HYPOXIA INDUCIBLE FACTORS REGULATE HYPOXIA-MEDIATED SPLICING
by
JOHNNY A. SENA
B.S., Eastern New Mexico University, 2005
M.S., Eastern New Mexico University, 2008
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Molecular Biology Program
2014
ii
This thesis for the Doctor of Philosophy degree by
Johnny A. Sena
has been approved for the
Molecular Biology Program
by
Rui Zhao, Chair
Cheng-Jun Hu, Advisor
David Bentley
Sean Colgan
Richard Davis
Cynthia Ju
Trevor Williams
Date _5/2/14__
iii
Sena, Johnny A. (Ph.D., Molecular Biology)
Hypoxia Inducible Factors Regulate Hypoxia-mediated Splicing
Thesis directed by Assistant Professor Cheng-Jun Hu.
ABSTRACT
In response to oxygen deprivation or hypoxia within a cell, HIF1α and
HIF2α hypoxia inducible transcription factors are stabilized and activate gene expression programs. Activation of HIF target genes promotes cell proliferation, apoptotic resistance, metabolic adaptation, and tissue angiogenesis in normal physiologic processes but these processes can be exploited by cancer cells to promote tumorigenesis.
Recent genome-wide studies suggest that 90% of human genes are alternatively spliced, producing RNA isoforms that encode functionally distinct proteins. Thus, effective hypoxia response requires regulation of gene transcription as well as RNA splicing. Interestingly, several reports suggest that hypoxia can regulate alternative splicing; however the mechanism was not known. Although it is well established that HIFs regulate transcription during hypoxia, it was not known if HIFs could regulate alternative splicing. We hypothesized that HIFs were responsible for regulating hypoxia-mediated splicing. The following work confirmed that hypoxia regulates alternative splicing of HIF and non-HIF target genes. Moreover, we determined that HIFs, through the activation of HIF target genes, regulates transcription and alternative splicing of a subset of HIF targets. Subsequently, we found that HIFs induced the expression of CDC-like kinases which phosphorylated and activated SR protein iv
splicing factors and that SC35 and SRp40 SR proteins regulated hypoxia- induced splicing of HIF target genes. Importantly, our findings provide a novel molecular mechanism through which HIFs regulate gene expression programs during hypoxia by regulating alternative splicing of HIF target genes. Importantly, we determined that CDC-like kinases regulate HIF-dependent tumorigenesis by regulating splicing of HIF targets. Excitingly, these findings identify a novel HIF signaling pathway that may be exploited to inhibit HIF activity in tumorigenesis.
The form and content of this abstract are approved. I recommend its publication.
Approved: Cheng-Jun Hu v
To My Family for Always Believing in Me.
vi
CONTENTS
CHAPTER
I. INTRODUCTION…………………………………………………………. …….1
Hypoxia……………...……………………………………………………. ….....1
Hypoxia Inducible Transcription Factors.....…………………………… …….2
Discovery of HIF………………………………………………….. …….4
HIF1α……………………………………………………… ….....4
HIF2α……………………………………………………… …….4
Structure of HIFα..……………………………………………….. …….5
Oxygen Regulation of HIFα……………………………………... …….6
HIF Regulation by Kinase Signaling…………………………… …...11
HIF in Normal Physiology……………………………………………….. …...12
Embryonic Development………………………………………… …...13
Wound Healing…………………………………………………… …...14
Stem Cell Maintenance………………………………………….. …...15
HIF in Tumorigenesis……………………………………………………. …...16
Tumor Initiation…………………………………………………… …...17
VHL disease………………………………………………. …...17
Cancer stem cells………………………………………… …...18
Tumor Angiogenesis…………………………………………….. …...19
Epithelial-Mesenchymal Transition and Metastasis …………. …...20
Tumor Metabolism……………………………………………….. …...21
Cancer Therapy Resistance…………………………………….. …...22 vii
Pre-mRNA Splicing………………………………..…………………….. …...24
Splicing Reaction..………………………..…………..………..… …...25
Alternative Splicing.……………………………………………… …...26
Role of Alternative Splicing in Development……………..…………31
Role of Alternative Splicing in Cancer……………….………… …...32
Alternative Splicing in Apoptosis…………………….…. …...32
Alternative Splicing in Cancer Metabolism …………… …...33
Alternative Splicing in the Regulation of Proto-oncogenes…..…………………………..……... …...33
Alternative Splicing in Metastasis and Invasion……...………………………………..……...... 33
Alternative Splicing in Angiogenesis…………………… …...34
Cis Elements Regulate Splicing…………………….………….. …...34
Trans-acting Factors Regulate Splicing………………..……… …...36
SR Protein Regulation…………………………………………... …...38
Coupling of Transcription and Splicing………………………… …...41
Cellular Stress Regulates Alternative Splicing………………... …...48
Regulation of Alternative Splicing by Heat Shock……………………………………………. …...48
Regulation of Alternative Splicing by Osmotic Stress………………………………………... …...49
Regulation of Alternative Splicing by Uv Irradiation………………………………………….. …...49
Regulation of Alternative Splicing by Hypoxia………………………………………………… …...49
viii
II. HIFS ENHANCE THE TRANSCRIPTIONAL ACTIVATION AND SPLICING OF ADRENOMEDULLIN……………. …...53
Abstract……………………………………………………………………. …...53
Introduction……………………………………………………………….. …...54
Results…………………………………………………………………….. …...55
Hypoxia Preferentially Increases Fully-spliced ADM Transcript Levels in Various Cell-lines..………………… …...55
The Increased ADM FL/I1-3 Ratio is Due to RNA Splicing but Not Due to Differential RNA Stability ……………………..………………… …...59
ADM I1-3 Transcripts are Primarily Located in the Nucleus …………………………...…………….. …...60
HIF Activity is Required for Increased Splicing of ADM Pre-mRNA…………………………………….. …...63
HIF Activity is Sufficient for Increased Splicing of ADM Pre-mRNA …………….……………………… …...65
ADM Splicing Reporters Recapitulate Splicing Changes Observed for the Endogenous ADM Gene.……...………………………………… …...66
The Transactivation Domain of HIFα Protein is Not Required for Increased RNA Splicing of the ADM Splicing Reporter.…..……………… …...71
Activation of Endogenous HIF Target Genes is Not Absolutely Required for Increased Splicing of the ADM Splicing Reporter…………….. …...74
Discussion………………………………………………………………… …...76
Acknowledgments………………………………………………………... …...82
III. HYPOXIA REGULATES ALTERNATIVE SPLICING OF HIF AND NON-HIF TARGET GENES……………………...... …...83
Abstract……………………………………………………………. …………...83 ix
Introduction……………………………………………………….. …………...84
Results………………………………………………………………………….85
Genome-wide Exon Array Analysis Determines that Hypoxia Alters RNA Splicing of HIF and Non-HIF Target Genes in Hep3B Cells…………………………………... …...85
Functional Clustering of Genes that Undergo Hypoxia-Mediated AS Reveals Novel Pathways Regulated By Hypoxia……………………….. …...90
RT-PCR and qRT-PCR Validation of Alternative Splicing for Select HIF Target Genes Identified in in Exon Array Analysis……………..……………………………. …...91
Differential Expression of HIF Target Genes During Hypoxia is Due to Alternative Splicing…………...……. …...96
HIF Activity, Not Hypoxia Per Se, is Necessary to Regulate AS of HIF Target Genes……………………………………………………...99
HIF Activity is Sufficient to Regulate AS of HIF Target Genes………………………………………… ….101
PDK1 Splicing Reporters Recapitulate Splicing Changes Observed for the Endogenous PDK1 Gene when Activated by HIF under Normoxia……………………………………………... ….104
Activation of Endogenous HIF Target Genes Contributes to the Increased FL/ΔE4 ratio of the PDK1 Splicing Reporter ……………………...107
Discussion…………………………………………………………………….110
Acknowledgments……..…………………………………………………. ….113
x
IV. HYPOXIC INDUCTION OF CDC-LIKE KINASES REGULATES TRANCRIPTION AND ALTERNATIVE SPLICING OF HIF TARGET GENES………………………………….. ….114
Abstract……………………………………………………………………. ….114
Introduction……………………………………………………………….. ….115
Results…………………………………………………………………….. ….116
HIFs Regulate Hypoxic Induction of CDC- Like Kinase 1, and 3 and Hypoxia Induces Phosphorylation of SR Proteins ……………………………...... ….116
TG003 Inhibits Hypoxia-induced Transcription and Splicing of HIF Targets………………….……….…………. ….119
SC35 and SRp40 Regulate AS of HIF Target Genes…………….……………………………….. ….121
SC35 Cooperates with HIF1α to Regulate AS of a PDK Minigene…………………….……………………... ….124
CLKs promote tumorigenesis of PRC3 Cells by Regulating AS of HIF Target Genes…………………. ….125
Discussion………………………………………………………………… ….128
Acknowledgments………………………………………………………... ….130
V. CONCLUSIONS AND PERSPECTIVES………………………………. ….132
Summary and Conclusions……………………………………………… ….132
Hypoxia, HIFs, and Pre-mRNA Splicing..…………………...… ….133
Hypoxic Induction of CLKs Regulates AS of HIF target Genes…………………..……………………… ….135
Perspectives and Future Directions……………………………………. ….137
Small Nucleolar RNAs……………..……………………………. ….138
RNA Helicases…………………………..……………………….. ….140 xi
Other SR Proteins…………..…………..……………………….. ….141
SMAD3 and RBPMS…………………………………………….. ….142
NOL3………………………………………………………………. ….143
Splicing Targeted Therapies……………………………………. …145
VI. MATERIALS AND METHODS………………………………………….. ….148
Cell Culture……………………………………………………………….. ….148
Knockdown of Endogenous mRNAs Using Small Interfering RNAs (siRNAs)………………………………………. ….149
Plasmid Constructs and Viral Transduction…….…………………….. ….150
RNA Stability Assays.……………………………………………………. ….152
ADM and PDK1 Splicing Reporter Assays…..………………………... ….153
Protein Analysis………………………………………………………….. ….153
RNA Preparation and Reverse Transcription PCR and Quantitative PCR………………..……………. ….154
Exon Array Analysis of Alternative Splicing...…………………………. ….155
Functional Clustering of Hypoxia Inducible Genes and Alternatively Spliced Genes using DAVID Bioinformatics Resources………………………………………. ….156
In Vitro Tumorigenic Assays……………………………………………. ….156
Statistical Analysis……………………………………………………….. ….157
REFERENCES……………………………………………………….………….. ….158
APPENDIX
A. BRG1 AND BRM CHROMATIN REMODELING COMPLEXES REGULATE THE HYPOXIA RESPONSE BY ACTING AS A CO-ACTIVATOR FOR A SUBSET OF HIF TARGET GENES…… ….192
Abstract……………………………………………………………………. ….192 xii
Introduction……………………………………………………………….. ….193
Materials and Methods………………………………………………...… ….195
Cell Culture……………………………………………………….. ….195
Knockdown of Endogenous mRNA Using Small Interfering RNAs (siRNAs) or Short Hairpin RNA (shRNAs)…………………………..……. ….196
Plasmid Constructs and Transient/ Stable Transfection……………………………………………… ….196
Protein Analysis…………………………………………………. ….197
RNA Preparation and Reverse Transcription Quantitative PCR………………………………… ….198
ChIP Experiments………………………………………………... ….198
Co-Immunoprecipitation…………………………………………. ….199
Nuclesome Scanning Assay (NUSA)………………………….. ….199
Luciferase Reporter Assay..…………………………………….. ….200
In vitro Tumorigenic Assays…………………………………….. ….201
Results………………………………………………………………………...202
BRG1 or BRM Transient Knockdown Reduces the Hx Induction of Several Known HIF1 and HIF2 Target Genes in Hep3B Cells.………...... …… ….201
Stable Knockdown of BRG1 Reduces HIF1α and HIF2α Gene Transcription in Hep3B Cells………………. ….203
The BRG1 Complex Promotes Hx Induction of HIF Target Genes in Hep3B Cells.…………………………….. ….203
Re-introduction of BRM or BRG1 into BRM/ BRG1 Deficient SW13 Cells Enhances Hx Induction of HIF1 Target Genes.………………………….... ….210
xiii
BRG1 Binds to HIF Target gene Promoters in an ARNT/HIF Dependent Manner in Hep3B Cells…………………………….……………………… ….213
BRM or BRG1 Complexes are not Required for Hx Induction of a Subset of HIF Target Genes…………...………………………………… ….216
Hx Induces Nucleosome Remodeling of the CA9 Promoter in a BRG1 Dependent Manner in Hep3B Cells…..……………………………………… ….217
