1.5

2

2.5

3

1 2 3 4 5

Days

Rela

tive

Norm

aliz

ed G

row

th

GFP siRNASLIRP siRNA

0

0.5

1

1.5

2

2.5

3

1 2 3 4 5 6Days

Norm

aliz

ed R

elat

ive

Gro

wth

GFP siRNASLIRP siRNA

A

B

Figure 5.5 A, B: SLIRP affects cell proliferation. Panel A: Cell titre analysis of cellproliferation over 6 days shows that when SLIRP is targeted with siRNA (see panel Cfor western blot) under normal growth conditions no change in cell proliferation isnoted. Panel B: This is not the case when cells are stimulated with 10 nM E2 on day1, in this case, knock-down of SLIRP protein increases cell proliferation on day 2,over 5 days. The * indicates statistical significance (assessed over 2 separateexperiments) at 95 % confidence intervals, p < 0.05. Each experiment was repeated intriplicate, representative experiments shown above.

147

**

*

GFP siRNA SLIRP siRNA

SLIRP β-Actin

1

2

3

Days

GFP siRNA SLIRP siRNA

HSP-60

β-Actin

SLIRP

C

D

Figure 5.5 C, D: SLIRP knock-down over time affects HSP-60 protein levels. PanelA: Western blot of the cells analysed during the cell titre analysis of cell proliferationshown in panel A and B. Pane D: When SLIRP protein is significantly reduced, as inthe previous 3 panels, HSP-60 levels are also reduced. These experiments were allperformed in triplicate and data shown is a representative from one of theseexperiments.

148

E. Hatchell Chapter 6: Discussion and Conclusions

149

Chapter 6

Discussion & Conclusions


150

6. DISCUSSION

6.1 Introduction

In this thesis I have presented a detailed characterisation of SLIRP. SLIRP is a novel,small, SRA-binding, predominantly mitochondrial, NR corepressor. Intriguingly,

SLIRP appears likely to have a bifunctional role in both the nucleus and the

mitochondria. In Chapter 3, I described the results for the basic characterisation of

SLIRP, including all bioinformatic analysis, expression profiling and SRA-SLIRP

interaction studies. In Chapter 4, I detailed the discovery of three fascinating aspects of

SLIRP biology: that SLIRP is a NR corepressor, that SLIRP is predominantly

mitochondrial and that SLIRP interacts with other NR corepressors. Chapter 5

presented data that highlights the complex role for SLIRP in the ER pathway of E2-

dependent breast cancer cells. In this Chapter, I will present an integrated view tohighlight the importance of both SLIRP and SRA in ER biology.

6.2 The emerging role for RNA and mitochondria in breast cancer

Recent research into the basic biology of breast cancer has offered two novel and

exciting observations which offer potential new avenues for the future development of

cancer therapeutics: the association of ncRNA molecules in breast cancer

tumorigenesis (such as SRA, as well as some miRNAs), and also the significant cross-

talk between the mitochondria and nucleus in breast cancer cells. These two areas of

research come together in this thesis through the study of SLIRP and its interactions

with SRA in breast cancer.

6.3 The discovery of SRA represents a paradigm shift in the perception of NR

coregulators

Until very recently, it has been understood that the role of RNA in the cell is to act

strictly as a messenger, transporting information and allowing the transformation ofDNA to protein. The apparent role of RNA has, with the advancement our

understanding of molecular biology, become increasingly more complex.


151

The sequencing of the genome was expected to define the differences between non- andmore-complex species. However, unexpectedly, it is the case that humans and mice

share commonality in over 99 % of their coding genes (Mattick, 2001; Kile & Hilton,2005). Furthermore, there is an inverse relationship between the complexity of an

organism and the percentage of its genome that is comprised of “junk” DNA

(Szymanski et al., 2005). Junk DNA accounts for only 10-20 % of single unicellularorganism genomes with this number increasing to 98 % in complex eukaryotes

(Szymanski et al., 2005). So, while many struggle to identify the differences in speciesthrough the functionality of expressed proteins, it has been hypothesised by some

scientists that a second level of regulation of gene expression by ncRNA occurs in

more complex organisms (Mattick, 2001).

In 2002, the FANTOM (Functional Annotation Of Mouse) consortium published theirfinding of several mRNA transcripts that were not translated into protein and yet

seemed to have functional importance (Okazaki et al., 2002 Tomaru & Hayashizaki,

2006). By 2005, FANTOM published that they had found over 100,000 sequences thatwere classified as ncRNA (Carninci et al., 2005). This is a significant number which

suggested that the number of ncRNA is also higher than first expected in humans(Tomaru & Hayashizaki, 2006). Increasing evidence indicates that RNA plays a much

more complicated role in biology than originally suspected.

The potential complexity of RNA is exemplified by the discovery of SRA, a major

player in NR biology. SRA coregulates NR pathways and has been implicated in breast

tumorigenesis. However, the mechanisms by which SRA mediates its effects remain tobe elucidated. RNA-binding proteins have long been accepted as crucial to cell biology.

Given the explosion of ncRNA families and the implications of the roles they play, it isfair to presume there will be RNA-protein interactions which are also crucial to basic

cell biology. Such molecules can essentially bridge the gap between RNA and protein.

6.4 SLIRP: a novel SRA-binding nuclear receptor corepressor

In this project’s initial investigation into SRA biology, a yeast III screen was performed

using a major structure of SRA, STR7 (Hatchell, 2002; Hatchell et al., 2006). From this

screen SLIRP was identified as a protein that binds to SRA and as a potent NRcorepressor. Detailed analysis of SLIRP in several NR pathways has shown SLIRP is a


152

repressor of E2, glucocorticoid, androgen, T3 and VitD action. In addition, SLIRPrepresses orphan NR activity as shown by its effects on PPARδ–mediated

transactivation. Initial transfection studies have shown that endogenous SLIRP protein

levels are tightly regulated. SLIRP, which is composed almost entirely of an RRM, iswidely expressed in normal human tissues while elevated in skeletal muscle, heart,

liver and testis. Furthermore, SLIRP is also widely expressed at the mRNA and protein

level in multiple cancer cell lines and IHC studies confirm its presence in primaryhuman breast tumors.

The studies presented in this thesis show that SLIRP interacts specifically with SRA in

vivo, and SRA STR7 in vitro. IP-RT-PCR experiments have shown clearly that SLIRP

protein closely associates with full-length SRA RNA in several different cancer celllines. The SRA STR7 stem loop is the longest and one of the most stable stem loops

critical for SRA coactivation function (Lanz et al., 2002). It has been reconfirmed thatSLIRP binds SRA STR7, and under the most stringent conditions, via REMSA.

Furthermore, when comparing SLIRP binding to either a wild-type STR7 or a mutant

SDM7 stem loop probe in REMSA, there is a significant decrease in SLIRP binding tothe mutant STR7 probe suggesting a direct interaction between SLIRP and STR7.

Consistent with the REMSA results, the transient transfection SRA mutation data

presented in this thesis shows that STR7 is required for SLIRP to act as a corepressor,which further strengthens the case for a direct interaction between STR7 and SLIRP in

vivo.

This thesis has shown the RRM of SLIRP is essential for SRA-binding. By making

discrete single and double aa substitutions of SLIRP’s RRM domain and assessing forSLIRP-STR7 binding in REMSA, SLIRP binding to wild-type STR7 was abolished.

Furthermore, these mutations significantly reduced SLIRP-corepression activity intransfection. Such data indicates that the RRM of SLIRP is essential for RNA-

(specifically SRA-STR7) interaction. It is of interest to note that Shi et al., (2001) also

showed that the SRA-binding and coregulator activity of SHARP, a NR corepressor,requires its RRM.

SHARP has been shown to bind SRA (Shi et al., 2001), however the exact SRA-

binding sequence or stem loop was not identified. After identifying that the SLIRP

genetic sequence also contained an RRM that shared significant homology to each of


153

the three RRMs located in SHARPs terminus, I hypothesised that SHARP would alsobind SRA STR7. This thesis has demonstrated that SHARP (via its RRM domain) can

interact with STR7, raising the possibility that SHARP and SLIRP may compete forbinding to SRA in vivo. The additive repressive effect of SHARP and SLIRP in

transfection, both of which bind to SRA STR7, suggests these proteins could function

to significantly down-regulate ER signaling in breast cancer cells. Further analysis todetermine if the nature of the interaction between SLIRP and SHARP is competitive or

independent will further elucidate the regulation of SRA transactivation in the ERpathway.

Importantly, RNAi studies show that a reduction of endogenous SLIRP expression, viasiRNA, increases SRA’s coactivation ability. This data: (i) is consistent with SLIRP

acting as a corepressor of E2-signaling; (ii) validates the results from the transfections;and (iii) suggests that the SRA-SLIRP interaction is functionally relevant.

6.5 Complex interactions exist between SLIRP and other NR coregulators

The studies presented in this thesis provide new insight into the mechanism ofinteraction between several NR coregulators: SRA, SLIRP, SRC-1 and N-CoR. Via

ChIP analysis, it was found that SLIRP can be recruited to both E2- and Dex-

responsive promoters. This is consistent with the transfection data, further indicatingSLIRP is a biologically functional NR corepressor. To assess the requirement of SRA

for SLIRP recruitment to the DNA, cells were treated with SRA siRNA and SLIRP

recruitment to the DNA again assessed. Here, it was particularly interesting to note thatSLIRP recruitment to an endogenous Dex-responsive promoter was regulated by the

amount of SRA in the cell. This suggests the interaction between SRA and SLIRP maybe crucial in NR mediated signaling.

It was next important to assess if SLIRP expression had any effect on other ERcoregulators in vivo. SRC-1 is a potent NR coactivator with fundamental roles in many

pathways. Using ChIP, it appeared that SRC-1 and SLIRP were competing forassociation with SRA. Specifically, when SLIRP levels are reduced, SRC-1 association

with SRA increases. This data is entirely consistent with the opposing function of these

two coregulators: the association of SRC-1 with SRA results in coactivation whilstassociation of SLIRP with SRA results in corepression. In order to determine if there is


154

competition, and if it is biologically relevant, it will be necessary to further investigatethis potential competitive-interaction with further in vitro and in vivo assessments of

SRC-1 and SLIRP levels in primary tumors.

Further mechanistic insights into SLIRP interaction with NR coregulators was achieved

through assessment of additional ChIP experiments performed in cells with reducedSLIRP expression. Results from these experiments suggest SLIRP is essential for

mediating N-CoR's association with the promoter in the absence of E2. In SLIRPsiRNA treated cells, binding of N-CoR is abrogated and the ER is strongly recruited.

Such data further suggests that SLIRP plays a fundamental role in ER corepression and

that removal of SLIRP from the cell can alter the promoter state from one of repressionto activation.

During the initial bioinformatic analysis of SLIRP, it was observed SLIRP colocalizes

to the same genomic localisation as SKIP, a powerful ER coactivator (Leong et al,

2004) on human Chr 14q24.3, and that this genomic distrubution is conserved acrossseveral species. While there has been no LOH described in this region to suggest that

genomic instability can lead to a predisposition to breast cancer (Martin et al., 1999),this intriguing observation led us to hypothesise that SLIRP and SKIP may be co-

ordinately regulated, as is the case for Grb7 and HER2 (Daly, 1998). However, detailed

expression analysis of the two proteins does not support this hypothesis. Furthermore,transfection and siRNA studies suggested that they have opposing and possibly

competitive effects on estrogen signaling, rather than working together in a coregulator

complex. There has, however, been several recent publications which indicate a directinteraction between SKIP and N-CoR in VDR transactivation (Leong et al., 2004) and

between SKIP and SHARP in the Notch pathway (Oswald et al., 2002). Thus, ourstudies of SLIRP, together with SKIP, N-CoR and SHARP, suggest there may be a role

for SLIRP in modulating VDR signaling, and this is something that the laboratory of

my Supervisor is currently investigating.

6.6 SLIRP is a predominantly mitochondrial NR corepressor

There is now mounting evidence that NR can reside in the mitochondrion. The role of

NRs in the mitochondrion, affecting cell survival and energy homeostasis, has recentlycome under close scrutiny. Mitochondrial cytochrome P450 has been shown to be


155

involved with E2, DHT and Vitamin D3 biosynthesis (Seliskar and Rozman, 2006).Several putative NR hormone response elements have been found in the mitochondrial

genome (Psarra et al., 2006) and T3 has been shown to activate mitochondrial TR(Scheller and Sekeris, 2003). Important for this thesis, are studies in breast cancer

tissues that suggest mitochondrial ER plays a role in tumor cell survival (Pedram et al.,

2006). Taken together, these data provide: (i) a foundation for NRs such as the ER(Pedram et al., 2006), GR (Sionov et al., 2006) and TR (Morrish et al., 2006) to reside

in the mitochondria; (ii) a basis for coregulator action in the mitochondria; and (iii) arationale for SLIRP’s presence there.

It was at first surprising to note that SLIRP is predominantly mitochondrial. However,together with the growing body of literature mentioned above, the existence of a

mitochondrial localisation sequence and the increased expression of SLIRP in high-energy demand and mitochondria-rich tissues, such as skeletal muscle, heart and liver

is consistent with its predominantly mitochondrial location. Confocal microscopy

suggests that more than 90 % of SLIRP is located in the mitochondria, raising thepossibility that it may function both in the nucleus and mitochondria to regulate NR

activity. Whether this is via interactions with SRA, or other mitochondrial RNA targets,is currently unknown, however, the subcellular localisation of SRA has recently been

identified as predominantly cytoplasmic, with some nuclear speckled staining (Zhao et

al., 2007). Such data is consistent with a complex role for both SLIRP and SRA in thenucleus and the mitochondria, as suggested in this thesis.

SLIRP’s capacity to repress PPARδ-mediated-signaling further suggests a potential

role in regulating lipid homeostasis in energy-rich tissues. It was particularly interestingto note that SLIRP’s repression was strongly dependent on an intact mitochondrial

localisation sequence. Taken together with data presented in this thesis indicatingSLIRP expression is tightly regulated in the cell and SLIRP repression is also

dependent on its RRM, such data suggests a complex and precisely timed role for

SLIRP in both the nucleus and the mitochondria.

A literature search indicates SLIRP is the first NR coregulator found to reside in themitochondria. However, there are significant parallels between it and another recently

discovered NR coregulator RIP140. RIP140 is a large NR corepressor (Augereau et al.,

2006) expressed in tissues similar to SLIRP and is a negative regulator of insulin-


156

responsive hexose uptake and oxidative metabolism (Powelka et al., 2006). Significantdecrease of RIP140 in adipocyte cells significantly upregulates genes involved in

glycolysis, mitochondrial biogenesis and oxidative phosphorylation, and causes anincrease in mitochondrial oxygen comsumption (Powelka et al., 2006). Essentially,

RIP140 seems to control a program of gene expression in the mitochondria that has

profound impact on energy homeostasis. Furthermore, the RIP140 knock-out mouse islean, and resistant to diet induced obesity (Powelka et al., 2006). As SLIRP can effect

PPARδ-mediated signaling, it will be of great interest to perform a similar assessment

of SLIRP-regulated genes in primary muscle and fat cells, followed by a detailedanalysis of the SLIRP knock-out mouse which is currently being developed.

6.7 SLIRP in breast cancer

Immunohistochemistry analysis identified SLIRP in the mitochondria of human breastcancer cells and this thesis has shown that SLIRP plays an important corepressive role

in the NR transactivation of both breast and prostate cancer cell-lines. Transient

transfections have shown SLIRP repression in breast cancer cells is both SRA- and E2-dependent, highlighting its dominant role in regulation of the ER pathway. Supporting

this notion further, when SLIRP protein is reduced via siRNA, both SRA coactivation

and cell proliferation are significantly increased.

Mitochondrial HSP-60 protein levels are perturbed in many diseases and disordersincluding breast cancer (Isidoro et al., 2005). HSP-60 expression in breast cancer is

associated with a more severe and advanced tumor phenotype where, furthermore,

women with tumors that overexpress HSP-60 show significantly less survival thanwomen with low levels of HSP-60 (Isidoro et al., 2004). After SLIRP was identified as

a predominantly mitochondrial protein, I developed the hypothesis that altering SLIRPlevels in breast cancer cells will affect the expression of other mitochondrial breast

cancer proteins. Because of the strong link between HSP-60, the mitochondria and

breast cancer, this was an ideal candidate gene to study. This thesis has demonstratedthat when SLIRP protein is reduced in breast cancer cells, HSP-60 protein levels are

also reduced. This data supports the hypothesis in this thesis that SLIRP has a role bothin the nucleus and in the mitochondria of breast cancer cells and suggests SLIRP as a

key bifunctional regulator of breast cancer cell signaling. Further cDNA microarray


157

studies assessing the gene-targets of SLIRP will reveal further mitochondrial proteinswhose expression levels are affected by SLIRP.

Unpublished data from our laboratory, from both immunohistochemistry and Kaplan-

myer survival analysis of a large cohort of women, show that women who have a breast

cancer with low levels of SLIRP expression have higher survival rates. The literaturepresents similar assessments of HSP-60 (Isidoro et al., 2004; Isidoro et al., 2005).

Substantially decreased levels of SLIRP decreases HSP60 levels which, predicted here,will in turn affect the tight regulation of several apoptotic pathways, and thus cell

proliferation. Further studies are clearly necessary and are currently under

consideration in the laboratory to elucidate the nature of the relationship betweenSLIRP and HSP-60. If the relationship is a direct one, it may potentially offer a new set

of therapeutic targets for the treatment of breast cancer patients.

6.8 The SRA story unfolds

Over time, the complexity of SRA biology is becoming apparent. SRA has been shown

to act as both a coactivator and corepressor under different cellular contexts (Zhao et

al., 2007) and also to be activated by hormones and MAPK (Lanz et al., 1999; Deblois

& Giguere, 2003). We, and others (Xu & Koeing, 2004; Zhao et al., 2004; Zhao et al.,

2007), have found that SRA can coactivate both type I and type II NR, and that itsaction is not specific only for steroid receptors as was originally thought (Lanz et al.,

1999). We hypothesise that SRA RNA acts as a backbone RNA coregulator to which

other proteins including SLIRP and SHARP can bind. It is hypothesised here that therate of transactivation of the ER, as well as other NRs, will depend on the

combinatorial properties of the coregulator group that is recruited. That is, the rate oftranscription will heavily depend on the timing, order and manner in which

coregulators are recruited.

Future studies are necessary, and are currently underway, in order to determine the

basic molecular mechanism of this recruitment and will hopefully address and clarifythe following issues. Currently, we do not understand the stoichiometry of the SRA-

protein interactions, how each of these various proteins is recruited and how that

regulates transactivation of a NR pathway. The structure of SRA, both alone andtogether with one or more of its protein binding-partners (including SLIRP), has yet to


158

be determined. The relative affinities of each coregulator for SRA is currentlyundefined, as is the overall maximal protein-binding capacity of SRA at any one time.

For example, can SLIRP and SHARP bind to STR7 simutaneously? Can they bindother SRA stem loops? What are the other coregulator proteins they interact with in

breast cancer cells? Do other coregulators interact with SRA (such as RTA and

CoAA)? What other, undiscovered, proteins interact with SRA?

SRA is a complex molecule and the work detailed in this thesis presents a complexpicture of what interactions may take place at only one stem-loop structure of SRA: i.e.

this is one small piece of a large puzzle. It is likely that there are complicated

interactions taking place at several other SRA stem loops, possibly simultaneously.Elucidation of these events will allow us to develop an in-depth understanding of this

remarkably fascinating molecule.

Thus, the data generated in this thesis provide a foundation for multiple new lines of

investigation to elucidate SLIRP’s biological function, it’s role in biological pathwaysand binding partners in physiological and pathophsyoiological conditions. They

include: (i) development of a knock-out mouse model of SLIRP; (ii) performing cDNAand miRNA microarray analysis where SLIRP is over- and under-expressed at various

time points after the addition of E2 in breast cancer, and after the addition of PPAR

ligands in muscle, fat and macrophages; (iii) performing a yeast II hybrid screen withSLIRP as bait in various cancer and normal cell systems; (iv) studying SLIRP’s role in

other pathways (including the Notch pathway) in hormone-dependent cancers; and (v)

examining the potential role of SLIRP as a regulator of energy homeostasis within themitochondria. This thesis offers a platform of knowledge from which these experiments

can be performed.

6.9 Final conclusion

The discovery of SRA, the first RNA coactivator, led to a paradigm shift in our

understanding of NR coregulation and hormone action. With the identification ofSLIRP, a novel SRA-binding protein, this thesis provides the most detailed

characterization of a direct SRA-protein interaction to date. SLIRP is a small

mitochondrial NR coregulator that represses E2-mediated signaling in breast cancercells and is expressed in primary breast cancers. The discovery of SLIRP is particularly


159

exciting as it has opened up several new avenues of research leading to a betterunderstanding of both SRA function and the complexity of NR coregulator signaling.

Further, due to its small size, its role as a powerful repressor and its effect on othercoregulators involved in breast cancer, SLIRP is a molecule that has remarkable

potential as a therapeutic target in hormone-dependent cancers. Finally, the discovery

of SLIRP highlights the importance of our growing understanding of ncRNA-proteininteractions, offers substantial new links in our understanding of the cross-talk between

NR signaling in the nucleus and mitochondria, and furthers our understanding of themechanisms involved in estrogen action in breast cancer.


160

E. Hatchell References

161

References


162

References

Adeyinka, A., Mertens, F., Idvall, I., Bondesson, L., Ingvar, C., Mitelman, F., et al.(1999). Different patterns of chromosomal imbalances in metastasising andnon-metastasising primary breast carcinomas. Int. Journal of Cancer, 84, 370-375.

Al-Hajj, M., Wicha, M. S., Benito-Hernandez, A., Morrison, S. J., & Clarke, M. F.(2003). Prospective identification of tumorigenic breast cancer cells. Proc NatlAcad Sci USA, 100(7), 3983 - 3988.

Anzick, S., Kononen, J., Walker, R., Azorsa, D., Tanner, M., Guan, Z., et al. (1997).AIB1, a steroid receptor coactivator amplified in breast and ovarian cancer.Science, 277(965-8).

Aravind, L., & Koonin, E., V. (1998). Novel predicted RNA-binding domainsassociated with the translation machinery. Journal of Molecular Evolution, 48,291-302.

Arya, R., Mallik, M., & Lakhotia, S. C. (2007). Heat shock genes - integrating cellsurvival and death. J Biosci, 32(3), 595-610.

Augereau, P., Badia, E., Balaguer, P., Carascossa, S., Castet, A., Jalaguier, S., et al.(2006). Negative regulation of hormone signaling by RIP140. J SteroidBiochem Mol Biol, 102(1-5), 51-59.

