Examination of the Role of the Hu Proteins, HuR and HuD, in Prostate Cancer Cells.
Christin Florence Down
This thesis is presented for the degree of Doctor of Philosophy
to the
School of Medicine and Pharmacology, University of Western Australia.
January 2007
i Declaration.
This thesis contains published work and/or work prepared for publication, some of which has been co-authored . The bibliographic details of the works and where they appear in the thesis are set out below. (The candidate must attach to this declaration a statement detailing the percentage contribution of each author to the work. This must been signed by all authors. Where this is not possible, the statement detailing the percentage contribution of authors should be signed by the candidate’s Coordinating Supervisor).
This thesis contains experimental data from the following co-authored work: Hu Proteins are Expressed and Regulate Androgen Receptor Expression and Cell Proliferation in Prostate Cancer. Christin F Down , Dianne J Beveridge, Michael R Epis, Ricky Lareu, Lisa M Stuart, Britt Granath, Dominic C Voon, Henry Furneaux, Cecily Metcalf, Jacqueline Bentel and Peter J Leedman. Submitted to Cancer Research .
The relative contribution of authors is as follows: Christin Down (the Candidate): 55% Dianne J Beveridge 6 % Michael R Epis 6 % Ricky Lareu 3 % Lisa M Stuart 3 % Britt Granath 3 % Dominic Voon 3 % Henry Furneaux 3 % Cecily Metcalf 3 % Jacqueline Bentel 5 % Peter J Leedman 10 %
Data from the manuscript described above are presented in this thesis in the following figures, and were performed by the investigator indicated:
Figure 3.1A : Dr Cecily Metcalfe, Department of Pathology, Royal Perth Hospital.
Figure 3.1C: Dr Ricky Lareu, School of Medicine and Pharmacology, University of Western Australia.
Figure 3.2A : Britt Granath, School of Medicine and Pharmacology, University of Western Australia.
Figures 3.2C and 3.4B : Dr Dominic Voon, School of Medicine and Pharmacology, University of Western Australia.
Figure 3.3: Dianne Beveridge, School of Medicine and Pharmacology, University of Western Australia.
Figure 6.10: Michael Epis, School of Medicine and Pharmacology, University of Western Australia.
Professor Peter J Leedman ………………………………
Christin Down ……………………………..
ii Thesis Summary .
Prostate cancer is a significant cause of morbidity and mortality in the male population, and despite its high prevalence, there have been few advances in prostate cancer management during the past 20 years. The primary treatment for prostate cancer centres around inhibition of androgen signalling, halting tumour growth through down-regulation of important proliferative genes under control of the androgen receptor (AR). Frequently, prostate cancer progresses to an androgen-independent state that is resistant to AR-targeting therapies and is characterised by persistent AR expression and re-expression of AR- regulated genes. Understanding the regulation of AR in prostate cancer is therefore of paramount importance for the development of new treatments to control advanced disease. The AR is regulated post-transcriptionally at the level of mRNA stability, and a UC-rich cis -element (“ARUC”) within the 3’ untranslated region (3’UTR) of the AR mRNA has been identified, which is bound by an mRNA-stabilising protein, HuR.
HuR is a member of a family of four related mRNA-binding proteins that includes HuD, which is typically expressed in neuronal tissue. In this thesis, HuD expression was examined immunohistochemically in a cohort of 12 prostate specimens and was detected at low levels in benign prostatic hyperplasia (BPH), with up-regulated expression in prostate cancers, in particular in high-grade tumours. HuD mRNA expression was also confirmed in the LNCaP prostate cancer cell line using reverse-transcription (RT)-PCR. Like HuR, HuD binds U-rich mRNA and RNA electrophoretic mobility shift assays (REMSA’s) confirmed HuD binding to the UC-rich ARUC motif in the AR mRNA 3’UTR, an association that could be disrupted by introduction of point mutations into the ARUC motif. In addition, the AR mRNA was found to co-immunoprecipitate with HuD and HuR from LNCaP cell lysates in immunoprecipitation RT-PCR (IP-RT-PCR) experiments, suggesting that both of these proteins are associated with the AR mRNA in vivo.
The ubiquitously expressed HuR protein was also detected in both normal and malignant prostate tissues using immunohistochemistry, as well as in the nuclei of LNCaP and 22Rv1 prostate cancer cells. As HuR was highly
iii expressed in these cell lines, effects of HuR knockdown on AR expression and activity was examined. Reduction of HuR was associated with a decrease in AR protein levels, evaluated using western blotting, however an effect on AR transcriptional activity could not be identified in androgen-responsive luciferase reporter assays. Quantitative PCR (qPCR) of actinomycin D treated cells was used to demonstrate that the half-life of the AR mRNA was significantly decreased following siRNA-mediated knockdown of HuR levels. Interestingly, reduction of HuR expression in 22Rv1 cells was also associated with a significant increase in cell proliferation.
To determine mRNA targets of HuR which may be important in prostate cancer proliferation or progression, an oligonucleotide micro-array analysis was performed on 22Rv1 cells that had been treated with siRNA directed against HuR. This identified a number of growth-associated signalling pathways in which multiple genes were down-regulated, including the mitogen-activated protein kinase (MAPK), focal adhesion, vascular endothelial growth factor (VEGF) and Wingless type/MMTV integration (Wnt) signalling pathways, as well as ubiquitin-mediated proteolysis. Further qPCR and IP-RT-PCR analysis of a subset of the regulated genes that contained 3’UTR HuR consensus binding sites confirmed mRNA down-regulation following HuR siRNA treatment and in vivo interaction with HuR of the CREBBP, PCAF, ELK1 and PAK2 transcripts.
In summary, results presented in this thesis have identified aberrant up- regulation of expression of the RNA-binding protein HuD in prostate cancer, which may contribute to the maintenance of AR expression via binding and stabilisation of the AR mRNA. HuR has been identified as an important protein for the control of AR gene expression and prostate cancer cell proliferation. Furthermore, HuR regulates expression of a range of key genes involved in transcription and cellular signalling, implicating it as an important potential target for therapeutics.
iv Table of Contents.
Declaration…………………………………………………………………………….ii Thesis Summary……………………………………………………………………..iii Table of Contents…………………………………………………………………….v Acknowlegements……………………………………………………………………x Awards Received…………………………………………………………………….xi Publications and Presentations Arising from this Thesis……………………….xii Index of Figures…………………………………………………………………….xiv Index of Tables…………. …………………………………………………………xvi Abbreviations……………………………………………………………………….xvii 1 Chapter 1: Introduction...... 2 1.1 The Prostate and Prostate Cancer...... 2 1.1.1 Structure and Function of the Prostate Gland...... 2 1.1.2 Androgen Effects on Prostate Tissue...... 4 1.1.3 Prostate Cancer...... 5 1.1.3.1 Treatment of Prostate Cancer...... 6 1.2 The Androgen Receptor...... 8 1.2.1 Structure...... 8 1.2.2 AR Activity...... 10 1.2.3 Regulation of Androgen Receptor Transcriptional Activity...... 11 1.2.4 Regulation of Androgen Receptor Protein Levels...... 15 1.3 The Androgen Receptor in Prostate Cancer...... 16 1.3.1 AR Gene Mutations...... 17 1.3.2 AR Gene Amplification...... 18 1.3.3 AR Co-Activators in Prostate Cancer...... 18 1.3.4 Cell Line Models of Prostate Cancer...... 22 1.4 Post-transcriptional Regulation of mRNA...... 22 1.4.1 Structural and Sequence Motifs in mRNA Processing and Turnover...... 23 1.4.2 RNA-Binding Proteins in Post-Transcriptional Regulation...... 31 1.4.3 Hu/ELAV Family of RNA Binding-Proteins...... 33 1.4.3.1 Hu Binding Sites...... 38 1.4.3.2 Modulation of Hu Protein Activity...... 39 1.4.4 Shared Targets, Different Outcomes...... 40
v 1.4.5 Mechanisms of AURE-Mediated Decay...... 43 1.5 Post-Transcriptional Regulation of the Androgen Receptor...... 44 1.5.1 Androgen Regulation of the AR mRNA Turnover...... 44 1.5.2 Structure of the AR mRNA...... 44 1.6 Summary and Statement of Aims...... 47 1.6.1 Summary...... 47 1.6.2 Hypotheses: ...... 47 1.6.3 Research Aims...... 48 2 Chapter 2: Materials and Methods...... 50 2.1 General Reagents...... 50 2.2 Tissue Culture Materials...... 52 2.3 Antibodies...... 53 2.4 Equipment and Software...... 53 2.5 Plasmid constructs...... 54 2.6 Preparation of Plasmid DNA...... 56 2.6.1 Preparation of E. coli...... 56 2.6.2 Plasmid DNA Extraction...... 56 2.6.3 DNA Sequencing...... 57 2.7 Cell Culture...... 58 2.7.1 Maintenance of 22Rv1 and LNCaP Cells...... 58 2.7.2 Charcoal-Stripped Foetal Calf Serum...... 58 2.7.3 Biochemical Analysis of Foetal Calf Serum...... 59 2.8 Transfections...... 59 2.8.1 Plasmid Transfections...... 59 2.8.2 Small-Interfering RNA Transfection...... 60 2.9 Cell Titration Assays...... 61 2.10 Luciferase Assays...... 61 2.11 mfold Prediction of mRNA Secondary Structures...... 62 2.12 Identification of HuR Binding Sites...... 62 2.12.1 Generation of Base-Frequency Plots...... 63 2.12.2 Calculation of Expected NNUUNNUUU Motif Frequencies...... 63 2.13 Analysis of Protein...... 63 2.13.1 Secreted PSA assays...... 63 2.13.2 Protein Extraction...... 63
vi 2.13.3 Protein Concentration Assay...... 64 2.13.4 Polyacrylamide Gel Electrophoresis of Proteins...... 64 2.13.5 Western Blotting...... 65 2.13.6 Chemiluminescent Detection of Proteins...... 65 2.13.7 Analysis of Western Blots...... 65 2.14 Preparation of RNA...... 66 2.14.1 RNA Extraction...... 66 2.14.2 RNA Extraction for Micro-Array Analysis...... 66 2.15 Analysis of mRNA expression...... 67 2.15.1 Micro-Array Analysis of HuR Knockdown in 22Rv1 Cells...... 67 2.15.2 cDNA Synthesis...... 69 2.15.3 Primer Design for Quantitative PCR...... 69 2.15.4 Quantitative Real-Time PCR...... 69 2.15.5 Actinomycin D Chase Assay...... 70 2.15.6 Nested PCR...... 72 2.16 Localisation of HuR...... 72 2.16.1 Sample Preparation...... 72 2.16.2 Microscopy...... 73 2.17 Analysis of RNA-Protein Interactions...... 73 2.17.1 Purification of GST fusion proteins...... 73 2.17.2 Riboprobe Synthesis...... 73 2.17.2.1 Linearisation and Purification of Riboprobe Templates...... 74 2.17.3 RNA Electrophoretic Mobility Shift Assay...... 75 2.18 Immunohistochemistry...... 76 2.19 Immunoprecipitation RT-PCR Assay (IP-RT-PCR)...... 76 2.20 Statistical Analysis...... 77 3 Chapter 3: Analysis of HuD Expression and Activity in Prostate Cancer . 79 3.1 Introduction...... 79 3.2 Results...... 81 3.2.1 HuD is Expressed in Human Prostate Tumours...... 81 3.2.2 HuD Binds the AR mRNA...... 83 3.2.3 Mutations in the ARUC Region Destabilise mRNA...... 85 3.2.4 Effects of HuD Expression and DHT on Luc-ARUC Function.... 90 3.2.5 HuD Effects on AR Expression and Activity...... 92
vii 3.3 Discussion...... 95 4 Chapter 4: The Regulation of AR by HuR...... 103 4.1 Introduction...... 103 4.2 Results...... 104 4.2.1 Characterisation of AR in 22Rv1...... 104 4.2.2 HuR Effects on AR Protein Expression and Activity...... 107 4.2.3 Effects of Reduced HuR Expression on AR Protein Levels and Activity...... 110 4.2.4 Localisation of HuR...... 116 4.2.5 HuR Regulation of Cell Proliferation of 22Rv1 Cells...... 120 4.3 Discussion...... 123 5 Chapter 5: Micro-Array Analysis of 22Rv1 Cells with Reduced HuR. .... 133 5.1 Introduction...... 133 5.2 Results...... 138 5.2.1 Identification of Genes with Altered Expression Following Knockdown of HuR...... 138 5.2.2 Genes Regulated by HuR Cluster in Growth-Associated Pathways...... 140 5.2.3 Effects of Reduced HuR on Genes Involved in Ubiquitin- Mediated Proteolysis...... 151 5.3 Discussion ...... 154 6 Chapter 6: Identification of Potential HuR Targets...... 166 6.1 Introduction...... 166 6.2 Results...... 167 6.2.1 Selection of a Consensus HuR Binding Motif...... 167 6.2.2 Identification of Potential Novel mRNA Targets of HuR...... 169 6.3 Discussion...... 185 7 Chapter 7: Discussion and Future Directions...... 196 7.1 General Discussion...... 196 7.2 Future Directions...... 202 7.3 Summary...... 207 8 References...... 209 Appendix 1: Buffers and Solutions ……………………………………………...234 Appendix 2: Plasmid Vectors…………………………………………………….244
viii Appendix 3: HuR-Regulated Genes……………………………………………..251 Appendix 4 Binary Logarithm (log2) Transformation……………………..…...266
ix Acknowledgements.
I thank my supervisors, Professor Peter Leedman and Dr Jacqueline Bentel, for the time they have spent guiding me through this project and developing permanent visual impairment reading this thesis.
I also acknowledge the assistance of Lisa Stuart, for generating the Luc- ARUC-MutA and –MutB plasmids, Diana Azzam and Michael Epis for assistance with qPCR, Violet Peeva for advice with micro-array preparation and analysis, Dr Paul Rigby for assistance with confocal microscopy, Dr Andrew Barker for help with radioactive probes, Dianne Beveridge for performing a key experiment in my absence and giving me hope, and co- workers on the 5 th and 6 th floors of the MRF building whose advice I have sought during the course of my studies. Thanks also to all the members of the Bentel lab, who adopted me during the last days of my thesis-writing extravaganza, and to members of the Leedman lab, for services rendered that I have otherwise neglected to mention.
Finally, I am deeply grateful for the unwavering friendship and support of my family, and of my fellow travellers, Kay Down, Esme Hatchell, Viki Russell and Rohan Lowe.
x
Awards Received:
Invitrogen Encouragement Award, Australian Society of Medical Research Symposium 2006, Perth, Western Australia, Australia.
Women in Endocrinology Travel Award, ENDO 2005 San Diego, California, United States of America.
Australasian Women in Endocrinology Merit Award, ENDO 2005 San Diego, California, United States of America.
Endocrine and Reproductive Biology Society of Western Australia Abstract Award, Endocrine Society of Australia 2005, Perth, Western Australia, Australia.
Royal Perth Hospital Young Investigator Encouragement Award, Royal Perth Hospital Young Investigators’ Research Symposium 2005, Royal Perth Hospital, Perth, Western Australia, Australia.
Australian and New Zealand Society for Cell and Developmental Biology/ Biochemical Journal Poster Prize, ComBio 2004, Perth, Western Australia, Australia.
Endocrine Society of Australia Travel Award, Endocrine Society of Australia 2004, Sydney, New South Wales, Australia.
National Health & Medical Research Council Biomedical (Dora Lush) Scholarship (2004-2006).
Royal Perth Hospital Medical Research Foundation Doctoral Scholarship (2003)
xi Publications and Presentations Arising from this Thesis:
Christin F Down , Dianne J Beveridge, Michael R Epis, Ricky Lareu, Lisa M Stuart, Britt Granath, Dominic C Voon, Henry Furneaux, Cecily Metcalf, Jacqueline Bentel and Peter J Leedman. Hu Proteins are Expressed and Regulate Androgen Receptor Expression and Cell Proliferation in Prostate Cancer. To be submitted to Journal of Biochemistry .
Christin F Down , Ricky Lareu, Dominic C Voon, Dianne J Beveridge, Britt Granath, Henry Furneaux, Jacqueline Bentel and Peter J Leedman.The Role of the Hu Proteins, HuR and HuD, in the Regulation of Androgen Receptor Expression and Cell Proliferation in Prostate Cancer (2006). Endocrine Abstracts 11 OC15 ECE 2006.
Christin F Down , Ricky Lareu, Dominic C Voon, Dianne J Beveridge, Britt Granath, Henry Furneaux, Jacqueline Bentel and Peter J Leedman. The Role of the Hu Proteins, HuR and HuD, in the Regulation of Androgen Receptor Expression and Cell Proliferation in Prostate Cancer. Presented to Australian Society of Medical Research Symposium 2006, Perth, Australia.
Christin F Down , Ricky Lareu, Dominic C Voon, Britt Granath, Dianne Beveridge, Henry Furneaux, Jacqueline Bentel and Peter J Leedman (2005). The Role of HuR, HuD and αCP1 in the Regulation of Androgen Receptor Expression and Activity in Prostate Cancer Cells. Proceedings of the Endocrine Society of Australia , Endocrine Journal (25) Suppl.: 88.
Christin F Down , Ricky Lareu, Dominic C Voon, Britt Granath, Dianne Beveridge, Henry Furneaux, Jacqueline Bentel and Peter J Leedman. The Role of HuR, HuD and αCP1 in the Regulation of Androgen Receptor Expression and Activity in Prostate Cancer Cells. Presented to Royal Perth Hospital Young Investigators’ Research Symposium 2005, Royal Perth Hospital, Perth, Australia.
xii
Christin F Down , Ricky Lareu, Dominic C Voon, Bu B Yeap, Dianne J Beveridge, Britt Granath, Henry Furneaux, Jacqueline Bentel and Peter J Leedman. The Role of HuR, HuD and αCP1 in the Regulation of Androgen Receptor Expression and Activity in Prostate Cancer Cells. Presented to Endocrine Society of Australia 2005, Perth, Western Australia, Australia.
Christin F Down , Ricky Lareu, Dianne J Beveridge, Britt Granath, Jacqueline Bentel and Peter J Leedman. The Role of the HuR and HuD in the Regulation of Androgen Receptor Expression and Activity in Prostate Cancer Cells. ENDO 2005 San Diego, California, United States of America. Poster 1-38.
Christin F Down , Ricky Lareu, Dominic C Voon, Dianne J Beveridge, Britt Granath, Henry Furneaux, Jacqueline Bentel and Peter J Leedman. The Role of the Hu Proteins, HuR and HuD in the Regulation of Androgen Receptor Expression and Activity in Prostate Cancer Cells. Presented to ComBio 2004, Perth, Western Australia, Australia.
Christin F Down , Ricky Lareu, Dominic C Voon, Bu B Yeap, Dianne J Beveridge, Britt Granath, Henry Furneaux, Jacqueline Bentel and Peter J Leedman. The Role of HuR and HuD in the Regulation of Androgen Receptor Expression and Activity in Prostate Cancer Cells. Presented to Endocrine Society of Australia 2004, Sydney, Australia.
xiii Index of Figures.
Figure: Title: Page: 1.1 Structure and location of the prostate gland. 3 1.2 Prostate tumour architecture: Gleason grading system. 7 1.3 Androgen receptor structure. 9 1.4 Androgen receptor signalling: Three modes of action. 12 1.5 Structural elements of messenger RNA. 25 1.6 Mechanisms of de-adenylation. 27 1.7 Mechanisms of mRNA decay. 28 1.8 Structure of Hu family proteins. 34 1.9 Structural elements of the androgen receptor mRNA. 45 2.1 The sequence of the 3’UTR of Luc-ARUC vector used for mfold structure 62 determination. 3.1 HuD is expressed in primary prostate tumours. 82 3.2 HuD binds the ARUC region of the AR mRNA. 84 3.3 HuD and HuR are closely associated with the AR mRNA in vivo . 86 3.4 Mutations in the ARUC region disrupt HuD binding and luciferase 87 activity. 3.5 Mutations in the ARUC region affect secondary structure. 89 3.6 Effects of HuD expression on luciferase reporter gene activity. 91 3.7 HuD effects on AR protein levels and PSA secretion. 93 3.8 Effects of exogenous HuD expression on AR transcriptional activity. 94 4.1 AR transcriptional activity in LNCaP and 22Rv1 cells. 105 4.2 DHT regulation of AR protein levels in LNCaP and 22Rv1 cells. 106 4.3 Expression of exogenous HuR does not affect AR protein levels. 108 4.4 HuR effects on AR transcriptional activity. 109 4.5 Effectiveness of HuR knockdown is dependent on cell density. 111 4.6 Knockdown of HuR using siRNA decreases AR protein levels. 112 4.7 Effects of siRNA on luciferase activity induced from the PSA-Luc vector. 114 4.8 Effects of HuR knockdown by siHuR-UTR siRNA on AR transcriptional 115 activity and PSA secretion. 4.9 Knockdown of HuR by siRNA significantly decreases AR mRNA half-life 117 in 22Rv1 cells. 4.10 Effects of HuR knockdown by siHuR-UTR siRNA on AR transcriptional 118 activity and PSA secretion. 4.11 Cellular localisation of HuR in DHT- and actinomycin D-treated cells. 119 4.12 siRNA and ActD effects on HuR intracellular localisation in 22Rv1 cells. 121 4.13 Effect of HuR knockdown on 22Rv1 cell proliferation. 122
xiv Index of Figures
Figure: Title: Page: 5.1 Sample image of a scanned oligonucleotide micro-array chip. 135 5.2 Overview of cRNA generation, labelling and hybridisation. 137 5.3 Analysis of mRNA used for oligonucleotide micro-array. 139 5.4 Convergence of MAPK, Wnt and focal adhesion pathways. 152 5.5 Ubiquitin-mediated proteolysis. 160 6.1 Convergence of MAPK and focal adhesion pathways. 174 6.2 Identification of potential HuR binding sites in CREBBP. 175 6.3 Identification of potential HuR binding sites in ELK1. 176 6.4 Identification of potential HuR binding sites in ITGB1. 177 6.5 Identification of potential HuR binding sites in PAK2. 178 6.6 Identification of potential HuR binding sites in PCAF. 179 6.7 Alignment of potential HuR binding sites. 180 6.8 Confirmation of reduction at the mRNA level of PAK2, ELK1, CREBBP 181 and PCAF. 6.9 Relative expression levels of CREBBP, ELK1 PAK2 and PCAF. 183 6.10 Interaction of HuR with the CREBBP, PCAF, PAK2 and ELK1 184 mRNA’s. 6.11 Positions of point mutations in ARUC-MutA and ARUC-MutB 186 sequences.
xv Index of Tables.
Table: Title: Page: 1.1 Co-regulators of the AR and their function in prostate cancer. 14 1.2 Targets of HuR. 36 1.3 Common targets of HuR/HuD and other AURE-binding proteins. 42 2.1 Typical analysis of hormone before and after stripping. 59 2.2 siRNA sequences. 60 2.3 Lysis volumes for protein extraction. 64 2.4 Primers and cycling conditions used for qPCR. 71 5.1 Genes potentially regulated by HuR. 141 5.2 Apoptosis associated genes. 144 5.3 Cell cycle-associated genes. 146 5.4 Proliferation-associated genes. 150 5.5 Genes associated with ubiquitin-mediated proteolysis. 153 6.1 Analysis of recent HuR target 3’UTR sequences. 168 6.2 Analysis of 3’UTRs from de Silanes et al. (2004). 170 6.3 Genes selected for follow-up analysis. 171 6.4 Evaluation of HuR binding sites in 3’UTR’s of selected genes. 172
xvi Abbreviations.
4-dione: androstenedione αCP: α complex protein A: adenosine Ab: antibody ActD: actinomycin D Adiol: androstenediol AI: androgen independence AF: activation function AMP: adenosine monophosphate ANAPC10: anaphase-promoting complex 10 ANOVA: analysis of variation APC/C: anaphase-promoting complex/cyclosome APS: ammonium persulphate AR: androgen receptor ARE: androgen response element ARED2.0: AU-rich element database version 2.0 ART27: AR-trapped clone 27 ARUC: androgen receptor uridine/cytosine rich ATCC: American Type Culture Collection AUF1: ARE/polyU-binding/degradation factor 1 AURE: AU-rich element BAD: BCL2-agonist of cell death BAG1: BCL2-associated athanogene 1 BFAR; bifunctional apoptosis regulator BIRC2: baculoviral IAP repeat containing 2 bp: base-pair BRF1: butyrate response factor 1 BSA: bovine serum albumin C: cytosine C-terminal: carboxyl-terminal cAMP: cyclic AMP CARM1: co-activator-associated arginine methyl transferase 1 CDK: cyclin-dependent kinase CDKN: cyclin-dependent kinase inhibitor cDNA : complementary DNA CEB: cytoplasmic extraction buffer cIAP2: cellular inhibitor of apoptosis cFLIP: inhibitor of Fas/FasL-mediated apoptosis CgA: chromogranin
xvii COX2: cyclo-oxygenase 2 CRE: cyclic-AMP response element CREB: cyclic-AMP response element binding protein CREBBP: cyclic-AMP response element binding protein-binding protein CRM1: chromosome maintenance region 1 cRNA: complementary RNA CT: cycle threshold CUL1: cullin 1 DBD: DNA-binding domain DBP: D-site of albumin promoter-binding protein DEPC: diethyl pyrocarbonate ddH 20: distilled deionised water DHEA: dehydroepiandrosterone DHT: 5 α-dihydrotestosterone DKK: Dickkopf homologue DMSO: dimethyl sulphoxide DNA: deoxyribonucleic acid DNase: deoxyribonuclease dNTP: deoxyribonucleotide triphosphate DOCK1: dedicator of cytokinesis 1 DPC: Diagnostic Products Corporation DSE: downstream element DTT: dithiothreitol DUSP1: dual-specificity phosphatase 1 EAP: EBER-associated protein EBER: Epstein-Barr Early Region EDTA: ethylene diamine tetra-acetic acid EGF: epidermal growth factor eIF: eukaryotic initiation factor ELAV: embryonic lethal abnormal visual ELK1: ETS-like protein kinase 1 ELMO2: elongation and mobility 2 ER: oestrogen receptor ERBB2: erythroblastic leukaemia viral oncogene homologue 2 ERK: extra-cellular receptor activated kinase ETS: E twenty-six EXT2: exostoses 2 FAK: focal adhesion kinase FABP3: fatty-acid binding protein 3 FBP: FUSE binding protein FBXO5: F box protein 5
xviii FCS: foetal calf serum FGF: fibroblast growth factor FGFR1: FGF receptor 1 FUSE: far up-stream element FZD: frizzled G: guanine GAP43: growth-associated protein 43 GAPDH: glyceraldehyde 3-phosphate dehydrogenase GC-MSF: granulocyte-macrophage stimulating factor GDF15: growth and differentiation factor 15 GDP: guanine diphosphate GPC4: glycipan 4 GMP: guanine monophosphate GNL3: guanine nucleotide binding protein-like 5 GnRH: gonadotropin-releasing hormone GO: Gene Ontology GRP: gastrin-releasing peptide GST: glutathione-S-transferase HEPES: N-[2-hydroxyethyl]piperazine-N’-[2-ethane sulphonic acid] HG: human genome HIF1 α: hypoxia inducible factor 1 α hK2: kallikrein 2 hnRNP: heterogeneous ribonucleoprotein HNS: HuR nucleo-cytoplasmic shuttling signal HRP: horseradish peroxidase HSP: heat-shock protein IGF: insulin-like growth factor IGFR1: IGF type I receptor IL: interleukin ILR1: interleukin receptor 1 IRE: iron responsive element ITGB1: integrin beta 1 IP-RT-PCR: immunoprecipitation RT-PCR IPTG: isopropyl β-D- thio-galactopyrosanide JNK: c-jun N-terminal kinase kb: kilobase KEGG: Kyoto Encyclopaedia of Genes and Genomes kDa: kiloDalton KH: hnRNP K homology KSRP: KH-type splicing regulatory protein LARII: luciferase assay reagent
xix LB: Luria Bertani LBD: ligand binding domain LCM: Laboratory for Cancer Medicine LEF1: lymphoid enhancer factor LH: luteinising hormone LOX-DICE: lipoxygenase differentiation control element LSMAF: LotteryWest State Micro-array Facility Luc: luciferase MAP3K: mitogen-activated protein kinase kinase kinase MAPK: mitogen-activated protein kinase MAPKAPK: mitogen-activated protein kinase-activated kinase MARCKS: myristoylated alanine-rich protein kinase C substrate MEK: mitogen/extra-cellular receptor activated kinase MOPS: 3-(N-morpholino)propanesulphonic acid mRNA: messenger RNA mRNP: messenger ribonucleoprotein complex MTA1: metastasis-associated protein 1 MTS: 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulphophenyl)-2H- tetrazolium N: degenerate base symbol for any nucleotide in DNA/RNA N-terminal: amino-terminal NADH: nicotinamide adenosine dinucleotide (reduced) NCBI: National Centre for Biotechnology Information NCoR: nuclear co-repressor NE: neuroendocrine NES: nuclear export signal NET: NaCl EDTA Tris NF1: nucleofibromin 1 NLS: nuclear localisation signal NMD: nonsense-mediated decay NSD: non-stop decay NTP: ribonucleotide triphosphate PABC: cytoplasmic polyA-binding protein PAGE: polyacrylamide gel electrophoresis PAK2: p21-activated kinase 2 PAN: polyA nuclease PAP: polyA polymerase PAP: prostatic acid phosphatase PARC: p53-associated parkin-like cytoplasmic protein PARN: polyA-specific ribonuclease PB: probasin
xx PBS: phosphate-buffered saline PCAF: p300/CREBBP-associated factor PCBP: polyC-binding protein PCR: polymerase chain reaction PEST: proline-, glutamate-, aspartate-, serine- and threonine-rich PIAS: protein inhibitor of activated STAT PLB: passive lysis buffer PMSF: phenylmethylsulphonyl fluoride Pol: polymerase PolyA: polyadenosine PolyC: polycytosine PolyG: polyglycine PolyP: polyproline PolyQ: polyglutamine POU3F2: POU domain, class 3, transcription factor 2 PP2A: protein phosphatase 2A PR: progesterone receptor PSA: prostate-specific antigen PTCH: Patched homologue PTEN: phosphatase and tensin homologue deleted from chromosome 10 PTP: permeability transition pore complex PVDF: polyvinyl difluoride qPCR: quantitative PCR R: degenerate base symbol for guanosine or adenosine RAF1: receptor activating factor 1 RBM5: RNA-binding motif 5 RBP: RNA-binding protein REMSA: RNA electrophoretic mobility shift assay RGC32: response gene to complement 32 RGG: arginine-glycine-glycine RGS2: regulator of G-protein signalling 2 RIN: RNA integrity number RINT1: RAD50-interacting protein 1 RPC: ribonucleoprotein complex RPMI: Roswell Park Memorial Institute RNA: ribonucleic acid Rnase: ribonuclease RRM: RNA-recognition motif rRNA: ribosomal RNA RT: reverse transcriptase S: serine
xxi S: degenerate base symbol for guanosine or cytosine SCF: Skp1-Cullin-Rbx-F box SCLC: small-cell lung cancer SDS: sodium dodecyl sulphate siRNA: small interfering RNA SLC7A1: solute carrier protein family 7 member 1A SMC: structural maintenance of chromosomes SMCL4: structural maintenance of chromosomes-like 4 SMRT: silencing mediator of retinoic acid and thyroid receptor SP1: specificity protein 1 SRC: steroid receptor co-activator SUMO1: SMT3 suppressor of mif two 3 homologue 1 SV40: simian virus 40 T: thymidine TAE: Tris Acetate EDTA TBS: Tris Buffered Saline TBS-T: Tris Buffered Saline-Tween 20 TCF: T-cell factor TCF: ternary complex factor TEMED: tetramethylethyldiamine TGF β: tumour growth factor β TIA1: T-cell intracellular antigen 1 TIAR: TIA-related protein TK: thymidine kinase TNF α: tumour necrosis factor α TTP: tristetraproline tRNA: transfer RNA U: uradine/uracil UDG: uracil-DNA-glycosylase UMP: ubiquitin-mediated proteolysis USE: upstream element UTP: uridine triphosphate UTR: untranslated region UVC: ultraviolet C VEGF: vascular endothelial growth factor W: degenerate base symbol for thymidine, adenosine, or uridine WAIMR: Western Australian Institute for Medical Research Wnt: Wingless/MMTV integration site WT: wild-type WTB: western transfer buffer Y: degenerate base symbol for cytosine or thymidine
xxii
CHAPTER 1: Introduction.
1 1 Chapter 1: Introduction.
1.1 The Prostate and Prostate Cancer. Prostate cancer is the most frequently diagnosed invasive cancer of Australian men, with 11 191 new cases reported in 2001 (Australian Institute of Health and Welfare). In 2000, prostate cancer accounted for 4 % of deaths in the male population (Magnus and Sadkowsky, 2006). Given the prevalence of this disease, there has been little progress in approaches to treatment over the past 20 years, and aggressive late stage disease continues to present a therapeutic obstacle.
1.1.1 Structure and Function of the Prostate Gland.
The prostate is a male-specific sex accessory gland that is situated at the base of the bladder, surrounding the urethra (Figure 1.1A). It is a discrete, encapsulated organ, composed of three zones. The transition zone, closest to the urethra, makes up ~ 5-10 % of prostate glandular tissue, the central zone ~ 25 %, and the peripheral zone the remaining ~ 70 % of the mass of the prostate (Dehm and Tindall, 2006).
The role of the prostate is to secrete prostatic fluid, which comprises about one-third of seminal plasma. Prostatic fluid is rich in proteins, organic solutes, lipids and cholesterol, and these components support the mobility and viability of sperm (Dehm and Tindall, 2006). Three major protein components of prostatic fluid are prostate-specific antigen (PSA), prostatic acid phosphatase (PAP) and kallikrein 2 (hK2). PSA is especially noteworthy, as it can also be detected in the blood of males with prostatic disease. Serum PSA levels are used clinically to monitor responses to prostate cancer treatments and to detect cancer relapse, often before other symptoms arise (Catalona and Loeb, 2005).
Prostatic fluid is secreted from epithelial cells which form glandular acini in the prostate. A basement membrane separates the glandular acini from the supporting stromal cells, which provide physical support as well as essential
2 A Bladder
Seminal vesicle
Ejaculatory duct Prostate
Transition zone Zones of the Central zone prostate Urethra Peripheral zone
B Secretory luminal cells
Basal cells Lumen Neuroendocrine cell
Basement membrane
Neuron
Stroma
Blood vessel
Figure 1.1. Structure and location of the prostate gland. A. Position of the prostate gland at the base of the bladder, surrounding the urethra. Transition, central and peripheral zones of the prostate are indicated. B. Structure of a secretory gland, showing luminal columnar secretory cells ( ), basal cells ( ), neuroendocrine cells ( ), stroma ( ), a blood vessel ( ) and a neurone communicating with a neuroendocrine cell ( ).
A. Reproduced from: http://www.prostate-cancer.org.uk/info/prostate_is.asp. B. Reproduced from Hansson and Abrahamsson (2001)
3 growth factors to the glandular epithelium (Section 1.1.2). There are three epithelial cell types in the acini: luminal columnar secretory cells that line the glandular acini, basal cells that sit against the basement membrane, and rare neuroendocrine cells which are located amongst the basal cells (Figure 1.1B). Neuroendocrine cells are terminally differentiated paracrine/endocrine secretory cells, which are thought to be involved in prostate development and in regulating secretory processes in the mature organ (Shariff and Ather, 2006).
1.1.2 Androgen Effects on Prostate Tissue.
Prostate function in the adult is critically dependent on androgens. Androgens are steroid hormones and include the testicular androgen testosterone and its biologically active metabolite, 5 α-dihydrotestosterone (DHT), as well as the less abundant adrenal androgens, including dehydroepiandrosterone (DHEA), which is metabolised to androstenedione (4-dione) and androstenediol (adiol) in prostate epithelial cells (Mo et al. , 2006). Androgen action is mediated by the androgen receptor (AR), a nuclear receptor that regulates pro-proliferative and anti-apoptotic signalling pathways in the prostate (Section 1.2). These pathways are critical both for development and maintenance of the normal prostate, and for prostate tumour cell proliferation. Withdrawal of androgens causes prostate involution and the apoptosis of columnar epithelial cells, while the stromal and basal cells are relatively unaffected (Heinlein and Chang, 2004). Re-administration of androgens results in rapid proliferation and differentiation of cells of the basal compartment. This effect is mediated directly by the AR expressed in epithelial cells, and indirectly via the stromal cells, which secrete paracrine factors supporting survival of epithelial cells (Dehm and Tindall, 2006). Most of these paracrine factors are up-regulated following androgen administration, and include the potent mitogens epidermal growth factor (EGF), insulin-like growth factor (IGF) and fibroblast growth factor (FGF) (Gnanapragasam et al. , 2000). Thus, the main function of AR in the prostate is to maintain the differentiated secretory function of epithelial cells.
It has been difficult to determine which genes are responsible for cell cycle arrest after androgen withdrawal, in part because of the different effects of 4 androgens when administered at various concentrations in experimental models. For example, in the AR-positive prostate cancer cell line LNCaP, 0.1 nM DHT is pro-proliferative, but many known androgen-induced genes are not up-regulated until the cells are exposed to 1 nM DHT, while 10 nM DHT, which inhibits proliferation, is the concentration at which differentiation occurs (Dehm and Tindall, 2006). Despite this difficulty, it has been suggested that inhibitor of Fas/FasL-mediated apoptosis (cFLIP) (Gao et al. , 2005), the cyclin- dependent kinase inhibitor (CDKI) p21 (Lu et al. , 1999), and Mdm2, a negative regulator of the tumour suppressor p53 (Nantermet et al. , 2004) are up- regulated by androgens, while Hox5a and Egr1 (positive regulators of the tumour suppressor, p53) (Nantermet et al. , 2004), as well as the pro-apoptotic protein caspase 2 (Rokhlin et al. , 2005), are down-regulated by androgens. All of these genes contain androgen response elements (ARE) in their promoters, and as such are direct androgen targets (Section 1.2.2).
It is estimated that ~ 30 % of genes up-regulated by androgens are involved in the synthesis or processing of secretory proteins that are involved in prostate function (DePrimo et al. , 2002; Nantermet et al. , 2004). Three of the best- characterised androgen-responsive genes are the secreted proteins PAP, hK2, and PSA (Section 1.1.1). Overall, androgens are thought to induce and repress similar proportions of genes, although most androgen-responsive genes are not well-characterised (Nantermet et al. , 2004).
1.1.3 Prostate Cancer.
Prostate cancer arises in the prostatic epithelium (Section 1.1.1), with more than 70 % of tumours forming in the peripheral zone, and ~ 20 % in the transition zone. Tumours arising in the central zone are relatively rare (Dehm and Tindall, 2006). Broadly speaking, prostate cancer can be divided into three categories, local/organ-confined disease, early metastatic disease that is still dependent on androgenic hormones, and late-stage metastatic disease that is independent of androgens. Prostate carcinomas are graded by the Gleason system, a histological description of the level of differentiation in the tumour. A tumour of Gleason grade 1 or 2 is similar in architecture to normal tissue, while a Gleason grade of 4 describes significant loss of glandular structure. Gleason grade 5 tumours exhibit complete loss of differentiation 5 (Figure 1.2). Combining the Gleason grades of the two most prevalent architectural types within the specimen ( i.e. the Gleason score) gives a better correlation with patient outcome than single Gleason grades (Gleason and Mellinger, 1974). Therefore, Gleason scores for patients are quoted as being between 2 and 10, and are commonly used in clinical practice.
1.1.3.1 Treatment of Prostate Cancer. Organ-confined prostate cancer may be treated by surgical removal of the prostate (radical prostatectomy) or irradiation, or there may be no intervention (“watchful waiting”), where serum PSA levels are monitored and therapeutic intervention instigated when necessary ( i.e. when the tumour progresses).
The treatment of metastatic prostate cancer has centred around the modification of androgen levels and/or activity for more than 60 years since Charles Huggins first reported the importance of testicular androgens in prostate tumour maintenance (Huggins et al. , 1940). Like the normal prostate, prostate cancer is critically dependent on AR signalling pathways, therefore the treatment of metastatic prostate cancer aims to inhibit these pathways (Section 1.1.2). Therapies that suppress AR signalling may do so by eliminating circulating androgens (“androgen ablation”), or by antagonising the AR itself (“AR blockade”).
Circulating testicular androgens can be eliminated by orchiectomy, as testosterone, produced by Leydig cells in the testes, constitutes ~ 90 % of serum androgen levels (Dehm and Tindall, 2006). Production of testosterone can also be down-regulated by the use of gonadotropin-releasing hormone (GnRH) agonists, which ultimately shut off luteinising hormone (LH) production by the cells of the anterior pituitary via feedback inhibition, reducing testicular production of testosterone. GnRH agonists may also directly affect prostate tumours (Kraus et al. , 2006).
AR antagonists, also called anti-androgens, include the non-steroidal drugs bicalutamide (Casodex) and flutamide, and the steroidal drug cyproterone acetate. AR antagonists bind competitively to the AR, preventing the binding of androgens. Bicalutamide promotes the formation of transcriptionally
6 1 2
3 4
5
Figure 1.2. Prostate tumour architecture: Gleason grading system. Prostate tumour specimens demonstrating the continuum of deteriorating prostate gland architecture. Specimens are numbered according to Gleason grade. Tumours of Gleason grade 1 and 2 closely resemble normal prostate. 3: The tumour is still well differentiated with glandular structures of more variable shape, and some invasion into the stroma. 4: Disruption and/or loss of the normal glandular unit are observed. 5: Gleason grade 5 tumours are characterised by loss of glandular structure and differentiation.
Images reproduced from: http://pathology2.jhu.edu/gleason/patterns.cfm#Gleason%20Pattern%205
7 inactive AR/co-activator complexes (Masiello et al. , 2002), and both flutamide and bicalutamide prevent AR from stably binding to target gene promoters (Farla et al. , 2005). AR antagonists are also useful to inhibit the action of adrenal androgens, as while orchiectomy decreases levels of circulating testicular androgens by >90 %, adrenal androgen levels are reduced by only 50 % (Mizokami et al. , 2004). Anti-androgens are usually administered in combination with GnRH agonists, especially at the commencement of GnRH agonist treatment when tumour flare may occur. While hormone treatments are generally effective initially, as the disease progresses, the tumour ceases to be dependent on androgens. Treatment options for androgen-independent prostate cancer are limited, being generally palliative in nature and as a consequence, the disease is often fatal (Trachtenberg and Blackledge, 2002).
1.2 The Androgen Receptor.
1.2.1 Structure.
The AR gene encodes a 110 kDa ligand-activated transcription factor of the nuclear steroid receptor family, and is located at chromosomal position Xq11.2-q12 (National Centre for Biotechnology Information [NCBI] Entrez Gene www.ncbi.nlm.nih.gov/entrez/query.fcgi?CMD=search&DB=gene). Other members of this family include the oestrogen receptor (ER) and the progesterone receptor (PR). Nuclear steroid receptors are composed of an amino(N)-terminal domain that contains an activation function (AF-1) domain, a DNA-binding domain (DBD), and a carboxyl (C)-terminal ligand-binding domain (LBD) (Agoulnik and Weigel, 2006). The LBD contains 12 α-helices, which form a ligand-binding pocket, and a second AF domain, AF-2 (Figure 1.3A) (He et al. , 2004). The DBD is comprised of two zinc finger motifs that determine sequence-specific binding to hormone response elements in gene promoters (Agoulnik and Weigel, 2006). The DBD is separated from the LBD by a hinge region, which contains both the nuclear localisation signal that allows nucleo- cytoplasmic shuttling, and a proline-, glutamate-, aspartate-, serine- and threonine-rich (PEST) ubiquitination/ degradation motif (Jenster et al. , 1992; Lee and Chang, 2003).
8 A PolyQ PolyP 58-78 371-379 Hinge DBD LBD AF1 AF2 1 142-337 PolyG 559-624 676 - 919 448-472
B
Figure 1.3. Androgen receptor structure. A. Schematic structure of the AR, showing the amino-acid repeat tracts ( ), activation function domains (AF; ), ligand-binding domain (LBD; ) and DNA-binding domain (DBD; ). Numbering starts from the first amino-acid and assumes 20 glutamines in the polyQ tract. B. Ribbon diagram of the LBD of the AR bound to the synthetic androgen R1881 (yellow and red space-filled atom diagram), showing the 12 component helices ( ). = AF-2 helices. . = FXXLF helix from AF-1, involved in the N/C terminal interaction. Helix 12 is folded over the ligand-binding pocket. (helix 12) and (helix 3) = residues critical for AF-1 binding.
A. Adapted from Agoulnik and Weigel (2006). B. Reproduced from He et al. (2004).
9 Unlike other nuclear steroid receptors, the N-terminal AF-1 is the major transactivation domain of the AR (Agoulnik and Weigel, 2006; Bevan et al. , 1999). Upon ligand binding in AR, ER and PR, helix 12 of the LBD/AF-2 folds over the ligand binding pocket (Figure 1.3B), creating a surface for AF-2 to interact with nuclear co-factors (He et al. , 2002). In most of these receptors, AF-2 binds to LXXLL motifs in co-regulators, but in the AR, AF-2 preferentially binds a FXXFL motif in the AF-1 domain (He et al. , 2000; He et al. , 2004). This N/C-interaction decreases ligand dissociation and stabilises the receptor (Doesburg et al. , 1997; Lee and Chang, 2003).
The AR also has three amino-acid repeat tracts in the first exon, a polyglutamine (polyQ) repeat near the N-terminus, and central polyproline (polyP) and polyglycine (polyG) repeats. The polyQ repeat is highly polymorphic (Lubahn et al. , 1988), with a normal range of between 18 and 22 repeats. Extremely long polyQ repeat lengths (> 40) are associated with spinal and bulbular muscular dystrophy (La Spada et al. , 1991), while shorter repeats (< 18) are associated with increased AR activity (Callewaert et al. , 2003; Feldman, 1997). There is limited evidence that shorter polyG repeats are associated with an increased prostate cancer risk, or earlier disease onset, although the significance of this association is currently unresolved (Bratt et al. , 1999; Correa-Cerro et al. , 1999; Edwards et al. , 1999; Giovannucci et al. , 1997; Hakimi et al. , 1995; Ingles et al. , 1997; Irvine et al. , 1995; Stanford et al. , 1997).
1.2.2 AR Activity.
In the absence of ligand, the AR is associated with heat shock proteins (HSP) 90, 70 and 56 and is localised in the cytoplasm (Dehm and Tindall, 2006). This association has three likely functions, to exclude the AR from the nucleus in the resting state, to protect the AR from proteolytic degradation, and to maintain the receptor in a conformation that allows ligand binding. Testosterone, which is hydrophobic, diffuses into the cell where it is converted to DHT by type II 5 α-reductase. DHT binds the AR with ~ 10 times higher affinity than testosterone, and the conformation taken by DHT-bound AR is thought to be more resistant to proteolysis (Dehm and Tindall, 2006). Androgens enter the ligand binding pocket of the AR LBD and induce a 10 conformational change that promotes dissociation of the HSPs and dimerisation of AR. The interaction between the N and C termini of the receptor occurs (Section 1.2.1), the AR dimer is phosphorylated on several sites (Section 1.2.3), and is transported into the nucleus (Edwards and Bartlett, 2005a).
In the classic mode of action, the AR dimer binds to an ARE in the gene promoter. The ARE may be canonical, composed of 5’-TGTTCT-3’-like inverted repeats, and these ARE’s are recognised by other type I receptors, particularly the glucocorticoid receptor, or it may be AR-selective, in which case the element appears as direct repeats (Callewaert et al. , 2003). Binding of the AR to canonical versus selective ARE’s is probably influenced by nuclear co-regulators (Section 1.2.3). Upon DNA binding, AR initiates transcription by recruiting a series of co-activators (Section 1.2.3) and RNA polymerase (pol) II. AR can also act non-classically (Figure 1.4), where it functions as a co-activator for other transcription factors, and can also participate in kinase cascades (“non-genomic” signalling), in which transcription is initiated by downstream phosphorylation targets independent of AR binding at the promoter (Agoulnik and Weigel, 2006).
1.2.3 Regulation of Androgen Receptor Transcriptional Activity.
Transcriptional activity of the AR is regulated by ligand binding, as well as phosphorylation, sumoylation (Poukka et al. , 2000), and the presence of nuclear co-factors. The AR is phosphorylated within ~ 15 minutes of synthesis (Kuiper et al. , 1991). This increases its apparent molecular mass from 110 kDa to ~ 112 kDa, and eventually to ~ 114 kDa (Jenster et al. , 1994). The purpose of this phosphorylation appears to be to stabilise the AR homodimer, promote nuclear translocation and to assist ligand binding (Edwards and Bartlett, 2005a). Unphosphorylated AR can bind hormone (Kuiper et al. , 1992), but its binding is impaired by dephosphorylation (Blok et al. , 1998).
11 RTK Androgens
G PCR
AR
HSP M A P Cytoplasm K
Nucleus
Kinase cascade TF Pol II 1 ARE
Transcription
2 TF
Non-genomic effects 3 TF
Figure 1.4. Androgen receptor signalling: Three modes of action. The AR ( ) is complexed with heat-shock proteins (HSP; ) in the cytoplasm. Androgens ( ) or extra-cellular stimuli, transmitted by receptor tyrosine kinases (RTK; ) such as ERBB2, or by G-protein coupled receptors (GPCR; ), stimulate dissociation of the AR and HSPs. AR is also phosphorylated ( ) by mitogen-activated protein kinases (MAPK; ) and is translocated into the nucleus. Ligand-bound, phosphorylated AR modulates RNA polymerase II (Pol II; ) transcriptional activity from ARE-containing promoters (1), or the activity of other transcription factors (TF; ) (2), or participates in non-genomic signalling (3).
Based on concepts described in Agoulnik and Weigel (2006).
12 Phosphorylation of the AR has been identified at serines (S) 16, 81, 94, 256, 308, 424 and 650 (Gioeli et al. , 2006), 210 and 790 (Murillo et al. , 2001), 514 (Yeh et al. , 1999b) and 515 (Bakin et al. , 2003). Specific functional effects for some of these residues have been elucidated. Recently, Goeli et al . (2006) showed that phosphorylation of S650 promotes nuclear translocation of AR, accounting for why mutation of this residue results in ~ 30 % decrease in AR transcriptional activity (Gioeli et al. , 2006; Zhou et al. , 1995). The sensitivity of the AR to the synthetic androgen R1881 is increased by mitogen-activated protein kinase (MAPK) phosphorylation of S515 (Bakin et al. , 2003). Phosphorylation of S210 and S790 by Akt (also known as protein kinase B, PKB) modulates AR transcriptional activity (Ghosh et al. , 2003), although this may be due to enhanced AR degradation associated with phosphorylation of these residues (Lin et al. , 2002). It appears that the MAPK and Akt target residues are phosphorylated transiently to directly modify AR activity and/or turnover, rather than being constitutively phosphorylated (Edwards and Bartlett, 2005a; Gioeli et al. , 2006; Zhu et al. , 2001b).
Phosphorylation of nuclear co-regulators also affects AR activity. Efficient transcription is dependent on the recruitment of basal transcription factors, template activating factors and TATA binding protein. AR must associate with co-activators to recruit these, and phosphorylation can affect both the interaction between AR and the co-regulator, and the activity of the co- regulator itself. Co-activators also recruit enzymes that are capable of modifying local chromatin architecture, such as the SWI/SNF complex, methyl transferases and acetyl transferases. In fact, many of the known AR co- activators are themselves histone acetyl-transferases, which is common among nuclear co-regulators (Agoulnik and Weigel, 2006; Smith and O'Malley, 2004). AR coregulators (summarised in Table 1.1) include the p160 co- activators of the steroid receptor regulator (SRC) superfamily (Bevan et al. , 1999; Gregory et al. , 1998; Linja et al. , 2004), cyclic-AMP (cAMP) response
13 Table 1.1. Co-regulators of the AR and their function in prostate cancer. Adapted from Culig et al. (2004)
Co - Function in prostate cancer: Reference: regulator: β-catenin Wnt ↑ AR transcriptional activity, ↓ AR ligand Song and signalling specificity. Gelmann (2005).
ARA54 Activates T877A mutant AR in presence of Kang et al. (1999). flutamide and 17 β-oestradiol.
ARA55 Causes AR agonist activity of flutamide. Yeh et al. (1999a).
ARA70 Causes AR agonist activity of flutamide and Yeh et al. (1999a). 17 β-oestradiol.
ART27 ↓ expression in prostate cancer. Taneja et al. (2004). CREBBP Potentiates flutamide-stimulated AR Comuzzi et al. activity; ↑ expression following androgen (2003,2004). withdrawal.
Cyclin D1 D1a: ↓ AR transcriptional activity, inhibits cell Reutens et al. cycle progression. (2001).
D1b: ↑ prostate cancer cell proliferation, ↑ Burd et al. (2006). expression in prostatic neoplasms.
NCoR Co-repressor Not established. p300 Expression correlates with poor prognostic Debes et al. indicators in prostate tumours; required for (2002, 2003), Fu ligand-independent activation of AR by IL6; et al. (2000). required for cell proliferation of androgen- dependent LNCaP cells; acetylates and activates AR.
PCAF Acetylates and activates AR; rescues cyclin Fu et al. (2002), D1a-mediated repression of AR. Reutens et al. (2001). PIAS1 ↓ mRNA expression in recurrent prostate Linja et al. (2004). tumours.
SMRT Co-repressor Transient over-expression ↓ AR activity in Liao et al. (2003) presence of flutamide.
SRC1 p160 family ↓ mRNA expression in recurrent prostate Linja et al. (2004); tumours; activates AR after phosphorylation Gregory et al. by MAPK; ↑ expression in recurrent (2001a); Ueda et prostate tumours; implicated in ligand- al. (2002). independent activation of AR by IL6.
SRC2 “ ↑ expression in recurrent prostate tumours; Gregory et al. (TIF2 ↑ AR activation by androstenedione; (2001a); Song and GRIP1) increases association of AR and β-catenin. Gelmann (2005).
SRC3 “ Expression correlates with prostate tumour Gnanapragasam (RAC3) grade. et al. (2001)
Tip60 ↑ nuclear localisation in prostate tumour Halkidou et al. cells following androgen withdrawal. (2003). 14 element binding protein-binding protein (CREBBP, CBP) (Linja et al. , 2004; Yu et al. , 2004), its functional homologue p300 (Fu et al. , 2000), and p300/CREBBP-associated factor (PCAF) (Debes et al. , 2005; Linja et al. , 2004; Yu et al. , 2004).
Among the other co-activators that interact with AR are the diverse set of co- regulators known as AR-associated proteins (ARA) (Edwards and Bartlett, 2005b), β-catenin (Yang et al. , 2002) and cyclin D1 (Petre-Draviam et al. , 2005). It is not presently clear which of these co-activators are always required for optimal AR activity, and which function in a gene- or tissue-specific manner (Agoulnik and Weigel, 2006). The relevance of these interactions in prostate cancer is further discussed in Section 1.3.3.
AR is also bound by co-repressors, including nuclear co-repressor (NCoR) and silencing mediator of retinoic acid- and thyroid receptor (SMRT). SMRT and NCoR inhibit both agonist- and antagonist-bound AR, partly by competing with co-activators. (Hodgson et al. , 2005; Yoon and Wong, 2006) Overexpression of SMRT further attenuates AR activation in the presence of flutamide (Liao et al. , 2003).
1.2.4 Regulation of Androgen Receptor Protein Levels.
The regulation of AR protein levels is complex, involving aspects of transcriptional and post-transcriptional regulation. The AR gene contains two exonic AREs (Grad et al. , 2001) and is transcriptionally repressed by androgens (Wolf et al. , 1993; Yeap et al. , 1999). The AR promoter is TATA- less, and this is compensated for by a specificity protein-1 (Sp1) site (Lee and Chang, 2003). A functional cAMP response element (CRE) also makes a positive contribution to promoter activity (Lee and Chang, 2003), but the mechanism by which either of these sites integrate with androgen regulation of the AR gene is unknown. AR is also a transcriptional target of lymphoid enhancer factor 1/T-cell factor (LEF1/TCF) transcription factors, which are activated by wingless type/MMTV integration (Wnt) signalling (Terry et al. , 2006). AR mRNA is highly upregulated by activation of Wnt signaling in prostate cancer cells. Paradoxically, Wnt activation also appears to stimulate phosphorylation of AR by Akt, promoting mdm2 E3-ligase-mediated 15 proteolysis, reducing AR protein in Wnt-stimulated prostate cancer cells (Yang et al. , 2006).
Androgen regulation of AR has been most well studied in the LNCaP prostate cancer cell line, which exhibits androgen-dependent growth and expresses the AR. In LNCaP, transcription of the AR gene and steady-state mRNA levels are decreased by exposure to androgen, but protein levels are increased ~ 2-fold (Krongrad et al. , 1991; Wolf et al. , 1993; Yeap et al. , 1999). This apparent paradox is solved by the observations that androgens stabilise both the AR protein and the AR mRNA, and increase translation of the AR transcript. Thus, although there is a net decrease in AR mRNA due to lower rates of transcription, the remaining mRNA is more persistent and gives rise to more protein per transcript (Yeap et al. , 1999). Contrastingly, in the breast cancer cell line MDA453, AR mRNA was destabilised by androgens, but transcription of the AR gene was unaffected (Yeap et al. , 1999).
The AR protein is degraded by systemic proteasomal pathways, including Akt/Mdm2 ubiquitin-mediated proteolysis (Lin et al. , 2002; Sheflin et al. , 2000) (Section 1.2.3) and the phosphatase and tensin homolog deleted from chromosome 10 (PTEN)/caspase 3 pathway (Lin et al. , 2004). This process is presumably inhibited by the presence of androgens, which extend AR protein half-life from ~ 3 h to > 10 h (Gregory et al. , 2001). Additionally, in the rat ventral prostate, androgens promote sequestration of AR mRNA in polyribosomes, consequently increasing mRNA stability, which in turn facilitates increased translation, and therefore protein levels (Mora and Mahesh, 1999). A similar mechanism may account for the increased AR mRNA stability detected in androgen-treated LNCaP cells (Yeap et al. , 1999). Stability of the AR mRNA is further discussed in Section 1.5.
1.3 The Androgen Receptor in Prostate Cancer. A cohort of androgen-regulated genes, such as PSA, are almost always expressed in androgen-independent prostate cancer, and oligonucleotides or antibodies targeted against the AR inhibit proliferation of androgen- independent prostate cancer cells (Haag et al. , 2005; Zegarra-Moro et al. , 2002). A significant similarity amongst androgen-independent prostate 16 cancers is that the majority continue to express the AR, therefore it is likely that the AR signalling axis remains active in cells of these tumours (Chen et al. , 2004; Taplin and Balk, 2004). A number of factors contribute to the maintenance of AR expression and signalling, and these include development of mutant AR that is stimulated by hormones other than DHT or testosterone, or by anti-androgens (Section 1.3.1), amplification of the AR which overcomes the presence of low androgen levels (Section 1.3.2), or altered expression of AR co-regulators, which increase the sensitivity of AR to very low levels of DHT, or potentiate AR transcriptional activity (Section 1.3.3).
1.3.1 AR Gene Mutations.
AR gene mutations have been detected more frequently in cultured prostate cancer cells than in vivo , however AR mutations are found in 10-20 % of prostate cancer specimens (Edwards and Bartlett, 2005a). These mutations are thought to develop in response to anti-androgen treatment, as a proportion of the mutant AR’s cause the receptor to be activated by intended antagonists. Thus in some advanced prostate cancers, tumour growth is actually stimulated by anti-androgens, and temporarily inhibited by discontinuation of anti- androgen treatment. This is known as anti-androgen withdrawal syndrome (Navarro et al. , 2002). AR gene mutations have been reported to occur more frequently in patients treated with combined androgen ablation regimens using the AR antagonist flutamide, with the AR LBD mutations in the tumour cells resulting in flutamide becoming an AR agonist (Taplin et al. , 1999). These patients still responded appropriately to the AR antagonist bicalutamide. Similar results were obtained following culture of LNCaP cells in the presence of bicalutamide and absence of androgen for extended periods. Proliferation of these cells was initially inhibited by bicalutamide, but after 6-13 weeks of continuous culture in the presence of bicalutamide, the cells became dependent on its presence, via W741C or W741L mutations in the LBD. Mutant AR’s containing W741C/W741L base substitutions could still be inhibited by flutamide, suggesting that the mutations were ligand-specific (Hara et al. , 2003).
AR gene mutations in prostate tumours may also significantly increase the sensitivity of AR to other steroid ligands, as was shown for the T877A mutant 17 AR, which exhibited increase transcriptional activity in response to the adrenal androgen, androstenediol (Mizokami et al. , 2004). The anti-androgen flutamide also exhibited aberrant agonist activity on this AR. Another mutation, L701H has been reported in two prostate cancer patients (Navone et al. , 1997; Zhao et al. , 1999), and resulted in AR being activated by cortisol, but not DHT. A double-mutant AR having both the T877A and L701H mutations was described in the cell line MDA-PCa 2b, which was established from a bone metastasis of a patient with androgen-independent prostate cancer (Zhao et al. , 2000a). This AR was strongly activated by cortisol, but was also stimulated by DHT, progesterone, oestradiol, and flutamide. AR gene mutations provide one mechanism by which androgen-independent disease could develop, however, the relatively low rate of AR mutation detection in patients indicates that other mechanisms are also associated with the development of hormone refractory tumours.
1.3.2 AR Gene Amplification.
Another mechanism leading to the development of androgen-independent prostate cancer is amplification of the AR gene. In a study by Edwards and colleagues (2003), an increase from an incidence of 0-5 % in androgen- sensitive to 20-30 % in androgen-independent disease was detected using matched samples of androgen-dependent and androgen-independent tumours from the same patients (Edwards et al. , 2003). In the same study, 80 % of tumours with AR gene amplification also exhibited significant up-regulation of AR protein, a correlation previously observed by Ford et al. (2003), however there was no difference in time to relapse between patients who presented with tumours carrying AR gene amplifications and those who did not (Edwards et al. , 2003; Ford et al. , 2003). A correlation has also been noted between AR gene amplification and AR mRNA expression (Linja et al. , 2001). Experiments in animal models suggest that increased AR expression may allow anti- androgens to act as weak agonists (Chen et al. , 2004).
1.3.3 AR Co-Activators in Prostate Cancer.
As described in Section 1.2.3, AR activity is regulated by factors other than absence or presence of ligand, including AR co-regulators, altered expression of which is thought to contribute to the development of androgen-independent
18 prostate cancer. Increased activity of some activators may stimulate ligand- independent activation of the AR, while the loss of others may result in inappropriate de-repression.
Members of the p160 family of co-activators (Section 1.2.3) are reported to be differentially expressed in advanced prostate cancers. In a study of eight recurrent prostate cancers, the majority were found to express high levels of SRC1 and SRC2 (also known as TIF2/GRIP1) as well as high levels of AR (Gregory et al. , 2001a). Co-expression of these proteins is likely to have functional consequences on AR activity in prostate tumours, as transient co- transfection of SRC2 and AR in the monkey kidney cell line CV1 substantially increased transactivation of AR at physiological concentrations of androstenedione (Gregory et al. , 2001a).
SRC2 also enhances the binding of the multifunctional oncoprotein β-catenin to the AF2 domain of the AR, an interaction which increases the sensitivity of AR for ligand while decreasing ligand specificity, thus stimulating AR transcriptional activity (Song and Gelmann, 2005). In a study of 37 patients, the expression of another AR co-activator, SRC3 (RAC3), in prostate tumours correlated significantly with tumour grade and stage of disease, but not with serum PSA levels. In addition, moderate or high SRC3 expression was associated with poorer disease-specific survival (Gnanapragasam et al. , 2001). This study indicated that SRC3 may have a role in the progression of prostate cancer.
In the presence of androgen, SRC1 is phosphorylated by MAPK and associates with and activates AR. When interleukin (IL) 6 levels are elevated causing activation of downstream signalling pathways, SRC1 activates AR in the absence of ligand (Ueda et al. , 2002), although evidence to the contrary has also been presented (Jia et al. , 2003). IL6 action is also mediated by p300, an AR co-activator which directly acetylates and activates AR (Fu et al. , 2000). The histone acetyl-transferase activity of p300 was shown to be necessary for IL6-induced, ligand-independent transactivation of the AR in androgen-independent and -dependent LNCaP cells and for proliferation of
19 androgen-dependent LNCaP cells (Debes et al. , 2002; Debes et al. , 2003). These findings have been supported in vivo, as histological detection of p300 in tumours from patients who had undergone radical prostatectomy correlated with in vivo proliferation, tumour volume, extra-prostatic extension (an unfavourable prognostic indicator), as well as prostate cancer progression after surgery (Debes et al. , 2003).
CREBBP, functionally homologous to p300, potentiates flutamide-stimulated activity of both mutant T877A and wild-type AR (Comuzzi et al. , 2003), and is up-regulated following androgen withdrawal (Comuzzi et al. , 2004). A protein frequently associated with both CREBBP and p300, PCAF, acetylates AR, increasing its transcriptional activity (Fu et al. , 2002). The histone acetyl- transferase activity of PCAF can also rescue cyclin D1a-mediated AR trans- repression. Cyclin D1a inhibits ligand-induced AR activation, possibly by competing with AR for PCAF binding, impedes cell-cycle progression in androgen-dependent prostate cancer cells and is rarely over-expressed in prostate cancer (Reutens et al. , 2001). In contrast, a mutated form, cyclin D1b, is often found in prostatic neoplasms and can stimulate cell proliferation in an AR-dependent fashion (Burd et al. , 2006). Tip60, another histone acetyl- transferase, is expressed at increased levels and is located in the nucleus following androgen withdrawal, with in vitro studies demonstrating its involvement in transcription of the PSA gene in androgen-independent and androgen-dependent cell lines (Halkidou et al. , 2003).
The ARA proteins, which recruit CREBBP and PCAF and assist with chromatin remodelling, have been implicated in various aspects of AR dysregulation (Yeh et al. , 1999a). When over-expressed in DU145 cells, ARA70 conferred androgenic activity on 17 β-oestradiol, while both ARA70 and ARA55 can cause agonist activity of the anti-androgen flutamide (Fujimoto et al. , 1999; Yeh et al. , 1999a). ARA54 co-activates the mutant T877A AR in LNCaP cells in the presence of 17 β-oestradiol or flutamide, but does not activate wild-type AR under the same conditions (Kang et al. , 1999). ARA proteins also interact with or recruit CREBBP and PCAF, assisting chromatin remodelling (Yeh et al. , 1999a).
20
In a study of 16 co-regulators, including SRC1, SRC2, β-catenin, protein inhibitor of activated STAT (PIAS1), CREBBP, NCoR, cyclin D1, p300, and ARA24, only PIAS1 and SRC1 exhibited decreased mRNA expression in hormone-refractory prostate tumours, but the protein levels were not determined, making these results difficult to interpret (Linja et al. , 2004). One co-regulator that does appear to be down-regulated in prostate cancer is ART27, which is expressed in differentiated columnar cells in normal tissue, but expressed at negligible levels in prostate cancer cells. Overexpression of ART27 in LNCaP cells inhibited androgen-mediated proliferation, but, curiously, enhanced androgen-induced transcription of the PSA gene (Taneja et al. , 2004). Surprisingly, loss of the co-repressors SMRT and NCoR has not been extensively reported in prostate tumours, perhaps because their function outside the context of AR regulation necessitates their continued expression.
AR co-activators have been demonstrated to be down-stream phosphorylation targets of MAPK signalling pathways that originate at cell surface receptors, including the erythroblastic leukemia viral oncogene homologue 2 (ERBB2, Her2/Neu), EGF receptor, IGF receptor and Wnt signalling pathways, and inappropriate activation of these pathways has been linked to carcinogenesis (Culig, 2004; Gioeli et al. , 2006; Godoy-Tundidor et al. , 2005; Rochette-Egly, 2003; Terry et al. , 2006).
EGF signalling increases the interaction between SRC2 and AR, and consequently androgen-dependent AR transactivation in the prostate cancer cell line CWR-R1 via a mechanism that involves phosphorylation and stabilisation of SRC2 (Gregory et al. , 2004). Over-expression of another protein of the EGF receptor family, ERBB2, in LNCaP prostate cancer cells activates Akt, thereby promoting cell survival and proliferation in the absence of androgens. This effect is proposed to occur via phosphorylation and inactivation of the pro-apoptotic factors BCL2-agonist of cell death (BAD) and pro-caspase 9, as well as phosphorylation and activation of AR by Akt (Wen et al. , 2000). The residues of AR observed to be phosphorylated by Akt in this study, S213 and S791, are distinct from those thought to promote AR
21 degradation. IGF type I receptor signalling has been shown to affect AR phosphorylation status and translocation to the nucleus, while IGFR1 levels are regulated by androgen via an apparent non-genomic effect of the AR (Wu et al. , 2006).