Nucleosome Remodeling of the CA9 Gene Requires BRG1’s ATPase Activity ………………………………………………...... ….220
Re-introduction of BRG1 into SW13 Cells Increases the HIF-mediated Hx Response and HIF-mediated Inhibition of Proliferation………………….. ….222
The HIF-mediated Hx Response Dictates the the Role of the BRG1 Complex in Hypoxic RCC4T Cells During Proliferation and Migration………………………. ….225
Discussion………………………………………………………………... ….229
Acknowledgments……………………………………………………….. ….234
xiv
TABLES
TABLE
5.1 Gene Expression Analysis of Exon Array Data For Genes Implicated in Hypoxia-mediated Alternative Splicing Regulation………..………..…………………...... ….144 xv
FIGURES
FIGURE
1.1 HIFα Protein Structure…..……...……………………………………….. …….7
1.2 HIFα Subunit Stability is Regulated by Oxygen……………..………... …….8
1.3 FIH Inhibits HIF Transactivation………………………………………... …...10
1.4 Two Step Pre-mRNA Splicing Reaction by the Spliceosome…..………………………………………. …...27
1.5 The Spliceosome Consists of Five Small Nuclear Ribonucleoproteins (snRNP) that Assemble onto the Intron to Facilitate Splicing…………..………………………….….. …...28
1.6 Alternative Splicing Patterns……………………………………………. …...30
1.7 Function of SR Proteins in Splicing Site Selection………………….... …...35
1.8 Phoshorylation-mediated Regulation of SR Protein Activity…………………………………………………..… …...40
1.9 Recruitment Model of Alternative Splicing…………………………….. …...44
1.10 Kinetic Model of Alternative Splicing…………………………………… …...46
2.1 Hypoxia Increases the Levels of Fully-Spliced ADM Transcripts in Cancer and Normal Cells………..………...…….. …...57
2.2 ADM FL and I1-3 Transcript Stability Does Not Account for the Hypoxia-mediated Increase in the ADM FL/I1-3 Ratio, and ADM FL and I1-3 Transcripts are Differentially Localized in Hep3B cells………………. …...62
2.3 HIF Activity is Required for Hypoxia-induced Splicing of ADM Pre-mRNA.…………….……………………………… …...64
2.4 HIF Activity is Sufficient to Promote Splicing of ADM Pre-mRNA…………..………………………………… …...67
2.5 HIFs Regulate Splicing of ADM Splicing Reporters………………….. …...70
xvi
2.6 Activation of the ADM Splicing Reporter is Sufficient to Increase Splicing of the ADM Reporter…………………………….. …...73
2.7 Activation of Endogenous HIF Target Genes is Not Required for Increased Splicing of an ADM Splicing Reporter….……………………………… …...77
3.1 Hypoxia Regulates Gene Expression and RNA Splicing in Hep3B Cells………..………………………………….. …...89
3.2 Hypoxia regulates AS of the HIF target gene, CA9, in Hep3B and SKNMC Cells……………………………………... …...94
3.3 Hypoxia Regulates AS of the HIF Target Gene, ANGPTL4, in Hep3B and UM-SCC-22B Cells……….…………….…. …...96
3.4 Hypoxia Regulates AS of Additional HIF Target Genes in Hep3B Cells………………………………………………….... …...97
3.5 HIF Activity, but Not Hypoxia Per Se is Necessary to Promote AS of HIF Target Genes ……………………………………... ….100
3.6 HIF Activity is Sufficient to Promote AS of HIF Target Genes in Hep3B Cells..…………………………….. ….102
3.7 HIF Activity is Sufficient to Promote AS of HIF Target Genes in Normoxic RCC4 Cell…………………………………. ….103
3.8 Transcription Activation is Not Sufficient to Regulate Splicing of a PDK1 Minigene…………………………………… ….105
3.9 Activation of Endogenous HIF Target Genes AS of a PDK1 Minigene…………………………………………..……... ….109
4.1 Hypoxia Induces the Expression of CLK1 and 3 in a HIF-dependent Manner and Induces Phoshorylation of SR proteins………………………………………….. ….119
4.2 CLKs Regulate Hypoxia-induced Gene Expression and Alternative Splicing of HIF Targets but Not of Non-HIF Targets.…………………………... ….121
4.3 SC35 and SRp40 Regulate Hypoxia-mediated AS of HIF Targets and SC35 Cooperates with HIF1 to Regulate AS of a PDK1 Minigene…..…………………… ….124 xvii
4.4 CLKs Regulate HIF2α-mediated Tumorigenesis in PRC3 Cell by Regulating AS………………………. ….128
4.5 Model of Hypoxia/HIF-mediated Alternative Splicing and Tumorigenesis…………………………………………….. ….131
A1 BRG1 is More Important than BRM Complexes in Regulating Hx-Induction of HIF Target Genes in Hep3B Cells……………………………………… ….204
A2 BRG1, but not BRM Complexes are Important in HIF1α and HIF2α Gene Transcription in Hep3B Cells………………. ….204
A3 BRG1 Knockdown Reduces Endogenous HIF1α and HIF2α mRNA Levels, but not Transfected Mouse HIF1α or HIF2α mRNA…………………………………………………... ….207
A4 BRG1 Promotes Hx Induction of HIF Target Genes in Hep3B Cells…………………………………………………… ….209
A5 Re-expression of BRG1 or BRM in BRM/BRG1 Deficient SW13 Cells Increases Expression of HIF2α and HIF Target Genes……………………………………………..…… ….211
A6 Re-expression of WT, not ATPase-dead BRG1 in BRM/BRG1 Deficient SW13 Cells Increases Expression of HIF2α and HIF Target Genes………………………….. ….212
A7 Hx Increases BRG1 Binding on the Promoters of CA9 and EPO in an ARNT-dependent Manner……………………… ….214
A8 Hx Induction of a Subset of HIF Target Genes is BRM/BRG1 Independent………………………………………………... ….218
A9 Hx Decreases Nucleosome Association on the CA9 Promoter in a BRG1-dependent Manner……………………………… ….219
A10 Reintroduction of wild-type but not ATPase-dead BRG1 in SW13 cells decreases nucleosome association on the CA9 promoter………………………………………. ….221
A11 Re-introduction of BRG1 into SW13 Cells Increases the HIF-mediated Hx Response and HIF-mediated Growth Suppressive Activity……………………………………………. ….223 xviii
A12 The HIF-mediated Hx Response Dictates the Role of the BRG1 Complex in Hypoxic RCC4T Cells for Proliferation and Migration………………………………………………. ….227 1
CHAPTER I
INTRODUCTION
Hypoxia
Oxygen is necessary for eukaryotic cell survival because it is required for cellular respiration and energy production via oxidative phosphorylation. Tissue access to oxygen is determined by cellular proximity to oxygenated blood (1), therefore all cells must reside within 2mm3 of a blood supply to survive (2). Thus, the extent of vascularization within a tissue determines the oxygen concentration experienced in the cellular microenvironment. Oxygen deprivation, referred to as hypoxia, occurs to different extents in tissues throughout the body and is a normal aspect of multicellular physiology. For instance, oxygen concentrations in arterial blood are maintained at 14% pO2, 10% in the myocardium, and around
5% in most tissues under normal physiological conditions. In contrast, other tissues such as cartilage, bone marrow, and thymus exhibit oxygen concentrations at or below 1% (3-6) and are therefore hypoxic. Physiologic hypoxia within different tissue niches drive important processes including embryonic development, wound healing, and stem cell niche maintenance (1, 7-
9). However, hypoxia, due to aberrant oxygen delivery or increased oxygen consumption, can also drive pathological processes including diabetic retinopathy, peripheral artery disease, ischemic heart disease, ischemic stroke, and tumorigenesis (7, 10, 11). 2
During hypoxia, cells activate gene expression programs that reduce
hypoxic stress and allow the cells and tissues to survive and adapt to the hypoxic
microenvironment (12, 13). These gene expression programs promote
neovascularization, the formation of new blood vessels, which results in
increased tissue vascularization and increased oxygenation of the cellular
microenvironment. Additionally, hypoxia activates gene expression programs
which shift cellular metabolism from oxidative phosphorylation during oxygen
replete conditions or normoxia, to glycolysis for ATP generation under low
oxygen conditions, prolonging cell survival until reoxygenation occurs (14, 15).
Interestingly, hypoxia-induced activation of these pathways ensures spatial and
temporal regulation of cellular proliferation, tissue growth and survival during
normal physiological process such as embryogenesis and wound healing.
However, hypoxia-induced gene expression of proliferation and pro-angiogenic promoting pathways can also be a driving force behind malignant tumor cell growth, increased angiogenesis and metastasis, resistance to anti-tumor therapies and a poor prognosis (16-20).
Hypoxia Inducible Transcription Factors
Hypoxia Inducible transcription Factors (HIFs) are master regulators of hypoxia-induced gene expression programs in mammalian physiological and
pathophysiological processes (12, 13). HIFs are heterodimeric transcription
factors composed of an oxygen-sensitive alpha subunit (HIFα) and a
constitutively expressed beta subunit (HIF1β, also called Aryl hydrocarbon 3
Receptor Nuclear Translocator or ARNT) (21). During hypoxia, HIFα protein subunits are stabilized and translocate from the cytoplasm to the nucleus of the cell where they dimerize with ARNT and activate transcription by binding to HIF binding sites (HBS) within hypoxia-responsive elements (HRE) in the promoters and/or enhancer regions of HIF target genes (21, 22). Three distinct HIFα subunits have been identified; HIF1α, HIF2α, and HIF3α (23, 24). Generally,
HIF1α is ubiquitiously expressed in mammalian tissues; however HIF2α expression is typically limited to specific cell types including endothelial cells, glial cells, type II pneumocytes, cardiomyocytes, and fibroblasts of the kidney, interstitial cells of the pancreas and duodenum, and hepatocytes. (25). On the other hand, the role of HIF3α in the hypoxic regulation of target gene expression in vivo is not well understood. HIF3α has multiple splice variants, of which the inhibitory domain PAS protein (IPAS) is the best characterized. IPAS is a truncated form of HIF3α that lacks a transactivation domain and functions as a dominant negative by binding to HIF1α and preventing the formation of HIF1α
/ARNT heterocomplexes. HIF3α mRNA can be detected in a variety of tissues, including the thymus, lung, brain, heart, kidney, liver, eye, and brain .
Three ARNT subunits have also been discovered; ARNT, ARNT2, and
ARNT3 (26), however ARNT is the only ubiquitously expressed isoform and is dominant in the hypoxic response (26). The roles of ARNT2 and ARNT3 in the hypoxic-induction of target genes have not been extensively characterized but they can form heterodimers with HIF1 and 2α subunits. However, unlike ubiquitiously expressed ARNT, ARNT2 and 3 expression is limited to the brain 4
and kidney or brain and skeletal muscle respectively (27, 28). Consistent with
ARNT2 expression in the brain, ARNT2 mediates hypoxic gene expression in
neurons. In addition, coexpression of ARNT3 and HIF1α in ARNT-deficient c4 cells enhanced the transcription of a HIF responsive reporter gene driven by a
hypoxia response element (HRE) sequence (29). Because ARNT is dominant in
the hypoxic response, the remainder of this dissertation will focus on
HIF1α/ARNT (HIF1) and HIF2α/ARNT (HIF2) heterodimers only.
Discovery of HIF
HIF1α. HIF1α was originally identified using DNA affinity purification as a
nuclear protein bound to a 18 nucleotide region of the 3’ flanking sequence of the
erythropoietin (EPO) gene (30). This region was later characterized as the EPO
enhancer region and the primary HIF-binding region. Purification of the HIF1α
protein and domain analysis determined that HIF1α functioned in a dimeric
complex with the ARNT protein (21). Both proteins were identified as basic-helix-
loop-helix transcription factors, containing a PAS (Per-ARNT-Sim) domain, a domain identified by its presence in the drosophila Per and Sim proteins. Further studies determined that HIF1α is expressed in all mammalian tissues and cell types (31) and regulates the hypoxic expression of hundreds, if not thousands, of
genes (32-35).