Australian Institute of Health and Welfare (AIHW), & Australasian Association ofCancer Registries (AACR). (2004). Cancer In Australia 2001. AIHW cat. no.CAN 23. Canberra: AIHW (Cancer Series no. 28).

Australian Institute of Health and Welfare (AIHW), & National Breast Cancer Centre(NBCC). (2006). Breast Cancer in Australia, 2006. AIHW cat. no. CAN 29.Canberra: AIHW (Cancer Series no. 34).

Bagchi, M. K. (1998). Molecular mechanisms of nuclear receptor-mediatedtranscriptional activation and basal repression. In L. P. Freedman (Ed.),Molecular biology of steroid and nuclear hormone receptors. Berlin:Birkhauser.

Barazi, H. O., Zhou, L., Templeton, N. S., Krutzsch, H. C., & Roberts, D. D. (2002).Identification of heat shock protein 60 as a molecular mediator of alpha 3 beta 1integrin activation. Cancer Res, 62(5), 1541-1548.

Barlev, N., Liu , L., Chehab, N., Mansfield, K., Harris, K., Halazonetis, T., et al.(2001). Acetylation of p53 activates transcription through recruitment ofcoactivators/histone acetyltransferases. Molecular Cell, 8, 1243-1254.

Barry, J. (2001). Vitamin D receptor function: investigation of regular proteininteractions at a distinct C-terminal interface. Unpublished Phd, University ofNew South Wales & Garvan Institute of Medical Research, Sydney.

Barry, J. B., Leong, G. M., Church, B., Issa, L. L., Eisman, J. A., & Gardiner, E. M.(2003). Interactions of SKIP/NCoA-62, TFIIB, and retinoid x receptor withvitamin D receptor helix H10 residues. Journal of Biological Chemistry,278(10), 8224-8228.

Baudino, T. A., Kraichely, D. M., Jefcoat, S. C. J., Winchester, S. K., Partridge, N. C.,& MacDonald, P. N. (1998). Isolation and characterization of a novelcoactivator protein NCoA-62, involved in vitamin -mediated trasncription.Journal of Biological Chemistry, 26, 16434-16441.

Beato, M., Herrlich, P., & Schutz, G. (1995). Steroid hormone receptors: many actorsin search of a plot. Cell, 83, 851-857.


163

Beatson, G. T. (1896). On the treatment of inoperable cases of carcinoma of themamma: suggestions for a new method of treatment with illustrative cases. TheLancet, ii, 104-107.

Bjorkqvist, A. M., Wolf, M., Nordling, S., Tammilehto, L., Knuuttila, A., Kere, J., etal. (1999). Deletions at 14q in malignant mesothelioma detected bymicrosatellite marker analysis. British Journal of Cancer, 81(7), 1111 - 1115.

Bouvet, P., Jain, C., Belasco, J., & Amalric, A. F. (1997). RNA recognition by jointaction of two nucleolin RNA-binding domains: genetic analysis and structuralmodeling. EMBO J, 16, 5235-5246.

Bres, V., Gomes, N., Pickle, L., & Jones, K. A. (2005). A human splicing factor, SKIP,associates with P-TEFb and enhances transcription elongation by HIV-1 Tat.Genes Dev, 19(10), 1211-1226.

Brooks, S. C., Locke, E. R., & Soule, H. D. (1973). Estrogen receptors in a human cellline (MCF-7) from breast carcinoma. Journal of Biological Chemistry, 248,6251-6253.

Brueggemeier, R. W. (2001). Aromatase, aromatase inhibitors, and breast cancer. Am JTher, 8(5), 333-344.

Brueggemeier, R. W., Hackett, J. C., & Diaz-Cruz, E. S. (2005). Aromatase inhibitorsin the treatment of breast cancer. Endocr Rev, 26(3), 331-345.

Burd, C., G, & Dreyfuss, G. (1994). Conserved structures and diversity of functions ofRNA-binding proteins. Science, 265, 615-621.

Burke, L. J., & Baniahmad, A. (2000). Co-repressors 2000. FASEB J, 14, 1876-1888.Burris, T. P., & Krishnan, V. (2005). Estrogen: a mitochondrial energizer that keeps on

going. Mol Pharmacol, 68(4), 956-958.Bussemakers, M. J., van Bokhoven, A., Verhaegh, G. W., Smit, F. P., Karthaus, H. F.,

Schalken, J. A., et al. (1999). DD3: a new prostate-specific gene, highlyoverexpressed in prostate cancer. Cancer Res, 59(23), 5975-5979.

Byrne, J., Nourse, C., Basset, P., & Gunning, P. (1998). Identification of homo- andheteromeric interactions between members of the breast carcinoma-associatedD52 protein family using the yeast two-hybrid system. Oncogene, 16(7), 873-881.

CanStat. (2006). Cancer in Victoria 2004. Melbourne, Australia: The Cancer CouncilVictoria Epidemiology Centre.

Cappello, F., Czarnecka, A. M., Rocca, G. L., Stefano, A. D., Zummo, G., & Macario,A. J. (2007). Hsp60 and Hsp10 as Antitumor Molecular Agents. Cancer BiolTher, 6(4).

Caretti, G., Schiltz, R. L., Dilworth, F. J., Di Padova, M., Zhao, P., Ogryzko, V., et al.(2006). The RNA helicases p68/p72 and the noncoding RNA SRA arecoregulators of MyoD and skeletal muscle differentiation. Dev Cell, 11(4), 547-560.

Carew, J. S., & Huang, P. (2002). Mitochondrial defects in cancer. Mol Cancer, 1, 9.Carninci, P., Kasukawa, T., Katayama, S., Gough, J., Frith, M. C., Maeda, N., et al.

(2005). The transcriptional landscape of the mammalian genome. Science,309(5740), 1559-1563.

Carruba, G. (2006). Estrogens and mechanisms of prostate cancer progression. Ann N YAcad Sci, 1089, 201-217.

Casas, F., Daury, L., Grandemange, S., Busson, M., Seyer, P., Hatier, R., et al. (2003).Endocrine regulation of mitochondrial activity: involvement of truncatedRXRalpha and c-Erb Aalpha1 proteins. Faseb J, 17(3), 426-436.

Casas, F., Rochard, P., Rodier, A., Cassar-Malek, I., Marchal-Victorion, S., Wiesner,R. J., et al. (1999). A variant form of the nuclear triiodothyronine receptor c-


164

ErbAalpha1 plays a direct role in regulation of mitochondrial RNA synthesis.Mol Cell Biol, 19(12), 7913-7924.

Cavarretta, I. T. R., Mukopadhyay, R., Lonard, D. M., Cowsert, L. M., Bennett, C. F.,O'Malley, B. W., et al. (2002). Reduction of coactivator expression by antisenseoligodeoxynucleotides inhibits ERα transcriptional activity and MCF-7proliferations. Molecular Endocrinology, 16(2), 253-270.

Cech, T. R., & Bass, B. L. (1986). Biological catalysis by RNA. Annual ReviewBiochemistry, 55, 599-629.

Chang, A. Y., Chan, J. Y., Chou, J. L., Li, F. C., Dai, K. Y., & Chan, S. H. (2006). Heatshock protein 60 in rostral ventrolateral medulla reduces cardiovascular fatalityduring endotoxaemia in the rat. J Physiol, 574(Pt 2), 547-564.

Chappell, J., Leitner, J. W., Solomon, S., Golovchenko, I., Goalstone, M. L., &Draznin, B. (2001). Effect of insulin on cell cycle progression in MCF-7 breastcancer cells. Direct and potentiating influence. J Biol Chem, 276(41), 38023-38028.

Chawla, A., Repa, J. J., Evans, R. M., & Mangelsdorf, D. J. (2001). Nuclear receptorsand lipid physiology: opening the x-files. Science, 294, 1866-1870.

Chen, J., Eshete, M., Alworth, W., & Yager, J. (2004). Binding of MCF-7 cellmitochondrial proteins and recombinant human estrogen receptors alph and betato human mitochondrial DNA estrogen response elements. Journal of CellularBiochemistry, 93(2), 358-373.

Chen, J., & Yager, J. (2004). Estrogen's effecs on mitochondrial gene expression:mechanisms and potential contributions to estrogen carcinogenesis. Ann N YAcad Sci, 1028, 258-272.

Chen, J. D., & Evans, R. M. (1995). A transcriptional co-repressor that interacts withnuclear hormone receptors. Nature, 377(6548), 454-457.

Cheng, S., & Balk, S. P. (2003). Steroid hormone receptor signaling in cancer. CancerTreat Res, 115, 293-318.

Chenna, R., Sugawara, H., Tadashi, K., Lopez, R., Gibson, T. J., Higgins, D. G., et al.(2003). Multiple sequence alignment with the Clustal series of programs.Nucleic Acids Research, 31(13), 3497-3500.

Cheung, E., Schwabish, M. A., & Kraus, W. L. (2003). Chromatin exposes intrinsicdifferences in the transcriptional activities of estrogen receptors alpha and beta.Embo J, 22(3), 600-611.

Chien, P., Ito, M., Park, Y., Taqami, T., Gehm, B., & Jameson, J. (1999). A fusionprotein of the estrogen receptor (ER) and nuclear receptor corepressor (NCoR)strongly inhibits estrogen-dependent responses in breast cancer cells. MolEndocrinol, 13(12), 2122 - 2136.

Chooniedass-Kothari, S., Emberly, E., Hamadani, M. K., Troup, S., Wang, X.,Czosnek, A., et al. (2004). The steroid receptor RNA activator is the firstfunctional RNA encoding a protein. FEBS Letters, 566, 43 - 47.

Chooniedass-Kothari, S., Hamedani, M. K., Troup, S., Hube, F., & Leygue, E. (2006).The steroid receptor RNA activator protein is expressed in breast tumor tissues.Int J Cancer, 118(4), 1054-1059.

Christian, M., White, R., & Parker, M. G. (2006). Metabolic regulation by the nuclearreceptor corepressor RIP140. TRENDS in Endocrinology and Metabolism,17(6), 243 - 250.

Chung, S., Jiany, L., Cheng, S., & Furneaux, H. M. (1996). Purification and propertiesof HuD, a neuronal RNA-binding protein. Journal of Biological Chemistry,271(19), 11518 - 11524.


165

Clarke, R. B., Howell, A., Potten, C. S., & Anderson, E. (1997). Dissociation betweensteroid receptor expression and cell proliferation in the human breast. CancerResearch, 57, 4987-4991.

Cogswell, P. C., Kashatus, D. F., Keifer, J. A., Guttridge, D. C., Reuther, J. Y.,Bristow, C., et al. (2003). NF-κB and IκBα are found in the mitochondria.Evidence for regulation of mitochondrial gene expression by NF-κB. Journal ofBiological Chemistry, 278(5), 2963 - 2968.

Cohen, R. N. (2006). Nuclear receptor corepressors and PPARgamma. Nucl ReceptSignal, 4, e003.

Coleman, K. M., Lam, V., Jaber, B. M., Lanz, R. B., & Smith, C. L. (2004). SRAcoactivation of estrogen receptor-alpha is phosphorylation-independent, andenhances 4-hydroxytamoxifen agonist activity. Biochem Biophys Res Commun,323(1), 332-338.

Cordera, F., & Jordan, V. C. (2006). Steroid receptors and their role in the biology andcontrol of breast cancer growth. Semin Oncol, 33(6), 631-641.

Costa, F. F. (2005). Non-coding RNAs: new players in eukaryotic biology. Gene,357(2), 83-94.

Cox, A., Dunning, A. M., Garcia-Closas, M., Balasubramanian, S., Reed, M. W.,Pooley, K. A., et al. (2007). A common coding variant in CASP8 is associatedwith breast cancer risk. Nat Genet, 39(3), 352-358.

Crick, F. (1970). Central dogma of molecular biology. Nature, 227, 561 - 563.Cuezva, J. M., Krajewska, M., de Heredia, M. L., Krajewski, S., Santamaria, G., Kim,

H., et al. (2002). The bioenergetic signature of cancer: a marker of tumorprogression. Cancer Res, 62(22), 6674-6681.

Czarnecka, A. M., Campanella, C., Zummo, G., & Cappello, F. (2006). Mitochondrialchaperones in cancer: from molecular biology to clinical diagnostics. CancerBiol Ther, 5(7), 714-720.

Daly, R. J. (1998). The Grb7 family of signalling proteins. Cell Signaling, 10(9), 613-618.

Darimont, B. D., Wagner, R. L., Apriletti, J. W., Stallcup, M. R., Kushner, P. J.,Baxter, J. D., et al. (1998). Structure and specificity of nuclear receptor-coactivator interactions. Genes & Development, 12, 3343-3356.

De Rienzo, A., Jhanwar, S. C., & Testa, J. R. (2000). Loss of heterozygosity analysis of13q and 14q in human malignant mesothelioma. Genes, Chromosomes &Cancer, 28, 337 - 341.

Deblois, G., & Giguere, V. (2003). Ligand-independent coactivation of ERα AF-1 bysteroid receptor RNA activator (SRA) via MAPK activation. Journal of SteroidBiochem. Molec. Biol, 85, 123-131.

Delsite, R., Kachhap, S., Anbazhagan, R., Gabrielson, E., & Singh, K. K. (2002).Nuclear genes involved in mitochondria-to-nucleus communication in breastcancer cells. Mol Cancer, 1, 6.

Di Croce, L., Okret, S., Kertsten, S., Gustafsson, J.-A., Parker, M. G., Walter, W., et al.(1999). Steroid and nuclear receptors. EMBO J, 18, 6201-6210.

Di Felice, V., David, S., Cappello, F., Farina, F., & Zummo, G. (2005). Is chlamydialheat shock protein 60 a risk factor for oncogenesis? Cell Mol Life Sci, 62(1), 4-9.

Dixon, M. J., & Morrow, M. (1999). Breast disease: a problem based approach. HongKong: W B Saunders.

Don, A. S., & Hogg, P. J. (2004). Mitochondria as cancer drug targets. Trends MolMed, 10(8), 372-378.


166

Dorgan, J. F., Longcope, C., Stephenson, H. E., Jr., Falk, R. T., Miller, R., Franz, C., etal. (1996). Relation of prediagnostic serum estrogen and androgen levels tobreast cancer risk. Cancer Epidemiol Biomarkers Prev, 5(7), 533-539.

Dowhan, D., Hong, E., Auboeuf, D., Dennis, A., Wilson, M., Berget, S., et al. (2005).Steroid hormone receptor coactivation and alternative RNA splicing byU2AF65-related proteins CAPERα and CAPERβ. Molecular Cell, 17, 429 - 439.

Draper, D. E. (1999). Themes in RNA-protein recognition. Journal of MolecularBiology, 293, 255-270.

Drees, B. L. (1999). Progress and variations in two-hybrid and three-hybridtechnologies. Current Opinion in Chemical Biology, 3, 64-70.

Dressel, U., Allen, T. L., Pippal, J. B., Rohde, P. R., Lau, P., & Muscat, G. E. (2003).The peroxisome proliferator-activated receptor beta/delta agonist, GW501516,regulates the expression of genes involved in lipid catabolism and energyuncoupling in skeletal muscle cells. Mol Endocrinol, 17(12), 2477-2493.

Dreyfuss, G., Matunis, M., J, Pinol-Roma, S., & Burd, C., G. (1993). hnRNP proteinsand the biogenesis of mRNA. Annual Review Biochemistry, 62, 289-321.

Dreyfuss, G., Philipson, L., & Mattaj, I. W. (1988). Riobonucleoprotein particles andcellular processes. The Journal of Cell Biology, 106, 1419-1425.

Duning, K., Buck, F., Barnekow, A., & Kremerskothen, J. (2007). SYNCRIP, acomponent of dendritically localized mRNPs, binds to the translationalregulator BC200 RNA. J Neurochem, Epub adead of print(Nov 27).

Easton, D. F., Pooley, K. A., Dunning, A. M., Pharoah, P. D., Thompson, D., Ballinger,D. G., et al. (2007). Genome-wide association study identifies novel breastcancer susceptibility loci. Nature.

Eddy, S. R. (2002). Computational genomics of noncoding RNA genes. Cell, 109(2),137-140.

Eggert, M., Michel, J., Schneider, S., Bornfleth, H., Baniahmad, A., Fackelmayer, F.O., et al. (1997). The glucocorticoid receptor is associated with the RNA-binding nuclear matrix protein hnRNP U. J Biol Chem, 272(45), 28471-28478.

Elenitoba-Johnson, K. S. J. (2001). Complex regulation of telomerase activity.American Journal of Pathology, 159(2), 405-410.

Emberly, E., Huang, G.-J., Hamedani, M. K., Czosnek, A., Ali, D., Grolla, A., et al.(2003). Identification of new human coding steroid receptor RNA activatorisofoms. Biochemical and Biophysical Research Communications, 301, 509 -515.

Emmens, C. W., Cox, R. I., & Martin, L. (1962). Antiestrogens. Recent Progress inHormone Research, 18, 415-459.

Endoh, H., Maruyama, K., Masuhiro, Y., Kobayashi, Y., Goto, M., Tai, H., et al.(1999). Purification and identification of p68 RNA helicase acting as atranscriptional coactivator specific for the activation function 1 of humanestrogen receptor α. Molecular and Cellular Biology, 19(8), 5363-5372.

Eneman, J. D., Wood, M. E., & Muss, H. B. (2004). Selecting adjuvant endocrinetherapy for breast cancer. Oncology (Williston Park), 18(14), 1733-1744,discussion 1744-1735, 1748, 1751-1734.

Evans, R. M. (1988). The steroid and thyroid hormone receptor superfamily. Science,240(4854), 889-895.

Felty, Q. (2006). Estrogen-induced DNA synthesis in vascular endothelial cells ismediated by ROS signaling. BMC Cardiovasc Disord, 6, 16.

Feng, J., Bi, C., Clark, B. S., Mady, R., Shah, P., & Kohtz, J. D. (2006). The Evf-2noncoding RNA is transcribed from the Dlx-5/6 ultraconserved region andfunctions as a Dlx-2 transcriptional coactivator. Genes Dev, 20(11), 1470-1484.


167

Fisher, B. (1999). From Halsted to prevention and beyond: advances in themanagement of breast cancer during the twentieth century. European Journal ofCancer, 35(14), 1963-1973.

Fisher, B., & Ore, L. (1993). On the underutilization of breast-conserving surgery forthe treatment of breast cancer. Annals of Oncology, 4, 96-98.

Fleming, F. J., Hill, A. D., McDermott, E. W., O'Higgins, N. J., & Young, L. S. (2004).Differential recruitment of coregulator proteins steroid receptor coactivator-1and silencing mediator for retinoid and thyroid receptors to the estrogenreceptor-estrogen response element by beta-estradiol and 4-hydroxytamoxifenin human breast cancer. J Clin Endocrinol Metab, 89(1), 375-383.

Fondell, J., Ge, H., & Roeder, R. (1996). Ligand induction of a transcriptionally activethyroid hormone receptor coactivator complex. Proc Natl Acad Sci USA, 93,8329-8333.

Frasor, J., Danes, J. M., Funk, C. C., & Katzenellenbogen, B. S. (2005). Estrogendown-regulation of the corepressor N-CoR: mechanism and implications forestrogen derepression of N-CoR-regulated genes. PNAS, 102(37), 13153 -13157.

Freedman, L. P. (1999). Increasing the complexity of coactivation in nuclear receptorsignaling. Cell, 97, 5-8.

Fuqua, S. A. W., Chamness, G. C., & McGuire, W. L. (1993). Estrogen receptormutations in breast cancer. Journal of Cellular Biochemistry, 51, 135-139.

Gailen, R., & Garcia, T. (1997). Estrogen receptor impairs interleuking-6 expression bypreventing protein binding on the NF-κB site. Nucleic Acids Research, 25(12),2424-2429.

Gamberi, C., Johnstone, O., & Lasko, P. (2006). Drosophila RNA binding proteins. IntRev Cytol, 248, 43-139.

Gavrilova-Jordan, L. P., & Price, T. M. (2007). Actions of steroids in mitochondria.Semin Reprod Med, 25(3), 154-164.

Ghisolfi-Nieto, L., Joseph, G., Puvion-Dutilleul, F., Amalric, F., & Bouvet, P. (1996).Nucleolin is a sequence-specific RNA-binding protein: characterization oftargets on pre-ribosomal RNA. Journal of Molecular Biology, 260, 34-53.

Giles, K. M., Daly, J. M., Beveridge, D. J., Thomson, A. M., Voon, D. C., Furneaux, H.M., et al. (2003). The 3'untranslated region of p21WAF1 mRNA is a compositecis-acting sequence bound by RNA-binding proteins from breast cancer cells,including HuR and poly(C)-binding protein. Journal of Biological Chemistry,278(5), 2937-2946.

Gill, R. K., & Bell, N. H. (2000). Steroid receptor co-activator-1 mediates 1,25-dihydroxyvitamin D3-stimulated alkaline phosphatase in human osteosarcomacells. Calcified Tissue International, 66(5), 370-374.

Giorgi, C., & Moore, M. J. (2007). The nuclear nurture and cytoplasmic nature oflocalized mRNPs. Semin Cell Dev Biol, 18(2), 186-193.

Goldenthal, M. J., Ananthakrishnan, R., & Marin-Garcia, J. (2005). Nuclear-mitochondrial cross-talk in cardiomyocyte T3 signaling: a time-course analysis.J Mol Cell Cardiol, 39(2), 319-326.

Goldenthal, M. J., Weiss, H. R., & Marin-Garcia, J. (2004). Bioenergetic remodeling ofheart mitochondria by thyroid hormone. Mol Cell Biochem, 265(1-2), 97-106.

Goodrich, J. A., & Tjian, R. (1994). Transcription Factors IIE and IIH and ATPhyroysis promoter clearance by RNA polymerase II. Cell, 77, 145-156.

Gorski, J., Welshons, W., & Sakai, D. (1984). Remodeling the estrogen receptor model.Molecular and Cellular Endocrinology, 36, 11-15.


168

Gray, J. W. (2001). Translational breast cancer research in the post-genomic world.R e t r i e v e d A p r i l , 2 0 0 1 , f r o mhttp://www.nature.com/nm/special_focus/bc/commentaries/

Green, S., & Chambon, P. (1991). The oestrogen receptor: from perception tomechanism. In M. G. Parker (Ed.), Nuclear hormone receptors: molecularmechanisms, cellular functions, clinical abnormalities. London: AcademicPress.

Green, S., Walter, P., Kumar, V., Krust, A., Bornert, J.-M., Argos, P., et al. (1986).Human oestrogen receptor cDNA: sequence, expression and homology to v-erb-A. Nature, 320, 134-139.

Gupta, G. P., Minn, A. J., Kang, Y., Siegel, P. M., Serganova, I., Cordon-Cardo, C., etal. (2005). Identifying site-specific metastasis genes and functions. Cold SpringHarb Symp Quant Biol, 70, 149-158.