1.3.4 Cell Line Models of Prostate Cancer.
The most widely-used prostate cancer cell line is LNCaP, an androgen- dependent cell line that was originally derived from a lymph-node metastasis (Dehm and Tindall, 2006; Horoszewicz et al. , 1983). LNCaP cells express high levels of the AR, which contains a point mutation that causes an amino- acid substitution in the LBD (T877A; Section 1.3.1), and other prostate-specific and androgen-regulated genes such as PSA and NKX3.1 (He et al. , 1997; Horoszewicz et al. , 1983; van Bokhoven et al. , 2003). A second model cell line, 22Rv1, is also in relatively frequent use. 22Rv1 cells were derived from an androgen-dependent xenograft (CWR22R-2152), and are themselves androgen-responsive, although not androgen-dependent for growth (Sramkoski et al. , 1999; van Bokhoven et al. , 2003). 22Rv1 cells express low levels of PSA, and two mutated AR’s, both of which have a duplication of exon 3 resulting in a third zinc finger motif (Section 1.2.1), and one of which is C- terminally truncated, omitting the LBD (Chlenski et al. , 2001; Tepper et al. , 2002; van Bokhoven et al. , 2003). The higher molecular weight form of the AR also contains a H874Y mutation in the LBD that was detected in the original tumour and which contributes to increased promiscuity of the AR (Attardi et al. , 2004; Tan et al. , 1997). Other commonly used prostate cancer cell lines are DU145, PC3 and LAPC4. DU145 and PC3 have variously been reported to express AR (Alimirah et al. , 2006) or not (van Bokhoven et al. , 2003), and are generally used as models of androgen-independent prostate cancer. LAPC4 cells express a wild-type AR and are androgen-dependent (van Bokhoven et al. , 2003). Studies in this thesis have used LNCaP and 22Rv1 cell lines to investigate the role of mRNA stability regulators in androgen action.
1.4 Post-transcriptional Regulation of mRNA. As an mRNA is being transcribed, it accumulates a coat of RNA-binding proteins, forming a heterogeneous nuclear ribonucleoprotein (hnRNP) complex, which becomes a messenger RNP (mRNP) in the cytoplasm. 22 HnRNP/mRNP complexes are unusually large, diverse, and dynamic. The component RNA-binding proteins are involved in various co- and post- transcriptional processes, such as 5’ and 3’ end processing, poly-adenylation, splicing, nuclear-cytoplasmic export, sub-cellular localisation, translation, stability and turnover; and at each step the composition of the hnRNP fluctuates (Dreyfuss et al. , 2002). Each of these steps contributes to post- transcriptional control of gene expression.
Regulated mRNA turnover is an important aspect of gene expression for antiviral defence and maintaining mRNA quality. It allows the detection and destruction of aberrant mRNAs that have been incorrectly processed, or which contain a premature stop codon or no stop codon (nonsense-mediated decay, NMD, and non-stop decay, NSD, respectively), as well as RNA of viral origin (Parker and Song, 2004). Pertinent to this thesis, modulation of mRNA stability allows the cell to adapt protein production rapidly in response to stimuli by stabilising or de-stabilising relevant transcripts (Meyer et al. , 2004). Many mRNAs, particularly those that encode potent effectors of critical cellular events, such as cyclins, (which regulate cell cycle progression) have very short half-lives; persistent translation of these transcripts could inappropriately alter cell viability. Conversely, proteins that are constitutively required may have mRNAs with extended half-lives. The distinguishing elements that identify transcripts with long or short half-lives are not well-characterised, however, both the sequence and structure of the mRNA as well as its cognate proteins contribute. At least three factors influence the composition of the mRNP: cis - acting sequence motifs in the mRNA (Section 1.4.1), the processes it has undergone (splicing, export, translation), and the status of the cell, which influences the availability and response of RNA-binding proteins (Dreyfuss et al. , 2002).
1.4.1 Structural and Sequence Motifs in mRNA Processing and Turnover.
The major elements of a mature cytoplasmic mRNA are the 5’ cap, 3’ poly- adenosine (polyA) tail, and 5’ and 3’ untranslated regions (UTRs), which can
23 be up to several kilobases (kb) in length (Figure 1.5). Pre-mRNAs, that is, nascent transcripts in the nucleus, have in addition sequences that define exon boundaries and sequences that identify cleavage and poly-adenylation sites (Figure 1.5). Just as ~ 75 % of transcripts are alternatively spliced (Johnson et al. , 2003), there are also alternative cleavage and poly-adenylation sites in ~ 30 % of mRNAs (Soller, 2006). In terms of mRNA turnover, however, the most important elements are those which remain in the mature mRNA.
The 5’ cap contains a guanine monophosphate (GMP), methylated on nitrogen 7 (m 7-GMP). It is added to the 5’ end of the nascent transcript with an unusual 5’-5’ phosphodiester bond, when the mRNA is 20-30 nucleotides long. The riboses of the two nucleotides immediately adjacent to the 5’ end may also be methylated. The cap has several functions. It stimulates splicing of the first exon and 3’ end processing, protects the mRNA from 5’-3’ exoribonucleases and makes interactions with the eukaryotic translation initiation factor eIF4G that are essential for translation (Soller, 2006).
The polyA tail is added during 3’ end processing. When RNA pol II has transcribed past the 3’ end of the gene, the transcript is cleaved at a CA dinucleotide downstream of an A(A/U)UAAA hexamer, which is distinguished from non-target hexamer sequences by the presence of U-rich upstream (USE) and downstream (DSE) sequence elements (Legendre and Gautheret, 2003). A tail of ~ 200 adenosines is then added by polyA polymerase (PAP), which is thought to be shortened to ~ 100 adenosines by the time the transcript leaves the nucleus (Soller, 2006).
The length of the polyA tail is an important determinant of mRNA stability. When the length of the tail falls below a critical level (~ 30-60 nucleotides in mammals), degradation is initiated (Chen et al. , 1994). De-adenylation is a cytoplasmic event, and at least three de-adenylating exoribonucleases exist in mammals, the heterodimeric polyA nuclease PAN2/PAN3, polyA-specific ribonuclease (PARN, DAN), and the mammalian homologue of the yeast CCR4-NOT complex (Meyer et al. , 2004).
24 A Non-coding regions Intron/exon (introns) Cleavage + boundary polyA sites First nucleotide m7-GpppA
5’ Cap 5’ UTR 3’ UTR Coding regions (exons)
B First nucleotide
7 m -GpppA AAAAA (200)
5’ Cap 5’ UTRCoding 3’ UTR polyA tail region
Figure 1.5. Structural elements of messenger RNA. A. Immature mRNA. The 5’ cap is added early during transcription. Cleavage and poly-adenylation occur late in transcription and signals for these processes are located near the end of the transcript ( ). Introns ( ) are delineated by intron/exon boundaries, and are spliced out during pre-mRNA processing. B. Mature mRNA. The 5’ cap and polyA tail protect the mRNA from degradation. The 5’ and 3’ UTRs ( ) contain other information relevant to RNA stability and translation. This information is sometimes located in the coding region ( ).
Based on concepts described in Meyer et al. (2004) and Mitchell and Tollervey (2001).
25
The activity of these exonucleases is influenced to differing extents by the cytoplasmic polyA-binding protein (PABPC) (Figure 1.6), a protein which specifically binds the polyA tail (Meyer et al. , 2004; Parker and Song, 2004). PABPC also binds to the cap-interacting translation initiation factor eIF4G, which may explain why the presence of a polyA tail is necessary for initiation of translation, even though the mRNA can be de-adenylated during translation (Meyer et al. , 2004).
Exoribonucleic degradation from the 3’ end is probably the major decay pathway in mammals and is carried out by the exosome, a multi-protein complex of 3’-5’ exonucleases and associated proteins (Chen et al. , 2001) (Figure 1.7). Although a nuclear exosome exists, it is discrete from the cytoplasmic complex, and is not involved in decay of cytoplasmic messages (Meyer et al. , 2004). Degradation by the exosome depends on prior de- adenylation, as the exosome appears to have no intrinsic de-adenylation activity (Chen et al. , 2001).
PolyA tail shortening also initiates de-adenylation-dependent decapping (Figure 1.7). The cap is removed by hydrolysis of the 5’-5’ phosphodiester bond, leaving a 5’ monophosphate on the transcript and releasing N 7-methyl- guanine diphosphate (m 7-GDP) (Cougot et al. , 2004). This reaction is catalysed by Dcp1/Dcp2. After cap hydrolysis, the mRNA can be rapidly degraded by the 5’-3’ exonuclease Xrn1 (Cougot et al. , 2004).
Although endonucleolytic cleavage of some transcripts has been reported, the cleavage products are essentially equivalent to de-capped and de-adenylated mRNAs, and so once cleaved, are probably degraded by the exosome and Xrn1 (Gallouzi et al. , 1998). Importantly, mRNAs missing either the polyA tail or 5’ cap cannot interact properly with translation factors, and therefore are unlikely to be translated at a significant rate (Meyer et al. , 2004).
26 A - PABPC m7-GpppA AAAAAA
B PABPC + m7-GpppA AAAAAA
C PABPC - m7-GpppA AAAAAA
+
Figure 1.6. Mechanisms of de-adenylation. A. The CCR4-NOT complex ( ) is inhibited (-) by cytoplasmic polyA- binding protein (PABPC; ), but is not affected by the presence of a 5’ cap. B. PAN2/PAN3 ( ) is stimulated by PABPC, and the effect of the 5’ cap is unknown. C. PARN ( ) is inhibited by PABPC (-), but stimulated by the 5’ cap (+).
Adapted from Parker and Song (2004).
27 7 m -GpppA AAAAA (200)
De-adenylation
m7-GpppA AA
De-adenylation 3’-5’ Degradation dependent decapping
Dpc1/Dcp2 m7-GpppA pA AA p 7 Gp Exosome m - 5’-3’ Degradation
AA Xrn1
Figure 1.7. Mechanisms of mRNA decay. The mRNA is first de-adenylated (Figure 1.6), then is either degraded from the 3’ end by the exosome ( ), or is decapped by Dcp1/Dcp2 ( ), then degraded from the 5’ end by Xrn1 ( ).
Adapted from Parker and Song (2004).
28 De-adenylation and subsequent degradation occurs in almost all cytoplasmic mRNAs, yet transcripts can have vastly different life-spans; from ~ 15 min for c-fos, to ~ 24 h in the case of β-globin (Shyu et al. , 1989). This can be explained by the presence of additional sequence motifs within the UTR’s of the transcript, and occasionally within the coding region (Meyer et al. , 2004). Such motifs are often involved with the regulation of stability, but also mediate enhancement (Antic et al. , 1999; Mazan-Mamczarz et al. , 2003) or repression (Baez and Boccaccio, 2005; Meng et al. , 2005) of translation when in the 5’UTR.
Two well-characterised motifs in humans are the AU-rich elements (AURE herein; more commonly ARE), and poly-cytosine (polyC) tracts. AURE- mediated decay has been suggested to explain the differences in half-lives observed in different transcripts, as this motif is characteristically found in mRNAs that undergo rapid decay (Chen and Shyu, 1995), although many relatively long-lived mRNAs also have AURE’s in their 3’UTR, such as β F1 ATPase (ATP5B), which has a half-life of 2-10 h (Izquierdo, 2006), and the AR mRNA, which has a half-life of 5-12 h in LNCaP cells (Yeap et al. , 1999). AURE-mediated decay is regulated, with stabilisation of AURE-containing mRNAs observed after heat-shock (Gallouzi et al. , 2000; Laroia et al. , 1999), hypoxia (Levy et al. , 1998), UV irradiation (Wang et al. , 2000b), cell proliferation (Wang et al. , 2000a), and exposure to lipopolysaccharides (LPS) (Lewis et al. , 1998) or pro-inflammatory cytokines (Piecyk et al. , 2000) (Suk and Erickson, 1996). The proteins that bind AURE’s and their functional significance are discussed in Section 1.4.2. Historically, AURE’s have been grouped into three classes, AURE classes I, II and III (Chen and Shyu, 1995; Chen et al. , 2002). AURE I motifs contain one to three AUUUA pentamers dispersed within a U-rich region, and are found in early-response genes such as c-fos. AURE II motifs contain multiple overlapping AUUUA pentamers within U-rich regions, and are found in cytokine mRNAs. AURE class III sequences do not have the AUUUA motif, but are U-rich. It has also been argued that the functional core motif involved in destabilisation of mRNA is a nonamer, UUAUUUA(U/A)(U/A), which includes the original pentamer (Lagnado et al. , 1994; Zubiaga et al. , 1995). This hypothesis is supported by
29 two recent studies that have also determined that a U-rich sequence with a minimum of nine residues is required for HuR binding (de Silanes et al. , 2004; Meisner et al. , 2004). This is further discussed in Chapter 6.
The AURE classes have been shown to be functionally distinct. The c-fos 3’UTR contains both a class I AURE, which destabilises the highly-stable β- globin mRNA when cloned into the 3’UTR; and a class III AURE that has no intrinsic destabilising effect, but enhances the effect of the AURE I when the two are combined (Chen et al. , 1994). A mechanistic insight was provided by the observation that the AURE III accelerated de-adenylation, and that mutation of the AURE I increased the stability of the body of the message without affecting the rate of de-adenylation (Chen et al. , 1994; Shyu et al. , 1991). This implied that there is a two-step pathway to AURE-mediated decay (Chen et al. , 1994). It also suggested that the different classes of AURE might be recognised by different proteins (Section 1.4.2).
The polyC motif (CCCYCCC, where Y is a pyrimidine) is found in the 3’UTRs of the highly stable globin genes, which can accumulate to ~ 95 % of cellular mRNA during terminal erythroid differentiation (Chkheidze et al. , 1999). The polyC motif is also associated with translational repression. PolyC tracts are bound by the polyC binding protein (PCBP) family, which includes the α- complex proteins ( αCP) 1-4 and hnRNPK (Makeyev and Liebhaber, 2002). The αCP’s were first identified as components of the α-complex, a mRNP that controls erythroid-specific accumulation of α-globin mRNA (Kiledjian et al. , 1999), and so are considered mRNA stabilisers, however PCBP’s are also involved in translational regulation of cellular and viral mRNAs (Makeyev and Liebhaber, 2002). αCP1 suppresses translation of 15-lipoxygenase (LOX), for example (Reimann et al. , 2002), and both αCP1 and αCP2 repress poliovirus mRNA translation (Silvera et al. , 1999).
Other sequence elements have been identified, but these for the most part are involved in quite specific processes. For example the iron response element (IRE), found in heavy- and light-polypeptide ferritin and in the transferrin receptor, is largely restricted to iron-regulation and response genes (Coulson
30 and Cleveland, 1993; Thomson et al. , 1999). In addition, there is some redundancy in the naming of RNA motifs. For example, the lipoxygenase differentiation control element (LOX-DICE), found in the LOX gene 3’UTR, could be described as 10 adjacent polyC motifs (Reimann et al. , 2002).
1.4.2 RNA-Binding Proteins in Post-Transcriptional Regulation.
RNA-binding proteins are typically modular in structure and contain one or more domains that specifically recognise mRNA. The commonest RNA- binding motifs are the RNA-recognition motif (RRM; Section 1.4.3), hnRNP K homology (KH) domain and arginine-glycine-glycine (RGG) box (Dreyfuss et al. , 2002). RNA-binding proteins are usually nucleo-cytoplasmic shuttling proteins, and while they tend to have a nuclear localisation in the steady state, they are in fact continuously shuttling (Dreyfuss et al. , 2002). These general features are also found in most of the proteins that recognise AURE and polyC tracts.
An increasing number of proteins have been reported to recognise AURE’s, including ARE/polyU-binding/ degradation factor 1 (AUF1, hnRNP D1) (Zhang et al. , 1993)), KH-type splicing regulatory protein (KSRP, FBP2) (Gherzi et al. , 2004), tristetraproline (TTP, ZPF36), butyrate response factor 1 (BRF1) (Lykke-Andersen and Wagner, 2005), T-cell intracellular antigen 1 (TIA1) (Izquierdo, 2006; Piecyk et al. , 2000), TIA-related protein (TIAR) (Izquierdo, 2006), and the Hu protein family (Wang and Tanaka-Hall, 2001). Of these, only the Hu proteins have a well-defined stabilising role (Section 1.4.3), and are the proteins of principle interest in this thesis. TIA1 and TIAR are involved in translational repression during the cellular stress response, but have not been shown to regulate mRNA turnover (Anderson and Kedersha, 2002; Kedersha and Anderson, 2002; Piecyk et al. , 2000). AUF1 exerts stabilising or destabilising effects depending on the cell type and physiological status (Loflin et al. , 1999a; Xu et al. , 2001), and the remainder of this group of proteins are generally regarded as promoting mRNA decay.
AUF1 was first isolated as two polypeptides (37 and 40 kDa) that bound specifically to the c-myc and granulocyte-macrophage colony-stimulating factor 31 (GM-CSF) 3'UTRs, and selectively accelerated degradation of c-myc mRNA in a cell-free decay system (Brewer, 1991). AUF1 exists as four isoforms, generated by alternative splicing: p37 AUF1 p40 AUF1 p42 AUF1 and p45 AUF1 and contains two RRM and one RGG box (Dreyfuss et al. , 2002; Wagner et al. , 1998; Zhang et al. , 1993). The 45 kDa and 42 kDa isoforms are primarily nuclear, while the 37 and 40 kDa isoforms continuously shuttle between the nucleus and cytoplasm (Arao et al. , 2000; Zhang et al. , 1993). AUF1 has also been isolated from the α-complex (Kiledjian et al. , 1997), although it is probably not directly associated with the CCCYCCC motif, as it has strict structural and sequence requirements for AURE’s (Fialcowitz et al. , 2005).
KSRP (FBP2) is one of four far-upstream element (FUSE) binding proteins (FBP), characterised by four repeated KH domains. FUSE is a regulatory DNA sequence far upstream of the c-myc promoter and FBP’s bind single-stranded DNA with an affinity and sequence specificity equal to or greater than those of RNA (Chung et al. , 2006). It has been suggested that FBP1 and 3 are restricted to the nucleus and have no role in regulating mRNA stability (Dean et al. , 2004). Conversely, KSRP has accepted roles in splicing (Chung et al. , 2006) and an emerging role as a regulator of mRNA stability (Briata et al. , 2005; Linker et al. , 2005).
TTP is a cysteine-cysteine-cysteine-histidine (CCCH) tandem zinc finger protein that binds to AURE II-containing transcripts and destabilises them, apparently by promoting de-adenylation (Lai et al. , 2003). It may also regulate transcription of tumour necrosis factor α (TNF α) (Zhu et al. , 2001a). Mice in which TTP has been knocked out have a severe inflammatory phenotype, which is caused by stabilisation of TNF α and GM-CSF mRNA’s and subsequent increased secretion of the protein (Blackshear, 2002; Carballo et al. , 1997). BRF1, another destabilising protein, shares the zinc finger domain, but few other sequence similarities (Dean et al. , 2004).
TIA and TIAR1 are RRM proteins, with closest homology to the Hu family (Section 1.4.3). There is limited evidence that they are regulators of cyclo- oxygenase 2 (COX2) RNA stability (Cok et al. , 2003). As TIA1 and TIAR have
32 important roles in other aspects of RNA biology, such as the stress response (Section 1.4.4), it is more likely their influence on COX2 protein levels is via translational repression (Dixon et al. , 2003).
1.4.3 Hu/ELAV Family of RNA Binding-Proteins.
The Hu antigen family of RNA-binding proteins are the human homologues of the Drosophila embryonic lethal abnormal visual (ELAV) gene, characterised by the presence of three very highly conserved RRM’s (~ 99 % amino-acid identity), with the second and third RRM being separated by a variable hinge region (Perrone-Bizzozero and Bolognani, 2002) (Figure 1.8A). RRM’s are a conserved single-stranded RNA-binding domain, with the RNA-interacting surface comprising four β-sheets packed against two α-helices (Figure 1.8B) (Chen and Varani, 2005; Maris et al. , 2005). The hinge region contains nuclear localisation (NLS) and nuclear export signals (NES), in HuR dubbed HNS for HuR Nucleo-Cytoplasmic Shuttling, and a similar region appears in HuD (Fan and Steitz, 1998b; Kasashima et al. , 1999). The hinge region probably also contains other functional elements which differ between Hu proteins. Replacing the hinge region of HuR with that of HuD in PC12 cells resulted in a protein that had cytoplasmic distribution, like wild-type HuD, but did not induce neurite formation as did wild-type HuD (Kasashima et al. , 1999). The Hu family contains four members, HuR (HuA, ELAVL1), which is ubiquitously expressed, and HuB (Hel-N1, ELAVL2), HuC (ELAVL3) and HuD (ELAVL4), which are typically restricted to neurones (Ma et al. , 1996).
Members of the Hu protein family have been implicated in mRNA splicing, export, translation and stability (Antic and Keene, 1998; Katsanou et al. , 2005; Mazan-Mamczarz et al. , 2003; Zhu et al. , 2006). The neuronal protein HuD, but not HuR, directly interacts with the major mRNA export apparatus, nuclear RNA export factor (NFX1, Tap) (Saito et al. , 2004). HuD has been shown to bind a diverse set of mRNAs, including those encoding neuroserpin (Cuadrado et al. , 2002), p21 Waf1 (Joseph et al. , 1998), acetyl-cholinesterases (Deschenes- Furry et al. , 2005a) and GAP-43 (Mobarak et al. , 2000). All of these transcripts are involved in neuronal differentiation, accounting for the observation that expression of neuronal Hu proteins is critical for neuronal differentiation
33 A Hinge
RRM1 RRM2 HNS RRM3
B
RRM2
c-fos 3’UTR
RRM1
Figure 1.8. Structure of Hu family proteins. A. Schematic diagram of a Hu family protein. The second and third RNA- recognition motifs (RRM; ) are separated by a variable hinge region ( ), which contains the nucleo-cytoplasmic shuttling signal (HNS). B. Ribbon diagram, showing an 11 nt fragment of the c-fos 3’UTR (ball-and-stick structure), interacting with the β-sheets ( ) of the first two RRM’s of HuD. . = α-helices. B. Reproduced from Wang et al. (2001).
34 (Aranda-Abreu et al. , 1999; Okano and Darnell, 1997). HuD has also been shown to bind the instability element in the 3’UTR of c-fos , although a functional consequence of this interaction has not been demonstrated (Chung et al. , 1996). Despite being a neuronal protein, HuD is also expressed in tumour types that have neuroendocrine features, such small cell lung cancer (SCLC) (Szabo et al. , 1991). SCLC patients frequently exhibit high serum titres of antibodies to this protein, resulting in the development of serious neurological autoimmune disorders (Dalmau et al. , 1990).
HuR also regulates the stability of a large number of mRNAs, including several genes that are important in cancer such as cyclin E1 (Guo and Hartley, 2006), β-catenin (de Silanes et al. , 2003), TNF α, vascular endothelial growth factor (VEGF) and IL8 (Nabors et al. , 2003). An extended list of known and proposed HuR targets is presented Table 1.2. Importantly, several of the targets of HuR are also genes which have been found to modulate AR expression or activity in prostate cancer (Section 1.3.3), such as cyclin D1, β- catenin, IL6, IGF receptor 1 (IGFR1), ERBB2, and the AR itself (Yeap et al. , 2002).
The stimuli for HuR binding to some of these transcripts has been identified. Ultra-violet C (UVC) irradiation stimulates the increased translation of the tumour suppressor p53, and HuR interacts with this mRNA in a UVC- dependent manner in RKO human colorectal cancer cells (Mazan-Mamczarz et al. , 2003). Another study has shown that HuR accumulates in the cytoplasm in response to UV light in normal keratinocytes, and this is associated with a 2- to 3-fold increase in the stability of the RhoB transcript (Westmark et al. , 2005). HuR also accumulates in the cytoplasm when RNAPII-mediated transcription is inhibited by treatment with actinomycin D
(ActD), and during S and G 2 phases of the cell cycle, a time when cyclins A and B exhibit greatest mRNA stability (Atasoy et al. , 1998; Fan and Steitz, 1998b; Wang et al. , 2000a). The finding that deletion of the hinge region results in HuR accumulation in the cytoplasm but does not affect the ability of HuR to stabilise target mRNA’s, and that cytoplasmic localisation of the neuronal Hu proteins stimulates neurite outgrowth suggests that cytoplasmic
35
Table 1.2. Targets of HuR . Reproduced from (Meisner et al. , 2004) and updated: 1: (Sheflin et al. , 2004). 2: (Leandersson et al. , 2006). 3: (Guo and Hartley, 2006). 4: (Cho et al. , 2006). 5: (Zhang et al. , 2006). 6: (Sommer et al. , 2005). 7: (Deschenes-Furry et al. , 2005b). 8: (Izquierdo, 2006). 9: (Jeyaraj et al. , 2005). 10: (Linker et al. , 2005). 11: (Song et al. , 2005). 12: (Westmark et al. , 2005). 13: (Kloss et al. , 2004). 14: (Mehta et al. , 2006). 15: (Meng et al. , 2005). 16: (Kawai et al. , 2006). 17: (Zou et al. , 2006). 18: (Okamoto et al. , 2004). 19: (Xu et al. , 2005). 20: (Pan et al. , 2005). 21: (Gantt et al. , 2005). 22: (Li et al. , 2005). *HuR binds 5’UTR. Gene Gene name: GenBank symbol: Accession: Cytokines, chemokines, growth factors
BMP6 Bone morphogenetic protein 6 NM_001718 CCL11 Chemokine (C-C motif) ligand 11, eotaxin NM_002986 CSF2 Colony stimulating factor 2, GMCSF NM_000758 1 EGF Epithelial growth factor NM_001963 FSHB Follicle stimulating hormone beta AH003599 IL1b Interleukin 1 beta NM_000576 IL2 Interleukin 2 NM_000586 IL3 Interleukin 3 NM_000588 IL4 Interleukin 4 NM_000589 IL6 Interleukin 6 NM_000600 IL8 Interleukin 8 NM_000584 MYOD1 Myogenic factor 3 NM_002478 MYOG Myogenin NM_002479 NF1 Neurofibromin 1 NM_000267
TNFa Tumor necrosis factor α NM_000594 VEGF Vascular endothelial growth factor NM_003376 WNT5A 2 Wingless-type MMTV integration site 5A NM_003392
Tumour suppressor genes, proto-oncogenes, cell-cycle regulators
CCNA2 Cyclin A NM_001237
CCNB1 Cyclin B1 NM_031966 CCND1 Cyclin D1 NM_053056 CCND2 Cyclin D2 NM_001759 3 CCNE1 Cyclin E1 NM_001238 CD83 CD83 antigen NM_004233 CDC2 4 Cell division cycle 2 NM_033379
CDKN1A Cyclin-dependent kinase inhibitor 1A, p21, NM_000389 Cip1 CDKN1B Cyclin-dependent kinase inhibitor 1B, NM_004064
p27,kip1 DEK DEK oncogene NM_003472 FOS v-fos FBJ murine osteosarcoma viral NM_005252
oncogene homologue, c-fos GADD45A 5 Growth arrest/DNA-damage inducible 45 α NM_001924 HLF Hepatic leukemia factor NM_002126
JUN v-jun sarcoma virus 17 oncogene NM_002228 homologue (avian), c-jun MYC v-myc myelocytomatosis viral oncogene NM_002467
homologue, c-myc MYCN v-myc myelocytomatosis viral related NM_005378 oncogene, neuroblastoma derived, n-myc TP53 Tumour protein p53 NM_000546 YES1 6 v-yes-1 Yamaguchi sarcoma viral M15990 oncogene homolog 1 Continued next page….
36
Gene Gene Name: GenBank Symbol: Accession: Enzymes ACHE 7 Acetylcholinesterase AY750146 ATP5B 8 F1 β ATPase NM_001686 ATP6V1E1 9 Vacuolar H +-translocating ATPase NM_001696 HDAC2 Histone deacetylase 2 NM_001527 INOS 10 Inducible nitrogen oxide synthase X73029 GCSh 11 γ-glutamylcysteine synthetase heavy chain M90656 MMP9 Matrix metalloproteinase 9 NM_004994 NDUFB6 NADH dehydrogenase (ubiquinone) 1 beta NM_002493 subcomplex NOS2A Nitric oxide synthase 2A NM_000625 PLAU Urokinase plasminogen activator NM_002658 PTGS2 Prostaglandin-endoperoxide synthase 2, NM_000963 COX2 REN Renin NM_000537 RHOB 12 RhoB GTPase NM_004040 SERPINB2 Serine (or cysteine) proteinase inhibitor, NM_002575 PAI-2 SGC 13 Soluble guanyl cyclase Y15723 UBE2N Ubiquitin-conjugating enzyme E2N NM_003348 Receptors, membrane proteins ADRB1 β1-adrenergic receptor NM_000684 ADRB2 β2-adrenergic receptor NM_000024 AR Androgen receptor NM_000044 CALCR Calcitonin receptor NM_001742 CDH2 Cadherin 2, type 1, N-cadherin NM_001792 ERBB2 14 v-erb-b2 erythroblastic leukemia viral NM_000448 oncogene homolog 2 (HER2/Neu) GAP43 Growth associated protein 43 NM_002045 IGFR1 15 * Insulin-like growth factor receptor 1 NM_000875 PLAUR Urokinase plasminogen activator receptor NM_002659 SLC2A1 Solute carrier family 2 member 1, GLUT1 NM_006516 SLC5A1 Solute carrier family 5, SGLT1 NM_000343 TNFSF5 Tumour necrosis factor (ligand) NM_000074 superfamily, member 5, CD154 Miscellaneous ACTG1 Actin, gamma 1 NM_001614 CTNNB1 Catenin (cadherin-associated protein), NM_001904 beta 1 CYTC 16 Cytochrome C BC024216 MARCKS Myristoylated alanine-rich protein kinase C NM_002356 substrate MTA1 Metastasis associated 1 NM_004689 NPM1 17 Nucleophosmin NM_002520 PITX2 Paired-like homeodomain transcription NM_000325 factor 2 SH2D1A 18 SH2 domain-containing 1A NM_002351 SLC7A1 Cationic amino acid transporter, CAT-1 NM_003045 SLC11A 19 Solute carrier family 11 member 1 NM_000578 Transcription Factors
ATF3 20 Activating transcription factor 3 NM_001030287 CEBP 21 CCAAT/enhancer binding protein NM_005194 HIF1A 1 Hypoxia-inducible factor 1 α NM_001530 HOX11 22 Homeobox protein 11 M75952
37 localisation of Hu proteins is associated with their RNA stabilising role (Aranda- Abreu et al. , 1999; Chen et al. , 2002; Kasashima et al. , 1999; Okano and Darnell, 1997). Recently, cytoplasmic localisation of HuR has been implicated in breast and ovarian cancers, where it was found to be an independent marker for poor prognosis (Erkinheimo et al. , 2003; Heinonen et al. , 2005). In a study of 455 ovarian cancer specimens, cytoplasmic HuR expression also correlated with COX2 expression, another marker of poor prognosis (Erkinheimo et al. , 2003). Similar results were observed in colon cancer, and RKO cells over- and under-expressing HuR were more and less tumorigenic in nude mice than control cells, respectively (de Silanes et al. , 2003).
1.4.3.1 Hu Binding Sites. Although initially thought to bind the AURE type I and II described in Section 1.4.1, increasing evidence suggests that HuR binding is relatively unrestrained, having affinity for sequences that are generally U-rich, i.e., class III AURE’s (de Silanes et al. , 2004; Fialcowitz et al. , 2005; Meisner et al. , 2004). To compare the effects of mRNA secondary structure on binding by AUF1 and HuR, Fialcowitz and colleagues modified the stability of an RNA hairpin structure based on that adopted by the TNF α 3’UTR, to which both HuR and AUF1 are known to bind. The hairpin fold was stabilised by introduction of U:C base pairs into the stem, or destabilised by introduction of non-canonical A:C base pairs into the stem. Increased stability of the hairpin was associated with decreased AUF1 binding affinity, and decreased stability with increased binding affinity. Contrastingly, HuR binding was minimally affected by changes in stability of the hairpin (Fialcowitz et al. , 2005). In another study comparing the binding requirements of AUF1 and HuR, it was demonstrated that AUF1 required the sequences flanking the AURE from its target 3’UTRs for high-affinity binding (dissociation constants of 30-90 nM), whereas in general, HuR bound short, isolated AU-rich sequences with dissociation constants of 20-50 nM (Blaxall et al. , 2002). Similar observations were made for HuD with co-crystal structures of the first two RRM’s of HuD with 11-nucleotide fragments of the class I AURE, c-fos, and the class II AURE of TNF α suggesting a preference for pyrimidine-rich sequences (Wang and Tanaka-Hall, 2001).
38 1.4.3.2 Modulation of Hu Protein Activity. Far less is known about regulation of Hu proteins than is known about their target mRNAs. Currently, it is thought a major mechanism regulating the stabilising activity of Hu proteins is their intracellular localisation. Many stimuli result in cytoplasmic accumulation of Hu proteins, including T-cell activation (Atasoy et al. , 1998; Wang et al. , 2006), and energy depletion (Jeyaraj et al. , 2006; Wang et al. , 2002), as well as those discussed in Section 1.4.3. Four protein ligands of HuR have been identified: SET α, SET β, phosphoprotein 32 (pp32) and acidic protein rich in leucine (APRIL) (Brennan et al. , 2000). These proteins also interact with chromosome region maintenance protein 1 (CRM1), a nuclear export factor. Inhibition of CRM1 export activity retains pp32 and APRIL in the nucleus, increases their association with HuR, and increases the association of HuR with nuclear polyA mRNA. These studies have elucidated the mechanisms regulating export of HuR from the nucleus, but not those regulating expression of HuR.
A limited number of studies have described regulation of expression of Hu family Hu proteins. HuD is transcriptionally repressed by thyroid hormone in rat PC12 cells and mouse N2a cells, and HuR protein is down-regulated by ~ 75 % following 48 h of DHT treatment in HepG2 human hepatocellular carcinoma cells (Cuadrado et al. , 2003; Sheflin et al. , 2001). In female mice, 6-48 h of DHT treatment causes total HuR levels in the sub-maxilliary salivary glands and kidney to fall progressively, while orchiectomy of male mice causes HuR levels to rise in these two tissues by 800 % and 200 %, respectively (Sheflin et al. , 2001). It was recently shown that HuR protein expression is increased during energy depletion due to increased translation of the HuR mRNA, despite overall HuR mRNA levels remaining unchanged (Jeyaraj et al. , 2006).
HuR protein can be methylated by co-activator-associated arginine methyl transferase 1 (CARM1) in response to LPS, although the consequences of this methylation have not been determined (Li et al. , 2002b). HuD is also a target of CARM1. When methylation of HuD is prevented by depletion of CARM1 in PC12 cells, the half-life of p21 Waf1 mRNA was extended, and the cells exhibited robust neuritogenesis and slow growth rates, which are typical
39 signs of neuronal differentiation. Treatment of the cells with nerve growth factor also decreased methylation of HuD (Fujiwara et al. , 2006). These studies indicate that methylation is a likely mechanism for modulation of the mRNA-stabilising activity of Hu proteins. Phosphorylation of Hu proteins has not been reported, however KSRP is phosphorylated in a p38 MAPK- dependent fashion, which impairs its ability to bind mRNA, thereby suggesting a mechanism by which the activity of any AURE-binding protein could be regulated (Briata et al. , 2005).
The Hu proteins also have different capabilities for forming multimers (Kasashima et al. , 2002). HuD and HuB can form multimers, an interaction independent of but enhanced by mRNA. HuC does not multimerise, and HuR can form dimers but not larger complexes. It is postulated that HuD/HuB form cytoplasmic complexes to facilitate the efficient up-regulation of stability and translation of target mRNAs (Kasashima et al. , 2002). As these interactions occur via the RRM’s, modification of this region could also affect the ability of Hu proteins to oligomerise, and may therefore affect the stabilising activity of different Hu proteins to different extents (Kasashima et al. , 2002).
1.4.4 Shared Targets, Different Outcomes.
AURE’s are generally associated with labile mRNAs (Section 1.4.1), yet the stabilising protein HuR binds these sequences. In a study of the relative affinities of p37 AUF1 and HuR for a diverse set of AURE-containing mRNAs, the relative affinities of AUF1 and HuR were found to be generally predictive of mRNA stability, with the exception of specific mRNA sequences (Blaxall et al. , 2002). Four explanations are possible: HuR may not bind the sequence; HuR binds the sequence but does not affect mRNA stability, or regulates mRNA stability only under certain circumstances; HuR and destabilising proteins bind discrete regions of the mRNA sequence; or HuR and destabilising proteins compete for the same sequence. Examples of each of these have been reported.
HuR does not stabilise c-jun mRNA, which contains a U-rich motif of the kind HuR would be expected to bind (Peng et al. , 1998). IL8, like c-fos mRNA (Section 1.4.1), has a core AURE class I destabilising domain and a decay- 40 enhancing AURE III auxiliary domain (Winzen et al. , 2004). HuR was shown to bind the auxiliary domain, but was not involved in stabilising IL8 mRNA in response to activation of the p38 MAPK pathway (Winzen et al. , 2004). In contrast, HuR binds and stabilises c-fos mRNA via interactions with the AURE III-containing auxiliary motif, while AUF1 binds the AUUUA-containing core decay element of this transcript (Chen et al. , 2002; DeMaria et al. , 1997). HuR and AUF1 can bind competitively to the same sites on p21 Waf1 and cyclin D1 mRNAs, as well as to distinct sites, while KSRP and HuR compete for the same binding site on iNOS mRNA (Lal et al. , 2004; Linker et al. , 2005). Antagonistic control of AURE-containing mRNA stability by BRF1 and HuR has also been demonstrated. When a cell line, stably transfected with a green fluorescent protein construct that contained an AURE, was treated with siRNA’s targeting HuR or BRF1, fluorescence decreased or increased, respectively, whereas this effect was abolished if the cells were treated with both siRNAs. This indicated that the observed changes in fluorescence were due to opposing, competitive functions of HuR and BRF1 (Raineri et al. , 2004). Further examples of common targets of HuR and other AURE-binding proteins are included in Table 1.3.
HuR also appears in mRNP complexes with TIA and TIAR on ATP5B mRNA (Izquierdo, 2006), with β-catenin on COX2 mRNA (Lee and Jeong, 2006), and with TIAR and KSRP on IL8 mRNA (Suswam et al. , 2005). The amount of HuR in the HuR/TIAR/KSRP/IL8 mRNP increased after stimulation with IL1 β, suggesting that the observed increase in IL8 mRNA stability might be due to enhanced HuR binding (Suswam et al. , 2005). Thus, as discussed in Section 1.4, the fate of an mRNA depends on the composition of the mRNP, which is dynamic, and the physiological state of the cell.
Under conditions of cellular stress, such as heat-shock or oxidative stress, translation is globally repressed, and transcripts necessary for cell survival receive the highest priority for translation (reviewed in (Kedersha and Anderson, 2002)). Other mRNAs are recruited to stress granules (aggregations of mRNP’s and translationally quiescent ribosomes), where they are not translated, but nor are they degraded (Anderson and Kedersha, 2002).
41
Table 1.3. Common targets of HuR/HuD and other AURE-binding proteins. Adapted from (Barreau et al. , 2005).
Target mRNA HuR HuD AUF1 BRF1 TIA1 TIAR TTP KSRP c-myc c-fos p21 Cyclin D1 NOSII/iNOS GM-CSF TNF α COX2 IL2 IL8 VEGF GLUT1 GAP43
42 TIA1 is critical for stress granule formation, and HuR and TTP, but not AUF1, are component proteins of stress granules (Kedersha and Anderson, 2002). This illustrates that the sub-cellular localisation of an mRNA also influences which proteins can bind it.
Finally, the tissue distribution of several of the AURE-binding proteins has been established in the mouse. HuR and AUF1 were both found in the thymus, spleen, and gonadal tissue, with HuR also in the intestine and AUF1 in the epithelial linings of lungs and nuclei of most neurons in the brain. TTP expression largely did not overlap with that of AUF1, while KSRP was widely distributed (Lu and Schneider, 2004). Therefore, the differential tissue distribution may limit which AURE-binding proteins a transcript may encounter. Moreover, AUF1 and HuR are differentially regulated by DHT treatment in mouse kidney cells. Cytosolic levels of AUF1 protein are decreased in mouse sub-maxillary salivary gland cells following DHT treatment, but not affected in kidney cells, whereas the HuR protein level is decreased by DHT treatment in both of these tissues (Sheflin et al. , 2001).
1.4.5 Mechanisms of AURE-Mediated Decay.
AURE-binding proteins are not themselves enzymes, therefore their role in mRNA decay is more likely to arise from interactions with the degradation machinery. Indeed, Chen and colleagues found that the mammalian exosome does not recognise ARE-containing mRNAs on its own but rather requires certain ARE binding proteins, such as KSRP and TTP that can interact with the exosome and recruit it to unstable mRNAs to promote rapid degradation (Chen et al. , 2001). This is reinforced by the finding that KSRP and TTP co- immunoprecipitate with the exosome (Linker et al. , 2005). However, mRNA degradation by either the exosome or Xrn1 depends on prior de-adenylation (Section 1.4.1). KSRP, which has been shown to interact with both PARN and the exosome via its KH domains, may sequentially regulate de-adenylation and 3’ exonucleic degradation (Gherzi et al. , 2004).
In cell-free systems, TTP and related proteins ( i.e. BRF1) stimulated the de- adenylation of ARE-containing, poly-adenylated transcripts by promoting the activity of PARN (Lai et al. , 2003). Another protein, CUG-BP, which was 43 thought to be primarily involved in splicing and translation, has recently been shown to recruit PARN to UG-rich transcripts as well as to AURE-containing transcripts (Marquis et al. , 2006; Moraes et al. , 2006). The stabilising activity of HuD has been suggested to depend on the length of the polyA tail, and to decrease the rate of de-adenylation in vitro (Beckel-Mitchener et al. , 2002). There has, however, been one report of an endonuclease that cleaves within an HuR binding site in the p27 Kip1 mRNA, and cleavage at this site is inhibited by HuR (Zhao et al. , 2000b). These observations are particularly interesting because there has been little direct evidence that PARN participates in regular mRNA turnover (Meyer et al. , 2004). PARN may be involved specifically in motif-mediated mRNA decay, where it is recruited by RNA-destabilisers, which then introduce the de-adenylated mRNA to the exosome (Chen et al. , 2001).
1.5 Post-Transcriptional Regulation of the Androgen Receptor.
1.5.1 Androgen Regulation of the AR mRNA Turnover.
The stability of the AR mRNA is increased by DHT in the prostate cancer cell line LNCaP, and decreased in the breast cancer cell line, MDA-MB 453 (Wolf et al. , 1993; Yeap et al. , 1999) (Section 1.2.4). The increased (2-fold) stability of the AR mRNA in LNCaP cells following DHT treatment has been attributed to elements in the 3’UTR, bound by αCP1, αCP2, and HuR (Yeap et al. , 2002). Interestingly, the sub-cellular localisation of HuR has been shown to be affected by DHT (Sheflin et al. , 2001). Treatment of HepG2 human hepatocellular carcinoma cells with DHT for 48 h increased the association of HuR with polyribosomes by ~ 3-fold, where it could potentially be involved in post-transcriptional regulation (increased stability and translation) of androgen- responsive transcripts (Sheflin et al. , 2001).
1.5.2 Structure of the AR mRNA.
The structure of the AR transcriptional unit was determined by Faber and colleagues, and is as follows: The main AR transcript is ~ 10.5 kb, with a second 8.5 kb product observed in LNCaP cells (Faber et al. , 1991) (Figure 1.9A). The shorter transcript is the result of alternative splicing of the 3’UTR,
44 A
5’ UTR Coding region 3’ UTR 4315 - 10500 1 1116 3878 4314 ARUC 21-149 4031 4082 4315-5650 8685-10500
GenBank sequence
B ARUC
CUGGGCUUUUUUUUUCUCUUUCUCUCCUUUCUUUUUCUUCUUCCCUCCCUA
U-rich sequences
Figure 1.9. Structural elements of the androgen receptor mRNA. A. The AR mRNA, showing the UTR’s ( ) and coding region ( ). The stem-loop involved in translation initiation is indicated ( ), as is the ARUC region. Features are numbered from the beginning of the GenBank transcript NM_000044, and the alternatively spliced 3’UTR’s ( ) are downstream of the GenBank sequence. B. Sequence of the ARUC region. The U-rich tracts to which HuR binds are indicated.
45 where 3 kb has been removed between nucleotides 5650 and 8685, down- stream of the transcription start site. Smaller products have been reported and are postulated to be the result of degradation. The coding region is 2.7 kb, and is flanked by long 5’ and 3’UTR’s of 1.1 kb and 6.8 kb, respectively (3.8 kb in the shorter transcript). The GenBank sequence (NM_000044) for the AR mRNA is not complete on the 3’ end, terminating at nucleotide 4314. There are two transcription initiation sites, 13 nucleotides apart, and two polyA signals, 221 nucleotides apart. The proximal polyA signal gives rise to multiple sites of polyA addition.
The major element in the first 436 nucleotides of the 3’UTR is postulated to be the androgen receptor UC-rich (ARUC) region (Figure 1.9B). The ARUC is a 51 nucleotide sequence composed of two AURE class III motifs, and one polyC motif, and is conserved in rats, mice and humans (Yeap et al. , 2002). Mutational analysis and RNA gel-shift experiments suggested that these sites are bound by two HuR proteins, and one of αCP1 or αCP2, however the effect of the binding of these proteins was not determined (Yeap et al. , 2002). There is a second polyC tract outside of the ARUC region, but this is not conserved between species (Yeap, 2000; Yeap et al. , 2002).
A stem-loop structure has been identified in the AR 5’UTR, which can act as a translation induction sequence in recombinant translation assays (Mizokami and Chang, 1994). The 5’UTR contains a class III AURE and two polyC motifs. Interestingly, the binding of αCP proteins to the 5’UTR of the ferritin mRNA has been associated with enhanced translation of that transcript (Thomson et al., 2005) although no such association has been found for the AR mRNA to date. Exon 1 of the AR contains CAG-repeats, which form the polyQ tract in the translated protein (Section 1.2.1). The CAG repeats form a stable stem-loop structure in the mRNA, and is bound by a number of cytoplasmic proteins in androgen-sensitive tissues, including Epstein-Barr virus early region (EBER)-associated protein (EAP) and the p110 subunit eIF3 (Yeap, 2000). Again, the functional significance of this on AR mRNA stability has not been determined.
46 1.6 Summary and Statement of Aims.
1.6.1 Summary.
The AR and AR signalling pathways are important regulators of prostate cancer cell proliferation and significant targets of current prostate cancer therapies. Persistent or recurrent AR signalling is associated with the development of androgen-independent prostate cancer and may be mediated via AR alterations including AR gene mutations and amplification, or changes in the relative expression of the AR and its co-regulators. AR expression is regulated transcriptionally and post-transcriptionally, with translational and in particular post-translational mechanisms contributing to final cellular AR levels and thus activity. The Hu family of RNA-binding proteins has been implicated in mRNA splicing, export, translation and stability and can exert very broad effects on cell function via interaction with multiple mRNA’s. HuR has been shown to bind a UC-rich sequence in the 3’UTR of the AR mRNA however the consequences of this binding are unknown. In addition, the expression of Hu proteins in the prostate and changes in their expression in prostate cancers are uncharacterised. The aims of this thesis were to investigate the activity of the Hu proteins in prostate cancer by determining (i) expression of HuD in human prostate cancer specimens, (ii) the interaction between HuR and HuR with the AR mRNA and (iii) HuR target mRNA’s in prostate cancer cells.
1.6.2 Hypotheses:
The hypotheses of this project were as follows: 1. HuD is aberrantly expressed in prostate tumours and contributes to regulation of AR expression and function. 2. HuR regulates AR expression and activity in prostate cancer cells. 3. HuR regulates the proliferation rate of prostate cancer cells. 4. mRNA targets of HuR in prostate cancer cells are involved in major growth regulatory and signalling pathways.
47 1.6.3 Research Aims.
In order to address these hypotheses, research was undertaken with the following aims: 1. To investigate the expression of HuD in primary prostate tumours and prostate cancer cell lines. 2. To identify HuD binding to the AR mRNA and characterise HuD effects on AR expression and function. 3. To determine the effects of HuR on AR mRNA and protein expression. 4. To elucidate the effects of changes in HuR expression on AR expression transcriptional activity. 5. To examine the intra-cellular localisation of HuR in prostate cancer cells. 6. To characterise changes in global gene expression following modulation of HuR protein level. 7. To identify novel targets of HuR in prostate cancer cells.
48
CHAPTER 2: Materials and Methods.
49 2 Chapter 2: Materials and Methods.
2.1 General Reagents. All general chemicals were of molecular biology grade or higher, and sourced from ICN Biomedicals, Sigma, or other reputable supplier. 1 kb plus DNA molecular weight marker Invitrogen, Australia
γ-32 P-uridine triphosphate (UTP) Amersham, Australia
Bio-Rad Protein Assay Reagent Bio-Rad, Australia
Bovine Serum Albumin (BSA) solution (100 Promega, Australia Mg/mL) BSA powder Sigma, USA
Deoxyribonucleotide triphosphates (dNTP’s), Invitrogen, Australia 100 mM Deoxyribonuclease (DNase) I Promega, Australia
DH5 α cells Stratagene, USA
Dithiothreitol (DTT), 0.1 M Invitrogen, Australia
Enhanced Chemiluminescence detection Amersham, Australia reagent E. coli BL21 Codon Plus Stratagene, USA
Filter paper 3M, USA
First Strand Buffer, 5 x Invitrogen, Australia
Gel purification kit, Wizard Promega, Australia
Glutathione-agarose beads Sigma, USA
Glycerol Sigma, Australia
GlycoBlue Ambion, USA
Glycogen Invitrogen, Australia
Heparin Sigma, USA
Hoescht 33258 Sigma, USA
Isopropyl β-D- thio-galactopyrosanide (IPTG) Sigma, USA
50 Kaleidoscope molecular weight markers Bio-Rad, Australia
Klenow fragment and buffer (10 x) Promega, Australia
LDS Sample Buffer Invitrogen, Australia
Low-fade mounting medium UWA Biomedical Imaging and Analysis Facility Light-sensitive film (Hyperfilm) Kodak, Australia
Luciferase Assay and Dual Glo kits Promega, Australia
MgCl 2, 50 mM Invitrogen, Australia
MicroSpin G-25 clean-up column Amersham, Australia
Plasmid purification kits, mini (small scale) Qiagen, Australia and maxi (large scale) NuPage 10 % Bis-Tris polyacrylamide gels Invitrogen, Australia
NuPage 3-(N-morpholino)propanesulphonic Invitrogen, Australia acid-sodium dodecyl sulphate (MOPS- SDS) running buffer (10 x) PCR tubes, flat-capped for quantitative PCR Axygen, Australia (qPCR) Platinum SYBR Green qPCR SuperMix- Invitrogen, Australia uracil DNA-glycosylase (UDG) Polyvinyldifluoride (PVDF)-plus membrane Roche, Australia
Primers GeneWorks, Australia
Protein A sepharose beads, pre-swollen Sigma, USA
Random hexamers Invitrogen, Australia
Restriction endonucleases and buffers Promega, Australia
Ribonucleotide triphosphates (NTP’s), 100 Invitrogen, Australia nM RNaseOUT Invitrogen, Australia
ROX reference dye Invitrogen, Australia
SuperScript II reverse transcriptase (RT) Invitrogen, Australia
T4 DNA ligase and buffer (5 x) Promega, Australia
51 T7 RNA polymerase and buffer (5 x) Invitrogen, Australia
Taq polymerase and buffer (5 x) Invitrogen, Australia
Triton X-100 Bio-Rad, Australia
Trizol Invitrogen, Australia
Trypan blue Sigma, USA
Tween-20 Sigma, USA
Yeast transfer RNA (tRNA) Sigma, USA
2.2 Tissue Culture Materials. 22Rv1 cell line Dr R Minchin, University of Queensland Actinomycin D-mannitol 1mg:49 mg Sigma, USA
Cell sieves BD Falcon, USA
CellTiter 96 Aqueous One Solution Promega, Australia cell proliferation assay reagent Dimethyl sulphoxide (DMSO) Sigma, Australia
LipoFectamine 2000 Invitrogen, Australia
LNCaP cell line American Type Culture Collection, USA Fibronectin Roche, Australia
Flasks (175 and 75 cm 2) Nunc, Denmark
Foetal calf serum (FCS) Gibco, Australia
Optimem medium Invitrogen, Australia
Penicillin Gibco, Australia
Phosphate-buffered saline (PBS) Invitrogen, Australia
Plates (6- and 24-well) TTP, Switzerland
Roswell Park Memorial Institute Invitrogen, Australia (RPMI) 1640 medium
52 Small interfering RNA (siRNA): Stealth siHuR-UTR and siCon Invitrogen, Australia siSully and siMut Dharmacon, USA
Streptomycin Gibco, Australia
Trypsin Invitrogen, Australia
2.3 Antibodies. All antibodies were used at experimentally established dilutions as follows, unless otherwise stated: Actin AC-15 mouse monoclonal Abcam 1/5 000
Alexa-Fluor 488-conjugated goat Molecular Probes 1/1 000 anti-mouse IgG secondary antibody AR 441 mouse monoclonal Santa Cruz 1/1 000
Horseradish peroxidase- Amersham 1/10 000 conjugated sheep anti-mouse HuD 16C12 Dr H Furneaux, University 1/500 of Connecticut HuR 3A2 mouse monoclonal Santa Cruz 1/5 000
M2 anti-FLAG mouse monoclonal Sigma 1/5 000
2.4 Equipment and Software. ChemiDoc XRS Bio-Rad, Australia
DP Controller Olympus, Australia
Fluorescence microscope Olympus, Australia
Fluostar Optima BMG Technologies, Germany
Gel dryer Bio-Rad, Australia
Gene Pulser Bio-Rad, Australia
53 GeneSifter VizX Labs, USA
mfold http://www.bioinfo.rpi.edu/ applications/mfold/rna/form1.cgi MRC 1000 confocal microscope Bio-Rad, Australia
PhosphorImager 445 Si Molecular Dynamics
Rotor-Gene and software Corbett Research, Australia
Scintillation counter Beckman, Australia
Thermal cyclers (PTC-200 DNA MJ Research, USA engines) Quantity One 1-D analysis Bio-Rad, Australia software
2.5 Plasmid constructs. pBS-ARUC-WT and pBS-ARUC mutants have been described previously (Yeap et al. , 2002). Briefly, a 51-nucleotide UC-rich section of the human AR 3’UTR, highly conserved in the rat, mouse and human AR genes, was selected for evaluation as a target for RNA-binding proteins. This sequence was cloned + into the multiple cloning site of the pBluescript KS vector (Appendix 2A). Mutations, listed in Appendix 2A and Figure 3.4A were introduced into the sequence, and these were also ligated into pBluescript KS + (Appendix 2A). To generate luciferase reporter vectors, the pGL3-MCS vector was modified as described in (Giles et al. , 2003) to facilitate directional cloning into the 3’UTR (Appendix 2B). The wild-type (WT) and mutant (MutA, MutB) AR UC-rich sequences were then excised from pBluescript KS + and ligated into pGL3- MCS to create the Luc-ARUC-WT, Luc-ARUC-MutA and Luc-ARUC-MutB vectors (Appendix 2C).
The pBS-c-fos -HuD plasmid, originally described in Giles et al. (2003), contained an AU-rich element from the human c -fos mRNA (5’- ATA TTT ATA TTT TTA TTT TAT TTT TTT –3’) cloned into pBluescript KS + (Appendix 2D).
54 This sequence was also subsequently cloned into pGL3-MCS to generate the Luc-c-fos vector (Appendix 2E). Lisa Stuart and Dominic Voon (Laboratory for Cancer Medicine, Western Australian Institute for Medical Research [LCM, WAIMR]) generated these clones. The Luc-c-fos vector was from Dr Keith Giles (Giles et al. , 2003).
The vector FLAG-HuR was a gift from Dr H Sakamoto, Kobe University, and comprised amino-acids 2-326 from the human HuR open reading frame cloned into the Hind III and Sal I sites of FLAG-CMV-2 (Kasashima et al. , 2002) (Appendix 2F). A FLAG-CMV-2 vector without insert (FLAG-empty) was generated by restriction endonuclease digestion with 1 U each of Hind III and Sal I for 30 min at 37 °C in 1 x Buffer D to remove the HuR sequence. The resultant overhangs were then filled with Klenow fragment (10 U, 37 °C for 30 min in 1 x Klenow buffer) and the ends of the vector religated (10 U of T4 ligase in 1 x ligase buffer overnight at room temperature) (Appendix 2F).
Amino-acids 2-373 of human HuD were subcloned by Britt Granath (LCM, WAIMR) from pGEX-2T-HuD (Appendix 2G) (Chung et al. , 1996) into the Bam HI sites of pGEX-6P2 and FLAG CMV7.1 to generate the plasmids pGEX- 6P-HuD (Appendix 2H) and FLAG-HuD (Appendix 2I), respectively.
The PB3-Luc vector, a gift from Dr R Matusik, Vanderbilt University, contained three copies of the androgen response elements (-244 to –96) from the rat probasin promoter, cloned into the T81-Luc vector with Hind III linkers (Appendix 2J) (Rennie et al. , 1993). Dr EM Wilson, University of North Carolina, provided the PSA-Luc plasmid, an androgen-responsive reporter that contains the 496 base-pair core enhancer from the PSA promoter, and the E4 TATA box, cloned into the Xho I and Sac I sites of the pGL3 Basic promoter (Appendix 2K). phRL-SV40 (Appendix 2L) and pRL-TK (Appendix 2M) Renilla luciferase vectors were from Promega.
55 2.6 Preparation of Plasmid DNA.
2.6.1 Preparation of E. coli.
Electro-competent E. coli DH5 α and XL-2 blue cells prepared by Lisa Stuart or Keith Giles (LCM, WAIMR) were used throughout for propagation of plasmids.
Competent cells were transformed with ~100 ng plasmid DNA using electroporation in clean, chilled electroporation cuvettes at 20 µF, 2.5 kV and 200 Ω in a Gene Pulser. After pulsing, cells with a time constant of > 4.5 ms were rescued by addition of 1 mL Luria Bertani (LB) broth 28 , and incubated for 30 min at 37 °C with shaking. Cells were plated on LB agar 27 containing 100 µg/mL ampicillin 4. After growth overnight at 37 °C, single colonies were streaked onto a fresh LB agar plate containing 100 µg/ µL ampicillin. Colonies from this plate were picked into fresh LB broth 28 containing 100 µg/mL ampicillin 4 for small- or large-scale plasmid preparations as required (Section 2.6.2).
For long-term storage, colonies were picked into 5 mL LB broth 28 , grown overnight at 37 °C with shaking, then 800 µL of culture was mixed with 200 µL of sterile glycerol and frozen at –80 °C in cryotubes.
2.6.2 Plasmid DNA Extraction.
Plasmid DNA was propagated as in Section 2.6.1. Large and small-scale preparations of plasmid were performed using Qiagen Plasmid Mini and Qiagen Plasmid Maxi plasmid purification kits as directed by the manufacturer. Briefly, 500 mL (large scale) or 5 mL (small scale) LB broth 28 containing 100 µg/mL ampicillin 4 was inoculated with a single bacterial colony and incubated overnight at 37 °C with shaking. The following day, cells were pelleted by centrifugation at 6 000 x g for 10 min (large scale) or 12 000 x g for 1 min (small scale). Medium was removed by aspiration and the pellet resuspended in 10 mL (large scale) or 250 µL (small scale) of P1 buffer. Cells were lysed in 10 mL (large scale) or 250 µl (small scale) of P2 lysis buffer, and incubated for
56 5 min. Ten mL (large scale) or 350 µL (small scale) neutralisation buffer (N3) was added and the solution was mixed by inversion. Insoluble debris was pelleted by centrifugation at 15 000 x g for 30 min (large scale) or 12 000 x g for 1 min (small scale), and the supernatants removed to fresh tubes. Supernatants were then applied to purification columns, and washed twice with QC buffer (30 mL for large scale and 500 µL for small scale). For small scale preparations, columns were centrifuged at 12 000 x g for 30 s between washes, then the DNA was eluted from the columns with 50 µL of deionised distilled (dd) H 20 and stored at –20 °C. For large scale preparations, DNA was eluted using E1 elution buffer (15 mL) and precipitated by the addition of 10.5 mL of isopropanol, followed by centrifugation for 30 min at 10 000 x g. The resulting DNA pellet was washed once with 70 % v/v ethanol, re-centrifuged for 30 min at 10 000 x g and the pellet allowed to air dry for 15 min before resuspension in 500 µL of ddH 20 and storage at –20 °C.
Plasmid DNA having a concentration of less than 200 ng/ µL was not used for transfection (Section 2.8).
2.6.3 DNA Sequencing.
Plasmid DNA or PCR products were diluted to 100 ng/µL, and primers were diluted to 1 pmol/ µL. Sequencing reactions and electrophoresis of products were performed by West Australian Genome Resource Centre, Department of Immunology, Royal Perth Hospital, using Big Dye Terminator V3.1. Chromatograms were checked by eye, and indeterminate base calls corrected where possible. Plasmid sequences were aligned with the published GenBank sequence using “BLAST 2 Sequences” (http://www.ncbi.nlm.nih.gov /blast/bl2seq/wblast2.cgi) and plasmids were used only if the entire coding region aligned perfectly with the GenBank sequence.
57 2.7 Cell Culture.
2.7.1 Maintenance of 22Rv1 and LNCaP Cells.
Both LNCaP human prostate carcinoma cells and 22Rv1 human prostate adenocarcinoma cells were cultured as advised by the American Type Culture Collection (ATCC). For routine maintenance, cells were cultured in phenol-red free RPMI 1640 media with 2 mM L-glutamine (pH 7.4), supplemented with 2 g/L sodium bicarbonate, 10 % v/v foetal calf serum (FCS), 50 U/mL penicillin and 50 mg/L streptomycin.
Cells were cultured at 37 °C, 5 % CO 2 in a humidified incubator and sub- cultured at 1:5 dilution following trypsinisation approximately once weekly, with the medium replaced twice weekly. For long-term storage, cells at 80 % confluence were trypsinised as normal, centrifuged at 1 000 x g for 1 min and resuspended in 3 mL of RPMI 1640 containing 50 % v/v FCS / 10 % v/v DMSO per 175 cm 2 flask. Cells (1 mL) were aliquoted into cryotubes and frozen under liquid nitrogen.
For experiments requiring low background levels of steroid hormones, cells were passaged into RPMI 1640 supplemented with 10 % charcoal stripped FCS at the time of plating (Section 2.7.2).
2.7.2 Charcoal-Stripped Foetal Calf Serum.
FCS (500 mL) was thawed overnight at 4 °C, and 2 mL was set aside for later biochemical assay. The remainder was incubated with dextran-coated charcoal 10 with periodic agitation for 3 h at room temperature. FCS was recovered from the suspension by centrifugation at 1 800 x g for 20 min at 4 °C. The supernatant was decanted and re-centrifuged. The FCS was then filtered in a laminar-flow hood, first through a 0.8 µm-pore membrane then through a 0.2 µm-pore membrane into a sterile bottle. An aliquot was set aside for biochemical analysis.
58 2.7.3 Biochemical Analysis of Foetal Calf Serum.
Samples of charcoal-stripped and unstripped FCS were assayed for testosterone, oestradiol and cortisol to ascertain the effectiveness of charcoal stripping. This was performed by Mario Taranto (Core and Clinical Biochemistry, Royal Perth Hospital). Typical analysis showed these hormones to be below detectable limits (Table 2.1).
Table 2.1. Typical analysis of hormone before and after stripping. Hormone Before stripping After stripping Cortisol <6 nmol/L <6 nmol/L Testosterone 0.3 nmol/L <0.3 nmol/L Progesterone <1 nmol/L <1 nmol/L Oestrogen 36 pmol/L <20 pmol/L
2.8 Transfections.
2.8.1 Plasmid Transfections.
LNCaP or 22Rv1 cells at 80 % confluence were trypsinised and resuspended in a small volume of FCS- or charcoal-stripped-FCS-supplemented growth medium, as appropriate. Where necessary, cell clumps were removed by straining through a 40 µm cell sieve. An aliquot of cells was diluted 2-fold in 0.4 % trypan blue stain 52 , and live cells were counted using a haemocytometer. Cells were then diluted in the appropriate medium to a final density of 5.5 x 10 4 cells/cm 2. LNCaP cells were incubated overnight before transfection the following day, while 22Rv1 cells were transfected at the time of plating.
LipoFectamine 2000 was used to deliver plasmid into cells. Transfection complexes were formed as directed by the manufacturer, using a ratio of 3 µL of Lipofectamine 2000 for every 2 µg of DNA. Typically, 1 µg of plasmid was transfected per well of a 6-well plate (9 cm 2) and scaled appropriately for wells of differing sizes, e.g. 0.225 µg per well of a 24-well plate. When Renilla luciferase was used as an internal transfection control, the plasmid was added to the transfection mixture at one-tenth of the concentration of the test plasmid. Cells were harvested 24 h after transfection (Section 2.10). The vector FLAG-
59 CMV7.1 without insert served as the empty vector control for experiments using FLAG-HuD, and the vector FLAG-CMV2 served as the empty vector control for experiments using FLAG-HuR (Section 2.5).
2.8.2 Small-Interfering RNA Transfection.
For small-interfering (si)RNA transfection, LNCaP or 22Rv1 cells were prepared as in Section 2.8.1, with the exception that cells were plated at a density of 2.5 x 10 4 cells/cm 2 (30-50 % confluence). LipoFectamine 2000 was used for siRNA transfection at a ratio of 1 µL of LipoFectamine 2000 for every 25 pmol of siRNA. The morning after siRNA transfection, media on 22Rv1 cells were routinely changed to remove LipoFectamine 2000, but media on LNCaP cells were left unchanged. Both LNCaP and 22Rv1 cells were left for 72 h following siRNA transfection before harvesting for protein or RNA extraction (Sections 2.13.2 and 2.14.1, respectively). Effective siRNA concentrations for each target were established experimentally, and were 20 nM for siHuR-Sully and siMut and 50 nM for Stealth siHuR-UTR and siCon. siRNA sequences are listed in Table 2.2.
For combined transfections, where siRNA and plasmids were used, the cells were transfected twice. At 72 h before harvest, the cells were transfected with siRNA as above, and the media changed at 24 h before harvest, when the plasmid DNA was transfected as per Section 2.8.1.
All samples were set up at least in duplicate, and most commonly in quadruplicate.
Table 2.2 SiRNA sequences. *https://rnaidesigner.invitrogen.com/rnaiexpress/ Name Sequence (sense strand) Manufacturer Reference BLOCK-iT RNAi siHuR-UTR CCA AUG GGA AUG GAC CAA AGA GUU U Invitrogen Designer* SiCon (non- BLOCK-iT RNAi CGU AUC ACU CCC UCC AAG ACA CAU G Invitrogen targeting) Designer SiHuR-Sully CUA CCU CCC UCA GAA CAU G Dharmacon (Sully et al. , 2004) SiMut (non- (van der Giessen AAG CCA AUU CAU CAG CAA UGG Dharmacon targeting) et al. , 2003)
60 2.9 Cell Titration Assays. The CellTiter 96 Aqueous One Solution cell proliferation assay was used to assess proliferation of 22Rv1 cells after treatment with siRNA. Cells were plated at a density of 2.5 x 10 4 cells/cm 2 in 96-well plates, in quintuplicate, and transfected at the time of plating as described in Section 2.8.2. In a variation to the manufacturer’s protocol, the cells were diluted to 2.72 x 10 4 cells/mL in RPMI 1640 supplemented with 10 % charcoal-stripped FCS, and 180 µL of cell suspension were added to each well. To each well, 20 µL of transfection mix, containing siHuR-UTR siRNA or siCon siRNA, was added to make a final concentration of 50 nM. To reduce handling and therefore well-to-well variation, media were not changed throughout. Cell proliferation was measured at 24 h intervals for 5 days, by adding 20 µL of CellTiter reagent, incubating under normal growth conditions for 1 h, and reading the absorbance at 490 nM in a Fluostar Optima platereader.
Doubling times of 22Rv1 cells were estimated by calculating the value of x in the growth equations in Figures 4.14A and B when y=0.1 and when y=0.2. Subtracting the value of x when y=0.1 from the value of x when y=0.2 was taken as the length of one doubling time, in h.