HIF2α. The HIF2α protein was first discovered in mouse and quail
endothelial cells and in cells expressing high levels of vascular endothelial
growth factor. Since many groups contributed to the discovery of HIF2α 5
simultaneously, it was referred to by several names including HRF(HIF-Related
Factor), HLF (HIF1α-Like Factor), MOP2 (Member Of the PAS superfamily-2), and EPAS1 (Endothelial PAS domain containing protein-1) (36-40). Protein structure analysis and characterization of the HRF/HLF/MOP2/EPAS1 basic- helix-loop-helix and PAS domains identified significant protein sequence similarity to the HIF1α protein and is commonly referred to as HIF2α (41). Like
HIF1α, HIF2α can be stabilized by low oxygen and binds as a heterodimer with
the ARNT protein to hypoxia responsive DNA elements within target gene
promoters (39, 42). Similar to HIF1α, HIF2α was also found to activate hypoxia-
dependent transcription of the VEGF and EPO genes (36, 39) and found to
regulate hypoxic induction of several HIF1α-independent target genes including
Flk-1 and tie-2, endothelial cell receptor tyrosine kinases (39, 43). However,
unlike HIF1α which is ubiquitously expressed in mammalian tissues, expression
of HIF2α is limited to several tissue and cell types including endothelial cells, type
II pneumocytes of the lung, liver hepatocytes, cardiomyocytes, macrophages,
astrocytes, and neuroendocrine cells of the organ of Zukerkandl located at the
bifurcation of the aorta of the heart (44, 45).
Structure of HIFα
HIF1α and HIF2α proteins are transcription factors that exhibit high protein
similarity, containing five protein domains including an N-terminal basic-helix-
loop-helix domain (bHLH) which facilitates DNA binding and a Per-ARNT-SIM
domain (PAS) which facilitates dimerization with the PAS domain of the ARNT 6
protein (Fig. 1.1). On the other hand, the C-terminus of the proteins contains an
N-terminal activation domain (N-TAD) and C-terminal activation domain (C-TAD), required for activation of target genes, and also an oxygen-dependent degradation domain (ODD) which is necessary for degradation by the ubiquitin- protesome, followed by an inhibitory domain (IH). Importantly, the N-TAD and C-
TAD of HIF1α and HIF2α contribute to the transcriptional activity of the proteins by facilitating protein interactions with CBP/p300 histone acetyltransferases (46,
47). The N-TAD has been suggested to play a role in determining HIF target
gene specificity through interaction with specific co-factors such as upstream
stimulatory factor 2 (USF2) and signal transducer and activator of transcription 3
(STAT3) (48-50).
Oxygen Regulation of HIFα
HIFα activity is regulated at multiple levels; however, the most studied
process of HIFα subunit regulation occurs at the posttranslational level. During
oxygen-rich conditions HIFα proteins are constitutively translated however they
are rapidly degraded by the 26s proteasome, with a half-life of 5 minutes,
resulting in silencing of HIF function (51, 52). Specifically, during normoxia a
family of prolyl hydroxylase enzymes (PHD1, PHD2, PHD3 in mammals) that
require oxygen, iron, ascorbate, and 2-oxoglutarate for enzymatic activity,
hydroxylate two conserved proline residues,P402 and P577 of HIF1α and P405
and P530 of HIF2α, in the ODD of the HIFα proteins (Fig. 1.1) (52-59).
7
Figure 1.1 ODD P402 P577 N813 HIF-1α bHLH PAS TAD IH TAD 836 aa P405 P530 N813851 HIF-2α bHLH PAS TAD IH TAD 874 aa % Amino Acid Sequence Similarity: 85% 70% 40% 19% 69% DNA binding ARNT binding Transactivation and stability
Figure 1.1: HIFα Protein Structure. HIF1α and HIF2α are bHLH and PAS proteins and exhibit a high degree of similarity in their DNA and ARNT binding domains, and the C-terminal transcriptional activation domains. The N-terminal transactivation domains overlap with the oxygen dependent degradation domains containing two hydroxylation-sensitive proline residues. Numbers refer to percentage of amino acid sequence similarity.
8
Figure 1.2 Normoxia Hypoxia A. B. HIFα ARNT O2 α α OH OH HIF HIF pVHL X X X OH OH HIFα ARNT HIFα proteasome pVHL VEGF Transcription PGK EPO
Figure 1.2: HIFα Subunit Stability is Regulated by Oxygen. A) Hydroxylation of proline residues facilitates binding of the von Hippel Lindau protein (pVHL), which targets HIFα for proteasomal degradation. Under hypoxic conditions, HIFα subunits are stabilized due to a lack of hydroxylation. Consequently, transcriptionally active heterodimers can form between HIFα and ARNT subunits resulting in activation of hypoxia inducible genes.
9
Subsequently, hydroxylation of HIFα proteins facilitates binding of the Von
Hippel-Lindau tumor suppress protein and polyubiquitination by E3 ubiquitin ligases, targeting HIFαs for degradation by the 26S proteasome (Fig. 1.2) (52,
60-65). However, because of a lack of oxygen substrate during hypoxia,
hydroxylation does not occur on the HIFα prolines. Thus, pVHL and E3 ubiquitin
ligases cannot recognize and bind to HIFαs, resulting in stabilization of the
proteins (66). Once stabilized by hypoxia, HIFα is transported to the nucleus by
importin 4 and 7 where it heterodimerizes with ARNT to form the active HIF
transcriptional complex. The HIF heterodimers then interact with CBP/p300
histone acetylases and bind to HREs within target gene promoters, activating
transcription (Fig. 1.2) (67-69).
In addition, studies suggest that HIF transactivation can also be
modulated by the “Factor Inhibiting HIF” (FIH) which is an aspariginyl
hydroxlyase that binds to the inhibitory (IH) domain of HIF1α and HIF2α proteins
and results in hydroxylation of an asparagine residue located in the HIFα C-TAD
(N813 of HIF1α, N851 of HIF2α) (Fig. 1.1), which disrupts the physical interaction
between HIFα and CBP/p300 co-factor proteins (Fig. 1.3) (70). Therefore, FIH
effectively inhibits HIF transcriptional activity since CBP/p300 co-factors mediate
protein interaction between transcription factors and the RNA polymerase II
general transcription machinery
Like proline hydroxylation, asparagine hydroxylation is dependent upon
substrate oxygen concentration and therefore, is not efficient under hypoxic
10
Figure 1.3
A B Hypoxia Normoxia
CBP/ P300 FIH CBP/ P300 POL II N-OH
HIFα ARNT HIFα ARNT
Transcription NoTranscription
Figure 1.3: FIH Inhibits HIF Transactivation. A) Under hypoxia HIF is in a complex with ARNT, CBP/P300 and RNA polymerase II (POL II) and is transcriptionally active. B) The presence of oxygen (Normoxia) allows for FIH- mediated asparaginyl hydroxylation of the HIFα protein interfering with CBP or P300 interaction with the HIFα C-TAD and preventing target gene transcription.
11 conditions. However, because PHDs and FIH have different sensitivities to hypoxic conditions, it has been suggested that HIF stability and activity can be differentially hydroxylated by a range of oxygen concentrations to fine-tune HIF activity and target gene expression (70).
HIF Regulation by Kinase Signaling
In addition to oxygen-dependent regulation of HIF activity, tumor specific genetic alterations in kinase signaling pathways such mitogen-activated protein kinase (MAPK) and phosphatidylinositol 3 kinase / AKT (PI3K/AKT) pathways can result in increased HIFα activity or even expression under normoxic conditions independent of oxygen concentration by inducing posttranslational modification of the HIFα proteins or synthesizing machinery (71-78).
MAPK family kinases including p38α, p38γ, and JNK have been shown to enhance transcriptional activity of HIFα by increasing HIFα protein synthesis (71-
74) or by increasing HIFα transactivation by phosphorylating residues in the HIFα
C-TAD and IH domains which is thought to stabilize the interaction between
CBP/p300 and the HIFα C-TAD (71, 72, 74). Additionally, it has been proposed that MAPK-mediated phosphorylation of the HIFα IH domain may antagonize
FIH-mediated transcriptional repression of HIF (71).
The PI3K/AKT pathway can enhance HIF activity in tumor cells by increasing the rate of HIFα protein synthesis through the downstream effector protein, “mammalian target of rapamycin” (mTOR) (75-77). mTOR phosphosphorylates p70 ribosomal protein S6 kinase (p70s6k), a kinase that 12
enhances the translation of mRNAs that have 5’-polypyrimidine tracts which can be found in HIF1α (78). Additionally, mTOR can hyperphosphorylate the translational regulatory protein, 4E-binding protein (4E-BP), which results in
increased translation rates of HIFα protein. Therefore, HIFα proteins can
accumulate and overwhelm PHD- and FIH- mediated inhibition, thus increasing
HIF target gene expression (56). Importantly, gain-of-function mutations in
signaling components of the PI3K/AKT pathway, or upstream signaling pathways
such as H-RAS, or loss-of-function mutations in PI3K/AKT inhibitory pathways
such as PTEN are often observed in tumor cells and can lead to increased HIFα
protein even under normoxic conditions, resulting in intratumoral HIF target gene
expression and tumorigenesis
HIF in Normal Physiology
Hypoxia occurs in a number of normal physiological processes.
Interestingly, but not surprisingly, HIFs are involved in promoting these
processes through activation of their target gene programs, often functioning to
increase cell proliferation and tissue vascularization. In fact, HIF1α and HIF2α
are required for the development, maintenance and repair of mammalian tissue
and misregulation of their downstream pathways can result in severe defects and
disease (1, 7-10).
13
Embryonic Development
Processes affecting mammalian embryo development are tightly regulated
spatially and temporally in order to maintain proper embryonic maturation.
Normally, oxygen is depleted in the developing embryo due to the immature state
of the developing vasculature and oxygen consumption by proliferating cells (45).
Thus, embryogenesis takes place in the hypoxic environment of the uterus,
which exhibits oxygen concentrations of 1-5% (45, 79). The hypoxic embryonic microenvironment results in the expression of HIFα proteins and consequent activation of HIFα target gene expression and acts as a morphogen for normal
tissue development. HIF target genes including vascular endothelial growth
factor (VEGF) promote increased vascularization of the embryonic tissue by
promoting endothelial cell proliferation and vascular permeability, thereby
increasing vascularization and oxygenation and affecting the proliferation of
embryonic cell populations (80, 81). Importantly, HIF1α and HIF2α are required
for embryonic development since embryonic lethality has been observed in
HIF1α or HIF2α knockout mice (81-83). Interestingly, HIF1α and HIF2α knockout
mice exhibit unique phenotypes, suggesting non-redundant roles for HIF1α and
HIF2α in development. For instance, HIF1α-null mouse embryos exhibit defects in neural tube closure beginning at embryonic day E8.0, with abrogation of
cephalic vascularization, decreased expression of the glycolytic enzyme
phosphoglycerate kinases (PGK), and increased neuroectodermal hypoxia and
apoptosis (80). Moreover, HIF1α knockout mice also exhibit embryonic lethality
on day E11 because of defective vessel formation in the placenta, yolk sac, and 14
branchial arches caused by decreased expression of VEGF (80, 84).
Additionally, conditional knockout of HIF1α in specific cell-types has
demonstrated important roles in the production of cartilage (85), adipogenesis
(86), osteogenesis (87), hematopoiesis (88), B lymphocyte development, T
lymphocyte differentiation (89), and innate immunity (90).
In contrast, HIF2α knockout mice, depending on the genetic background,
display defects in blood vessel formation, lung maturation, or embryonic lethality
caused by bradycardia as the result of defective catecholamine synthesis
observed between days E9.5 and E13.5 or die several months after birth due to
ROS-mediated multiorgan failure (81-83, 91).
Wound Healing
Initially, a wound undergoes disrupted vasculature and increased oxygen
consumption which results in a hypoxic microenvironment (92) and subsequent
stabilization of HIFα. Interestingly, HIF1α promotes the wound healing response
by inducing the expression of target genes such as angiopoietin 2 (ANGPT2),
platelet-derived growth factor-B (PDGF-B), placental growth factor (PLGF), and
VEGF that promote vascularization and vascular remodeling by inducing
mobilization of circulating angiogenic cells (CACs), and endothelial cell
proliferation (8, 9), resulting in reoxygenation of wounded tissue and a significant
decrease in wound healing time (93). Interestingly, high glucose concentrations
in humans due to diabetes mellitus can inhibit the hypoxic stabilization of HIF1α protein in dermal fibroblasts and result in reduced expression of HIF target genes 15 causing impaired wound healing responses in diabetic humans and in mouse models of diabetes (93-96).