Gupta, S., & Knowlton, A. A. (2005). HSP60, Bax, apoptosis and the heart. J Cell MolMed, 9(1), 51-58.

Haagensen, C. D. (1986). Diseases of the breast (3rd ed.). Philadelphia: W B Saunders.Halachmi, S., Marden, E., Martin, G., MacKay, H., Abbondanza, C., & Brown, M.

(1994). Estrogen receptor-associated proteins: possible mediators of hormone-induced transcription. Science, 264(5164), 1455-1458.

Hall, J. M., McDonnell, D. P., & Korach, K. S. (2002). Allosteric regulation ofestrogen receptor structure, function, and coactivator recruitment by differentestrogen response elements. Molecular Endocrinology, 16(3), 469-486.

Hanahan, D., & Weinberg, R. A. (2000). The hallmarks of cancer. Cell, 100(1), 57-70.Haslam, S. Z., Counterman, L. J., & Nummy, K. A. (1993). Effects of epidermal

growth factor, estrogen, and prgestin on DNA synthesis in mammary glands invivo are determined by the state of the gland. Journal of Cell Physiology, 155,72-78.

Hatchell, E. C. (2002). Mechanisms Involved In Oestrogen Action:A Study of RNA-Protein Interactions of the Steroid Receptor RNA Activator: SRA. UnpublishedHonours, University of Western Australia, Perth, Australia.

Hatchell, E. C., Colley, S. M., Beveridge, D. J., Epis, M. R., Stuart, L. M., Giles, K. M.,et al. (2006). SLIRP, a small SRA binding protein, is a nuclear receptorcorepressor. Mol Cell, 22(5), 657-668.

Hayashi, S. I., Eguchi, H., Tanimoto, K., Yoshida, T., Omoto, Y., Inoue, A., et al.(2003). The expression and function of estrogen receptor alpha and beta inhuman breast cancer and its clinical application. Endocr Relat Cancer, 10(2),193-202.

Hebbes, T. R., Thorne, A. W., & Crane-Robinson, C. (1988). A direct link betweencore histone acetylation and transcriptionally active chromatin. Embo J, 7(5),1395-1402.

Herbert, B.-S., Wright, W. E., & Shay, J. W. (2001). Telomerase and breast cancer.Breast Cancer Research, 3, 146-149.

Herynk, M., & Fuqua, S. (2004). Estrogen receptor mutations in human disease.Endocrine Reviews, 25(6), 869-898.

Hind, D., Ward, S., De Nigris, E., Simpson, E., Carroll, C., & Wyld, L. (2007).Hormonal therapies for early breast cancer: systematic review and economicevaluation. Health Technol Assess, 11(26), iii-iv, ix-xi, 1-134.

Ho, S. M., Leung, Y. K., & Chung, I. (2006). Estrogens and antiestrogens as etiologicalfactors and therapeutics for prostate cancer. Ann N Y Acad Sci, 1089, 177-193.

Holst, F., Stahl, P. R., Ruiz, C., Hellwinkel, O., Jehan, Z., Wendland, M., et al. (2007).Estrogen receptor alpha (ESR1) gene amplification is frequent in breast cancer.Nat Genet, 39(5), 655-660.


169

Hong, H., Kohli, K., Trivedi, A., Johnson, D., & Stallcup, M. (1996). GRIP1, a novelmouse protein that serves as a transcriptional coactivator in yeast for thehormone binding domains of steroid receptors. Proc Natl Acad Sci USA, 93,4948-4952.

Hood, D. A., Adhihetty, P. J., Colavecchia, M., Gordon, J. W., Irrcher, I., Joseph, A.M., et al. (2003). Mitochondrial biogenesis and the role of the protein importpathway. Med Sci Sports Exerc, 35(1), 86-94.

Horlein, A. J., Naar, A. M., Heinzel, T., Torchia, J., Gloss, B., Kurokawa, R., et al.(1995). Ligand-independent repression by the thyroid hormone receptormediated by a nuclear receptor co-repressor. Nature, 377(6548), 397-404.

Hortobagyi, G. N., Hung, M.-C., & Buzdar, A. U. (1999). Recent developments inbreast cancer therapy. Seminars in oncology, 26(issue 4, Supplement 12), 11-20.

Horwitz, K. B., Jackson, T. A., Bain, D. L., Richer, J. K., Takimoto, G. S., & Tung, L.(1996). Nuclear receptor coactivators and corepressors. M o l e c u l a rEndocrinology, 10(10), 1167-1177.

Howell, A., Clarke, R. B., Evans, G., Bundred, N., Cuzick, J., Santen, R., et al. (2007).Estrogen deprivation for breast cancer prevention. Recent Results Cancer Res,174, 151-167.

Huang, M., Weissman, J. T., Beraud-Dufour, S., Luan, P., Wang, C., Chen, W., et al.(2001). Crystal structure of Sar1-GDP at 1.7 A resolution and the role of theNH2 terminus in ER export. J Cell Biol, 155(6), 937-948.

Hube, F., Guo, J., Chooniedass-Kothari, S., Cooper, C., Hamedani, M. K., Dibrov, A.A., et al. (2006). Alternative splicing of the first intron of the steroid receptorRNA activator (SRA) participates in the generation of coding and noncodingRNA isoforms in breast cancer cell lines. DNA Cell Biol, 25(7), 418-428.

Hussein-Fikret, S., & Fuller, P. J. (2005). Expression of nuclear receptor coregulatorsin ovarian stromal and epithelial tumours. Mol Cell Endocrinol, 229(1-2), 149-160.

Huttenhofer, A., & Vogel, J. (2006). Experimental approaches to identify non-codingRNAs. Nucleic Acids Res, 34(2), 635-646.

Ide, T. (2006). [Mechanism of cell proliferation--cell cycle, oncogenes, andsenescence]. Yakugaku Zasshi, 126(11), 1087-1115.

Ignar-Trowbridge, D. M., Nelson, K. G., Bidwell, M. C., Curtis, S. W., Washburn, T.F., McLachlan, J. A., et al. (1992). Coupling of dual signaling pathways:epidermal growth factor action involves the estrogen receptor. Proc Natl AcadSci USA, 89, 4658-4662.

Ioannou, I. M., Tsawdaroglou, N., & Sekeris, C. E. (1988). Presence of glucocorticoidresponsive elements in the mitochondrial genome. Anticancer Res, 8(6), 1405-1409.

Isidoro, A., Casado, E., Redondo, A., Acebo, P., Espinosa, E., Alonso, A. M., et al.(2005). Breast carcinomas fulfill the Warburg hypothesis and provide metabolicmarkers of cancer prognosis. Carcinogenesis, 26(12), 2095-2104.

Isidoro, A., Martinez, M., Fernandez, P. L., Ortega, A. D., Santamaria, G., Chamorro,M., et al. (2004). Alteration of the bioenergetic phenotype of mitochondria is ahallmark of breast, gastric, lung and oesophageal cancer. Biochem J, 378(Pt 1),17-20.

Iwasaki, T., Chin, W. W., & Ko, L. (2001). Identification and characterization of RRM-containing coactivator activator (CoAA) as TRBP-interacting protein, and itssplice cariant as a coactivator modulator (CoAM). Journal of BiologicalChemistry, 276(36), 33375 - 33383.

Jaeger, J., A, SantaLucia, J., Jr, & Tinoco, I. J. (1993). Determination of RNA structureand thermodynamics. Annual Review Biochemistry, 62, 255-287.


170

James, S. Y., Mackay, A. G., & Colston, K. W. (1996). Effects of 1,25dihydroxyvitamin D3 and its analogues on induction of apoptosis in breastcancer cells. Journal of Steroid Biochem. Molec. Biol, 58(4), 395-401.

Jensen, E. V., & Jacobson, H. I. (1962). Basic guides to the mechanism of estrogenaction. Recent Progress in Hormone Research, 18, 387-414.

Jenster, G. (1998). Coactivators and corepressors as mediators of nuclear receptorfunction: an update. Molecular and Cellular Endocrinology, 143, 1-7.

Jepsen, K., Hermanson, O., Onami, T. M., Gleiberman, A. S., Lunyak, V., McEvilly, R.J., et al. (2000). Combinatorial roles of the nuclear receptor corepressor intranscription and development. Cell, 102, 753-763.

Johnson, B. A., Wilson, E. M., Li, Y., Moller, D. E., Smith, R. G., & Zhou, G. (2000).Ligand-induced stabilization of PPARgamma monitored by NMR spectroscopy:implications for nuclear receptor activation. J Mol Biol, 298(2), 187-194.

Jordan, V. C., & Morrow, M. (1999). Tamoxifen, raloxifene, and the prevention ofbreast cancer. Endocrine Reviews, 20(3), 253-278.

Jorde, L. B., Carey, J. C., Bamshad, M. J., & White, R. L. (1995). Medical Genetics(2nd ed.). St. Louis, USA: Mosby, Inc.

Katzenellenbogen, B. S., & Katzenellenbogen, J. A. (2002). Defining the "S" inSERMs. Science, 295, 2380-2381.

Kawashima, H., Takano, H., Sugita, S., Takahara, Y., Sugimura, K., & Nakatani, T.(2003). A novel steroid receptor co-activator protein (SRAP) as an alternativeform of steroid receptor RNA-activator gene: expression in prostate cancer cellsand enhancement of androgen receptor activity. Biochem J, 369, 163 - 171.

Kile, B. T., & Hilton, D. J. (2005). The art and design of genetic screens: mouse. NatRev Genet, 6(7), 557-567.

Kim, M. Y., Hsiao, S. J., & Kraus, W. L. (2001). A role for coactivators and histoneacetylation in estrogen receptor alpha-mediated transcription initiation. Embo J,20(21), 6084-6094.

Kim, S. H., Lee, S. U., Kim, M. H., Kim, B. T., & Min, Y. K. (2005). Mitogenicestrogen metabolites alter the expression of 17beta-estradiol-regulated proteinsincluding heat shock proteins in human MCF-7 breast cancer cells. Mol Cells,20(3), 378-384.

Klinge, C. M. (2001). Estrogen receptor interaction with estrogen response elements.Nucleic Acids Research, 29(14), 2905-2919.

Klinge, C. M., Jernigan, S. C., Mattingly, K. A., Risinger, K. E., & Zhang, J. (2004).Estrogen response element-dependent regulation of transcriptional activation ofestrogen receptors alpha and beta by coactivators and corepressors. J MolEndocrinol, 33(2), 387-410.

Ko, Y. J., & Balk, S. P. (2004). Targeting steroid hormone receptor pathways in thetreatment of hormone dependent cancers. Curr Pharm Biotechnol, 5(5), 459-470.

Koolman, & Roehm. (2005). Color Atlas of Biochemistry (2nd ed.). New York: ThemeMedical Publishers.

Kornberg, R. D. (1999). Eukaryotic transcriptional control. TIBS, 24(12), M46-M49.Kuiper, G., Enmark, E., Pelto-Huikkp, M., Nilsson, S., & Gustafsson, J.-A. (1996).

Cloning of a novel estrogen receptor expressed in rat prostrate and ovary. ProcNatl Acad Sci USA, 93, 5925-5930.

Kumar, V., & Chambon, P. (1988). The estrogen receptor binds tightly to its responsiveelement as a lighand-induced homodimer. Cell, 55, 145-156.

Kurisu, T., Tanaka, T., Ishii, J., Matsumura, K., Sugimura, K., Nakatani, T., et al.(2006). Expression and function of human steroid receptor RNA activator in


171

prostate cancer cells: role of endogenous hSRA protein in androgen receptor-mediated transcription. Prostate Cancer Prostatic Dis, 9(2), 173-178.

Kushner, P., Agard, D., Greene, G., Scanlan, T., Shiau, A., Uht, R., et al. (2000).Estrogen receptor pathways to AP-1. J Steroid Biochem Mol Biol, 74(5), 311 -317.

Lagos-Quintana, M., Rauhut, R., Winfred, L., & Thomas, T. (2001). Identification ofnovel genes encoding for small expressed RNAs. Science, 294, 853-858.

Lanz, R. B., Chua, S. A., Barron, N., Soder, B. M., DeMayo, F., & O'Malley, B. W.(2003). Steroid receptor RNA activator stimulates proliferation as well asapoptosis in vivo. Molecular and Cellular Biology, 23(20), 7163-7176.

Lanz, R. B., McKenna, N. J., Onate, S. A., Albrecht, U., Wong, J., Tsai, S. Y., et al.(1999). A steroid receptor coactivator, SRA, functions as an RNA and is presentin an SRC-1 complex. Cell, 97, 17-27.

Lanz, R. B., Razani, B., Goldberg, A. D., & O'Malley, B. W. (2002). Distinct RNAmotifs are important for coactivation of steroid hormone receptors by steroidreceptor RNA activator (SRA). Proc Natl Acad Sci USA, 99(25), 16081 -16086.

Lau, N. C., Lim, L. P., Weinstein, E. G., & Bartel, D. P. (2001). An abundant class oftiny RNAs with probably regulatory roles in Caenorhabditis elegans. Science,294, 858-862.

Lazennec, G., Bresson, D., Lucas, A., Chauveau, C., & Vignon, F. (2001). ERβ inhibitsproliferation and invasion of breast cancer cells. Endocrinology, 142(9), 4120.

Lee, K. C., Li, J., Cole, P. A., Wong, J., & Kraus, W. L. (2003). Transcriptionalactivation by thyroid hormone receptor-beta involves chromatin remodeling,histone acetylation, and synergistic stimulation by p300 and steroid receptorcoactivators. Mol Endocrinol, 17(5), 908-922.

Lee, R. C., & Ambros, V. (2001). An extensive class of small RNAs in Caenorhabditiselegans. Science, 294, 862-864.

Lees, J. A., Fawell, S. E., White, R., & Parker, M. G. (1990). A 22-amino-acid peptiderestores DNA-binding activity to dimerization-defective mutants of the estrogenreceptor. Mol Cell Biol, 10(10), 5529-5531.

Leo, C., Li, H., & Chen, J. D. (2000). Differential mechanisms of nuclear receptorregulation by receptor-associated coactivator 3. J Biol Chem, 275(8), 5976-5982.

Leonardsson, G., Steel, J., Christian, M., Pocock, V., Milligan, S., Bell, J., et al. (2004).Nuclear receptor corepressor RIP140 reguates fat accumulation. PNAS, 101(22),8437 - 8442.

Leong, G. M., Subramaniam, N., Issa, L. L., Barry, J. B., Kino, T., Driggers, P. H., etal. (2004). Ski-interacting protein, a bifunctional nuclear receptor coregulatorthat interacts with N-CoR/SMRT and p300. Biochem Biophys Res Commun,315(4), 1070-1076.

Leung, Y. K., Lau, K. M., Mobley, J., Jiang, Z., & Ho, S. M. (2005). Overexpression ofcytochrome P450 1A1 and its novel spliced variant in ovarian cancer cells:alternative subcellular enzyme compartmentation may contribute tocarcinogenesis. Cancer Res, 65(9), 3726-3734.

Levenson, A. S., & Jordan, C. V. (1997). MCF-7: The first hormone-responsive breastcancer cell line. Perspectives in Cancer Research, 57, 3071-3078.

Levenson, A. S., & Jordan, V. C. (1999). Selective oestrogen receptor modulation:molecular pharmacology for the millennium. European Journal of Cancer,35(14), 1974-1985.


172

Leygue, E., Dotzlaw, H., Watson, P. H., & Murphy, L. C. (1999). Expression of thesteroid receptor RNA activator in human breast tumors. Cancer Research, 59,4190-4193.

Li, D. Q., Wang, L., Fei, F., Hou, Y. F., Luo, J. M., Zeng, R., et al. (2006).Identification of breast cancer metastasis-associated proteins in an isogenictumor metastasis model using two-dimensional gel electrophoresis and liquidchromatography-ion trap-mass spectrometry. Proteomics, 6(11), 3352-3368.

Lien, E. A., & Lonning, P. E. (2000). Selective oestrogen receptor modifiers (SERMs)and breast cancer therapy. Cancer Treat Rev, 26(3), 205-227.

Lippman, M. E., & Bolan, G. (1975). Oestrogen-responsive human breast cancer inlong term culture. Nature, 256, 592-593.

Liu, M., Lee, M.-H., Cohen, M., Bommakanti, M., & Freedman, L. P. (1996).Transcriptional activation of the cdk inhibitor p21 by vitamin D3 leads to theinduced differentiation of the myelomoncytic cell line U937. Genes &Development, 10, 142-153.

Liu, Z., Wong, J., Tsai, S. Y., Tsai, M.-J., & O'Malley, B. W. (1999). Steroid receptorcoactivator-1 (SRC-1) enhances ligand dependent and receptor-dependent cellfree transcription of chromatin. Proc Natl Acad Sci USA, 96, 9485-9490.

Lodish, H., Baltimore, D., Berk, A., Zipusky, S. L., Matsudaira, P., & Darnell, J.(1995). Molecular Cell Biology (3rd ed.). New York: Scientific AmericanBooks Inc.

Lonard, D. M., & O'Malley, B. W. (2006). The expanding cosmos of nuclear receptorcoactivators. Cell, 125, 411 - 414.

Looijenga, L. H., Verkerk, A. J., De Groot, N., Hochberg, A. A., & Oosterhuis, J. W.(1997). H19 in normal development and neoplasia. Mol Reprod Dev, 46(3),419-439.

Look, T. A. (1997). Oncogenic transcription factors in the human acute leukemias.Science, 278, 1059-1064.

Lutz, C. S., & Alwine, J. C. (1994). Direct interaction of the U1 snRNP - a protein withthe upstream efficiency element of the SC40 late polyadenlyation signal. Genes& Development, 8, 576-586.

MacDonald, P. N., Dowd, D. R., Zhang, C., & Gu, C. (2004). Emerging insights intothe coactivator role of NCoA62/SKIP in vitamin D-mediated transcription.Journal of Steroid Biochem. Molec. Biol, 89-90, 179-186.

Maldonado, E., Ha, I., Cortes, P., Weis, L., & Reinberg, D. (1990). Factors involved inspecific transcription by mammalian RNA polymerase II: role of transcriptionfactors IIA, IID, and IIB during formation of a transcription-competentcomplex. Molecular and Cellular Biology, 10(12), 6335-6347.

Maminta, M. L. D., Molteni, A., & Rosen, S. T. (1991). Stable expression of the humanestrogen receptor in HeLa cells by infection: effect of estrogen on cellproliferation and c-myc expression. Molecular and Cellular Endocrinology, 78,61-69.

Mandal, K., Foteinos, G., Jahangiri, M., & Xu, Q. (2005). Role of antiheat shockprotein 60 autoantibodies in atherosclerosis. Lupus, 14(9), 742-746.

Mangelsdorf, D. J., Thummel, C., Beato, M., Herrlich, P., Schutz, G., Umesono, K., etal. (1995). The nuclear receptor superfamily: the second decade. Cell, 83, 835-839.

Maris, C., Dominguez, C., & Allain, F. H. (2005). The RNA recognition motif, aplastic RNA-binding platform to regulate post-transcriptional gene expression.Febs J, 272(9), 2118-2131.


173

Martin, M. D., Fischbach, K., Osborne, C. K., Moshsin, S. K., Allred, D. C., &O'Connell, P. (2001). Loss of heterozygosity events impeding breast cancermetastasis contain the MTA1 gene. Cancer Research, 61, 3578-3580.

Matthews, D. H., Sabina, J., Zuker, M., & Turner, D. H. (1999). Expanded sequencedependence of thermodynamic parameters improves prediction of RNAsecondary structure. Journal of Molecular Biology, 288, 911-940.

Mattick, J. S. (2001). Non-coding RNAs: the architects of eukaryotic complexity.EMBO Rep, 2(11), 986-991.

Mattick, J. S., & Makunin, I. V. (2006). Non-coding RNA. Hum Mol Genet, 15 SpecNo 1, R17-29.

McCarthy, J. E. G., & Kollmus, H. (1995). Cytoplasmic mRNA-protein interactions ineukaryotic gene expression. TIBS, 20, 191-197.

McDonnell, D. P., & Norris, J. D. (2002). Connections and regulation of the humanestrogen receptor. Science, 296(5573), 1642-1644.

McKenna, N. J., Lanz, R. B., & O'Malley, B. W. (1999). Nuclear receptor coregulators:cellular and molecular biology. Endocrine Reviews, 20(3), 321-344.

McKenna, N. J., & O'Malley, B. W. (2002a). Combinational control of gene expressionby nuclear receptors and coregulators. Cell, 108, 465-474.

McKenna, N. J., & O'Malley, B. W. (2002b). Minireview: nuclear receptor coactivators- an update. Endocrinology, 143(7), 2461-2465.

McKenna, N. J., & O'Malley, B. W. (2005). Teaching resources. An interactive coursein nuclear receptor signaling: concepts and models. Sci STKE, 2005(299), tr22.

Mendelsohn, J., & Baselga, J. (2000). The EGF receptor family as targets for cancertherapy. Oncogene, 19, 6550-6565.

Merrill, B. M., Stone, K. L., Cobianchi, F., Wilson, S. H., & Williams, K., R. (1988).Phenylalanies that are conserved among several RNA-binding proteins formpart of a nucleic acid-binding pocket in the A1 heterogeneous nuclearribonucleoprotein. Journal of Biological Chemistry, 263(7), 3307-3313.

Meyer, M. E., H, G., Turcotte, B., Bocquel, M. T., Tasset, D., & Chambon, P. (1989).Steroid hormone receptors compete for factors that mediate their enhancerfunction. Cell, 57, 433-442.

Mishra, S. K., Abhijit, M., Vadlamudi, R. K., Li, F., Wang, R.-A., Yu, W., et al.(2001). MICoA, a novel metastatis-associated protein 1 (MTA1) interactingprotein coactivator, regulates estrogen receptor-α transactivation functions.Journal of Biological Chemistry, 278(21), 19209-19219.

Morrish, F., Buroker, N. E., Ge, M., Ning, X.-H., Lopez-Guisa, J., Hockenbery, D., etal. (2006). Thyroid hormone receptor isofoms localize to cardiac mitochondrialmatrix with potential for binding to receptor elements on mtDNA.Mitochondrion, 6, 143 - 148.

Mosselman, S., Polman, J., & Dijkema, R. (1996). ERβ : identification andcharacterization of a novel human estrogen receptor. FEBS Letters, 392, 49-53.

Muller-Tidow, C., Diederichs, S., Thomas, M., & Serve, H. (2004). Genome-widescreening for prognosis predicting genes in early-stage non-small-cell lungcancer. Lung Cancer, 45(Suppl 2), S145 - 150.

Murphy, L. C., Simon, S. L., Parkes, A., Leygue, E., Dotzlaw, H., Snell, L., et al.(2000). Altered expression of estrogen receptor coregulators during humanbreast tumorigenesis. Cancer Research, 60, 6266-6271.