2.10 Luciferase Assays. All solutions were prepared as directed by the manufacturer. LNCaP or 22Rv1 cells that had been transfected with firefly and Renilla luciferase reporter vectors (Section 2.8.1) were lysed in 1 x Passive Lysis Buffer (PLB) to produce lysates for use in Dual Luciferase assays. For 6-well plates (9 cm 2 wells), 150 µL of PLB were used, and for 24-well plates (2 cm 2 wells), 45 µL of PLB were used. Twenty µL of lysate and 50 µL of Luciferase Assay Reagent II (LARII) were included in each sample. LARII was added by injection, and the luminescence over five seconds measured in a Fluostar Optima platereader. In those instances where Renilla luciferase was to be used as an internal control, 50 µL of Stop+Glo was added by injection and the luminescence measured again. LARII and Stop+Glo were prepared in accordance with the manufacturer's instructions. Relative luciferase readings were calculated by
61 dividing the firefly luciferase measurement for each sample by that of Renilla luciferase.
2.11 mfold Prediction of mRNA Secondary Structures. The online programme mfold (Zuker, 2003) (http://www.bioinfo.rpi.edu /applications/mfold/rna/form1.cgi) was used to predict secondary structure of ARUC-Luc constructs. The sequence input into the programme (Figure 2.1) included the entire 3’UTR from the ARUC-Luc construct, from the stop codon to the late poly(A) signal, to most closely represent the transcript in vivo. Structure was predicted under the following conditions: 5 % suboptimality; maximum interior/bulge loop size = 30; maximum asymmetry of interior bulge/loop = 30; maximum distance between paired bases = no limit; 37 °C; 1 M NaCl; no divalent ions.
Figure 2.1. The sequence of the 3’UTR of Luc-ARUC vector used for mfold structure determination. The stop codon of the Luc gene is bold and underlined, the restriction sites used in cloning (Giles et al. , 2003) are boxed and the polyA signal is underlined.
Xba I Spe I Bam HI Hin DIII LUC taa -ttctagaact agtggatcc-ARUC -aagcttatcg ataccggtcg Apa I Xba I acctcgaggg ggggccctct agagtcgggg cggccggccg cttcgagcag
acatgataag atacattgat gagtttggac aaaccacaac tagaatgcag
tgaaaaaaat gctttatttg tgaaatttgt gatgctattg ctttatttgt
aaccattata agctgcaata aacaagttaa caacaacaat tgcattcatt
ttatgtttca ggttcagggg gaggtgtggg aggtttttta aagcaagtaa
aacctctaca aatgtggta .
2.12 Identification of HuR Binding Sites. Potential HuR binding sites were identified by modelling 3’UTRs from genes identified by micro-array (Section 2.15.1) using mfold (Section 2.11). Single- stranded regions containing the motif NNUUNNNUUUU (Meisner et al. , 2004) were considered candidate targets of HuR. Where the accession number provided by GeneSifter (Section 2.15.1) corresponded to an expressed sequence tag rather than an annotated complete coding strand, the number was replaced with the appropriate GenBank accession used to retrieve the sequence for analysis. If possible, the transcript designated “complete on 3’ end” was used. 62 2.12.1 Generation of Base-Frequency Plots.
Base-frequency plots for graphical presentation of relative frequency of bases in the HuR binding motif were generated by manually aligning the sequences of interest, and inputting that alignment into the web-based programme WebLogo (http://weblogo.berkeley.edu/logo.cgi).
2.12.2 Calculation of Expected NNUUNNUUU Motif Frequencies.
The frequency with which the motif NNUUNNUUU was expected to occur was calculated by multiplying the frequency each base was expected to occur in the absence of selection. That is; 1 x 1 x 0.25 x 0.25 x 1 x 1 x 0.25 x 0.25 x 0.25 = 0.00009765, or 1/1024.
2.13 Analysis of Protein.
2.13.1 Secreted PSA assays.
Mario Taranto (Core and Clinical Biochemistry, Royal Perth Hospital) measured prostate-specific antigen (PSA) secreted from LNCaP and 22Rv1 cells using a solid-phase 2-site chemiluminescent immunometric assay (Diagnostic Products Corporation [DPC]). The chemiluminescence was measured using a DPC Immulite 2000 random access autoanalyser.
2.13.2 Protein Extraction.
Whole cell lysates were prepared from LNCaP or 22Rv1 cells by addition of mid-RIPA buffer 29 , supplemented with protease inhibitors 41 , or 1 x PLB directly to the drained culture plate according to Table 2.3. The plates were then placed at -80 °C for 15 min or until needed, thawed on ice and harvested by scraping. The contents were collected into microcentrifuge tubes and centrifuged at 4 °C for 1 min to pellet the cell debris. Supernatants were removed to a new tube, and frozen at -20 °C in aliquots if the final volume was more than 150 µL.
63
Table 2.3. Lysis volumes for protein extraction. Culture Vessel: Area (cm 2) Vol. of lysis buffer (µµµL) 10 cm dish 78 500 6-well plate 9 150 24-well plate 2 45
2.13.3 Protein Concentration Assay.
Protein concentrations of cell lysates were determined using the Bio-Rad protein assay (a modified Bradford assay). A protein standard curve was generated by diluting bovine serum albumin (BSA) (100 mg/mL) to 0.5, 1.0, 1.5, 2.0 and 4.0 mg/mL, and adding 1 µL of each solution, in duplicate, to a 96- well microtitre plate. In the same plate, 1 µL of each unknown sample was added in duplicate. Bio-Rad Protein Assay Reagent was diluted 5-fold with ddH 20 and 250 µL of this was added to each sample. The plate was then incubated for 5-10 min at room temperature before the absorbance at 590 nM was measured in a Fluostar Optima platereader. Concentrations of unknown samples were automatically calculated from the standard curve using the Fluostar software.
2.13.4 Polyacrylamide Gel Electrophoresis of Proteins.
Protein expression was studied using denaturing sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE). Polyacrylamide gels 39 (10 % w/v) were prepared, or NuPage 10 % Bis-Tris polyacrylamide were purchased. Equal microgram amounts (10-25 µg/lane) of extracts to be analysed were made up to an equal final volume with the buffer in which they had originally been prepared. These samples were mixed in a ratio of four parts sample to one part Invitrogen LDS Sample Buffer and heated to 95 ° C for 4 min. After denaturing, the samples were centrifuged briefly and cooled on ice. Samples were loaded into the gel, along with 7 µL of Kaleidoscope molecular weight marker and electrophoresed at 35 mA per gel until the dye front reached the bottom of the gel. Pre-cast gels were electrophoresed in 1 x MOPS-SDS buffer and prepared gels were electrophoresed in 1 x electrode buffer 16 .
64 2.13.5 Western Blotting.
After electrophoresis as described in Section 2.13.4, proteins were transferred onto PVDF-plus western blotting membranes. Membranes were dipped in 53 methanol, rinsed in ddH 20 and pre-equilibrated in western transfer buffer (WTB). The gel, two pieces of filter paper and two sponges were also pre- equilibrated in WTB. The blotting apparatus was assembled as follows: one sponge then one piece of filter paper were stacked on the black side of the blotting cassette, the gel was placed on the filter paper, followed by the PVDF- plus membrane, a second piece of filter paper and the second sponge. The cassette was closed and proteins were transferred to the membrane by electroblotting overnight in WTB (with stirring) at 30 V and 4 °C.
2.13.6 Chemiluminescent Detection of Proteins.
All incubations were performed at room temperature for at least 1 h, with rocking. PVDF-plus membranes with electroblotted proteins (Section 2.13.5) were blocked in Tris Buffered Saline (TBS) 49 containing 3 % skim milk powder. Following blocking, membranes were incubated with primary antibody, diluted in Tris Buffered Saline-Tween (TBS-T) 50 containing 1 % skim milk powder. Excess primary antibody was removed by three 10 min washes in TBS-T before incubation with secondary antibody diluted in TBS-T50 containing 1 % skim milk powder. Excess secondary antibody was removed by three 10 min washes in TBS-T50 . Enhanced chemiluminescence detection reagent was prepared by adding 50 µL of Solution B to 2 mL of Solution A in accordance with the manufacturer's instructions, and pipetted onto the membrane. After incubation for five minutes at room temperature, excess detection reagent was blotted off and the membrane was exposed to light-sensitive film or the image was recorded by a charge-coupled device camera, using a ChemiDoc XRS. Membranes were rinsed in TBS 49 , dried, and stored at 4 °C.
2.13.7 Analysis of Western Blots.
Western blots were analysed by densitometry, using Quantity One 1-D analysis software (Version 4.5.2, build 070). Band densities were measured by the “volume rectangle” method using local background subtraction, and the
65 average band density calculated by the intensity to volume ratio. The density of actin was used to normalise variations in loading and non-specific siRNA effects. Where possible, images generated by the charge-coupled device camera were used for quantitation in preference to film.
2.14 Preparation of RNA.
2.14.1 RNA Extraction.
Trizol reagent was used to extract RNA from LNCaP or 22Rv1 cells, substantially as described in the manufacturer’s protocol. Medium was aspirated from confluent cells in 9 cm 2 (6-well) tissue culture plates, 1 mL of Trizol was added per well, and cells were lysed for 5 min at room temperature. The plates were then scraped and the contents collected in sterile microcentrifuge tubes. Samples were processed to completion immediately or stored at -80 °C. To precipitate protein contaminants, 200 µL of chloroform was added to each sample, which was then shaken vigorously by hand for 15 sec before being centrifuged at 12 000 x g for 15 min at 4 °C. The aqueous phase containing the RNA was then removed to a new tube. One µL of carrier, GlycoBlue, and 500 µL of isopropanol were then added to each tube, and the RNA was precipitated for 10 min at room temperature. RNA was pelleted by centrifugation at 12 000 x g for 10 min at 4 °C, the supernatants were removed and discarded, and RNA pellets were washed with 1 mL of 75 9 % ethanol in diethyl pyrocarbonate (DEPC)-treated ddH 20 . The tubes were flicked to mix and re-centrifuged at 7 500 x g for 5 min at 4 °C. The ethanol was removed, the pellets were air-dried, resuspended in 15 µL of DEPC- 9 treated ddH 20 then stored at –80 °C.
2.14.2 RNA Extraction for Micro-Array Analysis.
The LotteryWest State Micro-array Facility (LSMAF) protocol, detailed below, was followed to produce RNA of very high quality (having a ratio of 28S to 18S ribosomal RNA [rRNA] of 1.6 or higher).
66 22Rv1 cells were treated with siRNA as described in Section 2.8. To obtain sufficient quantity of RNA, cells from three wells of a six-well plate were pooled, each having been lysed in 333 µL of Trizol reagent. Repeated pipetting ensured full lysis of samples. Samples were transferred into sterile microcentrifuge tubes, 200 µL of chloroform added and the tubes were shaken vigorously by hand for 15 s. Following 2 min incubation at room temperature, the samples were centrifuged for 15 min at 12 000 x g and 4 °C. The aqueous phases containing RNA were removed to fresh tubes, an equal volume of isopropanol was added and the tubes were vortexed. Samples were loaded into RNeasy Mini Spin columns and centrifuged for 15 s at 8 000 x g. This step was repeated until the entire sample had passed through the column. The columns were then washed and centrifuged as follows: 700 µL of RW1 buffer (15 s at 8 000 x g), 500 µL of RPE buffer (15 s at 8 000 x g), 500 µL of RPE buffer (2 min at 8 000 x g). Flow-throughs were discarded between washes and again after a final 1 min centrifugation at 12 000 x g. RNA was eluted by addition of 30 µL of RNase-free water, 1 min incubation at room temperature, and 1 min centrifugation at 8 000 x g. Two 1 µL aliquots were removed for quality assessment (Section 2.15.1), and the remaining RNA was snap-frozen on dry ice to minimise degradation before storage at –80 °C.
2.15 Analysis of mRNA expression.
2.15.1 Micro-Array Analysis of HuR Knockdown in 22Rv1 Cells.
HuR was knocked down in 22Rv1 cells using siHuR-UTR siRNA (Section 2.8.2). Knockdown of HuR was confirmed by western blotting (Section 2.13.5), and RNA was extracted as described in Section 2.14.2. A sample of the RNA was diluted 10-fold, and its quality and concentration assessed using the Agilent Bioanalyser at the LSMAF. RNA quality was determined using the ratio of the area under the 18S and 28S rRNA peaks, and the RNA integrity number (RIN) calculated by the Bioanalyser software. RIN is partially based on the rRNA ratio (Schroeder et al. , 2006). RNA samples having sufficient quality (ratio >1.6, RIN > 9) and quantity (>1 µg/ µL) were used for micro-array analysis.
67
RNA (10 µg) was hybridised to the Affymetrix GeneChip Human Genome U133 Plus 2.0 oligonucleotide micro-arrays at the LSMAF according to standard Affymetrix protocols (Section 5.1). Micro-arrays were performed in duplicate on siHuR-UTR-treated and siCon-treated samples. Data files in .CHP format, generated using GeneChip Operating System (GCOS, Affymetrix) by the LSMAF, were analysed using the web-based programme GeneSifter (www.genesifter.net). During creation of .CHP files, GCOS identified as “absent” those probes that failed to achieve a default threshold. Genes corresponding to probes that had been tagged as absent were excluded from further analysis in GeneSifter.
Data from each micro-array was normalised to the mean fluorescence intensity of that chip and transformed using binary logarithm by the GeneSifter software (Appendix 4). The two HuR siRNA-treated and two control siRNA-treated datasets were averaged, and a ratio of HuR:control performed using the pairwise comparison function in GeneSifter, including a Student’s T-test. The following settings in Preferences were used: quality cut-off: “all groups must pass”; Affymetrix Quality Selection “A,M,P”, therefore only genes that were detected as “present (P)” by GCOS in all four micro-arrays were included. This eliminated genes that were likely to have extremely low expression levels. Those genes that exhibited expression levels that were significantly different between the two treatments (p < 0.05), were expressed in all four chips, and were greater than 1.2-fold up- or down regulated were clustered into Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathways (Kanehisa, 1997; Kanehisa and Goto, 2000; Kanehisa et al. , 2006). GeneSifter automatically assigns a z score describing the significance of clustering within a given pathway. Genes, selected on the basis of being more than 1.2-fold up- or down-regulated, in a pathway that exhibited significant clustering (z score of >2), and having potential HuR binding sites in the 3’UTR (Section 2.12), were further analysed by qPCR to confirm mRNA changes (Section 2.15.4). NetAffx (http://www.affymetrix.com/analysis/index.affx) was used to retrieve Gene Ontology (GO) definitions for all regulated genes for further examination of processes potentially regulated by HuR.
68 2.15.2 cDNA Synthesis.
First-strand complementary DNA (cDNA) was synthesised using SuperScript II RT in accordance with the manufacturer's instructions. RNA (2 µg), 1 µg of random hexamers (2 µL of 0.5 µg/ µL) and 0.5 mM dNTPs (1 µL of 10 mM 4) were mixed in a final volume of 12 µL and heated to 65 °C for 5 min in a thermocycler before being cooled to 4 °C. Seven µL of a master mix consisting of 10 mM DTT (2 µL of 100 mM stock), 40 U of RNaseOUT (1 µL of 40 U/ µL) in 1 x First Strand Buffer (4 µL of 5 x stock) was added per reaction. After 2 min at room temperature, 200 U of SuperScript II RT (1 µL of 200 U/ µL) were added and the reactions were incubated in a thermal cycler at 25 °C for 10 min, 42 °C for 50 min then 70 °C for 15 min. cDNA was stored at –20 °C.
2.15.3 Primer Design for Quantitative PCR.
The AR forward and reverse primer sequences were sourced from (Latil et al. , 2001). Actin forward and reverse primer sequences were sourced from (Seth et al. , 1999). Ribosomal primers were designed using the Invitrogen OligoPerfect online designer (www.invitrogen.com). The 16S ribosomal RNA sequence from AP008515/gi:61287634 (GenBank) was input into the design software, and the primer set with the best match to the default settings was selected. Primers specific for CREB-binding protein (CREBBP), p300/CBP- associated factor (PCAF), p21 activated kinase 2 (PAK2) and ETS-like protein 1 (ELK1) were designed using the open-source software PerlPrimer v1.1.13 (http://perlprimer.sourceforge.net/). Each primer pair crossed at least one intron/exon boundary, with at least 7 basepair overlap, giving an amplicon of 150-250 basepairs in the 3’ half of the transcript. Primer sequences are listed in Table 2.4.
2.15.4 Quantitative Real-Time PCR.
Quantitative real-time PCR (qPCR) was used to determine levels of AR, CREBBP, p300/CAF, PAK2 and ELK1 mRNA relative to actin mRNA or 16S ribosomal RNA. qPCR reactions were prepared using the Platinum SYBR Green qPCR SuperMix-UDG, in accordance with the manufacturer's instructions. Briefly, a cocktail was prepared, using (per reaction): 10 µL of 2 x SuperMix, 200 nM forward and reverse primers (0.4 µL each of 10 µM14 ), 0.5 69 µL of ROX reference dye and ddH 2O to 15 µL. The cocktail was aliquoted into clean flat-capped 0.2 µL PCR tubes and 5 µL of diluted cDNA was added. cDNA was serially diluted in BSA (0.4 µg/ µL7), 10-fold for AR and actin, and 100-fold for the more abundant 16S rRNA. All samples were incubated at 50 °C for 2 min then at 94 °C for 2 min 30 s prior to amplification. Specific cycling conditions are listed in Table 2.4.
A Rotor-Gene was used to perform qPCR cycling and analysis. For each reaction, a standard curve was generated using cDNA serially-diluted in 10- fold steps from 5- to 50 000-fold. Concentrations of standards were selected to cover the range of cycle thresholds (CT) from ~3 to ~30 cycles, to establish a broad linear range. A minimum of three of five points was used to generate a standard curve with a regression coefficient of at least 0.98. Rotor-Gene software (Version 6.0, Build 19) automatically calculated cycle threshold (CT) using the standard curve to determine relative concentration of products. Reaction efficiencies were approximately equal to 1 with all primer sets. Product identity was confirmed in the first instance by sequencing, and by melting temperature as determined by the Rotor-Gene software thereafter.
2.15.5 Actinomycin D Chase Assay.
An actinomycin D (ActD) chase assay was used to determine the half-life of AR mRNA in cells treated with siRNA to HuR. 22Rv1 cells growing in 9 cm 2 plates were transfected with the either siHuR-UTR siRNA or siCon siRNA for 72 h, as described in Section 2.8. After 72 h, ActD 1 was added to a final concentration of 7.5 µg/mL to halt transcription. Samples were harvested in quadruplicate at 0, 4, 8 and 12 h. At each time-point, media were aspirated from the cells and 1 mL of Trizol reagent was added. RNA was extracted as described in Section 2.14.1. At the first time-point, protein was also extracted using 1 x PLB to confirm knockdown by western blot (Section 2.13.5). RNA from ActD-treated cells was reverse-transcribed, and relative AR mRNA levels were determined using qPCR (Sections 2.15.2, 2.15.4). Actin mRNA levels were used to normalise AR mRNA levels. Half-life of mRNA was estimated by solving the decay equation in Figure 4.8A for x when y=0.5.
70 Table 2.4. Primers and cycling conditions used for qPCR
Primer Name: Sequence: Cycling (x40):
AR forward CCT GGC TTC CGC AAC TTA CAC 94 °C 15 s
65 °C 30 s AR reverse GGA CTT GTG CAT GCG GTA CTC A 72 °C 60 s Actin forward GCC AAC ACA GTG CTG TCT GG 94 °C 15 s
60 °C 45 s Actin reverse TAC TCC TGC TTG CTG ATC CA 72 °C 60 s
16S forward CAC TGT CAA CCC AAC ACA GG 94 °C 15 s
55 °C 15 s 16S reverse GGC AGG TCA ATT TCA CTG GT 72 °C 45 s
CREBBP CAG CCA GCA TTG ATA ACA GAG 94 °C 15 s forward 60 °C 45 s CREBBP TCC TCC ATC ATC TCA GAC CT reverse 72 °C 60 s
PAK2 forward TAA GCT CAC TGA CTT TGG TT 94 °C 15 s
60 °C 45 s PAK2 reverse CTC TCC TTC TAC CAT CTC AAT AG 72 °C 60 s
PCAF forward GAT GGA GTT CGA CAG ATT CCT 94 °C 15 s
60 °C 45 s PCAF reverse CTT TGA TGG CTC TTC ACC TG 72 °C 60 s ELK1 forward TCC TAC GCA TAC ATT GAC CC 94 °C 20 s
62 °C 30 s ELK1 reverse ACT GGA TGG AAA CTG GAA GG 72 °C 30 s
BAG1 forward CAA GGA TTT GCA AGC TGA AG 94 °C 20 s
62 °C 30 s BAG1 reverse CTA GGA ATG CCT GAA CCT TT 72 °C 30 s
71 2.15.6 Nested PCR.
HuD cDNA was amplified from LNCaP, HepG2 or PC12 total RNA by Dr Ricky Lareu (LCM, WAIMR). For LNCaP and HepG2 samples, RNA was extracted (Section 2.14.1) and cDNA was generated using either a HuD-specific primer (5’-AAA GAT GCA CCA CCC AGT TC-3’) or oligo-dT as described in Section 2.15.2, followed by PCR using inner primers specific for the HuD1 (Pro1) isoform (5’-TGG TCC AGA CAT CAG TCT C-3’ and 5’-AGG CAG AAG AAG CCA TCA AA -3’) and the following cycling conditions: one cycle of 90 °C for 30 s; 34 cycles of 94 °C for 30 s, 53 °C for 30 s, 72 °C for 45 s. PCR products were electrophoresed in 1.5 % agarose gels 2 in 1 x Tris Acetate EDTA (TAE 47 ) buffer and visualised under UV transillumination.
2.16 Localisation of HuR.
2.16.1 Sample Preparation.
22Rv1 cells were plated at a density of 2.5 x 10 4 cells/cm 2 on sterile cover slips in 9 cm 2 dishes. Transfection of siRNA took place at the time of plating, as described in Section 2.8. After 72 h of siRNA treatment, a subset of samples were also treated with 7.5 µg/ µL ActD 1 for 4 h. In addition to these samples, 22Rv1 cells were also left untransfected, and treated with 10 nM DHT for 30 and 90 min or 4, 24 and 48 h. Cells were fixed onto the cover slips with freshly prepared 4 % v/v formaldehyde/PBS 19 (30 min). The fixed cells were rinsed twice with PBS and permeabilised with 0.01 % v/v Triton X-100/PBS 51 for 20 min. Cells were rinsed twice more with PBS then left to block overnight at 4 °C in 1 % w/v BSA/PBS 7. The following day, the cells were incubated with HuR 3A2 primary antibody diluted 1000-fold in PBS for 1 h. Primary antibody was washed off with three rinses of PBS, and the samples incubated with an Alexa- Fluor 488-conjugated secondary antibody (goat anti-mouse IgG) diluted 1 000- fold in PBS 37 for 1 h. Cover slips were protected from light from this step onwards. A secondary-antibody-only control was used on the first occasion to assess non-specific staining, which was very low. Secondary antibody was washed off with PBS, and the nuclei stained for 30 min with Hoescht 33258/ PBS 23 . The cover slips were washed a further 3 times in PBS 37 , then air-dried
72 briefly and mounted on slides with low-fade mounting medium. Cover slips were sealed onto the slides using clear nail varnish, and stored in the dark at 4 °C. LNCaP cells were plated as above, with the exception that fibronectin- coated cover slips 18 were used.
2.16.2 Microscopy.
Cells were visualised using a Bio-Rad MRC 1000 laser scanning confocal microscope using a 40 x oil-immersion objective, at the Biological Imaging and Analysis Facility, University of Western Australia. Images that were to be compared were captured using gain settings within 10 units of each other, to keep exposure times similar, and 3-4 Kalman scans were used to give high- definition images.
2.17 Analysis of RNA-Protein Interactions.
2.17.1 Purification of GST fusion proteins.
Gluathione-S-transferase (GST)-tagged fusion proteins were prepared by Britt Granath (LCM, WAIMR) essentially as described previously (Yeap et al. , 2002). Briefly, 500 mL cultures of E. coli BL21 Codon Plus cells transformed with pGEX-6P-HuD plasmid (Appendix H) were induced at an OD 600 of 0.6 with 0.2 mM IPTG (1 mL of 100 mM 24 ) at 30 °C for 2 h. Bacterial cultures were sonicated in NaCl EDTA Tris (NET 32 ) buffer and GST-fusion proteins were affinity purified from the lysed cultures using glutathione-agarose beads. Proteins were eluted in GST Elution Buffer 20 after washing with NET buffer. Protein purity was assessed by SDS-PAGE following quantitation using a Bio- Rad protein assay (Sections 2.13.3, 2.13.4)
2.17.2 Riboprobe Synthesis.
Labelled riboprobes were produced by transcribing the pBluescript-ARUC-WT (Yeap et al. , 2002) (Appendix 2A) or pBluescript-c-fos-HuD vector (Giles et al. , 2003) (Appendix 2D) in the presence of excess γ-32 P-UTP. Five µL of γ-32 P- UTP (3000 Ci/mmol) were concentrated by centrifugation under vacuum for 7
73 min. The following reagents were added to the dried pellet: 1 µg linearised template (Section 2.17.2.1), 1 x T7 RNA polymerase buffer (2 µL of 5 x), 20 mM DTT (2 µL of 100 mM), and 0.4 mM NTP’s minus UTP mix (2 µL of 2.5 mM stock 35 ). This was warmed to resuspend the isotope pellet and centrifuged briefly before being placed on ice. RNaseOUT (1 µL of 40 U/ µL) and 20 U of T7 RNA polymerase (1 µL of 20 U/ µL) were added, and the reaction incubated for 60 min at 37 °C. DNase I (1 µL of 1 U/ µL) was added and the reaction incubated for a further 10 min at 37 °C to remove the DNA template. An aliquot (1 µL) was retained for scintillation counting, and the 9 remaining 10 µL was diluted 5-fold in DEPC-treated ddH 20 before being applied to a MicroSpin G-25 clean-up column to remove unincorporated nucleotides. The column was centrifuged for 4 min at 2 000 x g to elute the probe. One µL of eluted probe was retained for scintillation counting.
To confirm probe integrity, the labelled probe was visualised by autoradiography. Equal amounts of probe (approx 150 000 counts per minute [cpm] each) from before and after column purification were mixed with RNA electrophoretic mobility shift assay (REMSA) loading buffer 42 , and loaded into a 6 % urea polyacrylamide gel 40 . The samples were electrophoresed at 150 V for 15-20 min. The apparatus was disassembled and the gel, left on one glass plate, wrapped in clingfilm before being exposed to light-sensitive film for 30 min at room temperature. Probes were only used if a major single product was detected.
2.17.2.1 Linearisation and Purification of Riboprobe Templates.
One µg of pBluescript-ARUCT-WT or pBluescript-c-fos -HuD was linearised by restriction endonuclease digestion with 1 U Hind III for 1 h at 37 °C in 1 x Buffer E in a final volume of 20 µL. To this was added 5 µL of DNA loading dye 13 , and the samples were loaded into a 1 % agarose gel 2, then electrophoresed in 1 x TAE buffer 47 at 100 V until the dye front was ~ 3 cm from the bottom of the gel. DNA fragments were visualised by UV transillumination. Sizes of DNA fragments were estimated by comparing their migration to the 1kb Plus ladder. The linear fragment was excised from the gel with a sterile scalpel blade, placed in a microcentrifuge tube and weighed. Gel purification was performed
74 using a Wizard gel purification kit as directed by the manufacturer. For every 10 µg of agarose, 10 µL of Membrane Binding Solution was added and the agarose was melted at 70 °C for 15 min. The dissolved gel mixture was applied to a minicolumn and incubated at room temperature for 1 min then centrifuged for 1 min at 12 000 x g. The column was then washed with 700 µL of Membrane Wash solution and recentrifuged for 1 min at 12 000 x g. A second wash of 500 µL was followed by centrifugation at 12 000 x g for 5 min 9 then the was DNA eluted by addition of 50 µL of DEPC-treated ddH 2O and centrifugation at 12 000 x g for 1 min.
2.17.3 RNA Electrophoretic Mobility Shift Assay.
Recombinant HuD protein was produced by Britt Granath, using the GST-HuD expression vector pGEX-6P-HuD. ARUC-WT and c-fos probes were produced 9 as described in Section 2.17.2 and diluted in DEPC-treated ddH 20 to 100 000 cpm/ µL. REMSA reactions were performed to test HuD binding to the ARUC- WT probe, and were prepared as follows: 200 ng HuD protein was mixed with 2 µg of yeast tRNA (1 µL of 2 mg/mL 54 ) and 50 µg heparin (1 µL of 50 mg/mL 21 ), in a final volume of 19 µL cytoplasmic extraction buffer (CEB) 8. This was incubated for 10 min at room temperature, while the probes were denatured at 80 °C for 5 min and cooled on ice. Denatured probe (1 µL) was added to the reaction, and incubated at room temperature for 30 min. Antibody (5 µL) against either HuD (16C12) or AR (441) were added as appropriate, and the reactions incubated for a further 30 min on ice. REMSA loading dye 42 (1 µL) was added and the reactions were loaded into 1 % agarose/0.5x TBE gels 2 and electrophoresed in 0.5 x Tris-borate EDTA (TBE) 48 for 1 h at 100 V. After electrophoresis, the gel was dried at 80 °C for 1 h in a gel dryer, wrapped in clingfilm and exposed overnight to a blanked phosphor-imaging plate. The plate was then scanned using a PhosphorImager 445 Si.
75 2.18 Immunohistochemistry. Immunohistochemistry was performed in the Department of Anatomical Pathology at Royal Perth Hospital and analysed by Dr Cecily Metcalf (Department of Anatomical Pathology, Royal Perth Hospital). Full-face sections of primary prostate tissue were immunostained for HuD. Sections were deparaffinised through 3 changes of xylene (3 min), rehydrated through graded alcohols to ddH 2O and subjected to antigen retrieval in 500 mM EDTA (pH 8.0) under pressure. After blocking endogenous peroxidase activity with 3
% (v/v) H 2O2, the HuD 16C12 antibody was applied at 1:10 000 dilution for 60 min. Immunoreactivity was detected by incubating slides with biotinylated sheep anti-mouse antibody (Chemicon IHC Select) followed by streptavidin- horseradish peroxidase and visualised with 3’,3’-diaminobenzidine (DAKO). Slides were counterstained with weak haematoxylin and mounted using DPX mounting medium.
2.19 Immunoprecipitation RT-PCR Assay (IP-RT-PCR). Dianne Beveridge and Micheal Epis (LCM, WAIMR) performed IP-RT-PCR as follows, using LNCaP or 22Rv1 cells and immunoprecipitating with HuD, HuR or IgG antibodies. Media were removed from cells at 70 % confluence and the cells were washed twice with ice-cold PBS 37 . Mid-RIPA buffer 29 (1 mL) containing 40 U of RNaseOUT was added and the cells were incubated for 30 min with gentle agitation every 10 min. The resultant lysate was collected in 2 mL microcentrifuge tubes and centrifuged at 10 000 x g for 10 min. The supernatants were transferred to new tubes and 10 µg of HuD 16C12 or HuR 3A2 antibody was added before incubation at 4 °C for 1 h. Pre-swollen protein A beads were washed with mid-RIPA buffer 29 three times then added to the supernatants and incubated for a further 60 min at 4 °C. Beads were collected by centrifugation at 10 000 x g for 2 min, and supernatants were removed to a fresh tube. The beads were then washed four times with mid- 29 9 RIPA buffer and resuspended in 200 µL of DEPC-treated ddH 20 .
RNA was extracted from the beads and from the supernatants using phenol:chloroform. An equal volume of 1:1 phenol:chloroform was added to
76 the beads or supernatant and the tubes were centrifuged for 10 min at 10 000 x g. The aqueous phase was removed and the RNA precipitated by the addition of 1 µg of glycogen (1 µg/ µL) along with 0.1 volume of 3 M sodium acetate 44 and 0.7 volume of isopropanol. After 1 h at –20 °C, RNA was collected by centrifugation at 12 000 x g for 15 min, followed by a wash with 500 µL of 70 % ethanol and recentrifugation at 12 000 x for 5 min. RNA was 9 air-dried and resuspended in 10 µL of DEPC-treated ddH 20 . cDNA was produced as described in Section 2.15.2. mRNA was detected by 4 PCR, using 1 µL of cDNA, 0.2 mM dNTPs (2 µL of 2 mM ), 1.5 mM MgCl 2 (0.6 µL of 50 mM), 50 ng of forward primer, 50 ng of reverse primer and 1 U of Taq polymerase (0.2 µL of 5 U/ µL) in 1 x PCR buffer (2 µL of 10 x) in a final volume of 20 µL. The primers used to detect the AR mRNA were: forward primer 5’- CTC TTC AGC ATT ATT CCA GTC-3’ and reverse primer 5’-ATC TTA GCT CTC TAA ACT TCC-3’. CREBBP, PCAF, ELK1 and PAK2 mRNA’s were detected using the primers listed in Table 2.4. Cycling conditions were as follows: one cycle of 95 °C for 5 min; 55 °C for 1 min; 72 °C for 1 min, then 40 cycles of 95 °C for 1 min; 55 °C for 1 min; 72 °C for 1 min, finishing with a long extension of 72 °C for 10 min. PCR reactions were electrophoresed in 1.5 % agarose gels 2 in 1 x TAE 47 buffer and visualised under UV transillumination on a ChemiDoc.
2.20 Statistical Analysis. Significance was determined using Student’s one-tailed T-test for paired means, and differences were considered significant if the p-value was < 0.05. For data sets containing multiple points over time, ANOVA was used, again using p < 0.05 as an indicator of significance. Microsoft Excel was used to perform these functions. Statistical analysis of micro-array data was performed using proprietary software integrated into GeneSifter (Section 2.15.1).
77
CHAPTER 3: Analysis of HuD Expression and Activity in Prostate Cancer.
78
3 Chapter 3: Analysis of HuD Expression
and Activity in Prostate Cancer
3.1 Introduction. HuD is an RNA-binding protein of the ELAV/Hu family, the expression of which is typically restricted to tissue of the central nervous system (Section 1.4.3). HuD binds AU-rich sequences in mRNA, regulating the stability of a number of transcripts involved in neuronal differentiation. Forced over-expression of HuD induces spontaneous neurite formation in PC12 phaeochromocytoma cells, suggesting an important contribution by HuD to the process of neuronal differentiation (Hollams et al. , 2002; Park-Lee et al. , 2003). Although largely restricted to neuronal tissue, HuD protein expression has been detected outside the central nervous system in tumours of neuroendocrine origin, such as neuroblastomas, and in tumours which express neuroendocrine markers, such as SCLC (Behrends et al. , 2002; Sekido et al. , 1994). Up-regulation of HuD mRNA has also been observed in metastatic neuroendocrine cell cancers arising from the prostates of CR2-TAg transgenic mice, which also express several other markers of neuronal or endocrine tissue (Hu et al. , 2002).
In the human prostate, neuroendocrine cells are found at a higher density in carcinomas, and while small cell neuroendocrine carcinomas of the prostate are rare, focal neuroendocrine differentiation is common in prostate tumours (Shariff and Ather, 2006). Neuroendocrine differentiation has been associated with increased tumour size and development of androgen-independence, however the clinical utility of neuroendocrine differentiation has not been established (Shariff and Ather, 2006). This is in part due to the use of diverse sampling methods and neuroendocrine markers, that have limited comparisons between studies. For example, measurement of serum levels of chromogranin A (CgA) in prostate cancer patients was shown to be of prognostic value (Berruti et al. , 2005; Taplin et al. , 2005), whereas immunohistochemical
79 analysis of CgA or neuron-specific enolase (NSE) expression in tumour specimens has been reported to be both of prognostic value (Theodorescu et al. , 1997; Weinstein et al. , 1996), or of no independent prognostic value (Ahlegren et al. , 2000; Cohen et al. , 1991; McWilliam et al. , 1997). Due to the lack of consistency in results between different studies, it has been suggested that new markers of neuroendocrine differentiation that correlate with aggressive tumour behaviour are needed for more accurate diagnosis of prostatic neuroendocrine disease. The CR2-TAg mouse model was generated to assist in the identification of such markers, and increased HuD mRNA expression in the metastatic neuroendocrine prostate tumours of the mice indicated that HuD may be a marker for neuroendocrine differentiation (Hu et al. , 2002).
Among the transcripts thought to be bound and stabilised by HuD are those encoding neuroserpin, acetyl-cholinesterase, GAP-43, and p21 Waf1 , and HuD has also been reported bind to and inhibit translation of p27 Kip1 mRNA (Cuadrado et al. , 2002; Deschenes-Furry et al. , 2005a; Joseph et al. , 1998; Kullmann et al. , 2002; Mobarak et al. , 2000). Many of the reported targets of HuD are not expressed in prostatic cells, or are expressed only at low levels, with the exception of p21 Waf1 and p27 Kip1 , which are both important cell cycle regulators. This suggests that expression of HuD in the prostate could result in aberrantly increased levels of mRNA’s and therefore proteins that are not normally expressed in the prostate. In addition, the RNA-binding domains of HuD are very similar to those of HuR (> 90 % identity), suggesting that HuD may bind mRNA’s that are normally targets of HuR in the prostate. HuR has been previously shown to bind the AR mRNA, and given the homology of HuR and HuD, it was of interest to investigate whether HuD also binds the AR mRNA (Yeap et al. , 2002).
The aim of the following studies was to establish the presence of HuD expression in primary prostate tumours and the prostate cancer cell line LNCaP, and to determine whether HuD is able to regulate the AR in prostate cancer cells.
80 3.2 Results
3.2.1 HuD is Expressed in Human Prostate Tumours.
A cohort of 12 primary prostate tumours was used for this preliminary immunohistochemical study. For each specimen, individual paraffin- embedded blocks were chosen which contained normal tissue, benign prostatic hypertrophy (BPH) and adenocarcinoma of various Gleason grades (2-5), thus allowing for simultaneous investigation of non-malignant and malignant tissues. Four µm sections from each of the blocks were stained with HuD antibody (Section 2.18). In these specimens, HuD immunostaining appeared as punctate cytoplasmic immunoreactivity, which in non-malignant and benign hyperplastic glands was uniformly weak (Figure 3.1A). Within malignant acini of Gleason grades 2-3, > 95 % of cells exhibited punctate cytoplasmic staining for HuD, which was of increased intensity in comparison to non-malignant glands. In the poorly differentiated Gleason grade 4-5 tumours, the intensity of immunostaining was higher and HuD immunoreactivity remained predominantly cytoplasmic (Figure 3.1A). Scoring of HuD immunostaining indicated that HuD was widely expressed in epithelial cells of the adult prostate with weak levels of expression in non-malignant tissues and elevated expression in areas of prostatic adenocarcinoma. Increased intensity of immunostaining was evident in regions of higher Gleason grade (Figure 3.1 B). HuR was ubiquitously expressed in all malignant and non-malignant cells of the prostate and staining was uniform in all tissues (data not shown).
As HuD protein was strongly expressed in malignant prostate cells, its expression was examined in the human prostate cancer cell line LNCaP. Using RT-PCR, HuD cDNA could be detected in LNCaP cells as well as in the PC12 rat phaeochromocytoma cell line which has been previously reported to express high levels of HuD (Section 2.12.6; Figure 3.1C) (Steller et al. , 1996). As expected, the transcript was not detected in HepG2 cells (Figure 3.1C). Despite the expression of HuD mRNA in LNCaP cells, HuD protein was undetectable by western blotting of LNCaP cell extracts (Section 2.13.5; results not shown).
81 1 A N
C
2 S
C
N
3
C Normal Adenocarcinoma Case BPH B prostate 2 3 4 5 1 +/- +/- ++ +++ 2 +/- +/- ++ +++ 3 +/- + +++ 4 +/- +/- ++ ++ +++ 5 +/- +/- ++ ++ 6 +/- +/- ++ +++ 7 +/- +/- ++ 8 +/- +/- ++ +++ 9 +/- +/- ++ 10 +/- +/- ++ +++ 11 +/- + +++ 12 +/- +/- +++
C
HuD (217 bp) LNCaP + + - - - + HepG2 - - + + - - PC12 - - - - + - Oligo-dT - + - + + + HuD primer + - + - - - RT + + + + + -
Figure 3.1. HuD is expressed in primary prostate tumours.
A. Immunohistochemical analysis of human prostate cancers stained with HuD antibody. A Gleason grade 3 carcinoma at 20 x magnification (1) and 200 x magnification (2) showing intense HuD immunostaining in malignant cells (C) compared to non-malignant prostate glands (N) and stroma (S). (3) A Gleason grade 5 prostate carcinoma, showing intense HuD immunostaining (200 x magnification). B. HuD immunostaining in prostate specimens was scored according to the intensity of immunostaining and the histopathology of that region of tissue. C. RT-PCR showing expression of HuD in LNCaP and PC12 but not HepG2 cell lines. Product derived from the HuD primer is not shown for PC12 cells.
82
3.2.2 HuD Binds the AR mRNA.
The RRMs of HuD have very high sequence identity with those of HuR, which has been previously reported to bind the AR UC-rich region (ARUC) (Yeap et al. , 2002). Given this, the ability of HuD to bind the ARUC sequence was assessed using an RNA electrophoretic mobility shift assay (REMSA; Section 2.17.3). This experiment was first performed using a ribonucleotide probe representing a 27 nucleotide AURE from the c-fos 3’UTR, as it has been reported previously that HuD binds this region with high affinity (Chung et al. , 1996). LNCaP cytoplasmic extracts formed an RPC with this probe, as did extracts from PC12 cells (Figure 3.2A). The mobility of the RPCs could be further retarded (“supershifted”) by incubating HuD antibody with the probe and extract prior to loading into the gel (Figure 3.2A). The ability of the HuD antibody to supershift the RPC formed from the c-fos AURE and LNCaP cell extract indicated that HuD protein was expressed in LNCaP cells, albeit at levels undetectable by western blotting (Section 3.2.1). An irrelevant antibody (eNOS) did not supershift the RPC from LNCaP extract, and an RPC did not form between either the HuD antibody or the irrelevant antibody and the c-fos AURE probe, demonstrating the specificity of the interactions (Figure 3.2A).
Recombinant GST-HuD bound both the ARUC and the c-fos probes (Figure 3.2B; Section 2.17.1), an effect that was not observed using GST alone (data not shown). An HuD antibody, but not an irrelevant AR antibody supershifted the RPCs, providing evidence that the proteins involved in binding to the ARUC probe included HuD and that the reactions were specific (Figure 3.2B). Having established that purified HuD could bind the ARUC probe, the ability of HuD from LNCaP cytoplasmic extracts to bind the ARUC probe was examined. Extracts from untransfected LNCaP cells and LNCaP cells transiently transfected to express FLAG-HuD readily formed RPC’s with the ARUC probe, but only extracts from cells expressing the FLAG-HuD vector formed an RPC that could be supershifted by the HuD antibody (Figure 3.2C; Sections 2.17.3, 2.8). This RPC could not be supershifted using an irrelevant V5 antibody, indicating that the reaction was specific (Figure 3.2C). These data confirmed
83 12 3 4 5 6 7 A SS
RPC
Free Probe
c-fos probe + + + + + + + LNCaP - - + + - - + PC12 - - - - + + - HuD Ab - + - + - + - eNOS Ab ------+
B 1 2 3 4 5 6 7 8 C 1 2 3 4 5 6 7
SS SS
RPC RPC
Free Probe Free Probe c-fos probe + - + - + - + - ARUC probe - + - + - + - + ARUC probe + + + + + + + GST-HuD - - + + + + + + pFLAG-HuD extract - - + + + - - HuD Ab - - - - + + - - LNCaP extract - - - - - + + AR Ab ------+ + HuD Ab - + - + - - + V5 Ab - - - - + - -
Figure 3.2. HuD binds the ARUC region of the AR mRNA. The binding of HuD to the ARUC or c-fos AURE was investigated using RNA electrophoretic shift assays (REMSA). A. The c-fos probe was incubated with LNCaP or PC12 cell extract, with or without HuD or eNOS (irrelevant) antibodies as indicated. B. The c-fos AU-rich or ARUC probe was incubated with purified recombinant GST-HuD, with or without HuD or AR (irrelevant) antibody. C. The ARUC probe was incubated with extracts from untransfected LNCaP cells or LNCaP cells transiently transfected to express FLAG-HuD, with or without HuD or V5 (irrelevant) antibody as indicated. RPC: RNA-protein complex. SS: supershift.
84 that HuD was able to bind the ARUC region of the AR mRNA, however, the failure of HuD antibodies to supershift RPC’s formed by the ARUC probe and extracts from untransfected LNCaP may indicate a lower affinity interaction between the ARUC sequence and HuD than the c-fos sequence and HuD, thus requiring a higher concentration of HuD to produce the supershift.
As HuD and HuR could bind the ARUC region of the AR mRNA in vitro and HuD was expressed in prostate cancer, the in vivo interaction between HuR/HuD and the AR mRNA was examined using IP-RT-PCR (Section 2.19) (Yeap et al. , 2002). Using this assay, AR mRNA was detected by RT-PCR in all samples prior to immunoprecipitation, and also among the fraction of mRNA’s that co-immunoprecipitated from LNCaP cell lysates with HuR and HuD antibodies (Figure 3.3). These results indicated a close in vivo association of HuR and HuD with the AR transcript. The AR transcript was not detected in the absence of reverse transcriptase or in the absence of antibodies, providing evidence that the interaction was specific (Figure 3.3).
3.2.3 Mutations in the ARUC Region Destabilise mRNA.
As discussed in Section 3.1, HuD and HuR bind U-rich tracts in mRNA, and for HuR at least, single-stranded conformation of the binding sequence is required (Meisner et al. , 2004). According to this rationale, U G point mutations were introduced into the U-rich tracts of the wild-type ARUC probe sequence, with the intention of disrupting binding by HuD (Figure 3.4A). Two mutants were studied, Mutant A (MutA), which has point mutations in the distal U-rich tract, and Mutant B (MutB), which has point mutations in both tracts. Both mutant constructs contain an additional U G mutation which contributes to deformation of the single-stranded RNA sequence. These mutant constructs were generated to investigate binding of HuR to the ARUC region (Yeap et al. , 2002). Consistent with prior observations, recombinant HuD bound the wild- type (WT) ARUC (Figure 3.4B; Section 0). Reduced binding of HuD to the MutA probe was detected while binding of HuD to the MutB probe was not evident using these methods (Figure 3.4B).
85 supt beads supt beads 123 4 5 6 7 89 10 1112 13 AR (718 bp) No Ab + - - + - - + - - + - - - HuD Ab - + - - + - - + - - + - - HuR Ab - - + - - + - - + - - + - RT + + + + + + ------AR vector ------+
Figure 3.3. HuD and HuR are closely associated with the AR mRNA in vivo .
Immunoprecipitation-reverse transcription PCR (IP-RT-PCR) was used to demonstrate binding of HuD to AR mRNA. RT-PCR was performed on extracts from LNCaP before (supt) and after immunoprecipitation (beads) using HuD or HuR antibodies or no antibody (lanes 1 to 6). A replicate experiment was performed in the absence of reverse transcriptase (lanes 7 to 12).
86 4031 U-rich U-rich 4082 A Wild type CUGGGCUUUUUUUUUCUCUUUCUCUCCUUUCUUUUUCUUCUUCCCUCCCUA
Mutant A CUGGGCUUUUUUUUUCUCU GUCUCUCCU GUCU GGG UCUUCUUCCCUCCCUA
Mutant B CUGGGCUU GUGUGUUCUCU GUCUCUCCU GUCU GGG UCUUCUUCCCUCCCUA
1 2 3 B
RPC
Free probe
ARUC WT probe + - - ARUC MutA - + - ARUC MutB - - + Relative Luciferase Activity (%) GST-HuD + + + D 0 20 40 60 80 100 C MCS-empty Hu Hu WT LUC WT * MutA LUC MutA *
X * MutB LUC X X MutB
Figure 3.4. Mutations in the ARUC region disrupt HuD binding and luciferase activity.
A. Sequences of ARUC WT and mutant probes. Major U-rich tracts are indicated and mutations are in bold and underlined. B. Recombinant HuD was incubated with the ARUC WT (lane 1), ARUC MutA (lane 2) and ARUC Mut B probes (lane 3). RPC: RNA-protein complex. C. Schematic diagrams of the Luc-ARUC-WT (WT), Luc-ARUC-MutA (MutA) and Luc-ARUC-MutB (MutB) vectors. Hu binding sites (Hu) and location of mutations in these sites are indicated ( X) . D. LNCaP cells were transfected with Luc-ARUC-WT (WT), Luc- ARUC-MutA (MutA) or Luc-ARUC-MutB (mutB) vectors or a vector with no insert (MCS-empty), and luciferase activity was assayed after 24 h. The data represent mean ± standard deviation (sd) of quadruplicate samples. * =p < 0.05.
87
Having elucidated the effects of mutations in the ARUC region on HuD binding, experiments to analyse the effects of these mutations on mRNA stability were performed. Wild-type ARUC, MutA and MutB sequences were cloned downstream of the Luc gene in the 3’UTR of the modified pGL3-control vector, 3.3 pGL3-MCS, to generate the Luc-ARUC-WT, Luc-ARUC-MutA and Luc- ARUC-MutB constructs, respectively (Figure 3.4C; Section 2.5). Induced luciferase activity from these constructs was measured as a correlate with mRNA stability. Consistent with observations from REMSA and the known mRNA stabilising function of HuD/HuR, reduced luciferase activity was induced from the mutated Luc-ARUC-MutA and Luc-ARUC–MutB constructs in comparison to the Luc-ARUC-WT vector, which contained the wild-type ARUC sequence. As predicted, Luc-ARUC-MutB produced the most significant reduction in induced luciferase activity (Figure 3.4C). Surprisingly, the wild- type ARUC sequence did not significantly increase luciferase activity compared to a control vector having no insert in the 3’UTR, indicating that the ARUC sequence had not increased luciferase mRNA stability.
The secondary structures of the 3’UTR of the Luc-ARUC and pGL3-MCS- empty constructs were predicted using mfold (Mathews et al. , 1999; Zuker, 2003) (Section 2.11; Figure 3.5). This illustrated that in the wild type sequence, the U-rich tract is single-stranded and therefore accessible to Hu proteins, whilst in the mutated sequences, a loss of single-strandedness was observed (Meisner et al. , 2004). Luc-ARUC-Mut A has partial single- strandedness, which correlates well with the observations of weak RPC formation in REMSA’s (Figure 3.4B) and reduced induction of luciferase activity (Figure 3.4C). Luc-ARUC-MutB exhibited minimal single-strandedness within this sequence, correlating with the absence of RPC formation in REMSA’s (Figure 3.4B) and lack of induced luciferase activity (Figure 3.4C). Of interest was the finding that introduction of the ARUC sequences into the 3’UTR markedly altered the folding of the pGL3-MCS-empty 3’UTR, which may have contributed to the lack of apparent increase in induced luciferase activity from the Luc-ARUC-WT in comparison to the pGL3-MCS-empty vector (Figure 3.5).
88 MCS-empty
WT
MutA
MutB
Figure 3.5. Mutations in the ARUC region affect secondary structure. Mfold secondary structure plots of the 3’UTRs of the pGL3-MCS-empty, Luc-ARUC-WT, Luc-MutA, and Luc-MutB vectors (Zuker et al. 2003). The UC-rich region bound by HuR and HuD is circled.
89
3.2.4 Effects of HuD Expression and DHT on Luc-ARUC Function.
Based on the observation that mutations in the HuD binding sites decreased luciferase activity, it was predicted that co-transfection of HuD with the Luc- ARUC constructs would increase luciferase activity from the Luc-ARUC-WT vector, but not the Luc-ARUC-MutB vector. However, co-transfection of FLAG-HuD plasmid did not significantly affect the luciferase activity from the pGL3-MCS-empty, Luc-ARUC Mut B or ARUC-WT constructs (Figure 3.6). FLAG-HuD was also co-transfected with a luciferase construct which contained the AU-rich element from c-fos in its 3’UTR (Luc-c-fos ). Although the interaction between HuD and the c-fos mRNA is well characterised, no functional effect of this interaction has been reported (Chung et al. , 1996; Wang and Tanaka-Hall, 2001). The observed reduction in induced luciferase activity from the Luc-c-fos vector when it was co-transfected into LNCaP cells with the FLAG-HuD vector indicated that the binding of HuD to this sequence may have a destabilising effect on luciferase mRNA (Figure 3.6).
The AR mRNA has been previously reported to exhibit increased stability following DHT treatment of LNCaP cells, but the mechanism by which this occurs in unknown. To determine whether the presence of the ARUC sequence could stabilise luciferase mRNA following DHT treatment, LNCaP cells were cultured in the presence and absence of DHT following transfection with the Luc-ARUC vectors. An increase in luciferase activity from the Luc- ARUC-WT construct, but not from the Luc-MutB construct following DHT treatment would be expected if the ARUC sequence was involved in DHT- mediated stabilisation of mRNA, however DHT treatment did not increase induced luciferase activity from the Luc-ARUC-WT construct, indicating the ARUC sequence is not involved in DHT-mediated stabilisation (Figure 3.6). To determine whether the ARUC-WT sequence required the presence of both HuD and DHT to mediate stabilisation of luciferase mRNA, the Luc-ARUC-WT or Luc-ARUC-MutB vectors were co-transfected into LNCaP cells with the FLAG-HuD vector, and DHT added to the culture medium. An increase in
90 Relative Luciferase Activity (%)
0 20 40 60 80 100 120 140
pGL3-MCS-empty
Luc-ARUC-WT
Luc-Mut B
Luc-c-fos *
pGL3-MCS-empty
+ DHT Luc-ARUC-WT
Luc-Mut B
Figure 3.6. Effects of HuD expression on luciferase reporter gene activity.
LNCaP cells were co-transfected with empty FLAG vector ( ) or FLAG-HuD ( ) and pGL3-MCS-empty, Luc-ARUC-WT, Luc-ARUC-MutB or Luc-c-fos as indicated. Luciferase activity was measured after 24 h. Cells were cultured in the presence of vehicle or DHT (10 nM) as indicated. The data represent mean ± sd of quadruplicate determinations, and the induced luciferase activity from pGL3-MCS- empty/FLAG-empty/no DHT was expressed as 100 %. Asterisks indicate means which are significantly different (p < 0.05).
91 induced luciferase activity from the Luc-ARUC-WT construct was not observed, indicating that binding of HuD to the ARUC sequence is not involved in DHT-mediated stabilisation of mRNA in this assay (Figure 3.6).
3.2.5 HuD Effects on AR Expression and Activity.
To assess effects of HuD on endogenous AR protein levels, LNCaP cells were transfected with increasing amounts (125-500 ng) of FLAG-HuD plasmid, and AR levels were determined by western blotting. FLAG-HuD levels were progressively elevated following transfection of cells with increasing amounts of the FLAG-HuD vector, however AR protein levels were increased only by ~ 5 % in cells transfected with the highest amount of FLAG-HuD compared to FLAG-empty-transfected cells, as determined by densitometry (Figure 3.7A). PSA secretion from FLAG-HuD-transfected LNCaP cells was also assayed as a readout of AR transcriptional activity, but no difference was observed between cells transfected with FLAG-HuD and those transfected with an empty FLAG vector (Figure 3.7B). PSA secretion increased following treatment with DHT, however these increases were not significantly different in cells transfected with the FLAG-HuD or FLAG-empty vectors (Figure 3.7B).
The effects of HuD on AR transcriptional activity were investigated in LNCaP cells using the androgen-responsive vectors, PSA-Luc and PB3-Luc, with induced firefly luciferase activity normalised to Renilla luciferase activity induced from the pRL-TK vector (Section 2.2; 2.7). In initial experiments it was found that co-transfection of cells with the FLAG-HuD vector resulted in markedly reduced levels of induced firefly and Renilla luciferase activities in comparison to cells co-transfected with the FLAG-empty vector. This was evident when using the PSA-Luc (Figure 3.8) and PB3-Luc vectors (data not shown). Despite extensive efforts to optimise transfection conditions and relative amounts of vectors used, this effect was always observed. Although the reasons for this are unknown, it is feasible that HuD over-expression may have had direct or indirect effects on luciferase expression, mediated via regulation of mRNA stability or protein translation. As induced Renilla luciferase activity was consistent between transfections including FLAG-empty or FLAG-HuD vectors (Figure 3.8A), firefly luciferase activity was normalised 92 B A ftaseto.Tedt ersn en±s of dupl sd transfection.of dataThe represent ± mean ( vehicle with treated were Cells vector. FLAG 500 with transfection after h 24 measured was cells A. PS and ARlevels protein on effects 3.7. HuD Figure by western blotting after 24 h. 24 after blotting western by pr actin and AR500 and 8; 7 and ng), lanes250 ng; 1 4, and 3 (lanes FLAG-HuD of amounts increasing or LNCaP were transfected with empty FLAG vector (500 n (500 vector FLAG empty with transfected were LNCaP
PSA secretion (ng/L) 100 120 40 20 60 80 0 FLAG-empty 8 7 6 5 4 3 2 1 LGepyFLAG-HuD FLAG-empty B. B. Secreted PSA present in media from LNCaP from media in present PSA Secreted FLAG-HuD ) or 10 nM DHT ( ) at time time at ) ( DHT nM 10 or ) otein levels were assessed assessed were oteinlevels icate measurements. icate measurements. ng FLAG-HuD or empty empty or FLAG-HuD ng A secretion. 5 g lns ad 6, and 5 lanes ng; 25 10 nM DHT Vehicle g, lanes 1 and 2)and 1 lanes g, FLAG-HuD Actin AR 93 PSA-Luc pRL-TK A * 300 * 250 * 200 150 100 50
Luciferase Activity (%) Luciferase Activity 0 --+ + DHT FLAG-empty FLAG-HuD
Vehicle B 10 nM DHT 800 700 600 500 400
Activity (%) Activity 300
Relative Luciferase RelativeLuciferase 200 100 0 FLAG-empty FLAG-HuD
Figure 3.8. Effects of exogenous HuD expression on AR transcriptional activity.
A. FLAG-HuD or empty FLAG vector was co-transfected into LNCaP with the androgen-responsive reporter PSA-Luc ( ) and thymidine kinase-Renilla (pRL-TK ) vectors. Cells were treated with vehicle (-) or 10 nM DHT (+) at time of transfection as indicated, and luciferase activity was assayed after 24 h. The data represent mean ± sd of quadruplicate measurements. B. The induced luciferase activity from the PSA-Luc vector in A was normalised to that from the pRL-TK vector. : vehicle-treated. : 10 nM DHT-treated. These data represent mean ± sd of quadruplicate samples. * = p < 0.05.
94 using Renilla luciferase activity to determine HuD effects on DHT-induced luciferase activity from the androgen responsive reporter vectors. Treatment of PSA-Luc-transfected cells with 10 nM DHT was found to result in a 3.1-fold increase in induced luciferase activity in cultures co-transfected with the FLAG- empty vector and a 3.8-fold increase in cultures co-transfected with FLAG-HuD (p > 0.05; Figure 3.8B). Similar results were obtained in transfections using the PB3-Luc vectors (data not shown). Therefore, using this assay, significant HuD effects on AR transcriptional activity could not be detected.
3.3 Discussion. In this study, expression of HuD protein was detected in 12 of 12 primary prostate tumours, and in the LNCaP prostate cancer cell line. Although the cohort of samples was small, the finding of HuD expression in all specimens was significant and the ability to compare HuD expression in non-malignant prostate glands and prostate tumours in individual sections, which identified consistently elevated HuD levels in malignant cells, provided convincing evidence of up-regulation of HuD expression in prostate cancer. Previous analyses of HuD protein expression have found it to be restricted to tissues from the central nervous system or tumours of neuroendocrine origin (Ma et al. , 1996; Manley, 1995). Immunohistochemical detection of a neuronal Hu protein in a single human prostate tumour (from a cohort of ten specimens) has been reported, but this study was carried out prior to the development of the HuD-specific 16C12 antibody, therefore the protein was not definitively identified (Dalmau et al. , 1992). Interestingly, in that study, the Hu protein was expressed predominantly in the nuclei of cells in positively-staining hippocampus and malignant lymph node specimens, in contrast to the present study where HuD was localised in the cytoplasm of both non-malignant prostate epithelial and prostate cancer cells (Dalmau et al. , 1992). Neuronal Hu proteins including HuD are hypothesised to play different roles in the cytoplasm and the nucleus of cells, with cytoplasmic neuronal Hu proteins regulating mRNA stability and/or translation and nuclear Hu proteins functioning as RNA processing regulators (Zhu et al. , 2006). As such, HuD activity may differ in neuronal tissues and in prostate cancer cells, and will require further investigation. 95
Splice variants of HuD have been detected in cDNA libraries derived from a range of tissues, including the normal prostate, however the detection method was not quantitative, and expression of the encoded protein(s) was not reported (Behrends et al. , 2002). Quantitative measurements of HuD mRNA expression in the publicly-available database SymAtlas (http://symatlas. gnf.org/SymAtlas/) indicated median expression of HuD cDNA in prostate tissue, while regions of the brain and pancreatic islet cells expressed HuD at > 5 x the median level (Su et al. , 2002). The LNCaP and ALVA-31 prostate cancer cell lines were reported to express median levels of HuD, which is consistent with the results of the present study indicating HuD mRNA expression in LNCaP cells. Although HuD mRNA is expressed in LNCaP cells, the lack of detectable HuD protein suggested that if expressed, endogenous HuD protein levels in this cell line are not high. In addition, the effect of passage number on HuD steady-state levels is uncharacterised. Furthermore, cellular localisation of endogenous HuD in LNCaP cells could not be carried out in this project and it remains unknown whether HuD is nuclear or cytoplasmic in LNCaP cells or whether the failure to detect HuD by western blotting in contrast to the evidence of HuD-containing complexes (RPC's) from LNCaP cell extracts is due to expression of HuD protein in only a small proportion of LNCaP cells.
The finding of increased HuD protein expression in higher Gleason grade prostate tumours was interesting, but the implications of this observation are unknown. Up-regulation of cellular HuD protein levels is likely to have wide- ranging effects, as HuD can bind numerous mRNA’s, and the inappropriate stabilisation of some of these targets could result in altered levels of the encoded proteins, thereby affecting many cellular processes. Clearly, the elucidation of HuD target mRNA’s in prostate cancer cells will be critical for understanding the consequences of HuD up-regulation in prostate tumours.
Proteins from LNCaP cell extracts bound the AU-rich element from the c-fos mRNA 3’UTR in REMSA’s. The ability of HuD antibodies to supershift a proportion of RPC’s formed between the c-fos probe and LNCaP extracts
96 suggested that HuD protein was expressed in LNCaP cells and that it bound this sequence, which is a canonical HuD target sequence (Wang and Tanaka- Hall, 2001). This observation was supported by the binding of purified GST- HuD to the c-fos probe.
Binding of HuD to the ARUC region was demonstrated by REMSA using extracts from LNCaP cells transfected to over-express FLAG-HuD, providing evidence that HuD, as well as HuR, is able to bind this sequence (Yeap et al. , 2002). The absence of detectable anti-HuD supershifted complexes in REMSA's using extracts from untransfected LNCaP cells indicated that endogenous HuD levels may be low. Furthermore, the prominence of RPC’s containing the ARUC probe and LNCaP cell extracts that were not supershifted by anti-HuD antibodies suggests that additional proteins also bound this sequence. Other proteins involved in the formation of the RPC’s may include HuR and αCP1/ αCP2, which have previously been reported to bind the ARUC region (Yeap et al. , 2002). It is also possible that the HuD- specific antibody used in this assay was not able to efficiently supershift the HuD/ARUC probe complex, although extracts from PC12 cells bound to the c- fos probe were efficiently supershifted using this antibody. The use of an anti- FLAG antibody may have been more effective in experiments utilising transfected LNCaP extracts that over-express FLAG-HuD and would have provided confirmation of supershift assays using anti-HuD antibodies. However, the binding of HuD to the AR mRNA was supported by the co- immunoprecipitation of HuD and the AR mRNA in IP-RT-PCR assays, together providing strong evidence of HuD binding to the ARUC sequence.
Because HuD is reported to stabilise mRNA following binding to target sequences in the 3’UTR, experiments were designed to evaluate the effects of the ARUC on mRNA stability (de Silanes et al. , 2003; Guo and Hartley, 2006; Nabors et al. , 2003). Mutations of the ARUC region of the ARUC probe resulted in reduced binding of purified HuD, suggesting that if the ARUC and HuD were involved in regulation of AR mRNA stability, these effects may be in part due to the amount of HuD bound to the ARUC sequence. Alternatively, other features of HuD binding such as its affinity and potentially the ability to
97 recruit other RNA stabilising factors may also contribute to HuD effects on a target mRNA.
Insertion of the ARUC sequence into the 3’UTR of a luciferase vector did not result in an increase in induced luciferase activity compared to that induced from the vector without insert. There are a number of potential explanations for this. In particular, the use of induced luciferase activity as a correlate of mRNA stability assumes a direct relationship between luciferase mRNA synthesis and stability, as well as mRNA stability and protein production. The stability of the luciferase protein in the transfected cells is also unknown. Future experiments could therefore use northern blotting or qPCR to detect the luciferase mRNA to assess stability, however there are technical difficulties associated with mRNA measurement following transfection due to poor RNA quality as a result of the transfection procedure (results not shown).
While investigating the apparent failure of the wild-type ARUC sequence to increase induced luciferase activity in comparison to induced luciferase activity from the pGL3-MCS-empty vector, it was found that introduction of the ARUC region into the 3’UTR of the pGL3-MCS vector altered the structure of the entire 3’UTR, potentially disrupting the binding of multiple RNA stabilising or processing proteins. The ARUC-WT sequence may not have been able to compensate for these losses, an effect enhanced by the loss of HuD binding in the MutA and MutB constructs. Furthermore, insertion of the ARUC sequence may attract both stabilising and destabilising AU-rich RNA binding proteins. Destabilising proteins such as AUF1 and KSRP have been found to bind concurrently and competitively to HuR target transcripts, so may also compete with HuD for binding to AU-rich sequences (Lal et al. , 2004; Linker et al. , 2005). AUF1 and KSRP are both expressed in LNCaP cells and in prostate tissue (http://symatlas.gnf.org/SymAtlas/) (Su et al. , 2002). It is also important to note that the mutations used disrupt not only HuD binding, but also equivalently disrupt HuR binding (Yeap et al. , 2002). Therefore, HuR may also contribute to stabilisation of ARUC-containing mRNA sequences, especially when it is considered that HuR is endogenously expressed in LNCaP cells at relatively high levels.
98
Prior experiments using a similar ARUC-WT sequence cloned into a different parent vector, pRSV-Luc, resulted in a small decrease in luciferase activity, suggesting a destabilising effect of the ARUC sequence in LNCaP cells (Yeap et al. , 2002). A major difference between pRSV-Luc and the pGL3-MCS vector used in the present study is that pGL3-MCS contains an SV40 enhancer element while pRSV-Luc does not. As presence of the strong SV40 enhancer may mask more subtle effects of the ARUC sequence, experiments could be repeated following insertion of the ARUC wild-type and mutant sequences into pRSV-Luc or a similar vector that does not contain a strong enhancer element. In addition, use of a vector which did not contain a 3’ enhancer element would avoid disruption of the structure of the 3’ region and its pre-existing RNA stabilising and destabilising proteins.
The effects of HuD on the ARUC were further investigated following transfection of cells to over-express FLAG-HuD, however induced luciferase activity was not altered in the presence of increased HuD expression. This indicates that HuD may not bind the ARUC region within the context of the pGL3-MCS 3’UTR or that HuD binding does not stabilise the luciferase mRNA. Alternatively, the inclusion of the ARUC sequence in the 3’UTR region or binding of HuD to the ARUC sequence may alter the binding of other proteins which are required for stabilising or processing of the luciferase mRNA, such that the net effect on luciferase activity is zero. In addition, levels of HuD may not have been high enough to exert any physiological affect. Previously, DHT treatment has been demonstrated to increase the stability of the AR mRNA in LNCaP cells, however the mechanism by which this occurs is unknown (Yeap et al. , 2002). As it was feasible that the ARUC sequence mediated the stabilisation of the AR mRNA, the effects of DHT treatment on luciferase activity induced from vectors containing the wild-type and mutant (MutB) ARUC sequences were also determined. The lack of effects on induced luciferase activity may have arisen because the effects of DHT on the AR mRNA are not mediated by HuD (or HuR) effects on the ARUC sequence, or it may be due to the fact that the technical issues already mentioned prevent detection of DHT-induced changes in luciferase activity.
99
An interesting finding in this study was that co-transfection of FLAG-HuD with the Luc-c-fos AU-rich element construct appeared to result in decreased luciferase activity from that construct. Previous studies of the relationship between HuD and the c-fos AURE have been limited to investigations of the binding interaction rather than of the functional effects of that binding (Chung et al. , 1996; Wang and Tanaka-Hall, 2001). Although it is feasible that insertion of the c-fos AURE into the luciferase 3’UTR disrupted the structure of the 3’UTR thereby preventing detection of increased stability due to effects on other mRNA processing of stabilising proteins, it is also possible that the binding of HuD to this element results in mRNA destabilisation.