Stem Cell Maintenance
Adult stem cells must endure throughout life to promote continuous replacement of dead or damaged cells in various tissues of the body.
Interestingly, studies have suggested that various types of stem cells reside in a hypoxic microenvironment, which may be conducive to stem cell longevity by reducing oxidative DNA damage (7). Moreover, hypoxia was found to regulate differentiation of stem/precursor cells in culture. For instance, hypoxia was found to increase proliferation and reduce apoptosis of CNS precursor cells (97) and promote survival and proliferation of neural crest stem cells coupled with a broader differentiation potential than in normoxia (98). Experiments also demonstrated that culturing human bone marrow hematopoietic stem cells under hypoxia promotes their engraftment and repopulation of the hematopoietic compartment in immunodeficient NOD/SCID mice (99). Additionally, HIF activity has been shown to block differentiation of myogenic satellite cells and neural stem cells in a Notch-dependent manner (100). For instance, the Notch intracellular domain interacts with HIF1α during hypoxia resulting in Notch stabilization and recruitment to Notch-responsive promoters, which results in activation of Notch target genes such as HES and HEY, that are involved in maintaining de-differentiation and pluripotency of hypoxic stem cell populations
(101). 16
Interestingly, HIF1α was also shown to be essential for telomere
maintenance in mouse embryonic stem cells since HIF1α knockdown resulted in
telomere shortening and cellular sensence (102).
HIF2α has also been shown to play an important role in stem cell
maintenance. For instance, a study demonstrates that the OCT4 gene (also
called pou class 5 homeobox 1 (POU5F1)), is a direct HIF2α target gene whose
expression may be increased in normal physiology to maintain hypoxic stem cell
niches (103). Oct4 forms a transcriptional network with Sox2 and Nanog to
regulate a number of genes which maintain an undifferentiated cell fate and promote pluripotetancy (104, 105). Consistent with a role of HIF2 in promoting
stem cell maintenance, HIF2α deficient mouse embryos exhibit reduced numbers of primordial germ cells which require Oct4 for survival and/or maintenance
(106). Interestingly, HIF2α overexpressing mice exhibit severe developmental abnormalities and phenotypes consistent with expanded expression of Oct4, including aberrant tissue patterning and gene expression (103), exemplifying the importance of maintaining tight regulation and expression of HIFs.
HIFs in Tumorigenesis
Large regions of chronic hypoxia exist within solid tumors (107, 108).
These areas of hypoxia form because of an increase in oxygen consumption by rapidly proliferating cells and disorganized tumor vasculature (108). Tumor hypoxia is an indicator of a poor prognosis due to increased tumor aggressiveness and reduces response to conventional tumor therapies (16-20, 17
109-112). Moreover, hypoxia is an important selective force that promotes the development of many cancers. In response to oxygen deprivation experienced within hypoxic regions of solid tumors, tumor cells activate specific gene programs which can promote tumorigenesis. Activation of hypoxia-induced programs in solid tumors result in an increase of VEGF (vascular endothelial growth factor) and subsequent vasculature formation in a process called angiogenesis (113, 114). Increased vascularization of the tumor partially restores oxygen tension, yet due to the irregular morphology and poor quality of the new vasculature, many regions within the solid tumor still experience a chronic state of hypoxia. Importantly, tumor angiogenesis, a “hallmarks of cancer”, is crucial to the formation and survival of tumors (115). Additionally, hypoxia, through HIF-mediated induction of HIF target genes also promotes the other five “hallmarks of cancer”, by driving cell survival, promoting proliferative signaling, promoting evasion of growth suppressors, inducing replicative immortality, and activating invasion and metastasis (110, 116).
Tumor Initiation
VHL disease. Patients afflicted with VHL disease lack functional VHL protein (pVHL), the tumor suppressor protein required for degradation of the
HIFα protein under normoxic conditions (52, 60, 62-65) (Fig. 1.2). Therefore, HIF target genes are constitutively activated in these patients. VHL disease is characterized by hemangioblastomas of the brain, spinal cord, and retina; pheochromocytoma, pancreatic cysts and neuroendocrine tumors; 18 endolymphatic sac tumors; and epididymal and broad ligament cysts, renal cysts and clear cell renal cell carcinoma (117-129). Importantly, 70% of patients with
VHL disease develop clear cell renal cell carcinoma (130). Moreover, the majority of sporadic renal cell carcinomas also have loss-of-function mutations in
VHL, exhibiting highly vascularized tumors similar to what is observed in VHL disease and demonstrating the importance of HIF activity in promoting tumor initiation (131-134).
Cancer stem cells. Several studies have demonstrated that cancers can grow from small subpopulations of cancer stem cells or malignant cells possessing stem-like qualities (135-138). Cancer stem cells are similar to normal stem cells in that they self renew and produce more committed progenitor cells whose progeny differentiate, although abnormally, to produce the bulk of the tumor (7) Cells exhibiting cancer stem cell-like properties have been identified in hematopoietic, brain, and breast cancers (135, 136, 139) and hypoxia has been shown to promote these characteristics in prostate cancer cell lines (140).
Constitutive HIF activity has been implicated in the promotion and maintenance of cancer stem cells through transcriptional activation of Oct4 and through activation of Notch signaling, which maintain an undifferentiated cell fate and promote pluripotetancy (7, 103, 141). Moreover, HIF2α may promote cell-cycle progression and transformation of cancer stem cells by enhancing transcriptional activity of c-Myc, a well established oncogene (7, 142) .
19
Tumor Angiogenesis
In the initial stages of tumor growth, oxygenation of the tumor mass can occur through diffusion. However, once the tumor reaches a diameter of 2mm, neovasculization must occur to supply oxygen and nutrient rich blood to the tumor to prevent cell death and tumor necrosis (2). Interestingly, the hypoxic environment within tumor cells promotes HIF stabilization and subsequent expression of HIF target genes such as VEGF, which promotes angiogenesis and tumor vascularization (113, 114). VEGF is a growth factor which stimulates vascularization by binding to the endothelial cell receptor tyrosine kinases, VEGF receptor-1 (VEGFR1/FLT-1; fms-like tyrosine kinase-1) and VEGFR2 (FLK-
1/KDR; kinase insert domain receptor) (143-146). Binding of VEGF to FLK-1 induces phosphorylation and activates the RAF/MEK/ERK pathway which promotes endothelial cell proliferation, the PI3K/AKT pathway promoting vascular permeability and endothelial cell survival, and the p38/MAPK pathway resulting in endothelial cell migration (147-153).
Interestingly, the FLK-1 gene is a HIF target gene in endothelial cells, suggesting that hypoxia stimulates endothelial proliferation by upregulating
VEGF signaling through increased production of both the VEGF ligand and receptor (154).
Matrix metallo-proteinase-2 (MMP2) is also a direct HIF target gene that is induced by hypoxia in endothelial cells which functions to degrade the cellular matrix, allowing endothelial cells to infiltrate and migrate into the tumor in response to hypoxia (155). Once invading endothelial cells adhere to one 20
another within the tumor, stalks and tubules begin to form making up the primitive
vasculature network (45). Vessel budding and branching during this stage of
angiogenesis occurs through a Notch-dependent mechanism which is enhanced
by activation of the HIF target gene and Notch ligand, delta-like-4 (DLL4) (100,
156-160). Tumor hypoxia also activates fibronectin, a HIF2α target gene, which
supports the newly developed vasculature through formation of a basement
membrane in a process known as vessel normalization (156).
However, tumor vascularization does not completely alleviate hypoxia
within the tumor since the rapidly proliferating cells consume large amounts of
oxygen and since the newly formed tumor vasculature is highly irregular, leaky,
and poorly functioning, containing arterio-venous shunts, blind ends, a lack of
smooth muscle or nerves, and incomplete endothelial linings and basement
membranes which hinders blood flow and nutrient delivery (16, 18, 108). As a
result, HIF is chronically activated and continues to promote tumor
vascularization and cellular proliferation by inducing the expression of growth
factor genes including EPO, VEGF, angiopoietin-2 (ANGPT2), platelet-derived growth factor-B (PDGFB), stem cell factor (SCF), stromal-derived factor-1
(SDF1), and placental growth factor (PLGF) (161-164).
Epithelial-Mesenchymal Transition and Metastasis
HIF activity has also been implicated in promoting epithelial-mesenchymal transition (EMT), a cellular phenotype associated with tumor metastasis (165-
168). EMT involves the transcriptional reprogramming of epithelial cells, 21
resulting in a loss of cell polarity and cell junction proteins and a gain in migratory
and invasive properties. Cells undergoing EMT have increased expression of gene programs associated with mesenchymal-like cell signaling pathways and cell markers including fibronectin, collagen I, and metalloproteinases resulting in
increased tumor cell migration, survival and metastasis (169, 170). Importantly,
the HIF target genes, Snail and Twist, are involved in promoting EMT (171, 172).
Hypoxia as well as HIFα overexpression have been shown to promote EMT and
a metastatic phenotype through increased expression of EMT-promoting
transcription factors, Snail, Twist, ZEB1, and ZEB2 (165, 166, 173, 174).
Tumor Metabolism
Increased HIF activity can result in an upregulation of glucose transporters
such as glucose transporter-1 (GLUT1) as well as a number of key glycolytic
genes including phosphoglycerate kinase-1 (PGK1) and lactate dehydrogenase
A (LDHA) (175) and also genes such as pyruvate dehydrogenase kinase (PDK1)
which inhibit the TCA cycle, thus promoting a shift in glucose metabolism from
oxidative phoshorylation via the TCA cycle to glycosis. While this metabolic shift
logically occurs during hypoxia in normal and tumor cells due to increased HIF
activity; tumor cells become increasingly dependent on glycolysis even when
they are in an oxygenated microenvironment. This dependence upon aerobic
glycolysis is referred to as the Warburg effect (176). Misregulation and
constitutive stabilization of HIFα protein subunits in tumor cells due to aberrant
kinase signaling or mutations in HIF regulatory factors such as VHL, likely 22
contribute to the propagation of this aberrant metabolic phenotype and tumor development and may provide a mechanism for the Warberg effect in tumors.
Interestingly, accumulation of TCA cycle intermediates such as succinate
may also promote HIF-mediated aerobic glycosis. For instance, studies have
shown that mutations in the succinate dehydrogenase tumor suppressor gene
can cause accumulation of succinate (177, 178). Consequently, succinate
escapes the mitochondria and binds to the cytoplasmic PHD proteins, out
competing 2-oxoglutarate which is necessary for PHD function. As a
consequence, HIFα proteins are stabilized under normoxia and transcription of
GLUT1, LDHA, PGK1, and PDK1 is activated, thus promoting aerobic glycosis
and the Warburg effect.
Cancer Therapy Resistance
Many studies have observed that hypoxic tumor cells are resistant to
conventional anti-cancer therapies including radiotherapy and chemotherapy (16-
20, 109-112, 179). For instance, overexpression of HIF1α in tumor biopsies is an
indicator of an increased risk of radiation therapy resistance (180, 181). Initially,
radiotherapy resistance of hypoxic tumors was attributed to reduced production
of radiation-induced free radicals (179). These free radical are required for the
DNA-damaging and cytotoxic effects of irradiation. However, more recent
discoveries have identified additional mechanisms of hypoxia-mediated
radiotherapy resistance involving HIF stabilization and target gene activation
(182, 183). For instance, HIF1α knockout in transformed mouse embryonic 23 fibroblast confers increased sensitivity to radiotherapy (184). Interestingly, reoxygenation of irradiated tumor cells increases the generation of cellular reactive oxygen species (ROS) (182). ROS can mediate stabilization of HIFα protein independent of tissue oxygen tension through depletion of ascorbate, a molecule required for the function of prolyl hydroxylases and consequent degradation of HIF (185). Consequently, HIF stabilization enhances expression of HIF target genes including VEGF and basic fibroblast growth factor (bFGF) which prevent radiation-induced endothelial cell death (182). Thus, HIF activation inhibits destruction of the tumor vasculature, one of the most effective mechanisms of radiation therapy (186). Interestingly, treatment of cancer cells with ascorbate ex vivo inhibits expression of HIF1α (187)
Additionally, tumor hypoxia is a negative indicator of chemotherapeutic response (18). The first barrier to effective chemotherapeutic treatments is their delivery to tumor cells. Because chemotherapeutic drug delivery is achieved through the bloodstream, delivery is dependent upon the vascular system, which is largely inadequate and abnormal within hypoxic regions of solid tumors (16).