Nagy, E., Henics, T., Eckert, M., Miseta, A., Lightowlers, R., N , & Kellermayer, M.(2000). Identification of the NAD+-binding fold of glyceraldehyde-3-phosphatedehydrogenase as as novel RNA-binding domain. Biochemical and BiophysicalResarch Communications, 275, 253-260.


174

Nawatz, Z., Lonard, D., Smith, C., Lev-Leham, E., Tsai, S., Tsai, M.-J., et al. (1999).The Angelman syndrome-associated protein, E6-AP, is a coactivator for thenuclear hormone receptor superfamily. Molecular and Cellular Biology,19(1182-89).

Nilsen, J., & Brinton, R. D. (2004). Mitochondria as therapeutic targets of estrogenaction in the central nervous system. Curr Drug Targets CNS Neurol Disord,3(4), 297-313.

Nilsson, S., Makela, S., Treuter, E., Tujague, M., Thomsen, J., Andersson, G., et al.(2001). Mechanisms of estrogen action. Physiological Reviews, 81(4), 1535-1565.

Nishihara, E., O'Malley, B. W., & Xu, J. (2004). Nuclear receptor coregulators are newplayers in nervous system development and function. Mol Neurobiol, 30(3),307-325.

Nobel Web. (2007). The nobel prize in physiology or medicine 2006. Retrieved 15May, 2007, from http://nobelprize.org/nobel_prizes/medicine/laureates/2006

Norris, J. D., Sherk, A., & McDonnell, D. P. (2002). A negative coregulator for thehuman ER. Molecular Endocrinology, 16(3), 459-468.

O'Connell, P., Fischbach, K., Hilsenbeck, S., Mohsin, S. K., Fuqua, S. A. W., Clark, G.M., et al. (1999). Loss of heterozygosity at D14S62 and metastatic potential ofbreast cancer. Journal of the National Cancer Institute, 91(16), 1391-1397.

O'Malley, B. W. (2006). Little molecules with big goals. Science, 313, 174 - 175.O'Malley, B. W., Tsai, S. Y., Bagchi, M. K., Weigel, N. L., Schrader, W. T., & Tsai,

M.-J. (1991). Molecular mechanism for the steroid hormone receptorsuperfamily. Science, 270, 1354-1357.

Ogbourne, S., & Antalis, T. M. (1998). Transcriptional control and the role of silencersin transcriptional regulation in eukaryotes. Biochem J, 331, 1-14.

Ogryzko, V. V., Schiltz, R. L., Russanova, V., Howard, B. H., & Nakatani, Y. (1996).The transcriptional coactivators p300 and CBP are histone acetyltransferases.Cell, 87, 953-959.

Okazaki, Y., Furuno, M., Kasukawa, T., Adachi, J., Bono, H., Kondo, S., et al. (2002).Analysis of the mouse transcriptome based on functional annotation of 60,770full-length cDNAs. Nature, 420(6915), 563-573.

Oliver, W. R., Jr., Shenk, J. L., Snaith, M. R., Russell, C. S., Plunket, K. D., Bodkin, N.L., et al. (2001). A selective peroxisome proliferator-activated receptor deltaagonist promotes reverse cholesterol transport. Proc Natl Acad Sci U S A, 98(9),5306-5311.

Onate, S. A., Tsai, S., Tsai, M.-J., & O'Malley, B. W. (1995). Sequence andcharacterization of a coactivator for the steroid hormone receptor superfamily.Science, 270(5240), 1354-1357.

Osborne, C. K., Schiff, R., Fuqua, S. A., & Shou, J. (2001). Estrogen receptor: currentunderstanding of its activation and modulation. Clin Cancer Res, 7(12 Suppl),4338s-4342s; discussion 4411s-4412s.

Oswald, F., Kostezka, U., Astrahantseff, K., Bourteele, S., Dillinger, K., Zechner, U., etal. (2002). SHARP is a novel component of the Notch/RBP-Jkappa signallingpathway. Embo J, 21(20), 5417-5426.

Pang, K. C., Frith, M. C., & Mattick, J. S. (2006). Rapid evolution of noncoding RNAs:lack of conservation does not mean lack of function. Trends Genet, 22(1), 1-5.

Parrella, P., Xiao, Y., Fliss, M., Sanchez-Cespedes, M., Mazzarelli, P., Rinaldi, M., etal. (2001). Detection of mitochondrial DNA mutations in primary breast cancerand fine-needle aspirates. Cancer Res, 61(20), 7623-7626.


175

Pedram, A., Razandi, M., Wallace, D. C., & Levin, E. R. (2006). Functional estrogenreceptors in the mitochondria of breast cancer cells. Mol Biol Cell, 17(5), 2125-2137.

Perez, D., Hoage, T., Pritchett, J., Ducharme-Smith, A., Halling, M., Ganapathiraju, S.,et al. (2007). Long, abundantly-expressed non-coding transcripts are altered incancer. Hum Mol Genet, Epub ahead of print(Nov 15).

Pettersson, K., & Gustafsson, J.-A. (2001). Role of estrogen receptor beta in estrogenaction. Annual Review of Physiology, 63, 165-192.

Powelka, A., Seth, A., Virbasius, J., Kiskinis, E., Nicoloro, S., Guilherme, A., et al.(2006). Suppression of oxidative metabolism and mitochondrial biogenesis bythe transcriptional corepressor RIP140 in mouse adipocytes. J Clin Invest,116(1), 135 - 136.

Prathapam, T., Kuhne, C., Hayman, M., & Banks, L. (2001). Ski interacts with theevolutionary conserved SNW domain of SKIP. Nucleic Acids Research, 29(17),3469-3476.

Psarra, A.-M. G., Solakidi, S., & Sekeris, C. (2006). The mitochondiron as a primarysite of action of steroid and thyroid hormones: Presence and action of steroidand thyroid hormone receptors in mitochondria of animal cells. Molecular andCellular Endocrinology, 246, 21 - 33.

Puigserver, P. (2005). Tissue-specific regulation of metabolic pathways through thetranscriptional coactivator PGC1-alpha. Int J Obes (Lond), 29(S1), S5 - S9.

Puigserver, P., & Spiegelman, B. (2003). Peroxisome proliferator-activated receptor-gamma coactivator 1 alpha (PGC-1alpha): transcriptional coactivator andmetabolic regulator. Endocrine Reviews, 24(1), 78 - 90.

Rachez, C., Suldan, Z., Ward, J., Change, C., Burakov, D., Erdjument-Bromage, H., etal. (1998). A novel protein complex that interacts with the vitamin D3 receptorin a ligand-dependent manner and enhances VDR transactivation in a cell-freesystem. Genes & Development, 12(1787-1800).

Rajendran, R. R., Nye, A. C., Frasor, J., Balsara, R. D., Martini, P. G. V., &Katzenellenbogen, B. S. (2003). Regulation of nuclear receptro transcriptionalactivity by a novel DEAD box RNA helicase (DP97). Journal of BiologicalChemistry, 278(7), 4628-4638.

Redfern, A. D., Ikeda, N., Epis, M., Beveridge, D. J., Stuart, L. M., Li, X., et al. (2007).PACT, PKR and TAR-binding protein: double stranded RNA-binding proteinsthat modulate SRA-dependent nuclear receptor signaling. Molecular Cell, InReview.

Robyr, D., & Wolffe, A. P. (1998). Hormone action and chromatin remodelling. CellMol Life Sci, 54, 113-124.

Robyr, D., Wolffe, A. P., & Wahli, W. (2000). Nuclear hormone receptor coregulatorsin action: diversity for shared tasks. Molecular Endocrinology, 14(3), 329-347.

Rogers, C. G., Parwani, A., Tekes, A., Schoenberg, M. P., & Epstein, J. I. (2005).Carcinosarcoma of the prostate with urothelial and squamous components. JUrol, 173(2), 439-440.

Russo, J., Ao, X., Grill, C., & Russo, I. H. (1999). Pattern of distribution of cellspositive for estrogen receptor alpha and progesterone receptor in relation toproliferating cells in the mammary gland. Breast Cancer Res Treat, 53(3), 217-227.

Russo, J., & Russo, I. H. (1999). Cellular basis of breast cancer susceptibility. OncolRes, 11(4), 169-178.

Russo, J., & Russo, I. H. (2006). The role of estrogen in the initiation of breast cancer.J Steroid Biochem Mol Biol, 102(1-5), 89-96.

Ruvkun, G. (2001). Glimpses of a tiny RNA world. Science, 294, 797-799.


176

Sadovsky, Y., Webb, P., Lopez, G., Baxter, J. D., Fitzpatrick, P. M., Gizang-Ginsberg,E., et al. (1995). Transcriptional activators differ in their responses tooverexpression of TATA-box-binding protein. Molecular and Cellular Biology,15(3), 1554-1563.

SantaLucia, J., Jr. (1998). A unified view of polymer, dumbell, and oligonulceotideDNA nearest neighbor thermodynamics. Proc Natl Acad Sci USA, 95, 1460-1465.

Santen, R. J., & Harvey, H. A. (1999). Use of aromatase inhibitor in breast carcinoma.Endocrine-Related Cancer, 6, 75-92.

Santen, R. J., Manni, A., Harvey, H. A., & Redmond, C. (1990). Endocrine treatmentof breast cancer in women. Endocrinology Reviews, 11, 221-265.

Sarkar, D. K., Liehr, J. G., & Singletary, K. W. (2001). Role of estrogen in alcoholpromotion of breast cancer and prolactinomas. Alcoholism: Clinical andExperimental Research, 25(5, Supplement), 230S-236S.

Sauer, F., Hanse, S. K., & Tjian, R. (1995). Multiple TAFIIs directing synergisticactivation of transcription. Science, 270, 1783-1787.

Savkur, R. S., Bramlett, K. S., Clawson, D., & Burris, T. P. (2004). Pharmacology ofnuclear receptor-coregulator recognition. Vitam Horm, 68, 145-183.

Scheller, K., & Sekeris, C. (2003). The effects of steroid hormones on the transcriptionof genes encoding enzymes of oxidative phosphorylation. ExperimentalPhysiology, 88(1), 129 - 140.

Scheller, K., Sekeris, C., Krhone, G., Hock, R., Hansen, I., & Scheer, U. (2000).Localization of gluticorticoid hormone receptors in mitochondria of humancells. European Journal of Cell Biology, 79(5), 299 - 307.

Schippinger, W., Regitnig, P., Dandachi, N., Wernecke, K. D., Bauernhofer, T.,Samonigg, H., et al. (2006). Evaluation of the prognostic significance ofandrogen receptor expression in metastatic breast cancer. Virchows Arch,449(1), 24-30.

Segersten, U., Holm, P. K., Bjorklund, P., Hessman, O., Nordgren, H., Binderup, L., etal. (2005). 25-Hydroxyvitamin D3 1alpha-hydroxylase expression in breastcancer and use of non-1alpha-hydroxylated vitamin D analogue. Breast CancerRes, 7(6), R980-986.

Sekeris, C. E. (1990). The mitochondrial genome: a possible primary site of action ofsteroid hormones. In Vivo, 4(5), 317-320.

Seliskar, M., & Rozman, D. (2007). Mammalian cytochromes P450--importance oftissue specificity. Biochim Biophys Acta, 1770(3), 458-466.

SenGupta, D. J., Zhang, B., Kraemer, B., Pochart, P., Fields, S., & Wickens, M. (1996).A three-hybrid system to detect RNA-protein interactions in vivo. Proc NatlAcad Sci USA, 93, 8496-8501.

Seol, W., Mahon, M., Lee, Y., & DD, M. (1996). Two receptor interaction domains inthe nuclear hormone receptor corepressor RIP/N-CoR. MolecularEndocrinology, 10, 1646-1655.

Shamoo, Y., Abdul-Manan, N., & Williams, K., R. (1995). Multiple RNA bindingdomains (RBDs) just don't add up. Nucleic Acids Research, 23(5), 725-728.

Shamovsky, I., & Nudler, E. (2006). Gene control by large noncoding RNAs. ScienceSTKE, 355, pe 40.

Shang, Y., & Brown, M. (2002). Molecular determinants for the tissue specificity ofSERMs. Science, 295, 2465-2468.

Shi, Y., Downes, M., Xie, W., Kao, H.-Y., Ordentlich, P., Tsai, C.-C., et al. (2001).Sharp, an inducible cofactor that integrates nuclear receptor repression andactivation. Genes & Development, 15, 1140-1151.


177

Shibata, H., Spencer, T. E., Onate, S. A., Jenster, G., Tsai, S. Y., Tsai, M.-J., et al.(1997). Role of co-activators and co-repressors in the mechanism ofsteroid/thyroid receptor action. Recent Progress in Hormone Research, 52, 141-165.

Simons, S. S. J. (1998). Structure and function of the steroid and nuclear receptorligand binding domain. In L. P. Freedman (Ed.), Molecular biology of steroidand nuclear hormone receptors (pp. 35-104). Berlin: Birkhauser.

Siomi, H., & Dreyfuss, G. (1997). RNA-binding proteins as regulators of geneexpression. Current Opinion in Genetics and Development, 7, 345-353.

Sionov, R., Cohen, O., Kfir, S., Zilberman, Y., & Yefenof, E. (2006). Role ofmitochondrial glucocorticoid receptor in glucocorticoid-induced apoptosis. JExp Med, 203(1), 189 - 201.

Solakidi, S., Psarra, A. M., & Sekeris, C. E. (2005). Differential subcellular distributionof estrogen receptor isoforms: localization of ERalpha in the nucleoli andERbeta in the mitochondria of human osteosarcoma SaOS-2 andhepatocarcinoma HepG2 cell lines. Biochim Biophys Acta, 1745(3), 382-392.

Spencer, T. E., Jenster, G., Burcin, M. M., Allis, C. D., Zhou, J., Mizzen, C. A., et al.(1997). Steroid recector coactivator-1 is a histone acetyltransferase.

Srikantan, V., Zou, Z., Petrovics, G., Xu, L., Augustus, M., Davis, L., et al. (2000).PCGEM1, a prostate-specific gene, is overexpressed in prostate cancer. ProcNatl Acad Sci U S A, 97(22), 12216-12221.

Staton, J. M., Thomson, A. M., & Leedman, P. J. (2000). Hormonal regulation ofmRNA stability and RNA-protein interactions in the pituitary. Journal ofMolecular Endocrinology, 25, 17-34.

Stockler, M., Wilken, N. R. C., Ghersi, D., & Simes, R. J. (2000). Systematic reviewsof chemotherapy and endocrine therapy in metastatic breast cancer. CancerTreatment Reviews, 26, 151-168.

Stormo, G. D., & Ji, Y. (2001). Do mRNAs acti as direct sensors of small molecules tocontrol their expression? Proc Natl Acad Sci USA, 98(17), 9465-9470.

Stothard, P. (2000). The sequence manipulation suite: JavaScript programs foranalyzing and formatting protein and DNA sequences. BioTechniques, 28,1102-1104.

Struhl, K., & Moqtaderi, Z. (1998). The TAFs in the HAT. Cell, 94, 1-4.Stryer, L. (1995). Biochemistry (2nd ed.). New York: W. H. Freeman and Co.Suen, C., Berrodin, T., Mastroeni, R., Cheskis, B., Lyttle, C., & Frail, D. (1988). A

transcriptional coactivator, steroid receptor coactivator-3, selectively augmentssteroid receptor transcriptional activity. Journal of Biological Chemistry, 273,27645-27653.

Szymanski, M., Barciszewska, M. Z., Erdmann, V. A., & Barciszewski, J. (2005). Anew frontier for molecular medicine: noncoding RNAs. Biochim Biophys Acta,1756(1), 65-75.

Szymanski, M., Erdmann, V. A., & Barciszewski, J. (2003). Noncoding regulatoryRNAs database. Nucleic Acids Res, 31(1), 429-431.

Tan, D. J., Bai, R. K., & Wong, L. J. (2002). Comprehensive scanning of somaticmitochondrial DNA mutations in breast cancer. Cancer Res, 62(4), 972-976.

Tao, Y., Williams-Skipp, C., & Scheinman, R. I. (2001). Mapping of gluticorticoidreceptor DNA binding domain surfaces contributing to transrepression of NF-κB and induction of apoptosis. Journal of Biological Chemistry, 276(4), 2329 -2332.

Thomson, A. M., Rogers, J. T., Walker, C. E., & Leedman, P. J. (1999). OptimizedRNA gelshift and UV cross linking assays for characterization of cytoplasmicRNA-protein interactions. BioTechniques, 27, 1032-1042.


178

Tiraby, C., & D, L. (2005). PGC-1alpha, a transcriptional coactivator involved inmetabolism. Med Sci (Paris), 21(1), 49 - 54.

Tobin, J. F., & Freedman, L. P. (2006). Nuclear receptors as drug targets in metabolicdiseases: new approaches to therapy. TRENDS in Endocrinology andMetabolism, 117(7), 284 - 290.

Toh, Y., Pencil, S., & Nicolson, G. (1994). A novel candidate metastasis-associatedgene, mta1, differentially expressed in highly metastatic mammaryadenocarcinoma cell lines. cDNA cloning, expression, and protein analyses.Journal of Biological Chemistry, 269(37), 22958-22963.

Tomaru, Y., & Hayashizaki, Y. (2006). Cancer research with non-coding RNA. CancerSci, 97(12), 1285-1290.

Tora, L., White, J., Brou, C., Tasset, D., Webster, N., Scheer, E., et al. (1989). Thehuman estrogen receptor has two independent nonacidic transcriptionalactivation functions. Cell, 59, 477-487.

Vadlamudi, R. K., Wang, R.-A., Mazumdar, A., Kim, Y.-s., Shin, J., Sahin, A., et al.(2001). Molecular cloning and characterization of PELP1, a novel humancoregulator of estrogen receptor α. Journal of Biological Chemistry, 276(41),38272-38279.

Varani, G., & Nagai, K. (1998). RNA recognition by RNP proteins during RNAprocessing. Annu. Rev. Biophys. Biomol. Struc., 27, 407-445.

Verkerk, A. J., Pieretti, M., Sutcliffe, J. S., Fu, Y. H., Kuhl, D. P., Pizzuti, A., et al.(1991). Identification of a gene (FMR-1) containing a CGG repeat coincidentwith a break point cluster region exhibiting length variation in fragile xsyndrome. Cell, 65, 905-914.

Vlastos, G., & Verkooijen, H. M. (2007). Minimally invasive approaches for diagnosisand treatment of early-stage breast cancer. Oncologist, 12(1), 1-10.

Waldman, F. M., DeVries, S., Chew, K. L., Moore, D. H. I., Kerlikowske, K., & Ljung,B.-M. (2000). Chromosomal alterations in ductal carcinomas in situ and their insitu recurrences. Journal of the National Cancer Institute, 92(4), 313-320.

Walter, P., Green, S., Greene, G., Krust, A., Bornert, J.-M., Jeltsch, J.-M., et al. (1985).Cloning of the human estrogen receptor cDNA. Proc Natl Acad Sci USA, 82,7889-7893.

Wang, C.-C., & Schimmel, P. (1999). Species barrier to RNA recognition overcomewith nonspecific RNA binding domains. The journal of Biological Chemistry,274, 16598-16512.

Wang, X., & Hall, T. M. T. (2001). Structural basis for recognition of AU-rich elementRNA by the HuD protein. Nature Structural Biology, 8(2), 141 - 145.

Wantanabe, M., Yanagisawa, J., Kitagawa, H., Takeyama, K.-i., Ogawa, S., Arao, Y.,et al. (2001). A subfamily of RNA-binding DEAD-box proteins acts as anestrogen receptor α coactivator through the N-terminal activation domain (AF-1) with an RNA coactivator, SRA. The EMBO Journal, 20(6), 1342-1352.

Ward, B. K., Mark, P. J., Ingram, D. M., Minchin, R. F., & Ratajczak, T. (1999).Expression of the estrogen receptor-associated immunophilins, cyclophilin 40and FKBP52, in breast cancer. Breast Cancer Research & Treatment, 58, 267-280.

Warnmark, A., Treuter, E., Wright, A. P., & Gustafsson, J. A. (2003). Activationfunctions 1 and 2 of nuclear receptors: molecular strategies for transcriptionalactivation. Mol Endocrinol, 17(10), 1901-1909.

Webster, D. J. T. (1986). A colour atlas of mastectomy with immediate reconstruction.United Kingdom: Wolfe Medical Publications.

Weintraub, H. (1985). Tissue-specific gene expression and chromatin structure. TheHarvey Lectures, Series 79, 217-245.


179

Westin, S., Rosenfeld, M. G., & Glass, C. K. (2000). Nuclear receptor coactivators.Advances in Pharmacology, 47, 89-112.

White, R., Leonardsson, G., Rosewell, I., Jacobs, M. A., Milligan, S., & Parker, M.(2000). The nuclear receptor co-repressor Nrip1 (RIP140) is essential for femalefertility. Nature Medicine, 6(12), 1368 - 1374.

Wong, C.-W., Komm, B., & Cheskis, B. J. (2001). Structure-function evaluation ofERα and β interplay with SRC family coactivators ER selective ligands.Biochemistry, 40, 6756-6765.

Xu, B., & Koenig, R. J. (2004). An RNA binding domain in the thyroid hormonereceptor enhances transcriptional activation. Journal of Biological Chemistry,279(2), 33051 - 33056.

Yanagisawa, J., Yasuo, Y., Masuhiro, Y., Suzawa, M., Wantanabe, M., Kashiwagi, K.,et al. (1999). Convergence of transforming growth factor-β and vitamin Dsignaling pathways on SMAD transcriptional coactivators. Science, 283, 1317-1321.

Yang, S.-H., Liu, R., Perez, E. J., Wen, Y., Stevens, S. M. J., Valencia, T., et al. (2004).Mitochondrial localization of estrogen receptor β. Proc Natl Acad Sci USA,101(12), 4130-4135.

Yang, T.-H., Tsai, W.-H., Lee, Y.-M., Lei, H.-Y., Lai, M.-Y., Chen, D.-S., et al.(1994). Purification and characterization of nucleolin and its identification as atranscriptional repressor. Molecular and Cellular Biology, 14(9), 6068-6074.

Yeap, B. (2000). Posttranscriptional regulation of androgen receptor gene expressionin human prostate and breast cancer cells: identification and characterisationof novel RNA-binding interactions. Unpublished Phd, University of WesternAustralia, Perth.

Yeap, B. B., Wilce, J. A., & Leedman, P. J. (2004). The androgen receptor mRNA.Bioessays, 26(6), 672-682.