HuD over-expression following transfection of LNCaP cells with a vector encoding FLAG-HuD resulted in modest increases in AR protein levels. The increased AR protein may have resulted from direct effects of HuD on AR mRNA and thereby protein, or alternatively may have occurred via HuD effects on other mRNA’s, leading to alteration in AR steady-state levels. In light of FLAG-HuD effects on AR protein levels, the absence of significant effects of HuD over-expression on basal or DHT-stimulated PSA secretion are likely to result from the lack of sensitivity of this assay which would not detect the small changes in AR protein levels and thus AR activity on the PSA promoter. Similarly, only modest increases in DHT-induced luciferase activity were detected from a PSA-Luc reporter vector following FLAG-HuD over- expression. These findings indicate that HuD effects on AR levels and activity are not marked using the assays employed in this thesis. However, the effects of long-term HuD over-expression either on AR levels and activity or on the levels and activity of other HuD targets are unknown. These initial experiments have therefore identified HuD over-expression in human prostate tumours and binding of HuD to the ARUC region, which suggests that one of the consequences of HuD over-expression is altered AR levels and therefore activity. While examination of HuD activity presented many technical difficulties associated with limitations of currently available assays, the co- expression of HuD and HuR in prostate cancer cells and the consequences of their expression on down-stream pathways are clearly important given the
100 wide range of mRNA targets bound by these two proteins and the finding that HuD levels are comparatively elevated in prostate cancer cells.
101
CHAPTER 4:
Regulation of the AR by HuR.
102 4 Chapter 4: The Regulation of AR by HuR.
4.1 Introduction. HuR is an ubiquitously expressed protein of the ELAV-like/Hu protein family that is thought to stabilise diverse mRNA’s encoding proteins with established roles in cancer, such as β-catenin, p53, several cyclins, and p21 Waf1 (Table 1.2). The role of HuR in cancer has been investigated in preliminary studies and HuR expression and/or cytoplasmic localisation are reported to be correlated with increasing malignancy in human breast, ovarian and colon cancers, with HuR expression in RKO cells positively correlating with tumour size in nude mice (de Silanes et al. , 2003; Erkinheimo et al. , 2003; Heinonen et al. , 2005) (Section 1.4.3). As HuR has been shown to bind the UC-rich region of AR mRNA in vitro and in vivo, it is postulated that HuR may contribute to the control of the AR levels in the prostate and in prostate cancer cells (Yeap et al. , 1999; Yeap et al. , 2002) (Section 1.5.1).
HuR has been observed to accumulate in polyribosomes following DHT treatment of HepG2 cells, and DHT regulates the binding of HuR to the hypoxia inducible factor 1 α (HIF1 α) and EGF mRNA’s, suggesting mechanisms by which HuR may be involved in DHT-mediated stabilisation of mRNA (Sheflin et al. , 2001; Sheflin et al. , 2004). These findings suggest that HuR may affect prostate tumour behaviour via regulation of AR protein levels and transcriptional activity, or by regulation of non-AR pathways such as Wnt signalling, as is postulated to be the case in colon cancer (de Silanes et al. , 2003). The aim of the following studies was to investigate the functional relationship between HuR expression and AR expression and activity, and to determine whether changes in HuR expression impacted prostate cancer cell behaviour.
103 4.2 Results.
4.2.1 Characterisation of AR in 22Rv1.
22Rv1 cells had not been used previously in this laboratory, therefore in preliminary experiments the androgen responsiveness of 22Rv1 cells was established. In order to compare the androgen responsiveness of 22Rv1 and LNCaP cells, both cell types were transfected with PSA-Luc or PB3-Luc and induced luciferase activity was measured following treatment with 10 nM DHT and increasing concentrations of the anti-androgen, bicalutamide. A strong (> 100-fold) up-regulation of luciferase activity induced from the PB3-Luc vector was observed in both cell lines in cultures containing 10 nM DHT (Figure 4.1). LNCaP cells exhibited a greater up-regulation of luciferase activity, but produced lower basal and overall luciferase counts, whereas 22Rv1 exhibited higher basal and total counts (data not shown). In both cell lines, luciferase activity induced from the PB3-Luc vector could be repressed by the addition of 100-fold excess bicalutamide, while 10-fold excess bicalutamide had little impact on DHT-induced luciferase activity (Figure 4.1). This is consistent with previously published data (Masiello et al. , 2002). Similar results were observed in experiments using the PSA-Luc vector (data not shown).
Androgens are known to stabilise and up-regulate AR protein levels in LNCaP cells (Krongrad et al. , 1991). To compare DHT regulation of the AR in LNCaP and 22Rv1 cells, the cells were plated at a cell density of 5.0 x 10 4 cells/cm 2 in RPMI 1640 medium containing 10 % charcoal-stripped-FCS (Section 2.7.2). Treatment of cultures with 10 nM DHT for 8 and 24 h increased AR protein levels to ~ 210 % and ~ 190 % of controls, respectively, in LNCaP cells (Figure 4.2A). In 22Rv1 cells, the 114 kDa AR protein increased to ~ 150 % of controls after 8 h, and ~ 250 % of controls after 24 h of DHT treatment while the 75 kDa AR variants were increased to ~ 130 % and ~ 140 % of that in control cultures, respectively (Figure 4.2B).
104 A 600
500
400
300 (foldchange) 200 Relative Luciferase Activity RelativeLuciferase Activity 100
0 Untreated 10 nM DHT 10 nM DHT 10 nM DHT 100 nM bical 1 000 nM bical B 140
120
100
80
(foldchange) 60
40 Relative Luciferase Activity RelativeLuciferase Activity 20
0 Untreated 10 nM DHT 10 nM DHT 10 nM DHT 1 000 nM bical 10 000 nM bical
Figure 4.1. AR transcriptional activity in LNCaP and 22Rv1 cells. LNCaP ( A) or 22Rv1 ( B) cells were transfected with the PB3-Luc vector and cultured in the presence of 10 nM DHT and increasing concentrations of bicalutamide as indicated. Luciferase activity was assayed at 24 h. Values are expressed relative to the luciferase activity of untreated cultures (fold induction). These data represent the mean ± SD of triplicate samples, normalised to Renilla luciferase activity.
105 A B LNCaP 22Rv1
8 h 24 h 8 h 24 h 12 3 4 56 7 8 1 2 3 4 56 7 8 AR (114 kDa)
AR (~75 kDa)
Actin (45 kDa)
DHT DHT DHT DHT
Figure 4.2. DHT regulation of AR protein levels in LNCaP and 22Rv1 cells. LNCaP ( A) or 22Rv1 ( B) cells were untreated (lanes 1, 2, 5 and 6) or were treated with 10 nM DHT (lanes 3, 4, 7 and 8). AR and actin protein levels were assessed by western blotting after 8 h (lanes 1 to 4) or 24 h (lanes 5 to 8).
106
4.2.2 HuR Effects on AR Protein Expression and Activity.
To determine if increased expression of HuR affected AR protein levels, LNCaP cells were transfected with the FLAG-HuR vector, and expression of HuR and AR were analysed by western blotting after 24 and 48 h. At 24 h following transfection, FLAG-HuR was barely detectable while at 48 h, prominent FLAG-HuR protein was evident in cell extracts (Figure 4.3). The later accumulation of FLAG-HuR protein in transfected LNCaP cells may be due to a low transfection efficiency or may reflect the long division time of these cells (~ 40 h; ATCC; www.atcc.org/common/catalog/numSearch/num Results.cfm?atccNum=CRL-2505). AR protein levels were not markedly altered in the transfected cells, however FLAG-HuR levels were always lower than endogenous HuR, and when total HuR (endogenous and FLAG-HuR) was evaluated using densitometry, transient transfection had only increased HuR levels by ~ 10 % (Figure 4.3).
The effects of increased HuR expression on AR transcriptional activity were also examined using luciferase activity induced from the androgen-responsive PSA-Luc reporter vector. Similar to previous experiments using these vectors (Section 3.2.5), transfection of LNCaP cells with the FLAG-HuR vector decreased basal luciferase and Renilla luciferase activity in vehicle-treated cultures (Figure 4.4A). As expected, DHT treatment of FLAG-empty and FLAG-HuR-transfected cells resulted in increased luciferase activity induced from the PSA-Luc vector. When normalised to Renilla luciferase activity, DHT had induced a 2.4-fold increase in induced luciferase activity from FLAG- empty-transfected cells and a 1.9-fold increase in induced luciferase activity from FLAG-HuR-transfected cells, differences that were not statistically significant (p > 0.05; Figure 4.4B). Because endogenous HuR levels in LNCaP cell were high and therefore the increase of overall HuR levels was not substantial following transfection with FLAG-HuR, transfection of siRNA directed against HuR was performed in subsequent experiments to examine the cellular effects of reduced HuR levels.
107 12 3 4 AR (114 kDa)
Actin (45 kDa) FLAG-HuR (42 kDa)
Endogenous HuR (36 kDa)
24 h 48 h 24 h 48 h
FLAG-empty FLAG-HuR
Figure 4.3. Expression of exogenous HuR does not affect AR protein levels. LNCaP cells were transfected with 500 ng empty FLAG (lanes 1 and 2) or FLAG- HuR plasmid (lanes 3 and 4) and expression of AR, actin, FLAG-HuR, and endogenous HuR was assessed by western blotting after 24 h (lanes 1 and 3) and 48 h (lanes 2 and 4). This experiment was performed in the absence of DHT.
108 A
300 PSA-Luc pRL-TK 250
200
150
Luciferase Activity (%) Luciferase Activity 100
50
0 --+ + DHT FLAG-empty FLAG-HuR
B Vehicle 350 10 nM DHT 300
250
200
150
100
50 Relative Luciferase Activity (%) RelativeLuciferase Activity
0 FLAG-empty FLAG-HuR
Figure 4.4. HuR effects on AR transcriptional activity.
A. FLAG-HuR or FLAG-empty vector was co-transfected into LNCaP cells with the androgen-responsive reporter PSA-Luc ( ) and thymidine kinase-Renilla vectors (pRL-TK ). Cells were cultured vehicle (-) or 10 nM DHT (+) as indicated at the time of transfection and induced luciferase activity was assayed after 24 h. Values are expressed relative to the luciferase activity of FLAG-empty-transfected cells, which was taken as 100 %. The data represent mean ± sd of triplicate measurements. B. Luciferase activity induced from the PSA-Luc vector was normalised using Renilla luciferase activity induced from the pRL-TK vector. : Vehicle-treated cells. : 10 nM DHT- treated cells. Data represent mean ± sd of triplicate measurements.
109
4.2.3 Effects of Reduced HuR Expression on AR Protein Levels and Activity.
Unlike HuD, which is expressed at very low levels in LNCaP cells, the highly expressed HuR protein is a reasonable target for siRNA-mediated knockdown (Section 3.2.1). In initial experiments, LNCaP cells plated at 2.5 x 10 4 cells/cm 2 (~ 30 % confluence) and 4.5 x 10 4 cells/cm 2 (~ 70 % confluence) were transfected with 20-200 nM siRNA and analysed by western blotting for actin and HuR expression (Section 2.8). The most effective concentrations of the two siRNAs were found to be 20 nM for siHuR-Sully and 50 nM for siHuR- UTR (Figure 4.5 and results not shown). HuR knockdown was found to be highly dependent on cell density, with cells plated at 2.5 x 10 4 cells/cm 2 exhibiting a greater and more consistent knockdown of HuR than 4.5 x 10 4 cells/cm 2 (Figure 4.5A, B). Maximal HuR knockdown was most consistently found at 72 h, but persisted until 120 h, the latest time-point analysed (Figure 4.5C).
Following optimisation of siRNA treatment, AR protein levels were measured in cells with a confirmed knockdown of HuR. Knockdown of HuR by ~ 70 % following siHuR-Sully siRNA treatment resulted in a concurrent decrease of ~20 % in AR protein in LNCaP cells compared to siMut siRNA-treated cells, as determined by western blotting followed by densitometry (Figure 4.6). Addition of DHT (10 nM) to the culture medium did not affect the magnitude of decrease of the AR protein (data not shown). This effect was contingent on a significant decrease of HuR; siRNA treatments that only knocked down HuR by ~ 30-40 % did not result in a detectable decrease in AR protein levels (data not shown).
To ascertain whether the reduction of AR after HuR knockdown was transmitted into a change in AR transcriptional activity, LNCaP cells were transfected with siHuR-Sully or control siMut, cultured for 48 h and then transfected with the PSA-Luc and pRL-TK plasmids (Section 2.8.2). Treatment
110 A 1 2 34 5 6 7 8 Actin (46 kDa) 2.5 x 10 4 cells/cm 2 HuR (36 kDa)
siMut siCon siHuR- siHuR- Sully UTR
B 12 3 4 5 6 7 8 Actin (45 kDa) 4.5 x 10 4 cells/cm 2 HuR (36 kDa)
siMut siCon siHuR- siHuR- Sully UTR
12 3 4 5 6 78 9 Actin (45 kDa) C 72 h HuR (36 kDa)
Actin (45 kDa) 96 h HuR (36 kDa)
Actin (45 kDa) 120 h HuR (36 kDa)
Lipo siHuR- siCon UTR
Figure 4.5. Effectiveness of HuR knockdown is dependent on cell density. LNCaP cells were transfected at a cell density of 2.5 x 10 4 cells/cm 2 (A) or 4.5 x 10 4 cells/cm 2 (B) with 50 nM siMut (lanes 1 and 2), siCon (lanes 3 and 4), siHuR- Sully (lanes 5 and 6) or siHuR-UTR (lanes 7 and 8), and assessed for HuR and actin expression after 72 h. C. LNCaP cells were treated with LipoFectamine 2000 alone (Lipo; lanes 1 to 3), or 50 nM siHuR-UTR (lanes 4 to 6) or siCon (lanes 7 to 9), and assessed by western blotting for HuR and actin expression after 72 h, 96 h and 120 h as indicated.
111 12 3 4 5 6 78 9 AR
Actin
HuR
siHuR- siMut Lipo Sully
Figure 4.6. Knockdown of HuR using siRNA decreases AR protein levels. LNCaP cells were transfected with 50 nM siHuR-Sully (lanes 1 to 3), siMut (lanes 3 to 6), or were treated with LipoFectamine 2000 alone (Lipo; lanes 7 to 9). After 72 h, levels of AR, actin, and HuR proteins were assessed by western blotting.
112 with either siHuR-Sully or the control (siMut) siRNA’s decreased luciferase activity induced from both reporters compared to control cultures treated with LipoFectamine 2000, and these effects were independent of the amount of siRNA transfected (Figure 4.7A). When the induced luciferase activity from the PSA-Luc vector was normalised using the Renilla luciferase activity induced from the pRL-TK vector, treatment of the cells with siHuR-Sully siRNA did not significantly affect induced luciferase activity from the PSA-Luc vector compared to cells treated with LipoFectamine alone. Conversely, treatment of the cells with siMut control siRNA significantly increased luciferase activity induced from the PSA-Luc vector (p<0.05; Figure 4.7B). Due to the pronounced effects of siRNA treatments on Renilla luciferase activity induced from the pRL-TK plasmid, a second Renilla control vector, phRL-SV40, was tested in identical experiments, however the results were similar (data not shown).
As the control siRNA, siMut, was associated with greater reductions in induced luciferase actively from the PSA-Luc and pRL-TK vectors in comparison to the HuR-targeting siRNA, siHuR-Sully, a second set of siRNA was trialled, using 22Rv1 cells and a different androgen responsive reporter, PB3-Luc, but no Renilla luciferase vector. The effects of these siRNAs on luciferase activity from the PB3-Luc vector were variable, however, both the HuR-specific siHuR- UTR and control siCon siRNA’s decreased induced luciferase activity from the PB3-Luc vector. The differences between the induced luciferase activity in siHuR-UTR-, siCon- or LipoFectamine-alone treated cells were not significant in the either the presence or absence of DHT (p > 0.05; Figure 4.8A). DHT- induced PSA secretion from siHuR-UTR-treated 22Rv1 cells was increased in comparison to cultures treated with siCon, however this increase was not significant (p > 0.05; Figure 4.8B).
HuR has been reported to bind and stabilise numerous mRNA’s, therefore it was hypothesised that the most likely mechanism by which HuR affected AR protein levels was by modulating the half-life of the AR transcript (Section 4.2.3). To determine whether this was the case, the degradation of AR mRNA
113 A PSA-Luc 120 pRL-TK 100
80
60
40
Relative Luciferase Activity (%) RelativeLuciferase Activity 20
0 20 50 80 20 50 80 Lipo siHuR-Sully (nM) siMut (nM) Treatment
B * 350 * 300 * 250
200
150
100
50 Relative Luciferase Activity (%) RelativeLuciferase Activity 0 20 50 80 20 50 80 Lipo siHuR-Sully (nM) siMut (nM) Treatment
Figure 4.7. Effects of siRNA on luciferase activity induced from the PSA- Luc vector.
A. LNCaP cells were treated with different concentrations of siHuR-Sully or siMut, or LipoFectamine 2000 alone (Lipo), as indicated. After 48 h of siRNA treatment the cells were co-transfected with the androgen-responsive reporter PSA-Luc ( ) and thymidine kinase-Renilla (pRL-TK; ) vectors. Cells were treated with 10 nM DHT at time of transfection and luciferase activity was assayed after 24 h. Values are expressed relative to the luciferase activity of LipoFectamine 2000 treated cells. B. The induced luciferase activity from the PSA-Luc vector in A was normalised to that from the pRL-TK vector. These data represent mean ± sd of quadruplicate samples. * = p < 0.05.
114 A 600 500 400 300 200
Luciferase Activity (%) Luciferase Activity 100 0 --- + + + DHT siHuR- siCon Lipo siHuR- siCon Lipo UTR UTR Treatment
B
0.30
0.20
0.10 PSA secretion (ng/L) PSA
0 siHuR-UTR siCon
siRNA
Figure 4.8. Effects of HuR knockdown by siHuR-UTR siRNA on AR transcriptional activity and PSA secretion.
A. 22Rv1 cells were treated with 50 nM siHuR-UTR, 50 nM siCon or Lipofectamine alone (Lipo) as indicated. After 48 h of siRNA treatment the cells were co-transfected with the androgen-responsive reporter PB3-Luc and cultured in the presence of vehicle (-) or 10 nM DHT (+) as indicated. Luciferase activity was assayed after 72 h of siRNA treatment, and normalised to the protein concentration of the cell lysates. Values are expressed relative to the luciferase activity of LipoFectamine 2000-treated cells, which was taken as 100 %. B. 22Rv1 cells were treated with 50 nM siHuR-UTR or siCon. After 48 h, 10 nM DHT was added, and 24 h after treatment with DHT, the levels of secreted PSA were measured. These data represent mean ± sd of triplicate samples.
115 in 22Rv1 cells was monitored following addition of the transcriptional inhibitor, actinomycin D (ActD), to the culture medium 72 h after treatment of the cells with siRNA directed against HuR (Section 2.15.5). qPCR was used to estimate the remaining level of AR mRNA relative to actin mRNA level at 4 h intervals. During the 12 h of ActD treatment, AR mRNA levels were not significantly decreased in siCon siRNA-treated cells, but the half-life of the AR mRNA in cells treated with siHuR-UTR was significantly decreased compared to siCon-treated cells to ~ 12 h (Figure 4.9A). In contrast to the effect on AR protein, a HuR knockdown of ~ 40 % was sufficient to decrease AR mRNA stability (Figure 4.9B).
Steady-state AR mRNA levels were also measured after 72 h of treatment with siHuR-UTR siRNA and although there was a trend towards decreased AR mRNA levels in both LNCaP (Figure 4.10A) and 22Rv1 cells (Figure 4.10B) this decrease was not significant (p > 0.05). Addition of DHT to the culture medium 24 h before collection of RNA did not affect the magnitude of the decrease in AR mRNA levels following treatment with siHuR-UTR siRNA (Figure 4.10A, B). The AR mRNA levels in cells treated with DHT versus vehicle-treated cells were not measured in the same qPCR experiment, and therefore could not be compared to determine the effect of DHT treatment alone on AR mRNA levels in the LNCaP and 22Rv1 cell lines.
4.2.4 Localisation of HuR.
Hu proteins are nucleo-cytoplasmic shuttling proteins, and move into the cytoplasm in response to a number of stimuli. Cytoplasmic accumulation of HuR and HuD is associated with stabilisation of target transcripts, and stimuli for the nucleo-cytoplasmic shuttling of these proteins include DHT and ActD (Fan and Steitz, 1998b; Sheflin et al. , 2001; Sheflin et al. , 2004). Because treatment of LNCaP cells with DHT was associated with an increase in mRNA stability, and HuR appears to be involved in the maintenance of AR mRNA stability in 22Rv1 cells, the effect of DHT treatment on the intracellular localisation of HuR in 22Rv1 cells was determined using confocal microscopy (Section 2.16). HuR was > 90 % nuclear in vehicle-treated cells and in cells treated with 10 nM DHT for 4 h (Figure 4.11A, B). Comparable results were 116 10.0 A siHuR-UTR siCON
y = 0.012x + 1
1.0 *
Remaining mRNA AR (log) y = -0.0428x + 1
0.1 0 4 8 12 Time (h)
1 2 B Actin HuR
Figure 4.9. Knockdown of HuR by siRNA significantly decreases AR mRNA half-life in 22Rv1 cells.
A. 22Rv1 cells were treated with 50 nM siHuR-UTR or siCon. After 72 h of siRNA treatment, 7.5 µg/µL actinomycin D was added and cells were harvested for RNA extraction every 4 h. AR and actin mRNA levels were measured by qPCR, and quantitated relative to AR mRNA levels at 0 h. These data represent mean ± sd of 5 replicates. * = p<0.05 (ANOVA). The formulae for the decay trendlines are given. B. Western blotting confirmed the knockdown of HuR in siHuR-UTR-treated cells (lane 2) compared to siCon- treated cells (lane 1).
117 A 140 120 100 80 60 40 20 Relative AR mRNA Relativelevel (%) AR 0 siHuR-Sully siMut siHuR-Sully siMut -- + + DHT Treatment B 120 100
80
60
40
20
Relative AR mRNA Relativelevel (%) AR 0 siHuR-UTR siCon siHuR-UTR siCon -- + + DHT Treatment
siMut + + - - siCon + - + - siHuR-Sully - - + + siHuR-UTR - + - + DHT - + - + DHT - - + + C Actin D Actin HuR HuR LNCaP 22Rv1
Figure 4.10. Effects of HuR knockdown by siRNA on steady-state AR mRNA levels in LNCaP and 22Rv1 cells.
A. LNCaP cells were treated with 20 nM siHuR-Sully or siMut as indicated. After 48 h of siRNA treatment, vehicle (-) or 10 nM DHT (+) was added as indicated. Following 24 h of DHT treatment, RNA was extracted and AR and actin mRNA levels were measured by qPCR, with the AR mRNA levels normalised using the actin mRNA levels. Relative AR mRNA level of siMut- treated cells is expressed as 100 %. These data represent mean ± sd of 3 replicates. B. The experiment described in A was repeated using 22Rv1 cells and 50 nM siHuR-UTR or siCon as indicated. Relative AR mRNA level of siCon-treated cells is expressed as 100 %. These data represent mean ± sd of 5 samples. Knockdown of HuR was confirmed by western blotting in LNCaP cells ( C) and in 22Rv1 cells ( D).
118 Nuclei HuR Merged
A Vehicle
B DHT DHT
C ActD
Figure 4.11. Cellular localisation of HuR in cells treated with DHT and actinomycin D. 22Rv1 cells were treated with for 4 h with (A) vehicle, (B) 10 nM DHT, or (C) 7.5 µg/µL actinomycin D. Cells were fixed and stained with HuR antibody (green) and Hoechst 33258 (red). Localisation of HuR was determined by confocal microscopy. Co-localisation of HuR and Hoechst 33258 dye is indicated by the yellow colour in the merged image. Magnification x 400.
119 observed following 30 and 90 min or 24 and 48 h of DHT treatment, suggesting that DHT does not affect sub-cellular distribution of HuR in 22Rv1 cells (data not shown). Similar results were observed with LNCaP cells (data not shown).
ActD has been reported previously to stimulate accumulation of HuR in the cytoplasm (Fan and Steitz, 1998b). Therefore, 22Rv1 cells were treated with ActD as a positive control for cytoplasmic accumulation of HuR. Following 4 h of ActD treatment, there was a substantial alteration of the distribution of HuR, with an increased proportion of the HuR immunostaining being localised to the cytoplasm (Figure 4.11C).
The intra-cellular localisation of HuR was also examined following treatment of 22Rv1 cells with siRNA (Figure 4.12). Neither siCon nor siHuR-UTR siRNA’s substantially altered the localisation of HuR, with HuR being predominantly (> 90 %) nuclear following 72 h of treatment with either siRNA (Figure 4.12 A, C). HuR was seen to accumulate in the cytoplasm of cells treated with both siRNA and ActD, with similar amounts of cytoplasmic HuR evident in cells treated with siHuR-UTR or siCon siRNA’s (Figure 4.12B, D). However, there appeared to be less HuR remaining in the nuclei in cells treated with both siHuR-UTR and ActD compared to cells treated with siCon and ActD (Figure 4.12B, D).
4.2.5 HuR Regulation of Cell Proliferation of 22Rv1 Cells.
Continued expression of AR is thought to maintain cell viability and proliferation of prostate cancer cells, and as HuR appears to have a role in regulating AR mRNA and protein levels, the effect of HuR knockdown on cell proliferation was investigated (Haag et al. , 2005; Zegarra-Moro et al. , 2002). 22Rv1 cells were treated with siHuR-UTR or siCon siRNA’s or with LipoFectamine 2000 alone. Cell proliferation was measured using a colorimetric assay daily from 24 h after transfection (Section 2.9). Somewhat unexpectedly, a significant increase in the rate of proliferation was observed in cells treated with siHuR-UTR siRNA compared to those treated with the non- targeted siCon siRNA or LipoFectamine alone, having doubling times of ~33 h, ~ 70 h and ~ 46 h, respectively (Figure 4.13). LNCaP cells treated with HuR
120 Nuclei HuR Merged
A siCon
B siCon, ActD
C siHuR-UTR
D siHuR-UTR, ActD siHuR-UTR,
Figure 4.12. siRNA and ActD effects on HuR intracellular localisation in 22Rv1 cells. 22Rv1 cells were treated with 50 nM siCon (A, B) , or siHuR-UTR (C, D) for 72 h, then with 7.5µg/µL for 4 h (B, D ). Cells were fixed and stained with HuR antibody (green) and Hoechst 33258 (red). Localisation of HuR was determined by confocal microscopy. Co-localisation of HuR and Hoechst 33258 dye is indicated by the yellow colour in the merged image. Magnification x 400.
121 A siHuR-UTR 7 Lipo 6 siCon * 5 *
4 *
3
Relative Absorbance (490 nm) (490 Relative Absorbance 2
1
0 0 24 48 72 96 120 Time (h) B Lipo: y=0.0753e 0.0099x , R 2=0.81 siHuR-UTR: y=0.0537e 0.0211x , R 2=0.90
siCon: y=0.0726e 0.015x , R 2=0.90
1 2 3 C Actin (45 kDa) HuR (36 kDa)
Figure 4.13. Effect of HuR knockdown on 22Rv1 cell proliferation.
A. 22Rv1 cells were treated with LipoFectamine 2000 alone ( ), siHuR-UTR siRNA ( ) or siCon siRNA ( ). Cell growth was measured daily for 5 days starting 24 h after treatment. Values are expressed relative to the absorbance at 24 h, which was set as 1. These data represent mean ± sd of 5 replicate samples. * = p < 0.05 (ANOVA). B. Formulae and R 2 values for growth rates of cultures in A. C. HuR and actin levels were measured by western blotting in siCon (lane 1), siHuR-UTR (lane 2) and LipoFectamine 2000 (lane 3) treated cells.
122 siRNA were not used for these experiments, because of poor signal-to-noise ratio in preliminary experiments (data not shown).
4.3 Discussion. The studies described in this chapter used both the LNCaP and 22Rv1 prostate cancer cell lines. LNCaP and 22Rv1 cells express the AR and are commonly used cell line models of prostate cancer, however, the LNCaP cell line was established in 1977, whereas the 22Rv1 cell line was established in 1999 and is less well-characterised (Horoszewicz et al. , 1983; Sramkoski et al. , 1999). 22Rv1 is a human prostate carcinoma cell line derived from a xenograft that was serially propagated in nude mice after castration-induced regression and relapse of the parental, androgen-dependent CWR22 xenograft (Cheng et al. , 1996; Sramkoski et al. , 1999). As described in Section 1.3.4, 22Rv1 cells express two forms of the AR, both of which have a duplication of exon three, resulting in a third zinc finger in the DBD. One expressed AR contains all functional domains, however the LBD contains an H874Y point mutation, while the other AR is C-terminally truncated and is detected by western blotting as a protein of ~ 75 kDa. This C-terminal truncation is thought to remove the LBD, and the resultant protein is distinct from the 87 kDa A- isoform of the AR, and unlike AR-A, which arises from initiation of translation from a low-efficiency internal methionine (M188), this novel AR reacts with an N-terminus-specific AR antibody (N20) (Chlenski et al. , 2001; Gao and McPhaul, 1998). The reason for the deletion of the LBD in the shorter AR protein in 22Rv1 is thought to be proteolysis of the AR or premature termination of translation, as only one AR transcript is detected by RT-PCR of whole cellular RNA from this cell line (Chlenski et al. , 2001). Despite these AR mutations, the androgen and bicalutamide responsiveness in luciferase assays, and the increased expression of AR protein following DHT treatment of the 22Rv1 cells was consistent with that detected in the LNCaP cells, which also express an AR containing a point mutation (Horoszewicz et al. , 1983). The confirmation that DHT regulation of AR protein was similar in 22Rv1 and LNCaP cells indicated that 22Rv1 cells were suitable as a model for studying HuR/AR interactions.
123 An additional protein of ~ 70 kDa was detected by AR western blotting of 22Rv1 cell extracts in this thesis. An AR protein of this molecular size has not been reported in previous studies of 22Rv1 cells which have used the N20 AR antibody (Chlenski et al. , 2001; van Bokhoven et al. , 2003). The present study used the 441 AR antibody, which was raised against an epitope between amino-acids 299 and 315 of the full-length AR, and therefore would detect N- and C-terminally truncated AR’s. As the ~ 70 kDa AR was not reported in previous studies (Chlenski et al. , 2001; van Bokhoven et al. , 2003), it may represent an N-terminally truncated AR that potentially contains other sequence abnormalities. Alternatively, the aberrantly sized AR (or other cross- reacting protein) may be unique to the 22Rv1 isolate obtained for use in the present study. Repeating the western blotting experiments with the N20 and C-terminal-specific C19 antibody may assist in the identification of this additional AR protein band.
When HuR expression was increased in LNCaP cells by transient over- expression of FLAG-HuR, no increases in AR protein levels were detected on AR western blots. This may have been because HuR protein expression was already high in this cell line, and therefore HuR target sites may already have been saturated, such that the small increase in HuR protein evident following transient transfection of the LNCaP cells with the FLAG-HuR vector may not have substantially altered HuR binding and/or AR mRNA or protein kinetics. In addition, the western blotting technique measured the level of AR (and HuR) proteins within the entire LNCaP cell culture, rather than the AR and HuR protein levels in cells that had been successfully transfected. As such, the methodology used may have masked the effects of increased HuR expression in individual cells. To determine the effects of increased expression of HuR on AR protein levels, it may be necessary to select cells that were transfected. This could be achieved using a green fluorescent protein (GFP)-tagged HuR expression vector in combination with fluorescence-activated cell sorting to isolate the population of HuR-expressing cells. In addition, homeostatic mechanisms may exist that limit the effects of moderate HuR over-expression on AR protein levels.
124 Cell density was found to be important for the effectiveness of siRNA-mediated knockdown of HuR. The knockdown of HuR was markedly greater when cells were plated at a density of 2.5 x 10 4 cells/cm 2 in comparison to cells plated at 4.5 x 10 4 cells/cm 2, which was consistent with the manufacturer’s recommendation of a cell confluence of 30 – 50 %. In these studies, siRNA- mediated knockdown of HuR protein resulted in a concurrent decrease in AR protein level. The control siRNA, siMut, also produced small decreases in HuR, actin and AR protein levels, which may have been due to the non- specific effects of siRNA transfection on cell physiology, or to off-target effects due to limited complementarity between the siRNA and multiple cellular mRNA’s. The marked reductions in AR protein levels were contingent on a large knockdown of HuR protein, perhaps due to homeostatic mechanisms maintaining AR protein levels, as protein turnover as well as mRNA turnover is known to contribute to AR protein levels (Krongrad et al. , 1991; Wolf et al. , 1993; Yeap et al. , 1999).
Transfection of androgen-responsive and control luciferase vectors following transfection of siRNA’s (siHuR-Sully and siMut) resulted in a decrease in the induced luciferase activity from those vectors in LNCaP cells. This effect could not be titrated over the concentrations of siRNA tested, and so may be due to non-specific effects of siRNA transfection. Moreover, when induced luciferase activity from the PSA-Luc vector was normalised to that from the pRL-TK vector, siMut siRNA-treatment of cells appeared to increase luciferase activity in comparison to both the siHuR-Sully and LipoFectamine only treated cells. This was not due to increased AR-mediated transcription, but appeared to be due to the greater inhibition of induced Renilla luciferase activity from the pRL- TK vector. Although luciferase activity is supposed to be constitutively induced from this vector via the thymidine kinase promoter, culture conditions, including the addition of DHT, have been reported previously to regulate induced luciferase activity from the vector, limiting its use (Mulholland et al. , 2004). When experiments were repeated using a different cell line (22Rv1), different siRNA’s (siHuR-UTR and siCon) and a different androgen-responsive reporter (PB3-Luc), in the absence of a Renilla expression plasmid, no effects of HuR knockdown on basal or androgen-induced luciferase activity in comparison to
125 control siRNA-treated cultures were observed. Moreover, measurement of secreted PSA protein in the cell culture medium following siRNA and DHT treatment revealed no significant difference between siHuR-UTR and siCon- treated cells.
Several factors may have affected the luciferase activity induced from the PSA-Luc, PB3-Luc and Renilla vectors. In these experiments, the cells were transfected twice, 48 h apart, and this may have exacerbated negative physiological effects of siRNA treatment. The use of stably-transfected short- hairpin vectors to express RNA duplexes targeting HuR mRNA may circumvent this problem, as the cells could be given a longer period to recover between transfections. However, the LipoFectamine 2000 transfected cells were also mock-transfected twice, and exhibited high luciferase activity, so it is likely that the introduced siRNA duplexes interfered with induction of luciferase activity, or with stability of the luciferase mRNA or protein. Measurement of the luciferase mRNA by northern or qPCR would address whether alterations in luciferase activity were due to changes in the luciferase mRNA, although there are technical difficulties associated with these methods. Alternatively, recruitment of the AR protein to an endogenous androgen-responsive promoter following transfection of an siRNA against HuR could be estimated by chromatin-immunoprecipitation (ChIP) assays, to determine whether the reduction of HuR protein levels affected this aspect of AR-mediated transcription.
Co-transfection of androgen-responsive (PSA-Luc) and control ( Renilla ) luciferase vectors with FLAG-HuR resulted in decreased induction of luciferase activity from those vectors. This effect was very similar to that observed when HuD was transiently over-expressed in LNCaP cells. Interestingly, whereas transient over-expression of HuD resulted in increased DHT-mediated induction of luciferase activity from the PSA-Luc vector, transfection of the cells to over-express HuR resulted in decreased DHT-mediated induction of luciferase activity from the PSA-Luc vector in comparison to FLAG-empty transfected cells. DHT effects were not significant, however, and the technical limitations of these assays, including their sensitivity for measurement of subtle
126 changes in AR levels and thus DHT-induced transcriptional activity restricted any conclusions regarding HuR effects on AR activity in prostate cancer cells.
Reduction of HuR protein levels in 22Rv1 cells using siRNA-mediated knockdown resulted in a significant decrease in the half-life of AR mRNA, compared to the half-life of AR mRNA in cells in which HuR had not been knocked down. Surprisingly, the estimated half-life of the AR mRNA in 22Rv1 in cultures transfected with the siCon siRNA was > 40 h, which is markedly longer than the half-life of AR mRNA of 5.5 h reported for LNCaP (Yeap et al. , 1999). The AR mRNA half-life was measured in cells that had been cultured in media supplemented with charcoal-stripped FCS, and therefore had been exposed to extremely low levels of steroid hormone. Under these conditions, the AR mRNA is thought to be less stable than in LNCaP cultures treated with DHT, however, the effect of DHT treatment on AR mRNA half-life in 22Rv1 cells has not been reported (Yeap et al. , 1999). AR mRNA half-life is expected to differ between cell lines and is reported to be 5.5 h in LNCaP cells and 3 h in MDA453 cells in the absence of DHT, however, both of these half-lives are much shorter than that reported herein (Yeap et al. , 1999). This discrepancy may also have been due to the different methods by which the mRNA half- lives were assessed. Previous studies have used northern blotting analysis to measure mRNA decay rather than qPCR. In northern blotting analysis of LNCaP AR mRNA, the density of the intact AR transcript of 10.5 kb was measured (Yeap et al. , 1999). In the present study, full-length and partially degraded transcripts could not be distinguished, because the amplicon was a region of exon 5 of the AR mRNA, approximately 7 kb from the 3’ end of the transcript. Therefore, the use of this amplicon in qPCR may have caused over-estimation of AR mRNA half-life, by measuring the combined levels of intact and partially degraded transcripts.
It is possible that the concentration of ActD used to determine the decay rate of AR mRNA (7.5 µg/ µL) was not sufficient to fully inhibit transcription, and therefore production of new transcripts may have also contributed to the apparent length of AR mRNA half-life. In addition, ActD has been previously observed to artificially increase the half-life of some transcripts, as well as to
127 substantially affect cell physiology, although mRNA’s containing a class III AURE, such as the ARUC, are thought not to be stabilised by ActD treatment (Loflin et al. , 1999b; Peng et al. , 1996; Zhao et al. , 2000b). Moreover, mRNA decay is a two-step kinetic process, involving de-adenylation and then decay of the mRNA body, and an approach using ActD does not attempt to measure de-adenylation rates (Section 1.4.1) (Loflin et al. , 1999b). However, these considerations apply equally to the AR mRNA from cells treated with siHuR- UTR siRNA and to the AR mRNA from cells treated with siCon siRNA, therefore the estimated difference in AR mRNA half-life between these samples is comparable. Future experiments could use an amplicon nearer the 3’ end of the AR mRNA for better specificity for intact transcripts, and the amount of ActD used could be titrated to ensure that transcription was fully inhibited.
It was interesting that a knockdown in HuR protein of ~ 40 % was sufficient to decrease the stability of the AR mRNA, while a decrease in AR protein was observed only following a knockdown of HuR protein of ~ 70 %. This may be due to mechanisms such as stabilisation of the AR protein as mRNA levels were decreased, insulating the AR protein from smaller fluctuations in the levels of AR mRNA. Differences in the sensitivity of measurement of AR mRNA turnover in comparison to steady-state AR mRNA levels may also have influenced results as changes in AR mRNA were more readily detected ( i.e. following ~ 40 % decreases in HuR protein) in the presence of ActD. The kinetics of AR mRNA and protein following HuR knockdown were also not evaluated, therefore while AR protein levels were decreased after 72 h of HuR siRNA treatment, AR mRNA levels were likely to have been decreased at earlier time-points. This could be assessed in future experiments with AR mRNA levels measured at 24 and 48 h and at later time-points to determine the nadir AR mRNA levels and times following siRNA treatments, as well as recovery time. Because the knockdown of HuR protein appeared to persist for up to 120 h following treatment of the cultures with HuR siRNA, the recovery of AR mRNA was likely to have occurred via de novo transcription of the AR gene, because the assumed stabilising effect of HuR would still be reduced
128 after 72 h of siRNA treatment, therefore the stability of the AR mRNA was likely to have also remained reduced.
Ligand-induced stabilisation of AR mRNA and protein has been reported previously and it is feasible that the mechanism accounting for this involves mRNA binding proteins such as HuR. Although DHT effects on steady-state mRNA levels could not be determined in the present studies, it will be important in future experiments to determine whether HuR is involved in stabilising the AR mRNA following exposure to androgens. Initially, DHT- induced increases in AR mRNA would need to be confirmed in 22Rv1 cells using either qPCR or northern blotting (to more accurately quantify levels of full-length mRNA). Steady-state AR mRNA levels and AR mRNA half-life (using ActD chase) following HuR knockdown in the absence and presence of DHT could then be calculated.
HuR did not accumulate in the cytoplasm of 22Rv1 cells treated with DHT, and was > 90 % nuclear in both DHT- and vehicle-treated cells. The predominant localisation of HuR in the nucleus in resting cells is consistent with previous reports of the localisation of HuR in other cell types (Atasoy et al. , 1998; Fan and Steitz, 1998b). Addition of ActD to the culture medium resulted in accumulation of the HuR in the cytoplasm, which is also consistent with previous reports, and indicated that HuR is able to shuttle in 22Rv1 cells (Atasoy et al. , 1998; Fan and Steitz, 1998b). Previous reports of HuR being enriched in polyribosomes of HepG2 cells following DHT treatment indicated that the intra-cellular localisation of HuR might also be regulated by DHT in prostate cancer cells, however those studies used sub-cellular fractionation rather than microscopy to determine the location of HuR (Sheflin et al. , 2001). As such, the approach taken in this study may not have had the resolution to detect any alteration in the localisation of HuR following DHT treatment. Future experiments could use sub-cellular fractionation of 22Rv1 or LNCaP cells to determine if HuR in these cell lines responds similarly following DHT treatment to HuR in HepG2 cells (Sheflin et al. , 2001). In HepG2 cells, DHT treatment for 48 h was found to decrease total cellular HuR levels by > 75 % (Sheflin et al. , 2001), an effect not seen in this study with the use of confocal
129 microscopy or western blotting of 22Rv1 or LNCaP cells (data not shown). Although the cytoplasmic localisation of HuR has been associated with stabilisation of target mRNA’s (Aranda-Abreu et al. , 1999; Chen et al. , 2002; Kasashima et al. , 1999; Okano and Darnell, 1997) studies have not been carried out previously in prostate cancer cells and RNA processing in the nucleus and stabilisation in the cytoplasm are likely to be coupled events.
Treatment of 22Rv1 cells with siRNA against HuR did not substantially alter the intracellular localisation of HuR compared to treatment with control siRNA, although there was less HuR immunostaining in siHuR-UTR siRNA-treated cells. Combined treatment of 22Rv1 cells with siRNA and ActD resulted in accumulation of similar amounts of HuR in the cytoplasm of siHuR-UTR- and siCon-treated cells. However, cells treated with siHuR-UTR siRNA exhibited less nuclear HuR immunostaining following ActD treatment than did cells treated with siCon and ActD. This suggests that the levels of cytoplasmic HuR were maintained at the expense of nuclear HuR in cells with reduced levels of HuR ( i.e. siHuR-UTR-treated cells). This is consistent with the proposition that the cytoplasmic localisation of HuR is associated with stabilisation of target mRNA’s and that HuR accumulates in the cytoplasm following transcriptional inhibition in order to stabilise target mRNA’s. These observations are also important because ActD was used to inhibit transcription in order to estimate the effect of reduction of HuR protein on AR mRNA half-life, and the ActD treatment was likely to have affected the intracellular distribution of HuR in those experiments. Because the confocal microscopy data were not quantitative, it was not possible to directly compare changes in cytoplasmic and nuclear HuR levels in siHuR-UTR-treated cells compared to siCon siRNA- treated cells, which may require sub-cellular fractionation in future studies. Experiments incorporating treatment of the cells with leptomycin B to inhibit nuclear export of HuR prior to treatment with ActD may also clarify the role of HuR in AR mRNA stabilisation. Leptomycin B has been previously reported to inhibit HuR export via inhibition of the nuclear export factor CRM1, thereby inhibiting export of HuR target mRNA’s, such as c-fos (Brennan et al. , 2000). Because a proportion of AR mRNA and HuR protein is likely to already exist in the cytoplasm, treatment with leptomycin B may allow effects of reduced HuR
130 protein on AR mRNA stability to be assessed without the complicating factor of re-distributed HuR protein induced by ActD treatment.
The stimulatory effect of siHuR-UTR siRNA compared to siCon siRNA on proliferation of 22Rv1 cells was unexpected. The decreased proliferation rate of siCon-treated cells compared to cells treated with LipoFectamine 2000 alone indicated that transfection of cells with this siRNA was inhibitory, which may have been specific to the siCon siRNA, or may reflect a general inhibitory effect of the presence of siRNA’s within the cell. However, the cells transfected with siRNA directed against HuR (siHuR-UTR) proliferated more rapidly than cells that had been treated with LipoFectamine 2000 alone. This suggested that reduction of HuR protein levels was growth-stimulatory in 22Rv1 cells, a finding that contrasts with previous reports, which indicate that HuR is a marker of more aggressive tumour behaviour (de Silanes et al. , 2003; Erkinheimo et al. , 2003; Heinonen et al. , 2005; Nabors et al. , 2001). Although previous reports have utilised different measurements to determine tumour growth, including immunohistochemical evaluation of human cancer specimens (de Silanes et al. , 2003; Erkinheimo et al. , 2003; Heinonen et al. , 2005; Nabors et al. , 2001) and growth of xenografts in nude mice (de Silanes et al. , 2003), the results of the present study were unusual considering the effects of HuR on the AR and importance of the AR in prostate cancer growth. These findings indicated that non-AR mRNA’s bound by HuR may be critical in controlling prostate cancer cell proliferation and presented an exciting possibility that these factors may be able to be targeted in prostate cancer treatment. Profiling of mRNA’s regulated following HuR knockdown could be used to isolate HuR-regulated pathways that potentially account for HuR effects on cell proliferation and other aspects of prostate tumour progression.
131
CHAPTER 5: Micro-Array Analysis of 22Rv1 Cells with
Reduced HuR.
132 5 Chapter 5: Micro-Array Analysis of 22Rv1
Cells with Reduced HuR.
5.1 Introduction. HuR is expressed both in normal prostate and in prostate cancer cells, with no marked differences in HuR expression or localisation observed in 12 prostate tumours compared to adjacent benign tissues (Section 3.2.1). A number of HuR targets have been identified in several cell types however the biological activity of HuR has not been investigated in prostate tumours (Blaxall et al. , 2000; de Silanes et al. , 2003; Erkinheimo et al. , 2003; Heinonen et al. , 2005; Nabors et al. , 2001). One of the reported targets of HuR is the AR mRNA, and results described in this thesis indicate that HuR stabilises the AR transcript in prostate cancer cells (Yeap et al. , 2002) (Chapter 4). The central role of the AR in the control of prostate cancer cell proliferation and its widespread use as a therapeutic target in prostate cancer management indicate that HuR levels and activity may play a significant role in prostate tumour growth. HuR effects in prostate cancer cells would be expected to be mediated by the AR as well as by other HuR targets, including COX2, cytokines and angiogenic factors (Denkert et al. , 2004; Nabors et al. , 2001) (Table 1.2). As high levels of HuR expression are detected in prostate cancer cells, the identification of HuR targets was carried out in this thesis by comparison of gene expression in 22Rv1 cells treated with siRNA directed against HuR and 22Rv1 cells treated with control siRNA. To facilitate the identification of large numbers of genes, this study utilised oligonucleotide micro-arrays.
High-density oligonucleotide micro-arrays have rapidly developed since they were first reported in 1995, and modern oligonucleotide micro-array analysis allows the simultaneous measurement of mRNA expression from a large number of genes (Schena et al. , 1995; Stoughton, 2005). Oligonucleotide micro-array technology involves a high-density array of DNA oligonucleotides (probes) on a solid support, which is hybridised with labelled nucleic acid fragments derived from mRNA (Stoughton, 2005). The specific hybridisation
133 between an mRNA-derived fragment and a probe with defined sequence can be recorded as an analogue signal approximating the transcript abundance for a particular gene (Stoughton, 2005).
The GeneChip Human Genome (HG) U133 plus 2.0 array used in this study comprises 47 000 transcripts corresponding to 38 500 well-defined genes and an additional 8 500 uncharacterised transcripts (www.affymetrix.com/products /arrays/specific/hgu133plus.affx). Technical information regarding the GeneChip micro-arrays is detailed in the “Expression Analysis Technical Manual with Specific Protocols for the Use with Hybridisation, Wash and Stain Kit” and information specific to the design of the Human Genome U133 plus 2.0 array is provided in the Affymetrix technical note “Design and Performance of the Human Genome U133 Plus 2.0 and Human Genome U133A Arrays” (www.affymetrix.com/index.affx). Briefly, HG U133 Plus 2.0 probe sequences are based on publicly-available data from the expressed-sequence tag database dbEST, UniGene, RefSeq and GenBank, and multiple probes have been designed for each gene, generating a specific probe set. Probe sets consisted of 11 pairs of perfectly-matched and mismatched probes, 25 nucleotides in length, arrayed on a glass chip at a density of 1 x 10 5-10 6 copies per probe in an 11 x 11 µm probe “cell” (Figure 5.1). Mismatched probes contained single-base mismatches with the target gene sequence, to control for non-specific annealing.
Unlike conventional spotted oligonucleotide micro-arrays which use fluorescently-labelled oligonucleotide fragments, Affymetrix arrays use complementary (c)RNA “targets” derived from the initial input RNA by a series of amplification steps that involve generation of double-stranded oligonucleotides followed by reverse transcription in the presence of modified ribonucleotides to create single-stranded biotinylated cRNA fragments of ~ 1 500 nucleotides in length. The cRNA fragments are hydrolysed into targets of ~ 100 nucleotides and hybridised to the chip. The chip is then washed and stained with a streptavidin-phycoerythrin conjugate, and scanned with a specialised scanner (GeneChip Scanner 3000). The light emitted at 570 nm is proportional to the amount of target bound at each probe cell, and
134 Figure 5.1. Sample image of a scanned oligonucleotide micro-array chip. Each 11 x 11 um probe cell ( ) contains 10 5-10 6 copies of an oligonucleotide probe. Annealing of cRNA sample fragments and subsequent fluorescent staining generates varying intensities that are proportional to the amount of annealed sample cRNA, and therefore are proportional to the expression level of the transcript from which that cRNA was derived. Several (usually 11) pairs of perfect-match and mismatched probes are distributed over the chip control for non-specific annealing of cDNA fragments. Reproduced from Affymetrix technical note: Design and Performance of the Human Genome U133 Plus 2.0 and Human Genome U133A Arrays (2003).
135 therefore proportional to the level of expression of that target gene. The GeneChip operating system (GCOS) averages the intensity data from across the 11 probe pairs, applies a number of statistical algorithms to identify absent and false-positive probes, then assigns a signal value to each probe set, corresponding to an expression value for the target gene.
Although oligonucleotide micro-arrays are the most rapid method for analysis of transcriptional or post-transcriptional changes in gene expression across an entire genome, the quality of the micro-array data generated is highly dependent on both the input RNA quality and on the correct handling of the nucleic acids derived from the input RNA. In addition, some HG U133 Plus 2.0 probe sets are not entirely specific for the intended target gene. Certain probe sets contain sequences common to more than one gene (s-suffixed sets), or contain some probes that detect transcripts other than the intended transcript, but collectively are specific (x-suffixed probe sets), and some probe sets cannot differentiate between alternatively spliced isoforms of single genes (a-suffixed probe sets). It is for these reasons that another method, such as qPCR, must be employed to confirm changes in mRNA expression.
To assess the effects of variability of input RNA quality and handling, quality controls are included at two points during the sample processing and hybridisation (Figure 5.2). The experimental RNA sample is “doped” with control polyA RNA’s encoding the Bacillus subtilis proteins lys, phe, thr and dap to monitor the entire labelling process, and in the final data set the signal values for the probe sets corresponding to these control genes should all be detected as “present”, with increasing signal values in the order lys < phe < thr < dap . Synthetic cRNA’s are added to the experimental sample prior to hybridisation to assess the range and level of assay sensitivity. The hybridisation controls correspond to the bioB, bioC, bioD and cre genes of E. coli , and are also in staggered concentrations, such that bioB should only be detected on 50 % of scans ( i.e. is on the detection threshold) and the signal values for the probe sets should have increasing intensities in the order bioB < bioC < bioD < cre . Internal controls for quality assessment also include the signal values for the β-actin and GAPDH probe sets, which are derived from
136 *
*
Figure 5.2. Overview of cRNA generation, labelling and hybridisation. The total cellular RNA sample is doped with artificial polyA RNA to monitor biotin labelling and subsequent handling ( *) before first-strand synthesis of cDNA. Double-stranded cDNA is generated and column-purified, then transcribed into biotinylated cRNA and re-purified. cRNA hybridisation controls (*) are added to the sample to monitor hybridisation, washing/staining and scanning, then the cocktail of sample and control cRNA’s are hybridised to the micro-array chip, washed and stained with a fluorescent antibody dye. The intensity of staining at each probe cell is scanned and signal values assigned to each gene for which the chip contains a probe set. Adapted from Expression Analysis Technical Manual with Specific Protocols for Use with Hybridisation, Wash and Stain Kit (www.affymetrix.com)
137 the experimental sample. The signal intensity ratio of the 3’ probe pairs to the 5’ probe pairs for each transcript is generally no more than ~ 3:1. This ratio is caused by the 3’ bias introduced by the use of oligo(dT) in the generation of the cDNA. Lower 3’:5’ ratios may indicate poor sample quality. The final quality assessment is the percentage of genes present, which is highly dependent on the cell/tissue type and the experimental treatment, but should be similar between replicate samples.
The large data sets generated by oligonucleotide micro-array analysis can be manipulated by a number of approaches with varying levels of sophistication, from pairwise comparison of data derived from two experimental conditions, to multivariate analysis of multiple treatments or disease states (for example) over time. The present micro-array was a pairwise comparison of gene expression between two treatments of 22Rv1 cells, siRNA mediated knockdown of HuR vs. nonsense siRNA treatment, with sample collection at a single time-point, 72 h after siRNA treatment. Therefore, the following studies aimed to identify genes that exhibited increased or decreased expression subsequent to treatment of 22Rv1 cells with siHuR-UTR compared to those treated with siCon, using the web-based micro-array analysis software GeneSifter (Section 2.15.1).
5.2 Results.
5.2.1 Identification of Genes with Altered Expression Following Knockdown of HuR.
To identify potential mRNA targets of HuR, 22Rv1 cells were treated with siRNA against HuR or with control siRNA (Section 2.8.2). After 72 h, total RNA was extracted and its integrity ascertained by Bioanalyser (Sections 2.14.2, 2.15.1) (Figure 5.3). High-quality mRNA was analysed by GeneChip HG U133 Plus 2.0 oligonucleotide micro-array, and data that passed quality assessment were analysed using GeneSifter to isolate genes that exhibited significant alterations in expression levels of ≥ 1.2-fold (p < 0.05; Section 2.15.1). This analysis identified 582 regulated genes, of which 84 were increased and 498 decreased in expression (Appendix 3). Consistent with 138 A 18S 28S B
C D
Marker
E F
Figure 5.3. Analysis of mRNA used for oligonucleotide micro-array. A. mRNA purified from duplicate samples of siHuR-UTR-treated 22Rv1 cells was analysed by Bioanalyser to determine the ratio of 18S to 28S ribosomal RNA as a measure of RNA integrity. The absence of small degradation products also indicated high quality mRNA. Ribosomal RNA peaks are indicated. Fluorescence units (FU) are on the y axis, and time (s) is on the x axis. B. The mRNA was purified from siCon-treated 22Rv1 cells, and constitutes the control experiment for A. C. Replicate experiment; a prominent marker peak is indicated; otherwise, as for A. D. Replicate experiment; as for B. E. mRNA of poor quality. F. RNA ladder.
139 previous indications, no change in AR mRNA expression was detected following treatment of the cells with siHuR-UTR siRNA (Section 4.2.3). As expected, HuR mRNA expression was significantly decreased 2-fold (p < 0.05). The largest alteration in gene expression was observed for an uncharacterised open reading frame on chromosome 21, which exhibited a 2.95-fold decrease in siHuR-UTR siRNA-treated cells. The largest decrease in a characterised gene was detected in the glutaminase transcript, which was down-regulated 2.83-fold. Relatively fewer genes were found to have significantly increased expression, with the greatest up-regulation of 1.93-fold being observed in the transcript encoding claudin domain-containing protein 1, a membrane protein of unknown function.
5.2.2 Genes Regulated by HuR Cluster in Growth-Associated Pathways.
The genes that had been identified as being significantly up- or down- regulated following treatment of the cells with siHuR-UTR compared to those treated with siCon siRNA were clustered into Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathways to identify cellular processes that were affected by knockdown of HuR (Section 2.15.1). Significant clustering (z-score of > 2) of down-regulated genes occurred in pathways involved in the regulation of the actin cytoskeleton, focal adhesion, VEGF, MAPK and Wnt signalling, axon guidance, adherens junctions, ubiquitin-mediated proteolysis, glutamate metabolism, nitrogen metabolism and response to pathogenic E. coli infection (Table 5.1). Up-regulated genes, of which there were fewer, clustered in glycerophospholipid metabolism and glycine, serine and threonine metabolism (Table 5.1). Steroid signalling pathways are not included in the KEGG database, therefore it could not be determined whether significant clustering occurred in AR-mediated signalling pathways. However, using GO Biological Process definitions to search for AR-regulated or AR-associated genes among the transcripts with altered expression following knockdown of HuR retrieved three genes, PIAS2, CREBBP and PCAF.
Because of the apparent increase in growth rate of 22Rv1 cells treated with siRNA directed against HuR, genes that were associated with apoptosis, cell
140 Table 5.1. Genes potentially regulated by HuR . Genes that were up- and down-regulated ≥1.2-fold after HuR knockdown were grouped according to KEGG pathway. Z scores >2 indicate significant clustering. Up-regulated genes are in purple . Genes may appear in more than one pathway. X : value derived from x-suffixed probe set. S: value derived from s-suffixed probe set. Z Score: Pathway and Gene Name Fold Change: UP DOWN Focal adhesion -1.13 5.01 Baculoviral IAP repeat-containing 2 1.32 BCL2-antagonist of cell death 1.41 Dedicator of cytokinesis 1 1.53 Diaphanous homolog 1 ( Drosophila ) 1.85 ELK1, member of ETS oncogene family X 1.63 Filamin A, α (actin binding protein 280) x 1.23 Filamin B, β (actin binding protein 278) S 1.32 Integrin, β1 (fibronectin receptor, β) 1.88 P21 (CDKN1A)-activated kinase 2 S 1.80 PTK2 protein tyrosine kinase 2 1.45 Receptor activator 1, RAF1 1.66 Related RAS viral (r-ras) oncogene homolog 2 1.21 Thrombospondin 1 1.35 Vav 3 oncogene 1.33 MAPK signaling pathway 0.33 2.47 Dual specificity phosphatase 1 S 1.39 ELK1, member of ETS oncogene family X 1.63 Fibroblast growth factor receptor 1 1.89 Filamin A, α (actin binding protein 280) X 1.23 Filamin B, β (actin binding protein 278) S 1.32 Mitogen-activated protein kinase kinase kinase 2 1.35 Mitogen-activated protein kinase-activated protein kinase 3 S 1.56 Neurofibromin 1 1.25 P21 (CDKN1A)-activated kinase 2 S 1.80 Phospholipase A2, group VI 1.44 Protein phosphatase 1A 1.55 Receptor activation factor 1 (RAF1) 1.66 Related RAS viral (r-ras) oncogene homolog 2 1.21 Regulation of actin cytoskeleton -1.13 3.99 Dedicator of cytokinesis 1 1.53 Diaphanous homolog 1 ( Drosophila ) 1.85 Diaphanous homolog 2 ( Drosophila ) 1.44 Fibroblast growth factor receptor 1 1.89 Integrin, β1 (fibronectin receptor, β) 1.88 P21 (CDKN1A)-activated kinase 2 S 1.80 Protein phosphatase 1, regulatory (inhibitor) subunit 12B 1.41 PTK2 protein tyrosine kinase 2 1.45 Receptor activation factor 1 (RAF1) 1.66 Related RAS viral (r-ras) oncogene homolog 2 1.21 Vav 3 oncogene 1.33 Wiskott-Aldrich syndrome-like 1.75 Continued next page…
141 Wnt signaling pathway 0.18 3.12 CREB binding protein 1.59 Cullin 1 1.29 CXXC finger 4 1.63 Dickkopf homolog 1 ( Xenopus laevis ) 1.42 Dishevelled associated activator of morphogenesis 1 1.72 Frizzled homolog 1 ( Drosophila ) 1.26 Frizzled homolog 8 ( Drosophila ) 1.59 Protein phosphatase 2A (catalytic subunit) 2.02 Protein phosphatase 2A (reg. subunit) 2.32 Seven in absentia homolog 1 ( Drosophila )S 1.31 Axon guidance -0.9 2.12 Integrin, β1 (fibronectin receptor, β) 1.88 Neurofibromin 1 1.25 P21 (CDKN1A)-activated kinase 2 S 1.80 PTK2 protein tyrosine kinase 2 1.45 Related RAS viral (r-ras) oncogene homolog 2 1.21 SLIT-ROBO Rho GTPase activating protein 1 1.69 Glycerophospholipid metabolism 2.33 2 1-acylglycerol-3-phosphate O-acyltransferase 3 1.32 Alkylglycerone phosphate synthase 1.37 Hydroxysteroid (17-β) dehydrogenase 12 1.39 Lysophospholipase 3 (lysosomal phospholipase A2) 1.38 Phospholipase A2, group VI (cytosolic, calcium-independent ) 1.44 Retinol dehydrogenase 11 (all-trans and 9-cis) S 1.44 Ubiquitin mediated proteolysis 1.44 3.29 Anaphase promoting complex subunit 10 1.51 Cullin 1 1.29 Hypothetical protein LOC222070 1.25 Neural precursor cell expressed, developmentally down-regulated 4 2.19 WW domain containing E3 ubiquitin protein ligase 1 S 1.54 VEGF signaling pathway -0.65 3.04 BCL2-antagonist of cell death 1.41 Mitogen-activated protein kinase-activated protein kinase 3 S 1.56 Phospholipase A2, group VI (cytosolic, calcium-independent) 1.44 PTK2 protein tyrosine kinase 2 1.45 Receptor activation factor (RAF1) 1.66 Adherens junction -0.68 2.03 CREB binding protein 1.59 Fibroblast growth factor receptor 1 1.98 Synovial sarcoma, X breakpoint 2 interacting protein 1.5 Wiskott-Aldrich syndrome-like 1.75 Glycine, serine and threonine metabolism 5.36 0.1 Dihydrolipoamide dehydrogenase 1.22 Hydroxysteroid (17-β) dehydrogenase 12 1.39 Phosphoserine phosphatase X 1.76 Retinol dehydrogenase 11 (all-trans and 9-cis) S 1.44 Glutamate metabolism -0.43 3.04 Glutamine synthetase 1.30 Glutaminase 2.83 Guanine monophosphate synthetase 1.66 Nitrogen metabolism -0.38 3.66 Carbonic anhydrase VIII 1.83 Glutamine synthetase 1.30 Glutaminase 2.83
142 cycle, or proliferation by GO Biological Process definitions were also identified (Section 4.2.5). Twelve apoptosis-associated genes were identified as having altered expression following siHuR-UTR siRNA treatment. Of these, three were up-regulated and nine were down-regulated (Table 5.2). This set of genes included up-regulated pro-apoptotic genes ( e.g. BCL2-associated athanogene 1; BAG1), down-regulated anti-apoptotic genes ( e.g. baculoviral IAP repeat-containing 2; BIRC2), and down-regulated pro-apoptotic genes (e.g. BCL2-agonist of cell death; BAD), as well as a number of genes with less well-defined roles in apoptosis.
Of the regulated genes associated with apoptosis, only BIRC2, BFAR and BAD appear to be directly involved in the regulation of apoptosis. BIRC2 and BFAR are thought to interact directly with and inhibit pro-apoptotic caspases and proteases, whereas BAD inhibits the formation of heterodimers between BCLXL and BCL2, thereby reversing their death repressor activity (Chattopadhyay et al. , 2001; Rothe et al. , 1995; Shu et al. , 1996; Stegh et al. , 2002; Uren et al. , 1996; Zhang et al. , 2000). Seven in absentia homologue 1 (SIAH1) regulates apoptosis in a less direct manner than BIRC2, BFAR and BAD. SIAH1 is one of a family of ubiquitin ligases that are involved in proteasome-mediated degradation of proteins regulating diverse processes including the cellular response to hypoxia and the induction of apoptosis. siRNA-mediated knockdown of SIAH proteins partially protects cells from apoptosis induced by trophic factor deprivation and DNA damage, via decreased activation of the JNK pathway (Xu et al. , 2006).
Both receptor activation factor 1 (RAF1, c-RAF) and zinc finger DHHC type- containing 16 (ZDHHC16, Ah2) are involved in growth factor signalling, however RAF1 was down-regulated, while ZDHHC16 was up-regulated. RAF1 is a protein kinase that activates MEK1 and MEK2, and is therefore implicated in the control of gene expression involved in the cell cycle, apoptosis, cell differentiation and cell migration, although RAF1 is not itself an effector of apoptosis (Garnett et al. , 2005). ZDHHC16, which was up-regulated following knockdown of HuR, acts downstream of c-Abl, a non-receptor tyrosine kinase implicated in DNA damage-induced cell death and in growth factor receptor
143
Table 5.2. Apoptosis associated genes . Up-regulated genes are in purple . X : value derived from x-suffixed probe set. S: value derived from s-suffixed probe set.
Apoptosis: Gene ID: Accession Fold GO Biological Process: no.: change:
Baculoviral IAP repeat- Anti-apoptosis; surface-linked signal BIRC2 NM_001166 1.32 transduction; positive regulation of I- containing 2 kappaB kinase/NF -kappaB cascade. BCL2-antagonist of cell BAD U66879 1.41 Induction of apoptosis death
BCL2-associated Apoptosis; anti-apoptosis; protein BAG1 AA394039 1.23 folding; surface-linked signal athanogene 1 transduction. Bifunctional apoptosis BFAR NM_016561 1.71 Apoptosis; anti-apoptosis regulator
D-site of albumin promoter- Anti-apoptosis; cell proliferation; S DBP U79283 1.27 transcription; protein kinase C binding protein activation. Apoptosis; phagocytosis; integrin- Dedicator of cytokinesis 1 DOCK1 NM_001380 1.53 mediated signalling
Engulfment and cell motility ELMO2 NM_022086 2.42 Apoptosis; phagocytosis. 2
Apoptosis; ubiquitin cycle; protein Fused-toes homologue FTS NM_022476 1.39 modification.
Apoptosis; cell proliferation; signal Receptor activation factor 1 RAF1 BF823525 1.66 transduction.
Seven in absentia homolog Apoptosis; ubiquitin cycle; cell cycle; S SIAH1 AI953523 1.31 1 (Drosophila) axon guidance
Apoptosis; electron transport; signal Thioredoxin-like 1 TXNL1 NM_004786 1.28 transduction.
Zinc finger, DHHC-type ZDHHC16 BC004535 1.25 Apoptosis. containing 16
144 signaling, and is pro-apoptotic when over-expressed, but concurrent activation of c-Abl appears to be necessary for this function of ZDHHC16 (Li et al. , 2002a). Thioredoxin-like 1 (TXNL1) was also up-regulated following treatment of 22Rv1 cells with siHuR-UTR siRNA, and in HEK293 cells, has been reported to protect against glucose-deprivation induced cytotoxicity (Jimenez et al. , 2006).
The role of the transcription factor D-site of albumin promoter binding protein (DBP) in apoptosis in the prostate has not been characterised, however it has been implicated in cell cycle control, carcinogenesis, circadian gene regulation, tissue regeneration and apoptosis in the liver (Schrem et al. , 2004). Similarly, the function of the down-regulated gene, human homologue of murine fused- toes (FTS) in apoptosis is unknown. Human FTS is thought to play a role in apoptosis because mice carrying a mutant copy of the FTS gene exhibit fused toes and these morphological abnormalities are caused by altered patterns of programmed cell death (Lesche et al. , 1997). Interestingly, FTS is similar in structure to E3 ubiquitin ligases, however it lacks the conserved cysteine residue that enables those enzymes to conjugate ubiquitin to the target protein (Lesche et al. , 1997).
Dedicator of cytokinesis 1 (DOCK1) and engulfment and mobility 2 (ELMO2) mRNA’s were identified as being down-regulated following knockdown of HuR. Both of the proteins encoded by these mRNA’s are signalling molecules involved in cytoskeletal rearrangements in the focal adhesion pathway, and their role in apoptosis appears to be in the cell surface extension of an engulfing cell around a dying cell (Santy et al. , 2005). Interestingly, ELMO2 is also involved in neurite outgrowth, a process in which Hu proteins also play a role (Katoh and Negishi, 2003).