Additionally, the effectiveness of chemotherapeutic drugs is limited by the activation of HIF target genes that antagonize the mechanism of drug action, which is the case for chemotherapeutics such as carboplatin and etoposide.
Carboplatin inhibits DNA replication and induces tumor cell apoptosis by forming
DNA adducts. Etoposide is a topoisomerase II inhibitor. Both Carboplatin and etoposide (and also radiotherapy) create double-stranded DNA breaks leading to apoptosis. However, hypoxic conditions reduce the effectiveness of these drugs 24 by inducing HIF target genes such as the multidrug resistance gene (MDR1) which enhances double strand break repair (184, 188).
Pre- mRNA Splicing
Pre-mRNA splicing occurs in nearly all human genes and is a crucial regulatory stage in gene expression. Pre-mRNA splicing, like 5’ capping and cleavage/ polyadenylation of mRNA, is a co-transcriptional process in which introns are removed from nascently transcribing mRNAs and exons are ligated or spliced together to form mature RNAs. RNA splicing was first discovered in the
1970’s creating a paradigm shift in the field of gene expression which previously believed that DNA directly encoded the mature RNA sequence in a one-to-one correspondence of bases between the gene and the mature mRNA which is typically the case for gene expression in bacteria. However, in the late 1970’s researchers studying adenoviruses indentified a series of RNA molecules which contained sequences from non-adjoining sites in the viral genome, which were termed mosaics (189, 190). These RNA mosaics were found in late viral infection; however, studies of early infection revealed long primary transcripts that contained all of the sequences from the late viral RNAs as well as intervening or intron sequences.
Subsequently, introns were discovered in many other viral and eukaryotic
RNAs including hemoglobin and immunoglobulin (191). Moreover, splicing of
RNA transcripts was observed in in vitro systems derived from eukaryotic cells including intron removal from RNAs in yeast cell-free extracts (192). Together, 25 these findings established that intron removal and splicing of large precursor mRNA (pre-mRNA) was responsible for the generation of mature mRNA.
Splicing Reaction
The biochemical mechanism by which mRNA splicing occurs is now well characterized. Introns are removed from nascent pre-mRNA by cleavage at conserved intron/exon junctions called splice sites. The 5’ splice site designates the exon/intron junction at the 5’ end of the intron and typically includes a GU dinucleotide within a larger, less conserved, GURAGU consensus sequence (Fig.
1.4) (193). At the opposite end of the intron are three conserved sequence elements, the branch point (consensus sequence:YNCURAC) which is 18-40 nucleotides upstream from the 3’ splice site, followed by the polypyrimidine tract and by the 3’ splice site that contains a terminal AG dinucleotide consensus sequence. mRNA splicing is carried out by a large macromolecular machine called the spliceosome which assembles onto these RNA sequences and catalyzes two transesterification splicing reactions (Fig 1.4) (194-196). First, the
2’-hydroxyl group of the branch point adenosine residue attacks the phosphate at the 5’ splice site which results in cleavage of the 5’ splice site from the 5’ exon followed by ligation of the 5’ splice site to the branch point. This produces two intermediates, a detached 5’ exon and a 3’ exon/intron lariat structure. During the second transesterification reaction, the 3’-hydroxyl residue of the detached 5’ exon attacks the phosphate residue at the 3’ splice site resulting in ligation of the two exons and release of the lariat structured intron. 26
The spliceosome assembles onto each intron using five small nuclear ribonucleoproteins (snRNPs), U1, U2, U4, U5, and U6. Each snRNP consists of a small nuclear RNA (snRNA), a common set of seven Sm proteins (B/B’, D3,
D2, D1, E, F, and G) and a variable number of other proteins (197). During the splicing reaction, the U1 snRNP binds to the 5’ splice site of the intron through base pairing of the U1 snRNA with the splice site (Fig. 1.4). Simultaneously, the branch point sequence is bound by the splicing factor 1(SF1) protein, while the polypyrimidine tract sequence is bound by a 65-kDa subunit of the dimeric U2 auxiliary factor (U2AF), forming the earliest defined spliceosome complex called the E complex (Fig 1.5) (198). Subsequently, the U2 snRNP displaces the SF1 protein in an ATP dependent manner and binds to the branch point sequence through base-pairing of its snRNA, forming the A complex. Next, the U4/U5/U6 tri-snRNP displaces the U2AF complex and binds to the polypyrimidine tract sequence forming the B complex. The B complex then undergoes a series of complicated rearrangements in which the U1 and U4 snRNPs are removed from the complex allowing the U6 snRNP to interact with the 5’ splice site, forming the
C complex. The C complex is responsible for catalyzing the two transesterification splicing reactions described above.
Alternative Splicing
Most human genes, over 90%, are expressed as pre-mRNAs that must be spliced to produce mature mRNAs. Splicing is an essential process of gene expression that is necessary for embryonic development and survival. Whereas 27
Figure 1.4
Figure 1.4: Two Step Pre-mRNA Splicing Reaction by the Spliceosome. Splicing takes place in two transesterification steps. The first step results in a detached 5’ exon (exon1) and an intron/3’ exon (exon2) lariat structure. The second step ligates exons 1 and 2 and releases the intron lariat. Boxes and solid lines represent exons and introns respectively. Conserved sequences at the 5’ and 3’ splices, branch point and polypyrimidine tract (Y(n)) are indicated. R= purine nucleotides, Y= pyrimidine nucleotides 28
Figure 1.5
Figure 1.5: The Spliceosome Consists of Five Small Nuclear Ribonucleoproteins (snRNP) that Assemble onto the Intron to Facilitate Splicing. The E (early) complex contains the U1 snRNP bound to the 5’ splice site (GU), the SF1 protein bound to the branch point (A), the U2AF 65 complex bound to the poly pyrimidine tract (Y(n)) and the U2AF 35 bound to the 3’ splice site (AG). The A complex is formed by the replacement of SF1 with U2 snRNP at the branch point. Subsequently, U2AF is replaced by the U4/U5/U6 tri-snRNP forming the B complex. Through a series of rearrangements the C complex is formed, producing the catalytically active spliceosome which carries out the two- transesterification steps described in Figure 1.4. 29 some exons are present in every mRNA of a particular RNA, and are said to be constitutively spliced, many exons are alternatively spliced (AS) or spliced together in various combinations to generate variable forms of mature mRNA from a single pre-mRNA species. Typical multi-exon mRNAs can be spliced in many different patterns (Fig 1.6). Most exons are constitutive, meaning that they are always spliced or included in the mature mRNA. However, some exons are regulated and are sometimes included or excluded in the mature mRNA and are called cassette exons. Alternative splicing of cassette exons is the most common and most studied form of AS. Occasionally, cassette exons are mutually exclusive meaning that only one of multiple alternative exons is included in the mature mRNA but never all exons together. Additionally, mRNAs can be spliced such that an exon can be lengthened or shortened by altering the position of either the 5’or 3’ splice site. Also, the 5’-terminal exon of an mRNA can be switched through the use of alternative promoters, however alternative promoter usage is primarily regulated at the level of transcription. Likewise, the 3’-terminal exon of an mRNA can be switched by combining alternative splicing and alternative polyadenylation sites, polyadenylation is mechanistically similar to splicing. On rare occasion, some mRNAs undergo a splicing pattern called intron retention in which an intron is not spliced out of the mature mRNA. If the retained intron is in the coding region of the mRNA, it must encode amino acids in frame with the neighboring exons or a frame shift may lead to the expression of aberrant or non functional protein, or a premature stop codon located in the intron may cause the mRNA to undergo non-sense mediated decay (NMD). 30
Figure 1.6
Figure 1.6: Alternative Splicing Patterns. A) Cassette exons can be included or excluded in the mRNA. B) Mutually exclusive exons occur when two or more adjacent cassette exons are spliced in which only one of the exons is included in the mRNA at one time but never all exons together. C-D) Alternative exons can be shortened or lengthened through the use of alternative 5’ or 3’ splice sites. Alternative promoter E) or poly adenylation F) occurs through the use of different 5’ and 3’ terminal exons. G) Introns can be removed or retained in the mature mRNA. Constitutive exons are in blue boxes. Alternative exons or introns may be included or excluded in mature mRNA and indicated by purple boxes. Dotted lines represent splice site usage.
31
Intron retention is the rarest type of splicing in mammals but is one of the most
common forms of alternative splicing in plants (199). Importantly, particular
mRNAs commonly have multiple sites of AS and give rise to a family of related
proteins from a single gene (196).
The Role of Alternative Splicing in Development
AS is prevalent in higher eukaryotes and it is estimated that over 90% of
human genes are alternatively spliced. Moreover understanding splicing at the
molecular level is important for comprehending gene expression programs.
Importantly, the primary functions of AS is to enhance protein diversity by
increasing the number of proteins that can be expressed by a single gene (200).
For example, it is estimated that the human body is made up of over 100,000
proteins; however, with the completion of the human genome project it was found
that the human genome only encodes about 20,000-25,000 genes. Therefore,
AS can account for this apparent discrepancy by producing multiple proteins from
a single pre-mRNA. Moreover, splicing is absolutely required for development
since gene knockout of splicing factors such as SRp20, AF/SF2, SC35, hnRNP
U, and hnRNP C in mice results in embryonic lethality or in lethality 1-2 weeks
after birth which is the case for nova-1, a neuron-specific splicing factor (201-
206).
Not only is AS essential for normal development and protein diversity, it is
also of medical importance since abnormal pre-mRNA splicing is a cause of
many human diseases or contributes to their severity, including spinal muscular 32
atrophy, dementia, cystic fibrosis, and cancer (207-211). In fact, at least 15% and as many as 50% of genetic mutations that cause disease alter pre-mRNA
splicing (212, 213).
The Role of Alternative Splicing in Cancer
Thus far, aberrant or cancer specific alternative splicing has been
observed for over a hundred different genes, however there is still debate as to
whether mis-splicing is a cause of cancer or a consequence of the disease.
However, recent studies suggest that the ASF/SF2 splicing factor acts as a
proto-oncogene and is capable of transforming immortalized cells, while
reduction of ASF/SF2 levels in transformed cells reduces the malignant
phenotype (214). Moreover, ASF/SF2 is up-regulated in many cancers,
suggesting that mis-regulation of splicing may potentially cause tumorigenesis.
Regardless, it is clear that cancers can utilize alternatives splicing pathways to
their advantage.
Alternative Splicing in Apoptosis. For instance, many genes that
regulate cell apoptosis are alternatively spliced, producing isoforms with pro-
apoptotic or anti-apoptotic functions. This is exemplified by the Bcl-x gene which
is alternatively spliced to produce a pro-apoptotic (Bcl-xs) and an anti-apoptotic
isoform (Bcl-xL) (215-217). Interestingly, various cancer types express high
ratios of anti-apoptotic to pro-apoptotic isoforms (Bcl-xL/Bcl-xs) consistent with
an important role of Bcl-xL in cancer cell survival (218). 33
Alternative Splicing in Cancer Metabolism. AS also plays an important
role in the control of cancer metabolism through the regulation of pyruvate
kinase, muscle (PKM), a key metabolic gene. PKM is expressed in all
mammalian tissues except for the liver and erythrocytes (219). The PKM gene
produces two major isoforms, PKM1 and PKM2, through mutually exclusive AS.
Additionally, PKM isoform expression is regulated during development and
embryonic cells express PKM2 whereas differentiated cell express PKM1.
Interestingly, the PKM2 isoforms is re-expressed in various cancers and studies demonstrated that replacement of PKM2 with PKM1 in cancer cells reduced the production of lactate and enhanced oxidative phoshorylation, implicating PKM2 as a promoter of the Warburg effect (220).