Yi, P., Driscoll, M. D., Huang, J., Bhagat, S., Hilf, R., Bambara, R. A., et al. (2002).The effects of estrogen-response element- and ligand-induced structuralchanges on the recruitment of cofactors and transcriptional responses by ERαand ERβ. Molecular Endocrinology, 16(4), 674-693.

Zhang, C., Dowd, D. R., Staal, A., Gu, C., Lian, J. B., van Wijnen, A. J., et al. (2003).Nuclear coactivator-62 kDa/Ski-interacting protein is a nuclear matrix-associated coactivator that may couple vitamin D receptor-mediatedtranscription and RNA splicing. J Biol Chem, 278(37), 35325-35336.

Zhao, X., Patton, J. R., Davis, S. L., Florence, B., Ames, S. J., & Spanjaard, R. A.(2004). Regulation of nuclear receptor activity by a pseudouridine synthasethrough posttranscriptional modification of steroid receptor RNA activator. MolCell, 15(4), 549-558.

Zhao, X., Patton, J. R., Ghosh, S. K., Fischel-Ghodsian, N., Shen, L., & Spanjaard, R.A. (2007). Pus3p- and Pus1p-dependent pseudouridylation of steroid receptorRNA activator controls a functional switch that regulates nuclear receptorsignaling. Mol Endocrinol, 21(3), 686-699.

Zhou, S., Fujimuro, M., Hsieh, J. J., Chen, L., Miyamoto, A., Weinmaster, G., et al.(2000). SKIP, a CBF1-associated protein, interacts with the ankyrin repeatdomain of NotchIC To facilitate NotchIC function. Mol Cell Biol, 20(7), 2400-2410.

Zhou, S., & Hayward, S. D. (2001). Nuclear localization of CBF1 is regulated byinteractions with the SMRT corepressor complex. Mol Cell Biol, 21(18), 6222-6232.

Zucker, M. (2003). Mfold web server for nucleic acid folding and hybridizationprediction. Nucleic Acids Research, 31(13), 3406 - 3415.


180

E. Hatchell Appendix I: Buffers & Solutions

181

Appendix I

Buffers and Solutions


182

Appendix I: Recipes for Buffers and Solutions

1. Agarose gel (0.8 %)Agarose 0.8 gTAE x 1 100 mLEthidium bromide 5 µLThe agarose was mixed with 1 x TAE and heated in a microwave until dissolved. Theethidium bromide was added to a final concentration of 0.5 µg/mL.

2. Agarose gel (1.2%)Agarose 1.2 gTAE x 1 100 mLEthidium bromide (10 mg/mL) 5 µLThe agarose was mixed with 1 x TAE and heated in a microwave until dissolved. Theethidium bromide was added to a final concentration of 0.5 µg/mL.

3. Ammonium acetate (1M)Ammonium acetate 3.845 gDdH20 up to 10 mLAmmonium acetate was dissolved, the solution was stored at room temperature.

4. Ammonium persulphate, APS (10 %)Ammonium persulphate 1 gddH2o up to 10 mLAmmonium Persulphate crystals were dissolved, and the solution stored at 4ºC for upto 6 months.

5. Ampicillin (100 mg/mL) Ampicillin powder 1gddH20 up to 10 mLAmpicillin was dissolved, and the solution aliquoted and stored at -20ºC for up to 3months.

6. Aprotinin (1 mg/mL)Aprotinin powder 10 gddH20 up to 10 mLAprotinin was dissolved, and the solution aliquoted and stored at -20ºC for up to 3months.

7. Benzamadine (0.1 M)Benzamadine powder 313.2 mgBaxter water up to 20 mLSolution was dissolved, aliquoted and stored at –20ºC for up to 3 months.

8. Bovine Serum Albumin, BSA (2 mg/mL)BSA solution supplied by Pierre Ltd, Illinois, USA.

9. Buffer CSupplied with Bam HI (10 mM Tris-HCl, 10 mM MgCl2, 50 mM NaCl, 1 mM DTT,pH 7.9 at 37ºC)


183

10. Buffer HSupplied with Eco RI (90 mM Tris-HCl, 10 mM MgCl2, 50 mM NaCl, pH 7.5 at 37ºC)

11. Chloramphenicol solution (34 mg/mL)Chloramphenicol powder 34 mgEthanol (100 %) 1 mLSolution was dissolved, aliquoted and stored at –20ºC for up to 3 months.

12. Cytoplasmic extraction buffer (CEB)Hepes 100 mM, pH 7.1 10 mL (10 mM)MgCl2 1 M 300 µL (3 mM)KCl 1 M 4 mL (40 mM)Glycerol 5 mL (5 %)NP-40 200 µL (0.2 %)DEPC ddH20 up to 100 mLThe solution was autoclaved, aliquoted and stored at –20 ºC for up to 1 month.The following was added fresh prior to use of 5 mL CEB:DTT (1 M) 5 µL (1 mM)PMSF (0.1 M) 25 µL (0.5 mM)Leupeptin (1 mg/mL) 50 µL (10 µg/mL)Aprotinin (1 mg/mL) 10 µL (2 µg/mL)

13. DHT (4, 5 – dihydrotestosterone)DHT 50 mg100% Ethanol 10 mLDHT was weighed on a fine balance, into 100% ethanol. The sample was then diluted5.8 uL in 994.2 uL 100 % ethanol to make 100 uM. DHT is serially diluted to either 10uM or 1 uM concenctration.

14. Dithiothretol (DTT) solution (0.1 M)Dithiothretol powder 1 mgDEPC ddH20 up to 10 mLSolution was dissolved, aliquoted and stored at –20ºC for up to 3 months.

15. DNA loading dyeXylene Cyanol 10 mg (1 mg/mL)Bromophenol Blue 10 mg (1mg/mL)Sucrose 7 g (0.7 g/mL)EDTA (0.1 M) 10 mLThe solution was stored at room temperature.

16. ddH2ONormal tap water was slowly passed through a USF purifier that eliminates all ionsfrom the water, producing double distilled (ddH20) water.

17. dNTP Mix (10 mM)dATP (100 mM) 10 µLdCTP (100 mM) 10 µLdGTP (100 mM) 10 µLdUTP (100mM) 10 µL


184

ddH20 up to 100 µLSolution was mixed thoroughly and stored at -20°C.

18. DEPC-treated waterddH20 1000 mLDiethyl pyrocarbonate 1 mLDEPC was added to the water in the fume hood. The solution was left to mix overnighton the magnetic stirrer, and autoclaved the next day to remove any further DEPC. Thesolution was stored at room temperature.

19. DexamethasoneEstradiol 10 mg100% Ethanol 10 mLEstradiol was weighed on a fine balance, into 100% ethanol to make 2.5 x 10 –4 mol/L.The sample was then diluted 1 in 10 in water, followed by a final 4 uL in 1000 uLdilution in water. Two microlitres were then added to a 6 well plate for a final volumeof 10 nM.

20. Dextran-Coated Charcoal2.5 g of Dextran T-7025 g of acid-washed charcoalCharcoal and dextran-70 was added 500 mL of ddH20 and incubated overnight at 4 °Cwith stirring. Serum was recovered from suspension by centrifugation at 4 °C for 20min at 1 800 x g, washed twice by resuspension in fresh ddH20 and re-centrifuged for10 min.

21. DMEM / F12 SolutionDMEM / F12 powder mix GibcoNaHCO3 2.438 gddH20 up to 1000 mLSolution was stirred thoroughly and the pH adjusted to 7.36. The solution was thenfilter sterilised and stored at 4°C. Prior to use 100 U / mL penicillin and 100 mg / mLstreptomycin were added to the solution.

22. EDTA 0.5 M, pH 8.0EDTA 186.12 gddH20 up to 1000 mLEDTA was added to 700 mL water, and the pH adjusted using concentrated NaOH.The solution was then made up to 1 L with ddH20, and stored at room temperature.

23. Electrode Buffer (5 x)Tris.HCl 14.9 g 0.123 MGlycine 72.1 g 0.96 MSDS 4.9 g 17 mMIngredients were dissolved in ~980 mL of ddH20, and the pH adjusted to 8.0 withNaOH and make to 1 L with ddH20. The buffer was stored at room temperature.

24. EstrogenEstradiol 10 mg100% Ethanol 10 mLEstradiol was weighed on a fine balance, into 100% ethanol to make 3.7 x 10 –4 mol/L.The sample was then diluted 1 in 10 in water, followed by a final 2.7 uL in 1000 uL


185

dilution in water. Two microlitres were then added to a 6 well plate for a final volumeof 10 nM.

25. 70% EthanolAbsolute Ethanol 70 mLddH20 30 mLThe ethanol was made up to a total volume of 100 mL with ddH20, and stored either atroom temperature or at -20°C.

26. 75% EthanolAbsolute ethanol 75 mLddH20 25 mLThe ethanol was made up to a total volume of 100 mL with ddH20, and stored either atroom temperature or at -20°C.

27. 90% EthanolAbsolute ethanol 90 mLddH20 10 mLThe ethanol was made up to a total volume of 100 mL with ddH20, and stored either atroom temperature or at -20°C.

28. Ethidium bromide (10 mg/mL)Ethidium bromide 1 gddH20 up to 100 mLThe ethidium bromide was dissolved, and stored in the dark at room temperature.

29. 10% glycerol solutionGlycerol 10 mLddH20 90 mLThe solution was mixed well, and stored at 4ºC.

30. GST cleavage bufferTris – HCl, pH 7.0 0.3 mL (50 mM)NaCl (5 M) 0.2 mL (150 mM)EDTA 0.1 mL (1 mM)DTT (1 M) 5 µL (1 mM)PMSF (0.1 M) 200 µL (1 mM)Leupeptin (1 mg/mL) 50 µL (10 µg/mL)Aprotinin (1 mg/mL) 10 µL (2 µg/mL)ddH20 up to 10 mLSolution was thoroughly mixed and stored at 4ºC.

31. GST elution bufferTris, pH 8.0 (1 M) 2.5 mL (50 mM)EDTA (0.5 M) 100 µL (1 mM)Reduced glutathione 50 mg (10 mg/mL)ddH20 up to 10 mLSolution was made fresh and stored at 4ºC.

32. GW501516GW501516 1 mg100% DMSO 1 mL


186

GW501516 was weighed on a fine balance, into 100% ethanol. The sample was thendiluted 4.5 uL in 1 mL DMSO. One microlitre was then added to a 6 well plate for afinal volume of 500 nM, or 2 uL for a final concentration of 1 uM.

33. Heparin (50 mg/mL)The 100 mg vial of heparin was dissolved in 2 mL of DEPC-treated ddH20, aliquotedand stored at –20 °C.

34. High salt bufferHepes pH 7.8 (2 M) 10 mL (20 mM)Glycerol 50 mL (25 %)MgCl2 (1 M) 1.5 mL (15 mM)KCl (5 M) 24 mL (1.2 M)EDTA (0.5 M) 40 µL (0.2 mM)PMSF (0.1 M) 10 µL (0.2 mM)DTT (1 M) 5 µL (1 mM)DEPC ddH20 up to 100 mLSolution was dissolved in DEPC ddH20, and aliquoted and stored at -20ºC.

35. ICI182780ICI182780 10 mg100% Ethanol 10 mLICI182780 was weighed on a fine balance, into 100% ethanol to make 1.64 x 10 –4

mol/L. The sample was then diluted 1 in 10 in water, followed by a final 6.13 uL in1000 uL dilution in water. Two microlitres were then added to a 6 well plate for a finalvolume of 10 nM.

36. InsulinInsulin 10 mg / mL 1 uLRPM1 100 mLInsulin was made up to 0.1 mg / mL in RPM1, then added at 1 in 100 dilution to thestock RPM1, final concentration of insulin added to growing MCF-7 cells was 0.01 mg/ mL, as according to ATCC recommendations.

37. IPTG isopropyl-β-d-thiogalactoside (0.4 mM)IPTG powder 0.2383g (100 mM)ddH20 10 mLIPTG was dissolved in the water and the solution was aliquoted and stored at -20ºC.

38. KCl (1 M)KCl 74.6 gddH20 up to 1000 mLSolution stored at room temperature.

39. LB (Luria Bertani) brothBacto trypotone 5 gYeast extract 2.5 gNaCl 5 gddH20 up to 500 mLThe solution was mixed until dissolved, and then autoclaved. Appropriate antibioticswere added fresh prior to use after the solution had cooled.


187

40. LB agar platesBacto trypotone 5 gYeast extract 2.5 gNaCl 5 gAgar 7.5 gddH20 up to 500 mLThe solution was mixed until dissolved, and then autoclaved. Once the solution hadcooled to approximately 60ºC, appropriate antibiotics were added, the solution pouredinto plates, and the plates left to cool for 1 hour.

41. Leupeptin solution ( 1 mg/mL)Leupeptin powder 10 mgDEPC ddH20 up to 10 mLSolution was dissolved, aliquoted and stored at –20ºC for up to 3 months.

42. 10 x ligation buffer / 5 x transcription buffer / 10 x PCR bufferSupplied with the appropriate enzyme by Promega, Australia and Gibco, Australia.

43. Low salt bufferHepes pH 7.8 (2 M) 10 mL (20 mM)Glycerol 50 mL (25 %)MgCl2 (1 M) 1.5 mL (15 mM)KCl (1 M) 2 mL (0.02 M)EDTA (0.5 M) 40 µL (0.2 mM)PMSF (0.1 M) 10 µL (0.2 mM)DTT (1 M) 5 µL (1 mM)DEPC ddH20 up to 100 mLSolution was dissolved in DEPC ddH20, aliquoted and stored at -20ºC.

44. Low salt LBBacto trypotone 5 gYeast extract 2.5 gNaCl 2 gddH20 up to 500 mLThe solution was mixed until dissolved, and then autoclaved. Appropriate antibioticswere added fresh prior to use after the solution has cooled.

45. Mid-RIPA (moderate whole cell extraction buffer)NaCl (5 M) 1.5 mL 150 mMTris pH 8.0 (500 mM) 3 mL 25 mMIgepal-CA 6300.(100 %) 5 mL 1 % (v/v)Deoxycholate 0.25 g 0.12 mMIngredients were mixed and make up to a final volume of 46 mL with ddH20, andstored at –20oC.

46. MgCl2 (250 mM)MgCl2 23.83 gddH20 up to 250 mLSolution stored at room temperature.

47. NaCl (5 M)NaCl 292.5 g


188

ddH20 up to 1000 mLSolution stored at room temperature.

48. NETN bufferTris, pH 7.6 (1 M) 10 mLNaCl (5 M) 15 mLEDTA, pH 8.0 (0.5M) 1 mLddH20 up to 500 mLThe solution was mixed well, autoclaved, and stored at 4ºC.

49. 5% Non-denaturing polyacrylamide gelAcrylamide solution 30% (29:1) 1.64 mL10 x TBE 0.5 mLDEPC ddH20 7.83 mLAPS (10 %) 150TEMED 15Solution was mixed well and poured into the appropriate gel plates immediately as thesolution sets within 2 minutes. The gel was allowed to polymerase for at least one hourprior to use.

50. Phenylmethlsulfonyl fluoride (PMSF) solution ( 0.1 mg/mL)Phenylmethlsulfonyl fluoride 1 mgDEPC ddH20 up to 10 mLSolution was dissolved, aliquoted and stored at –20ºC for up to 3 months.

51. Phospate buffered saline (PBS)PBS powder GibcoddH20 up to 1000 mLpH was adjusted to 7.3 and then the solution filter sterilised, or autoclaved, and storedat 4ºC.

52. PIC SolutionPhenol 25 mLChloroform 24 mLIsoamylalcohol 1 mLPIC solution was made in this ratio: [25] Phenol : [24] Chloroform : [1]Isoamylalochol.Solution was stored at 4ºC.

53. 8% polyacrylamide separating gel (PAGE)Acrylamide solution 30 % (29:1) 2.7 mLResolving gel buffer 2 mLddH20 5.3 mLAmmonium persulphate (10 %) 150 µLTEMED 15 µLSolution was mixed well and poured into the appropriate gel plates immediately as thesolution sets within 2 minutes. The solution was poured up to the 3/4 mark only. Thegel was then allowed to dry for at least one hour. Once dry, a 4 % stacking gel was seton top.

54. 10% polyacrylamide separating gel (PAGE)Acrylamide solution 30 % (29:1) 3.3 mLResolving gel buffer 2 mL


189

ddH20 4.66 mLAmmonium persulphate (10 %) 150 µLTEMED 15 µLSolution was mixed well and poured into the appropriate gel plates immediately as thesolution sets within 2 minutes. The solution was poured up to the 3/4 mark only. Thegel was then allowed to set for at least one hour. Once set, a 4 % stacking gel was seton top.

55. 12.5% polyacrylamide separating gel (PAGE)Acrylamide solution 30 % (29:1) 4 mLResolving gel buffer 2 mLddH20 4 mLAmmonium persulphate (10 %) 150 µLTEMED 15 µLSolution was mixed well and poured into the appropriate gel plates immediately as thesolution sets within 2 minutes. The solution was poured up to the 3/4 mark only. Thegel was then allowed to set for at least one hour. Once set, a 4 % stacking gel was seton top.

56. 15% polyacrylamide separating gel (PAGE)Acrylamide solution 30 % (29:1) 5 mLResolving gel buffer 2 mLddH20 3 mLAmmonium persulphate (10 %) 150 µLTEMED 15 µLSolution was mixed well and poured into the appropriate gel plates immediately as thesolution sets within 2 minutes. The solution was poured up to the 3/4 mark only. Thegel was then allowed to set for at least one hour. Once set, a 4 % stacking gel was seton top.

57. PreScission protease cleavage bufferCleavage buffer 335 µL (95.5 %)PreScission protease enzyme 15 µL (4.5 %)Solution was made up fresh immediately prior to use and stored at 4ºC and on ice whilein use.

58. Resolving gel bufferTris 182 gSDS 4 gddH20 up to 1000 mLpH was adjusted to 8.8 with HCl, and the solution stored at 4ºC.

59. REMSA fixing solutionAcetic acid 100 mL (10%)Isopropanol 100 mL (10%)DEPC ddH20 up to 1000 mLSolution was mixed thoroughly and stored at room temperature.

60. REMSA Loading DyeFormamide 8 mL 80 % v/vEDTA (500 mM) 400 µL 20 mM


190

Sterile glycerol (100 %) 900 µL 9 % v/vBromophenol blue 10 mg 0.1% w/vXylene cyanol 10 mg 0.1 % w/vIngredients were dissolved and made to 10 mL with 5x TBE, aliquoted and store at –20°C.

61. Riboprobe elution bufferAmmonium acetate (5 M) 10 mL (5 M)EDTA, pH 8.0 (0.5 M) 200 µL (1 mM)Solution was mixed thoroughly and stored at 4ºC.

62. rNTP solution (2.5 mM; no UTP)ATP (100 mM) 2.5 µLCTP (100 mM) 2.5 µLGTP (100 mM) 2.5 µLDEPC ddH20 up to 100 µLSolution was mixed thoroughly and stored at -20ºC.

63. rNTP solution (2.5 mM; with UTP)ATP (100 mM) 2.5 µLCTP (100 mM) 2.5 µLGTP (100 mM) 2.5 µLUTP (100 mM) 2.5 µLDEPC ddH20 up to 1 mLSolution was mixed thoroughly and stored at -20ºC.

64. RPM1 SolutionRPMI powder mix GibcoNaHCO3 2.438 gddH20 up to 1000 mLSolution was stirred thoroughly and the pH adjusted to 7.36. The solution was thenfilter sterilised and stored at 4°C. Prior to use 100 U / mL penicillin and 100 mg / mLstreptomycin were added to the solution.

65. SOC mediumBactotryptone 1 g (2 %)Bacto yeast extract 0.25 g (0.5 %)NaCl (5 M) 100 µL (10 mM)KCl (1 M) 125 µL (2.5 mM)MgCl2 (250 mM) 500 µL (10 mM)MgSO4 500 µL (100 mM)Glucose (20 % solution) 1 mL (20 mM)ddH20 up to 50 mLSolution was mixed well and filter sterilised. SOC medium was kept at roomtemperature for up to one month.

66. Sodium AcetateNaAc powder 82.03 gddH20 up to 1000 mLSolution was mixed well and brought to pH 4.0 with acetic acid prior to autoclaving.Solution was stored at room temperature.


191

67. 10 x soduim dodecyl sulphate (SDS) loading bufferTris 151.5 gGlycine 720 gSDS 5 gddH20 up to 5000 mLSolution was mixed well and stored at room temperature.

68. Sodium Hydroxide (1 M)NaOH 40 gddH20 up to 1000 mLSolution was mixed well, autoclaved and stored at room temperature.

69. Stacking gel bufferTris 30 gSDS 2 gThe solution was made up with 400 mL ddH20 and the pH adjusted to 6.8 with HCl.The following was then added:ddH20 up to 500 mLand the solution was stored at 4ºC.

70. Stop SolutionFormamide 8 mL (80 %)EDTA (1 M) 200 µL (10 mM)Xylene Cyanol 10 mg (1 mg/ mL)Bromophenol Blue 10 mg (1 mg/mL)ddH20 up to 10 mLSolution was mixed thoroughly and stored at 4ºC.

71. T3 (Thyroid hormone)T3 10 mg100% Ethanol 10 mLT3 was weighed on a fine balance, into 100% ethanol to make 1.53 x 10 –4 mol/L. Thesample was then diluted 1 in 10 in water, followed by a final 6.5 uL in 1000 uL dilutionin water. Two microlitres were then added to a 6 well plate for a final volume of 10nM.

72. TamoxifenTamoxifen 10 mg100% Ethanol 10 mLTamoxifen was weighed on a fine balance, into 100% ethanol to make 1.77 x 10 –4


73. TE BufferTris 60.57 g (10 mM)EDTA (500 mM) 1 mL (1 mM)ddH20 up to 800 mLThe mix was thoroughly dissolved in ddH20, then the pH altered to choice (7.5 or 8.0)with HCl / NaOH. The solution was topped up with ddH20 to 1 L and autoclaved. Thesolution was then stored at room temperature.


192

74. Tetracycline solution ( 5 mg/mL)Tetracycline powder 5 mgEthanol (100 %) 1 mLSolution was dissolved, aliquoted and stored at –20ºC for up to 3 months.

75. 50 x tris acetate EDTA (TAE) bufferTris base 242 gGAA 57.1 mLEDTA 37.2 gddH20 up to 1000 mLSolution was mixed well and stored at room temperature.

76. 10 x tris boric EDTA (TBE) bufferTris base 216 gBoric acid 110 gEDTA pH 8.0 (5 M) 80 mLddH20 up to 2 LSolution was mixed thoroughly and stored at room temperature for up to 3 months.

77. TBS (Tris-buffered saline)Tris pH 8.0 (500mM) 100 mL 50 mMNaCl (5 M) 30 mL 150 mMddH20 up to 1000 mLSolution stored at room temperature.