Sixteen gene products involved in the regulation of the cell cycle were identified as having altered expression in 22Rv1 cells in which HuR had been reduced with siRNA (Table 5.3). As with the apoptosis-associated genes, the cell cycle-associated gene set comprised both direct and indirect cell cycle regulators with both inhibitory and stimulatory functions. The expression of
145 Table 5.3. Cell cycle-associated genes. Up-regulated genes are in purple . Genes previously identified as being associated with apoptosis do not appear in this table. X : value derived from x-suffixed probe set. S: value derived from s-suffixed probe set.
Cell Cycle: Gene ID: Accession Fold GO Biological Process: no: change:
Anaphase promoting G2/M transition; metaphase/anaphase ANAPC10 DQ304649 1.51 complex 10 transition; ubiquitin cycle.
G2/S transition; negative regulation of cell Cullin 1 CUL1 AA682674 1.29 proliferation; cell cycle arrest; ubiquitin cycle.
Cyclin-dependent kinase Cell cycle arrest; negative regulation of cell CDKN2B AW444761 1.82 inhibitor 2B (p15) proliferation; and cyclin-dependent kinase 4.
Dual specificity Response to oxidative stress; cell cycle; DUSP1 NM_004417 1.39 phosphatase 1S dephosphorylation of proteins.
S Negative regulation of cell cycle; signal Exostoses (multiple) 2 EXT2 NM_000401 1.48 transduction; skeletal development.
Guanine nucleotide GNL3 NM_014366 1.23 Cell cycle; cell proliferation. binding protein-like 3
P300/CREBBP Negative regulation of cell proliferation; PCAF AV727449 1.75 chromatin remodelling; cell cycle arrest; associated factor transcription. P53-associated parkin- Cell cycle; metaphase/anaphase transition; PARC AB014608 1.43 like cytoplasmic protein. ubiquitin cycle.
Patched homolog Negative regulation of cell cycle; cell PTCH BG054916 1.24 proliferation; morphogenesis; signal (Drosophila ) transduction. RAD1 homolog ( S. X RAD1 NM_002853 1.33 Cell cycle checkpoint; DNA repair pombe )
Cell cycle; ER/Golgi vesicle-mediated RAD50 interactor 1 RINT1 NM_021930 1.28 transport.
Regulator of G-protein Cell cycle; negative regulation of GPCR RGS2 NM_002923 1.21 signalling 2 S signalling.
Response gene to Regulation of cyclin-dependent protein RGC32 NM_014059 1.31 complement 32 kinase activity.
RNA binding motif protein Negative regulation of cell cycle S RBM5 U23946 1.31 5 progression; RNA processing.
Cell cycle; cell proliferation; negative Stratifin SFN X57348 1.27 regulation of protein kinase.
Structural maintenance of Cell cycle; chromosome condensation, SMC4L1 NM_005496 1.31 chromosomes 4-like 1 S segregation.
146 cyclin-dependent kinase inhibitor 2B (CDKN2B; p15) mRNA was up-regulated following knockdown of HuR. CDKN2B is an inhibitor of cell cycle progression through G 1 phase, via complex formation with the cyclin-dependent kinases (CDK)4 or CDK6, preventing their activation (Hannon and Beach, 1994). Inhibition of CDK4 or 6 by CDKN2B is thought to be stimulated by tumour growth factor β (TGF β), however the activity of TGF β signalling was not investigated in the present study (Rennie and Nelson, 1998).
Other transcripts that were identified as being up-regulated following knockdown of HuR that encode proteins considered to be inhibitors of the cell cycle included stratifin (14-3-3δ), Patched homologue (PTCH) and regulator of G-protein signalling 2 (RGS2). Stratifin is a negative regulator of cell cycle progression that exhibits increased expression in hormone-refractory prostate cancer and drug-resistant breast cancer (Han et al. , 2006; Laronga et al. , 2000). PTCH is the receptor for the human Sonic Hedgehog ligand, a signalling molecule involved in the Notched pathway, a pathway that did not exhibit significant clustering of genes in the current study (Stone et al. , 1996). RGS2 is an inhibitor of G-protein coupled receptor signalling, which also appears to inhibit androgen-independent AR activation in androgen- independent LNCaP cells and CWR22Rv1 cells, and silencing of RGS2 in androgen-sensitive LNCaP cells has been reported to confer a growth advantage under steroid-reduced conditions (Cao et al. , 2006).
The proteins encoded by the response gene to complement 32 (RGC32), dual- specificity phosphatase 1 (DUSP1) and guanine nucleotide binding protein-like 3 (GNL3, nucleostemin) mRNA’s have been reported to stimulate cell cycle progression, and these transcripts were found to be up-regulated in 22Rv1 cells in which HuR had been reduced. RGC32 is a substrate and regulator of CDC2, the over-expression of which induces cell cycle activation (Fosbrink et al. , 2005). Over-expression of RGC32 protein has been reported in colon, bladder, breast, lung and prostate tumours, and is co-expressed with the proliferation marker Ki-67 in malignant epithelial cells of the colon (Fosbrink et al. , 2005). DUSP1 is an inhibitor of kinase signalling, the mRNA and protein of
147 which is down-regulated in a high proportion of prostate carcinomas, although its expression was up-regulated in the present micro-array (Rauhala et al. , 2005). GNL3 is a recently-characterised stem cell protein, which is also essential for cancer cell proliferation, and is up-regulated in malignant renal and breast tumours (Fan et al. , 2006; Sijin et al. , 2004).
A number of genes with roles in cell cycle checkpoints were down-regulated following reduction of HuR using siRNA treatment. RAD1 is the human homologue of Schizosaccharomyces pombe rad1+, a cell cycle checkpoint control gene required for S-phase and G 2/M arrest in response to both DNA damage and incomplete DNA replication (Bluyssen et al. , 1998). Similarly,
RAD50 interacting protein 1 (RINT1) regulates late S- and G 2/M phase transition via interactions with RAD50, a structural maintenance of chromosomes (SMC) protein family member that participates in a variety of cellular processes, including DNA double-strand break repair, cell cycle checkpoint activation, telomere maintenance, and meiosis (Xiao et al. , 2001). SMC4-like 1 (SMC4L1) is a protein very similar to SMC4, which is involved in chromosome condensation at mitosis, suggesting that it may have a similar role (Schmiesing et al. , 1998). RNA-binding motif 5 (RBM5, Luca15, H37) is thought to be a tumour suppressor as the RBM5 gene chromosomal locus is frequently deleted in lung cancer. RBM5 participates in both cell cycle arrest in
G1-phase and in apoptosis (Oh et al. , 2006).
The mRNA’s encoding three E3 ubiquitin ligase complex proteins were among those that were down-regulated. Anaphase-promoting complex 10 (ANAPC10), Cullin 1 (CUL1) and p53-associated parkin-like cytoplasmic protein (PARC), are involved in the degradation of cyclins and cyclin- dependent kinases during the cell cycle (Nakayama and Nakayama, 2006). In addition to its E3 ligase activity, PARC acts as an anchor protein maintaining the cytoplasmic localisation of p53 and p53-associated protein complexes, inhibiting nuclear localisation of p53 and p53-dependent apoptosis (Nikolaev et al. , 2003).
148 The exostoses 2 (EXT2) mRNA was also down-regulated, however the contribution of this gene makes to cell cycle regulation is unknown. Defects in the gene encoding EXT2 are to thought cause the hereditary condition multiple exostoses, characterised by bony outgrowths at the ends of the long bones, which are probably the result of loss of cell cycle control (Duncan et al. , 2001; Stickens et al. , 1996). PCAF is a nuclear co-regulator, which can inhibit cyclin D1-mediated inhibition of the AR, and it is likely that its inclusion in the GO Biological Process database stems from its role in the transcriptional regulation of genes more directly involved in the cell cycle (Reutens et al. , 2001).
Both increased and decreased expression of transcripts encoding proteins involved in cell growth-associated pathways were observed, with pro- and anti- proliferative genes being represented roughly equally (Table 5.4). The transcripts encoding negative regulators of cell proliferation were dickkopf 1 (DKK1), fatty acid binding protein 3 (FABP3), mortality factor 4 (MORF4) and MORF4-like (MORF4L). DKK1 expression was up-regulated in this study, and is an inhibitor of the growth-promoting Wnt signalling pathway. DKK1 has been reported to be down-regulated in melanoma cell lines and malignant melanomas. DKK1 and other members of the DKK family are also thought to be involved in cell-cell adhesion and cell migration (Kuphal et al. , 2006). FABP3 (also known as mammary-derived growth inhibitor [MDGI]), which was down-regulated in the present study, is a small (~ 12 kDa) protein involved in fatty-acid uptake that has been reported to be anti-proliferative in breast cancer cell lines (Huynh et al. , 1995). The expression of MORF4 and MORF4 mRNA’s was increased, and these genes are thought to promote senescence in immortalised cell lines, although their functions are not yet well- characterised (Bertram et al. , 1999).
The positive regulators of cell proliferation were POU domain class 3- containing transcription factor 2 (POU3F2), growth differentiation factor 15 (GDF15) and MAPK kinase kinase (MAP3K) 2 and 11. POU3F2 is a down- stream transcriptional target of Wnt signalling in neural development, which exhibits increased expression in melanoma, however, POU3F2 mRNA levels were decreased in 22Rv1 cells with reduced HuR in the present study
149 Table 5.4. Proliferation-associated genes. Genes that were not otherwise included in the apoptosis or cell cycle sets are included in this table. Up-regulated genes are in purple . X : value derived from x-suffixed probe set. S: value derived from s-suffixed probe set.
Cell Proliferation: Gene ID: Accession Fold GO Biological Process : no: change:
Dickkopf homologue 1 DKK1 NM_012242 1.42 Negative regulation of Wnt signalling.
Fatty acid binding FABP3 AI219891 1.51 Negative regulation of cell growth; transport. protein 3
Fibroblast growth factor Fibroblast-growth factor receptor signalling; FGFR1 M63889 1.89 receptor 1 MAPK signalling; cell growth.
Growth differentiation Transforming growth factor beta signalling; X GDF15 AF003934 1.44 factor 15 cell-cell signalling.
G1 phase; cell proliferation; JNK and MAPK MAPK kinase kinase 11 MAP3K11 NM_002419 1.62 activity; microtubule-based process.
MAPK kinase kinase 2 MAP3K2 N50665 1.35 MAPK signalling.
S Regulation of cell growth; transcription; Mortality factor 4 MORF4 NM_006792 1.44 chromatin remodelling.
S Regulation of cell growth; transcription; Mortality factor 4 like 1 MORF4L NM_006791 1.54 chromatin remodelling; aging.
POU domain, class 3, Positive regulation of cell proliferation; POU3F2 BE855760 1.69 transcription factor 2 transcription.
150
(Goodall et al. , 2004). The expression of GDF15 mRNA was increased in the presence of reduced HuR, which was particularly interesting as GDF15 is a growth factor that has been reported to be up-regulated in prostatic intra- epithelial neoplasias and prostate tumours compared to normal prostate tissue (Cheung et al. , 2004). However, up-regulation of GDF15 protein is thought to be independent of increases in GDF15 mRNA (Cheung et al. , 2004). MAP3K2 and MAP3K11 were up- and down-regulated, respectively, following HuR knockdown in 22Rv1 cells. These proteins are down-stream signalling partners of p38 MAP kinase and c-jun N-terminal kinase (JNK), and therefore impinge on a range of cellular processes. The fibroblast growth factor receptor (FGFR1) is an up-stream signalling protein in the same pathways, being a membrane-associated receptor (Figure 5.4).
5.2.3 Effects of Reduced HuR on Genes Involved in Ubiquitin- Mediated Proteolysis.
A substantial proportion of the cell cycle-, apoptosis- and proliferation- associated genes were associated with ubiquitin-meditated proteolysis, which was also one of the KEGG pathways identified in which significant clustering of genes was observed (Table 5.1). Because defects or alterations in ubiquitin- mediated proteolysis are postulated to have a role in cell cycle dysregulation in some cancers, GO Biological Process definitions were used to retrieve ubiquitin-mediated proteolysis-associated genes from amongst the significantly up- and down-regulated transcripts identified by the present micro-array analysis (Gstaiger et al. , 2001; Lim et al. , 2002; Signoretti et al. , 2002). Twenty-three genes with known or predicted functions in ubiquitin-mediated proteolysis were identified, of which eight were E3 ubiquitin ligases, the enzyme type that is thought to direct the specificity of ubiquitin-mediated proteolysis for protein targets (Table 5.5) (Nakayama and Nakayama, 2006). Of these genes, the E3 ligase Cullin 1 (CUL1) and anaphase-promoting complex 10 (ANAPC10) are part of two of the best-characterised E3 complexes involved in cell cycle regulation, which have been reported to exhibit altered expression in several types of cancer (Nakayama and Nakayama, 2006).
151 Wnt signalling Focal adhesion MAPK signalling DKK1 Fn FGF
FZD ITGB1 FGFR1
CXXC4 +P FAK
DOCK1 PI3K
Rac VAV3 RAF1
+P
PAK2
+P
MEK
PP2A ß-catenin +P
ERK1/2
+P
ELK1
+P
Cell cycle Cell proliferation
Figure 5.4. Convergence of MAPK, Wnt and focal adhesion pathways. Transcripts belonging to the MAPK and focal adhesion pathways that were decreased after HuR knockdown are indicated in green . Genes exhibiting increased expression are in pink . DKK1; Dickkopf homologue 1. ELK1: ETS- like kinase 1. ERK: extracellular regulated kinase. FAK: Focal adhesion kinase. FGF: Fibroblast growth factor. FGFR1: FGF receptor 1. Fn: Fibronectin. FZD: Frizzled homologue 1/8. ITGB1: integrin ß1. MEK: Mitogen- or extracellular-regulated kinase. PAK2: p21-activated kinase 2. PI3K: Phosphatidyl inositol-3 kinase. PP2A: protein phosphatase 2A. Vav; Vav 3 oncogene. +P; phosphorylate. : activation. : inhibition. :multiple steps omitted for clarity. Modified from http://www.genome.jp/kegg/pathway/hsa/hsa04510.html.
152
Table 5.5. Genes associated with ubiquitin-mediated proteolysis. Up-regulated genes are in purple . The function of the protein in the ubiquitin-mediated proteolysis pathway is indicated if known. E2: E2 ubiquitin-conjugating enzyme. E3: E3 ubiquitin ligase. X : value derived from x-suffixed probe set. S: value derived from s-suffixed probe set.
Gene name: Gene ID: Accession Fold Function: no: change: Anaphase promoting complex 10 ANAPC10 AW025455 1.55 E3
Cullin 1 CUL1 AA682674 1.29 E3 Cullin 4B S CUL4B NM_003588 1.27 E3
F-box protein 5 X FBXO5 AK026197 1.22 E3
Fused-toes homologue FTS NM_022476 1.39
Hect-domain and RLD 4 HERC4 AB046813 1.96 E3
Leucine rich repeat and sterile α domain LRSAM1 BG029705 1.48 containing 1 Membrane-associated ring finger 7 MARCH7 NM_022826 1.22 (CH3HC4) Membrane-associated ring finger 8 MARCH8 BE781914 2.34 (CH3HC4) Mindbomb homologue 1 MIB1 AI798164 1.70
Neural precursor cell expressed, NEDD4 D42055 2.19 E3 developmentally down-regulated 4 p53-associated parkin-like cytoplasmic PARC AB014608 1.43 E3 protein Protein inhibitor of STAT 2 PIAS2 AF361054 1.77
Seven in absentia 1 S SIAH1 AI953523 1.31 E3
SMT3 suppressor of mif two 3 S SUMO1 U67122 1.27 homologue 1
SUMO1/sentrin specific peptidase 5 SENP5 N32782 1.56
Ubiquitin-conjugating enzyme E2H UBE2H AI280328 1.26 E2
Ubiquitin-conjugating enzyme E2 J1 S UBE2J1 AF151039 1.60 E2
Ubiquitin-conjugating enzyme E2 variant S UBE2V1 U39361 1.59 E2 1
Ubiquitin-specific peptidase 40 USP40 AA522888 1.32
Ubiquitin-specific peptidase 30 USP30 BC004868 1.40 WW domain containing E3 ubiquitin S WWP1 BF131791 1.54 E3 protein ligase 1
Zinc and RING finger 1 ZNRF1 AL136903 1.52
153
5.3 Discussion Analysis of gene expression changes in 22Rv1 cells treated with siHuR-UTR siRNA compared to cells treated with siCon siRNA using oligonucleotide micro-arrays detected substantially more down-regulated transcripts than up- regulated transcripts. The relatively large proportion of down-regulated genes compared to up-regulated genes was expected in light of the known stabilising role of HuR (Fan and Steitz, 1998a). The increases and decreases in gene expression following siHuR-UTR siRNA treatment were relatively modest, ranging between a 2.95-fold decrease and a 1.9-fold increase. This may have been because the changes detected were due to alterations in mRNA stability rather than to changes in transcriptional activation or repression. This range of increases/decreases was comparable to the change in expression of HuR mRNA, which was determined to have decreased 2-fold following siHuR-UTR siRNA treatment. The results, obtained 72 h following siRNA treatment, are also consistent with the observation that knockdown of the HuR protein persisted until 120 h, and suggest that an alteration of mRNA level of this magnitude is sufficient to result in marked alterations in protein expression (Section 4.2.3).
Significant proportions of genes that exhibited altered expression following treatment of the cells with siHuR-UTR siRNA encoded proteins that are involved in growth-associated signal transduction pathways, such as MAPK, VEGF and Wnt signalling pathways (Table 5.1). The MAPK, VEGF and Wnt signalling pathways are kinase cascades that transmit extracellular stimuli detected at the cell membrane by membrane-anchored receptors to transcriptional effectors via reversible phosphorylation of intracellular proteins (Wilkinson and Millar, 2000). The extra-cellular signals are integrated and modulated by the relative activities of kinases and phosphatases, arranged into complex networks that may be autonomous or subject to cross-regulation. Growth factor signals or stress stimuli can be communicated by MAPK cascades including the p38, JNK and extracellular-receptor regulated kinase (ERK) signalling pathways. MAPK signalling regulates a number of cellular processes, including the cell cycle, and failure to correctly integrate 154 extracellular stimuli with cell cycle regulation can result in accumulation of DNA damage and genomic instability (Wilkinson and Millar, 2000). The integration of a diverse set of extra-cellular signals by the MAPK pathways may account for the observations that many of the up- and down-regulated genes from the present micro-array were common to the MAPK, focal adhesion and regulation of actin cytoskeleton pathways (Table 5.1).
The decreases detected in the expression of mRNA’s encoding proteins associated with MAPK signalling pathways are interesting because pharmacological inhibition of p38 MAPK reduces expression of the anti- apoptotic gene cellular inhibitor of apoptosis (c-IAP2), activates the “executioner” caspase-3, and enhances TNF-induced cell death in murine fibroblast cell lines L929 and NIH3T3 (Luschen et al. , 2004). In addition, it has been proposed that activation of p38 MAPK by IL1 and the IL1 receptor (IL1R) may promote prostate cancer cell proliferation as expression of IL1, IL1R, PAK1 and ELK1 proteins are increased in prostate cancer tissue samples, suggesting that decreased expression of MAPK signalling proteins may be inhibitory to the growth of prostate cancer cells (Ricote et al. , 2006). However, sustained p38 MAPK and JNK signalling has also been reported to induce apoptosis in hepatocytes, and as these signalling pathways are not well- characterised in 22Rv1 cells, it is difficult to predict the outcome of reduction of MAPK protein expression in this cell line (Wang et al. , 2004). ELK1 mRNA levels were also decreased after treatment of the cells with siRNA directed against HuR, however alterations in IL1, IL1R and PAK1 mRNA’s were not observed. As the ELK1 transcription factor is a distal effector of MAPK signalling, it will be interesting to investigate whether steady-state ELK1 protein levels are regulated by HuR and thus whether MAPK signalling overall is modified by HuR levels and activity.
Interestingly, the p38 MAPK cascade has been reported to be involved in regulation of the stability of COX2 and TNF α mRNA’s, which are also targets of HuR (Dean et al. , 2003; Dean et al. , 2004). Stabilisation of COX2 and TNF α mRNA’s by p38 MAPK activation is thought to act via inhibition of mRNA de-adenylation by MAPK-activated protein kinase 2 (MAPKAPK2), which was
155 up-regulated following knockdown of HuR in the present micro-array analysis. This may indicate that the loss of the stabilising activity of HuR could be partially compensated for by up-regulation of MAPKAPK2. Future experiments could measure the effect of over-expression of MAPKAP2 protein combined with knockdown of HuR on COX2 mRNA half-life to determine whether this is the case.
Several genes specific to focal adhesion signalling were significantly down- regulated in 22Rv1 cells that had been treated with siRNA directed against HuR. Focal adhesions are specialised structures formed at the contact point of the cell and the extracellular matrix (Petit and Thiery, 2000). At the focal adhesion, transmembrane adhesion receptors ( eg . integrins) provide a structural link between the actin cytoskeleton and the extracellular matrix components, and these interactions are essential for the control of tissue remodelling, embryogenesis and cell migration (Petit and Thiery, 2000). Therefore, focal adhesion signalling affects the regulation of the actin cytoskeleton, a biological process also identified as being significantly down- regulated (Table 5.1). Focal adhesion signalling is initiated at the cell surface through an αβ integrin complex and Rho-GTPases, and is mediated largely by focal adhesion kinase (FAK), which did not exhibit altered expression in this micro-array (Cox et al. , 2006). There is no integrin-specific signalling mechanism and the down-stream effectors of FAK signalling are shared by other pathways, such as MAPK (Brakebusch and Fassler, 2005). Interestingly, loss of integrin β1 (ITGB1) expression, which was down-regulated 1.88-fold in this micro-array, is rarely fatal to the cell, perhaps because it is one of at least 8 β-integrins commonly expressed in cell and there is redundancy in the use of β1-integrins (Brakebusch and Fassler, 2005). However, nude mice that over- express ITGB1 form larger primary tumours and have increased liver and lung metastases (Brakebusch et al. , 1999). In recent studies of human prostate cancer, it has been reported that the loss of expression of the ITGB1 C-isoform may promote prostate cancer cell invasion by decreasing cell adhesion via down-regulation of IGFII, a factor which increases cell adhesion, whereas the A-isoform of ITGB does not affect IGFII levels (Goel et al. , 2006). It was not possible to determine from the present micro-array data which isoform of
156 ITGB1 was down-regulated, however down-regulation of either isoform may result in changes to 22Rv1 cell morphology or adhesion.
The clustering of genes in the axon guidance and adherens junctions pathways may also relate to modulations of cell morphology. Adherens junctions are cell-to-cell adhesion sites where cadherin-catenin complexes function as cell adhesion molecules, and where the actin-based cytoskeleton is assembled (Nagafuchi, 2001). Although no alteration of the expression of either cadherins or catenins was observed following knockdown of HuR, both up-stream (FGFRI) and down-stream effectors of this pathway (synovial sarcoma X-breakpoint 2 interacting protein and Wiskott-Aldrich syndrome-like) exhibited decreased mRNA expression, which may indicate that alterations to cell adhesion occurred following HuR knockdown (Table 5.1) (de Bruijn et al. , 2002; Miki et al. , 1996). Axon guidance also involves modulation of cell shape, and the finding that neurofibromin 1 expression was down-regulated was particularly interesting as this transcript has previously been reported to be bound by HuR (Haeussler et al. , 2000). The neurofibromin 1 gene encodes a tumour suppressor, the deficiency of which causes the autosomal dominant disease neurofibromatosis type 1. Although neurofibromin 1 is predominantly expressed in neuronal cells, its expression in prostate cancer has not been investigated and it is feasible that it is biologically active in prostate tumours and that its decreased expression in prostate cancer cells may also be pro- proliferative (Trovo-Marqui and Tajara, 2006).
The Wnt signalling pathway was another of the growth-associated signalling pathways in which there was significant clustering of down-regulated genes. The Wnt proteins are secreted ligands that bind to and activate their cognate receptors, human homologues of the frizzled (FZD) proteins, and initiate intra- cellular signalling that stimulates cell proliferation (Polakis, 2000). Although down-regulation of the FZD homologue 1 and 8 mRNA’s suggests that Wnt signalling may be decreased following siHuR-UTR siRNA treatment, CXXC4 and the regulatory and catalytic subunits of protein phosphatase 2A (PP2A) mRNA’s were also down-regulated, and these mRNA’s encode repressors of Wnt signalling (Figure 5.4). As such, the activity of the Wnt signalling pathway
157 could not be predicted by examination of genes with altered expresion following siHuR-UTR siRNA treatment and further studies will need to be performed to examine interactions between HuR and the Wnt pathway.
HuR may affect both the basal levels of Wnt signalling and the activity of the pathway in response to external stimuli. The AR gene is a transcriptional target of Wnt signalling and the AR protein is a down-stream phosphorylation target of the Wnt signalling pathway, therefore changes in Wnt signalling may alter the androgen responsiveness of prostate cancer cells (Terry et al. , 2006; Yang et al. , 2006). Interestingly, two previously reported targets of HuR, β- catenin, the major down-stream effector of Wnt signalling, and the ligand Wnt- 5a were not identified as having altered expression following siRNA-mediated knockdown of HuR in this study, however the cell type and other experimental conditions differed between these studies and it is likely that HuR exhibits tissue-specific effects with relation to its activity and the expression of its target genes (de Silanes et al. , 2003; Leandersson et al. , 2006; Polakis, 2000).
Glycerophospholipid and glycine, serine and threonine metabolism pathways exhibited a significant proportion of up-regulated genes in 22Rv1 cells that had been treated with siHuR-UTR siRNA, whereas the glutamate and nitrogen metabolism pathways exhibited a significant proportion of down-regulated genes, suggesting general metabolism in these cells was affected. This may have been directly due to the knockdown of HuR, or may have occurred as consequence of alterations in the activity of other signalling pathways. The finding of defined alterations in metabolic pathways was interesting however, in view of the altered proliferation of the cells following HuR knockdown.
Examination of the expression levels of apoptosis-, cell cycle- and proliferation- associated genes identified changes in mRNA levels of both pro- and anti- proliferative genes. Among the down-regulated genes were BAD, CUL1, FABP3 and RBM5, which are all implicated in negative regulation of cell proliferation or are pro-apoptotic, according to GO Biological Process definitions. Conversely, the down-regulated genes BIRC2, POU3F2, and RAF1 are all pro-proliferative or anti-apoptotic by the same definitions, while the anti-
158 proliferative genes CDKN2B, DKK1 and stratifin were up-regulated. However, BAG1, a potent suppressor of apoptosis implicated in aggressive prostate cancer (Krajewska et al. , 2006), was also up-regulated, as were the mRNA levels of GDF15, the expression of which has also been linked to prostate carcinogenesis (Cheung et al. , 2004). As the sum of these pro- and anti- proliferative signals cannot be predicted without further investigation, the altered expression of these transcripts is difficult to interpret.
Twenty-two of twenty-three transcripts encoding proteins involved in ubiquitin- mediated proteolysis were down-regulated in cells treated with siHuR-UTR siRNA, indicating that these cells may have alterations in the general cellular proteolysis pathways. The finding that HuR might regulate ubiquitin-mediated proteolysis is an intriguing one, as defects in ubiquitin-mediated proteolysis are associated with changes in cell proliferation, genome instability and cancer (Ciechanover and Schwartz, 2004). This is because ubiquitin-mediated proteolysis is a critical component of cell cycle regulation, being responsible for the appropriate degradation of cyclins and cyclin-dependent kinase inhibitors to facilitate cell cycle progression (Nakayama and Nakayama, 2006). Ubiquitin also targets such oncoproteins as M-myc, N-myc, c-Jun, β-catenin and SRC for proteolysis. Tumour suppressor genes such as p53 and p27 are also targets of ubiquitin-mediated proteolysis, and destabilisation of these has also been implicated in oncogenesis (Ciechanover and Schwartz, 2004). Thus, either gain or loss of function of ubiquitin-mediated proteolysis can have profound impacts on the cell.
Ubiquitination is a three-step enzymatic process, illustrated in Figure 5.5. The major enzyme, ubiquitin-activating enzyme (E1), plays a general role, activating the cellular pool of ubiquitin molecules, which are then covalently attached to target proteins by the ubiquitin-conjugating enzyme (E2). The E3 enzyme determines specificity of the E2 enzyme by binding to the target protein and to E2, bringing them into close proximity. The poly-ubiquitinated protein is then degraded by the proteasome (Guardavaccaro and Pagano, 2004). Eight of the ubiquitin-mediated proteolysis-associated genes decreased after HuR knockdown were E3 ubiquitin ligases, and three were E2
159 E3
26S proteasome E1 E2 Target protein S S Ub O= C O= C Ub Ub Ub Ub Ub AMP Ub Ub ATP Degraded Ub product ATP AMP Ub Ub Ub Recycling
Figure 5.5. Ubiquitin-mediated proteolysis . Ubiquitin (Ub) is a small 8-kDa protein, which is first transferred to the ubiquitin-activating enzyme, E1, in an ATP-dependent manner. Activated ubiquitin is then transferred to the ubiquitin-conjugating enzyme, E2. Finally, the ubiquitin is covalently attached to the target protein by E3 ubiquitin ligase, leading to the formation of a polyubiquitin chain. The polyubiquitylated protein is recognised by the 26S proteasome, and is destroyed in an ATP-dependent manner. Ub is released for recycling. Reproduced from Nakayama and Nakayama (2006).
160 conjugating enzymes. Interestingly, another E2, UBE2N, has previously been identified as a transcript to which HuR binds (de Silanes et al. , 2004).
The E3 ligase protein, CUL1, the transcript for which was down-regulated in this study, is a structural component of the Skp1-CUL1-Rbx-F-box (SCF) complex, which is involved in G 1-S phase transition of the cell cycle (Nakayama et al. , 2000). Substrate specificity of SCF complexes is directed by the variable F-box protein, and the over-expression of the F-box protein Skp2 has been reported in a number of cancers and correlates with decreased p27 Kip1 expression, indicating loss of cell cycle control (Gstaiger et al. , 2001; Lim et al. , 2002; Nakayama et al. , 2000). Another F-box protein, F-box protein 5 (FBXO5), was down-regulated in the current study, and although it is likely to be analogous to Skp2, the substrate specificity of this protein is unknown, and therefore the effect of down-regulation of FBXO5 cannot be predicted.
ANAPC10 was also down-regulated in the present micro-array analysis and is a subunit of the anaphase promoting complex/cyclosome (APC/C), which is necessary for separation of sister chromatids during mitosis and for M-G1 phase transition (Nakayama et al. , 2000; Nakayama and Nakayama, 2006). The down-regulation of this gene may therefore also be significant, although fewer alterations to the APC/C have been reported in cancer than alterations in SCF complexes (Nakayama and Nakayama, 2006).
Down-regulation of E1 ubiquitin ligases was not detected in this study, and loss or mutation of E1 in cancer has not been reported. However, loss of E1 ubiquitin ligase activity is likely to be fatal to the cell, whereas mutation/altered expression of several E3 ubiquitin ligases have been reported in human cancers (Guardavaccaro and Pagano, 2004). Interestingly, AR protein is degraded by ubiquitin-mediated proteolysis, although this is via the E3 ligase mdm2, which was not altered in this micro-array (Gaughan et al. , 2005). Down-regulation was also observed of the transcript encoding SUMO1, another protein involved in AR modification (sumoylation) and thus regulation of AR protein levels (Poukka et al. , 2000). However, a decrease in the expression of proteins involved in the degradation of AR would be likely to
161 result in an increase in AR protein levels, which was not observed (Section 4.2.3).
Recent studies have shown that proteasome inhibitors might be beneficial in certain pathologies, although they may affect many processes non-specifically (Ciechanover and Schwartz, 2004). Inhibition of specific SCF complexes might confer more target specificity, and therefore have more therapeutic benefit (Ciechanover and Schwartz, 2004). For these reasons, it may be useful to determine whether HuR regulates the mRNA stability of proteins involved in proteolysis, particularly E3 ubiquitin ligases, and whether this has a functional effect. This could be achieved by comparing levels of ubiquitination across a spectrum of proteins using an anti-ubiquitin antibody, because the substrates for most E3 ligases have not yet been identified, therefore it would be difficult to assess the effect of HuR knockdown on specific targets of ubiquitination (Nakayama and Nakayama, 2006).
To develop a clear picture of the processes occurring in the siHuR-UTR- treated cells, micro-array analysis would have to be performed at earlier and later time-points over the course of the experiment, and correlated with both HuR protein levels and other markers of the status of key target pathways ( i.e. signalling pathways, cell cycle, apoptosis). The time-point of 72 h after siRNA treatment was chosen because HuR protein knockdown was greatest at this time-point and it was anticipated that HuR targets would also be most evident at that time. It is also possible, however, that earlier time-points may have produced a better noise-to-signal ratio, perhaps at the expense of identifying HuR targets that exhibited more subtle regulation. Similarly, later time-points may have identified down-stream effects of HuR knockdown that would better reflect the long-term consequences of altering HuR levels and/or activity and that are physiologically or clinically more relevant.
Many important effectors of cellular responses are phosphoproteins, including the AR and members of kinase cascades, and oligonucleotide micro-arrays cannot measure the activation status of such proteins. In addition, changes in translation of mRNA’s are not detected by oligonucleotide micro-arrays, a point
162 pertinent to the current studies, as both HuR and HuD have been shown to translationally repress a variety of genes, including p27 Kip1 (Kullmann et al. , 2002) and inflammatory cytokines (Katsanou et al. , 2005), and HuR translationally enhances p53 in response to UV (Mazan-Mamczarz et al. , 2003). In addition, if HuR regulates pathways such as ubiquitin-mediated proteolysis as was suggested by the regulation of a number of mRNA’s encoding members of this pathway, effects on activity of the pathway may not be able to be predicted using oligonucleotide micro-array assays and would need to be confirmed using functional assays. Therefore, alterations in the levels of proteins encoded by HuR-regulated mRNA’s should be confirmed using techniques such as western blotting or immunohistochemistry and biological assays for specific pathways need to be employed to elucidate functional effects. Interpretation of results should also take into account that minor changes in transcript levels following knockdown of HuR that are confirmed by the detection of small changes in protein levels and/or functional assays may be biologically significant when sustained and may have profound effects on tumour growth, invasion or metastasis.
This micro-array analysis resulted in the identification of a number of candidate HuR targets, most of which had not been reported previously, however the spectrum of genes isolated did not establish a simple explanation for the increased proliferation of 22Rv1 cells with reduced HuR. Examination of the differentially expressed genes revealed both pro- and anti-proliferative and pro- and anti-apoptotic factors as well as several genes with bifunctional roles with respect to these cellular functions (Wang et al. , 1999).
The findings highlight the central role of HuR in regulating fundamental pathways essential for normal physiology as well as the importance of cross- talk between pathways to maintain and co-ordinate cell activity. The identification of HuR regulation of signalling molecules was of particular interest due to the central role of signalling pathways in maintaining normal functions of cells and the well-documented dysregulation of these pathways in malignancy (Bradham and McClay, 2006). Due to their potential importance in prostate cancers, and the degree of caution that had to be applied to data
163 gained from a small number of replicates (2), HuR regulation of a subset of these candidate target mRNA’s was further investigated.
164
CHAPTER 6: Identification of Potential HuR Targets.
165
6 Chapter 6: Identification of Potential HuR
Targets.
6.1 Introduction. Previous reports have identified 78 mRNA targets of HuR (summarised in Chapter 1; Table 1.1), many of which are implicated in cancer. HuR is thought to recognise target mRNA’s via interactions between its RRM’s and single- stranded U-rich sequences present in the 3’UTR (Maris et al. , 2005). Although the binding site for HuR has historically been described as a U-rich AURE type III (Section 1.4.1), two refinements on this motif have been suggested recently, a highly degenerate motif (U/G)(U/G)N(W/G)(W/G)(W/G)NNNU(U/S)NNW (W/G)N(W/G)NN(W/C) described by de Silanes et al. (2004) and the nonamer NNUUNNUUU, described by Meisner et al. (2004). These consensus sequences are relatively similar, being U-rich, but the uridine bases appear in fixed positions in the motif of Meisner et al. , whereas the motif of de Silanes et al. contains only one invariant base. These two consensus sequences were derived by different approaches. In the report by Meisner et al., the dissociation constant (K d) for HuR binding to mRNA sequences containing variations of the canonical AUUUA pentamer were calculated and the sequence NNUUNNUUU was bound with the highest affinity. Meisner et al. (2004) also found this motif in 49 of 50 mRNA’s shown experimentally to be bound by HuR (including the AR), with a single U to C mismatch present in the 50 th . de Silanes et al. (2004) determined their consensus sequence using a oligonucleotide micro-array to identify transcripts that co-immunopurified with HuR, and aligned mRNA 3’UTR’s to identify a conserved stem-loop motif (de Silanes et al. , 2004). It is probable that additional mRNA’s bound by HuR remain to be isolated, however the identification of consensus sequences present in the 3’UTR’s of reported HuR targets indicates that such sequences may be useful for predicting whether a given transcript is bound by HuR.
166 Using oligonucleotide micro-arrays, the expression of a large number of mRNA’s was found to be decreased following treatment of 22Rv1 cells with siRNA directed against HuR. These genes are likely to be either direct targets of HuR, or secondary targets, the expression of which was perturbed by alterations in other genes. To investigate whether individual mRNA’s may be direct HuR targets, the 3’UTR’s of the down-regulated genes were screened for the presence of HuR consensus sequences.
6.2 Results.
6.2.1 Selection of a Consensus HuR Binding Motif.
To select one of the two recently-proposed HuR consensus motifs, the 3’UTR’s of 21 HuR target mRNA’s reported subsequent to the investigation by Meisner et al. (2004) were examined for the presence of the motifs (Cho et al. , 2006; Deschenes-Furry et al. , 2005b; Gantt et al. , 2005; Guo and Hartley, 2006; Izquierdo, 2006; Jeyaraj et al. , 2005; Kawai et al. , 2006; Kloss et al. , 2004; Leandersson et al. , 2006; Li et al. , 2005; Linker et al. , 2005; Mehta et al. , 2006; Meng et al. , 2005; Okamoto et al. , 2004; Pan et al. , 2005; Sheflin et al. , 2004; Sommer et al. , 2005; Song et al. , 2005; Westmark et al. , 2005; Xu et al. , 2005; Zhang et al. , 2006; Zou et al. , 2006). Because of the degeneracy of the motif proposed by de Silanes et al. (2004), it was not possible to screen 3’UTR’s for these sequence variants. The HuR binding motif characterised by Meisner et al. (2004) was located in 19 of 21 of the 3’UTR’s investigated, with the 3’UTR of one of the proposed targets being undefined on the NCBI website (Table 6.1). Therefore, this motif was present in the 3’UTR’s of all but 4 of the 78 mRNA’s to which HuR has been reported to bind to date (Section 1.4.3; Table 1.2).
The frequency with which the putative HuR binding motif was expected to occur was calculated for the 3’UTR’s of the 21 recently-reported targets of HuR, and compared to the actual frequencies with which it occurred (Section 2.9.2). The motif had an expected frequency of 1/1024 nucleotides but occurred an average of ~ 4 times more often than predicted (1/259
167
Table 6.1. Analysis of recent HuR target 3’UTR sequences. 3’UTRs were analysed for the presence of the NNUUNNUUU motif, and the expected and actual frequency of this motif calculated for the individual UTRs and for the total length of UTR. The minimum requirement for inclusion in this table was an IP-RT-PCR assay indicating an in vivo association between HuR and the transcript. 1: (Deschenes-Furry et al. , 2005b). 2: (Pan et al. , 2005). 3: (Izquierdo, 2006). 4. (Jeyaraj et al. , 2005). 5. (Guo and Hartley, 2006). 6: (Cho et al. , 2006). 7: (Gantt et al. , 2005). 8: (Kawai et al. , 2006). 9: (Sheflin et al. , 2004). 10: (Zhang et al. , 2006). 11: (Song et al. , 2005). 12: (Li et al. , 2005). 13: (Meng et al. , 2005). 14: (Linker et al. , 2005). 15: (Zou et al. , 2006). 16: (Westmark et al. , 2005). 17: (Kloss et al. , 2004). 18: (Okamoto et al. , 2004). 19: (Xu et al. , 2005). 20: (Leandersson et al. , 2006). 21: (Sommer et al. , 2005). a: 3’UTR not defined in GenBank. *HuR is reported to bind to the 5’UTR of this transcript, and the motif was present in both 5’ and 3’UTRs.
Gene Gene name: GenBank Length NNUUNNUUU: symbol: Accession: UTR (nt): Expected: Actual: ACHE 1 Acetylcholinesterase AY750146 3987 3.8 4 ATF3 2 Activating transcription factor 3 NM_001030287 1236 1.2 2 ATP5B 3 F1 Beta ATPase NM_001686 162 0.2 1 ATP6V1E1 4 Vacuolar H +-translocating ATPase NM_001696 538 0.5 1 CCNE1 5 Cyclin E1 NM_001238 544 0.5 2 CDC2 6 Cell division cycle 2 NM_033379 212 0.2 1 CEBPB 7 CCAAT/enhancer binding protein NM_005194 604 0.6 4 CYTC 8 Cytochrome C BC024216 - - a EGF 9 Epithelial growth factor NM_001963 811 0.8 7 GADD45A 10 Growth arrest/DNA damage inducible NM_001924 626 0.6 3 gene 45 α GCSh 11 γ-glutamylcysteine synthetase heavy M90656 628 0.6 5 chain HIF1A 9 Hypoxia-inducible factor 1 α NM_001530 2765 2.7 6 HOX11 12 Homeobox protein 11 M75952 867 0.8 4 IGF1R 13 Insulin-like growth factor receptor 1 NM_000875 46, 840 0.04,0.8 1,1* INOS 14 Inducible nitrogen oxide synthase X73029 477 0.5 1 NPM1 15 Nucleophosmin NM_002520 325 0.3 3 RHOB 16 RhoB GTPase NM_004040 1401 1.4 9 SGC 17 Soluble guanyl cyclase Y15723 419 0.4 0 SH2D1A 18 SH2-domain containing 1A NM_002351 1788 1.7 9 SLC11A 19 Solute carrier family 11 member 1 NM_000578 1872 1.8 2 WNT5A 20 Wnt homologue 5A NM_003392 4385 4.3 30 YES1 21 Yamaguchi sarcoma viral oncogene M15990 2678 2.6 9 homolog 1 Totals: 27210 26.5 105
168 nucleotides) in the 3’UTR’s listed in Table 6.1. Analysis of the 3’UTR’s of the proposed HuR targets described by de Silanes et al. (2004) also revealed that the motif of Meisner et al. (2004) was present in 15 of the 17 proposed targets, with the 3’UTR sequences of the remaining transcripts not being available on the NCBI website (Table 6.2). In addition, RNA foot-printing analysis of one of these sequences, the metastasis-associated protein 1 (MTA1) 3’UTR following incubation of this region with HuR indicated that the sequence proposed by de Silanes et al. (2004) was not protected from cleavage by RNases, whereas the NNUUNNUUU region was intact (Dr Andrew Barker; personal communication). Based on these observations, the motif of Meisner et al. (NNUUNNUUU) was selected for use as the consensus site to identify potential mRNA targets of HuR from amongst the genes found to be down-regulated in the micro-array analysis of siHuR-UTR treated 22Rv1 cells (Chapter 5).
6.2.2 Identification of Potential Novel mRNA Targets of HuR.
Nineteen genes were selected for 3’UTR analysis from amongst the 582 genes identified by micro-array as being significantly down-regulated in 22Rv1 cells following treatment with siHuR-UTR siRNA (Table 6.3; Chapter 5). These 19 genes were selected on the basis that they exhibited > 30 % (1.4-fold) down- regulation of the mRNA following siHuR-UTR treatment, and that the proteins encoded by the mRNA’s acted in pathways from which multiple mRNA’s had been down-regulated, including cell cycle and proliferation, or AR signalling, according to GO Biological Process or KEGG definitions (Chapter 5). Of the 19 genes, 12 were involved in apoptosis, cell proliferation, or cell cycle according to GO Biological Process definitions. Seven genes belonged to the focal adhesion/MAPK signalling pathways (which exhibit considerable overlap), of which five had not already been included in the cell proliferation set. Three genes involved in AR signalling or implicated in prostate cancer were also identified, of which two were unique to that set (Table 6.3).
To identify transcripts that were likely to be targets of HuR, the 3’UTR’s of the 19 selected genes were examined for the nonamer NNUUNNUUU. The 3’UTRs were then plotted using mfold to determine which of those nonamers were likely to be presented in a single-stranded confirmation (Table 6.4)
169
Table 6.2. Analysis of 3’UTRs from de Silanes et al. (2004). 3’UTRs, from 17 selected genes that were enriched in HuR immuno-precipitates, were examined for the presence of the NNUUNNUUU motif. a: 3’UTR not formally defined. ACTG1, UBE2N and MTA1 were also included in the analysis of Meisner et al. (2004).
Gene Gene name: GenBank NNUUNNUUU: symbol: accession: ACTG1 Actin gamma 1 NM_001614 CRYZL1 Crystallin, zeta (quinone reductase)-like 1 BC013155 CXCL6 Chemokine (C-X-C motif) ligand 6 BC013744 FKBP1A FK506 binding protein 1A, 12 kDa BC119733 KPNA4 Karyopherin alpha 4 (importin alpha 3) BC028691 LETM1 Leucine zipper-EF-hand transmembrane BC021208 MCM10 Minichromosome maintenance deficient 10 BC009108 MMP7 Matrix metalloprotease 7 BC003635 MTA1 Metastasis-associated 1 NM_004689 NFE2L2 Nuclear factor (erythroid-derived 2)-like 2 BC011558 PC4 RNA polymerase II transcription cofactor 4 BC008883 a PP1CB Protein phosphatase 1, catalytic subunit BC002697 PTMA Prothymosin, alpha BC034921 PVR Poliovirus receptor NM_006505 RPF1 RNA processing factor 1 AF322053 a RPL14 Ribosomal protein L14 BC000606 SLC1A5 Solute carrier family 1 AF105230 SNX3 Sorting nexin 3 BC016863 UBE2N Ubiquitin conjugating enzyme E2 NM_003348 UCHL1 Ubiquitin thiolesterase NM_004181
170
Table 6.3. Genes selected for follow-up analysis. Genes down-regulated >1.4-fold that are involved in cell cycle, AR/prostate, MAPK signalling and focal adhesion, according to combined KEGG and GO definitions. Genes may appear twice if they fall into more than one category.
GenBank Fold Gene Name: Gene Title: Accession No.: Change: Cell Cycle/Proliferation: Anaphase promoting complex subunit 10 ANAPC10 DQ304649 1.51 BCL2-antagonist of cell death BAD BC095431 1.41 Bifunctional apoptosis regulator BFAR NM_016561 1.71 Dedicator of cytokinesis DOCK1 NM_001380 1.53 Engulfment and cell motility 2 ELMO2 NM_022086 2.42 Fatty acid binding protein 3 FABP3 NM_004102 1.51 Fibroblast growth factor receptor 1 FGFR1 AY585209 1.89 Glypican 4 GPC4 NM_001448 1.53 MAPK kinase kinase 11 MAP3K11 NM_002419 1.62 p300/CBP-associated factor PCAF NM_014409 1.75 POU domain, class 3, transcription factor 2 POU3F2 NM_005604 1.69 Receptor activation factor 1 RAF1 NM_002485 1.66 AR/Prostate: CREB binding protein CREBBP AW183655 1.59 p300/CBP-associated factor PCAF NM_014409 1.75 Protein inhibitor of activated STAT, 2 PIAS2 AF361054 1.77 MAPK Signalling and Focal Adhesion: ELK1, member of ETS oncogene family ELK1 M25269 1.63 Fibroblast growth factor receptor 1 FGFR1 AY585209 1.89 Integrin β1 ITGB1 NM_002211 1.88 MAPK-activated protein kinase 3 MAPKAPK3 NM_004635 1.56 P21 (CDKN1A)-activated kinase 2 PAK2 AF092132 1.8 Protein phosphatase 2A, PP2A NM_018444 1.55 Receptor activation factor 1 RAF1 BF823525 1.66
171
Table 6.4. Evaluation of HuR binding sites in 3’UTRs of selected genes. Genes from Table 6.3 were analysed for single-stranded NNUUNNUUU motifs using mfold. ARED2.0 refers to the AU-rich element database classifications of AU-rich motifs: III: AUUUAUUUAUUUA. IV: WAUUUAUUUAW. V: WAUUUAW, where W= A/U.
Gene Title: Length of 3’UTR: Total no. of Single-stranded ARED2.0 cluster nonamers: nonamers:
ANAPC10 2196 15 1 - BAD 372 0 0 - BFAR 824 4 0 - CREBBP 1254 14 5 III DOCK1 1189 7 1 - ELK1 1307 10 2 IV + V ELMO2 1337 2 0 - FABP3 633 4 1 - FGFR1 2504 5 0 - GPC4 1831 15 3 V ITGB1 1177 12 2 - MAP3K11 537 0 0 - MAPKAPK3 1242 4 0 - PAK2 3892 23 4 - PCAF 1893 10 3 - PIAS2 147 0 0 - PP2A 2411 13 4 - POU3F2 2670 21 2 - RAF1 904 3 0 -
172 (Zuker, 2003). This eliminated BAD, BFAR, ELMO2, FGFR1, MAP3K11, MAPKAPK3, PIAS2, and RAF1 from further analysis. The 2 nd version of the AU-rich element database (ARED2.0; http://rc.kfshrc.edu.sa/ARED/) was also searched, to see whether any of these transcripts had been previously identified as containing one of the canonical AU-rich motifs (Bakheet et al. , 2003). Only CREBBP, ELK1 and glycipan 4 (GPC4) were included in the ARED2.0 database (Table 6.4).
From the remaining 11 genes, final selection for analysis of HuR/mRNA interactions was based on the inclusion of the protein encoded by the down- regulated mRNA in converging pathways in which multiple members were decreased, or its implication as an AR co-regulator in prostate cancer. Specifically, ITGB1, PAK2 and ELK1 were selected, as they appear in the MAPK/focal adhesion pathways along with RAF1 and FGFR1 (Figure 6.1). CREBBP and PCAF were selected because they have been reported to be AR co-activators (Smith and O'Malley, 2004). The secondary structure plots of the CREBBP, PCAF, ITGB1, PAK2 and ELK1 3’UTRs are shown in Figures 6.2- 6.6. The sequences of the single-stranded regions of the 3’UTRs of CREBBP, PCAF, ITGB1, PAK2 and ELK1 that contained NNUUNNUUU motifs were aligned manually, and the alignment used to generate a base-frequency plot using WebLogo (Figures 6.7A, B; Section 2.12.1). The base-frequency plot generated from the present alignment was compared to that derived by de Silanes et al. (2004) (Figure 6.7B). Comparison of the two consensus sequences demonstrated that the central U-rich regions of the sequences were similar, although the consensus of de Silanes et al. (2004) had a high proportion of guanine bases in the central and flanking regions, which were absent from the consensus sequence generated from the present study (Figure 6.7B).
The down-regulation of the CREBBP, PCAF, PAK2 and ELK1 mRNA’s following treatment of 22Rv1 cells with siHuR-UTR siRNA was confirmed by qPCR (Figure 6.8). Confirmation of the decrease in ITGB1 mRNA expression was not attempted due to time constraints. Levels of CREBBP, PCAF, PAK2 and ELK1 transcripts were measured relative to actin mRNA level in the same
173 +P +P Fn ITGB1 FAK PI3K
Focal adhesion +P
PAK2
+P +P FGF FGFR1 RAF1 MEK
MAPK signalling +P
ERK1/2
+P Nucleus ELK1
+P CREBBP Transcription of proliferative genes PCAF
Cell proliferation
Figure 6.1. Convergence of MAPK and focal adhesion pathways. Transcripts belonging to the MAPK and focal adhesion pathways that were decreased after HuR knockdown are indicated in green. CREBBP and PCAF are involved in transcriptional regulation by ELK1 on some promoters. Those genes that also contain a potential HuR binding site are circled. ERK: extracellular receptor-regulated kinase. FAK: Focal adhesion kinase. FGF: Fibroblast growth factor. Fn: Fibronectin. MEK: Mitogen- or extracellular- regulated kinase. PI3K: Phosphotidyl inositol-3 kinase; +P; phosphorylates. For other abbreviations, see Table 6.3. Broken lines on arrows denote multiple steps omitted for clarity. Modified from www.genome.jp/kegg/pathway/hsa/hsa04510.html.
174 8010-8019
7800-7806
7730-7738 7598-7608
8158-8166
CREBBP
Figure 6.2. Identification of potential HuR binding sites in CREBBP. The entire 3’UTR from CREBBP was plotted using mfold. Single-stranded HuR binding sites are indicated by arrows. Numbers refer to single-stranded bases in the stem-loops. Numbering starts from the beginning of the GenBank transcript (Table 6.3).
175 1909-1926
2760-2768
ELK1
Figure 6.3. Identification of potential HuR binding sites in ELK1. The entire 3’UTR from ELK1 was plotted using mfold. Single-stranded HuR binding sites are indicated by arrows. Numbers refer to single-stranded bases in the stem-loops. Numbering starts from the beginning of the GenBank transcript (Table 6.3).
176 3520-3527
3321-3340 ITGB1
Figure 6.4. Identification of potential HuR binding sites in ITGB1. The entire 3’UTR from ITGB1 was plotted using mfold. Single-stranded HuR binding sites are indicated by arrows. Numbers refer to single-stranded bases in the stem-loops. Numbering starts from the beginning of the GenBank transcript (Table 6.3).
177 3433 -3448
1923-1943
4623-4634
5662-5693
PAK2
Figure 6.5. Identification of potential HuR binding sites in PAK2. The entire 3’UTR from PAK2 was plotted using mfold. Single-stranded HuR binding sites are indicated by arrows. Numbers refer to single-stranded bases in the stem-loops. Numbering starts from the beginning of the GenBank transcript (Table 6.3).
178 3133- 3147
3633-3663
4071-4078
PCAF
Figure 6.6. Identification of potential HuR binding sites in PCAF. The entire 3’UTR from PCAF was plotted using mfold. Single-stranded HuR binding sites are indicated by arrows. Numbers refer to single-stranded bases in the stem-loops. Numbering starts from the beginning of the GenBank transcript (Table 6.3).
179 NNUUNNUUU A 7598 7608 CREBBP UUCUUUUUAUU 7730 UUUUUUUUU 7738 7800 UUUGUUU 7806 8010 GAUUUUUUUC 8014 8158UGUUGCUUU 8166
1923 PAK2 GAUUUAUUAAUUUU 1945 3433 UUUAUUUCCAUUAUUU 3448 4623UUAUUUAUUUUA 4634 5662 UUUUAUGCCUUUUAUUU… …UUAAUUUAAUCCCGU 5693
1909 1926 ELK1 CGACCUUUCUUUCUUUA 2760CUUUUUUU 2768
3133 3147 PCAF CAUUUUUAUUUUAUG 3633 AUUUUAAUAAUUAUUUUUCUCUACA 3663 4071 UUUAUUUC 4078
ITGB1 3321 UUUAUUUUA… …UUAUUUUUAUU 3340 3520 CUUACUUU 3527
AR 4031 CUGGGCUUUUUUUUUCUCUUUCUCUC… …CUUUCUUUUUCUUCUUCCCUCCCUA 4082
Flanking regionCentral region Flanking region B BITS NNUUNNUUU
5’ 3’
2
1 BITS De Silanes et al. (2003) 0 7 8 9 2 4 5 1 3 6 12 17 5’ 10 11 13 14 15 16 18 19 20 3’
Figure 6.7. Alignment of potential HuR binding sites. A. Single-stranded regions of the 3’UTRs that contained the NNUUNNUUU motif from the CREBBP, PCAF, ELK1 and PAK2 mRNA’s were aligned. The ARUC region of the AR mRNA is included for comparison. B. The single- stranded NNUUNNUUU regions from A (except the ARUC sequence) were used to generate a base-frequency plot (top), and compared to the consensus from de Silanes et al. (2004) (bottom). See Table 6.3 for abbreviations of gene names.
180 ABPAK2 ELK1 * * 120 120 100 100 80 80 60 60 40 40
Relativelevel mRNA 20 Relativelevel mRNA 20 0 0 siCon siHuR-UTR siCon siHuR-UTR
C PCAF D CREBBP * * 120 120
100 100
80 80
60 60 40 40 20 20 Relativelevel mRNA Relativelevel mRNA 0 0 siCon siHuR-UTR siCon siHuR-UTR
E % Change % Change Fold Change Fold Change Gene Name: (qPCR) (micro-array) (qPCR) (micro-array) CREBBP 20 37 1.25 1.59 ELK1 40 37.5 1.7 1.6 PAK2 35 45 1.54 1.8 PCAF 40 57 1.7 1.75
Figure 6.8. Confirmation of reduction at the mRNA level of PAK2, ELK1, CREBBP and PCAF. mRNA levels of PAK2 (A), ELK1 (B), PCAF (C) and CREBBP (D) in 22Rv1 cells were measured by qPCR and expressed relative to actin mRNA levels, following 72 h of treatment of the cells with siCon or siHuR-UTR siRNA, as indicated. Data represent mean ± sd of quadruplicate samples, from at least 2 independent experiments. Asterisks represent differences which are significant (p < 0.05). See Table 6.3 for abbreviations of gene names. E. Comparison of percentage and fold decreases in CREBBP, ELK1, PCAF, and PAK2 mRNA’s as determined by qPCR and cDNA micro-array.
181 samples that were used in the micro-array, as well as in additional independent samples 72 h after transfection of the 22Rv1 cells with siRNA directed against HuR (Section 2.8.2). Decreases of ~ 40 % (1.7-fold) in ELK1 mRNA, ~ 40 % (1.7-fold) in PCAF mRNA levels, ~ 35 % (1.54-fold) in PAK2 mRNA levels, and ~ 20 % (1.25-fold) in CREBBP mRNA levels were observed in cells transfected with siHuR-UTR compared to cells transfected with siCon (Figure 6.8A-D). The decreases in CREBBP, PCAF and PAK2 were smaller than those determined by micro-array, while the decrease in ELK1 mRNA was slightly larger (Figure 6.8E). The expression levels of CREBBP, PCAF, PAK2 and ELK1 mRNA’s as measured by micro-array analysis were also compared to the median gene expression level of all genes in the micro-array, to determine the relative abundance of these transcripts (Chapter 5). Gene expression data was scaled and transformed using binary logarithm (log2) such that the mean expression level across all genes was equal to zero, and up-regulated genes were assigned positive values and down-regulated genes negative values (Figure 6.9; Section 2.12; Appendix 4). Of CREBBP, PCAF, PAK2 and ELK1, only PAK2 had high (> median) expression in control siRNA- treated cells as determined by the micro-array, and ELK1 had the lowest relative expression of the four genes (Figure 6.9).
As the CREBBP, PCAF, PAK2 and ELK1 mRNA’s contained a putative HuR binding site in the 3’UTR, the in vivo interaction between HuR and these mRNA’s in 22Rv1 cells was examined using IP-RT-PCR (Section 2.19). Using this assay, CREBBP, PCAF, PAK2 and ELK1 mRNA’s were detected by RT- PCR in all samples prior to immunoprecipitation, and also among the fraction of mRNA’s that co-immunoprecipitated from 22Rv1 cell lysates with HuR antibody (Figure 6.10 and data not shown). These results indicated a close in vivo association of HuR with the CREBBP, PCAF, PAK2 and ELK1 transcripts. The transcripts were not detected in the absence of reverse transcriptase or in the absence of HuR secondary antibody ( i.e. IgG alone), providing evidence that the interaction was specific (Figure 6.10 and data not shown).
The down-regulated genes from this micro-array were also compared to the lists of reported HuR targets in Meisner et al . (2004) and others (Table 1.1;
182 ABCREBBP PAK2 0 0.5
-0.5 0 -1
-1.5 -0.5 Intensity (log2) Intensity
-2 (log2) Intensity
-2.5 -1 siCon siHuR siCon siHuR -UTR -UTR
C ELK1 D PCAF 0 0
-0.5 -0.5 -1 -1 -1.5 -1.5 -2 Intensity (log2) Intensity (log2) Intensity
-2.5 -2 siCon siHuR siCon siHuR -UTR -UTR
Figure 6.9. Relative expression levels of CREBBP, ELK1 PAK2 and PCAF. Log-transformed expression data was plotted for CREBBP ( A), ELK1 ( B), PAK2 ( C) and PCAF ( D) using GeneSifter. The value zero represents the mean intensity value of all genes in the micro-array, i.e . the mean gene expression level. Data represent average intensity ± sd for the gene indicated, from two independent micro-array experiments. siCon-treated cells: . siHuR-UTR-treated cells: . For abbreviations of gene names, see Table 6.3.
183 supt. beads HuR IgG HuR IgG 1 2 3 4 CREBBP
PCAF
PAK2
ELK1
Figure 6.10. Interaction of HuR with the CREBBP, PCAF, PAK2 and ELK1 mRNA’s. IP-RT-PCR was used to demonstrate binding of HuR to CREBBP, ELK1, PCAF, and PAK2 mRNA’s. RT-PCR was performed on 22Rv1 cell extracts before (supt.) and after immunoprecipitation (beads) using an antibody specific for HuR or IgG. These data represent a single experiment.
184 Table 6.1). From those lists, three were found in this micro-array to be down- regulated, nucleofibromin 1 (NF1), cationic amino-acid transporter (SLC7A1) and myristoylated alanine-rich protein kinase C substrate (MARCKS). None of the subset of genes reported by de Silanes et al. (2004) to be enriched after HuR co-immunoprecipitation were in identified in the present study as being down-regulated (Table 6.2).
6.3 Discussion. In this study, bioinformatic analyses identified the NNUUNNUUU sequence in the 3’UTR’s of 74 out of 78 transcripts (94 %) that have been previously reported to be bound by HuR (Table 1.2; Table 6.1)(Meisner et al. , 2004). The NUUNNUUU motif was also present in 88 % of the 3’UTR’s of transcripts reported to be enriched in an mRNA population that had been immunoprecipitated from RKO colorectal cells using an HuR antibody (de Silanes et al. , 2004) (Table 6.2). The presence of NNUUNNUUU sequence in such a high proportion of the 3’UTR’s reportedly bound by HuR, and the apparent failure of HuR to protect from RNase digestion the alternative motif proposed by de Silanes et al. (2004) suggested that the NNUUNNUUU motif, which was protected from cleavage, may be the real HuR consensus recognition motif. Interestingly, the mutations introduced into the ARUC region that disrupted HuD (Section 3.2.3) and HuR binding (Yeap et al. , 2002) were in the two NNUUNNUUU motifs in the AR (Figure 6.11).
The absence of the NNUUNNUUU motif from a small number of the 3’UTR’s of mRNA’s to which HuR has been reported to bind may be due to the fact that RNA-binding proteins may bind to sequences in addition to their highest- affinity site if the protein is in molar excess over the target mRNA, which could occur with a high-abundance protein such as HuR (Dreyfuss et al. , 2002). Therefore, transcripts such as the renin mRNA, which contains a single U C mismatch in the NNUUNNUUU motif (NNCUNNUUU) in the 3’UTR, may still be bound by HuR (Adams et al. , 2003; Morris et al. , 2004). It is noteworthy that a crystal structure of the first two RRM’s of HuD in complex with 11- nucleotide fragments of the c-fos and TNF α AURE’s indicated that a cytosine residue was well-tolerated in this position, and that the consensus recognition 185 NNUUNNUUU 4031 CUGGGCUUUUUUUUUCUCUUUCUCUC… ARUC WT …CUUUCUUUUUCUUCUUCCCUCCCUA 4082
4031 CUGGGCUUUUUUUUUCUCU GUCUCUC… ARUC MutA …CU GUCU GGG UCUUCUUCCCUCCCUA 4082
4031 CUGGGCUU GUGUGUUCUCU GUCUCUC… ARUC MutB …CU GUCU GGG UCUUCUUCCCUCCCUA 4082
Figure 6.11. Positions of point mutations in ARUC-MutA and ARUC-MutB sequences. Alignment of the ARUC WT and mutant sequences. U G point mutations are in pink .
186 for HuD binding to mRNA was likely to be an octomer, N(U/C)UNNUU(U/C), which is very similar to the motif proposed for HuR recognition by Meisner et al. (2004) (Wang and Tanaka-Hall, 2001). The study by Wang and colleagues (2001) also indicated that purines would not be well-tolerated within the region of RNA bound by HuD due to steric hindrance, whereas the alignment proposed by de Silanes et al. (2004) had a high proportion of guanines, which were not present in the alignment generated in the present study. Given the homology of the RRM’s of HuR and HuD, it seems likely that purines would also be poorly tolerated in the RNA motif bound by HuR.
The motif proposed by Meisner et al. (2004) appears in the 3’UTR’s of reported HuR target mRNA’s at ~ 4 times more frequently than is expected, assuming random occurrence of bases (Table 6.2). This provides support to the hypothesis that this motif is a functional non-random element. However, other mRNA sequence motifs such as the U-rich upstream element involved in polyA/cleavage site recognition may have contributed to the apparent frequency of NNUUNNUUU motifs detected in this screen (Section 1.4.1). Therefore, the frequency of the motif in 3’UTR’s reported to be bound by HuR needs to be compared to the frequency with which this motif appears in the general population of 3’UTR’s, to determine if the NNUUNNUUU motif is enriched in all 3’UTR’s or only in 3’UTR’s that are bound by HuR. In addition, it may also be useful to compare the frequency with which the motif occurs in coding regions with its frequency in 3’UTR’s, to determine if this sequence is likely to be involved in mRNA turnover, as elements that regulate mRNA decay are most frequently found in the 3’UTR’s of transcripts (Meyer et al. , 2004).
The presence of the NNUUNNUUU motif in the 3’UTR was used to select a subset of transcripts, previously identified by oligonucleotide micro-array analysis, to be investigated using qPCR and IP-RT-PCR. An initial subset was generated by limiting selection for follow-up analysis to transcripts which exhibited > 30 % down-regulation following treatment of the cells with siRNA directed against HuR. This criterion was intended to ensure that the decrease would be detectable by qPCR, which preliminary investigations indicated was subject to considerable sample-to-sample variation (data not shown). Only
187 down-regulated transcripts were included because HuR is thought to be a stabilising protein, thus a reduction in HuR expression was less likely to result in an increase in mRNA levels of direct HuR targets (Hollams et al. , 2002). Initial selection criteria also included the appearance of the gene product in a growth-associated pathway from which multiple members were down- regulated following HuR knockdown, in order to investigate the potential of HuR to co-ordinately regulate signalling pathways. mRNA's that encoded gene products involved in the regulation of the cell cycle or proliferation were also included in the subset because of the contribution of altered cell cycle regulation to tumorigenesis (Dash and El-Deiry, 2004). Finally, mRNA’s encoding known or putative AR co-regulators were included, because altered expression of AR co-regulators is thought to contribute to the development of androgen-independence in prostate cancer (Culig et al. , 2004).
The 3’UTR’s of the 19 “short-listed” genes were examined for the presence of the NNUUNNUUU sequence. Although a limited number of reported HuR targets contain U C mismatches in the NNUUNNUUU sequence, mRNA’s which contained such a mismatch were excluded from further analysis because they were considered less likely to be high-affinity targets of HuR than mRNA’s that contained perfect NNUUNNUUU motifs (Meisner et al. , 2004). The secondary structures of the 3’UTR’s that included a perfect NNUUNNUUU sequence were predicted using mfold, to determine whether the NNUUNNUUU sequence was likely to be single-stranded ( i.e. contained within a bulge or stem-loop structure), which is thought to be required for binding of HuR (Maris et al. , 2005; Meisner et al. , 2004; Zuker, 2003). mfold predicts the most stable secondary structure of an RNA sequence based on optimal base- pairing ( i.e. largest decrease in free energy) between ribonucleotides within the sequence, however the stability of these structures could be altered by the binding of proteins to the RNA. As mRNA’s always exist as part of mRNP complexes within the cell, the structures predicted by mfold may not represent the intra-cellular conformation of the mRNA’s. However, sequences that are persistently single-stranded ( i.e. bulges) across a range of free energies may never participate in base-pairing, due to the absence of suitable base-pairing partner sequences in the 3’UTR. In addition, many proteins involved in RNA
188 processing and degradation contain an RRM, KH domain or RGG box, and therefore are likely to recognise single-stranded RNA (Dreyfuss et al. , 2002; Maris et al. , 2005). This suggests that the mRNA contained within a mRNP is likely to be maintained in a single-stranded conformation, which may facilitate HuR binding, and also that some of the NNUUNNUUU motifs predicted to be double-stranded using mfold may be single-stranded in vivo.