Alternative Splicing in the Regulation of Proto-Oncogenes. Cyclin D1
is a well-known regulator of the cell-cycle through its interaction with cyclin-
dependent kinase 4 or 6 and functions to phosphorylate the retinoblastoma (Rb)
tumor suppressor protein thereby relieving Rb’s repression of E2F target genes
which results in cell-cycle progression. Cyclin D1 undergoes AS to produce a
full-length cyclin D1a and a shorter cyclin D1b isoform (221). Although both
isoforms are expressed in normal tissue, the cyclin D1b isoform is frequently
overexpressed in some cancers including breast and prostate cancers.
Interestingly, cyclin D1b was found to be a more potent inducer of transformation
than cyclin D1a when overexpressed in immortalized cells.
Alternative Splicing in Metastasis and Invasion. Several genes,
including CD44, FGFR2, RAC1 and Ron, that play important roles in promoting 34
invasive behavior are regulated by AS (222). Interestingly, EMT was shown to
be accompanied by a reprogramming of AS (169) which may be due to the
down-regulation of epithelial-specific splicing factors, ESRP1 and ESRP2 (223).
Alternative Splicing in Angiogenesis. Angiogenesis is another
important hallmark of cancer that was shown to be regulated by AS (224). For example; the VEGFA gene, which is a major inducer of angiogenesis, undergoes
extensive alternative splicing in normal cells producing pro-angiogenic isoforms,
VEGF121a, 165a and 185a, and anti-angiogenic isoforms, VEGF121b, 165b,
and 185b. Interestingly, VEGFa isoforms were up-regulated in kidney carcinoma
which is consistent with these tumors being highly vascularized (224). However,
the VEGFb isoforms were not detected. Conversely, down-regulation of VEGFb
isoforms was observed in prostate carcinoma and melanoma (225).
Cis Elements Regulate Splicing
Constitutive and alternative splicing of exons is carried out by the
spliceosome and facilitated by trans-acting factors, such as serine-arginine (SR)
proteins, heterogenous nuclear ribonucleoproteins (hnRNPs) and other RNA
binding proteins that recognize and bind to short highly degenerate cis-acting
sequence elements within exons or introns (196). Cis-acting elements located in
exons can either promote splicing and exon inclusion and are known as exonic
splicing enhancers (ESE) or can inhibit splicing and promote exon skipping and
are known as exonic splicing silencers (ESS) (Fig 1.7). Likewise, cis-acting
elements located in introns can either promote splicing and exon inclusion and 35
Figure 1.7
Figure 1.7: Function of SR Proteins in Splice Site Selection. A) SR proteins bound to exonic splicing enhancers (ESE) promote exon inclusion by recruiting U1 and U2AF to 5’ and 3’ splice sites. B) SR proteins bound to ESEs promote exon inclusion by inhibiting hnRNPs bound to exonic splicing silencers (ESS). hnRNPs inhibit the bind of spliceosomal components. C) SR proteins promote exon inclusion by facilitating bridging interaction across introns by binding to intronic splicing enhancers (ISE). D) SR proteins and hnRNPs bound to intronic splicing silencers (ISS) can promote exon skipping by inhibiting the binding of spliceosomal components.
36 are called intronic splicing enhancers (ISE) or can inhibit splicing and promote exon skipping and are called intronic splicing silencers (ISS).
Trans-acting Factors Regulate Splicing
Generally, ESEs and ISE are recognized and bound by SR proteins.
Classical SR proteins regulate constitutive and alternative splicing and contain at least one RNA recognition motif (RRM) that is important for RNA binding and an arginine and serine rich (RS) domain that is important for protein-protein interactions and RNA binding. Classical SR proteins include, SF2/ASF, SC35,
SRp20, SRp75, SRp40, SRp55, and 9G8.
SR proteins are prominent components of nuclear speckles and are localized in nuclear interchromatin granules, (ICGs) which are storage and/or reassembly sites for these and other splicing factors, or localized in perichromatin fibrils (PCFs) which are the sites of actively transcribing genes and co-transcriptional splicing. The intranuclear localization and organization of SR proteins is highly dynamic. SR proteins are activated and recruited from the ICG storage clusters to the sites of co-transcriptional splicing (PCFs) via phosphorylation of their RS domains by a number of protein kinases including SR protein kinase 1 and 2 (SRPK1, SRPK2) and CDC-like kinase 1, 2, 3 and 4
(CLK1, CLK2, CLK3, CLK4). Once activated by phosphorylation and recruited to the active sites of transcription, SR proteins enhance or inhibit the binding of the spliceosome on nascent pre-mRNAs and regulate constitutive and alternative splicing. 37
SR proteins are thought to regulate splice site selection and promote exon
inclusion by facilitating recruitment of spliceosomal components such as U1
snRNP and U2AF to 5’ and 3’ splice sites respectively (Fig. 1.7) (226).
Alternatively, SR proteins may promote exon inclusion by antagonizing the
function of splicing silencers such as hnRNPs bound to ESSs or ISSs that block
the recruitment of spliceosomal components to the 5’ and 3’ splice sites (Fig.
1.7B). Additionally, SR proteins may promote exon inclusion by forming a
network of protein-protein interactions thus facilitating intron bridging interactions
with U1 and U2AF thus positioning the 5’ and 3’ splice sites next to one another
during early spliceosomal assembly (Fig. 1.7C). Interestingly, SR proteins can
also bind to ISSs and promote exon skipping by perturbing the binding of U2AF
on the polypyrimidine tract; therefore SR proteins can facilitate exon inclusion or
exon skipping depending on their binding location (Fig. 1.7D). In addition to regulating splicing of pre-mRNA, a subset of SR proteins such as SF2/ASF,
SRp20 and 9G8 shuttle between the cytoplasm and nucleus and promote nuclear mRNA export and enhance translation.
On the other hand, ESSs and ISSs are bound by hnRNPs which inhibit binding of spliceosomal components to the 5’ and 3’ splice sites, thus promoting exon skipping. HnRNPs are some of the most abundant proteins in the nucleus and regulate pre-mRNA processing including splicing, mRNA export, mRNA localization, mRNA stability and translation (227).
38
SR Protein Regulation
Generally speaking, SR proteins are ubiquitously expressed in all tissues
and cells types, however differential expression of some SR proteins has been
observed for specific cell types in response to cell signaling. SR proteins directly
control splicing and are regulated at the transcriptional, splicing and post-
translational levels. Very little is known about how SR protein expression is controlled at the transcriptional level. However, SR proteins have been found to
undergo non-productive splicing, leading to NMD through auto-regulation or
regulation in trans by other SR proteins. Interestingly, regulation of SR protein
function primarily occurs at the post-translational level. SR proteins are
phosphorylated, acetylated, methylated, ubiquitylated and sumoylated which
controls the subcellular localization, protein-protein and protein-RNA interactions,
and activity of these proteins (228-231). These control mechanisms of SR
protein expression and activity act to maintain SR protein homeostasis in most
cells. In this thesis, I will focus on phosphorylation of SR proteins since the
consequences of SR protein phosphorylation on splicing regulation are well
characterized.
SR proteins undergo highly dynamic cycles of phosphorylation and
dephosphorylation that is required for pre-mRNA splicing (Fig. 1.8) (232).
Studies indicate that SR protein phoshorylation is required for spliceosome
assembly where as dephosphorylation is essential for splicing catalysis within the
spliceosome. SR proteins are phosphorylated in their RS domains by two major
protein kinase families in vivo, including the SR protein kinase family (SRPK1-3) 39 and the dual-specificity CDC-like kinase family (CLK1-4). Additionally, several other signaling kinases have been shown to phosphorylate SR proteins in vitro, including protein kinase A (PKA) and protein kinase C (PKC) (233), AKT (234,
235), topoisomerase I (236), and dual specificity tyrosine phosphorylation- regulated kinases (DYRKs) (237, 238).
SRPKs are primarily localized in the cytoplasm with a small fraction localized in the nucleus during interphase of the cell cycle (239, 240). SRPK1 was the first protein discovered to phosphorylate the RS domains of SR proteins that is necessary for spliceosomal assembly and function (241) (Fig. 1.8).
Importantly, SRPK-induced phosphorylation of a subset of SR proteins that shuttle between the cytoplasm and the nucleus, such as ASF/SF2, SRp20, 9G8 and SRp40, induces their nuclear import by enhancing binding affinity for transportin-SR, a nuclear import receptor (242, 243). Following nuclear import, hypophosphorylated SR proteins accumulate in nuclear stress granules called nuclear interchromatin granules (ICGs). ICGs act as storage and/or reassembly sites for SR proteins.
In contrast to SRPKs, CLKs are localized in the cell nucleus. Importantly,
CLKs are activated through autophosphorylation and once activated, they translocate to ICGs where they further phosphorylate SR proteins (SRps), resulting in ICG disassembly and activation of both nuclear and shuttling SR proteins. Free, hyperphosphorylated SR proteins then relocalize within the nucleus to sites of actively transcribing genes called perichromatin fibrils (PCFs), where they participate in spliceosomal assembly and co-transcriptional splicing. 40
Figure 1.8
Figure 1.8: Phosphorylation-mediated Regulation of SR Proteins Activity 1. Reversible phosphorylation of their RS domain profoundly affect SR protein (SRps) activity and subcellular localization. Newly synthesized SRps need SRPK-mediated phosphorylation in order to enter the nucleus and assemble in nuclear speckles. CLKs mediate SRps hyperphosphorylation and induce their release from nuclear speckles (ICG) and their recruitment to transcription sites (PCF). Dephosphorylation of SRps is successively required for proper splicing catalysis. Moreover, dephosphorylated SRps facilitate export of spliced mRNA in the cytosol, where they enhance protein translation.
1This Figure is reprinted with the permission of the International Journal of Cell Biology.
Naro C, Sette C. Phosphorylation-Mediated Regulation of Alternative Splicing in Cancer. Int J Cell Biol. 2013;2013:151839. 41
While SR protein phosphorylation is necessary for spliceosome assembly on pre-mRNA, dephosphorylation is critical for splicing catalysis. For instance,
inhibitors that block protein phosphatase 2 (PP2A) activity were found to inhibit
the second catalysis step of splicing whereas inhibition of protein phosphatase
1(PP1) and PP2A blocked both steps of splicing catalysis (244). Unfortunately,
very little is known about how and when dephosphorylation is triggered in vivo.
Following dephosphorylation during splicing catalysis, nuclear SR proteins
can be rephosphorylated by nuclear CLKs and recycled for subsequent rounds of
splicing (245). Furthermore, dephosphorylated shuttling SR proteins remain
associated with mature mRNA and facilitate interactions with the Tap protein, an
mRNA export factor (Fig. 1.8) (246). Thus, SR proteins couple pre-mRNA
splicing with mRNA export to the cytoplasm (247). Once in the cytoplasm, these
mRNA associated SR proteins are directed to the ribosome where they play a
direct role in mRNA translation (248). Shuttling SR proteins can be
rephosphorylated by cytoplasmic SRPKs and re-enter the nucleus where they re-
accumulate in ICGs and once re-phosphorylated by CLKs are used in
subsequent splicing reactions (245).
Coupling of Transcription and Splicing
While transcription is necessary for RNA production, in vitro studies using
pre-synthesized pre-mRNA templates mixed with nuclear extracts (249-251) or
introduced into living cells(252) suggest that active transcription is not necessary
for spliceosome assembly and intron removal. However, in vivo, transcription 42
and splicing are highly connected and it is generally accepted that splicing occurs
while the pre-mRNA is being synthesized by RNA polymerase II (pol II) or on pre-
mRNA that has been fully synthesized but remains tethered to the chromatin
template (253-255). Therefore, transcription and splicing occur in the cell
nucleus at the same time and place and are functionally coupled since a number
of factors which regulate transcription, including pol II and its Carboxyl-Terminal
Domain (CTD) (256), promoter sequences (257), transcriptional activators (258), and chromatin remodelers (259), can alter splicing outcomes. Additionally, splicing has also been observed to regulate transcription (260, 261), suggesting
that complex cross-talk occurs between these independent but highly
interconnected processes to promote proper gene expression. It is clear that
transcription and splicing are functionally coupled; however, the biochemical
mechanisms of coupling are still unknown. Two non-mutually exclusive models,
the recruitment and the kinetic models, describe how transcription and splicing
might be coupled.
The recruitment model of co-transcriptional splicing proposes that splicing
regulators such as U1 snRNP or splicing factors such as SR proteins or other
RNA binding proteins are recruited to the active site of transcription where they
are able to interact with nascent pre-mRNA and facilitate splicing reactions (Fig.