78. TBS-T (Tris-buffered saline-Tween 20)Tween-20 1 mLTBS 1000 mLSolution stored at room temperature.

79. Tris- HCl (1 M)Tris 24.22 gddH20 up to 100 mLThe solution was mixed, the pH adjusted to 8.0, and the solution topped off with ddH20to 200 mL.

80. 0.2% Triton-XTriton-X 0.2 mLddH20 up to 100 mLAutoclaved and stored at 4oC

81. Trypan Blue (0.4 %)Trypan blue 40 mgPBS 10 mLSolution stored at room temperature.. 82. 6% (5M) urea gelUrea 4.2 g10 x TBE 1 mLAcrylamide (30 %) solution 19:1 1.67 mLDEPC ddH2O up to 10 mL


193

The solution was mixed well until completely dissolved. Then the following wasadded:APS (10 %) 100 µLTEMED 12 µLThe resulting solution was then added immediately to the appropriate gel plates, as thesolution sets within 2 minutes. The gel was allowed to dry for at least one hour prior touse.

83. UV–cross linking fixer solutionAcetic acid 100 mL (10 %)Methanol (100 %) 500 mL (50 %)DEPC ddH20 up to 1000 mLSolution was mixed thoroughly and stored at room temperature.

84. Vitamin D3Vitamin D3 10 mg100% Ethanol 10 mLVitamin D3 was weighed on a fine balance, into 100% ethanol to make 2.599 x 10 –4


85. WTB (western transfer buffer)Tris 3.03 g 25 mMGlycine 14.4 g 190 mMSDS (10 %) 2 mL 0.2 %(w/v)Methanol (100 %) 100 mL 10 %(v/v)Tris and glycine were dissolved in 900 mL of ddH20, the methanol and SDS were thenadded. This solution was stored at room temperature and used within one month.


194

E. Hatchell Appendix II: Prizes and Awards

195

Appendix II

Prizes and Awards

E. Hatchell Appendix II: Prizes and Awards

196

Appendix II: Prizes and Awards

The following awards and prizes were awarded during the course of this candidacy:

1. Dora Lush NHMRC Scholarship (2004 – 2007).

2. Novartis Junior Scientist Award at the Endocrine Society of Australia Annual

General Meeting (2004).

3. ESA Travel Award (2004).

4. Young Investigator Encouragement Award (2004).

5. ANZSCDB David Walsh Student Prize, ComBio Conference (2004).

6. ENDO Travel Grant, ENDO (2005).

7. ASBMB Student Poster Prize, ComBio Conference (2005).

8. UWA Postgraduate Travel Award (2006).

9. WAIMR Travel Award (2006).

10. ESA Travel Award (2006).

11. ASMR AAIR Award (2006).

E. Hatchell Appendix III: Publications and Patents

197

Appendix III

Publications and Patents


198

Appendix III: Publications and Patents

The following is a list of published journal (or submitted publications), publishedabstracts, patents, and invited talks arising from the work presented in this thesis:

REFEREED JOURNAL ARTICLES:

1. Esme C Hatchell, Shane M Colley, Dianne J Beveridge, Michael R Epis, Lisa MStuart, Keith M Giles, Andrew D Redfern, Lauren EC Miles, Andrew Barker, LouisaMacDonald, Peter G Arthur, James CK Lui, Jemma L Golding, Ross K McCulloch,Cecily B Metcalf, Jackie A Wilce, Matthew CJ Wilce, Rainer B Lanz, Bert WO’Malley, Peter J Leedman (2006) SLIRP, a Small SRA Binding Protein, Is a NuclearReceptor Corepressor, Molecular Cell 22, 657 – 688.

2. Andrew D Redfern, Naoya Ikeda, Michael Epis, Dianne J Beveridge, Lisa M Stuart,Xiaotao Li, Esme C Hatchell, Rainer B Lanz,, Louisa M MacDonald, Shane M Colley,Cecily Metcalf, Anne Gatignol, Bert W O’Malley and Peter J Leedman (2007) PACT,PKR and TAR-binding protein: double stranded RNA-binding proteins that modulateSRA-dependent nuclear receptor signaling Molecular Cell (In Review, 2007).

3. Jodi M Saunus, Stacey L Wardrop, Juliet D French, Dianne J Beveridge, Esme CHatchell, Alexandra A Gason, Kaylene J Simpson, kConFab, Peter J Leedman, MelissaA Brown (2006) Post-transcriptional regulation of the breast cancer susceptibility geneBRCA1 by the RNA binding protein HuR (Manuscript to be resubmitted, 2008).

PUBLISHED SCIENTIFIC ABSTRACTS:

1. Hatchell E.C., Colley S.M., Beveridge D.J., Epis M.R., Stuart L.M., Lanz R.B.,O'Malley B.W. and Leedman P.J. (2006) Complex Interactions Between SLIRP, SRAand Other Nuclear Receptor Coregulators. Australian Society for Medical ResearchWeek Symposium, June 2006, Perth, Australia.

2. Hatchell E.C., Colley S.M., Beveridge D.J., Redfern A.D., Stuart L.M., McCullochR.K., Wilce M.C., Lanz R.B., O'Malley B.W. and Leedman P.J. (2005) Investigation ofthe Cellular Function and Interactions of SLIRP, a Novel Nuclear Corepressor of theEstrogen Receptor Pathway. ComBio Conference, September 2005, Adelaide,Australia.

3. Hatchell E.C., Colley S.M., Beveridge D.J., Redfern A.D., Stuart L.M., McCullochR.K., Wilce M.C., Lanz R.B., O'Malley B.W. and Leedman P.J. (2005) SLIRP: a novelcorepressor and SRA-binding protein. Endocrine Society of Australia (ESA) AnnualScientific Meeting, September 2005, Perth, Australia.

4. Hatchell E.C., Colley S.M., Beveridge D.J., Redfern A.D., Stuart L.M., McCullochR.K., Wilce M.C., Lanz R.B., O'Malley B.W. and Leedman P.J. (2005)Characterisation of SLIRP, a novel RRM-containing SRA-binding protein, thatregulates estrogen action. ENDO Meeting, San Diego, USA.


199

5. Hatchell E.C., Colley S.M., Beveridge D.J., Stuart L.M., Lanz R.B., O'Malley B.W.and Leedman P.J. (2004) Cloning of a Novel SRA-Binding Protein that RegulatesEstrogen Action in Cancer Cells. ComBio Conference, Perth, Australia.

6. Redfern A., Hatchell E., Beveridge D.J., Lanz R.L., Stuart L., Wilce M.C.J. andLeedman P.J. (2004) SRA-Binding Proteins and Hormone Action in Cancer. ComBioConference, Perth, Australia.

7. Esme C Hatchell, Shane M Colley, Andrew D Redfern, Dianne J Beveridge, Lisa MStuart, Rainer B Lanz*, Bert W O’Malley* and Peter J Leedman (2004) Cloning of aNovel SRA-Binding Protein that Regulates Estrogen Action in Cancer Cells. MedicalResearch Foundation, Royal Perth Hospital Young Investigator Finalist, Perth,Australia.

8. Esme C Hatchell, Shane M Colley, Andrew D Redfern, Dianne J Beveridge, Lisa MStuart, Rainer B Lanz, Bert W O’Malley and Peter J Leedman (2004) Cloning of aNovel SRA-Binding Protein that Regulates Estrogen Action in Cancer Cells. EndocrineSociety of Australia, Novartis Junior Scientist Award, Sydney, Australia.

PATENT APPLICATIONS / INTELLECTUAL PROPERTY:

Part of the contents of this thesis were submitted for an International Patent entitled‘ANovel SRA-Binding Protein’, filed in Perth, Western Australia, 22nd July & 23rd

December, 2005. (WO/2007/009194). The patent is due to enter National Phase in2008.

INVITED TALKS:

1. The Monash Institute for Medical Research & Monash Centre for Excellence,Melbourne Australia, 20th of July 2006.2. The Australian Asthma and Allergy Institute, QEII Hospital, Perth Australia, 28th ofJuly 2006.

ORGANISED CONFERENCES:

1. Local Organising Committee for the Endocrine Society of Australia annual nationalconference, Perth, September 2005.

INVITED CHAIRS:

1. Endocrine Society of Australia, annual national conference, Novartis Junior ScientistAward Session, 2005, Perth.2. Endocrine Society of Australia, annual national conference, Gene Regulation andConference Session, 2006, Gold Coast.


200

Molecular Cell 22, 657–668, June 9, 2006 ª2006 Elsevier Inc. DOI 10.1016/j.molcel.2006.05.024

SLIRP, a Small SRA Binding Protein,Is a Nuclear Receptor Corepressor

Esme C. Hatchell,1,2,7 Shane M. Colley,1,7

Dianne J. Beveridge,1,2 Michael R. Epis,1,2

Lisa M. Stuart,1,2 Keith M. Giles,1,2

Andrew D. Redfern,1,2 Lauren E.C. Miles,1,2

Andrew Barker,1,2 Louisa M. MacDonald,1

Peter G. Arthur,3 James C.K. Lui,3

Jemma L. Golding,1,2 Ross K. McCulloch,4

Cecily B. Metcalf,5 Jackie A. Wilce,3

Matthew C.J. Wilce,2 Rainer B. Lanz,6

Bert W. O’Malley,6 and Peter J. Leedman1,2,*1Laboratory for Cancer MedicineThe University of Western Australia Centre

for Medical ResearchWestern Australian Institute for Medical Research2School of Medicine and Pharmacology and3School of Biomedical and Chemical SciencesUniversity of Western Australia4Research Centre and5Department of PathologyRoyal Perth HospitalPerth, Western AustraliaAustralia6Division of Molecular and Cellular BiologyBaylor College of MedicineHouston, Texas 77030

Summary

Steroid receptor RNA activator (SRA), the only knownRNA coactivator, augments transactivation by nuclear

receptors (NRs). We identified SLIRP (SRA stem-loopinteracting RNA binding protein) binding to a func-

tional substructure of SRA, STR7. SLIRP is expressedin normal and tumor tissues, contains an RNA recogni-

tion motif (RRM), represses NR transactivation ina SRA- and RRM-dependent manner, augments the

effect of Tamoxifen, and modulates association ofSRC-1 with SRA. SHARP, a RRM-containing corepres-

sor, also binds STR7, augmenting repression withSLIRP. SLIRP colocalizes with SKIP (Chr14q24.3), an-

other NR coregulator, and reduces SKIP-potentiatedNR signaling. SLIRP is recruited to endogenous pro-

moters (pS2 and metallothionein), the latter in a SRA-dependent manner, while NCoR promoter recruitment

is dependent on SLIRP. The majority of the endoge-nous SLIRP resides in the mitochondria. Our data

demonstrate that SLIRP modulates NR transactiva-tion, suggest it may regulate mitochondrial function,

and provide mechanistic insight into interactions be-

tween SRA, SLIRP, SRC-1, and NCoR.

Introduction

Coregulators, functioning as coactivators or corepres-sors of nuclear receptor (NR) activity, play pivotalroles in mediating hormone action via the regulation of

transcriptional efficiency (McKenna and O’Malley, 2002).The discovery of steroid receptor coactivator-1 (SRC-1),a broad-spectrum coactivator (Onate et al., 1995), pro-vided important insight into the mechanisms under-lying transcriptional activation by NRs. Since then,a large family of coregulators has been discovered,each of which is selectively recruited by specific NRs ina ligand- and tissue-specific manner to cognate re-sponse elements in the genome (Robyr et al., 2000;McKenna and O’Malley, 2002).

For the estrogen receptor (ER), a NR that plays a keyrole in the proliferation of breast cancer cells (Herynkand Fuqua, 2004), a large number of coregulators havebeen identified. These include SRC-1 (Onate et al.,1995), SHARP (Shi et al., 2001), and SRA (Lanz et al.,1999). Remarkably, SRA (steroid receptor RNA activa-tor) is the only known coregulator that has the capacityto coactivate as an RNA and for this reason stands alonein its functional characteristics.

SRA plays an important role in mediating 17b-estra-diol (E2) action (Lanz et al., 1999, 2003). Its expressionis aberrant in many human breast tumors, suggestinga potential role in tumorigenesis (Murphy et al., 2000).Despite evidence that an alternative splice variant ofSRA exists as a protein (Chooniedass-Kothari et al.,2004), it has been conclusively shown that SRA can func-tion as an RNA transcript to coactivate NR transcription(Lanz et al., 1999, 2002, 2003). While it is currently hypoth-esized that SRA acts as an RNA scaffold for other co-regulators at the transcription initiation site, the precisemechanism by which SRA augments ER activity remainsunclear.

Recent findings have identified protein interactors ofSRA and provided insight into the putative mechanismsunderlying SRA’s transcriptional coactivation ability.Specifically, SHARP (SMRT/HDAC1 associated repres-sor protein) is a NR corepressor that interacts with SRAin vitro and contains three RNA recognition motifs(RRMs) (Shi et al., 2001). These RRMs are required bySHARP to repress SRA-augmented E2-induced trans-activation (Shi et al., 2001). Another ER coregulator thatbinds SRA in vitro and copurifies with SRA from cellextracts is p72 (Wantanabe et al., 2001). Thus, althoughSRA-protein interactions impact significantly on NR ac-tivity and signaling, the specifics of interactions remainunclear, and the identity of SRA binding proteins whosefunction is dependent upon targeting specific SRA sub-structures is unknown.

We sought to identify SRA binding proteins using aspecific stem-loop structure of SRA (stem-loop struc-ture 7, STR7) that was identified both as important forits coactivator function (Lanz et al., 2002) and also asa target for proteins from breast cancer cell extracts.Here we describe the isolation and characterization ofSLIRP (SRA stem-loop interacting RNA binding protein),a widely expressed small SRA binding protein, thatis a repressor of NR signaling. SLIRP functions in anadditive manner with SHARP to further repress ERactivity and augments the estrogen antagonistic effectsof 4-hydroxytamoxifen (Tam) and ICI 182780 (ICI).

*Correspondence: [email protected] These authors contributed equally to this work.

mailto:[email protected]

Although SLIRP is localized predominantly to the mito-chondria, it is actively recruited to hormone-responsivepromoters where it modifies NR cofactor recruitmentand transactivation. Interestingly, SLIRP lies adjacentto SKIP (Ski-interacting protein), another NR coregulator(MacDonald et al., 2004), in the genome and antagonisesSKIP’s coactivation of the ER. Taken together, thesedata suggest a key role for SLIRP as a corepressor,modulating several NR pathways.

Results

SRA STR7 Is a Target for Proteinsin Human Breast Cancer Cells

SRA is a complex RNA molecule predicted to containmultiple stable stem-loop structures (Figure 1A) (Lanzet al., 2002). STR7, an 89 nucleotide (nt) sequence, isthe largest and one of the most stable stem-loop SRAstructures that functions in a cooperative manner withother stem loops to augment ER transactivation (Lanzet al., 2002). Based on these observations, we firstinvestigated if SRA STR7 was a target for RNA bindingproteins in RNA electrophoretic mobility shift assay(REMSA) studies. While weak RNA-protein complexes(RPCs) formed with cytoplasmic extracts from MCF-7,MDA-MB-468, and HeLa cell lines, two strong RPCswere evident in the nuclear extracts of each (Figure 1B).Addition of excess unlabeled STR7 effectively abro-gated RPC formation (Figure 1C, lane 3). However, addi-tion of w100-fold excess of either unlabeled vector tran-script (pBlue) or yeast tRNA competitor RNA did notdiminish the formation of RPCs in HeLa cell nuclear ex-tracts (Figure 1C, lanes 4–7). Taken together, these dataindicate a highly specific interaction between SRA STR7and nuclear proteins from human cancer cells.

UV crosslink (UVXL) studies were performed to furthercharacterize RPC formation with the STR7 riboprobe.Multiple STR7-protein complexes were identified fromeach of the nuclear cell extracts (Figure 1D). Althoughmany bands were common, there were some significantdifferences between cell types. For example, two RPCs,of w39 and 40 kDa, were more prominent in extractsfrom MDA-MB-468 cells than in MCF-7 and HeLa ex-tracts (Figure 1D). These data show that an array ofnuclear proteins bind SRA STR7 in vitro.

Cloning of SLIRP, a SRA Binding ProteinTo isolate SRA binding proteins, we used SRA STR7 asbait in a yeast three-hybrid screen (SenGupta et al.,1996) of a primary human breast cancer cDNA library(Byrne et al., 1998). Mfold secondary (2º) structure anal-ysis indicated that the STR7 stem-loop structure waspreserved in the hybrid RNA transcribed from the pIIIA/MS2-2 bait construct. From the screen, we isolated acDNA clone that contained an open reading frame witha 30 untranslated region (UTR) and polyadenylated(poly A) tail (Figure 2A). The cDNA sequence predicteda protein of 109 amino acids (aa), with a Mr of 12.7 kDa.Database analysis revealed that this clone, which wetermed SLIRP, was identical to human sequences Hypo-thetical Protein DC50 (GenBank accession number[GAN], AF271779) and Chr 14 Open Reading Frame 156(C14orf156) (GAN, BC017895).

The SLIRP mRNA sequence (GAN, AY860853) pre-dicts a protein composed almost entirely of an RRM(Figure 2A) containing RNP1 and RNP2 submotifs(Burd and Dreyfuss, 1994). The RRM domain in SLIRPshares substantial aa homology with SHARP (Figure 2B),a SRA corepressor (Shi et al., 2001), and nucleolin, a ca-nonical RRM-containing protein (Bouvet et al., 1997).Significantly, homology between SLIRP and SHARP,particularly within key interacting residues of theirRRMs (Figure 2C), suggests these molecules may bindthe same RNA targets, i.e., SRA STR7.

The aa sequence of SLIRP is highly conserved acrosshuman, rat, and mouse species (Figure 2D). Curiously,rat and mouse SLIRP homologs are surrounded by thesame genes as human SLIRP. This striking sequenceconservation suggests an important function for SLIRPin multiple species. Of interest, human SLIRP is posi-tioned within 1750 nt of SKIP, on Chr 14q24.3 with no in-tervening genes. SKIP is a vitamin D receptor (VDR) co-repressor implicated in oncogenesis (Barry et al., 2003;MacDonald et al., 2004). SKIP is expressed in breastcancer tissue and regulates ER transactivation (Barryet al., 2003). Colocalization of these genes suggestedthat SLIRP and SKIP may participate in the same NRpathways and could be coordinately expressed.

SLIRP Is Expressed Widely in Human Tissuesand Cancer Cells

In normal human tissue, SLIRP mRNA is ubiquitously ex-pressed, but in varying amounts, with the highest levelsin heart, liver, skeletal muscle, and testis (Figure 3A).SLIRP was readily detected in a variety of cell lines, in-cluding SK-BR-3, MCF-7, HMEC, MDA-MB-468, LNCaP,and COS-7 (Figure 3B) and increased in HeLa, Calu-6,and HepG2 cells. Notably, we found that SLIRP expres-sion across multiple tissues (Figure 3A) and cell lines(Figure 3B) was similar to that of SRA (refer to Lanzet al. [1999], Figure 1B).

We generated polyclonal SLIRP antisera (ab) anddemonstrated SLIRP protein (Mr w12.7 kDa) expressionin multiple human cell lines, including those derivedfrom breast, prostate, and lung carcinomas (Figure 3C).The ab was highly specific for human SLIRP, with virtu-ally no cross reactivity with a number of other species(see Figure S1 in the Supplemental Data available withthis article online). Expression of SLIRP protein variedacross different breast cancer cell lines, and in somecells, discordant levels of SLIRP mRNA and proteinwere observed (e.g., HeLa cells). While expression ofSLIRP and SKIP was similar, there was little evidencethat their expression at the mRNA or protein level wascoordinately regulated.

Immunohistochemistry (IHC) with SLIRP ab on humanprimary breast cancer tissue showed SLIRP staining innormal ductal tissue, but little in the surrounding stroma(Figures 3Da and 3Db). Intense SLIRP staining wasnoted in carcinoma tissue (Figure 3Dd) compared tocontrol (Figure 3Dc). Staining was evident throughoutthe cell, but predominantly with punctate, cytoplasmicdistribution (Figures 3Db and 3Dd).

Characterization of SLIRP’s Interaction with SRATo confirm SLIRP’s interaction with SRA in vivo, weperformed immunoprecipitation-RT-PCR (IP-RT-PCR)

Molecular Cell658

assays with SLIRP ab. Using HeLa (Figure 4A),MCF-7, and MDA-MB-468 cells (data not shown), SRAcoimmunopurified with SLIRP (Figure 4A, lane 5), butnot b actin (Figure 4A, lane 6). Thus, SLIRP closely inter-acts with SRA mRNA in several different cancer cell lines.As SRA also copurifies with SRC-1 (Lanz et al., 1999), weexamined the effects of reducing intracellular SLIRPlevels with siRNA and found a corresponding increasein SRC-1 associated with SRA (Figures 4B and 4C).This suggests that competition may exist betweenSRC-1 and SLIRP for association with SRA in vivo, whichcould directly impact on their coregulator effects.

Using recombinant SLIRP proteins (Figure 4D), bothGST-SLIRP (Figure 4E, lane 2) and cleaved SLIRP (datanot shown) bound STR7 avidly, while GST alone did not(Figure 4G, lane 3). Addition of increasing amounts of un-labeled (cold) STR7 probe efficiently competed out thecomplex (Figure 4E, lanes 3 and 4). In contrast, neitheraddition of excess unlabeled pBlue (Figure 4E, lanes 5and 6) nor high amounts of tRNA significantly affectedSLIRP-STR7 complex formation (data not shown). Takentogether, these results indicate that SLIRP binds STR7 invitro with a high degree of specificity.

We found binding of GST-SLIRP to the SRA SDM7probe (a SRA STR7 mutant containing several stem-loop point mutations and having reduced transactiva-tion activity compared to wild-type) (Lanz et al., 2002)was consistently reduced compared to the STR7 probe(up to 2.9-fold) (Figure 4F, lane 6). SLIRP binding to theSDM7 probe could also be overcome with excess unla-beled SRA STR7 (lanes 7 and 8).

Given the homology between SLIRP and SHARPwithin their RRM domains, we next examined if SHARP

could also bind SRA STR7. A GST-SHARP fusion protein(Shi et al., 2001) containing SHARP’s three RRMdomains (GST-SHARP-RRM) bound STR7 avidly (Fig-ure 4G, lane 1), while GST-SHARP-RD (containing onlyits repression domain) did not (Figure 4G, lane 2). Thesedata indicate that the SHARP RRMs may compete withSLIRP for binding to SRA STR7.