Most of the transcripts identified as having a NNUUNNUUU motif in the 3’UTR were not included in the ARED2.0 database (Table 6.4), because ARED2.0 is a collection of genes that contain variations of the canonical AURE pentamer, AUUUA, whereas adenosines were not specified in the NNUUNNUUU sequence. Interestingly, the ARED2.0 classifications of AURE (III, IV and V) comply with the consensus of Meisner et al. The presence of canonical AURE’s in the 3’UTR’s of the CREBBP, ELK1 and GPC4 in ARED2.0 suggests that these transcripts may be targets of AURE-binding proteins other than HuR, such as AUF1 and TTP, although these interactions have not been reported.
Elimination of transcripts that did not contain the motif of Meisner et al. in a single-stranded conformation reduced the subset of potential HuR targets to 11 candidates. From these 11 candidate transcripts, ITGB1, PAK2, ELK1, CREBBP and PCAF were selected for confirmation by qPCR of down- regulation following HuR knockdown, because the proteins they encode have been reported to act directly in the same pathways (Chu et al. , 2005). In addition, these transcripts all had more than one single-stranded NNUUNNUUU motif, which may increase the likelihood that one of the motifs was available for HuR binding.
The single-stranded NNUUNNUUU sequences from the ITGB1, PAK2, ELK1, CREBBP, PCAF and AR 3’UTR’s were aligned manually and this alignment demonstrated that a low proportion of guanine residues were present in the regions flanking the NNUUNNUUU motifs. Moreover, only three of the sequences contained guanine residues within the motif itself ( eg. NGUUGNUUU), and these were all in the CREBBP 3’UTR. Conversely, the
189 base-frequency graph of the alignment of 3’UTR’s that were enriched after immunoprecipitation with HuR did contain a relatively high proportion of guanine residues, both in the flanking region and within the central U-rich region (de Silanes et al. , 2004). As previously discussed, guanine residues may not be well-tolerated within the sequence bound by the RRM’s of Hu proteins (Wang and Tanaka-Hall, 2001). This premise is also supported by the observation that introduction of U G point mutations into the NNUUNNUUU motif in the UC-rich region of the AR mRNA 3’UTR disrupted both HuD and HuR binding (Yeap et al. , 2002).
Several of the proposed HuR binding sites in the alignment in Figure 6.7A are missing the initial two bases (NN), because those residues were involved in base-pairing. This was permitted on the basis that the identity of these bases is not critical and as binding of HuR to stable higher order structures has been previously reported, HuR may be able to melt the secondary structure around the site upon binding, thereby freeing flanking nucleotides from base-pairing (Fialcowitz et al. , 2005).
The reduction of PCAF, CREBBP, PAK2 and ELK1 mRNA levels following treatment of the 22Rv1 cells with siRNA directed against HuR was confirmed by qPCR. As the quality of gene expression data generated by oligonucleotide micro-array analysis is highly dependent on the integrity of the RNA initially collected, it was important to validate the results by an independent method (Schroeder et al. , 2006). The decreases in mRNA expression levels of PCAF, CREBBP and PAK2 as determined by qPCR were smaller than those determined by micro-array analysis, perhaps due to the differences between the technologies used (amplification in PCR vs. probe hybridisation in oligonucleotide micro-array). Comparison of the mRNA expression levels of PCAF, CREBBP, PAK2 and ELK1 with the mean level of gene expression in 22Rv1 indicated that CREBBP, PAK2 and ELK1 exhibited lower-than-average expression levels. However, the mean gene expression level may have been over-estimated, as those genes with sufficiently low expression to be flagged as “absent” in by the GCOS software were excluded from analysis. The default settings on the GCOS software were used, and were conservative,
190 therefore some of the gene expression values that were excluded may have been genuine data points that would have lowered the mean value had they been included. Moreover, the relative expression of PCAF, CREBBP, PAK2 and ELK1 transcripts in control-siRNA treated 22Rv1 cells may not reflect the relative importance of the proteins that they encode.
Following the confirmation of down-regulation of expression of the PCAF, CREBBP, PAK2 and ELK1 mRNA’s, the in vivo association of these transcripts with HuR was investigated using IP-RT-PCR. All four of these mRNA’s could be detected using RT-PCR from immunoprecipitates generated using an HuR antibody. The co-immunoprecipitation of PCAF, CREBBP, PAK2 and ELK1 mRNA’s with HuR in IP-RT-PCR assays, together with their down-regulation following siHuR-UTR siRNA treatment and the presence of the HuR consensus motif in the 3’UTR’s, suggests that these transcripts may be genuine targets of HuR. Further experimentation, such as REMSA using 22Rv1 cell lysates, will be needed to confirm that the PCAF, CREBBP, PAK2 and ELK1 transcripts are bound by HuR. It will be of interest to map the HuR binding sites on these transcripts in the future, to determine if they are indeed the NNUUNNUUU sequence. This could be achieved using at least two approaches. RNAse footprinting assays with an extended stretch of the 3’UTR would be ideal, as this would give HuR the “option” of binding to a preferred site. Alternatively, REMSA assays could be performed with short sequences to determine the minimal site; however, it is likely that HuR would bind this U- rich sequence in vitro if presented in isolation, so this technique may have limited usefulness. In addition, determination of the effect of HuR knockdown on the mRNA half-lives and protein expression levels of ELK1, CREBBP, PAK2 and PCAF would be useful to elucidate whether HuR is involved in their post-transcriptional regulation.
The indication that HuR may bind the CREBBP mRNA is of particular interest in prostate cancer because the CREBBP is a transcriptional co-regulator that exhibits increased expression in tumour specimens sampled after androgen ablation (Comuzzi et al. , 2004). In addition, CREBBP enhances AR activity after stimulation by androgens and selectively enhances the agonistic action of the anti-androgen hydroxyflutamide (Comuzzi et al. , 2003; Fronsdal et al. ,
191 1998). Overexpression of either p300 or CREBBP enables activated protein kinase A (PKA) to stimulate PSA-promoter driven transcription in the absence of androgen, and this transcriptional activation is not inhibited by bicalutamide, suggesting that it is independent of the AR (Kim et al. , 2005). Interestingly, expression of CREBBP appears to be regulated by the AR, as CREBBP is down-regulated in LNCaP cells following treatment with IL6 or the synthetic androgen R1881, while CREBBP expression is unaffected by IL6 or R1881 treatments of the AR-negative cell line PC3 (Comuzzi et al. , 2004). As the micro-array analysis was performed in the presence of low levels of androgens (i.e. in charcoal-stripped FCS), the decreased expression of CREBBP mRNA dectected in the micro-array was unlikely to be due directly to AR-mediated repression, providing support to the results of the present study that the level of this transcript is regulated by HuR.
There is also evidence that PCAF, the mRNA of which was down-regulated following knockdown of HuR, is a co-regulator of AR. PCAF exhibits histone- acetyl transferase activity and acetylates AR, thereby decreasing the interaction of AR and the transcriptional repressor NCoR, and increasing DHT- dependent interactions with co-activators (Fu et al. , 2002). In addition, the histone acetyltransferase activity of PCAF rescues cyclin D1-mediated AR trans-repression (Reutens et al. , 2001). PCAF associates with CREBBP in vitro and in vivo , and CREBBP and PCAF regulate ERK1 gene expression at both the transcriptional and post-transcriptional level, indicating that these proteins may also influence MAPK signalling, another pathway that was affected by knockdown of HuR (Chu et al. , 2005).
ELK1 mRNA was decreased ~ 40 % (1.7-fold) following siHuR-UTR siRNA treatment of 22Rv1 cells. ELK1 is a transcription factor that is a nuclear target for all three MAPK signalling cascades as well as Wnt signalling. It is a member of the ternary complex factor (TCF) subfamily, and E twenty-six (ETS) superfamily of transcription factors (Yordy and Muise-Helmericks, 2000). TCF’s interact with CREBBP and p300 and when phosphorylated by MEK/ERK, and activate transcription of genes containing serum-response elements in their promoters, such as c-fos and TNF α (Buchwalter et al. , 2004). Trancriptional
192 activation by ELK1 is thought to be an important regulator of diverse functions such as neuronal differentiation, cell proliferation and tumorigenesis, and associates with the mitochondrial permeability transition pore complex (PTP), a structure involved in both apoptotic and necrotic cell death, although the function of this interaction has not been characterised (Barrett et al. , 2006). Transcriptional activation by ELK1 is also induced by mouse bombesin (gastrin-releasing peptide [GRP] in humans), a mitogen which is thought to be secreted by prostatic neurendocrine cells (Xiao et al. , 2002). Activation of ELK1 by bombesin results in a significant increase of cell proliferation in both PC3 and DU145 androgen-independent prostate cancer cell lines, suggesting that ELK1 expression may be of particular interest in prostate tumours that exhibit neuroendocrine foci, although widespread (rather than focal) increases in ELK1 immunoreactivity have been reported in prostate tumour specimens (Ricote et al. , 2006; Xiao et al. , 2002).
The expression of mRNA encoding PAK2 was observed to be significantly decreased in 22Rv1 cells in which HuR had been reduced. PAK2 belongs to a family of four p21-activated kinases, which are conserved non-receptor serine/threonine kinases that integrate a range of signalling pathways that originate from extra-cellular stimuli. A major function of PAK proteins is as effector molecules of the Rho GTPases that participate in focal adhesion signalling (Section 5.3) (Daniels and Bokoch, 1999; Kumar et al. , 2006). PAK’s are implicated in cytoskeletal remodelling, which is critical for cell adhesion, spreading, and motility. PAK kinase activity and subcellular distribution are tightly regulated by transient localised Rho GTPase activation, and by interactions mediated by Rho GTPase adapter proteins. In some breast cancer cell lines, PAK is constitutively activated, and mislocalised to atypical focal adhesions in the absence of activated Rho GTPases (Stofega et al. , 2004). It has been suggested that this abnormal PAK localisation and activity may contribute to tumorigenesis, although PAK mislocalisation has not been reported in prostate cancer (Stofega et al. , 2004). PAK2 is a negative regulator of the oncoprotein c-myc and, uniquely among PAK isoforms, has both pro- and anti-apoptotic functions (Huang et al. , 2004); (Kumar et al. , 2006). PAK2 is proteolytically cleaved and activated during apoptosis by
193 caspases and is proposed to be a pro-apoptotic factor in stress-induced cell death (Bokoch, 1998; Rudel and Bokoch, 1997). Conversely, activation of full- length PAK2 promotes cell survival and suppresses stress-induced cell death, but does not stimulate cell proliferation. This suppression of apoptosis is thought to be via phosphorylation and inhibition of BAD, the transcript levels of which were also decreased following HuR knockdown (Jakobi et al. , 2001). It may be interesting in future experiments to determine whether alterations in the transcript level of PAK2 influence activation of the full-length protein or PAK2 cleavage, and whether this contributes to apparent differences in growth or survival of 22Rv1 cells following knockdown of HuR.
In summary, the widespread identification of the NNUUNNUUU motif in reported targets of HuR provides evidence for the motif proposed by Meisner et al. (2004) being the HuR consensus binding site. This is further supported by the finding that four mRNA’s, chosen because of the presence of the proposed HuR consensus motif in the 3’UTR, co-immunoprecipitated with HuR in preliminary experiments. These initial findings also suggest that PCAF, CREBBP, PAK2 and ELK1 mRNA’s are novel targets of HuR, which may be confirmed in further analyses of the levels and localisation of the respective encoded proteins. In addition, the presence of the protein products of the PCAF, CREBBP, PAK2 and ELK1 mRNA’s in converging signalling pathways supports a hypothesis that HuR co-ordinately regulates such pathways.
194
CHAPTER 7: General Discussion.
195 7 Chapter 7: Discussion and Future
Directions.
7.1 General Discussion. The Hu proteins are a family of four RNA-binding proteins which exhibit high affinity for U-rich or AU-rich mRNA sequences, and which perform roles in mRNA processing, nuclear export and the regulation of mRNA stability (Antic and Keene, 1998; Katsanou et al. , 2005; Mazan-Mamczarz et al. , 2003; Zhu et al. , 2006). Both the ubiquitously expressed HuR and the neuronally-expressed HuD are thought to stabilise mRNA’s encoding a range of proteins that are important in development, regulation of the cell cycle, and in cancer. One of the transcripts to which HuR has been reported to bind is the AR mRNA, a receptor that is central to prostate cancer growth as well as an important therapeutic target utilised in the management of locally advanced and metastatic disease (Agoulnik and Weigel, 2006; Culig and Bartsch, 2006). HuD protein is expressed at very high levels in neuroblastomas and in small cell lung cancer, a cancer type that expresses neuroendocrine markers (Ball and King, 1997; Sekido et al. , 1994). This thesis has identified the expression of HuD in human prostate tumours, and investigated the potential role of HuR and HuD in modulating the expression and activity of the AR in prostate cancer cells. In addition, the effects of decreased HuR protein on global gene expression in prostate cancer cells were examined to identify HuR target genes.
Cancer is a result of multiple changes to regulatory systems controlling cell division, cell proliferation and programmed cell death. Although the initial triggers for malignant transformation of cells may be genetic mutations, additional alterations in gene expression and in regulatory pathways are also necessary for malignant cells to survive and proliferate. Mechanisms that contribute to uncontrolled cell proliferation include epigenetic alterations, ( e.g. DNA methylation) (Jaffe, 2003; Koeneman, 2006; Luczak and Jagodzinski, 2006; Rennie and Nelson, 1998), dysregulation of ubiquitin-mediated
196 proteolysis (Ciechanover and Schwartz, 2004; Guardavaccaro and Pagano, 2004), failure of cell cycle checkpoint-induced growth arrest and apoptosis (Nakanishi et al. , 2006; Zhivotovsky and Orrenius, 2006), loss of tumour suppressors such as p53 and the retinoblastoma protein (Cowell, 1990; Hickman et al. , 2002; Levine, 1997), unbalanced alternative splicing of mRNA’s (Venables, 2006), altered expression of micro-RNA’s (Dalmay and Edwards, 2006), constitutive or inappropriate activation of MAPK pathways (Bradham and McClay, 2006; Ricote et al. , 2006; Sebolt-Leopold, 2000), and changes in the stability of mRNA’s encoding signalling molecules (Audic and Hartley, 2004; Dibbens et al. , 1999; Hollams et al. , 2002; Ross et al. , 1991).
The regulation of mRNA stability and translation contributes significantly to the control of gene expression in both normal tissues and in cancer cells. In normal tissues, modulation of mRNA stability/translation is important for basal gene regulation as well as for differentiation and development (Antic and Keene, 1997; Audic and Hartley, 2004; Perrone-Bizzozero and Bolognani, 2002). The importance of post-transcriptional regulation in human disease is exemplified in the inherited fragile X mental retardation syndrome, which results from the genetic absence of fragile X mental retardation protein (FMRP), an RNA-binding protein that regulates the translation of key transcripts during brain development (Brown et al. , 2001; Vanderklish and Edelman, 2005). The syndrome is characterised by altered differentiation of neural stem cells, leading to impaired synaptic plasticity and spinal dysmorphogenesis, highlighting the critical role of mRNA processing during development. While inherited disorders involving Hu proteins have not been described, the potency of post-transcriptional regulation by molecules such as HuR and HuD is illustrated by the observations that overexpression of HuD and HuR can stimulate precocious neurite outgrowth in PC12 cells and myogenic differentiation of C2C12 myoblasts, respectively, due to stabilisation of the mRNA’s encoding key signalling molecules and transcription factors involved in neurogenic and myogenic differentiation (Akamatsu et al. , 1999; Aranda-Abreu et al. , 1999; Cherry et al. , 2006; Clayton et al. , 1998; Figueroa et al. , 2003; Mobarak et al. , 2000; Okano and Darnell, 1997; van der Giessen et al. , 2003). These studies demonstrate that co-ordinated post-transcriptional
197 regulation of genes is sufficient to bring about a unified cellular response with profound impacts on cell behaviour. If the post-transcriptional regulation of specific mRNA’s encoding proteins with major roles in cell cycle control or proliferation are perturbed, the result may contribute to pathological cell growth including cancer.
The increased total or cytoplasmic expression of HuR has been implicated in human colon, breast, ovarian and high grade brain tumours and in murine lung neoplasms (de Silanes et al. , 2003; Dixon et al. , 2001; Erkinheimo et al. , 2003; Heinonen et al. , 2005; Nabors et al. , 2001). In the present study it was found that HuR was widely expressed in both the normal prostate and in prostate cancer cells, although variations in its levels and/or intracellular localisation were not noted. While the contribution of HuR to the steady-state levels of its target genes is uncharacterised, increased stability of several cytokine and cyclin mRNA’s which are targets of HuR have been reported in tumours, as have elevated mRNA levels of c-fos , c-jun , and c-myc , which are thought to be targets of HuR and HuD (Audic and Hartley, 2004; Brewer, 1991; Chen et al. , 1994; Keyomarsi and Pardee, 1993; Peng et al. , 1996; Ross et al. , 1991). A range of transcripts encoding proteins that have altered expression in malignant cells have also been reported to be post-transcriptionally regulated by HuR, including β-catenin, COX2, p21 Waf1 , p53 and p27 Kip1 (de Silanes et al. , 2003; Galban et al. , 2003; Kullmann et al. , 2002; Lasa et al. , 2000; Lee and Jeong, 2006; Zou et al. , 2006). Although HuR is likely to function in collaboration with other regulators of transcription, mRNA stability and translation to determine levels of target gene transcripts, the widespread changes in gene expression following HuR knockdown in the present and in previous studies suggest that its involvement is significant (Brennan and Steitz, 2001; Cherry et al. , 2006; Deschenes-Furry et al. , 2005b; Figueroa et al. , 2003; van der Giessen et al. , 2003). As such, changes in the levels of multiple regulatory and signalling molecules evident in many cancer types may result from the activity of Hu proteins on their target mRNA’s.
The findings in this thesis that both HuR and HuD bind the AR mRNA and that HuR stabilises AR mRNA have identified novel mechanisms of AR regulation
198 that may be significant in prostate cancers. Consequences of the co- expression of HuR and HuD in prostate cancer cells and their binding to the ARUC sequence in the 3’UTR of the AR mRNA are unknown. Given the similarities between the RRM’s of HuR and HuD, it is likely that at least a proportion of the HuR target mRNA’s identified in this and in previous studies will also be bound by HuD (Akamatsu et al. , 1999; Kasashima et al. , 1999; Perrone-Bizzozero and Bolognani, 2002; Wakamatsu and Weston, 1997). However, the up-regulation of HuD in prostate cancers, in particular in high grade tumours and the likelihood of subsets of mRNA’s regulated by HuD and not HuR, suggest that gene expression profiles driven by the differential levels of Hu proteins in non-malignant prostate and in malignant prostate tumours of low and high grade delineate the behaviour of cells within these lesions.
Strategies that induce moderate decreases in the levels of specific mRNA’s may be useful as therapeutic interventions for cancer and for other proliferative disorders. The interactions between proteins such as HuR and their target mRNA’s could potentially be disrupted with the aim of destabilising the target mRNA’s. Several studies have shown that it is possible to inhibit the function of RNA-binding proteins using decoy RNA oligonucleotides that comprise the consensus binding site for the target protein. While many of these studies have focused on RNA-protein interactions during human immunodeficiency virus 1 infection and replication, two have focused on the interactions between mRNA’s and the αCP and HuR proteins (Bohjanen et al. , 1997; Konopka et al. , 2000; Makeyev et al. , 2002; Meisner et al. , 2004). These include demonstration of a significant and specific in vivo inhibition of αCP functions using an in vitro designed RNA decoy to saturate the KH domains of αCP1 and 2, and use of RNA oligonucleotides to alter the secondary structure of HuR target mRNA’s to modulate HuR binding and mimic IL2 activation in T- cells (Makeyev et al. , 2002; Meisner et al. , 2004). These investigations have provided proof of principle that RNA binding proteins represent valid therapeutic targets, and highlight the importance of clearly defining the HuR and HuD binding sites and their target mRNA’s.
199 Considerable efforts have been made in recent years to determine the nature of Hu binding sites, as well as those of other AURE-binding proteins (Brewer et al. , 2004; de Silanes et al. , 2004; Fialcowitz et al. , 2005; Lal et al. , 2004; Meisner et al. , 2004). These studies have indicated that the binding site for HuR is the U-rich single-stranded motif described as an AURE type III, a definition which includes the NNUUNNUUU motif proposed by Meisner et al. (2004). The AR mRNA ARUC element conforms to both the consensus sequence reported by Meisner et al (Meisner et al. , 2004) and the more degenerate sequence proposed by de Silanes et al (de Silanes et al. , 2004). Further studies of the structural basis of Hu/mRNA interactions will be needed to establish the requirements for high affinity binding such that strategies for specific disruption of particular mRNA/protein interactions can be developed. The potential to disrupt HuR/AR mRNA or HuD/AR mRNA interactions will be of particular interest in the context of prostate cancer, as suppression of AR protein expression and the activity of signalling pathways that exhibit crosstalk with the AR remain the major therapeutic goals of prostate cancer treatment.
The development of inhibitors for specific Hu/mRNA interactions will potentially be useful for targeted destabilisation of mRNA’s, however the potency of mRNA stabilising proteins appears to stem from their ability to regulate multiple mRNA’s in a way that can be stimulus-specific. This is analogous to the regulation of gene expression by transcription factors and co-factors, an analogy that carries with it the implication that mechanisms for both basal and induced regulation of mRNA stability exist. Therefore, one of the major questions that remains regarding post-transcriptional regulation of mRNA is the mechanism by which mRNA’s are sorted into subsets which are co-ordinately regulated in differentiation and development and in response to external stimuli (e.g. UV irradiation). While motifs within 3’ and 5’UTR’s undoubtedly play a role in the identification of individual mRNA’s as belonging to a particular subset, these motifs alone are probably not sufficient to explain the apparent co-ordinate post-transcriptional regulation of genes observed in myogenesis or in cellular responses to heat shock, for example.
200 AURE’s were initially thought be restricted to mRNA’s with high turnovers, such as growth factors, cytokines and oncogenes, and were therefore proposed to be a key determinant of mRNA half-life (Chen and Shyu, 1994, 1995). However, AURE’s are now understood to be present in transcripts that encode functionally diverse proteins involved in differentiation, immune responses, signal transduction, transcriptional and translational control, haematopoiesis, apoptosis, nutrient transport and metabolism (Khabar, 2005). The elements were detected in an estimated 8 % of human mRNA’s in this study, which was restricted to the AU-rich 3’UTR elements described in the ARED2.0 database. This database does not include U-rich elements that lack adenosines ( i.e. class III AURE’s), including the UC-rich motif in the AR mRNA 3’UTR to which HuR binds in vitro , therefore the estimate of 8 % is likely to be conservative. In addition, many of these transcripts do not have short half- lives, supporting the proposal that although the presence of an AURE may be necessary for regulated mRNA turnover, it is not the sole determinant of mRNA degradation or stabilisation under either basal conditions or in response to external stimuli.
The diversity of mRNA’s to which HuR has been reported to bind and the spectrum of mRNA’s that exhibited decreased expression in 22Rv1 cells following treatment with HuR siRNA in this thesis provide support to the hypothesis that HuR is involved in basal gene regulation as well as induced regulation of specific subsets of genes, a role consistent with its ubiquitous tissue distribution (Audic and Hartley, 2004; Fan and Steitz, 1998a; Gorospe, 2003; Ma et al. , 1996; Meisner et al. , 2004). In addition to the Hu proteins themselves, proteins that interact with AURE-binding proteins and modify their binding affinities or sub-cellular localisation will also play major roles in the directed (induced) regulation of mRNA turnover. Several of these proteins have been identified, for example CARM1, which methylates both HuD and HuR and is thought to inhibit their stabilising activity (Fujiwara et al. , 2006; Li et al. , 2002b). In addition, as RNP complexes are very large and contain diverse collections of RNA-binding proteins, it is feasible that the combination of RNA- binding proteins associated with a transcript also serves as an identification tag designating transcripts for stabilisation, turnover or sequestration.
201 Therefore, a major challenge in the understanding of post-transcriptional regulation will be to identify the additional mRNA cis -elements, proteins and protein combinations that definitively assign mRNA fate.
In the context of prostate cancer, it remains to be determined whether the stabilisation of mRNA’s by HuR and HuD relates to basal or induced transcript regulation. Characterisation of the contribution of these factors to the regulation of AR levels in the presence and absence of ligand and in response to current prostate cancer treatments is also highly relevant. Studies demonstrating co-regulation of AR and HuR following treatment of cells with DHT or selenium which result in increased stability of the AR mRNA and increased cytoplasmic abundance of HuR have provided evidence of involvement of HuR in the androgen responsiveness of cells (Chun et al. , 2006; Sheflin et al. , 2001; Yeap et al. , 1999). However, these mechanisms have not yet been investigated in human prostate tumours. In conclusion, the in vitro studies performed in this thesis have indicated that HuR and HuD contribute to the regulation of the AR and other important pathways in prostate cancer. Whether this is a consequence of changes to basal or directed gene regulation by HuR and HuD, or whether cancer cells develop the ability to “hijack” the post-transcriptional regulation of subsets of beneficial genes is at present unknown but will form part of the future characterisation of this important family of mRNA regulatory proteins.
7.2 Future Directions. HuD protein expression was detected in a cohort of 12 primary prostate tumour specimens, and specimens of higher Gleason grade exhibited stronger HuD immunostaining. HuD expression has not been reported previously in prostate cancer, therefore this finding is novel and should be investigated further using large cohorts of both primary and metastatic prostate tumours. As HuD expression appeared to be correlated with Gleason grade, it will be interesting to investigate whether this association is maintained in larger, well-defined cohorts of specimens. The relationship between HuD expression and expression of other predictive and prognostic markers, in particular AR, p27 Kip1 and other HuD targets will also be useful as strong correlation between these 202 markers would provide evidence supporting HuD regulation of the steady-state levels of these transcripts/proteins.
Although no cancer-specific changes in HuR levels were detected by immunostaining of prostate specimens in this study, only 12 specimens could be examined. Analysis of larger numbers of primary and metastatic prostate tumours will be required to verify these results and may reveal subsets of prostate tumours with aberrant HuR protein levels and/or localisation. In addition to potential associations with other prostate cancer biomarkers, relationships between expression of Hu proteins and clinical aspects of disease including tumour size and stage, responses to treatments and survival could also be examined.
Previous studies have reported four isoforms of HuD that arise from differential mRNA splicing (Behrends et al. , 2002). Expression of the different HuD isoforms in prostate cancers and association of expression of specific isoforms with clinical aspects of the tumours or disease outcomes could be examined. These studies would require the development of isoform-specific antibodies to be carried out using immunohistochemistry, alternatively, an RNA-based approach ( eg. qPCR) could be used. Given the unusual expression of HuD in prostate cancer cells, preliminary analyses of HuB and HuC mRNA expression in prostate tumours could be undertaken, to determine whether other neuronal Hu proteins are also aberrantly expressed in prostate cancer cells. HuB- and HuC-specific antibodies had not been generated at the time of writing this thesis, but could be used in later studies to augment mRNA procedures.
A significant finding in this thesis was the co-expression of high levels of HuR and HuD in human prostate tumours. Both HuR and HuD appear to bind the same UC-rich region of the AR mRNA, suggesting that the aberrant expression of HuD in prostate cancer cells may contribute to dysregulation of the AR mRNA and other transcripts which contain a similar U-rich sequence in the 3’UTR. The relative affinities of HuR and HuD for U-rich sequences such as the ARUC are uncharacterised and therefore it is not known whether HuD competes with HuR for binding to this sequence. The relative binding affinities
203 of HuR and HuD could be assessed using in vitro assays such as REMSA, although it would also be useful to identify and compare by micro-array the mRNA’s bound to each of HuR and HuD following specific immunoprecipitation from LNCaP or 22Rv1 cell lysates (de Silanes et al. , 2004). In addition, the proteins with which HuR and HuD interact need to be identified, as these are likely to be different, given that the hinge regions of HuR and HuD are dissimilar and have been reported to confer differing abilities to stabilise neuronal mRNA’s (Kasashima et al. , 1999). Protein partners of HuD and HuR could be also identified using yeast two-hybrid screens in which the full-length proteins or the hinge regions of HuD or HuR were used as the bait epitope. Expression of HuR/HuD interacting proteins could be examined in prostate cancer specimens, to determine their usefulness as predictive or prognostic indicators, alone or in combination with HuR/HuD. The functional activity of the target proteins and their effects on HuR/HuD activity could also be evaluated in vitro following over-expression or knockdown of either or both HuD/HuR and individual interacting proteins.
HuD binds the ARUC region of the AR mRNA in vitro and the AR mRNA co- immunoprecipitates with HuD-specific antibodies, however the effect of increased HuD expression on AR mRNA stability was not assessed and only a modest effect on AR protein expression was shown. In order to more fully elucidate the functional effects of HuD expression in prostate cancer, it may be useful to generate a stable, inducible HuD over-expressing LNCaP cell line. As endogenous HuD protein expression appears to be very low in LNCaP cells, this approach is likely to be more informative than siRNA-mediated knockdown of HuD protein expression. The Tet-On Advanced Inducible Gene Expression System (ClonTech) could be used to induce high levels of HuD expression in LNCaP or 22Rv1 cells to determine the effects of long-term HuD expression on AR mRNA stability, as well as AR protein expression and secretion of PSA (Urlinger et al. , 2000). This is of interest because it may allow assessment of the effects of HuD expression on androgen signalling in prostate cancer, which is one of the main pathways promoting proliferation, both in early and advanced prostate tumours.
204 The effects of HuD over-expression on global mRNA levels in 22Rv1 cells could also be assessed using an oligonucleotide micro-array approach similar to the one used in this thesis, to identify potential targets of HuD in prostate cancer cells. A comparison could then be made between the mRNA’s identified as having significantly altered expression following induction of HuD expression with those that were altered following reduction of HuR expression, to determine whether the two Hu proteins bind a similar set of transcripts, and to identify common and distinct targets of HuD and HuR in prostate cancer. This information and subsequent functional studies investigating specific HuD- regulated genes could then be used to estimate the likely impact of HuD up- regulation in human prostate tumours.
Micro-array analysis of gene expression in 22Rv1 cells in which HuR had been reduced with siRNA revealed a number of pathways that contained multiple down-regulated genes. To determine whether these changes were a consequence of multiple mRNA’s being directly destabilised by loss of HuR expression, it will be useful to first ascertain which of the down-regulated genes in each pathway contain HuR consensus binding sites. The presence of an HuR binding site has been identified within the 3’UTR of PAK2 and ELK1 mRNA’s, which encode a kinase and a transcription factor, respectively, in the MAPK pathway, as well as within the PCAF and CREBBP mRNA’s, which encode transcriptional co-activators that are involved in a number of transcriptional responses, including androgen signalling. However, the presence of HuR binding motifs was not investigated in other potentially important genes including those involved in regulation of the cell cycle, proliferation, apoptosis and ubiquitin-mediated proteolysis. HuR regulation of these genes could be investigated by qPCR following treatment of cells with HuR siRNA and changes in protein levels determined using western blotting. As multiple genes within pathways were found to be regulated following HuR siRNA treatment, functional assays may be informative. These could include proliferation studies, and assays of MAPK activity or ubiquitin-mediated protein degradation.
205 To identify a set of genes that are regulated by HuR specifically in the prostate, HuR could be knocked down in the androgen-dependent cell line LNCaP and the androgen-independent cell line PC3 as well as in a non-prostate cell line such as HeLa, and oligonucleotide micro-arrays performed. The subset of down-regulated genes that are confirmed by qPCR, contain an HuR consensus binding site and are common to 22Rv1, LNCaP and PC3 cells but not HeLa cells may be prostate-specific targets of HuR. Those genes commonly down-regulated in 22Rv1 and LNCaP cell lines may be HuR targets specific to androgen-responsive prostate cells, although this would have to be interpreted with caution as 22Rv1 cells are androgen-responsive, but not androgen-dependent for growth.
Although the experiments described above could be performed using siRNA- mediated knockdown of HuR, stable transfection of 22Rv1 or LNCaP cells with an inducible or constitutive HuR anti-sense vector would allow elucidation of the long-term effects of decreased HuR, which may better represent gene regulation changes in tumours in vivo. In particular, the injection of stably- under-expressing HuR prostate cancer cell lines into nude mice may allow investigation of the long-term effects of HuR reduction not only on tumour growth, but also on androgen and anti-androgen responsiveness. The response of 22Rv1 or LNCaP cell lines that stably under-express HuR to mitotic stimuli could also be investigated, given the suggestion that the MAPK and focal adhesion signalling pathways appeared to be down-regulated following transient knockdown of HuR. Stable knockdown of HuR would also facilitate examination of the effects of HuR expression on AR mRNA stability in the presence and absence of DHT, as well as the long-term effects on AR expression and PSA secretion, which are clinically useful prostate cancer biomarkers.
The AR mRNA is likely to be bound by protein regulators of mRNA stability other than HuR and HuD. αCP1 and αCP2 have been reported to bind the ARUC region of the AR 3’UTR, and although the functional effects of this binding have not been established, αCP1 and 2 are considered to promote mRNA stability (Makeyev and Liebhaber, 2002; Yeap et al. , 2002). The
206 binding of the destabilising proteins AUF1, BRF1 and TTP to the ARUC region has not been assessed, however AUF1 has an affinity for AU-rich mRNA (AURE’s of type I and II), and therefore may not bind to the UC-rich ARUC region (Blaxall et al. , 2002; Fialcowitz et al. , 2005). The requirements for binding of TTP and BRF1 are less well-defined, and therefore their relative affinity for the ARUC could be tested using REMSA “supershift” assays involving the ARUC-WT probe and LNCaP or 22Rv1 cell lysates with or without BRF1-specific or TTP-specific antibodies. Competition assays between HuR and BRF1/TTP could then be used to determine whether these proteins are likely to have contributed to the decrease in AR mRNA stability observed in 22Rv1 cells following knockdown of HuR by siRNA treatment. If BRF1 or TTP do possess affinity for the ARUC region, it will be interesting to assess the expression of these proteins in prostate tumour samples, as alterations in the levels of mRNA destabilising proteins may have as significant an impact on gene regulation as alterations in the expression of proteins such as HuR and HuD.
7.3 Summary.
Modulation of mRNA stability is an important contributor to regulation of gene expression, with significant impacts on normal cell behaviour (neurite outgrowth, myogenesis) as well as purported roles in the onset, maintenance, and progression of cancers. Regulation of mRNA stability is predicted to be important in the context of prostate cancer, given that a key therapeutic target in prostate cancer, the AR, in known to be post-transcriptionally regulated (Yeap et al. , 1999). This thesis has investigated the expression and function of two RNA-binding proteins, HuD and HuR, in prostate cancer cells. Findings of the aberrant up-regulation of HuD expression in human prostate tumours and the binding of HuD and HuR to the AR mRNA have provided evidence that these Hu proteins contribute to the androgen (and potentially the anti- androgen) responsiveness of prostate cancer cells. Isolation and functional description of candidate HuR-regulated genes has suggested that HuR contributes significantly to the proliferation and biological activity of prostate cancer cells via regulation of genes in multiple pathways. These results underscore the importance of post-transcriptional regulation in prostate cancer
207 and indicate the potential for targeting HuD and/or HuR in the management of the disease.
208 8 References.
Adams, D.J., Beveridge, D.J., van der Weyden, L., Mangs, H., Leedman, P.J., and Morris, B.J. (2003) HADHB, HuR, and CP1 bind to the distal 3'-untranslated region of human renin mRNA and differentially modulate renin expression. Journal of Biological Chemistry. 278 : 44894-44903. Agoulnik, I.U., and Weigel, N.L. (2006) Androgen receptor action in hormone- dependent and recurrent prostate cancer. Journal of Cellular Biochemistry 99 : 362-372. Ahlegren, G., Pedersen, K., Lundberg, S., Aus, G., Hugosson, J., and Abrahamsson, P. (2000) Neuroendocrine differentiation is not prognostic of failure after radical prostatectomy but correlates with tumor volume. Urology 56 : 1011- 1015. Akamatsu, W., Okano, H.J., Osumi, N., Inoue, T., Nakamura, S., Sakakibara, S., Miura, M., Matsuo, N., Darnell, R.B., and Okano, H. (1999) Mammalian ELAV- like neuronal RNA-binding proteins HuB and HuC promote neuronal development in both the central and the peripheral nervous systems. Proceedings of the National Academy of Sciences of the United States of America. 96 : 9885-9890. Alimirah, F., Chen, J., Basrawala, Z., Xin, H., and Choubey, D. (2006) DU-145 and PC-3 human prostate cancer cell lines express androgen receptor: implications for the androgen receptor functions and regulation. FEBS Letters 580 : 2294-2300. Anderson, P., and Kedersha, N. (2002) Visibly stressed: the role of eIF2, TIA-1, and stress granules in protein translation. Cell Stress Chaperones. 7: 213-221. Antic, D., and Keene, J.D. (1997) Embryonic lethal abnormal visual RNA-binding proteins involved in growth, differentiation, and posttranscriptional gene expression. American Journal of Human Genetics 61 : 273-278. Antic, D., and Keene, J.D. (1998) Messenger ribonucleoprotein complexes containing human ELAV proteins: interactions with cytoskeleton and translational apparatus. Journal of Cell Science 111 : 183-197. Antic, D., Lu, N., and Keene, J.D. (1999) ELAV tumor antigen, Hel-N1, increases translation of neurofilament M mRNA and induces formation of neurites in human teratocarcinoma cells. Genes & Development 13 : 449-461. Aranda-Abreu, G.E., Behar, L., Chung, S., Furneaux, H., and Ginzburg, I. (1999) Embryonic lethal abnormal vision-like RNA-binding proteins regulate neurite outgrowth and tau expression in PC12 cells. Journal of Neuroscience 19 : 6907-6917. Arao, Y., Kuriyama, R., Kayama, F., and Kato, S. (2000) A nuclear matrix-associated factor, SAF-B, interacts with specific isoforms of AUF1/hnRNP D. Archives of Biochemistry & Biophysics 380 : 228-236. Atasoy, U., Watson, J., Patel, D., and Keene, J.D. (1998) ELAV protein HuA (HuR) can redistribute between nucleus and cytoplasm and is upregulated during serum stimulation and T cell activation. Journal of Cell Science 111 : 3145- 3156. Attardi, B.J., Burgenson, J., Hild, S.A., and Reel, J.R. (2004) Steroid hormonal regulation of growth, prostate specific antigen secretion, and transcription mediated by the mutated androgen receptor in CWR22Rv1 human prostate carcinoma cells. Molecular & Cellular Endocrinology 222 : 121-132. Audic, Y., and Hartley, R.S. (2004) Post-transcriptional regulation in cancer. Biology of the Cell. 96 : 479-498. Baez, M.V., and Boccaccio, G.L. (2005) Mammalian Smaug is a translational repressor that forms cytoplasmic foci similar to stress granules. Journal of Biological Chemistry 280 : 43131-43140.
209 Bakheet, T., Williams, B.R.G., and Khabar, K.S.A. (2003) ARED 2.0: an update of AU- rich element mRNA database. Nucleic Acids Research. 31 : 421-423. Bakin, R.E., Gioeli, D., Sikes, R.A., Bissonette, E.A., and Weber, M.J. (2003) Constitutive activation of the Ras/mitogen-activated protein kinase signaling pathway promotes androgen hypersensitivity in LNCaP prostate cancer cells. Cancer Research 63 : 1981-1989. Ball, N.S., and King, P.H. (1997) Neuron-specific hel-N1 and HuD as novel molecular markers of neuroblastoma: a correlation of HuD messenger RNA levels with favorable prognostic features. Clinical Cancer Research 3: 1859-1865. Barreau, C., Paillard, L., and Osborne, H.B. (2005) AU-rich elements and associated factors: are there unifying principles? Nucleic Acids Research 33 : 7138-7150. Barrett, L.E., Van Bockstaele, E.J., Sul, J.Y., Takano, H., Haydon, P.G., and Eberwine, J.H. (2006) Elk-1 associates with the mitochondrial permeability transition pore complex in neurons. Proceedings of the National Academy of Sciences of the United States of America 103 : 5155-5160. Beckel-Mitchener, A.C., Miera, A., Keller, R., and Perrone-Bizzozero, N.I. (2002) Poly(A) tail length-dependent stabilization of GAP-43 mRNA by the RNA- binding protein HuD. Journal of Biological Chemistry. 277 : 27996-28002. Behrends, U., Jandl, T., Golbeck, A., Lechner, B., Muller-Weihrich, S., Schmid, I., Till, H., Berthold, F., Voltz, R., and Mautner, J.M. (2002) Novel products of the HUD, HUC, NNP-1 and alpha-internexin genes identified by autologous antibody screening of a pediatric neuroblastoma library. International Journal of Cancer 100 : 669-677. Berruti, A., Mosca, A., Tucci, M., Terrone, C., Torta, M., Tarabuzzi, R., Russo, L., Cracco, C., Bollito, E., Scarpa, R.M., Angeli, A., and Dogliotti, L. (2005) Independent prognostic role of circulating chromogranin A in prostate cancer patients with hormone-refractory disease. Endocrine Related Cancer 12 : 109- 117. Bertram, M.J., Berube, N.G., Hang-Swanson, X., Ran, Q., Leung, J.K., Bryce, S., Spurgers, K., Bick, R.J., Baldini, A., Ning, Y., Clark, L.J., Parkinson, E.K., Barrett, J.C., Smith, J.R., and Pereira-Smith, O.M. (1999) Identification of a gene that reverses the immortal phenotype of a subset of cells and is a member of a novel family of transcription factor-like genes. Molecular & Cellular Biology 19 : 1479-1485. Bevan, C.L., Hoare, S., Claessens, F., Heery, D.M., and Parker, M.G. (1999) The AF1 and AF2 domains of the androgen receptor interact with distinct regions of SRC1. Molecular & Cellular Biology 19 : 8383-8392. Blackshear, P.J. (2002) Tristetraprolin and other CCCH tandem zinc-finger proteins in the regulation of mRNA turnover. Biochemical Society Transactions 30 : 945- 952. Blaxall, B.C., Dwyer-Nield, L.D., Bauer, A.K., Bohlmeyer, T.J., Malkinson, A.M., and Port, J.D. (2000) Differential expression and localization of the mRNA binding proteins, AU-rich element mRNA binding protein (AUF1) and Hu antigen R (HuR), in neoplastic lung tissue. Molecular Carcinogenesis 28 : 76-83. Blaxall, B.C., Pende, A., Wu, S.C., and Port, J.D. (2002) Correlation between intrinsic mRNA stability and the affinity of AUF1 (hnRNP D) and HuR for A+U-rich mRNAs. Molecular & Cellular Biochemistry 232 : 1-11. Blok, L.J., de Ruiter, P.E., and Brinkmann, A.O. (1998) Forskolin-induced dephosphorylation of the androgen receptor impairs ligand binding. Biochemistry 37 : 3850-3857. Bluyssen, H.A., van Os, R.I., Naus, N.C., Jaspers, I., Hoeijmakers, J.H., and de Klein, A. (1998) A human and mouse homolog of the Schizosaccharomyces pombe rad1+ cell cycle checkpoint control gene. Genomics. 54 : 331-337. Bohjanen, P.R., Liu, Y., and Garcia-Blanco, M.A. (1997) TAR RNA decoys inhibit tat- activated HIV-1 transcription after preinitiation complex formation. Nucleic Acids Research 25 : 4481-4486.
210 Bokoch, G.M. (1998) Caspase-mediated activation of PAK2 during apoptosis: proteolytic kinase activation as a general mechanism of apoptotic signal transduction? Cell Death & Differentiation 5: 637-645. Bradham, C., and McClay, D.R. (2006) p38 MAPK in development and cancer. Cell Cycle. 5: 824-828. Brakebusch, C., Wennerberg, K., Krell, H.W., Weidle, U.H., Sallmyr, A., Johansson, S., and Fassler, R. (1999) Beta1 integrin promotes but is not essential for metastasis of ras-myc transformed fibroblasts. Oncogene. 18 : 3852-3861. Brakebusch, C., and Fassler, R. (2005) Beta 1 integrin function in vivo: Adhesion, migration and more. Cancer and Metastasis Reviews 24 : 403-411. Bratt, O., Borg, A., Kristoffersson, U., Lundgren, R., Zhang, Q.X., and Olsson, H. (1999) CAG repeat length in the androgen receptor gene is related to age at diagnosis of prostate cancer and response to endocrine therapy, but not to prostate cancer risk. British Journal of Cancer 81 : 672-676. Brennan, C.M., Gallouzi, I.E., and Steitz, J.A. (2000) Protein ligands to HuR modulate its interaction with target mRNAs in vivo. Journal of Cell Biology 151 : 1-14. Brennan, C.M., and Steitz, J.A. (2001) HuR and mRNA stability. Cellular & Molecular Life Sciences. 58 : 266-277. Brewer, B.Y., Malicka, J., Blackshear, P.J., and Wilson, G.M. (2004) RNA sequence elements required for high affinity binding by the zinc finger domain of tristetraprolin: conformational changes coupled to the bipartite nature of Au- rich MRNA-destabilizing motifs. Journal of Biological Chemistry 279 : 27870- 27877. Brewer, G. (1991) An A + U-rich element RNA-binding factor regulates c-myc mRNA stability in vitro. Moecular and Cellular Biology. 11 : 2460-2466. Briata, P., Forcales, S.V., Ponassi, M., Corte, G., Chen, C.-Y., Karin, M., Puri, P.L., and Gherzi, R. (2005) p38-Dependent Phosphorylation of the mRNA Decay- Promoting Factor KSRP Controls the Stability of Select Myogenic Transcripts. Molecular Cell 20 : 891-903. Brown, V., Jin, P., Ceman, S., Darnell, J.C., O'Donnell, W.T., Tenenbaum, S.A., Jin, X., Feng, Y., Wilkinson, K.D., Keene, J.D., Darnell, R.B., and Warren, S.T. (2001) Microarray identification of FMRP-associated brain mRNAs and altered mRNA translational profiles in fragile X syndrome. Cell 107 : 477-487. Buchwalter, G., Gross, C., and Wasylyk, B. (2004) Ets ternary complex transcription factors. Gene 324 : 1-14. Burd, C.J., Petre, C.E., Morey, L.M., Wang, Y., Revelo, M.P., Haiman, C.A., Lu, S., Fenoglio-Preiser, C.M., Li, J., Knudsen, E.S., Wong, J., and Knudsen, K.E. (2006) Cyclin D1b variant influences prostate cancer growth through aberrant androgen receptor regulation. Proceedings of the National Academy of Sciences of the United States of America 103 : 2190-2195. Callewaert, L., Christiaens, V., Haelens, A., Verrijdt, G., Verhoeven, G., and Claessens, F. (2003) Implications of a polyglutamine tract in the function of the human androgen receptor. Biochemical & Biophysical Research Communications. 306 : 46-52. Cao, X., Qin, J., Xie, Y., Khan, O., Dowd, F., Scofield, M., Lin, M.F., and Tu, Y. (2006) Regulator of G-protein signaling 2 (RGS2) inhibits androgen-independent activation of androgen receptor in prostate cancer cells. Oncogene. 25 : 3719- 3734. Carballo, E., Gilkeson, G.S., and Blackshear, P.J. (1997) Bone marrow transplantation reproduces the tristetraprolin-deficiency syndrome in recombination activating gene-2 (-/-) mice. Evidence that monocyte/macrophage progenitors may be responsible for TNFalpha overproduction. Journal of Clinical Investigation 100 : 986-995. Catalona, W.J., and Loeb, S. (2005) The PSA era is not over for prostate cancer. European Urology 48 : 541-545.
211 Chattopadhyay, A., Chiang, C.W., and Yang, E. (2001) BAD/BCL-[X(L)] heterodimerization leads to bypass of G0/G1 arrest. Oncogene. 20 : 4507- 4518. Chen, C.D., Welsbie, D.S., Tran, C., Baek, S.H., Chen, R., Vessella, R., Rosenfeld, M.G., and Sawyers, C.L. (2004) Molecular determinants of resistance to antiandrogen therapy. Nature Medicine 10 : 33-39. Chen, C.Y., Chen, T.M., and Shyu, A.B. (1994) Interplay of two functionally and structurally distinct domains of the c-fos AU-rich element specifies its mRNA- destabilizing function. Molecular & Cellular Biology 14 : 416-426. Chen, C.Y., and Shyu, A.B. (1994) Selective degradation of early-response-gene mRNAs: functional analyses of sequence features of the AU-rich elements. Molecular & Cellular Biology 14 : 8471-8482. Chen, C.Y., and Shyu, A.B. (1995) AU-rich elements: characterization and importance in mRNA degradation. Trends in Biochemical Sciences 20 : 465-470. Chen, C.Y., Gherzi, R., Ong, S.E., Chan, E.L., Raijmakers, R., Pruijn, G.J., Stoecklin, G., Moroni, C., Mann, M., and Karin, M. (2001) AU binding proteins recruit the exosome to degrade ARE-containing mRNAs. Cell 107 : 451-464. Chen, C.-Y.A., Xu, N., and Shyu, A.-B. (2002) Highly Selective Actions of HuR in Antagonizing AU-Rich Element-Mediated mRNA Destabilization. Molecular and Cellular Biology. 22 : 7268-7278. Chen, Y., and Varani, G. (2005) Protein families and RNA recognition. FEBS Journal 272 : 2088-2097. Cheng, L., Sun, J., Pretlow, T.G., Culp, J., and Yang, N.S. (1996) CWR22 xenograft as an ex vivo human tumor model for prostate cancer gene therapy. Journal of the National Cancer Institute 88 : 607-611. Cherry, J., Karschner, V., Jones, H., and Pekala, P.H. (2006) HuR, an RNA-binding protein, involved in the control of cellular differentiation. In Vivo 20 : 17-23. Cheung, P.K., Woolcock, B., Adomat, H., Sutcliffe, M., Bainbridge, T.C., Jones, E.C., Webber, D., Kinahan, T., Sadar, M., Gleave, M.E., and Vielkind, J. (2004) Protein profiling of microdissected prostate tissue links growth differentiation factor 15 to prostate carcinogenesis. Cancer Research 64 : 5929-5933. Chkheidze, A.N., Lyakhov, D.L., Makeyev, A.V., Morales, J., Kong, J., and Liebhaber, S.A. (1999) Assembly of the alpha -Globin mRNA Stability Complex Reflects Binary Interaction between the Pyrimidine-Rich 3' Untranslated Region Determinant and Poly(C) Binding Protein alpha CP. Molecular and Cellular Biology. 19 : 4572-4581. Chlenski, A., Nakashiro, K., Ketels, K.V., Korovaitseva, G.I., and Oyasu, R. (2001) Androgen receptor expression in androgen-independent prostate cancer cell lines. The Prostate 47 : 66-75. Cho, N.H., Kang, S., Hong, S., An, H.J., Choi, Y.H., Jeong, G.B., and Choi, H.K. (2006) Elevation of cyclin B1, active cdc2, and HuR in cervical neoplasia with human papillomavirus type 18 infection. Cancer Letters 232 : 170-178. Chu, B.Y., Tran, K., Ku, T.K., and Crowe, D.L. (2005) Regulation of ERK1 gene expression by coactivator proteins. Biochem Journal. 392 : 589-599. Chun, J.Y., Nadiminty, N., Lee, S.O., Onate, S.A., Lou, W., and Gao, A.C. (2006) Mechanisms of selenium down-regulation of androgen receptor signaling in prostate cancer. Molecular Cancer Therapeutics 5: 913-918. Chung, H.J., Liu, J., Dundr, M., Nie, Z., Sanford, S., and Levens, D. (2006) FBPs are calibrated molecular tools to adjust gene expression. Molecular and Cellular Biology 26 : 6584-6597. Chung, S., Jiang, L., Cheng, S., and Furneaux, H. (1996) Purification and properties of HuD, a neuronal RNA-binding protein. Journal of Biological Chemistry 271 : 11518-11524. Ciechanover, A., and Schwartz, A.L. (2004) The ubiquitin system: pathogenesis of human diseases and drug targeting. Biochimica et Biophysica Acta (BBA) - Molecular Cell Research
212 The Ubiquitin-Proteasome System 1695 : 3-17. Clayton, G.H., Perez, G.M., Smith, R.L., and Owens, G.C. (1998) Expression of mRNA for the elav-like neural-specific RNA binding protein, HuD, during nervous system development. Developmental Brain Research 109 : 271-280. Cohen, R.J., Glezerson, G., and Haffejee, Z. (1991) Neuro-endocrine cells--a new prognostic parameter in prostate cancer. British Journal of Urology 68 : 258- 262. Cok, S.J., Acton, S.J., and Morrison, A.R. (2003) The proximal region of the 3'- untranslated region of cyclooxygenase-2 is recognized by a multimeric protein complex containing HuR, TIA-1, TIAR, and the heterogeneous nuclear ribonucleoprotein U. Journal of Biological Chemistry 278 : 36157-36162. Comuzzi, B., Lambrinidis, L., Rogatsch, H., Godoy-Tundidor, S., Knezevic, N., Krhen, I., Marekovic, Z., Bartsch, G., Klocker, H., Hobisch, A., and Culig, Z. (2003) The transcriptional co-activator cAMP response element-binding protein- binding protein is expressed in prostate cancer and enhances androgen- and anti-androgen-induced androgen receptor function. American Journal of Pathology 162 : 233-241. Comuzzi, B., Nemes, C., Schmidt, S., Jasarevic, Z., Lodde, M., Pycha, A., Bartsch, G., Offner, F., Culig, Z., and Hobisch, A. (2004) The androgen receptor co- activator CBP is up-regulated following androgen withdrawal and is highly expressed in advanced prostate cancer. Journal of Pathology 204 : 159-166. Correa-Cerro, L., Wohr, G., Haussler, J., Berthon, P., Drelon, E., Mangin, P., Fournier, G., Cussenot, O., Kraus, P., Just, W., Paiss, T., Cantu, J.M., and Vogel, W. (1999) (CAG)nCAA and GGN repeats in the human androgen receptor gene are not associated with prostate cancer in a French-German population. European Journal of Human Genetics 7: 357-362. Cougot, N., van Dijk, E., Babajko, S., and Seraphin, B. (2004) 'Cap-tabolism'. Trends in Biochemical Sciences 29 : 436-444. Coulson, R.M., and Cleveland, D.W. (1993) Ferritin synthesis is controlled by iron- dependent translational derepression and by changes in synthesis/transport of nuclear ferritin RNAs. Proceedings of the National Academy of Sciences of the United States of America 90 : 7613-7617. Cowell, J.K. (1990) The nuclear oncoproteins: RB and p53. Seminars in Cancer Biology. 1: 437-446. Cox, B.D., Natarajan, M., Stettner, M.R., and Gladson, C.L. (2006) New concepts regarding focal adhesion kinase promotion of cell migration and proliferation. Journal of Cellular Biochemistry. 99 : 35-52. Cuadrado, A., Navarro-Yubero, C., Furneaux, H., Kinter, J., Sonderegger, P., and Munoz, A. (2002) HuD binds to three AU-rich sequences in the 3'-UTR of neuroserpin mRNA and promotes the accumulation of neuroserpin mRNA and protein. Nucleic Acids Research. 30 : 2202-2211. Cuadrado, A., Navarro-Yubero, C., Furneaux, H., and Munoz, A. (2003) Neuronal HuD gene encoding a mRNA stability regulator is transcriptionally repressed by thyroid hormone. Journal of Neurochemistry. 86 : 763-773. Culig, Z. (2004) Androgen receptor cross-talk with cell signalling pathways. Growth Factors 22 : 179-184. Culig, Z., Comuzzi, B., Steiner, H., Bartsch, G., and Hobisch, A. (2004) Expression and function of androgen receptor coactivators in prostate cancer. Journal of Steroid Biochemistry & Molecular Biology 92 : 265-271. Culig, Z., and Bartsch, G. (2006) Androgen axis in prostate cancer. Journal of Cellular Biochemistry 99 : 373-381. Dalmau, J., Furneaux, H.M., Gralla, R.J., Kris, M.G., and Posner, J.B. (1990) Detection of the anti-Hu antibody in the serum of patients with small cell lung cancer--a quantitative western blot analysis. Annals of Neurology. 27 : 544- 552.
213 Dalmau, J., Furneaux, H.M., Cordon-Cardo, C., and Posner, J.B. (1992) The expression of the Hu (paraneoplastic encephalomyelitis/sensory neuronopathy) antigen in human normal and tumor tissues. American Journal of Pathology 141 : 881-886. Dalmay, T., and Edwards, D.R. (2006) MicroRNAs and the hallmarks of cancer. Oncogene 25 : 6170-6175. Daniels, R.H., and Bokoch, G.M. (1999) p21-activated protein kinase: a crucial component of morphological signaling? Trends in Biochemical Sciences 24 : 350-355. Dash, B.C., and El-Deiry, W.S. (2004) Cell cycle checkpoint control mechanisms that can be disrupted in cancer. Methods Mol Biol. 280 : 99-161. de Bruijn, D.R., dos Santos, N.R., Kater-Baats, E., Thijssen, J., van den Berk, L., Stap, J., Balemans, M., Schepens, M., Merkx, G., and van Kessel, A.G. (2002) The cancer-related protein SSX2 interacts with the human homologue of a Ras-like GTPase interactor, RAB3IP, and a novel nuclear protein, SSX2IP. Genes Chromosomes Cancer. 34 : 285-298. de Silanes, I.L., Fan, J., Yang, X., Zonderman, A.B., Potapova, O., Pizer, E.S., and Gorospe, M. (2003) Role of the RNA-binding protein HuR in colon carcinogenesis. Oncogene 22 : 7146-7154. de Silanes, I.L., Zhan, M., Lal, A., Yang, X., and Gorospe, M. (2004) Identification of a target RNA motif for RNA-binding protein HuR. Proceedings of the National Academy of Sciences of the United States of America 101 : 2987-2992. Dean, J.L., Sarsfield, S.J., Tsounakou, E., and Saklatvala, J. (2003) p38 Mitogen- activated protein kinase stabilizes mRNAs that contain cyclooxygenase-2 and tumor necrosis factor AU-rich elements by inhibiting deadenylation. Journal of Biological Chemistry 278 : 39470-39476. Dean, J.L., Sully, G., Clark, A.R., and Saklatvala, J. (2004) The involvement of AU- rich element-binding proteins in p38 mitogen-activated protein kinase pathway- mediated mRNA stabilisation. Cellular Signalling 16 : 1113-1121. Debes, J.D., Schmidt, L.J., Huang, H., and Tindall, D.J. (2002) p300 mediates androgen-independent transactivation of the androgen receptor by interleukin 6. Cancer Research 62 : 5632-5636. Debes, J.D., Sebo, T.J., Lohse, C.M., Murphy, L.M., Haugen de, A.L., and Tindall, D.J. (2003) p300 in prostate cancer proliferation and progression. Cancer Research 63 : 7638-7640. Debes, J.D., Comuzzi, B., Schmidt, L.J., Dehm, S.M., Culig, Z., and Tindall, D.J. (2005) p300 regulates androgen receptor-independent expression of prostate- specific antigen in prostate cancer cells treated chronically with interleukin-6. Cancer Research 65 : 5965-5973. Dehm, S.M., and Tindall, D.J. (2006) Molecular regulation of androgen action in prostate cancer. Journal of Cellular Biochemistry 99 : 333-344. DeMaria, C.T., Sun, Y., Long, L., Wagner, B.J., and Brewer, G. (1997) Structural determinants in AUF1 required for high affinity binding to A + U-rich elements. Journal of Biological Chemistry 272 : 27635-27643. Denkert, C., Weichert, W., Winzer, K.J., Muller, B.M., Noske, A., Niesporek, S., Kristiansen, G., Guski, H., Dietel, M., and Hauptmann, S. (2004) Expression of the ELAV-like protein HuR is associated with higher tumor grade and increased cyclooxygenase-2 expression in human breast carcinoma. Clinical Cancer Research 10 : 5580-5586. DePrimo, S.E., Diehn, M., Nelson, J.B., Reiter, R.E., Matese, J., Fero, M., Tibshirani, R., Brown, P.O., and Brooks, J.D. (2002) Transcriptional programs activated by exposure of human prostate cancer cells to androgen. Genome Biology 3: 14. Deschenes-Furry, J., Angus, L.M., Belanger, G., Mwanjewe, J., and Jasmin, B.J. (2005a) Role of ELAV-like RNA-binding proteins HuD and HuR in the post-
214 transcriptional regulation of acetylcholinesterase in neurons and skeletal muscle cells. Chemico Biological Interactions 158 : 43-49. Deschenes-Furry, J., Belanger, G., Mwanjewe, J., Lunde, J.A., Parks, R.J., Perrone- Bizzozero, N., and Jasmin, B.J. (2005b) The RNA-binding protein HuR binds to acetylcholinesterase transcripts and regulates their expression in differentiating skeletal muscle cells. Journal of Biological Chemistry 280 : 25361-25368. Dibbens, J.A., Miller, D.L., Damert, A., Risau, W., Vadas, M.A., and Goodall, G.J. (1999) Hypoxic regulation of vascular endothelial growth factor mRNA stability requires the cooperation of multiple RNA elements. Molecular Biology of the Cell 10 : 907-919. Dixon, D.A., Tolley, N.D., King, P.H., Nabors, L.B., McIntyre, T.M., Zimmerman, G.A., and Prescott, S.M. (2001) Altered expression of the mRNA stability factor HuR promotes cyclooxygenase-2 expression in colon cancer cells. Journal of Clinical Investigation 108 : 1657-1665. Dixon, D.A., Balch, G.C., Kedersha, N., Anderson, P., Zimmerman, G.A., Beauchamp, R.D., and Prescott, S.M. (2003) Regulation of cyclooxygenase-2 expression by the translational silencer TIA-1. Journal of Experimental Medicine 198 : 475- 481. Doesburg, P., Kuil, C.W., Berrevoets, C.A., Steketee, K., Faber, P.W., Mulder, E., Brinkmann, A.O., and Trapman, J. (1997) Functional in vivo interaction between the amino-terminal, transactivation domain and the ligand binding domain of the androgen receptor. Biochemistry 36 : 1052-1064. Dreyfuss, G., Kim, V.N., and Kataoka, N. (2002) Messenger-RNA-binding proteins and the messages they carry. Nature Reviews Molecular Cell Biology 3: 195- 205. Duncan, G., McCormick, C., and Tufaro, F. (2001) The link between heparan sulfate and hereditary bone disease: finding a function for the EXT family of putative tumor suppressor proteins. Journal of Clinical Investigation. 108 : 511-516. Edwards, J., Krishna, N.S., Grigor, K.M., and Bartlett, J.M. (2003) Androgen receptor gene amplification and protein expression in hormone refractory prostate cancer. British Journal of Cancer. 89 : 552-556. Edwards, J., and Bartlett, J.M. (2005a) The androgen receptor and signal-transduction pathways in hormone-refractory prostate cancer. Part 1: Modifications to the androgen receptor. British Journal of Urology International 95 : 1320-1326. Edwards, J., and Bartlett, J.M. (2005b) The androgen receptor and signal-transduction pathways in hormone-refractory prostate cancer. Part 2: Androgen-receptor cofactors and bypass pathways. British Journal of Urology International 95 : 1327-1335. Edwards, S.M., Badzioch, M.D., Minter, R., Hamoudi, R., Collins, N., Ardern-Jones, A., Dowe, A., Osborne, S., Kelly, J., Shearer, R., Easton, D.F., Saunders, G.F., Dearnaley, D.P., and Eeles, R.A. (1999) Androgen receptor polymorphisms: association with prostate cancer risk, relapse and overall survival. International Journal of Cancer 84 : 458-465. Erkinheimo, T.L., Lassus, H., Sivula, A., Sengupta, S., Furneaux, H., Hla, T., Haglund, C., Butzow, R., and Ristimaki, A. (2003) Cytoplasmic HuR expression correlates with poor outcome and with cyclooxygenase 2 expression in serous ovarian carcinoma. Cancer Research 63 : 7591-7594. Faber, P., van Rooij, H., van der Korput, H., Baarends, W., Brinkmann, A., Grootegoed, J., and Trapman, J. (1991) Characterization of the human androgen receptor transcription unit. Journal of Biological Chemistry. 266 : 10743-10749. Fan, X.C., and Steitz, J.A. (1998a) Overexpression of HuR, a nuclear-cytoplasmic shuttling protein, increases the in vivo stability of ARE-containing mRNAs. EMBO Journal 17 : 3448-3460.
215 Fan, X.C., and Steitz, J.A. (1998b) HNS, a nuclear-cytoplasmic shuttling sequence in HuR. Proceedings of the National Academy of Sciences of the United States of America 95 : 15293-15298. Fan, Y., Liu, Z., Zhao, S., Lou, F., Nilsson, S., Ekman, P., Xu, D., and Fang, X. (2006) Nucleostemin mRNA is expressed in both normal and malignant renal tissues. British Journal of Cancer. 94 : 1658-1662. Farla, P., Hersmus, R., Trapman, J., and Houtsmuller, A.B. (2005) Antiandrogens prevent stable DNA-binding of the androgen receptor. Journal of Cell Science 118 : 4187-4198. Feldman, D. (1997) Androgen and vitamin D receptor gene polymorphisms: the long and short of prostate cancer risk. Journal of the National Cancer Institute. 89 : 109-111. Fialcowitz, E.J., Brewer, B.Y., Keenan, B.P., and Wilson, G.M. (2005) A hairpin-like structure within an AU-rich mRNA-destabilizing element regulates trans-factor binding selectivity and mRNA decay kinetics. Journal of Biological Chemistry 280 : 22406-22417. Figueroa, A., Cuadrado, A., Fan, J., Atasoy, U., Muscat, G.E., Munoz-Canoves, P., Gorospe, M., and Munoz, A. (2003) Role of HuR in skeletal myogenesis through coordinate regulation of muscle differentiation genes. Molecular & Cellular Biology 23 : 4991-5004. Ford, O.H., 3rd, Gregory, C.W., Kim, D., Smitherman, A.B., and Mohler, J.L. (2003) Androgen receptor gene amplification and protein expression in recurrent prostate cancer. Journal of Urology 170 : 1817-1821. Fosbrink, M., Cudrici, C., Niculescu, F., Badea, T.C., David, S., Shamsuddin, A., Shin, M.L., and Rus, H. (2005) Overexpression of RGC-32 in colon cancer and other tumors. Experimental Molecular Pathology. 78 : 116-122. Fronsdal, K., Engedal, N., Slagsvold, T., and Saatcioglu, F. (1998) CREB binding protein is a coactivator for the androgen receptor and mediates cross-talk with AP-1. Journal of Biological Chemistry 273 : 31853-31859. Fu, M., Wang, C., Reutens, A.T., Wang, J., Angeletti, R.H., Siconolfi-Baez, L., Ogryzko, V., Avantaggiati, M.L., and Pestell, R.G. (2000) p300 and p300/cAMP-response element-binding protein-associated factor acetylate the androgen receptor at sites governing hormone-dependent transactivation. Journal of Biological Chemistry 275 : 20853-20860. Fu, M., Wang, C., Wang, J., Zhang, X., Sakamaki, T., Yeung, Y.G., Chang, C., Hopp, T., Fuqua, S.A., Jaffray, E., Hay, R.T., Palvimo, J.J., Janne, O.A., and Pestell, R.G. (2002) Androgen receptor acetylation governs trans activation and MEKK1-induced apoptosis without affecting in vitro sumoylation and trans- repression function. Molecular & Cellular Biology 22 : 3373-3388. Fujimoto, N., Yeh, S., Kang, H.Y., Inui, S., Chang, H.C., Mizokami, A., and Chang, C. (1999) Cloning and characterization of androgen receptor coactivator, ARA55, in human prostate. Journal of Biological Chemistry 274 : 8316-8321. Fujiwara, T., Mori, Y., Chu, D.L., Koyama, Y., Miyata, S., Tanaka, H., Yachi, K., Kubo, T., Yoshikawa, H., and Tohyama, M. (2006) CARM1 Regulates Proliferation of PC12 Cells by Methylating HuD. Molecular & Cellular Biology. 26 : 2273-2285. Galban, S., Martindale, J.L., Mazan-Mamczarz, K., Lopez de Silanes, I., Fan, J., Wang, W., Decker, J., and Gorospe, M. (2003) Influence of the RNA-Binding Protein HuR in pVHL-Regulated p53 Expression in Renal Carcinoma Cells. Molecular & Cellular Biology. 23 : 7083-7095. Gallouzi, I.E., Parker, F., Chebli, K., Maurier, F., Labourier, E., Barlat, I., Capony, J.P., Tocque, B., and Tazi, J. (1998) A novel phosphorylation-dependent RNase activity of GAP-SH3 binding protein: a potential link between signal transduction and RNA stability. Molecular & Cellular Biology 18 : 3956-3965. Gallouzi, I.-E., Brennan, C.M., Stenberg, M.G., Swanson, M.S., Eversole, A., Maizels, N., and Steitz, J.A. (2000) HuR binding to cytoplasmic mRNA is perturbed by
216 heat shock. Proceedings of the National Academy of Sciences of the United States of America 97 : 3073-3078. Gantt, K., Cherry, J., Tenney, R., Karschner, V., and Pekala, P.H. (2005) An early event in adipogenesis, the nuclear selection of the CCAAT enhancer-binding protein {beta} (C/EBP{beta}) mRNA by HuR and its translocation to the cytosol. Journal of Biological Chemistry 280 : 24768-24774. Gao, S., Lee, P., Wang, H., Gerald, W., Adler, M., Zhang, L., Wang, Y.F., and Wang, Z. (2005) The androgen receptor directly targets the cellular Fas/FasL- associated death domain protein-like inhibitory protein gene to promote the androgen-independent growth of prostate cancer cells. Molecular Endocrinology 19 : 1792-1802. Gao, T., and McPhaul, M.J. (1998) Functional activities of the A and B forms of the human androgen receptor in response to androgen receptor agonists and antagonists. Molecular Endocrinology 12 : 654-663. Garnett, M.J., Rana, S., Paterson, H., Barford, D., and Marais, R. (2005) Wild-type and mutant B-RAF activate C-RAF through distinct mechanisms involving heterodimerization. Molecular Cell. 20 : 963-969. Gaughan, L., Logan, I.R., Neal, D.E., and Robson, C.N. (2005) Regulation of androgen receptor and histone deacetylase 1 by Mdm2-mediated ubiquitylation. Nucleic Acids Research 33 : 13-26. Gherzi, R., Lee, K.Y., Briata, P., Wegmuller, D., Moroni, C., Karin, M., and Chen, C.Y. (2004) A KH domain RNA binding protein, KSRP, promotes ARE-directed mRNA turnover by recruiting the degradation machinery. Molecular Cell 14 : 571-583. Ghosh, P.M., Malik, S., Bedolla, R., and Kreisberg, J.I. (2003) Akt in prostate cancer: possible role in androgen-independence. Current Drug Metabolism 4: 487-496. 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 3'-untranslated 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. Journal of Biological Chemistry. 278 : 2937-2946. Gioeli, D., Black, B.E., Gordon, V., Spencer, A., Kesler, C.T., Eblen, S.T., Paschal, B.M., and Weber, M.J. (2006) Stress kinase signaling regulates androgen receptor phosphorylation, transcription, and localization. Molecular Endocrinology 20 : 503-515. Giovannucci, E., Stampfer, M.J., Krithivas, K., Brown, M., Dahl, D., Brufsky, A., Talcott, J., Hennekens, C.H., and Kantoff, P.W. (1997) The CAG repeat within the androgen receptor gene and its relationship to prostate cancer.[erratum appears in Proc Natl Acad Sci U S A 1997 Jul 22;94(15):8272]. Proceedings of the National Academy of Sciences of the United States of America 94 : 3320- 3323. Gleason, D.F., and Mellinger, G.T. (1974) Prediction of prognosis for prostatic adenocarcinoma by combined histological grading and clinical staging. The Journal of Urology 111 : 58-64. Gnanapragasam, V.J., Robson, C.N., Leung, H.Y., and Neal, D.E. (2000) Androgen receptor signalling in the prostate. British Journal of Urology International. 86 : 1001-1013. Gnanapragasam, V.J., Leung, H.Y., Pulimood, A.S., Neal, D.E., and Robson, C.N. (2001) Expression of RAC 3, a steroid hormone receptor co-activator in prostate cancer. British Journal of Cancer 85 : 1928-1936. Godoy-Tundidor, S., Cavarretta, I.T., Fuchs, D., Fiechtl, M., Steiner, H., Friedbichler, K., Bartsch, G., Hobisch, A., and Culig, Z. (2005) Interleukin-6 and oncostatin M stimulation of proliferation of prostate cancer 22Rv1 cells through the signaling pathways of p38 mitogen-activated protein kinase and phosphatidylinositol 3-kinase. Prostate 64 : 209-216.