1.9). Recruitment of splicing factors to the activate site of transcription has been
proposed to occur by various mechanisms. The most studied mechanism
suggests that during transcription elongation, the CTD of RNA pol II is
phosphorylated by PTEFb on serine 2 which allows the CTD to act as a landing 43 pad for multiple mRNA processing proteins including 5’-capping factors, splicing factors and poly adenylation factors. Thus, as RNA pol II synthesizes nascent pre-mRNA, the splicing factors bound to the CTD are able to bind to pre-mRNAs and facilitate spliceosomal assembly and splicing. This model was generated from several observations. First, studies demonstrated that transcriptional activation induced re-localization of splicing factors and accumulation near areas of active transcription (262). Secondly, splicing factor re-localization is disrupted when transcription is initiated by RNA pol II lacking a CTD, implying that the CTD mediates re-localization of splicing factors during transcription (256).
Additionally, RNA pol II has been shown to interact with splicing factors in solution using co-IP experiments (263). In another study, a transcriptionally active pol II-complex purified over an SII column from Hela cell extract contained the U1, U2 and U4 snRNPs, U2AF65 and many SR proteins (264).
Although it is plausible that RNA pol II can directly recruit splicing factors to the active site of transcription thereby coupling transcription and splicing, more recent studies demonstrate that RNA pol II’s ability to interact with splicing factors is RNA dependent since RNase treatment prevents co-IP (255). This suggests that the pol II and splicing factor interactions are through RNA tethering rather than direct protein-protein interactions. Moreover, these data may imply that the presence or availability of pre-mRNA substrate is responsible for recruiting splicing factors to the active site of transcription and not RNA pol II.
Other variations of the recruitment model have been proposed in which gene promoter structure or transcription factors may recruit splicing factors to the 44
Figure 1.9
Figure 1.9: Recruitment Model of Alternative Splicing. A) The C-terminal domain (CTD) of RNA polymerase II facilitates binding of SR protein (SRp) on nascent RNA during transcription when bound to promoter X, allowing SR proteins to regulate alternative splicing by promoting exon inclusion. B) In contrast, when RNA polymerase II is bound to promoter Y it does not recruit SRps to nascent mRNA, therefore exon skipping occurs. 2
2 This Figure is reprinted with the permission of Trends in Genetics.
Cáceres JF, Kornblihtt AR. Alternative splicing: multiple control mechanismsand involvement in human disease. Trends Genet. 2002 Apr;18(4):186-93. 45 active site of transcription. For instance, studies using minigene splicing reporters show that promoter swapping regulates alternative splicing of the minigene. It was not determined how different promoters regulated splicing, but it was suggested that cis-elements within the promoters recruit different splicing factors to the promoter through transcription factors or co-factors (reviewed in
(265). However, the promoter specific effect on alternative splicing was not due to differences in reporter activation strength. The recruitment idea is also plausible and studies have shown that transcription factors or co-factors can interact with components of the spliceosome or with splicing factors (266, 267).
However, direct evidence that links the recruitment model with splicing in vivo is lacking.
An alternative but not exclusive model that could explain the coupling of transcription and splicing is the kinetic model of splicing. This model proposes that the rate of RNA pol II elongation or processivity across the body of the template DNA during transcription, determines alternative splicing outcomes (Fig.
1.10) (265). In other words, a slow pol II elongation rate or elongation in the presence of internal pauses, promotes exon inclusion, while a fast elongation rate or elongation in the absence of internal pauses promotes exon skipping. It is generally observed that alternative exons contain suboptimal 3’ splice sites compared to the downstream constitutive exons, therefore during slow elongation or in the presence of internal pause sites only the weak alternative 3’ splice site is transcribed and available to the spliceosome, thus promoting exon inclusion of
46
Figure 1.10
Figure 1.10: Kinetic Model of Alternative Splicing. In the kinetic model, A) promoter X promotes a slow or less processive Pol II that allows the preferential elimination of the upstream intron, leading to the inclusion of the alternative exon. B) On the other hand promoter Y promotes a fast or more processive Pol II. In these conditions the strong 3’ splice site of the downstream intron outcompetes the weak 3’ splice site of the upstream intron, resulting in exclusion of the alternative exon 3.
3 This Figure is reprinted with the permission of Trends in Genetics.
Cáceres JF, Kornblihtt AR. Alternative splicing: multiple control mechanismsand involvement in human disease. Trends Genet. 2002 Apr;18(4):186-93. 47 the alternative exon. Conversely, during fast elongation, RNA pol II transcribes the alternative exon containing a weak 3’ splice site as well as the downstream constitute e exon containing a strong 3’ splice site, thus, during fast elongation the splicing machinery has the option of using the weak 3’ splice site of the alternative exon or the stronger 3’ splice site of the constitutive exon. In this scenario, the strong 3’ splice site of the constitutive exon out-competes the weaker 3’ splice site of the alternative exon, resulting in exon skipping.
The kinetic model of splicing is based on observations which demonstrate that treatment of cells with dichlororibofuranosylbenzimidazole (DRB) or flavopiridol inhibits pol II elongation and promotes exon inclusion of a transfected minigene splicing reporter (268, 269). However, caution should be used when interpreting splicing data using DRB since additional studies suggest that DRB can inhibit the activity of the splicing regulators, CLK1 and 4 (270), thus the effects of DRB on elongation and splicing may be coincidental and unrelated.
More direct proof for the kinetic model of splicing comes from studies utilizing a mutant RNA pol II that exhibits a slower elongation rate. Cells were transfected with a mutant RNA pol II which is resistant to α-amanitin and minigene splicing reporters. After treatment with α-amanitin, the endogenous pol
II is inhibited and the minigenes are transcribed by the elongation-defective RNA pol II, resulting in enhanced exon inclusion (271).
Further support comes from studies which show that Brahma (BRM), a chromatin remodeler, causes RNA pol II pausing within the body of the E- cadherin, BIM, cyclin D1 and CD44 genes during transcription and that the 48
phosphorylation pattern of RNA pol II shifts from serine 2 to serine 5 during
pausing. Moreover, this pausing resulted in increased alternative exon inclusion
(259), clearly demonstrating a link between RNA pol II elongation rate and alternative splicing regulation. This data is consistent with the observation that
serine 2 phosphorylated RNA pol II is indicative of elongating or processive pol II
whereas serine 5 phosphorylated RNA pol II is indicative of non-elongating or paused pol II. Moreover, this study implies that chromatin structure or other factors which alter RNA pol II processivity can potentially regulate alternative splicing outcomes.
On the contrary, RNA pol II elongation rate becomes irrelevant during
constitutive splicing when two consecutive strong 3’ splice sites are available to
the spliceosome, since the upstream exon will always be included in the mRNA
because it contains strong splice sites and is not out competed by the upstream
exon. Therefore, while the recruitment model can explain the coupling of
transcription with constitutive and alternative splicing, the kinetic model can only
account for the coupling of transcription with alternative splicing.
Cellular Stress Regulates Alternative Splicing
Regulation of Alternative Splicing by Heat Shock. Previous studies
suggest that cellular stress regulates AS by causing relocalization of splicing
factors to cytoplasmic stress granules, which is the case for heat shock and
oxidative stress (272). Heat shock was shown to inhibit global constitutive
splicing by increasing intron retention due to dephosphorylation of the SRp38 49 splicing factor by protein phosphatase. As a result, dephosphorylated SRp38 interacts with U1 snRNP, preventing U1 association with pre-mRNA and inhibiting intron removal, consistent with the recruitment model of splicing (273).
Regulation of Alternative Splicing by Osmotic Stress. On the other hand, osmotic stress was shown to regulate alternative splicing of an E1A splicing reporter by causing dissociation of SRPK1 with Hsp70 and Hsp90 molecular chaperones, resulting in SRPK1 translocation from the cytoplasm to the nucleus and subsequent differential phosphorylation of SR proteins and alteration of splice site choice (274).
Regulation of Alternative Splicing by Uv Irradiation. Uv irradiation was shown to regulate AS of a splicing reporter and also of endogenous apoptotic genes including Bcl-x and caspase 9 by inhibiting RNA polymerase II elongation
(275). Moreover, this study showed Uv-mediated splicing was not due to SR protein relocalization suggesting that Uv-mediated AS occurs through the kinetic model.
Regulation of Alternative Splicing by Hypoxia. Several examples of hypoxia-dependent regulation of AS have been reported for individual genes
(276-280). For example, one study found that the BNIP3 gene is induced by hypoxia in rat cardiomyocytes and that the BNIP3 gene produced a full-length isoform containing exons 1-6 and a shorter isoform in which exon 3 was skipped
(280). Additionally, the shorter isoform was preferentially induced by hypoxia.
The full-length isoform was found to be pro-apoptotic whereas the shorter isoform was found to inhibit apoptosis suggesting that rat cardiomyocytes 50 selectively express the anti-apoptotic isoform which acts to reduce mitochondria- mediate apoptosis during hypoxia.
Recent studies in human umbilical vein endothelial cells (HUVECs) investigated changes in splicing in the physiological response to hypoxia on a genome-wide scale (281). Using exon arrays, these studied demonstrated that
19 genes were predicted to undergo hypoxia-mediated splicing events; ten genes had AS of cassette exons, three genes had alternative poly adenylation sites, two genes underwent intron retention and four genes underwent alternative promoter usage. The researchers investigated AS splicing of six of the ten genes predicated to undergo AS of cassette exons and were able to validate AS for three of the genes; CASK, SPTAN1 and PIGN. Further analysis revealed that all three of these genes underwent hypoxia-mediated exon skipping, although these genes were not induced by hypoxia.
In addition to regulating AS in normal cells and tissues, studies suggest that hypoxia regulates AS in neuroblastoma cancer cells. Neurotrophin tyrosine kinase receptor type 1 (TrkA) is a member of the tyrosine kinase neurotrophin receptor family that is the preferred receptor for nerve growth factor (NGF) and is critical for development and maturation of central and peripheral nervous systems, regulating proliferation, differentiation, and programmed cell death
(282-284). Studies demonstrated that the TrkA gene produces two mRNA isoforms, TrkAI and TrkAII in differentiated neurons and a third isoform, TrkAIII, was expressed in undifferentiated neuronal progenitor cells and in some neuroblastoma cell-lines and was expressed in primary human neuroblastomas 51
(276). Interestingly, the TrkAI and II isoforms are typically considered to have tumor-suppressive properties (279). However, the researchers discovered that
TrkAIII was preferentially induced by hypoxia and was able to antagonize the function of the TrkAI and II isoforms. This suggested a hypoxia-regulated mechanism for oncogenic TrkA activation in neuroblastoma cells as characterized by generation of a constitutively active TrkAIII splice variant that exhibits oncogenic properties by antagonizing tumor-suppressive NGF/TrkAI signaling.
Cysteine rich 61 (Cyr61) is a secreted protein involved in development, wound repair, angiogenesis, inflammation, cell survival, vascular diseases, and endometriosis that is induced under hypoxic conditions (285) . Hypoxia was shown to regulate AS of the Cyr61 gene in breast cancer cell-lines, resulting in the preferential induction of a protein coding (intron 3 skipping) isoform over an intron 3 retaining isoform (94). In addition, these studies determined that the protein coding isoform was highly expressed in primary breast tumors and revealed a stage-dependent induction of Cyr61 mRNA and protein in breast cancer tumorigenesis which was accompanied by a shift from an intron 3 retaining toward an intron 3 skipping mRNA phenotype, which lead to expression of the biologically active Cyr61 protein.