SLIRP Represses SRA-Mediated Nuclear ReceptorCoactivation

As SLIRP was identified from a human breast cancer li-brary and SRA is an activator of ER signaling (Lanz et al.,1999), we were particularly interested in assessingSLIRP’s potential role as a modulator of E2 action. Intransfection assays in HeLa cells using an E2-respon-sive reporter, we found that SRA coactivated reporteractivity approximately 3- to 4-fold (Figure 5A), as previ-ously reported. When cotransfected with SRA, SLIRP re-pressed SRA-augmented coactivation by up to 3-foldin a dose-dependent manner (Figure 5A). These datadefine SLIRP as an ER corepressor. Addition of SLIRPto cells cotransfected with SRA and treated with Tamor ICI further enhanced the E2-antagonistic activitiesof these compounds (Figure 5A).

We next investigated whether SLIRP repressed otherNRs. Cotransfection of SLIRP with NR reporters andSRA into HeLa cells resulted in strong repression of glu-cocorticoid (GR), androgen (AR), thyroid (TR), and VDR-mediated transactivation (Figure 5B), indicating thatSLIRP can modulate several different NR signaling path-ways. To determine if SLIRP could act on orphan NRs,we used a peroxisome proliferator-activated receptord (PPARd) agonist GW501516 (Oliver et al., 2001) and a

Figure 1. SRA STR7 Is Bound by Human

Breast Cancer Cell Proteins

(A) Mfold secondary structure plot of SRA

(Zucker, 2003). DE value for the full-length

SRA structure was 2243.1 kJ mol21. STR7

(labeled) is one of the most stable substruc-

tures of SRA.

(B) Comparison of nuclear and cytoplasmic

MCF-7, HeLa, and MDA-MB-468 extract

binding to SRA STR7 via REMSA.

(C) REMSA with SRA STR7 showing reduced

complex formation with unlabeled ‘‘cold’’

competitor STR7, but not excess cold pBlue

vector or tRNA. RPC, RNA-protein complex.

+–+++, increasing cold competitor RNA.

(D) UV crosslink assay with nuclear extracts

from cell lines in (B). Cell extract (30 mg) was

incubated with labeled STR7, UV irradiated,

RNase A digested, resolved by SDS-PAGE,

and detected by PhosphorImager after trans-

fer to PVDF membrane. [14C] molecular

weight markers were used as size standards.

Arrows with asterisk highlight 39 and 40 kDa

RPCs in MDA-MB-468 cells.

SLIRP, a SRA Binding Nuclear Receptor Corepressor659

Figure 2. SLIRP Is an RRM-Containing SRA Binding Protein

(A) Nucleotide and amino acid (aa) sequence of SLIRP. The entire mRNA is shown. Arrow denotes sequence isolated via yeast three-hybrid

screen, start and stop codons are in capitals, italics denote poly A signal. SLIRP contains a highly conserved RRM (underlined) with

consensus RNP2 and RNP1 submotifs (highlighted). *, ^, and 2 denote putative N-myristoylation, protein kinase C phosphorylation, and casein

kinase II phosphorylation sites, respectively.

(B) Sequence alignment comparing SLIRP, SHARP and nucleolin RRMs. Black boxes indicate aas conserved with consensus RRM sequence

described by Burd and Dreyfuss (1994).

(C) SLIRP and SHARP functional domains. RRM, RNA recognition motif. RID, receptor interaction domain. SID/RD, repression domain. Numbers

denote aa sequence position.

(D) Alignment of human, mouse, and rat SLIRP aa sequences illustrates high degree of homology between species. Black, aa identity; gray, sim-

ilarity; white, no homology.

Molecular Cell660

PPAR-Luc reporter (PPARE) (Dressel et al., 2003). SRAaugmented the activation by the agonist w2-fold, whichwas repressed (up to 2.9-fold) by SLIRP (Figure 5B).These data suggest that SLIRP has broad corepressoractivity within the NR superfamily.

To complement our SLIRP overexpression studies,we investigated the effects of SLIRP siRNA on dexa-methasone (Dex)-responsive reporter activity. In cellswith reduced endogenous SLIRP expression, we founda 10-fold increase in GRE-luc activity, further confirmingthat SLIRP acts as a NR corepressor (Figure 5C).

SLIRP Modulates SHARP- and SKIP-MediatedCoregulation of NR Activity

The high aa sequence homology between SHARP andSLIRP and their avid binding to SRA STR7 in vitro sug-gested that a functional interaction may exist betweenthese molecules in vivo. When cotransfected with SRA,SHARP repressed SRA-mediated coactivation of theE2-responsive reporter (Figure 5D, lane 4) as previouslyreported (Shi et al., 2001). When SLIRP was cotrans-fected with SHARP and SRA, an additional 2-fold repres-sion of SRA-augmented coactivation was observed (Fig-ure 5D, lane 5). Thus, SHARP and SLIRP appear to actin an additive fashion to enhance repression of the E2-responsive reporter.

We investigated the effects of SKIP on SLIRP repres-sion in further transfections (Figure 5D, lanes 6–12). Inthe presence of transfected SKIP alone, reporter activitywas increased w2-fold, consistent with SKIP functioningas a coactivator of ER transactivation. In the presenceof cotransfected SRA, an additive effect was observedwith a total increase in activity of w6-fold. When SLIRPwas added, reporter activity was reduced by more than5-fold. Thus, in the presence of SRA, SLIRP is a potentrepressor of SKIP-mediated coactivation. When wereduced endogenous SLIRP expression using siRNA,SLIRP repression was abrogated. Reduction of SKIPexpression with siRNA reversed SKIP’s coactivationeffect. When expression of SLIRP and SKIP were both re-duced, an intermediate reporter activity resulted. Takentogether, these data validate the functional role of eachof these coregulators on ER transactivation and suggestthat a competitive interaction exists between SLIRP andSKIP in NR signaling.

SLIRP Function Requires an Intact RRM DomainTo investigate the structural and functional significanceof the RRM domain within SLIRP, we assessed the prop-erties of proteins with mutations to this motif (Figure 5E).Based on binding predictions from other RRM-containingproteins, arginine 24 and 25 were mutated to alanines(R24,25A) in the RNP2 submotif, and within the RNP1domain, leucine 62 was mutated to alanine (L62A). Adouble mutant (DM) containing both the R24,25A andL62A substitutions was also prepared. In REMSA stud-ies, each of the mutations markedly reduced bindingto the SRA STR7 probe (Figure 5F, lanes 4–9 and Fig-ure S2). In transfections, each mutant partially relievedthe SLIRP-mediated repression (Figure 5G, lanes 4–6),indicating the requirement of an intact RRM domain forSLIRP to function as a repressor of E2-induced SRAcoactivation.

Figure 3. SLIRP Is Widely Distributed in Normal Human Tissues and

Cancer Cell Lines

(A) Northern analysis of SLIRP compared with b-actin in normal hu-

man tissues.

(B) Northern analysis of SLIRP in cancer cell lines. mRNA from hu-

man breast (SK-BR-3, MCF-7, MDA-MB-468), prostate (LNCaP),

lung (Calu-6), cervical (HeLa), and liver (HepG2) cell lines and normal

mammary (HMEC), and monkey kidney (COS-7) cells probed with

SLIRP, SKIP, and GAPDH probes.

(C) Immunoblot of protein lysates from breast (SK-BR-3, MCF-7,

MDA-MB-468, T47D), cervical (HeLa), prostate (LNCaP, PC3), lung

(Calu-6), and fibrosarcoma (HT1080) cells using SLIRP, SKIP, or b-

actin abs.

(D) SLIRP in primary human breast cancer tissue. Sections (20 3 a,c;

40 3 b,d) from a human breast ductal cancer were probed with

SLIRP ab (Da, Db, and Dd) and compared with sections from the

same tumor with no ab (Dc). Arrows denote stroma, ducts, and tu-

mor tissue. Box in (Da) denotes region magnified in (Db) (403).


Figure 4. SLIRP Associates with SRA In Vivo, Regulates SRC-1-SRA Association, and Binds SRA STR7 In Vitro

(A) SLIRP associates with SRA in vivo. SRA was detected by RT-PCR in supernatant samples or following immunoprecipitation with beads plus

SLIRP but not b-actin ab. No product generated from RT2 samples. P, SRA expression plasmid; W, no template. Arrow denotes 260 bp SRA-

specific PCR product.

(B and C) SLIRP knockdown augments SRC-1 association with SRA. (B) Lysates of MCF-7 cells treated with SLIRP siRNA were incubated with no

ab or SRC-1 ab and SRA detected as above. SLIRP was significantly knocked down without affecting b-actin or SRA. Substantially more SRA

copurified with SRC-1 in SLIRP siRNA-treated cells than nonsense controls. (C) Immunoblot confirming SLIRP protein knockdown without

affecting SRC-1.

(D) Schematic of plasmids used in REMSA studies were the following: GST alone, GST-SLIRP (wild-type), GST-SHARP-RRM (SHARP aas 1–608),

and GST-SHARP-RD (SHARP aas 3420–3651).

(E) Binding of recombinant GST-SLIRP to SRA STR7. REMSA demonstrating specific binding of STR7 by GST-SLIRP (lane 2) is reduced with

unlabeled ‘‘cold’’ STR7 (lanes 3 and 4, up to 100-fold excess), but not excess cold pBlue (lanes 5 and 6).

(F) REMSA demonstrating that GST-SLIRP binds SRA STR7 more avidly than SDM7 mutant probe. Binding to both probes reduced following

addition of up to 100-fold excess ‘‘cold’’ STR7 competitor (lanes 3 and 4 and 7 and 8).

(G) STR7 is bound by GST-SHARP-RRM (lane 1), but not with either GST-SHARP-RD (lane 2) or GST alone (lane 3).

Molecular Cell662

To examine the functional specificity of the SLIRP-STR7 interaction in vivo, we utilized the SRA-SDM7 mu-tant (Lanz et al., 2002), in which the stem-loop structureis mutated but preserved (Figure 5G). This mutation de-creased SRA-mediated coactivation to w70% of wild-type levels. Furthermore, when we cotransfected SLIRPwith SRA-SDM7, SLIRP was unable to function as a re-pressor. These data suggested that a direct interactionbetween SLIRP and STR7 is critical for SLIRP’s repres-sive activity.

SLIRP Is Recruited to Endogenous NR TargetPromoters

To determine if SLIRP is recruited to E2- and Dex-re-sponsive promoters, we performed chromatin immuno-precipitation (ChIP) assays. We found that SLIRP wasrecruited to the E2-responsive pS2 promoter within60 min of ligand treatment (Figure 6A), but this associa-tion had returned to undetectable levels by 120 min. ERa

binding increased in response to ligand returning tobasal levels within 120 min. In contrast, HuD, another

Figure 5. SLIRP Is a SRA-Dependent Repressor of NR Transactivation that Interacts with SHARP and SKIP and Requires an Intact RRM for

Coregulation

(A) SLIRP represses ER transactivation, which is augmented by Tam and ICI. HeLa cells were cotransfected with ERE-luciferase (Luc), expres-

sion vectors for ERa 6 SRA, and increasing amounts of SLIRP. After 24 hr, cells were treated for 8 hr with E2 prior to assessment of Luc activity

(normalized to protein). Tam or ICI were added where indicated at the same time as E2. All results are representative of triplicate experiments;

error bars represent standard deviation.

(B) SLIRP represses signaling of multiple NR pathways. HeLa cells were cotransfected with either a GRE-Luc, ARE-Luc, TRE-Luc, VitD-Luc, or

PPARE-Luc reporter plus corresponding AR, TR, VDR, PPARd, SRA, and SLIRP expression vectors and incubated with ligand (Dex, DHT, T3,

VitD, GW501516) for 8 hr, and Luc activity was determined as above.

(C) Targeted reduction of SLIRP expression potentiates GR transcription. HeLa cells were cotransfected with GRE-Luc and siRNA directed

against either SLIRP or a nonsense target. After 48 hr, cells were treated with Dex (8 hr) prior to assessment of Luc activity. Immunoblot con-

firmed reduced endogenous SLIRP expression in SLIRP siRNA-treated cells (lane 1) compared with nonsense (lane 2) relative to b-actin.

(D) SLIRP augments SHARP’s repression and antagonizes SKIP’s coactivation of ER. HeLa cells were cotransfected with ERE-Luc and expres-

sion vectors for ERa alone, and/or empty, SHARP, SLIRP, or SKIP vectors, 6siRNA (nonsense, SLIRP, or SKIP). RT-PCR confirmed reduced

SLIRP and SKIP expression in siRNA-treated cells (right panel).

(E) Plasmids for expression of wild-type and mutant SLIRP with carboxy-terminal FLAG epitope.

(F) Mutation of the RRM domain abrogates binding of SLIRP to SRA. REMSA using labeled SRA STR7 probe and increasing amounts of GST-

SLIRP fusion proteins (wild-type, mutants R24,25A, L62A, or double mutant R24,25A,L62A).

(G) SLIRP mutants have reduced ability to repress ER activity. HeLa cells were transfected with ERE-Luc and either wild-type (SLIRP-FLAG) or

mutated SLIRP-FLAG (R24,25A, L62A, DM) expression vectors together with SRA or SRA SDM7 (stem-loop mutant) and reporter activity as-

sessed as in (A), above.


well-characterized RRM-containing RNA binding pro-tein (Chung et al., 1996), was not recruited to the DNA.These data confirmed that SLIRP can closely associatewith the response element of an E2-regulated gene.

To investigate the mechanism by which SLIRP mightmediate its effect at the transcriptional level, we per-formed ChIP assays in HeLa cells treated with SRAsiRNA. Interestingly, in cells with reduced SRA expres-sion, w50% less SLIRP was recruited to the Dex-re-sponsive metallothionein promoter (Figure 6B, lane 10).This suggests that the presence of SRA is critical forrecruiting SLIRP to the promoter and consequently me-diating SLIRP’s repressive effects.

To investigate interactions of SLIRP with other core-pressors, we performed ChIP studies in cells treatedwith SLIRP siRNA. In the absence of E2, NCoR is re-cruited to the pS2 promoter together with a small amountof ER (Figure 6C). However, in cells with reduced SLIRP,NCoR could no longer be detected on the promoter, andER recruitment was significantly higher. This suggestsa key role for SLIRP in facilitating recruitment of NCoRto the promoter.

SLIRP Is Predominantly MitochondrialBased on our transfection and ChIP data, we envisagedthat SLIRP would be a predominantly nuclear protein.However, imaging studies using the SLIRP ab revealedendogenous SLIRP to have a filamentous cytoplasmicdistribution confined predominantly to the mitochondria(Figure 7A, top panel). A similar pattern was observed us-ing ab to HSP-60, a mitochondrial-specific protein (Fig-ure 7A, second row) (Gupta and Knowlton, 2005). In cellstransfected with FLAG-tagged SLIRP, we found thatSLIRP colocalized with another mitochondrial-specific

protein cytochrome c oxidase (Figure 7A, third row). Se-quence analysis of SLIRP revealed an N-terminal 26 aadomain highly predictive of an amphipathic a-helical mi-tochondrial targeting sequence conserved between themouse, rat, and human genomes (see Figure 2C). Thismitochondrial signal sequence is evident in the 3D pre-dicted structure of SLIRP as an independent helix linkedto the RRM (Figure 7C). To evaluate the importance ofthe N-terminal signal sequence, we compared the intra-cellular localization of SLIRP-FLAG versus FLAG-SLIRPconstructs. Interestingly, SLIRP-FLAG localized to themitochondria, whereas FLAG-SLIRP was pancellular(Figure 7A, bottom two rows), consistent with the notionthat the N-terminal mitochondrial signal sequence is crit-ical for targeting SLIRP to the mitochondria.

To further investigate the mitochondrial location of en-dogenous SLIRP, we examined primary human breasttissue with SLIRP and HSP-60 abs. A punctate cytoplas-mic staining pattern, characteristic of mitochondria, wasobserved with both SLIRP and HSP-60 abs (Figure 7B).Taken together, these data confirm that SLIRP residespredominantly in the mitochondria and that interferencewith the N-terminal signal sequence can substantiallyalter the intracellular distribution of the protein.

To evaluate the functional importance of the mito-chondrial signal sequence, we generated a triple mutantof the first three arginines in SLIRP (R7,13,14,A) (see Fig-ure 5F). When cotransfected into HeLa cells with SRA,this SLIRP mutant had reduced ability to function as arepressor (Figure 7D). These data suggested that themitochondrial signal sequence is required for maintain-ing corepressor function, raising the possibility thatSLIRP has bifunctional capacity as a NR corepressorin the nucleus and mitochondria.

Figure 6. SLIRP Is Recruited to Endogenous

Promoters and Modulates NCoR Recruit-

ment

(A) ChIP assay demonstrating recruitment

of SLIRP and ER, but not HuD, to the pS2

promoter of MCF-7 cells in response to E2.

Sheared, genomic, MCF-7 DNA used as input

control.

(B) Recruitment of SLIRP to the metallothio-

nein promoter is regulated by SRA. HeLa cells

treated with SRA siRNA or nonsense siRNA

(NS siRNA) for 3 days were incubated with

Dex, and then ChIP assays were performed

with either GR, SRC-1, or SLIRP ab (left).

RT-PCR shows knockdown of SRA without

affecting b-actin (right).

(C) SLIRP regulates NCoR association with

the pS2 promoter. MCF-7 cells were treated

with either SLIRP siRNA or NS siRNA (3

days) followed by E2 for 45 min before ChIP

assay using ER or NCoR ab as above (left).

RT-PCR demonstrating SLIRP knockdown

is shown (right).

Molecular Cell664

Discussion

SRA coregulates NR pathways and has been implicatedin tumorigenesis. However, the mechanisms by whichSRA mediates its effects remain to be elucidated. Weidentified SLIRP as a protein that binds to SRA in vitroand in vivo and is a potent repressor of E2,glucocorticoid,androgen, thyroid hormone (T3), and VitD action. In addi-tion, SLIRP represses orphan NR activity, as shown by itseffects on PPARd-mediated transactivation. SLIRP,which is composed almost entirely of an RRM region, iswidely expressed in normal human tissues while elevatedin skeletal muscle, heart, liver, and testis. Furthermore,SLIRP is also widely expressed at the mRNA and proteinlevel in multiple cancer cell lines, and IHC studies confirmits presence in primary human breast tumors.

Here we show that SLIRP interacts specifically withSTR7 in vitro and with endogenous SRA in vivo. Shiet al. (2001) showed that the coregulator activity ofSHARP requires its RRM. We observed similar findingswith SLIRP in that discrete single and double aa substi-tutions of the RRM domain significantly reduced its SRAbinding and corepression activities. We also demon-strate that SHARP (via its RRM domain) interacts withSTR7, raising the possibility that SHARP and SLIRP

may compete for binding to SRA, consequently affect-ing ER-regulated gene expression.

The SRA STR7 stem loop is the longest and one of themost stable identified by 2º structure predictions andaccounts for a substantial proportion of SRA’s overallcoactivator activity (Lanz et al., 2002). Our SRA mutationdata show that STR7 is required for SLIRP to act as a co-repressor, further strengthening the case for a direct in-teraction between STR7 and SLIRP in vivo. Reduction ofendogenous SLIRP expression increases SRA’s coacti-vation ability, suggesting not only that this interactionis functionally relevant, but also that SLIRP could playan important tumor-suppressor role in SRA-activatedNR pathways. The additive repressive effect of SHARPand SLIRP, both of which bind to SRA STR7, suggeststhese proteins could function to significantly downregu-late ER signaling in breast cancer cells.

Complex Interactions between SLIRP and Other NRCoregulators at Hormone-Responsive Promoters

Our studies provide new insight into the mechanism ofinteraction between SRA, SLIRP, and SRC-1. In particu-lar, we found that recruitment of SLIRP to an endoge-nous Dex-responsive promoter is regulated by theamount of SRA in the cell. Furthermore, we have shown

Figure 7. SLIRP Localizes Predominantly to the Mitochondria

(A) Simultaneous mitochondrial (red, Mitotracker), nuclear (blue, Hoescht 33258), and endogenous SLIRP protein staining (green, rabbit poly-

clonal sera, Alexa Fluor 488 secondary ab) of HeLa cells. Overlaying of confocal images reveals colocalization of SLIRP and the mitochondria

(yellow) (top row). Endogenous SLIRP also colocalizes with mitochondria-specific HSP-60 (second row). Transfected SLIRP-FLAG colocalized

with cytochrome c (middle row) and Mitotracker stain (fourth row). Transfected FLAG-SLIRP was pancellular and did not colocalize with the

Mitotracker stain (bottom row).

(B) SLIRP and HSP-60 stain similarly in human breast cancer tissue. IHC of primary human breast cancer tissue using either SLIRP or HSP-60

abs. Ducts stained readily with both abs in a punctate cytoplasmic pattern, consistent with a mitochondrial location for HSP-60 and SLIRP.

(C) Three-dimensional modeling of the SLIRP protein predicts a mitochondrial localization signal in the amino terminal 26 aas. Residues

subjected to point mutation are indicated.

(D) Mutations in the SLIRP mitochondrial sequence relieve its repressive activity. HeLa cells were transfected as in Figure 5A with ERE-luc, ERa,

wild-type SRA, and either wild-type SLIRP or the R7,13,14A mutant. Results are representative of triplicate experiments; error bars represent

standard deviation.


that SRC-1 and SLIRP appear to compete for associa-tion with SRA. Specifically, when SLIRP levels are re-duced, SRC-1 association with SRA increases. This isconsistent with the opposing function of these two cor-egulators: the association of SRC-1 with SRA resultsin coactivation, while association of SLIRP with SRA re-sults in corepression.

Our ChIP data in cells with reduced SLIRP expressionprovide additional mechanistic insights. The results sug-gest that SLIRP is essential for mediating NCoR’s asso-ciation with the promoter in the absence of E2. Mostinterestingly, in SLIRP siRNA-treated cells, not only isbinding of NCoR abrogated, but the ER is strongly re-cruited, suggesting that removal of SLIRP from thecell alters the promoter state from one of repression toactivation.

We were intrigued to observe that SKIP and SLIRP co-localize to human Chr 14q24.3 and that this genomic dis-tribution is conserved across species. It raised the pos-sibility that they may be coordinately regulated, as is thecase for Grb7 and HER2 that lie adjacent to each otheron human Chr 17 (Daly, 1998). However, our expressiondata did not support this idea. Furthermore, our trans-fection and siRNA studies suggested that they have op-posing and possibly competitive effects on estrogensignaling rather than working in concert. Given the inter-action between SKIP and NCoR in VDR transactivation(Leong et al., 2004), our studies with SKIP and NCoR sug-gest a complex role for SLIRP in modulating VDR signal-ing. Although loss of heterozygosity (LOH) has beendescribed at Chr 14q31.2 in breast tumors (Martin et al.,2001; O’Connell et al., 1999), LOH in the genomic areaof SLIRP/SKIP in breast cancer has not been described.