217 Goel, H.L., Moro, L., King, M., Teider, N., Centrella, M., McCarthy, T.L., Holgado- Madruga, M., Wong, A.J., Marra, E., and Languino, L.R. (2006) Beta1 integrins modulate cell adhesion by regulating insulin-like growth factor-II levels in the microenvironment. Cancer Research. 66 : 331-342. Goodall, J., Martinozzi, S., Dexter, T.J., Champeval, D., Carreira, S., Larue, L., and Goding, C.R. (2004) Brn-2 expression controls melanoma proliferation and is directly regulated by beta-catenin. Molecular and Cellular Biology. 24 : 2915- 2922. Gorospe, M. (2003) HuR in the mammalian genotoxic response: post-transcriptional multitasking. Cell Cycle 2: 412-414. Grad, J.M., Lyons, L.S., Robins, D.M., and Burnstein, K.L. (2001) The androgen receptor (AR) amino-terminus imposes androgen-specific regulation of AR gene expression via an exonic enhancer. Endocrinology. 142 : 1107-1116. Gregory, C.W., Hamil, K.G., Kim, D., Hall, S.H., Pretlow, T.G., Mohler, J.L., and French, F.S. (1998) Androgen receptor expression in androgen-independent prostate cancer is associated with increased expression of androgen-regulated genes. Cancer Research 58 : 5718-5724. Gregory, C.W., Johnson, R.T., Jr., Mohler, J.L., French, F.S., and Wilson, E.M. (2001) Androgen receptor stabilization in recurrent prostate cancer is associated with hypersensitivity to low androgen. Cancer Research 61 : 2892-2898. Gregory, C.W., He, B., Johnson, R.T., Ford, O.H., Mohler, J.L., French, F.S., and Wilson, E.M. (2001a) A mechanism for androgen receptor-mediated prostate cancer recurrence after androgen deprivation therapy. Cancer Research 61 : 4315-4319. Gregory, C.W., Fei, X., Ponguta, L.A., He, B., Bill, H.M., French, F.S., and Wilson, E.M. (2004) Epidermal growth factor increases coactivation of the androgen receptor in recurrent prostate cancer. Journal of Biological Chemistry 279 : 7119-7130. Gstaiger, M., Jordan, R., Lim, M., Catzavelos, C., Mestan, J., Slingerland, J., and Krek, W. (2001) Skp2 is oncogenic and overexpressed in human cancers. Proceedings of the National Academy of Sciences of the United States of America 98 : 5043-5048. Guardavaccaro, D., and Pagano, M. (2004) Oncogenic aberrations of cullin- dependent ubiquitin ligases. Oncogene. 23 : 2037-2049. Guo, X., and Hartley, R.S. (2006) HuR contributes to cyclin E1 deregulation in MCF-7 breast cancer cells. Cancer Research 66 : 7948-7956. Haag, P., Bektic, J., Bartsch, G., Klocker, H., and Eder, I.E. (2005) Androgen receptor down regulation by small interference RNA induces cell growth inhibition in androgen sensitive as well as in androgen independent prostate cancer cells. Journal of Steroid Biochemistry and Molecular Biology. 96 : 251-258. Haeussler, J., Haeusler, J., Striebel, A.M., Assum, G., Vogel, W., Furneaux, H., and Krone, W. (2000) Tumor antigen HuR binds specifically to one of five protein- binding segments in the 3'-untranslated region of the neurofibromin messenger RNA. Biochemical and Biophysical Research Commununications. 267 : 726-732. Hakimi, J.M., Schoenberg, M.P., Rondinelli, R.H., Piantadosi, S., and Barrack, E.R. (1995) Androgen receptor variants with short glutamine or glycine repeats may identify unique subpopulations of men with prostate cancer. Clinical Cancer Research 3: 1599-1608. Halkidou, K., Gnanapragasam, V.J., Mehta, P.B., Logan, I.R., Brady, M.E., Cook, S., Leung, H.Y., Neal, D.E., and Robson, C.N. (2003) Expression of Tip60, an androgen receptor coactivator, and its role in prostate cancer development. Oncogene 22 : 2466-2477. Han, B., Xie, H., Chen, Q., and Zhang, J.T. (2006) Sensitizing hormone-refractory prostate cancer cells to drug treatment by targeting 14-3-3sigma. Molecular Cancer Therapies. 5: 903-912.
218 Hannon, G.J., and Beach, D. (1994) p15INK4B is a potential effector of TGF-beta- induced cell cycle arrest. Nature. 371 : 257-261. Hara, T., Miyazaki, J., Araki, H., Yamaoka, M., Kanzaki, N., Kusaka, M., and Miyamoto, M. (2003) Novel mutations of androgen receptor: a possible mechanism of bicalutamide withdrawal syndrome. Cancer Research. 63 : 149- 153. He, B., Kemppainen, J.A., and Wilson, E.M. (2000) FXXLF and WXXLF Sequences Mediate the NH2-terminal Interaction with the Ligand Binding Domain of the Androgen Receptor. Journal of Biological Chemistry 275 : 22986-22994. He, B., Lee, L.W., Minges, J.T., and Wilson, E.M. (2002) Dependence of Selective Gene Activation on the Androgen Receptor NH2- and COOH-terminal Interaction. Journal of Biological Chemistry. 277 : 25631-25639. He, B., Gampe, R.T., Jr., Kole, A.J., Hnat, A.T., Stanley, T.B., An, G., Stewart, E.L., Kalman, R.I., Minges, J.T., and Wilson, E.M. (2004) Structural basis for androgen receptor interdomain and coactivator interactions suggests a transition in nuclear receptor activation function dominance. Molecular Cell 16 : 425-438. He, W.W., Sciavolino, P.J., Wing, J., Augustus, M., Hudson, P., Meissner, P.S., Curtis, R.T., Shell, B.K., Bostwick, D.G., Tindall, D.J., Gelmann, E.P., Abate- Shen, C., and Carter, K.C. (1997) A novel human prostate-specific, androgen- regulated homeobox gene (NKX3.1) that maps to 8p21, a region frequently deleted in prostate cancer. Genomics 43 : 69-77. Heinlein, C.A., and Chang, C. (2004) Androgen receptor in prostate cancer. Endocrine Reviews 25 : 276-308. Heinonen, M., Bono, P., Narko, K., Chang, S.-H., Lundin, J., Joensuu, H., Furneaux, H., Hla, T., Haglund, C., and Ristimaki, A. (2005) Cytoplasmic HuR expression is a prognostic factor in invasive ductal breast carcinoma. Cancer Research 65 : 2157-2161. Hickman, E.S., Moroni, M.C., and Helin, K. (2002) The role of p53 and pRB in apoptosis and cancer. Current Opinions in Genetics and Development. 12 : 60- 66. Hodgson, M.C., Astapova, I., Cheng, S., Lee, L.J., Verhoeven, M.C., Choi, E., Balk, S.P., and Hollenberg, A.N. (2005) The androgen receptor recruits nuclear receptor CoRepressor (N-CoR) in the presence of mifepristone via its N and C termini revealing a novel molecular mechanism for androgen receptor antagonists. Journal of Biological Chemistry 280 : 6511-6519. Hollams, E.M., Giles, K.M., Thomson, A.M., and Leedman, P.J. (2002) MRNA stability and the control of gene expression: implications for human disease. Neurochemical Research 27 : 957-980. Horoszewicz, J.S., Leong, S.S., Kawinski, E., Karr, J.P., Rosenthal, H., Chu, T.M., Mirand, E.A., and Murphy, G.P. (1983) LNCaP model of human prostatic carcinoma. Cancer Research 43 : 1809-1818. Hu, Y., Ippolito, J.E., Garabedian, E.M., Humphrey, P.A., and Gordon, J.I. (2002) Molecular characterization of a metastatic neuroendocrine cell cancer arising in the prostates of transgenic mice. Journal of Biological Chemistry 277 : 44462-44474. Huang, Z., Traugh, J.A., and Bishop, J.M. (2004) Negative control of the Myc protein by the stress-responsive kinase Pak2. Molecular & Cellular Biology. 24 : 1582- 1594. Huggins, C., Stephens, R., and Hodges, C. (1940) Studies on prostatic cancer: The effects of castration on advanced carcinoma of the prostate gland. Archives of Surgery 43 : 209. Huynh, H.T., Larsson, C., Narod, S., and Pollak, M. (1995) Tumor suppressor activity of the gene encoding mammary-derived growth inhibitor. Cancer Research. 55 : 2225-2231.
219 Ingles, S.A., Ross, R.K., Yu, M.C., Irvine, R.A., La Pera, G., Haile, R.W., and Coetzee, G.A. (1997) Association of prostate cancer risk with genetic polymorphisms in vitamin D receptor and androgen receptor. Journal of the National Cancer Institute. 89 : 166-170. Irvine, R.A., Yu, M.C., Ross, R.K., and Coetzee, G.A. (1995) The CAG and GGC microsatellites of the androgen receptor gene are in linkage disequilibrium in men with prostate cancer. Cancer Research 55 : 1937-1940. Izquierdo, J.M. (2006) Control of the ATP synthase beta subunit expression by RNA- binding proteins TIA-1, TIAR, and HuR. Biochemical & Biophysical Research Communications 348 : 703-711. Jaffe, L.F. (2003) Epigenetic theories of cancer initiation. Advances in Cancer Research 90 : 209-230. Jakobi, R., Moertl, E., and Koeppel, M.A. (2001) p21-activated protein kinase gamma- PAK suppresses programmed cell death of BALB3T3 fibroblasts. Journal of Biological Chemistry 276 : 16624-16634. Jenster, G., van der Korput, J.A., Trapman, J., and Brinkmann, A.O. (1992) Functional domains of the human androgen receptor. Journal of Steroid Biochemistry & Molecular Biology 41 : 671-675. Jenster, G., de Ruiter, P.E., van der Korput, H.A., Kuiper, G.G., Trapman, J., and Brinkmann, A.O. (1994) Changes in the abundance of androgen receptor isotypes: effects of ligand treatment, glutamine-stretch variation, and mutation of putative phosphorylation sites. Biochemistry 33 : 14064-14072. Jeyaraj, S., Dakhlallah, D., Hill, S.R., and Lee, B.S. (2005) HuR stabilizes vacuolar H+-translocating ATPase mRNA during cellular energy depletion. Journal of Biological Chemistry 280 : 37957-37964. Jeyaraj, S.C., Dakhlallah, D., Hill, S.R., and Lee, B.S. (2006) Expression and distribution of HuR during ATP depletion and recovery in proximal tubule cells. American Journal Physiology & Renal Physiology: 00440.02005. Jia, L., Kim, J., Shen, H., Clark, P.E., Tilley, W.D., and Coetzee, G.A. (2003) Androgen receptor activity at the prostate specific antigen locus: steroidal and non-steroidal mechanisms. Molecular Cancer Research: MCR 1: 385-392. Jimenez, A., Pelto-Huikko, M., Gustafsson, J.A., and Miranda-Vizuete, A. (2006) Characterization of human thioredoxin-like-1: potential involvement in the cellular response against glucose deprivation. FEBS Letters. 580 : 960-967. Johnson, J.M., Castle, J., Garrett-Engele, P., Kan, Z., Loerch, P.M., Armour, C.D., Santos, R., Schadt, E.E., Stoughton, R., and Shoemaker, D.D. (2003) Genome-wide survey of human alternative pre-mRNA splicing with exon junction microarrays. Science 302 : 2141-2144. Joseph, B., Orlian, M., and Furneaux, H. (1998) p21(waf1) mRNA contains a conserved element in its 3'-untranslated region that is bound by the Elav-like mRNA-stabilizing proteins. Journal of Biological Chemistry 273 : 20511-20516. Kanehisa, M. (1997) A database for post-genome analysis. Trends in Genetics 13 : 375-376. Kanehisa, M., and Goto, S. (2000) KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Research 28 : 27-30. Kanehisa, M., Goto, S., Hattori, M., Aoki-Kinoshita, K.F., Itoh, M., Kawashima, S., Katayama, T., Araki, M., and Hirakawa, M. (2006) From genomics to chemical genomics: new developments in KEGG. Nucleic Acids Research 34 : 1. Kang, H.Y., Yeh, S., Fujimoto, N., and Chang, C. (1999) Cloning and characterization of human prostate coactivator ARA54, a novel protein that associates with the androgen receptor. Journal of Biological Chemistry 274 : 8570-8576. Kasashima, K., Terashima, K., Yamamoto, K., Sakashita, E., and Sakamoto, H. (1999) Cytoplasmic localization is required for the mammalian ELAV-like protein HuD to induce neuronal differentiation.[erratum appears in Genes Cells 2000 Mar;5(3):235-7]. Genes to Cells 4: 667-683.
220 Kasashima, K., Sakashita, E., Saito, K., and Sakamoto, H. (2002) Complex formation of the neuron-specific ELAV-like Hu RNA-binding proteins. Nucleic Acids Research 30 : 4519-4526. Katoh, H., and Negishi, M. (2003) RhoG activates Rac1 by direct interaction with the Dock180-binding protein Elmo. Nature. 424 : 461-464. Katsanou, V., Papadaki, O., Milatos, S., Blackshear, P.J., Anderson, P., Kollias, G., and Kontoyiannis, D.L. (2005) HuR as a negative posttranscriptional modulator in inflammation. Molecular Cell 19 : 777-789. Kawai, T., Lal, A., Yang, X., Galban, S., Mazan-Mamczarz, K., and Gorospe, M. (2006) Translational control of cytochrome c by RNA-binding proteins TIA-1 and HuR. Molecular & Cellular Biology 26 : 3295-3307. Kedersha, N., and Anderson, P. (2002) Stress granules: sites of mRNA triage that regulate mRNA stability and translatability. Biochemical Society Transactions 30 : 963-969. Keyomarsi, K., and Pardee, A.B. (1993) Redundant cyclin overexpression and gene amplification in breast cancer cells. Proceedings of the National Academy of Sciences of the United States of America. 90 : 1112-1116. Khabar, K.S. (2005) The AU-rich transcriptome: more than interferons and cytokines, and its role in disease. Journal of Interferon & Cytokine Research. 25 : 1-10. Kiledjian, M., DeMaria, C.T., Brewer, G., and Novick, K. (1997) Identification of AUF1 (heterogeneous nuclear ribonucleoprotein D) as a component of the alpha- globin mRNA stability complex. Molecular & Cellular Biology 17 : 4870-4876. Kiledjian, M., Day, N., and Trifillis, P. (1999) Purification and RNA binding properties of the polycytidylate-binding proteins alphaCP1 and alphaCP2. Methods 17 : 84-91. Kim, J., Jia, L., Stallcup, M.R., and Coetzee, G.A. (2005) The role of protein kinase A pathway and cAMP responsive element-binding protein in androgen receptor- mediated transcription at the prostate-specific antigen locus. Journal of Molecular Endocrinology 34 : 107-118. Kloss, S., Srivastava, R., and Mulsch, A. (2004) Down-regulation of soluble guanylyl cyclase expression by cyclic AMP is mediated by mRNA-stabilizing protein HuR. Molecular Pharmacology 65 : 1440-1451. Koeneman, K.S. (2006) Prostate cancer stem cells, telomerase biology, epigenetic modifiers, and molecular systemic therapy for the androgen-independent lethal phenotype. Urologic Oncology 24 : 119-121. Konopka, K., Lee, N.S., Rossi, J., and Duzgunes, N. (2000) Rev-binding aptamer and CMV promoter act as decoys to inhibit HIV replication. Gene 255 : 235-244. Krajewska, M., Turner, B.C., Shabaik, A., Krajewski, S., and Reed, J.C. (2006) Expression of BAG-1 protein correlates with aggressive behavior of prostate cancers. Prostate 66 : 801-810. Kraus, S., Naor, Z., and Seger, R. (2006) Gonadotropin-releasing hormone in apoptosis of prostate cancer cells. Cancer Letters 234 : 109-123. Krongrad, A., Wilson, C.M., Wilson, J.D., Allman, D.R., and McPhaul, M.J. (1991) Androgen increases androgen receptor protein while decreasing receptor mRNA in LNCaP cells. Molecular & Cellular Endocrinology 76 : 79-88. Kuiper, G.G., de Ruiter, P.E., Grootegoed, J.A., and Brinkmann, A.O. (1991) Synthesis and post-translational modification of the androgen receptor in LNCaP cells. Molecular & Cellular Endocrinology 80 : 65-73. Kuiper, G.G., de Ruiter, P.E., and Brinkmann, A.O. (1992) Androgen receptor heterogeneity in LNCaP cells is caused by a hormone independent phosphorylation step. Journal of Steroid Biochemistry & Molecular Biology 41 : 697-700. Kullmann, M., Gopfert, U., Siewe, B., and Hengst, L. (2002) ELAV/Hu proteins inhibit p27 translation via an IRES element in the p27 5'UTR. Genes & Development 16 : 3087-3099.
221 Kumar, R., Gururaj, A.E., and Barnes, C.J. (2006) p21-activated kinases in cancer. Nature Reviews. Cancer. 6: 459-471. Kuphal, S., Lodermeyer, S., Bataille, F., Schuierer, M., Hoang, B.H., and Bosserhoff, A.K. (2006) Expression of Dickkopf genes is strongly reduced in malignant melanoma. Oncogene. 25 : 5027-5036. La Spada, A.R., Wilson, E.M., Lubahn, D.B., Harding, A.E., and Fischbeck, K.H. (1991) Androgen receptor gene mutations in X-linked spinal and bulbar muscular atrophy. Nature 352 : 77-79. Lagnado, C.A., Brown, C.Y., and Goodall, G.J. (1994) AUUUA is not sufficient to promote poly(A) shortening and degradation of an mRNA: the functional sequence within AU-rich elements may be UUAUUUA(U/A)(U/A). Molecular & Cellular Biology 14 : 7984-7995. Lai, W.S., Kennington, E.A., and Blackshear, P.J. (2003) Tristetraprolin and its family members can promote the cell-free deadenylation of AU-rich element- containing mRNAs by poly(A) ribonuclease. Molecular & Cellular Biology 23 : 3798-3812. Lal, A., Mazan-Mamczarz, K., Kawai, T., Yang, X., Martindale, J.L., and Gorospe, M. (2004) Concurrent versus individual binding of HuR and AUF1 to common labile target mRNAs. EMBO Journal 23 : 3092-3102. Laroia, G., Cuesta, R., Brewer, G., and Schneider, R.J. (1999) Control of mRNA decay by heat shock-ubiquitin-proteasome pathway. Science. 284 : 499-502. Laronga, C., Yang, H.Y., Neal, C., and Lee, M.H. (2000) Association of the cyclin- dependent kinases and 14-3-3 sigma negatively regulates cell cycle progression. Journal of Biological Chemistry 275 : 23106-23112. Lasa, M., Mahtani, K.R., Finch, A., Brewer, G., Saklatvala, J., and Clark, A.R. (2000) Regulation of Cyclooxygenase 2 mRNA Stability by the Mitogen-Activated Protein Kinase p38 Signaling Cascade. Molecular & Cellular Biology. 20 : 4265-4274. Latil, A., Bieche, I., Vidaud, D., Lidereau, R., Berthon, P., Cussenot, O., and Vidaud, M. (2001) Evaluation of androgen, estrogen (ER alpha and ER beta), and progesterone receptor expression in human prostate cancer by real-time quantitative reverse transcription-polymerase chain reaction assays. Cancer Research 61 : 1919-1926. Leandersson, K., Riesbeck, K., and Andersson, T. (2006) Wnt-5a mRNA translation is suppressed by the Elav-like protein HuR in human breast epithelial cells. Nucleic Acids Research 34 : 3988-3999. Lee, D.K., and Chang, C. (2003) Expression and Degradation of Androgen Receptor: Mechanism and Clinical Implication. Journal of Clinicical Endocrinology & Metabolism 88 : 4043-4054. Lee, H.K., and Jeong, S. (2006) {beta}-Catenin stabilizes Cyclooxygenase-2 mRNA by interacting with AU-rich elements of 3'-UTR. Nucleic Acids Research : gkl698. Legendre, M., and Gautheret, D. (2003) Sequence determinants in human polyadenylation site selection. BMC Genomics. 4: 7. Lesche, R., Peetz, A., van der Hoeven, F., and Ruther, U. (1997) Ft1, a novel gene related to ubiquitin-conjugating enzymes, is deleted in the Fused toes mouse mutation. Mammalian Genome. 8: 879-883. Levine, A.J. (1997) p53, the cellular gatekeeper for growth and division. Cell. 88 : 323- 331. Levy, N.S., Chung, S., Furneaux, H., and Levy, A.P. (1998) Hypoxic stabilization of vascular endothelial growth factor mRNA by the RNA-binding protein HuR. Journal of Biological Chemistry 273 : 6417-6423. Lewis, T., Gueydan, C., Huez, G., Toulme, J.J., and Kruys, V. (1998) Mapping of a minimal AU-rich sequence required for lipopolysaccharide-induced binding of a 55-kDa protein on tumor necrosis factor-alpha mRNA. Journal of Biological Chemistry 273 : 13781-13786.
222 Li, B., Cong, F., Tan, C.P., Wang, S.X., and Goff, S.P. (2002a) Aph2, a protein with a zf-DHHC motif, interacts with c-Abl and has pro-apoptotic activity. Journal of Biological Chemistry 277 : 28870-28876. Li, H., Park, S., Kilburn, B., Jelinek, M.A., Henschen-Edman, A., Aswad, D.W., Stallcup, M.R., and Laird-Offringa, I.A. (2002b) Lipopolysaccharide-induced methylation of HuR, an mRNA-stabilizing protein, by CARM1. Journal of Biological Chemistry. 277 : 44623-44630. Li, Y., Jiang, Z.Z., Chen, H.X., Leung, W.K., Sung, J.J., and Ma, W.J. (2005) The existence of a putative regulatory element in 3'-untranslated region of proto- oncogene HOX11's mRNA. Journal of Biochemistry & Molecular Biology 38 : 500-506. Liao, G., Chen, L.Y., Zhang, A., Godavarthy, A., Xia, F., Ghosh, J.C., Li, H., and Chen, J.D. (2003) Regulation of androgen receptor activity by the nuclear receptor corepressor SMRT. Journal of Biological Chemistry 278 : 5052-5061. Lim, M.S., Adamson, A., Lin, Z., Perez-Ordonez, B., Jordan, R.C.K., Tripp, S., Perkins, S.L., and Elenitoba-Johnson, K.S.J. (2002) Expression of Skp2, a p27Kip1 ubiquitin ligase, in malignant lymphoma: correlation with p27Kip1 and proliferation index. Blood 100 : 2950-2956. Lin, H.K., Wang, L., Hu, Y.C., Altuwaijri, S., and Chang, C. (2002) Phosphorylation- dependent ubiquitylation and degradation of androgen receptor by Akt require Mdm2 E3 ligase. E MBO Journal. 21 : 4037-4048. Lin, H.K., Hu, Y.C., Lee, D.K., and Chang, C. (2004) Regulation of androgen receptor signaling by PTEN (phosphatase and tensin homolog deleted on chromosome 10) tumor suppressor through distinct mechanisms in prostate cancer cells. Molecular Endocrinology. 18 : 2409-2423.. Linja, M.J., Savinainen, K.J., Saramaki, O.R., Tammela, T.L., Vessella, R.L., and Visakorpi, T. (2001) Amplification and overexpression of androgen receptor gene in hormone-refractory prostate cancer. Cancer Research 61 : 3550-3555. Linja, M.J., Porkka, K.P., Kang, Z., Savinainen, K.J., Janne, O.A., Tammela, T.L., Vessella, R.L., Palvimo, J.J., and Visakorpi, T. (2004) Expression of androgen receptor coregulators in prostate cancer. Clinical Cancer Research 10 : 1032- 1040. Linker, K., Pautz, A., Fechir, M., Hubrich, T., Greeve, J., and Kleinert, H. (2005) Involvement of KSRP in the post-transcriptional regulation of human iNOS expression-complex interplay of KSRP with TTP and HuR. Nucleic Acids Research 33 : 4813-4827. Loflin, P., Chen, C.Y., and Shyu, A.B. (1999a) Unraveling a cytoplasmic role for hnRNP D in the in vivo mRNA destabilization directed by the AU-rich element. Genes & Development 13 : 1884-1897. Loflin, P.T., Chen, C.Y., Xu, N., and Shyu, A.B. (1999b) Transcriptional pulsing approaches for analysis of mRNA turnover in mammalian cells. Methods 17 : 11-20. Lu, J.Y., and Schneider, R.J. (2004) Tissue distribution of AU-rich mRNA-binding proteins involved in regulation of mRNA decay. Journal of Biological Chemistry 279 : 12974-12979. Lu, S., Liu, M., Epner, D.E., Tsai, S.Y., and Tsai, M.J. (1999) Androgen regulation of the cyclin-dependent kinase inhibitor p21 gene through an androgen response element in the proximal promoter. Molecular Endocrinology 13 : 376-384. Lubahn, D.B., Joseph, D.R., Sar, M., Tan, J., Higgs, H.N., Larson, R.E., French, F.S., and Wilson, E.M. (1988) The human androgen receptor: complementary deoxyribonucleic acid cloning, sequence analysis and gene expression in prostate. Molecular Endocrinology 2: 1265-1275. Luczak, M.W., and Jagodzinski, P.P. (2006) The role of DNA methylation in cancer development. Folia Histochemica et Cytobiologica 44 : 143-154. Luschen, S., Scherer, G., Ussat, S., Ungefroren, H., and Adam-Klages, S. (2004) Inhibition of p38 mitogen-activated protein kinase reduces TNF-induced
223 activation of NF-kappaB, elicits caspase activity, and enhances cytotoxicity. Experimental Cell Research. 293 : 196-206. Lykke-Andersen, J., and Wagner, E. (2005) Recruitment and activation of mRNA decay enzymes by two ARE-mediated decay activation domains in the proteins TTP and BRF-1. Genes & Development 19 : 351-361. Ma, W.J., Cheng, S., Campbell, C., Wright, A., and Furneaux, H. (1996) Cloning and characterization of HuR, a ubiquitously expressed Elav-like protein. The Journal of Biological Chemistry 271 : 8144-8151. Magnus, P., and Sadkowsky, K. (2006) Mortality over the twentieth century in Australia : Australian Institute for Health and Wefare. Makeyev, A.V., Eastmond, D.L., and Liebhaber, S.A. (2002) Targeting a KH-domain protein with RNA decoys. RNA: A Publication of the RNA Society 8: 1160- 1173. Makeyev, A.V., and Liebhaber, S.A. (2002) The poly(C)-binding proteins: a multiplicity of functions and a search for mechanisms. RNA: A Publication of the Rna Society 8: 265-278. Manley, G.T., Smitt, P.S., Dalmau, J., Posner, J.B. (1995) Hu antigens: reactivity with Hu antibodies, tumor expression, and major immunogenic sites. Annals of Neurology 38 : 102-110. Maris, C., Dominguez, C., and Allain, F.H.-T. (2005) The RNA recognition motif, a plastic RNA-binding platform to regulate post-transcriptional gene expression. FEBS Journal 272 : 2118-2131. Marquis, J., Paillard, L., Audic, Y., Cosson, B., Danos, O., Le Bec, C., and Osborne, H.B. (2006) CUG-BP1/CELF1 requires UGU-rich sequences for high affinity binding. Biochemical Journal 29 : 29. Masiello, D., Cheng, S., Bubley, G.J., Lu, M.L., and Balk, S.P. (2002) Bicalutamide Functions as an Androgen Receptor Antagonist by Assembly of a Transcriptionally Inactive Receptor. Journal of Biological Chemistry. 277 : 26321-26326. Mathews, D.H., Sabina, J., Zuker, M., and Turner, D.H. (1999) Expanded sequence dependence of thermodynamic parameters improves prediction of RNA secondary structure. Journal of Molecular Biology 288 : 911-940. Mazan-Mamczarz, K., Galban, S., de Silanes, I.L., Martindale, J.L., Atasoy, U., Keene, J.D., and Gorospe, M. (2003) RNA-binding protein HuR enhances p53 translation in response to ultraviolet light irradiation. Proceedings of the National Academy of Sciences of the United States of America 100 : 8354- 8359. McWilliam, L.J., Manson, C., and George, N.J. (1997) Neuroendocrine differentiation and prognosis in prostatic adenocarcinoma. British Journal of Urology 80 : 287- 290. Mehta, A., Trotta, C.R., and Peltz, S.W. (2006) Derepression of the Her-2 uORF is mediated by a novel post-transcriptional control mechanism in cancer cells 10.1101/gad.1388706. Genes & Development. 20 : 939-953. Meisner, N.C., Hackermuller, J., Uhl, V., Aszodi, A., Jaritz, M., and Auer, M. (2004) mRNA openers and closers: modulating AU-rich element-controlled mRNA stability by a molecular switch in mRNA secondary structure. Chembiochem 5: 1432-1447. Meng, Z., King, P.H., Nabors, L.B., Jackson, N.L., Chen, C.Y., Emanuel, P.D., and Blume, S.W. (2005) The ELAV RNA-stability factor HuR binds the 5'- untranslated region of the human IGF-IR transcript and differentially represses cap-dependent and IRES-mediated translation. Nucleic Acids Research 33 : 2962-2979. Meyer, S., Temme, C., and Wahle, E. (2004) Messenger RNA turnover in eukaryotes: pathways and enzymes. Critical Reviews in Biochemistry & Molecular Biology 39 : 197-216.
224 Miki, H., Miura, K., and Takenawa, T. (1996) N-WASP, a novel actin-depolymerizing protein, regulates the cortical cytoskeletal rearrangement in a PIP2-dependent manner downstream of tyrosine kinases. EMBO Journal. 15 : 5326-5335. Mizokami, A., and Chang, C. (1994) Induction of translation by the 5'-untranslated region of human androgen receptor mRNA. Journal of Biological Chemistry. 269 : 25655-25659. Mizokami, A., Koh, E., Fujita, H., Maeda, Y., Egawa, M., Koshida, K., Honma, S., Keller, E.T., and Namiki, M. (2004) The adrenal androgen androstenediol is present in prostate cancer tissue after androgen deprivation therapy and activates mutated androgen receptor. Cancer Research 64 : 765-771. Mo, Q., Lu, S.-f., and Simon, N.G. (2006) Dehydroepiandrosterone and its metabolites: Differential effects on androgen receptor trafficking and transcriptional activity. The Journal of Steroid Biochemistry and Molecular Biology 99 : 50-58. Mobarak, C.D., Anderson, K.D., Morin, M., Beckel-Mitchener, A., Rogers, S.L., Furneaux, H., King, P., and Perrone-Bizzozero, N.I. (2000) The RNA-binding protein HuD is required for GAP-43 mRNA stability, GAP-43 gene expression, and PKC-dependent neurite outgrowth in PC12 cells. Molecular Biology of the Cell 11 : 3191-3203. Mora, G.R., and Mahesh, V.B. (1999) Autoregulation of the androgen receptor at the translational level: testosterone induces accumulation of androgen receptor mRNA in the rat ventral prostate polyribosomes. Steroids. 64 : 587-591. Moraes, K.C., Wilusz, C.J., and Wilusz, J. (2006) CUG-BP binds to RNA substrates and recruits PARN deadenylase. RNA: A Publication of the Rna Society 12 : 1084-1091. Morris, B.J., Adams, D.J., Beveridge, D.J., van der Weyden, L., Mangs, H., and Leedman, P.J. (2004) cAMP controls human renin mRNA stability via specific RNA-binding proteins. Acta Physiologica Scandinavica 181 : 369-373. Mulholland, D.J., Cox, M., Read, J., Rennie, P., and Nelson, C. (2004) Androgen responsiveness of Renilla luciferase reporter vectors is promoter, transgene, and cell line dependent. The Prostate 59 : 115-119. Murillo, H., Huang, H., Schmidt, L.J., Smith, D.I., and Tindall, D.J. (2001) Role of PI3K signaling in survival and progression of LNCaP prostate cancer cells to the androgen refractory state. Endocrinology 142 : 4795-4805. Nabors, L.B., Gillespie, G.Y., Harkins, L., and King, P.H. (2001) HuR, a RNA stability factor, is expressed in malignant brain tumors and binds to adenine- and uridine-rich elements within the 3' untranslated regions of cytokine and angiogenic factor mRNAs. Cancer Research. 61 : 2154-2161. Nabors, L.B., Suswam, E., Huang, Y., Yang, X., Johnson, M.J., and King, P.H. (2003) Tumor necrosis factor alpha Induces angiogenic factor up-regulation in malignant glioma cells: A role for RNA stabilization and HuR. Cancer Research 63 : 4181-4187. Nagafuchi, A. (2001) Molecular architecture of adherens junctions. Current Opinion in Cell Biology 13 : 600-603. Nakanishi, M., Shimada, M., and Niida, H. (2006) Genetic instability in cancer cells by impaired cell cycle checkpoints. Cancer Science 97 : 984-989. Nakayama, K., Nagahama, H., Minamishima, Y.A., Matsumoto, M., Nakamichi, I., Kitagawa, K., Shirane, M., Tsunematsu, R., Tsukiyama, T., Ishida, N., Kitagawa, M., and Hatakeyama, S. (2000) Targeted disruption of Skp2 results in accumulation of cyclin E and p27(Kip1), polyploidy and centrosome overduplication. EMBO Journal 19 : 2069-2081. Nakayama, K.I., and Nakayama, K. (2006) Ubiquitin ligases: cell-cycle control and cancer. Nature Reviews Cancer. 6: 369-381. Nantermet, P.V., Xu, J., Yu, Y., Hodor, P., Holder, D., Adamski, S., Gentile, M.A., Kimmel, D.B., Harada, S., Gerhold, D., Freedman, L.P., and Ray, W.J. (2004) Identification of genetic pathways activated by the androgen receptor during
225 the induction of proliferation in the ventral prostate gland. Journal of Biological Chemistry 279 : 1310-1322. Navarro, D., Luzardo, O.P., Fernandez, L., Chesa, N., and Diaz-Chico, B.N. (2002) Transition to androgen-independence in prostate cancer. Journal of Steroid Biochemistry & Molecular Biology. 81 : 191-201. Navone, N.M., Olive, M., Ozen, M., Davis, R., Troncoso, P., Tu, S.M., Johnston, D., Pollack, A., Pathak, S., von Eschenbach, A.C., and Logothetis, C.J. (1997) Establishment of two human prostate cancer cell lines derived from a single bone metastasis. Clinical Cancer Research 3: 2493-2500. Nikolaev, A.Y., Li, M., Puskas, N., Qin, J., and Gu, W. (2003) Parc: a cytoplasmic anchor for p53. Cell. 112 : 29-40. Oh, J.J., Razfar, A., Delgado, I., Reed, R.A., Malkina, A., Boctor, B., and Slamon, D.J. (2006) 3p21.3 tumor suppressor gene H37/Luca15/RBM5 inhibits growth of human lung cancer cells through cell cycle arrest and apoptosis. Cancer Research. 66 : 3419-3427. Okamoto, S., Ji, H., Howie, D., Clarke, K., Gullo, C., Manning, S., Coyle, A.J., and Terhorst, C. (2004) Expression of the SH2D1A gene is regulated by a combination of transcriptional and post-transcriptional mechanisms. European Journal of Immunology 34 : 3176-3186. Okano, H.J., and Darnell, R.B. (1997) A hierarchy of Hu RNA binding proteins in developing and adult neurons. Journal of Neuroscience 17 : 3024-3037. Pan, Y.X., Chen, H., and Kilberg, M.S. (2005) Interaction of RNA-binding proteins HuR and AUF1 with the human ATF3 mRNA 3'-untranslated region regulates its amino acid limitation-induced stabilization. Journal of Biological Chemistry 280 : 34609-34616. Parker, R., and Song, H. (2004) The enzymes and control of eukaryotic mRNA turnover. Nat Struct Mol Biol 11 : 121-127. Park-Lee, S., Kim, S., and Laird-Offringa, I.A. (2003) Characterization of the interaction between neuronal RNA-binding protein HuD and AU-rich RNA. Journal of Biological Chemistry 278 : 39801-39808. Peng, S.S., Chen, C.Y., and Shyu, A.B. (1996) Functional characterization of a non- AUUUA AU-rich element from the c-jun proto-oncogene mRNA: evidence for a novel class of AU-rich elements. Molecular & Cellular Biology 16 : 1490-1499. Peng, S.S., Chen, C.Y., Xu, N., and Shyu, A.B. (1998) RNA stabilization by the AU- rich element binding protein, HuR, an ELAV protein. EMBO Journal. 17 : 3461- 3470. Perrone-Bizzozero, N., and Bolognani, F. (2002) Role of HuD and other RNA-binding proteins in neural development and plasticity. Journal of Neuroscience Research. 68 : 121-126. Petit, V., and Thiery, J.-P. (2000) Focal adhesions: structure and dynamics. Biology of the Cell 92 : 477-494. Petre-Draviam, C.E., Williams, E.B., Burd, C.J., Gladden, A., Moghadam, H., Meller, J., Diehl, J.A., and Knudsen, K.E. (2005) A central domain of cyclin D1 mediates nuclear receptor corepressor activity. Oncogene 24 : 431-444. Piecyk, M., Wax, S., Beck, A.R., Kedersha, N., Gupta, M., Maritim, B., Chen, S., Gueydan, C., Kruys, V., Streuli, M., and Anderson, P. (2000) TIA-1 is a translational silencer that selectively regulates the expression of TNF-alpha. EMBO Journal 19 : 4154-4163. Polakis, P. (2000) Wnt signaling and cancer. Genes Dev. 14 : 1837-1851. Poukka, H., Karvonen, U., Janne, O.A., and Palvimo, J.J. (2000) Covalent modification of the androgen receptor by small ubiquitin-like modifier 1 (SUMO-1). Proceedings of the National Academy of Sciences of the United States of America 97 : 14145-14150. Raineri, I., Wegmueller, D., Gross, B., Certa, U., and Moroni, C. (2004) Roles of AUF1 isoforms, HuR and BRF1 in ARE-dependent mRNA turnover studied by RNA interference. Nucleic Acids Research 32 : 1279-1288.
226 Rauhala, H.E., Porkka, K.P., Tolonen, T.T., Martikainen, P.M., Tammela, T.L., and Visakorpi, T. (2005) Dual-specificity phosphatase 1 and serum/glucocorticoid- regulated kinase are downregulated in prostate cancer. International Journal of Cancer 117 : 738-745. Reimann, I., Huth, A., Thiele, H., and Thiele, B.J. (2002) Suppression of 15- lipoxygenase synthesis by hnRNP E1 is dependent on repetitive nature of LOX mRNA 3'-UTR control element DICE. Journal of Molecular Biology 315 : 965- 974. Rennie, P.S., Bruchovsky, N., Leco, K.J., Sheppard, P.C., McQueen, S.A., Cheng, H., Snoek, R., Hamel, A., Bock, M.E., MacDonald, B.S., and et al. (1993) Characterization of two cis-acting DNA elements involved in the androgen regulation of the probasin gene. Molecular Endocrinology 7: 23-36. Rennie, P.S., and Nelson, C.C. (1998) Epigenetic mechanisms for progression of prostate cancer. Cancer & Metastasis Reviews 17 : 401-409. Reutens, A.T., Fu, M., Wang, C., Albanese, C., McPhaul, M.J., Sun, Z., Balk, S.P., Janne, O.A., Palvimo, J.J., and Pestell, R.G. (2001) Cyclin D1 binds the androgen receptor and regulates hormone-dependent signaling in a p300/CBP-associated factor (P/CAF)-dependent manner. Molecular Endocrinology 15 : 797-811. Ricote, M., Garcia-Tunon, I., Bethencourt, F., Fraile, B., Onsurbe, P., Paniagua, R., and Royuela, M. (2006) The p38 transduction pathway in prostatic neoplasia. Journal of Pathology 208 : 401-407. Rochette-Egly, C. (2003) Nuclear receptors: integration of multiple signalling pathways through phosphorylation. Cellular Signalling 15 : 355-366. Rokhlin, O.W., Taghiyev, A.F., Guseva, N.V., Glover, R.A., Chumakov, P.M., Kravchenko, J.E., and Cohen, M.B. (2005) Androgen regulates apoptosis induced by TNFR family ligands via multiple signaling pathways in LNCaP. Oncogene 24 : 6773-6784. Ross, H.J., Sato, N., Ueyama, Y., and Koeffler, H.P. (1991) Cytokine messenger RNA stability is enhanced in tumor cells. Blood 77 : 1787-1795. Rothe, M., Pan, M.G., Henzel, W.J., Ayres, T.M., and Goeddel, D.V. (1995) The TNFR2-TRAF signaling complex contains two novel proteins related to baculoviral inhibitor of apoptosis proteins. Cell. 83 : 1243-1252. Rudel, T., and Bokoch, G.M. (1997) Membrane and morphological changes in apoptotic cells regulated by caspase-mediated activation of PAK2. Science 276 : 1571-1574. Saito, K., Fujiwara, T., Katahira, J., Inoue, K., and Sakamoto, H. (2004) TAP/NXF1, the primary mRNA export receptor, specifically interacts with a neuronal RNA- binding protein HuD. Biochemical & Biophysical Research Communications 321 : 291-297. Santy, L.C., Ravichandran, K.S., and Casanova, J.E. (2005) The DOCK180/Elmo complex couples ARNO-mediated Arf6 activation to the downstream activation of Rac1. Current Biology. 15 : 1749-1754. Schena, M., Shalon, D., Davis, R.W., and Brown, P.O. (1995) Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science. 270 : 467-470. Schmiesing, J.A., Ball, A.R., Jr., Gregson, H.C., Alderton, J.M., Zhou, S., and Yokomori, K. (1998) Identification of two distinct human SMC protein complexes involved in mitotic chromosome dynamics. Proceedings of the National Academy of Sciences of the United States of America. 95 : 12906- 12911. Schrem, H., Klempnauer, J., and Borlak, J. (2004) Liver-enriched transcription factors in liver function and development. Part II: the C/EBPs and D site-binding protein in cell cycle control, carcinogenesis, circadian gene regulation, liver regeneration, apoptosis, and liver-specific gene regulation. Pharmacology Reviews. 56 : 291-330.
227 Schroeder, A., Mueller, O., Stocker, S., Salowsky, R., Leiber, M., Gassmann, M., Lightfoot, S., Menzel, W., Granzow, M., and Ragg, T. (2006) The RIN: an RNA integrity number for assigning integrity values to RNA measurements. BMC Molecular Biology 7: 3. Sebolt-Leopold, J.S. (2000) Development of anticancer drugs targeting the MAP kinase pathway. Oncogene. 19 : 6594-6599. Sekido, Y., Bader, S.A., Carbone, D.P., Johnson, B.E., and Minna, J.D. (1994) Molecular analysis of the HuD gene encoding a paraneoplastic encephalomyelitis antigen in human lung cancer cell lines. Cancer Research 54 : 4988-4992. Seth, D., Shaw, K., Jazayeri, J., and Leedman, P.J. (1999) Complex post- transcriptional regulation of EGF-receptor expression by EGF and TGF-alpha in human prostate cancer cells. British Journal of Cancer. 80 : 657-669. Shariff, A.H., and Ather, M.H. (2006) Neuroendocrine differentiation in prostate cancer. Urology 68: 2-8. Sheflin, L., Keegan, B., Zhang, W., and Spaulding, S.W. (2000) Inhibiting proteasomes in human HepG2 and LNCaP cells increases endogenous androgen receptor levels. Biochemical and Biophysical Research Communications. 276 : 144-150. Sheflin, L.G., Zhang, W., and Spaulding, S.W. (2001) Androgen regulates the level and subcellular distribution of the AU-rich ribonucleic acid-binding protein HuR both in vitro and in vivo. Endocrinology 142 : 2361-2368. Sheflin, L.G., Zou, A.-P., and Spaulding, S.W. (2004) Androgens regulate the binding of endogenous HuR to the AU-rich 3'UTRs of HIF-1[alpha] and EGF mRNA. Biochemical and Biophysical Research Communications 322 : 644-651. Shu, H.B., Takeuchi, M., and Goeddel, D.V. (1996) The tumor necrosis factor receptor 2 signal transducers TRAF2 and c-IAP1 are components of the tumor necrosis factor receptor 1 signaling complex. Proceedings of the National Academy of Sciences of the United States of America. 93 : 13973-13978. Shyu, A.B., Greenberg, M.E., and Belasco, J.G. (1989) The c-fos transcript is targeted for rapid decay by two distinct mRNA degradation pathways. Genes & Development 3: 60-72. Shyu, A.B., Belasco, J.G., and Greenberg, M.E. (1991) Two distinct destabilizing elements in the c-fos message trigger deadenylation as a first step in rapid mRNA decay. Genes & Development 5: 221-231. Signoretti, S., Di Marcotullio, L., Richardson, A., Ramaswamy, S., Isaac, B., Rue, M., Monti, F., Loda, M., and Pagano, M. (2002) Oncogenic role of the ubiquitin ligase subunit Skp2 in human breast cancer. Journal of Clinical Investigation 110 : 633-641. Sijin, L., Ziwei, C., Yajun, L., Meiyu, D., Hongwei, Z., Guofa, H., Siguo, L., Hong, G., Zhihong, Z., Xiaolei, L., Yingyun, W., Yan, X., and Weide, L. (2004) The effect of knocking-down nucleostemin gene expression on the in vitro proliferation and in vivo tumorigenesis of HeLa cells. Journal of Experimental & Clinical Cancer Research. 23 : 529-538. Silvera, D., Gamarnik, A.V., and Andino, R. (1999) The N-terminal K homology domain of the poly(rC)-binding protein is a major determinant for binding to the poliovirus 5'-untranslated region and acts as an inhibitor of viral translation. Journal of Biological Chemistry 274 : 38163-38170. Smith, C.L., and O'Malley, B.W. (2004) Coregulator function: a key to understanding tissue specificity of selective receptor modulators. Endocrine Reviews 25 : 45- 71. Soller, M. (2006) Pre-messenger RNA processing and its regulation: a genomic perspective. Cellular & Molecular Life Sciences 63 : 796-819. Sommer, S., Cui, Y., Brewer, G., and Fuqua, S.A. (2005) The c-Yes 3'-UTR contains adenine/uridine-rich elements that bind AUF1 and HuR involved in mRNA
228 decay in breast cancer cells. Journal of Steroid Biochemistry & Molecular Biology 97 : 219-229. Song, I.S., Tatebe, S., Dai, W., and Kuo, M.T. (2005) Delayed mechanism for induction of gamma-glutamylcysteine synthetase heavy subunit mRNA stability by oxidative stress involving p38 mitogen-activated protein kinase signaling. Journal of Biological Chemistry 280 : 28230-28240. Song, L.N., and Gelmann, E.P. (2005) Interaction of beta-catenin and TIF2/GRIP1 in transcriptional activation by the androgen receptor. Journal of Biological Chemistry 280 : 37853-37867. Sramkoski, R.M., Pretlow, T.G., 2nd, Giaconia, J.M., Pretlow, T.P., Schwartz, S., Sy, M.S., Marengo, S.R., Rhim, J.S., Zhang, D., and Jacobberger, J.W. (1999) A new human prostate carcinoma cell line, 22Rv1. In Vitro Cellular & Developmental Biology 35 : 403-409. Stanford, J.L., Just, J.J., Gibbs, M., Wicklund, K.G., Neal, C.L., Blumenstein, B.A., and Ostrander, E.A. (1997) Polymorphic repeats in the androgen receptor gene: molecular markers of prostate cancer risk. Cancer Research 57 : 1194- 1198. Stegh, A.H., Barnhart, B.C., Volkland, J., Algeciras-Schimnich, A., Ke, N., Reed, J.C., and Peter, M.E. (2002) Inactivation of caspase-8 on mitochondria of Bcl-xL- expressing MCF7-Fas cells: role for the bifunctional apoptosis regulator protein. Journal of Biological Chemistry. 277 : 4351-4360. Steller, U., Kohls, S., Muller, B., Soller, R., Muller, R., Schlender, J., and Blohm, D.H. (1996) The RNA binding protein HuD: rat cDNA and analysis of the alternative spliced mRNA in neuronal differentiating cell lines P19 and PC12. Molecular Brain Research 35 : 285-296. Stickens, D., Clines, G., Burbee, D., Ramos, P., Thomas, S., Hogue, D., Hecht, J.T., Lovett, M., and Evans, G.A. (1996) The EXT2 multiple exostoses gene defines a family of putative tumour suppressor genes. Nat Genet. 14 : 25-32. Stofega, M.R., Sanders, L.C., Gardiner, E.M., and Bokoch, G.M. (2004) Constitutive p21-activated kinase (PAK) activation in breast cancer cells as a result of mislocalization of PAK to focal adhesions. Molecular Biology of the Cell. 15 : 2965-2977. Stone, D.M., Hynes, M., Armanini, M., Swanson, T.A., Gu, Q., Johnson, R.L., Scott, M.P., Pennica, D., Goddard, A., Phillips, H., Noll, M., Hooper, J.E., de Sauvage, F., and Rosenthal, A. (1996) The tumour-suppressor gene patched encodes a candidate receptor for Sonic hedgehog. Nature. 384 : 129-134. Stoughton, R.B. (2005) Applications of DNA microarrays in biology. Annual Reveiws in Biochemistry 74 : 53-82. Su, A.I., Cooke, M.P., Ching, K.A., Hakak, Y., Walker, J.R., Wiltshire, T., Orth, A.P., Vega, R.G., Sapinoso, L.M., Moqrich, A., Patapoutian, A., Hampton, G.M., Schultz, P.G., and Hogenesch, J.B. (2002) Large-scale analysis of the human and mouse transcriptomes. Proceedings of the National Academy of Sciences of the United States of America : 012025199. Suk, K., and Erickson, K.L. (1996) Differential regulation of tumour necrosis factor- alpha mRNA degradation in macrophages by interleukin-4 and interferon- gamma. Immunology 87 : 551-558. Sully, G., Dean, J.L., Wait, R., Rawlinson, L., Santalucia, T., Saklatvala, J., and Clark, A.R. (2004) Structural and functional dissection of a conserved destabilizing element of cyclo-oxygenase-2 mRNA: evidence against the involvement of AUF-1 [AU-rich element/poly(U)-binding/degradation factor-1], AUF-2, tristetraprolin, HuR (Hu antigen R) or FBP1 (far-upstream-sequence-element- binding protein 1). Biochemical Journal 377 : 629-639. Suswam, E.A., Nabors, L.B., Huang, Y., Yang, X., and King, P.H. (2005) IL-1beta induces stabilization of IL-8 mRNA in malignant breast cancer cells via the 3' untranslated region: Involvement of divergent RNA-binding factors HuR, KSRP and TIAR. International Journal of Cancer 113 : 911-919.
229 Szabo, A., Dalmau, J., Manley, G., Rosenfeld, M., Wong, E., Henson, J., Posner, J.B., and Furneaux, H.M. (1991) HuD, a paraneoplastic encephalomyelitis antigen, contains RNA-binding domains and is homologous to Elav and Sex-lethal. Cell 67 : 325-333. Tan, J., Sharief, Y., Hamil, K.G., Gregory, C.W., Zang, D.Y., Sar, M., Gumerlock, P.H., deVere White, R.W., Pretlow, T.G., Harris, S.E., Wilson, E.M., Mohler, J.L., and French, F.S. (1997) Dehydroepiandrosterone activates mutant androgen receptors expressed in the androgen-dependent human prostate cancer xenograft CWR22 and LNCaP cells. Molecular Endocrinology 11 : 450- 459. Taneja, S.S., Ha, S., Swenson, N.K., Torra, I.P., Rome, S., Walden, P.D., Huang, H.Y., Shapiro, E., Garabedian, M.J., and Logan, S.K. (2004) ART-27, an androgen receptor coactivator regulated in prostate development and cancer. Journal of Biological Chemistry 279 : 13944-13952. Taplin, M.E., Bubley, G.J., Ko, Y.J., Small, E.J., Upton, M., Rajeshkumar, B., and Balk, S.P. (1999) Selection for androgen receptor mutations in prostate cancers treated with androgen antagonist. Cancer Research 59 : 2511-2515. Taplin, M.E., and Balk, S.P. (2004) Androgen receptor: a key molecule in the progression of prostate cancer to hormone independence. Journal of Cellular Biochemistry 91 : 483-490. Taplin, M.E., George, D.J., Halabi, S., Sanford, B., Febbo, P.G., Hennessy, K.T., Mihos, C.G., Vogelzang, N.J., Small, E.J., and Kantoff, P.W. (2005) Prognostic significance of plasma chromogranin a levels in patients with hormone- refractory prostate cancer treated in Cancer and Leukemia Group B 9480 study. Urology 66 : 386-391. Tepper, C.G., Boucher, D.L., Ryan, P.E., Ma, A.H., Xia, L., Lee, L.F., Pretlow, T.G., and Kung, H.J. (2002) Characterization of a novel androgen receptor mutation in a relapsed CWR22 prostate cancer xenograft and cell line. Cancer Research 62 : 6606-6614. Terry, S., Yang, X., Chen, M.-W., Vacherot, F., and Buttyan, R. (2006) Multifaceted interaction between the androgen and Wnt signaling pathways and the implication for prostate cancer. Journal of Cellular Biochemistry 99 : 402-410. Theodorescu, D., Broder, S.R., Boyd, J.C., Mills, S.E., and Frierson, H.F., Jr. (1997) Cathepsin D and chromogranin A as predictors of long term disease specific survival after radical prostatectomy for localized carcinoma of the prostate. Cancer 80 : 2109-2119. Thomson, A.M., Rogers, J.T., and Leedman, P.J. (1999) Iron-regulatory proteins, iron- responsive elements and ferritin mRNA translation. International Journal of Biochemistry & Cell Biology 31 : 1139-1152. Thomson, A.M., Cahill, C.M., Cho, H.H., Kassachau, K.D., Epis, M.R., Bridges, K.R., Leedman, P.J., Rogers, J.T. (2005) The acute box cis-element in human heavy ferritin mRNA 5'-untranslated region is a unique translation enhancer that binds poly(C)-binding proteins. Journal of Biological Chemistry. 280 : 30032-30045. Trachtenberg, J., and Blackledge, G. (2002) Looking to the Future: Advances in the Management of Hormone-Refractory Prostate Cancer. European Urology Supplements 1: 44-53. Trovo-Marqui, A., and Tajara, E. (2006) Neurofibromin: a general outlook. Clinical Genetics 70 : 1-13. Ueda, T., Mawji, N.R., Bruchovsky, N., and Sadar, M.D. (2002) Ligand-independent activation of the androgen receptor by interleukin-6 and the role of steroid receptor coactivator-1 in prostate cancer cells. Journal of Biological Chemistry 277 : 38087-38094. Uren, A.G., Pakusch, M., Hawkins, C.J., Puls, K.L., and Vaux, D.L. (1996) Cloning and expression of apoptosis inhibitory protein homologs that function to inhibit apoptosis and/or bind tumor necrosis factor receptor-associated factors.
230 Proceedings of the National Academy of Sciences of the United States of America 93 : 4974-4978. Urlinger, S., Baron, U., Thellmann, M., Hasan, M.T., Bujard, H., and Hillen, W. (2000) Exploring the sequence space for tetracycline-dependent transcriptional activators: novel mutations yield expanded range and sensitivity. Proceedings of the National Academy of Sciences of the United States of America 97 : 7963-7968. van Bokhoven, A., Varella-Garcia, M., Korch, C., Johannes, W.U., Smith, E.E., Miller, H.L., Nordeen, S.K., Miller, G.J., and Lucia, M.S. (2003) Molecular characterization of human prostate carcinoma cell lines. Prostate 57 : 205-225. van der Giessen, K., Di-Marco, S., Clair, E., and Gallouzi, I.E. (2003) RNAi-mediated HuR depletion leads to the inhibition of muscle cell differentiation. Journal of Biological Chemistry 278 : 47119-47128. Vanderklish, P.W., and Edelman, G.M. (2005) Differential translation and fragile X syndrome. Genes Brain and Behaviour. 4: 360-384. Venables, J.P. (2006) Unbalanced alternative splicing and its significance in cancer. BioEssays 28 : 378-386. Wagner, B.J., DeMaria, C.T., Sun, Y., Wilson, G.M., and Brewer, G. (1998) Structure and genomic organization of the human AUF1 gene: alternative pre-mRNA splicing generates four protein isoforms. Genomics 48 : 195-202. Wakamatsu, Y., and Weston, J.A. (1997) Sequential expression and role of Hu RNA- binding proteins during neurogenesis. Development. 124 : 3449-3460. Wang, E., Marcotte, R., and Petroulakis, E. (1999) Signaling pathway for apoptosis: a racetrack for life or death. Journal of Cellular Biochemistry. Suppl 33 : 95-102. Wang, J.G., Collinge, M., Ramgolam, V., Ayalon, O., Fan, X.C., Pardi, R., and Bender, J.R. (2006) LFA-1-Dependent HuR Nuclear Export and Cytokine mRNA Stabilization in T Cell Activation. Journal of Immunology 176 : 2105- 2113. Wang, W., Caldwell, M.C., Lin, S., Furneaux, H., and Gorospe, M. (2000a) HuR regulates cyclin A and cyclin B1 mRNA stability during cell proliferation. EMBO Journal 19 : 2340-2350. Wang, W., Furneaux, H., Cheng, H., Caldwell, M.C., Hutter, D., Liu, Y., Holbrook, N., and Gorospe, M. (2000b) HuR regulates p21 mRNA stabilization by UV light. Molecular & Cellular Biology 20 : 760-769. Wang, W., Fan, J., Yang, X., Furer-Galban, S., Lopez de Silanes, I., von Kobbe, C., Guo, J., Georas, S.N., Foufelle, F., Hardie, D.G., Carling, D., and Gorospe, M. (2002) AMP-activated kinase regulates cytoplasmic HuR. Molecular & Cellular Biology 22 : 3425-3436. Wang, W.H., Gregori, G., Hullinger, R.L., and Andrisani, O.M. (2004) Sustained activation of p38 mitogen-activated protein kinase and c-Jun N-terminal kinase pathways by hepatitis B virus X protein mediates apoptosis via induction of Fas/FasL and tumor necrosis factor (TNF) receptor 1/TNF-alpha expression. Molecular and Cellular Biology. 24 : 10352-10365. Wang, X., and Tanaka-Hall, T.M. (2001) Structural basis for recognition of AU-rich element RNA by the HuD protein. Nature Structural Biology 8: 141-145. Weinstein, M.H., Partin, A.W., Veltri, R.W., and Epstein, J.I. (1996) Neuroendocrine differentiation in prostate cancer: enhanced prediction of progression after radical prostatectomy. Human Pathology 27 : 683-687. Wen, Y., Hu, M.C., Makino, K., Spohn, B., Bartholomeusz, G., Yan, D.H., and Hung, M.C. (2000) HER-2/neu promotes androgen-independent survival and growth of prostate cancer cells through the Akt pathway. Cancer Research. 60 : 6841- 6845. Westmark, C.J., Bartleson, V.B., and Malter, J.S. (2005) RhoB mRNA is stabilized by HuR after UV light. Oncogene 24 : 502-511. Wilkinson, M.G., and Millar, J.B. (2000) Control of the eukaryotic cell cycle by MAP kinase signaling pathways. Faseb Journal 14 : 2147-2157.
231 Winzen, R., Gowrishankar, G., Bollig, F., Redich, N., Resch, K., and Holtmann, H. (2004) Distinct domains of AU-rich elements exert different functions in mRNA destabilization and stabilization by p38 mitogen-activated protein kinase or HuR. Molecular & Cellular Biology 24 : 4835-4847. Wolf, D., Herzinger, T., Hermeking, H., Blaschke, D., and Horz, W. (1993) Transcriptional and posttranscriptional regulation of human androgen receptor expression by androgen. Molecular Endocrinology 7: 924-936. Wu, J.D., Haugk, K., Woodke, L., Nelson, P., Coleman, I., and Plymate, S.R. (2006) Interaction of IGF signaling and the androgen receptor in prostate cancer progression. Journal of Cellular Biochemistry 99 : 392-401. Xiao, D., Qu, X., and Weber, H.C. (2002) GRP receptor-mediated immediate early gene expression and transcription factor Elk-1 activation in prostate cancer cells. Regulatory Peptides. 109 : 141-148. Xiao, J., Liu, C.C., Chen, P.L., and Lee, W.H. (2001) RINT-1, a novel Rad50- interacting protein, participates in radiation-induced G(2)/M checkpoint control. Journal of Biological Chemistry. 276 : 6105-6111. Xu, N., Chen, C.Y., and Shyu, A.B. (2001) Versatile role for hnRNP D isoforms in the differential regulation of cytoplasmic mRNA turnover. Molecular & Cellular Biology 21 : 6960-6971. Xu, Y.Z., Di Marco, S., Gallouzi, I., Rola-Pleszczynski, M., and Radzioch, D. (2005) RNA-binding protein HuR is required for stabilization of SLC11A1 mRNA and SLC11A1 protein expression. Molecular & Cellular Biology 25 : 8139-8149. Xu, Z., Sproul, A., Wang, W., Kukekov, N., and Greene, L.A. (2006) Siah1 interacts with the scaffold protein POSH to promote JNK activation and apoptosis. Journal of Biological Chemistry. 281: 303-312. Yang, F., Li, X., Sharma, M., Sasaki, C.Y., Longo, D.L., Lim, B., and Sun, Z. (2002) Linking beta-catenin to androgen-signaling pathway. Journal of Biological Chemistry 277 : 11336-11344. Yang, X., Chen, M.-W., Terry, S., Vacherot, F., Bemis, D.L., Capodice, J., Kitajewski, J., de la Taille, A., Benson, M.C., Guo, Y., and Buttyan, R. (2006) Complex regulation of human androgen receptor expression by Wnt signaling in prostate cancer cells. Oncogene 25 : 3436-3444. Yeap, B.B., Krueger, R.G., and Leedman, P.J. (1999) Differential posttranscriptional regulation of androgen receptor gene expression by androgen in prostate and breast cancer cells. Endocrinology. 140 : 3282-3291. Yeap, B.B. (2000) Post-transcriptional regulation of androgen receptor gene expression in human prostates and breast cancer cells: Identification and characterisation of novel RNA-protein interactions. In University Department of Medicine Perth: University of Western Australia. Yeap, B.B., Voon, D.C., Vivian, J.P., McCulloch, R.K., Thomson, A.M., Giles, K.M., Czyzyk-Krzeska, M.F., Furneaux, H., Wilce, M.C., Wilce, J.A., and Leedman, P.J. (2002) Novel binding of HuR and poly(C)-binding protein to a conserved UC-rich motif within the 3'-untranslated region of the androgen receptor messenger RNA. Journal of Biological Chemistry. 277 : 27183-27192. Yeh, S., Kang, H.Y., Miyamoto, H., Nishimura, K., Chang, H.C., Ting, H.J., Rahman, M., Lin, H.K., Fujimoto, N., Hu, Y.C., Mizokami, A., Huang, K.E., and Chang, C. (1999a) Differential induction of androgen receptor transactivation by different androgen receptor coactivators in human prostate cancer DU145 cells. Endocrine 11 : 195-202. Yeh, S., Lin, H.K., Kang, H.Y., Thin, T.H., Lin, M.F., and Chang, C. (1999b) From HER2/Neu signal cascade to androgen receptor and its coactivators: a novel pathway by induction of androgen target genes through MAP kinase in prostate cancer cells. Proceedings of the National Academy of Sciences of the United States of America. 96 : 5458-5463. Yoon, H.G., and Wong, J. (2006) The corepressors silencing mediator of retinoid and thyroid hormone receptor and nuclear receptor corepressor are involved in
232 agonist- and antagonist-regulated transcription by androgen receptor. Molecular Endocrinology 20 : 1048-1060. Yordy, J.S., and Muise-Helmericks, R.C. (2000) Signal transduction and the Ets family of transcription factors. Oncogene 19 : 6503-6513. Yu, J., de Belle, I., Liang, H., and Adamson, E.D. (2004) Coactivating factors p300 and CBP are transcriptionally crossregulated by Egr1 in prostate cells, leading to divergent responses. Molecular Cell 15 : 83-94. Zegarra-Moro, O.L., Schmidt, L.J., Huang, H., and Tindall, D.J. (2002) Disruption of androgen receptor function inhibits proliferation of androgen-refractory prostate cancer cells. Cancer Research 62 : 1008-1013. Zhang, H., Xu, Q., Krajewski, S., Krajewska, M., Xie, Z., Fuess, S., Kitada, S., Pawlowski, K., Godzik, A., and Reed, J.C. (2000) BAR: An apoptosis regulator at the intersection of caspases and Bcl-2 family proteins. Proceedings of the National Academy of Sciences of the United States of America. 97 : 2597- 2602. Zhang, W., Wagner, B.J., Ehrenman, K., Schaefer, A.W., DeMaria, C.T., Crater, D., DeHaven, K., Long, L., and Brewer, G. (1993) Purification, characterization, and cDNA cloning of an AU-rich element RNA-binding protein, AUF1. Molecular & Cellular Biology 13 : 7652-7665. Zhang, Y., Bhatia, D., Xia, H., Castranova, V., Shi, X., and Chen, F. (2006) Nucleolin links to arsenic-induced stabilization of GADD45alpha mRNA. Nucleic Acids Research 34 : 485-495. Zhao, X.Y., Boyle, B., Krishnan, A.V., Navone, N.M., Peehl, D.M., and Feldman, D. (1999) Two mutations identified in the androgen receptor of the new human prostate cancer cell line MDA PCa 2a. Journal of Urology 162 : 2192-2199. Zhao, X.Y., Malloy, P.J., Krishnan, A.V., Swami, S., Navone, N.M., Peehl, D.M., and Feldman, D. (2000a) Glucocorticoids can promote androgen-independent growth of prostate cancer cells through a mutated androgen receptor.[see comment][erratum appears in Nat Med 2000 Aug;6(8):939]. Nature Medicine 6: 703-706. Zhao, Z., Chang, F.C., and Furneaux, H.M. (2000b) The identification of an endonuclease that cleaves within an HuR binding site in mRNA. Nucleic Acids Research 28 : 2695-2701. Zhivotovsky, B., and Orrenius, S. (2006) Carcinogenesis and apoptosis: paradigms and paradoxes. Carcinogenesis 27 : 1939-1945. Zhou, Z., Kemppainen, J., and Wilson, E. (1995) Identification of three proline-directed phosphorylation sites in the human androgen receptor. Molecular Endocrinology 9: 605-615. Zhu, H., Hasman, R.A., Barron, V.A., Luo, G., and Lou, H. (2006) A Nuclear Function of Hu Proteins as Neuron-specific Alternative RNA Processing Regulators. Molecular Biology of the Cell 11 : 11. Zhu, W., Brauchle, M.A., Di Padova, F., Gram, H., New, L., Ono, K., Downey, J.S., and Han, J. (2001a) Gene suppression by tristetraprolin and release by the p38 pathway. American Journal of Physiology Lung Cellular & Molecular Physiology 281 . Zhu, Z., Becklin, R.R., Desiderio, D.M., and Dalton, J.T. (2001b) Identification of a novel phosphorylation site in human androgen receptor by mass spectrometry. Biochemical & Biophysical Research Communications 284 : 836-844. Zou, T., Mazan-Mamczarz, K., Rao, J.N., Liu, L., Marasa, B.S., Zhang, A.H., Xiao, L., Pullmann, R., Gorospe, M., and Wang, J.Y. (2006) Polyamine depletion increases cytoplasmic levels of RNA-binding protein HuR leading to stabilization of nucleophosmin and p53 mRNAs. Journal of Biological Chemistry 281 : 19387-19394. Zubiaga, A.M., Belasco, J.G., and Greenberg, M.E. (1995) The nonamer UUAUUUAUU is the key AU-rich sequence motif that mediates mRNA degradation. Molecular & Cellular Biology 15 : 2219-2230.