SLIRP Is a Predominantly Mitochondrial NR

CorepressorThe increased expression of SLIRP in high-energy de-mand and mitochondria-rich tissues such as skeletalmuscle, heart, and liver is consistent with its predomi-nantly mitochondrial location. Multiple imaging studiessuggest that more than 90% of SLIRP is located in themitochondria, raising the possibility that it may functionboth in the nucleus and mitochondria to regulate NR ac-tivity. Whether this is via interactions with SRA or othermitochondrial RNA targets is unknown. In addition,SLIRP’s capacity to represses PPARd-mediated signal-ing suggests a potential role in regulating lipid homeo-stasis in energy-rich tissues.

The role of NRs in the mitochondrion, affecting cellsurvival and energy homeostasis, has recently come un-der close scrutiny. Studies in breast cancer tissues sug-gest that mitochondrial ER plays a role in tumor cell sur-vival (Pedram et al., 2006). Additionally, T3 can inducetranscription in the absence of nuclear factors acting viamitochondrial TR (Scheller and Sekeris, 2003). GR inter-acts with NFkB subunits (Tao et al., 2001) present in themitochondria (Cogswell et al., 2003), and putative GREsexist in some key components of the mitochondrialgenome-encoded oxidative phosphorylation pathway(Psarra et al., 2006). Taken together, these data providea foundation for NR and coregulator action in the mito-chondria and a rationale for SLIRP’s presence there.

The discovery of SRA, the first RNA coactivator, led toa paradigm shift in our understanding of NR coregula-

tion and hormone action. With the identification ofSLIRP, a new SRA binding protein, we provide the mostdetailed characterization of a direct SRA-protein inter-action to date. The expression of SLIRP in a variety ofcancers, its functional corepression of multiple NR sig-naling pathways, and its capacity to regulate NCoR pro-moter recruitment and SRC-1 association with SRA aswell as interactions with SHARP and SKIP suggest thatSLIRP may play an important role in regulating a broadrange of NR activities and therefore potentially tumori-genesis. Moreover, its expression in energy-rich tissues,mitochondrial location, and repression of PPARd-medi-ated signaling suggest that SLIRP may participate incontrolling lipid and energy metabolism, with roles inboth the nucleus and mitochondria. Further studies toelucidate the role of SLIRP in each of these cellular loca-tions should contribute substantially to the biology un-derlying hormone-dependent tumor growth and thecontrol of body metabolism.

Experimental Procedures

Cell Culture

MCF-7,MDA-MB-468, SK-BR-3, T47D, LNCaP, PC-3,HepG2,Calu-6,

HT1080, COS-7, HeLa cell lines obtained from the American Type

Culture Collection, HMEC from Dr. Roger Reddel. Cells were grown

as per supplier’s recommendations or as previously described (Giles

et al., 2003) and used within 20 passages of original stock for all

experiments.

Yeast Three-Hybrid Screen of Human Breast Cancer Library

Screening and plasmid isolation protocols were as described (Sen-

Gupta et al., 1996). Analysis of SRA and STR7 structures was per-

formed using Mfold (Zucker, 2003). pIIIA/MS2-2 and S. cerevisiae

L40coat were gifts of Dr. Marvin Wickens. Dr. Jennifer Byrne do-

nated the human breast cancer cDNA library (Byrne et al., 1998). Col-

onies underwent stringency tests as described (Park et al., 1999).

SLIRP Plasmid Constructs, Sequences, and Vectors

SLIRP coding domain was amplified from the yeast three hybrid

clone with SLIRP (sense) 50 cgc gga tcc gcg gcc tca gca gca 30,

SLIRP (reverse) 50 gcg cgg atc cta ggc tgc agt ctca 30 primers and

subcloned into BamHI cut pCMV-FLAG 7.1 for transfection assays

and pGEX-6P2. STR7 oligonucleotides (sense) 50 agg agg cag gta

tgt gat gac atc agc cga cgc ctg gca ctg ctg cag gaa cag tgg gct

gga gga aag ttg tca ata cct gta aag aa 30 and (reverse) 50 ttc ttt aca

ggt att gac aac ttt cct cca gcc cac tgt tcc tgc agc agt gcc agg cgt

cgg ctg atg tca tca cat acc tgc ttc ct 30 were cloned into EcoRV-di-

gested pBluescript II KS+ (Stratagene) to generate labeled ribo-

probes.Togenerate SRASDM7, underlined residues above weremu-

tated to tcc, ctc, and ctc as outlined (Lanz et al., 2002). pBluescript

vectors were linearized and riboprobes generated as described

(Thomson et al., 1999). SLIRP mutants (R24,25A and L62A) were pre-

pared by PCR-based mutagenesis and subcloned into BamHI-cut

pCMV-FLAG 7.1 and pGEX-6P2 as above. Sequence alignments of

SLIRP and SHARP and SLIRP interspecies comparisons were per-

formed using algorithms described by Stothard (2000).

REMSA and UVXL

REMSA and UVXL were performed as described (Thomson et al.,

1999). In competition REMSA, up to 100-fold excess unlabeled

pBlue, STR7, or tRNA was used. Large-scale recombinant GST pro-

tein expression was performed as described (Giles et al., 2003).

Recombinant proteins were digested with PreScission Protease

(Sigma).

Transfection and Luciferase Assays

HeLa cells were transfected in RPMI containing 5% stripped serum

using FuGENE6 (Roche) with equal molar ratios of control empty

and or cDNA expression plasmids plus the following: 50 ng/well

ERa, TRb, AR, PPARd; 0.6 mg/well ERE-Luc, GRE-Luc, TRE-Luc,

Molecular Cell666

ARE-Luc, VDRE-Luc, PPARE-Luc; 0.0625 mg/well pSCT or 0.08 mg/

well pSCT-SRA, pSCT-SRA-SDM7a; 0.049 mg/well pCMV-FLAG-

7.1 or 0.05-0.5 mg/well FLAG-SLIRP, FLAG-SLIRP-R24,25A, FLAG-

SLIRP-L62A, FLAG-SLIRP-DM; 0.3 mg/well pCMX or 0.5 mg/well

pCMX-SHARP; 100 ng/well pCGN or pCGN-SKIP. After transfection,

cells were cultured for 24 hr prior to addition of E2, Dex, DHT (10 nM),

T3 (1 nM), VitD (100 nM), GW501516 (500 nM, Alexis Biochemicals),

Tam (1 mM), ICI (1 mM, Tocris), or equal volume of absolute ethanol or

DMSO. After 8 hr, lysates assayed for luciferase activity relative to

protein levels using Luciferase (Promega) and Protein Assays (Bio-

Rad) on a FluorStar Optima (BMG).

RNA Interference

Cells were transfected with siRNA complexes directed against SRA,

SLIRP, SKIP, or control (nonsense) (Dharmacon RNA Technologies,

final concentration 20–200 nM) using Lipofectamine 2000 (Invitro-

gen). To assess SLIRP knockdown on GRE-luc activity, reporter

plus pSCT or pSCT-SRA and siRNA conjugates were added simulta-

neously. After 48 hr, cells were induced with Dex (10 nM) and lysates

processed as described above. For SLIRP and SKIP siRNA studies,

cells were transfected with siRNA for 72 hr, ERa and ERE-Luc co-

transfected at 48 hr, and E2 (10 nM) added 8 hr prior to harvest.

SKIP (sense) 50-cat tca act ctg gag cta aac aga-30 and (reverse) 50-

tcc gat cag caa tgt aga ggg ct-30; SLIRP (sense) 50-gcg ctg cgt aga

agt atc aa-30 and SLIRP (reverse) 50-tcg att ccg aag tcc ttc tt-30;

and b-actin (sense) 50-gcc aac aca gtg ctg tct gg-30 and b-actin (re-

verse) 50-tac tcc tgc ttg ctg atc ca-30 primers were used to quantitate

targets by RT-PCR.

Northern Analysis

Human tissue blots obtained from BD Biosciences (7759-1, 7760-1).

Total RNA isolated using TRIzol LS (GIBCO-BRL), Poly A RNA using

Poly ATtract (Promega). Samples (2 mg) were electrophoresed and

transferred to Hybond-N membrane (Amersham). Human SLIRP,

SKIP, or b-actin cDNAs were random primer labeled ([32P]dCTP)

and hybridized with membranes for 2–5 hr, washed to 0.1 3 SSC/

0.1% SDS at 65ºC, and visualized by PhosphoImager.

Immunoblotting

Cell lysates were resolved by SDS-PAGE and transferred onto PVDF

Membrane (Roche). SLIRP was detected with rabbit polyclonal anti-

sera raised against GST-SLIRP protein (1:250). SRC-1 (SC6096),

SKIP, ERa (SC7207), and b-actin abs (Abcam ab 6276) were used

with HRP conjugated anti-rabbit or anti-mouse secondary abs and

ECL Plus (Amersham).

Immunohistochemistry

Full-face sections of primary human breast cancer tissue were

immunostained using standard protocols. SLIRP polyclonal ab or

HSP-60 ab (SC13115) was used at 1:2500, with biotinylated goat

anti-rabbit and anti-mouse ab (Chemicon IHC Select) followed by

streptavidin-horseradish peroxidase as secondary and tertiary

reagents. Sections were vizualised with diaminobenzidine (DAKO)

followed by a light counterstain with haematoxylin.

Immunoprecipitation RT-PCR Assay

Method was as described (Giles et al., 2003). Using MCF-7, MDA-

MB-468, or HeLa cells with either SLIRP ab, SRC-1 ab, b-actin ab,

or no ab, coimmunopurifying SRA detected using the following

primers: SRAEIV(sense), 50-tga tga cat cag ccg acg cct-30 and

SRAEIV(reverse) 50-gct gca gat ttc tct tca ttg-30. SLIRP and b-actin

mRNAs were detected using primers as above.

ChIP Assay

Assay was as described (Dowhan et al., 2005). MCF-7 and HeLa cells

were treated with 100 nM E2 for 0-120 min or Dex 100 nM for 15 min

prior to fixation. Soluble chromatin incubated with 4 mg ERa, SLIRP

polyclonal, HuD (SC5979), SRC-1, NCoR, or GR (a 50:50 combina-

tion of SC1002 and SC1004) abs. Recovered DNA fragments were

amplified with either pS2 (sense) 50-ggc cat ctc tca cta tga atc act

tct gc-30 and (reverse) 50-ggc agg ctc tgt ttg ctt aaa gag cg-30 or met-

allothionein (MTA2) (sense) 50-act cgt ccc ggc tct ttc ta-30 and (re-

verse) 50-agg agc agt tgg gat cca t-30primers.

Imaging Studies

HeLa cells cultured on glass cover slips were incubated with Mito-

Tracker (Molecular Probes), SLIRP, HSP-60, or cytochrome c

(SC7159) ab and Alexa Fluor 488 goat anti-rabbit (Molecular Probes)

secondary ab added. Cover slips were bathed in 100 ng/ml Hoechst

33258/PBS, washed, and mounted in Vectashield (Vector Laborato-

ries). For FLAG imaging, cells were transfected 24 hr prior to staining

as above and visualized using a BioRad MRC1000 confocal micro-

scope.

Supplemental Data

Supplemental Data include one figure and can be found with

this article online at http://www.molecule.org/cgi/content/full/22/5/

657/DC1/.

Acknowledgments

The authors express their thanks to Jennifer Byrne (Children’s Hos-

pital, Westmead, Sydney, Australia) for the human breast cancer

cDNA library and relevant clinical details, Xiatao Li (Baylor College

of Medicine) for technical advice, Ronald M. Evans and Michael

Downes (Salk Institute) for the SHARP clones, Gary Leong (Univer-

sity of Sydney, Australia) for the SKIP constructs, Marvin Wickens

(University of Wisconsin) for the yeast three-hybrid system, George

Muscat (Institute for Molecular Bioscience, Queensland, Australia)

for the PPARd and PPARE vectors and advice, Dennis Dowhan (Cen-

tre for Immunology and Cancer Research, Queensland, Australia) for

assistance with the ChIP assay and SKIP ab, Chris Glass (University

of California) for advice on the ChIP assay, Roger Reddel (West-

mead, Sydney, Australia) for the HMEC cells, Paul Rigby (University

of WA, Australia) for his assistance with the imaging studies, and

Evan Ingley (Western Australian Institute for Medical Research

(WAIMR), Australia) for his technical assistance. This work was

funded by National Health and Medical Research Council (NHMRC),

National Breast Cancer Foundation (NBCF), Cancer Council of WA

(CCWA), WAIMR and Royal Perth Hospital Medical Research Foun-

dation grants. E.C.H. is the recipient of a NHMRC Dora Lush Ph.D.

scholarship, and A.D.R. is the recipient of a NBCF Ph.D. scholarship.

Received: July 23, 2005

Revised: March 21, 2006

Accepted: May 19, 2006

Published: June 8, 2006

References

Barry, J.B., Leong, G.M., Church, B., Issa, L.L., Eisman, J.A., and

Gardiner, E.M. (2003). Interactions of SKIP/NCoA-62, TFIIB, and

retinoid x receptor with vitamin D receptor helix H10 residues.

J. Biol. Chem. 278, 8224–8228.

Bouvet, P., Jain, C., Belasco, J., and Amalric, A.F. (1997). RNA rec-

ognition by joint action of two nucleolin RNA-binding domains: ge-

netic analysis and structural modeling. EMBO J. 16, 5235–5246.

Burd, C.G., and Dreyfuss, G. (1994). Conserved structures and diver-

sity of functions of RNA-binding proteins. Science 265, 615–621.

Byrne, J., Nourse, C., Basset, P., and Gunning, P. (1998). Identifica-

tion of homo- and heteromeric interactions between members of the

breast carcinoma-associated D52 protein family using the yeast

two-hybrid system. Oncogene 16, 873–881.

Chooniedass-Kothari, S., Emberly, E., Hamadani, M.K., Troup, S.,

Wang, X., Czosnek, A., Hube, F., Mutawe, M., Watson, P.H., and

Leygue, E. (2004). The steroid receptor RNA activator is the first

functional RNA encoding a protein. FEBS Lett. 566, 43–47.

Chung, S., Jiany, L., Cheng, S., and Furneaux, H.M. (1996). Purifica-

tion and properties of HuD, a neuronal RNA-binding protein. J. Biol.

Chem. 271, 11518–11524.

Cogswell, P.C., Kashatus, D.F., Keifer, J.A., Guttridge, D.C.,

Reuther, J.Y., Bristow, C., Roy, S., Nicholson, D.W., and Baldwin,

A.S.J. (2003). NF-kB and IkBa are found in the mitochondria. Evi-

dence for regulation of mitochondrial gene expression by NF-kB.

J. Biol. Chem. 278, 2963–2968.


http://www.molecule.org/cgi/content/full/22/5/657/DC1/

http://www.molecule.org/cgi/content/full/22/5/657/DC1/

Daly, R.J. (1998). The Grb7 family of signalling proteins. Cell Signal.

10, 613–618.

Dowhan, D., Hong, E., Auboeuf, D., Dennis, A., Wilson, M., Berget,

S., and O’Malley, B.W. (2005). Steroid hormone receptor coactiva-

tion and alternative RNA splicing by U2AF65-related proteins

CAPERa and CAPERb. Mol. Cell 17, 429–439.

Dressel, U., Allen, T.L., Pippal, J.B., Rohde, P.R., Lau, P., and Mus-

cat, G.E.O. (2003). The peroxisome proliferator-activated receptor

b/d agonist, GW501516, regulates the expression of genes involved

in lipid catabolism and energy uncoupling in skeletal muscle. Mol.

Endocrinol. 17, 2477–2493.

Giles, K.M., Daly, J.M., Beveridge, D.J., Thomson, A.M., Voon, D.C.,

Furneaux, H.M., Jazayeri, J.A., and Leedman, P.J. (2003). The 30 un-

translated region of p21WAF1 mRNA is a composite cis-acting

sequence bound by RNA-binding proteins from breast cancer cells,

including HuR and poly(C)-binding protein. J. Biol. Chem. 278, 2937–

2946.

Gupta, S., and Knowlton, A.A. (2005). HSP60, Bax, apoptosis and the

heart. J. Cell. Mol. Med. 9, 51–58.

Herynk, M., and Fuqua, S. (2004). Estrogen receptor mutations in

human disease. Endocr. Rev. 25, 869–898.

Lanz, R.B., McKenna, N.J., Onate, S.A., Albrecht, U., Wong, J., Tsai,

S.Y., Tsai, M.-J., and O’Malley, B.W. (1999). A steroid receptor

coactivator, SRA, functions as an RNA and is present in an SRC-1

complex. Cell 97, 17–27.

Lanz, R.B., Razani, B., Goldberg, A.D., and O’Malley, B.W. (2002).

Distinct RNA motifs are important for coactivation of steroid hor-

mone receptors by steroid receptor RNA activator (SRA). Proc.

Natl. Acad. Sci. USA 99, 16081–16086.

Lanz, R.B., Chua, S.A., Barron, N., Soder, B.M., DeMayo, F., and

O’Malley, B.W. (2003). Steroid receptor RNA activator stimulates

proliferation as well as apoptosis in vivo. Mol. Cell. Biol. 23, 7163–

7176.

Leong, G.M., Subramanian, N., Issa, L.L., Barry, J.B., Kino, T., Drig-

gers, P.H., Hayman, M.J., Eisman, J.A., and Gardiner, E.M. (2004).

Ski-interacting protein, a bifunctional nuclear receptor corepressor

that interacts with N-CoR/SMRT and p300. Biochem. Biophys.

Res. Commun. 315, 1070–1076.

MacDonald, P.N., Dowd, D.R., Zhang, C., and Gu, C. (2004). Emerg-

ing insights into the coactivator role of NCoA62/SKIP in vitamin

D-mediated transcription. J. Steroid Biochem. Mol. Biol. 89–90,

179–186.

Martin, M.D., Fischbach, K., Osborne, C.K., Moshsin, S.K., Allred,

D.C., and O’Connell, P. (2001). Loss of heterozygosity events imped-

ing breast cancer metastasis contain the MTA1 gene. Cancer Res.

61, 3578–3580.

McKenna, N.J., and O’Malley, B.W. (2002). Combinational control of

gene expression by nuclear receptors and coregulators. Cell 108,

465–474.

Murphy, L.C., Simon, S.L., Parkes, A., Leygue, E., Dotzlaw, H., Snell,

L., Troup, S., Adeyinka, A., and Watson, P.H. (2000). Altered expres-

sion of estrogen receptor coregulators during human breast tumor-

igenesis. Cancer Res. 60, 6266–6271.

O’Connell, P., Fischbach, K., Hilsenbeck, S., Mohsin, S.K., Fuqua,

S.A.W., Clark, G.M., Osborne, C.K., and Allred, D.C. (1999). Loss of

heterozygosity at D14S62 and metastatic potential of breast cancer.

J. Natl. Cancer Inst. 91, 1391–1397.

Oliver, W.R., Jr., Shenk, J.L., Snaith, M.R., Russell, C.S., Plunket,

K.D., Bodkin, N.L., Lewis, M.C., Winegar, D.A., Sznaidman, M.L.,

Lambert, M.H., et al. (2001). A selective peroxisome proliferator-ac-

tivated receptor d agonist promotes reverse cholesterol transport.

Proc. Natl. Acad. Sci. USA 98, 5306–5311.

Onate, S.A., Tsai, S., Tsai, M.-J., and O’Malley, B.W. (1995). Se-

quence and characterization of a coactivator for the steroid hor-

mone receptor superfamily. Science 270, 1354–1357.

Park, Y.W., Tan, S.-L., and Katze, M.G. (1999). Differential sensitivity

to 5-fluoro-orotic acid as a screen for bait RNA-independent false

positives in a yeast three-hybrid system. Biotechniques 26, 1102–

1106.

Pedram, A., Razandi, M., Wallace, D.C., and Levin, E.R. (2006). Func-

tional estrogen receptors in the mitochondria of breast cancer cells.

Mol. Biol. Cell 17, 2125–2137. Published online February 22, 2006.

10.1091/mbc.E05-11-1013.

Psarra, A.-M.G., Solakidi, S., and Sekeris, C.E. (2006). The mitochon-

drion as a primary site of action of steroid and thyroid hormones:

presence and action of steroid and thryoid hormone receptors in

mitochondria of animal cells. Mol. Cell. Endocrinol. 246, 21–33.

Robyr, D., Wolffe, A.P., and Wahli, W. (2000). Nuclear hormone re-

ceptor coregulators in action: diversity for shared tasks. Mol. Endo-

crinol. 14, 329–347.

Scheller, K., and Sekeris, C. (2003). The effects of steroid hormones

on the transcription of genes encoding enzymes of oxidative phos-

phorylation. Exp. Physiol. 88, 129–140.

SenGupta, D.J., Zhang, B., Kraemer, B., Pochart, P., Fields, S., and

Wickens, M. (1996). A three-hybrid system to detect RNA-protein

interactions in vivo. Proc. Natl. Acad. Sci. USA 93, 8496–8501.

Shi, Y., Downes, M., Xie, W., Kao, H.-Y., Ordentlich, P., Tsai, C.-C.,

Hon, M., and Evans, R.M. (2001). Sharp, an inducible cofactor that

integrates nuclear receptor repression and activation. Genes Dev.

15, 1140–1151.

Stothard, P. (2000). The sequence manipulation suite: JavaScript

programs for analyzing and formatting protein and DNA sequences.

Biotechniques 28, 1102–1104.

Tao, Y., Williams-Skipp, C., and Scheinman, R.I. (2001). Mapping of

gluticorticoid receptor DNA binding domain surfaces contributing to

transrepression of NF-kB and induction of apoptosis. J. Biol. Chem.

276, 2329–2332.

Thomson, A.M., Rogers, J.T., Walker, C.E., and Leedman, P.J.

(1999). Optimized RNA gelshift and UV cross linking assays for char-

acterization of cytoplasmic RNA-protein interactions. Biotech-

niques 27, 1032–1042.

Wantanabe, M., Yanagisawa, J., Kitagawa, H., Takeyama, K.-i.,

Ogawa, S., Arao, Y., Suzawa, M., Kobayashi, Y., Yano, T., Yoshi-

kawa, H., et al. (2001). A subfamily of RNA-binding DEAD-box

proteins acts as an estrogen receptor a coactivator through the

N-terminal activation domain (AF-1) with an RNA coactivator, SRA.

EMBO J. 20, 1342–1352.

Zucker, M. (2003). Mfold web server for nucleic acid folding and

hybridization prediction. Nucleic Acids Res. 31, 3406–3415.

Accession Numbers

The GenBank accession number for the SLIRP sequence reported in

this paper is AY860853.

Molecular Cell668

Insight into estrogen action in breast cancer via the study of a … · Andrew Redfern, Lisa Stuart, Andrew Barker, Louisa MacDonald, Dominic Voon and Keith Giles, who between them

Documents