233 Zuker, M. (2003) Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Research 31 : 3406-3415.
234 Appendix 1: Buffers and Solutions.
1. Actinomycin D Add 1 mL of sterile PBS 37 to 1 mg vial of Actinomycin D. Store at 4 °C for up to 2 weeks.
2. Agarose Gel 1 % w/v Add 1 g of agarose to 100 mL of 1 x Tris acetate EDTA (TAE) 47 or 0.5x Tris Borate EDTA (TBE) 48 . 1.5 % w/v Add 1.5 g of agarose to 100 mL of 1x TAE 47 . 3 % w/v Add 3 g of agarose to 100 mL of 1x TAE. Microwave ingredients to dissolve agarose and add 5 µL of 10 mg/mL of ethidium bromide 17 . Prior to use, melt in microwave, cool to ~ 70 °C and pour into casting tray. Store at room temperature.
3. APS (10 % w/v)
Weigh 100 mg of APS and add 1 mL of ddH 20. Make fresh before use.
4. Ampicillin (100 mg/mL)
Dissolve 1 g vial of ampicillin in 10 mL of ddH 20. Aliquot and store at –20 °C for up to 3 months.
5. Aprotinin (1 mg/mL)
Weigh 10 mg of aprotinin and add 10 mL of ddH 20. Aliquot and store at –20 °C for up to 3 months.
6. Bicalutamide
Bicalutamide is almost insoluble in water (<0.005 mg/mL or ~ 1 µM) but soluble in acetone. Weigh ~ 50 mg of bicalutamide and make to 5 mg/mL in ~ 5 mL of acetone
234 (11.6 mM). Dilute 86.2 µL with 913.8 µL of acetone to make 1 mM. Add to cell culture as appropriate. Store at – 20 °C for up to 6 months.
7. BSA (Bovine Serum Albumin) 1 % w/v Dissolve 500 mg of BSA in 50 mL of PBS 37 . Store at 4 °C for up to a week. 0.4 µg/ µL for qPCR
Dilute 40 µL of 100 mg/mL BSA stock (Promega) in 10 mL of sterile ddH 20. Aliquot and store at –20 °C.
8. CEB (Cytoplasmic Extraction Buffer) Final Concentration: HEPES 5 mL of 1 M (pH 7.6) 22 10 mM KCl 2 mL of 1 M 25 40 mM 30 MgCl 2 150 µL of 250 mM 3 mM Glycerol 2.5 mL of 100 % 5 %(v/v) Igepal-CA 630 100 µL of 100 % 0.2 %(v/v)
Mix ingredients and add to final volume of 46 mL in ddH 20. Store –20 °C in 4.6 mL aliquots and add protease inhibitors 41 prior to use.
9. DEPC-Treated ddH 20
Add 1 mL of DEPC to 1 L ddH 20. Leave overnight with stirring, then autoclave. Store at room temperature.
10. Dextran-Coated Charcoal 2.5 g of Dextran T-70 25 g of acid-washed charcoal
Add to 500 mL of ddH 20 and incubate overnight at 4 °C with stirring. Recover from suspension by centrifugation at 4 °C for 20 min at 1 800 x g. Wash twice by resuspension in fresh ddH 20 and re-centrifuge for 10 min. Use immediately for charcoal-stripping of FCS (Section 2.7.2).
235 11. DHT (4,5 α-Dihydrotestosterone)
Weigh 50 mg of DHT and dissolve in 10 mL of 100 % ethanol (17.2 mM). Dilute 5.8 µL in 994.2 µL of 100 % ethanol to make 100 µM. Dilute serially in ddH 20 to make 10 µM and 1 µM. Store at –20 °C for up to 6 months.
12. DTT (1 M)
Dissolve 1.54 g of DTT in 10 mL of ddH 20. Aliquot and store at –20 °C indefinitely.
13. DNA Loading Dye (10 x)
Weigh 0.2 g of Orange G dye and dissolve in 10 mL of ddH 2O. Add 1 mL of this solution to the following and mix well:
Glycerol 5 mL EDTA (500 mM 15 ) 1 mL Orange G solution 1 mL ddH 2O 3 mL Total volume: 10 mL Store at room temperature.
14. dNTP’s 10 mM dNTPs
Add 10 µL of each 100 mM stock of dTTP, dATP, dCTP and dGTP to 60 µL of ddH 20. Aliquot and store at –20 °C. 2.5 mM dNTPs
Add 2.5 µL of each 100 mM stock of dTTP, dATP, dCTP and dGTP to 90 µL of ddH 20. Aliquot and store at –20 °C.
15. EDTA (500 mM)
Add 186.17 g of EDTA to 980 mL ddH 20. Adjust pH to 8.0 with 10 M NaOH, make to final volume of 1 L with ddH 2O and autoclave. Store at room temperature.
16. Electrode Buffer (5 x) Final concentration: Tris.HCl 14.9 g 0.123 M Glycine 72.1 g 0.96 M SDS 4.9 g 17 mM
Dissolve ingredients in ~980 mL of ddH 20. Adjust pH to 8.0 with 10 M NaOH and make to 1 L with ddH 20. Store at room temperature. 236
17. Ethidium Bromide (10 mg/mL)
Dissolve 1 g of ethidium bromide in 100 mL of ddH 20. Store at room temperature protected from light.
18. Fibronectin-Coated Coverslips Incubated sterile coverslips in 9 cm 2 dishes with 1 mL of human fibronectin (1 mg/mL in sterile PBS 37 ) for 3 h at 37 °C. Aspirate excess fibronectin and air-dry the cover slips before use.
19. Formaldehyde (4 % ) Dilute 5 mL of 40 % v/v formaldehyde stock in 50 mL of PBS 37 . Make fresh before use.
20. GST Elution Buffer Final concentration: Tris 1 mL of 500 mM (pH 8.0) 46 50 mM Reduced glutathione 76 mg 25 mM 46 Dissolve reduced glutathione in 9 mL of ddH 20, add 1 mL of 500 mM Tris (pH 8.0) to make a final volume of 10 mL. Make fresh each time and store at 4 °C.
21. Heparin (50 mg/mL)
Dissolve 100 mg vial of heparin in 2 mL of DEPC-treated ddH 20. Aliquot and store at –20 °C.
22. HEPES (1 M)
Dissolve 238.3 g of HEPES free acid in 1 L of ddH 20. Adjust pH to 7.6 with NaOH and autoclave.
23. Hoechst 33258 (10 mg/mL)
Dissolve 100 mg of Hoechst 33258 in 10 mL ddH 20. Store at 4 °C.
237 24. IPTG (100 mM)
Dissolve 238.8 mg of IPTG in 10 mL of ddH 20. Store at –20 °C in aliquots.
25. KCl (1 M)
Dissolve 74.6 g of KCl in 1 L of ddH 20. Store at room temperature.
26. Leupeptin (1 mg/mL)
Dissolve 10 mg of leupeptin in 10 mL of ddH 20 and store at –20 °C.
27. LB (Luria Bertani)-Agar Dissolve 0.75 g of bacteriological agar in 500 mL of LB broth 28 and autoclave. Store in the dark at room temperature. To use LB-agar, melt in a microwave, cool to ~70 °C, add 1 µL of 100 mg/mL ampicillin 4 per mL of agar before use (if required) and pour into plates.
28. LB (Luria Bertani) Broth Final concentration: Tryptone 10 g 10 % w/v Yeast extract 5 g 5 % w/v NaCl 5 g 5 % w/v
Dissolve ingredients in 1 L of ddH 20 and autoclave. Store in the dark at room temperature. Add 1 µL of 100 mg/mL ampicillin 4 per mL of agar before use (if required).
29. Mid-RIPA Final concentration: NaCl 1.5 mL of 5 M stock 31 150 mM Tris pH 8.0 3 mL of 500 mM stock 46 25 mM Igepal-CA 630 0.5 mL of 100 % 1 % (v/v) Deoxycholate 0.25 g 0.12 mM
Mix ingredients and make up to a final volume of 46 mL with ddH 20. Store at –20 °C in 4.6 mL aliquots. Add 400 µL of protease inhibitors 41 before use.
238 30. MgCl 2 (250 mM)
Dissolve 23.83 g in 250 mL of ddH 20 and store at room temperature.
31. NaCl (5 M)
Dissolve 292.5 g in 1 L of ddH 20 and store at room temperature.
32. NET Buffer Final concentration: NaCl 20 mL of 5 M 31 100 mM EDTA pH 8.0 2 mL of 500 mM 15 1 mM Tris pH 8.0 40 mL of 500 mM 46 20 mM
Mix ingredients and make up to 1 L with ddH 20 then autoclave. Add protease inhibitors below and store at 4 °C.
Final concentration: Triton X-100 1 mL of 100 % 0.1% Aprotinin 1 mL of 1 mg/mL 5 1 µg/mL Leupeptin 5 mL of 1 mg/mL 26 5 µg/mL Pepstatin 700 µL of 1 mg/mL 36 0.7 µg/mL
33. NaF (1 M)
Dissolve 420 mg of NaF in 10 mL of ddH 20 and store at –20 °C in aliquots.
34. NaVO 3 (100 mM)
Dissolve 121.9 mg of NaVO 2 in 10 mL of ddH 20 and store –20 °C in aliquots.
35. NTP’s minus UTP (2.5 mM)
Add 2.5 µL of each 100 mM stock (ATP, CTP, GTP) in 92.5 µL of DEPC-treated ddH 20. Aliquot and store at –20 °C.
36. Pepstatin (1 mg/mL)
Dissolve 10 mg of pepstatin in 10 mL of methanol. Store at –20 °C in aliquots.
239 37. PBS (phosphate-buffered saline)
Dissolve 1 sachet of 1x PBS powder in 980 mL of ddH20 and adjust pH to 7.2 with 10 M HCl. Make to 1 L and autoclave. Store at 4 °C.
38. PMSF (100 mM)
Dissolve 17.4 mg of PMSF in 1 mL of methanol. Store at – 20 °C.
39. Polyacrylamide Gel, Denaturing Resolving gel (10 %), per gel: 3.3 mL 30 % 29:1 acrylamide/bisacrylamide
4.1 mL ddH 2O 2.5 mL Resolving Gel Buffer 43 150 µL 10 % APS 3 15 µL TEMED
Mix ingredients, adding APS and TEMED last. Pour between clean glass plates to 0.5 cm below the final position of the comb. Overlay with ddH 20 and leave 15 min to polymerise, before tipping off ddH 20 and pouring stacking gel.
Stacking gel (6 %), per gel: 1 mL 30 % 29:1 acrylamide/bisacrylamide 2.7 mL water 1.25 mL of Stacking Gel Buffer 45 37.5 µL 10 % APS 3 3.75 µL TEMED Mix ingredients, adding APS and TEMED last. Pour on top of resolving gel to top of glass plates and carefully insert comb. Leave 15 min to polymerise.
40. Polyacrylamide-Urea Gel
Per gel: Dissolve 4.2 g of urea in 8.33 mL of 1 x TBE 48 at 37 °C. Add: 1.67 mL 30 % 19:1 acrylamide/bisacrylamide 100 µL 10 % APS 3 12.5 µL TEMED Pour between clean glass plates, insert comb and leave to polymerise 15 min.
240
41. Protease Inhibitors Final concentration: Aprotinin 10 µL of 1 mg/mL 5 2 µg/mL Pepstatin 3.5 µL of 1 mg/mL 36 0.7 µg/mL Leupeptin 25 µL of 1 mg/mL 26 5 µg/mL EDTA 20 µL of 500 mM 15 2 mM NaF 250 µL of 1 M 33 50 mM
34 NaVO 3 50 µL of 100 mM 1 mM PMSF 50 µL of 100 mM 38 1 mM DTT 5 µL of 1 M 12 1 mM Add to 4.6 mL of mid-RIPA 29 or CEB 8 buffer prior to use.
42. REMSA Loading Dye Final concentration: Formamide 8 mL 80 % v/v EDTA 400 µL of 500 mM 15 20 mM Sterile glycerol 900 µL 100 % 9 % v/v Bromophenol blue 10 mg 0.1% w/v Xylene cyanol 10 mg 0.1 % w/v Mix ingredients and make to 10 mL with 5x TBE 48 . Aliquot and store at –20 °C.
43. Resolving Gel Buffer (1.5 M Tris pH 8.8)
Dissolve 90.9 g of Tris.HCl in 500 mL of ddH 20. Adjust pH to 8.8 with 10 M NaOH. Store at 4 °C.
44. Sodium Acetate (3 M)
Dissolve 82.03 g of sodium acetate in 1 L of ddH 20. Adjust pH to 4.0 with glacial acetic acid and autoclave. Store at 4 °C.
45. Stacking Gel Buffer (0.5 M Tris, pH 6.8)
Dissolve 30.3 g of Tris.HCl in 500 mL of ddH 20. Adjust pH to 6.8 with 10 M NaOH. Store at 4 °C.
241 46. Tris (500 mM, pH 8.0)
Dissolve 60.57 g of Tris.HCl in 980 mL of ddH 20.
Adjust pH to 8.0 with 10 M NaOH, and make up to final volume of 1 L with ddH 20. Store at room temperature.
47. TAE (Tris-Acetate EDTA; 50 x) Final concentration: EDTA 37.2 g 0.1 M Tris 242 g 2 M Glacial acetic acid 57.1 mL 5.71 %
Dissolve ingredients in ddH 20 and make up to 1 L with ddH 2O. Dilute 20 mL in 980 mL of ddH 20 (1x TAE) for use with agarose gel electrophoresis. Store at room temperature.
48. TBE (Tris-Borate EDTA; 10 x) Final concentration: Tris base 108 g 0.89 M Boric acid 55 g 0.89 M EDTA 40 mL of 500 mM 15 0.02 M
Dissolve ingredients in 900 mL of ddH 20 and adjust pH to 8.0 with 10 M HCl. Make up to final volume of 1 L with ddH 20. Store at room temperature. Dilute with ddH 20 as appropriate prior to use.
49. TBS (Tris-Buffered Saline) Final concentration: Tris pH 8.0 100 mL of 500 mM 46 50 mM NaCl 30 mL of 5 M 31 150 mM
Make to 1 L with ddH 20. Store at room temperature.
50. TBS-T (Tris-Buffered Saline-Tween 20) Add 1 mL of Tween-20 to 1 L of TBS 49 . Store at room temperature.
51. Triton X-100 (0.01 % v/v)
Add 500 µL of Triton X-100 to 50 mL of PBS 37 . Dilute 500 µL of this solution in 50 mL of PBS to make a 0.01 % (v/v) solution. Store at room temperature.
242
52. Trypan Blue (0.4 %) Dissolve 40 mg of Trypan blue in 10 mL of PBS 37 . Store at room temperature.
53. WTB (Western Transfer Buffer) Final concentration: Tris 3.03 g 25 mM Glycine 14.4 g 190 mM SDS 2 mL of 10 % 0.2 %(w/v) methanol 100 mL of 100 % 10 %(v/v)
Dissolve Tris and glycine in 900 mL of ddH 20. Add the methanol and SDS and use within one month. Store at room temperature.
54. Yeast tRNA (2 mg/mL)
Dissolve 25 mg vial of yeast tRNA in 12.5 mL of DEPC-treated ddH 20 and aliquot. Store at –20 °C.
243 Appendix 2: Plasmid Vectors.
A
pBS-ARUC-WT, pBS-ARUC-MutA, MCS ARUC sequences: pBS-ARUC-MutB. (3051 bp) WT CTGGGCTTTTTTTTTCTCTTTCTCTCCTTTCTTTTTCTTCTTCCCTCCCTA MutA CTGGGCTTTTTTTTTCTCT GTCTCTCCT GTCT GGG TCTTCTTCCCTCCCTA MutB CTGGGCTT GTGTGTTCTCT GTCTCTCCT GTCT GGG TCTTCTTCCCTCCCTA
ARUC
pBS-ARUC constructs : pBS-ARUC-WT, -MutA and -MutB vectors were generated by ligating ARUC PCR products (Figure 3.4A) with Bam HI/HindIII overhangs into pBluescript KS+, cut with the same enzymes (Yeap et al ., 2002).
B
pGL3-MCS (5334 bp)
Xba I 1934
3’-UTR multiple cloning site (5’-3’) Xba ISpe I BamH I Sma I Pst I EcoR I EcoR V Hind III Cla I Hinc II Xho I Apa I Xba I TCTAGAACTAGTGGATCCCCCGGGCTGCAGGAATTCGATATCAAGCTTATCGATACCGTCGACCTCGAGGGGGGGCCCTCTAGA
pGL3-MCS : Modified pGL3-control luciferase reporter vector with multiple cloning site (MCS) for the analysis of 3’-UTR elements. Giles et al. (2003). Kpn I / Apa I-cut pBluescript II KS+ vector backbone was ligated with a Kpn I / Apa I linker from pcDNA3. An Xba I fragment was then excised from this plasmid and ligated with Xba I-cut pGL3-control vector. The resulting multiple cloning site contains unique Spe I and Apa I sites that simplify directional insertion of p21 3’-UTR elements downstream of the luciferase reporter gene.
244 C
Luc-ARUC-WT, Luc-ARUC-MutA, Luc-ARUC-MutB. (5385 bp)
Xba I 1934
Xba I Spe I BamH I Hind III Cla I Hinc II Xho I Apa I Xba I TCTAGAACTAGTGGATCC ARUC 51 nt AAGCTTATCGATACCGTCGACCTCGAGGGGGGGCCCTCTAGA
ARUC sequences: WT CTGGGCTTTTTTTTTCTCTTTCTCTCCTTTCTTTTTCTTCTTCCCTCCCTA MutA CTGGGCTTTTTTTTTCTCT GTCTCTCCT GTCT GGG TCTTCTTCCCTCCCTA MutB CTGGGCTT GTGTGTTCTCT GTCTCTCCT GTCT GGG TCTTCTTCCCTCCCTA
Luc-ARUC constructs : ARUC-WT, -MutA and –MutB were excised from the pBS-ARUC plasmids (Appendix 2A) with Spe I and Apa I, and ligated into pGL3- MCS (Appendix 2B), cut with the same enzymes.
D
pBS-c-fos -HuD MCS (3027 bp)
c-fos HuD sequence: ATATTTATATTTTTATTTTATTTTTTT
c-fos
pBS-c-fos -HuD : This vector was generated by ligating the indicated c-fos -HuD PCR product with Bam HI/HindIII overhangs into pBluescript KS+, cut with the same enzymes (Giles et al ., 2003). 245 E
Luc-c-fos. (3561 bp)
Xba I 1934
Xba I Spe I BamH I Hind III Cla I Hinc II Xho I Apa I Xba I TCTAGAACTAGTGGATCC c-fos 27 nt AAGCTTATCGATACCGTCGACCTCGAGGGGGGGCCCTCTAGA
c-fos ATATTTATATTTTTATTTTATTTTTTT
Luc-c-fos : The c-fos -HuD sequence was excised from the pBS-c-fos -HuD plasmid (Appendix 2D) with Spe I and Apa I, and ligated into pGL3-MCS (Appendix 2B), cut with the same enzymes.
F HuR
FLAG-HuR (5675 bp) FLAG-empty (4700 bp)
HuR
FLAG-HuR and FLAG-empty : HuR cDNA (bases 3-978) was generated by PCR with Hind III and Sal I-linked primers, then ligated into pFLAG-CMV2 (Sigma), cut with the same enzymes. FLAG-empty was generated by removing the HuR gene from this vector with the same enzymes, end-filling, and blunt-end ligating.
246 G
pGEX-2T-HuD (6013 bp)
pGEX-2T-HuD : Bases 3-1119 (amino-acids 2-373) of HuD were amplified by PCR with Bam HI-linked primers and ligated into pGEX-2T-HuD (GE Healthcare Life Sciences), cut with Bam HI (Chung et al 1996).
H
pGEX-6P-HuD (6013 bp)
pGEX-6P-HuD : Bases 3-1119 (amino-acids 2-373) of HuD were excised from pGEX-2T-HuD (Appendix 2G) and ligated into pGEX-6P-HuD (GE Healthcare Life Sciences), cut with Bam HI .
247 I
FLAG-HuR (5816 bp) FLAG-empty (4700 bp)
FLAG-HuD and FLAG-empty : HuD (bases 3-1119) was cleaved out of pGEX-2T (Appendix 2G) with Bam HI and religated into pFLAG-CMV7.1, cut with the same enzymes. FLAG-empty was unmodified pFLAG-CMV7.1.
J
Hind III Probasin promoter Hind III -244 to -96 -244 to -96 -244 to -96
Xho I PB3-Luc Sac I (pT81-Luc) 5 500 bp Kpn I Sma I Sal I Hind III Bam HI Luciferase TK promoter
PB3-Luc : Three copies of the –244 to –96 promoter region of the rat probasin gene were cloned into the Hind III sites of pT81-Luc. This region of the promoter contains two androgen response elements, and is described in Rennie et al. (1993). Bam HI and Sac I are unique sites suitable for excision of the insert.
248 K
PSA-Luc (5 400 bp)
PSA-Luc : A 496 bp core region of the PSA enhancer was cloned into the pSP72 e4T CAT vector. PSA-Luc was created by PCR subcloning the E4 TATA box and PSA enhancer region and placing it into the Xho I and Sac I sites of the pGL3-basic (Promega) vector.
L
phRL-SV40 Renilla Luciferase . This vector was purchased from Promega and encodes Renilla luciferase under the control of the SV40 early enhancer/promoter.
249 M
pRL-TK Renilla Luciferase. This vector was purchased from Promega and encodes Renilla luciferase under the control of the HSV-TK enhancer/promoter.
250 Appendix 3: HuR-Regulated Genes.
DOWN-REGULATED GENES, Alphabetical by gene name. Ratio p- value Identifier Gene Name 1.32 0.02239 BC004219 1-acylglycerol-3-phosphate O-acyltransferase 3 1.23 0.02495 AF049895 2,4-dienoyl CoA reductase 1, mitochondrial 1.33 0.04028 NM_014362 3-hydroxyisobutyryl-Coenzyme A hydrolase 1.28 0.00305 AB044805 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 2 1.27 0.02867 NM_016821 8-oxoguanine DNA glycosylase 1.31 0.02853 AB019691 A kinase (PRKA) anchor protein (yotiao) 9 1.44 0.03191 NM_014371 A kinase (PRKA) anchor protein 8-like 1.48 0.04313 AI557319 Abhydrolase domain containing 2 1.35 0.03332 NM_001108 acylphosphatase 2, muscle type 1.56 0.04624 AW242701 ADAM metallopeptidase domain 22 1.44 0.04518 AB002364 ADAM metallopeptidase with thrombospondin type 1 motif, 3 1.28 0.00337 BE545756 Adducin 3 (gamma) 1.34 0.02775 AK026966 Adenylate kinase 3-like 1 1.69 0.04529 AF038193 ADP-ribosylation factor-like 3 1.73 0.0282 AI651212 Adrenergic, beta, receptor kinase 2 1.45 0.00023 NM_001633 Alpha-1-microglobulin/bikunin precursor 1.51 0.04227 AW025455 Anaphase promoting complex subunit 10 1.29 0.01577 AU160676 Ankyrin repeat and KH domain containing 1 1.71 0.02922 AU148042 Ankyrin repeat and KH domain containing 1 1.55 0.01611 X80821 Ankyrin repeat domain 12 1.49 0.03773 NM_022096 Ankyrin repeat domain 5 1.42 0.02118 AU132185 Antigen identified by monoclonal antibody Ki-67 1.37 0.04618 BG260519 Aquarius homolog (mouse) 1.49 0.0045 AA977481 ARP3 actin-related protein 3 homolog (yeast) 1.37 0.0147 AI768497 Aryl hydrocarbon receptor nuclear translocator 1.32 0.04003 NM_001166 Baculoviral IAP repeat-containing 2 1.2 0.02004 AI813772 Bardet-Biedl syndrome 4 1.34 0.03036 AI813772 Bardet-Biedl syndrome 4 1.23 0.03717 AW873330 B-cell receptor-associated protein 29 1.41 0.04551 U66879 BCL2-antagonist of cell death 1.71 0.04912 NM_016561 Bifunctional apoptosis regulator 1.34 0.02794 AA573805 Bobby sox homolog (Drosophila) 1.27 0.03084 H09739 BRCA1/BRCA2-containing complex, subunit 3 1.22 0.00725 X95152 breast cancer 2, early onset 1.49 0.03433 AL043482 Breast cancer antigen 32004 mRNA, partial sequence 1.31 0.04182 AW205585 Bromodomain adjacent to zinc finger domain, 2A
251 1.38 0.0273 NM_006696 Bromodomain containing 8 1.58 0.01114 NM_007049 Butyrophilin, subfamily 2, member A1 1.44 0.03835 D87742 C219-reactive peptide 1.25 0.03716 AL575747 Calcium regulated heat stable protein 1, 24kDa 1.61 0.00333 NM_006366 CAP, adenylate cyclase-associated protein, 2 (yeast) 1.83 0.01135 NM_004056 Carbonic anhydrase VIII 1.44 0.049 NM_003793 Cathepsin F 1.27 0.0094 NM_002414 CD99 antigen 1.67 0.03501 NM_017913 CDC37 cell division cycle 37 homolog (S. cerevisiae)-like 1 1.26 0.01638 AI797678 CDNA clone IMAGE:3683736 2.2 0.03047 AA534039 CDNA clone IMAGE:5181407 2.53 0.0393 BF115777 CDNA clone IMAGE:5303499 1.56 0.04724 AK022980 CDNA FLJ12918 fis, clone NT2RP2004580 1.3 0.03362 AV724183 CDNA FLJ31362 fis, clone NB9N41000011 1.45 0.04258 AI982758 CDNA FLJ33653 fis, clone BRAMY2024715 1.26 0.02448 AA609183 CDNA FLJ46097 fis, clone TESTI2021112 1.27 0.0378 NM_001790 Cell division cycle 25C 1.52 0.02653 AB020635 Cell division cycle 26 1.72 0.04627 AB018325 Centaurin, delta 2 1.4 0.0121 BC002703 Centromere protein A, 17kDa 1.28 0.00585 NM_001812 Centromere protein C 1 1.9 0.00227 AW955956 Centrosomal protein 152kDa 1.5 0.00693 AB020719 Centrosomal protein 152kDa 1.62 0.02893 NM_006584 Chaperonin containing TCP1, subunit 6B (zeta 2) 1.7 0.0481 AK025562 Chloride channel 5 (nephrolithiasis 2, X-linked, Dent disease) 1.34 0.02584 AF288208 Choline kinase alpha 1.4 0.02161 NM_018265 Chromosome 1 open reading frame 106 1.38 0.02924 AW512122 Chromosome 1 open reading frame 25 1.28 0.01932 AF288392 Chromosome 1 open reading frame 26 2.35 0.02452 AK025270 Chromosome 10 open reading frame 137 1.41 0.04799 BF110370 Chromosome 12 open reading frame 30 1.99 0.00624 NM_024857 Chromosome 17 open reading frame 41 1.34 0.02079 BE669553 Chromosome 2 open reading frame 26 1.4 0.01751 NM_013310 Chromosome 2 open reading frame 27 1.24 0.02342 NM_018840 Chromosome 20 open reading frame 24 1.27 0.032 NM_018244 Chromosome 20 open reading frame 44 1.3 0.01245 AL117574 Chromosome 20 open reading frame 80 2.95 0.02406 BM194465 Chromosome 21 open reading frame 34 1.71 0.02453 AI394199 Chromosome 4 open reading frame 16 1.37 0.00425 AL110193 Chromosome 9 open reading frame 114
252 1.3 0.00479 BF590861 Chromosome 9 open reading frame 40 1.27 0.00009 AF161411 Chromosome 9 open reading frame 80 1.31 0.00662 NM_003492 Chromosome X open reading frame 12 1.4 0.04074 NM_024810 chromosome X open reading frame 45 1.25 0.00297 AW080832 CKLF-like MARVEL transmembrane domain containing 8 1.32 0.02328 NM_007097 Clathrin, light polypeptide (Lcb) 1.49 0.03088 NM_001324 Cleavage stimulation factor, 3' pre-RNA, subunit 1, 50kDa 1.37 0.04169 AI186145 Clone 24583 mRNA sequence 1.61 0.02086 NM_000130 Coagulation factor V (proaccelerin, labile factor) 1.23 0.01622 AI458313 Coiled-coil domain containing 34 1.59 0.03145 AW183655 CREB binding protein (Rubinstein-Taybi syndrome) 1.29 0.04444 AA682674 Cullin 1 1.27 0.04487 NM_003588 Cullin 4B 1.71 0.02273 AL390176 Cut-like 1, CCAAT displacement protein (Drosophila) 1.63 0.04794 NM_025212 CXXC finger 4 1.89 0.0108 NM_020348 Cyclin M1 1.37 0.00721 NM_012110 Cysteine-rich hydrophobic domain 2 1.88 0.03516 AF142573 Cysteine-rich secretory protein LCCL domain containing 1 1.43 0.03792 R06827 Cytochrome P450, family 20, subfamily A, polypeptide 1 1.48 0.01861 AI796334 Cytochrome P450, family 39, subfamily A, polypeptide 1 1.21 0.04672 BC035554 Cytoskeleton associated protein 5 1.27 0.04773 U79283 D site of albumin promoter (albumin D-box) binding protein 1.96 0.00732 AI873425 DCP2 decapping enzyme homolog (S. cerevisiae) 1.47 0.02261 AI023398 DDEF1 intronic transcript 1 1.53 0.01083 NM_001380 Dedicator of cytokinesis 1 1.28 0.02241 AK022894 Dendritic cell-derived ubiquitin-like protein 1.25 0.03411 NM_018630 Der1-like domain family, member 1 1.85 0.04674 AK023345 Diaphanous homolog 1 (Drosophila) 1.44 0.04409 NM_007309 Diaphanous homolog 2 (Drosophila) 1.4 0.04514 NM_000791 Dihydrofolate reductase 1.9 0.02464 AW104373 Dihydrofolate reductase-like 1 1.72 0.03976 AL043148 Dishevelled associated activator of morphogenesis 1 1.66 0.04247 N74444 Dmx-like 1 1.5 0.0211 BC011863 DNA helicase HEL308 1.3 0.03103 NM_021067 DNA replication complex GINS protein PSF1 1.25 0.04071 NM_004125 DnaJ-like protein 1.29 0.0328 AB029040 Dopey family member 1 1.4 0.00802 AK024863 Dynein, cytoplasmic 1, light intermediate chain 2 1.28 0.04611 BF445163 Ectonucleotide pyrophosphatase/phosphodiesterase 5 2.01 0.00642 BC003376 ELAV-like 1 (Hu antigen R)
253 1.63 0.02657 M25269 ELK1, member of ETS oncogene family 2.18 0.03434 BF063657 Elongation factor, RNA polymerase II, 2 1.49 0.01289 AU148611 Endogenous retrovirus pHE.1 (ERV9) 1.37 0.02926 W22924 Engulfment and cell motility 2 2.42 0.03035 NM_022086 Engulfment and cell motility 2 1.2 0.00454 N30904 Enhancer of yellow 2 homolog (Drosophila) 1.54 0.04149 BF437750 Epithelial V-like antigen 1 1.35 0.00959 AI760464 ERGIC and golgi 2 1.23 0.00305 AI277336 ESTs 1.31 0.02089 BG539238 ESTs 1.43 0.02519 N58278 ESTs 1.51 0.03397 N25986 ESTs 1.6 0.03399 BE546873 ESTs 1.95 0.04511 AW272167 ESTs 1.71 0.04731 H14374 ESTs 1.37 0.03172 AI701905 ESTs, Weakly similar to ALU1_HUMAN ALU SUBFAMILY J 1.9 0.03748 T93113 ESTs, Weakly similar to ALU4_HUMAN ALU SUBFAMILY SB2 1.71 0.00615 AK025271 Eukaryotic translation elongation factor 1 gamma 1.48 0.04756 NM_000401 Exostoses (multiple) 2 1.3 0.00956 BE350122 Exportin 1 (CRM1 homolog, yeast) 1.26 0.0135 AI478747 Family with sequence similarity 102, member B 1.23 0.04458 BC006456 Family with sequence similarity 21, member C 1.23 0.03128 BC000039 Family with sequence similarity 26, member B 1.22 0.03951 NM_003902 Far upstream element (FUSE) binding protein 1 1.51 0.03948 AI219891 Fatty acid binding protein 3 1.22 0.03983 AK026197 F-box protein 5 2.14 0.03808 AA947475 FGFR1 oncogene partner 2 1.89 0.01306 M63889 Fibroblast growth factor receptor 1 1.22 0.02711 NM_014923 Fibronectin type III domain containing 3A 1.23 0.02281 AI625550 Filamin A, alpha (actin binding protein 280) 1.32 0.0349 M62994 Filamin B, beta (actin binding protein 278) 1.72 0.01169 AA683602 FK506 binding protein 7 1.2 0.04195 AF338193 FKSG49 1.64 0.00053 NM_017669 FLJ20105 protein 1.26 0.04736 NM_003505 Frizzled homolog 1 (Drosophila) 1.59 0.0406 AB043703 Frizzled homolog 8 (Drosophila) 1.28 0.00012 AW305300 Full-length cDNA clone CS0DF003YC20 1.87 0.03507 AI651510 Full-length cDNA clone CS0DF007YJ21 1.52 0.02747 AI300571 Full-length cDNA clone CS0DJ002YF02 1.39 0.00898 BC001134 Fused toes homolog (mouse)
254 1.29 0.03973 NM_022476 Fused toes homolog (mouse) 1.21 0.03848 AW206419 G protein-coupled receptor 137B 1.69 0.0047 T16257 G protein-coupled receptor 37 (endothelin receptor type B-like) 1.57 0.0401 NM_000153 Galactosylceramidase 1.32 0.00087 NM_013430 Gamma-glutamyltransferase 1 1.3 0.01023 AL161952 glutamate-ammonia ligase (glutamine synthetase) 2.83 0.03065 NM_014905 Glutaminase 1.33 0.03009 NM_002103 Glycogen synthase 1 (muscle) 1.59 0.00999 AI052020 Glycolipid transfer protein 1.53 0.04162 AF064826 Glypican 4 1.24 0.03578 N53479 Golgi autoantigen, golgin subfamily b, macrogolgin 1.69 0.02364 NM_017577 GRAM domain containing 1C 1.83 0.03207 AB008790 Growth factor receptor-bound protein 7 1.66 0.02439 BE466139 Guanine monphosphate synthetase 1.22 0.01262 BF343852 H2A histone family, member V 1.25 0.02078 AA534860 H2A histone family, member V 1.22 0.03483 NM_016299 Heat shock 70kDa protein 14 1.71 0.03936 N54448 Heat-responsive protein 12 1.76 0.03026 AI819938 Hect domain and RLD 4 1.96 0.04302 AB046813 Hect domain and RLD 4 2.14 0.04194 AI922198 Hermansky-Pudlak syndrome 3 1.39 0.03433 W74620 Heterogeneous nuclear ribonucleoprotein D (AUF1) 1.3 0.04065 AL044078 Heterogeneous nuclear ribonucleoprotein L 1.45 0.03738 AL523158 Hexosaminidase A (alpha polypeptide) 1.26 0.02238 NM_005336 High density lipoprotein binding protein (vigilin) 1.48 0.00944 NM_003528 Histone 2, H2be 1.58 0.02048 AF230097 Histone deacetylase 8 1.49 0.03508 BC000166 HIV-1 Tat interacting protein, 60kDa 1.6 0.03905 AI052747 Homeobox B5 1.25 0.02783 BF676081 Homo sapiens cDNA FLJ11174 fis, clone PLACE1007367 1.37 0.01183 AK023816 Homo sapiens cDNA FLJ13754 fis, clone PLACE3000362. 1.27 0.04865 AA554430 Homo sapiens cDNA FLJ14343 fis, clone THYRO1000916 1.34 0.02198 AC074331 Homo sapiens chromosome 19, BAC CIT-HSPC_204F22 1.43 0.0473 BE788667 Homo sapiens full length insert cDNA clone YQ50D11 1.67 0.03893 NM_017696 Homo sapiens hypothetical protein FLJ20170 (FLJ20170) 1.27 0.04284 BG250721 Homo sapiens mRNA; cDNA DKFZp564C2063 2.29 0.03784 NM_014114 Homo sapiens PRO0097 protein (PRO0097) 1.42 0.01438 BG500396 Homo sapiens, clone IMAGE:4052238 2.55 0.00722 AW827204 Homo sapiens, clone IMAGE:5277883, mRNA 1.5 0.02042 BC035700 Homo sapiens, clone IMAGE:5550275, mRNA.
255 1.27 0.0339 NM_013320 Host cell factor C2 1.27 0.04405 AL121936 Human DNA sequence from clone CTA-14H9 on chromosome 6 Human DNA sequence from clone RP1-155G6 on chromosome 1.34 0.00952 AL121903 20 Human DNA sequence from clone RP4-534K7 on chromosome 1.43 0.02098 AL109925 1p31.2-32.3. 1.32 0.04888 NM_007076 Huntingtin interacting protein E 1.25 0.00902 U50532 Hypothetical gene CG012 1.41 0.04889 NM_014887 Hypothetical gene CG012 1.22 0.00467 AV724415 Hypothetical gene supported by AK096066 1.21 0.04837 AV751709 Hypothetical gene supported by AL449243 1.27 0.03865 AL355708 Hypothetical LOC388134 1.46 0.0181 BC039495 Hypothetical LOC401074 1.36 0.03471 AL136879 Hypothetical protein DKFZp434N035 1.22 0.02615 NM_022492 Hypothetical protein FLJ12788 1.32 0.00136 W60806 hypothetical protein FLJ14005 /FL=gb:NM_024607.1 1.41 0.0192 R33170 Hypothetical protein FLJ20397 1.73 0.01798 AK000684 Hypothetical protein FLJ22104 1.32 0.02393 W19668 Hypothetical protein FLJ22222 2.32 0.01786 NM_153016 hypothetical protein FLJ30672 1.67 0.0234 BF593867 Hypothetical protein FLJ37078 1.48 0.04237 AI634532 Hypothetical protein LOC138046 1.25 0.04685 AI052701 Hypothetical protein LOC144438 1.5 0.03996 AI469960 Hypothetical protein LOC153561 1.22 0.00802 NM_153358 Hypothetical protein LOC163049 1.46 0.02601 AI276663 Hypothetical protein LOC201725 1.33 0.01716 AI669947 Hypothetical protein LOC203107 1.27 0.02424 AA706480 Hypothetical protein LOC286260 1.73 0.00027 AI129320 Hypothetical protein LOC339751 1.25 0.03938 BC006479 Hypothetical protein MGC2749 1.6 0.04728 W87523 Hypothetical protein similar to KIAA0187 gene product 2.07 0.01924 NM_014333 Immunoglobulin superfamily, member 4 1.43 0.01311 AW269792 Importin 9 1.61 0.00899 BG500301 Integrin, beta 1 (fibronectin receptor, beta polypeptide) 1.88 0.0145 NM_133376 Integrin, beta 1 (fibronectin receptor, beta polypeptide) 1.24 0.01081 BF968057 Interferon regulatory factor 2 binding protein 2 1.45 0.03421 BF671187 IQ motif and WD repeats 1 1.39 0.04075 NM_018442 IQ motif and WD repeats 1 1.28 0.02877 BC005806 IQ motif containing B1 1.43 0.0269 NM_000216 Kallmann syndrome 1 sequence
256 1.24 0.04642 AI436268 Kelch repeat and BTB (POZ) domain containing 3 1.57 0.02561 AA808694 Kelch-like 20 (Drosophila) 1.39 0.01493 BF591270 Kelch-like 8 (Drosophila) 1.25 0.04951 AW138594 Kelch-like 9 (Drosophila) 1.74 0.00061 AI420074 KIAA0182 1.9 0.03835 NM_014734 KIAA0247 1.44 0.02072 NM_014686 KIAA0355 1.41 0.02547 BE566023 KIAA0372 1.23 0.02929 AI823592 KIAA0423 1.33 0.04285 AK026025 KIAA0515 1.91 0.01996 BE138647 KIAA0741 gene product 1.41 0.02852 BF940192 KIAA0776 1.43 0.01478 BF515597 KIAA1005 protein 1.31 0.04028 AB037791 KIAA1370 1.43 0.01085 AB040957 KIAA1524 1.89 0.02782 AB051499 KIAA1712 1.23 0.03057 AI621225 KIAA1737 1.3 0.01889 AA608629 KIAA1853 2.54 0.04336 BF003148 KIAA1961 gene 1.32 0.04364 NM_005733 Kinesin family member 20A 1.31 0.0219 AI985709 L(3)mbt-like 2 (Drosophila) 1.48 0.02318 AW005866 LAG1 longevity assurance homolog 5 (S. cerevisiae) 1.48 0.03683 BG029705 Leucine rich repeat and sterile alpha motif containing 1 1.37 0.00015 NM_014834 Leucine rich repeat containing 37A 1.25 0.02464 NM_020347 Leucine zipper transcription factor-like 1 1.35 0.0139 BF969397 LOC387882 hypothetical protein 1.38 0.03892 AK026684 LOC90693 protein 1.38 0.03262 AA043552 LUC7-like 2 (S. cerevisiae) 1.38 0.02829 AL110209 Lysophospholipase 3 (lysosomal phospholipase A2) 1.61 0.0421 U84744 Lysosomal trafficking regulator 1.77 0.02379 X63381 MADS box transcription enhancer factor 2, polypeptide A 1.39 0.00478 AK024956 Mastermind-like 3 (Drosophila) 1.56 0.03943 N53862 Matrin 3 1.22 0.00861 NM_022826 Membrane-associated ring finger (C3HC4) 7 2.34 0.04353 BE781914 Membrane-associated ring finger (C3HC4) 8 1.29 0.0162 AF072814 Metal response element binding transcription factor 2 2.16 0.01789 AW272411 Methylenetetrahydrofolate dehydrogenase 2-like 1.52 0.02525 AA251906 Methyltransferase 5 domain containing 1 1.63 0.04654 NM_024723 MICAL-like 2 1.7 0.00079 AI798164 Mindbomb homolog 1 (Drosophila)
257 1.62 0.01555 NM_002419 Mitogen-activated protein kinase kinase kinase 11 1.56 0.03086 U43784 Mitogen-activated protein kinase-activated protein kinase 3 1.21 0.04002 AI692878 MOB1, Mps One Binder kinase activator-like 2B (yeast) 1.83 0.02423 AL536899 MRNA; clone CD 43T7 1.33 0.04936 AI692322 MRNA; clone CD 43T7 1.34 0.04371 AI939472 Myoneurin 1.41 0.00937 BF593509 Myosin binding protein C, slow type 1.24 0.01959 BF215996 Myosin IB 1.25 0.0145 X98507 myosin IC 1.47 0.00893 NM_000259 Myosin VA (heavy polypeptide 12, myoxin) 1.34 0.03724 AI827941 Myosin, heavy polypeptide 9, non-muscle 1.63 0.00943 AL080178 Myotubularin related protein 9 1.42 0.04717 M68956 Myristoylated alanine-rich protein kinase C substrate 1.24 0.04367 NM_005385 Natural killer-tumor recognition sequence 2.19 0.02198 D42055 NEDD 4 1.25 0.02723 AW054826 Neurofibromin 1 1.28 0.04786 BF056507 Neutral sphingomyelinase activation associated factor 1.8 0.01573 BE046449 NHL repeat containing 2 1.27 0.0444 NM_024618 NOD9 protein 2.03 0.01747 BG108193 Nuclear cap binding protein subunit 1, 80kDa 1.22 0.01227 AB037925 Nuclear factor of kappa light polypeptide 1.44 0.00585 AW007746 Nuclear fragile X mental retardation protein interacting protein 2 1.36 0.00825 NM_005123 Nuclear receptor subfamily 1, group H, member 4 1.25 0.04946 NM_003297 Nuclear receptor subfamily 2, group C, member 1 1.34 0.04793 NM_000176 Nuclear receptor subfamily 3, group C, member 1 1.71 0.04366 AW080999 Nuclear receptor subfamily 3, group C, member 2 1.83 0.01589 AW206115 Nucleoporin 205kDa 1.36 0.04189 NM_030927 Nucleoporin like 1 1.4 0.02995 BE671532 Nudix (nucleoside diphosphate linked moiety X)-type motif 9 1.51 0.01314 BC040701 Olfactory receptor, family 2, subfamily A, member 9 pseudogene 1.36 0.03582 NM_017784 Oxysterol binding protein-like 10 1.34 0.00523 BF796470 P21 (CDKN1A)-activated kinase 2 1.8 0.03482 AF092132 P21 (CDKN1A)-activated kinase 2 1.75 0.01042 AV727449 P300/CBP-associated factor 1.43 0.01172 AB014608 P53-associated parkin-like cytoplasmic protein 1.62 0.04195 AB020690 Paraneoplastic antigen MA2 1.21 0.04528 AF001602 Paraoxonase 2 1.69 0.01797 NM_018282 Paraspeckle component 1 1.38 0.0032 AL569804 PDZ domain containing RING finger 3 1.57 0.03954 BF038548 Peptidylglycine alpha-amidating monooxygenase
258 1.38 0.04516 AI703342 Per1-like domain containing 1 1.34 0.03328 BC000496 Peroxisomal biogenesis factor 19 1.21 0.00117 BF430956 PHD finger protein 3 1.3 0.01428 NM_015937 Phosphatidylinositol glycan, class T 1.27 0.0219 AI680192 phosphoinositide-3-kinase, regulatory subunit, polypeptide 1 1.44 0.01494 AF102988 Phospholipase A2, group VI (cytosolic, calcium-independent) 1.33 0.04618 AI207338 Phosphoribosylglycinamide formyltransferase 2.39 0.0382 BC005379 Plasminogen-like B2 1.68 0.01926 AI264121 Plexin domain containing 2 1.41 0.01755 AI419423 Poly(A) polymerase gamma 1.22 0.01344 NM_007221 Polyamine-modulated factor 1 1.69 0.03885 BE855760 POU domain, class 3, transcription factor 2 1.34 0.03313 NM_024653 PRKR interacting protein 1 (IL11 inducible) 1.33 0.02969 AI754404 Procollagen-lysine, 2-oxoglutarate 5-dioxygenase 2 1.26 0.01581 AB011118 Proline/serine-rich coiled-coil 2 1.48 0.00993 NM_006223 Protein (peptidylprolyl cis/trans isomerase) NIMA-interacting, 4 1.32 0.02539 AI825800 Protein disulfide isomerase family A, member 3 1.4 0.00347 NM_006810 Protein disulfide isomerase family A, member 5 1.77 0.04615 AF361054 Protein inhibitor of activated STAT, 2 1.9 0.00291 AF118089 Protein kinase N2 1.41 0.02531 AF324888 Protein phosphatase 1, regulatory (inhibitor) subunit 12B 1.55 0.00795 AA886888 Protein phosphatase 1A Mg2+dependent, alpha isoform 2.02 0.02492 BF030448 Protein phosphatase 2 catalytic subunit, alpha isoform 2.32 0.01548 AI760130 Protein phosphatase 2 regulatory subunit B, alpha 1.39 0.03288 L42375 Protein phosphatase 2, regulatory subunit B, gamma isoform 1.33 0.01699 BG542521 Protein phosphatase 2C, Mg2+-dependent, catalytic subunit 1.32 0.02014 NM_007171 Protein-O-mannosyltransferase 1 1.58 0.04296 BC002752 PTD015 protein 1.45 0.00213 NM_005607 PTK2 protein tyrosine kinase 2 1.52 0.00891 NM_003622 PTPRF interacting protein, binding protein 1 (liprin beta 1) 1.36 0.00558 D87078 Pumilio homolog 2 (Drosophila) 1.38 0.02376 N25931 Purine-rich element binding protein B 1.9 0.0488 AI684747 PX domain containing serine/threonine kinase 2.13 0.04864 AW367507 RAB18, member RAS oncogene family 1.32 0.0168 AA349595 RAB6 interacting protein 1 1.33 0.01405 NM_002853 RAD1 homolog (S. pombe) 1.28 0.04653 NM_021930 Rad50-interacting protein 1 1.33 0.01785 NM_014636 Ral GEF with PH domain and SH3 binding motif 1 1.45 0.04709 NM_018037 Ral GEF with PH domain and SH3 binding motif 2 1.28 0.01765 AF064606 RAN binding protein 9
259 1.25 0.04389 U64675 RANBP2-like and GRIP domain containing 8 1.34 0.01424 AI928037 Rap2-binding protein 9 2.28 0.01818 BG251692 Ras and Rab interactor 2 1.57 0.03648 NM_012419 Regulator of G-protein signalling 17 1.21 0.04369 AI431643 Related RAS viral (r-ras) oncogene homolog 2 1.22 0.04662 NM_002902 Reticulocalbin 2, EF-hand calcium binding domain 1.81 0.0074 NM_006915 Retinitis pigmentosa 2 (X-linked recessive) 1.44 0.03629 NM_016026 Retinol dehydrogenase 11 (all-trans and 9-cis) 1.28 0.04948 AI761728 Ribonuclease, RNase A family, 4 1.41 0.0054 BC002574 Ring finger and WD repeat domain 3 1.31 0.01206 U23946 RNA binding motif protein 5 2.01 0.0458 AI580100 RNA binding motif, single stranded interacting protein 1 1.47 0.00777 BG032035 RYK receptor-like tyrosine kinase 1.27 0.0062 BC002847 SAR1 gene homolog B (S. cerevisiae) 1.27 0.00125 NM_013243 Secretogranin III 1.3 0.03355 AF131749 Seizure related 6 homolog (mouse)-like 2 1.46 0.03443 NM_004713 Serologically defined colon cancer antigen 1 1.29 0.03116 NM_022743 SET and MYND domain containing 3 1.46 0.03877 W65369 SET domain, bifurcated 2 1.31 0.03398 AI953523 Seven in absentia homolog 1 (Drosophila) 1.41 0.01164 NM_003022 SH3 domain binding glutamic acid-rich protein like 1.5 0.01943 AU147591 Similar to bA110H4.2 (similar to membrane protein) 1.34 0.03864 AW002079 Similar to CDNA sequence BC021608 1.33 0.02702 BC041636 Similar to FLJ40113 protein 1.23 0.02025 BE618393 Similar to hypothetical protein FLJ10307 1.69 0.01745 BF508615 SLIT-ROBO Rho GTPase activating protein 1 1.25 0.03994 NM_014267 Small acidic protein 1.49 0.04077 BG413612 Small nuclear ribonucleoprotein polypeptide N 1.29 0.04996 NM_003082 Small nuclear RNA activating complex, polypeptide 1, 43kDa 1.31 0.0289 NM_005496 SMC4 structural maintenance of chromosomes 4-like 1 (yeast) 1.27 0.01393 U67122 SMT3 suppressor of mif two 3 homolog 1 (yeast) 1.31 0.0465 NM_030777 Solute carrier family 2 (glucose transporter), member 10 1.36 0.00981 NM_024594 Solute carrier family 2, member 3 pseudogene 1 1.35 0.00346 BC001689 Solute carrier family 25 member 20 2.61 0.02476 NM_013386 Solute carrier family 25, member 24 1.56 0.00935 AI826268 Solute carrier family 25, member 29 1.3 0.02593 AI826268 Solute carrier family 25, member 29 1.36 0.04812 BF056007 Solute carrier family 27 (fatty acid transporter), member 1 1.57 0.04323 AW452623 Solute carrier family 7 (cationic aminoacid transporter) member 1 1.37 0.03544 BF432635 Solute carrier family 7, member 6 opposite strand
260 1.33 0.01992 BF447105 Sortilin 1 1.3 0.0237 AA908770 Sorting nexin 13 1.3 0.02837 AL578668 Sorting nexin 14 1.68 0.02958 AI539783 Spectrin domain with coiled-coils 1 1.68 0.03943 BE748802 SRY (sex determining region Y)-box 6 2.34 0.02238 W46994 staufen (Drosophila, RNA-binding protein) homolog 2 1.52 0.01024 BC017275 Staufen, RNA binding protein, homolog 1 (Drosophila) 1.32 0.00638 AA683501 Sulfatase modifying factor 1 1.56 0.04453 N32782 SUMO1/sentrin specific peptidase 5 1.42 0.02925 AI494567 Suppressor of Ty 6 homolog (S. cerevisiae) 1.75 0.02248 NM_002999 Syndecan 4 (amphiglycan, ryudocan) 1.57 0.00339 BC001003 Synovial sarcoma, X breakpoint 1 1.5 0.02815 R52678 Synovial sarcoma, X breakpoint 2 interacting protein 1.31 0.04976 AU152583 Synovial sarcoma, X breakpoint 2 interacting protein 1.56 0.0032 BF509528 Syntaxin 4A (placental) 1.36 0.02221 BF060683 Tankyrase, TRF1-interacting ankyrin-related ADP-ribose pol 2 1.7 0.02495 AW294686 Tau tubulin kinase 2 1.44 0.00536 BF195608 TBC1 domain family, member 2B 1.34 0.03983 NM_017489 Telomeric repeat binding factor (NIMA-interacting) 1 2.15 0.01733 NM_012339 Tetraspanin 15 1.5 0.03004 AF054841 Tetraspanin 4 1.47 0.0037 NM_017868 Tetratricopeptide repeat domain 12 1.22 0.00807 AW274240 Tetratricopeptide repeat domain 18 1.36 0.0231 BE779448 Three prime histone mRNA exonuclease 1 1.35 0.02297 BF055462 Thrombospondin 1 1.39 0.02309 NM_021992 Thymosin-like 8 1.45 0.0288 AI961231 Thymus high mobility group box protein TOX 1.66 0.04045 AW975051 Thyroid hormone receptor associated protein 2 1.25 0.01901 NM_016213 Thyroid hormone receptor interactor 4 1.28 0.01525 BC032474 Toll-interleukin 1 receptor domain containing adaptor protein 1.54 0.02033 NM_005802 Topoisomerase I binding, arginine/serine-rich 1.44 0.04961 AI147211 TRAF-interacting protein with a forkhead-associated domain 2 0.01267 AW629399 Transcribed locus 1.86 0.01657 AI659426 Transcribed locus 1.3 0.01948 AU145336 Transcribed locus 1.49 0.0201 AW263497 Transcribed locus 1.31 0.02265 BF112104 Transcribed locus 2.33 0.02413 CA314425 Transcribed locus 1.69 0.0283 AI768144 Transcribed locus 2.19 0.03074 AI791303 Transcribed locus
261 1.33 0.03188 AA664350 Transcribed locus 1.35 0.03401 AA420614 Transcribed locus 1.55 0.03925 BM676963 Transcribed locus 2.01 0.04103 AA678564 Transcribed locus 1.27 0.0421 AI305170 Transcribed locus 1.93 0.04385 AI190292 Transcribed locus 1.9 0.04486 AI042187 Transcribed locus 1.26 0.0363 BF035279 Transcribed locus, moderately similar to NP_055301.1 1.49 0.04232 BF116042 Transcribed locus, moderately similar to NP_659411.1 1.69 0.00064 R67076 Transcribed locus, moderately similar to NP_997277.1 2.88 0.01862 AI668620 Transcribed locus, moderately similar to XP_209041.2 2 0.0007 D31421 Transcribed locus, strongly similar to XP_526782.1 1.49 0.044 AI733297 Transcribed locus, weakly similar to XP_375410.1 1.4 0.0299 D81004 Transcribed locus, weakly similar to XP_517655.1 1.66 0.04774 BG248313 Transcribed locus, weakly similar to XP_517655.1 1.54 0.04384 AW057520 Transcription factor 12 (helix-loop-helix transcription factors 4) 1.35 0.02649 NM_012460 Translocase of inner mitochondrial membrane 9 homolog (yeast) 1.22 0.03602 AI986461 Translocation associated membrane protein 2 1.32 0.03853 NM_030938 Transmembrane protein 49 1.33 0.03246 AI004375 Transmembrane protein 56 1.35 0.0346 AI769794 Transmembrane protein 87B 1.37 0.03888 AI684626 Trinucleotide repeat containing 6A 1.2 0.00513 AK027071 TSC22 domain family, member 1 1.4 0.03022 AW129593 Tudor domain containing 7 1.22 0.01506 AF114012 Tumor necrosis factor (ligand) superfamily, member 12 1.24 0.01348 BF338045 Tumor necrosis factor, alpha-induced protein 8-like 1 1.52 0.00346 AW850555 TYRO3P protein tyrosine kinase pseudogene 1.34 0.01636 NM_018319 Tyrosyl-DNA phosphodiesterase 1 1.25 0.04936 AI761518 Ubiquitin protein ligase E3 component n-recognin 2 1.25 0.00944 AA528138 Ubiquitin specific peptidase 30 1.4 0.01166 BC004868 Ubiquitin specific peptidase 30 1.32 0.03254 AA522888 Ubiquitin specific peptidase 40 1.6 0.02514 AF151039 Ubiquitin-conjugating enzyme E2, J1 (UBC6 homolog, yeast) 1.26 0.04649 AI280328 Ubiquitin-conjugating enzyme E2H (UBC8 homolog, yeast) 1.24 0.03937 NM_003359 UDP-glucose dehydrogenase 1.38 0.04492 NM_014233 Upstream binding transcription factor, RNA polymerase I 1.31 0.00492 AW167087 Vaccinia related kinase 2 1.25 0.00221 AF165513 Vacuolar protein sorting 45A (yeast) 1.27 0.01789 AW296039 VAMP-associated protein A, 33kDa 1.33 0.01163 NM_006113 Vav 3 oncogene
262 1.66 0.0276 BF823525 V-raf-1 murine leukemia viral oncogene homolog 1 1.33 0.04787 BG177759 WD repeat domain 26 1.75 0.02064 AK025323 Wiskott-Aldrich syndrome-like 1.37 0.01569 U79457 WW domain binding protein 1 1.28 0.0125 AW003119 WW domain containing adaptor with coiled-coil 1.54 0.00355 BF131791 WW domain containing E3 ubiquitin protein ligase 1 1.21 0.02922 BC000836 Yippee-like 5 (Drosophila) 2.52 0.00598 AI990739 YTH domain containing 1 1.52 0.04099 AL136903 Zinc and ring finger 1 1.87 0.04964 BE465894 Zinc binding alcohol dehydrogenase, domain containing 1 2.03 0.02382 AI969773 Zinc finger CCCH-type containing 10 1.38 0.01977 AB051513 Zinc finger CCCH-type containing 12C 1.26 0.0295 AI669304 Zinc finger protein 148 (pHZ-52) 1.74 0.00001 AU154474 Zinc finger protein 20 (KOX 13) 1.48 0.01108 NM_012256 Zinc finger protein 212 1.3 0.02131 AA744771 Zinc finger protein 22 (KOX 15) 1.82 0.03847 AF533251 Zinc finger protein 396 1.34 0.01893 T89120 Zinc finger protein 462 1.38 0.03461 BC005868 Zinc finger protein 551 1.22 0.00852 NM_017652 Zinc finger protein 586 1.86 0.02846 AW300140 Zinc finger protein 599 1.6 0.00643 AL120354 Zinc finger protein 654 1.33 0.04207 AI741051 Zinc finger, matrin type 1 1.21 0.02748 AI040009 Zinc finger, RAN-binding domain containing 3 UP-REGULATED GENES Ratio p- value Identifier Gene Name 1.52 0.01431 BC002571 Abhydrolase domain containing 14A 1.37 0.01878 AA927724 Adenine phosphoribosyltransferase 1.37 0.00477 BF688144 Alkylglycerone phosphate synthase 1.31 0.0022 AB038950 Amyotrophic lateral sclerosis 2 chromosome region, candidate 2 1.21 0.04667 BC001719 Ankyrin repeat and SOCS box-containing 6 1.23 0.01375 NM_001645 Apolipoprotein C-I 1.41 0.02148 AV701177 Arrestin domain containing 4 1.2 0.03464 BF982002 AYP1 protein 1.23 0.00051 AA394039 BCL2-associated athanogene 1.34 0.01546 NM_016098 Brain protein 44-like 1.21 0.01199 M24915 CD44 antigen (Indian blood group) 1.22 0.01725 NM_014914 Centaurin, gamma 2 1.29 0.03761 BC002873 Chromosome 3 open reading frame 60 1.93 0.00973 AF161522 Claudin domain containing 1
263 1.56 0.03452 NM_001853 Collagen, type IX, alpha 3 1.82 0.04353 AW444761 Cyclin-dependent kinase inhibitor 2B (p15, inhibits CDK4) 1.39 0.03647 NM_000785 Cytochrome P450, family 27, subfamily B, polypeptide 1 1.35 0.01137 BF941492 DALR anticodon binding domain containing 3 1.4 0.03055 NM_014764 DAZ associated protein 2 1.25 0.0392 AL133598 DEAD (Asp-Glu-Ala-Asp) box polypeptide 20 1.42 0.02004 NM_012242 Dickkopf homolog 1 (Xenopus laevis) 1.22 0.02145 J03620 Dihydrolipoamide dehydrogenase 1.39 0.01755 NM_004417 Dual specificity phosphatase 1 1.21 0.03036 NM_019002 ETAA16 protein 1.37 0.00727 NM_002032 Ferritin, heavy polypeptide 1 1.21 0.04598 BF196572 Golgi associated g-adaptin ear containing ARF binding protein 2 1.49 0.0452 NM_016099 Golgi autoantigen, golgin subfamily a, 7 1.54 0.04045 AF003934 Growth differentiation factor 15 1.23 0.04267 NM_014366 Guanine nucleotide binding protein-like 3 (nucleolar) 1.55 0.00499 AL121900 clone RP11-379J5 on chromosome 20 1.34 0.01514 AL117354 clone RP5-976O13 on chromosome 1p21.2-22.2 1.39 0.01128 NM_016142 Hydroxysteroid (17-beta) dehydrogenase 12 1.29 0.01738 BF507383 Hypothetical gene supported by AK026843; BX640678 1.55 0.03876 AL527334 Hypothetical protein DKFZp564O0523 1.37 0.02602 T79568 hypothetical protein FLJ10656 /FL=gb:NM_018170.1 1.23 0.03208 AI110850 Hypothetical protein FLJ39370 1.38 0.00368 NM_016401 Hypothetical protein HSPC138 1.8 0.04247 AK055559 Hypothetical protein LOC148189 1.25 0.02747 NM_015983 Hypothetical protein LOC222070 1.76 0.03664 BC039295 Hypothetical protein MGC39518 1.61 0.01906 AI741392 Importin 7 1.37 0.04692 AA126419 Inositol polyphosphate-4-phosphatase, type I, 107kDa 1.31 0.00021 NM_013417 Isoleucine-tRNA synthetase 1.36 0.00111 NM_006855 KDEL endoplasmic reticulum protein retention receptor 3 1.33 0.00678 AA029155 Mannosidase, alpha, class 2A, member 1 1.38 0.04172 AV700323 Mannosidase, alpha, class 2A, member 1 1.88 0.0331 NM_021647 Microfibrillar-associated protein 3-like 1.28 0.00498 AL563572 Mitochondrial methionyl-tRNA formyltransferase 1.33 0.01478 AV726260 Mitochondrial ribosomal protein L41 1.35 0.03578 N50665 Mitogen-activated protein kinase kinase kinase 2 1.44 0.03583 NM_006792 Mortality factor 4 1.54 0.03833 NM_006791 Mortality factor 4 like 1 1.49 0.03751 AL109722 MRNA full length insert cDNA clone EUROIMAGE 31619 1.36 0.00962 AF054589 MyoD family inhibitor domain containing
264 1.23 0.01406 H05010 Nedd4 family interacting protein 1 1.29 0.03617 AI096706 NIPA-like domain containing 2 1.24 0.04167 BG054916 Patched homolog (Drosophila) 1.42 0.0467 NM_005038 Peptidylprolyl isomerase D (cyclophilin D) 1.76 0.00935 AI936197 Phosphoserine phosphatase 1.31 0.03471 AA199881 Potassium channel tetramerisation domain containing 1 1.6 0.03946 NM_015387 Preimplantation protein 3 1.22 0.0328 NM_002874 RAD23 homolog B (S. cerevisiae) 1.22 0.03234 AV706343 Ras-associated protein Rap1 1.55 0.04672 BF446578 RasGEF domain family, member 1A 1.21 0.04004 NM_002923 Regulator of G-protein signalling 2, 24kDa 1.31 0.01243 NM_014059 Response gene to complement 32 1.36 0.01749 AB018283 Rho-related BTB domain containing 1 1.21 0.02797 BE274422 Ribosomal protein L22-like 1 1.21 0.04442 NM_000978 Ribosomal protein L23 1.28 0.03611 AA911739 Selenocysteine lyase 1.21 0.00993 AI554514 Sine oculis homeobox homolog 4 (Drosophila) 1.22 0.04748 U46837 SRB7 suppressor of RNA polymerase B homolog (yeast) 1.77 0.04911 AB028976 Sterile alpha motif domain containing 4A 1.27 0.01749 X57348 stratifin 1.66 0.04646 NM_006285 Testis-specific kinase 1 1.28 0.01913 NM_004786 Thioredoxin-like 1 1.55 0.03991 AI740460 Transcribed locus 1.31 0.01217 NM_003198 Transcription elongation factor B (SIII), polypeptide 3 1.26 0.03285 AA191576 TRK-fused gene 1.59 0.03998 U39361 Ubiquitin-conjugating enzyme E2 variant 1 1.54 0.02341 NM_016079 Vacuolar protein sorting 24 (yeast) 1.32 0.03164 BC004419 Vacuolar protein sorting 24 (yeast) 1.43 0.00984 AW590651 Zinc finger protein 141 (clone pHZ-44) 1.25 0.04334 BC004535 Zinc finger, DHHC-type containing 16
265 Appendix 4: Binary Logarithm (log2) Transformation. Log2
100 %
Initial value
- 50 % Linear
Log2 values vs linear values. Binary logarithm transformation is useful for plotting micro-array data because it compacts the expression data, and allows fold-change data (which is not linear) to be plotted on a graph with a linear y axis. For example, a fold-increase of 2 represents a 100 % increase in numerical value, while a fold-decrease of 2 represents only a 50 % decrease in numerical value. Binary logarithmic transformation of those values allows expression of a fold-increase of 2 as “1” and a fold-decrease of 2 as “-1”.
266