Although hypoxia has been shown to regulate alternative splicing of several genes, the molecular mechanism or signaling pathways involved have not been determined. We hypothesize that HIF transcription factors regulate AS of a subset of hypoxia-induced HIF target genes since mRNA splicing is 52 functionally coupled to transcription, likely through transcription factors. The following chapters will present results supporting this hypothesis and identify the molecular signaling pathway by which HIFs regulate constitutive splicing of the
HIF target gene adrenomedullin (ADM) and alternative splicing of a sub-set of
HIF target genes including carbonic anhydrase 9 (CA9), angiopoietin-like 4
(ANGPTL4), pyruvate dehydrogenase kinase 1 (PDK1), WNK lysine deficient kinase 1 (WNK1), prolyl 4-hydroxylase, alpha polypeptide II (P4HA2), procollagen-lysine, 2-oxoglutarate 5-dioxygenase 2 (PLOD2) and enolase 2
(ENO2). I will discuss how CDC-like kinases (CLKs), which are induced by hypoxia, regulate HIF-mediated mRNA splicing and expression of HIF target genes. Finally, I will discuss the role of CLKs in HIF-mediated tumorigenesis. 53
CHAPTER II
HIFS ENHANCE THE TRANSCRIPTIONAL ACTIVATION AND SPLICING OF
ADRENOMEDULLIN 4
Abstract
Adrenomedullin (ADM) is important for tumor angiogenesis, tumor cell
growth and survival. Under normoxic conditions, the ADM gene was found to produce two alternative transcripts, a fully-spliced transcript that produces AM
and PAMP peptides and a intron-3-retaining transcript that produces a less
functionally significant PAMP peptide only. ADM is a well-established hypoxia
inducible gene; however, it is not clear which ADM isoform is induced by hypoxia.
In this study, it was determined that various cancer and normal cells express two
predominant types of ADM transcripts, a AM/PAMP peptide producing FL
transcript in which all introns are removed, and a non-protein producing I1-3
transcript in which all introns are retained. Interestingly, hypoxia preferentially
induced the FL isoform. Moreover, HIFs, but not hypoxia per se are necessary
and sufficient to increase splicing of ADM pre-mRNA. ADM splicing reporters
confirmed that transcriptional activation by HIF or other transcription factors is
sufficient to enhance splicing. However, HIFs are more potent in enhancing
ADM pre-mRNA splicing than other transcriptional activators. Thus, ADM intron
4 This chapter is reprinted with the permission of Molecular Cancer Research.
Sena JA, Wang L, Pawlus MR, Hu CJ. HIFs Enhance the Transcriptional Activation and Splicing of Adrenomedullin. Mol Cancer Res. 2014 Feb 12. PubMed PMID: 24523299 54
retention is not a consequence of abnormal splicing, but is an important
mechanism to regulate ADM expression. These results demonstrate a novel
function of HIFs in regulating ADM expression by enhancing its pre-mRNA
splicing. Importantly, using endogenous and cloned ADM gene, further evidence
is provided for the coupling of transcription and RNA splicing. Implications: Here,
a novel function of HIFs in regulating ADM gene expression is identified by
enhancing ADM pre-mRNA splicing.
Introduction
Hypoxia is a common characteristic of many solid tumors. The hypoxic
intratumoral microenvironment stabilizes hypoxia-inducible transcription factor 1α
(HIF1α) and 2α (HIF2α) that are normally degraded by the 26S proteasome upon
pVHL-mediated ubiquitination under normoxia. Stabilized HIF1α and HIF2α proteins translocate to the nucleus and heterodimerize with a constitutive nuclear protein, the aryl hydrocarbon receptor nuclear translocator (ARNT, also called
HIF1β) to form HIF1α/ARNT (HIF1) and HIF2α/ARNT (HIF2) heterodimers.
Then, HIF1 and HIF2 bind to HIF binding sites (HBS) on HIF target gene
promoters and/or enhancers and activate genes involved in neovascularization
Adrenomedullin (ADM) is a well-established hypoxia induced gene. The
ADM gene codes for a 185 amino acid propreadrenomedullin protein that is
cleaved into a 52 amino acid AM peptide and a 20-amino acid peptide called
“proadrenomedullin N-terminal 20 peptide” or PAMP (Fig. 2.1D). AM peptide
plays important roles in tumorigenesis by inducing tumor angiogenesis, 55 enhancing tumor cell proliferation, and reducing tumor cell apoptosis (286-294).
However, PAMP appears to be less important in tumorigenesis because PAMP has no activity in tumor cell proliferation and survival, although PAMP is a stronger vasodilator and angiogenic factor than the AM peptide (295-297). Thus, hypoxia induced ADM gene expression is an important component of the hypoxia response that is crucial for tumor progression and metastasis.
Interestingly, various cancer cells cultured under normoxia were found to produce two isoforms (298). One isoform is devoid of introns (full-length, FL) and produces both PAMP and AM peptides. A second isoform, in which the third intron is retained (I3), produces only the PAMP peptide due to a premature stop codon in intron 3 (298). Moreover, the relative ratio of I3/FL was increased by hypoxia, resulting in an increased PAMP/AM peptide ratio even though both isoforms were induced by hypoxia (298). This data suggested that hypoxia favors intron 3 retention and expression of PAMP peptide. The goal of our study is to clarify if hypoxia favors PAMP generation and to determine how hypoxia regulates the ADM isoform ratio change.
Results
Hypoxia Preferentially Increases Fully-spliced ADM Transcript Levels in
Various Cell-lines
The ADM gene was reported to produce two isoforms, one isoform devoid of introns (ADM FL) and a second isoform in which the third intron is retained
(ADM I3) (298). To determine if hypoxia differentially regulates the levels of 56
these ADM transcripts, RNA prepared from normoxic and hypoxic Hep3B cells
was used for RT-PCR. ADM transcripts were detected using forward and
reverse primers located in the first exon (exon 1) and the last exon (exon 4),
respectively, of the ADM gene. Interestingly, normoxic Hep3B cells, a
hepatocellular carcinoma cell-line, expressed multiple ADM transcripts (Fig.
2.1A), a fully spliced transcript in which all introns were removed (Fig.1A, FL), a
transcript in which all three introns were included (Fig. 2.1A, I1-3), a transcript in
which intron 1 was included (Fig. 1A, I1), and two minor isoforms in which intron
2 was included (Fig. 1A, I2) and a previously reported intron 3-containing isoform
(Fig. 2.1A, I3) (298). Interestingly, hypoxia induced the levels of ADM FL but not
the other intron-containing isoforms (Fig. 2.1A). To better quantify the ADM
isoforms, qRT-PCR was used to measure the relative ratios of the ADM FL and
I1-3 transcripts since they were the dominantly expressed transcripts in Hep3B
cells. Using transcript specific qPCR primers for ADM FL (E1-E2/E3-E2, detects
both FL and I3) and ADM I1-3 (I1-E2/I2-E2, detects I1-3 only), it was found that
Hep3B cells exhibited an ADM FL/I1-3 ratio of 3.28 and 15.95 under normoxia and hypoxia (Fig. 2.1A, bar graph), suggesting that hypoxia favors ADM FL expression even though both isoforms are induced. We also found that normoxic
MDA-MB-231, Hela, and HEK293 cells expressed ADM FL, I1-3, and I1 isoforms and that hypoxia preferentially induced the levels of ADM FL in these cells (data not shown).
To assess ADM expression in relatively normal cells, HK2 cells, an HPV-
16 transformed kidney cell-line was analyzed. Using RT-PCR, HK2 cells were 57
Figure 2.1
Figure 2.1: Hypoxia Increases the Levels of Fully Spliced ADM Transcripts in Cancer and Normal Cells. RT-PCR (left) and qRT-PCR (right) analysis of ADM transcripts in normoxic or hypoxic Hep3B (A), HK2 (B), and HUVEC (C) cells. FL indicates ADM transcripts in which all introns are removed. I1 refers to ADM RNA containing intron 1. I1-3 refers to ADM RNA in which all three introns are retained. I2 or I3 refers to ADM RNA retaining intron 2 or 3. In the qRT-PCR panel, the numbers next to the sample labels represent the expression ratio of ADM FL to ADM I1-3, ± SD. (D), schematic diagram of the ADM gene including 4 exons (boxes) and three introns (lines). The start and stop codons for the proper ADM protein were indicated. The exon regions that encode the PAMP and the AM peptides are indicated by black boxes and striped boxes, respectively. E), Western blot analysis of HIF1α, HIF2α, b-actin, and ADM proteins in normoxic and hypoxic Hep3B cell.
58
found to express the FL, I1, and I1-3 transcripts with FL and I1-3 being the
primarily expressed transcripts under normoxia. Interestingly, hypoxia increased
the ADM FL, but not the I1-3 transcript levels (Fig. 2.1B). According to qRT-
PCR, hypoxia increased the levels of ADM FL and I1-3 transcript by 5.98 and
1.16 fold (Fig. 2.1B, bar graph). As a result, the ADM FL/I1-3 ratio was increased
from 11.7 under normoxia to 62.16 under hypoxia.
Similarly, HUVEC cells, a primary umbilical vein vascular endothelial cell-
line, also exhibited splicing patterns similar to those observed in Hep3B and HK2
cells under normoxia (Fig. 2.1C). Furthermore, hypoxia induced the levels of
ADM FL, but not I1-3 transcripts (Fig. 2.1C, right). Again, qRT-PCR confirmed
that the levels of ADM FL transcript was increased by 6.52 fold but the levels of
ADM I1-3 transcript was not induced (0.91 fold) (Fig. 2.1B, bar graph), thus
hypoxia enhanced the FL/I1-3 ratio from 5.17 under normoxia to 37.62 under hypoxia. Since ADM FL and I1-3 transcripts are the dominantly expressed
transcripts in the cell lines tested here; the FL and I1-3 transcripts are the focus
for the remainder of this study.
The ADM FL transcript codes for a 182-amino acid pro-pre-
adrenomedullin precursor protein that is cleaved to produce a 52-amino acid AM
peptide and a 20-amino acid PAMP peptide (Fig. 2.1D). In contrast, the ADM I3
transcript only codes for the PAMP peptide due to a premature stop codon in
intron 3, whereas the ADM I1, I2, and I1-3 isoforms do not code for either the AM
or PAMP peptides due to a premature stop codon in introns 1 and 2 (Fig. 2.1D).
After establishing that hypoxia increased the expression of the ADM FL 59
transcript, western blots for the AM peptide were performed. As expected, high
levels of HIF1α and HIF2α proteins were detected in hypoxic but not in normoxic
Hep3B cells (Fig. 2.1E). Importantly, the ADM precursor (pro, 22 kDa) and the
AM peptides (6 kDa, doublet likely due to post-translational modifications),
derived from the ADM FL transcript, were detected under normoxia.
Interestingly, hypoxia significantly increased the levels of AM peptide (Fig. 2.1E,
anti-AM). However, the levels of the ADM precursor protein were not
significantly different between normoxic and hypoxic Hep3B cells (Fig. 2.1E,
Pro), likely due to its cleavage which generates the AM and PAMP peptides.
These results indicated that hypoxia increases the levels of ADM FL transcript in
cancer and normal cell lines.
The Increased ADM FL/I1-3 Ratio is Due to RNA Splicing but Not Due to
Differential RNA Stability
Intron-retaining transcripts frequently contain premature termination codons and are targeted for nonsense-mediated decay (NMD) (299, 300). As
stated above, ADM I1-3, I1, I2, and I3 transcripts contain premature stop codons in introns 1, 2, and 3. These intron-retaining ADM transcripts are predicted to
undergo NMD when they are exported to cytoplasm, which might explain why
they exhibit lower induction under hypoxia. To test this, Hep3B cells were placed
under hypoxia for 16 hrs to increase the levels of both ADM FL and I1-3, followed
by treatment with actinomycin D to inhibit de novo transcription. Cells were then
placed back under normoxia or hypoxia for 0, 2, 4, or 8 hours and harvested for 60
RNA isolation and cDNA synthesis. Using qRT-PCR , both ADM FL and I1-3 transcripts were found to be very unstable since both transcripts were reduced by
82-97% 2 hrs post actinomycin D treatment under normoxia or hypoxia (Fig.
2.2A). In addition, the ADM FL and I1-3 mRNA levels did not change significantly from 2-8 hrs following actinomycin D treatment (Fig. 2.2A).
Interestingly, we found that the mRNA stability of ADM FL did not significantly differ between normoxia and hypoxia at 2, 4 and 8 hrs (Fig. 2.2A). However, hypoxia reduced the stability of ADM I1-3. For instance, ADM I1-3 was reduced by 82, 83, and 84% at 2, 4, and 8 hrs, respectively under normoxia but was reduced by 91-93% under hypoxia at 2, 4, and 8 hrs (Fig. 2.2A). It is not clear why ADM transcripts degrade rapidly although these results are consistent with a previous report, which demonstrates similar trends in ADM mRNA stability (301).
Overall, ADM I1-3 was more stable than ADM FL both under normoxia and hypoxia; therefore, this data suggests that the hypoxia induced increase of the
ADM FL/I1-3 ratio is not likely due to preferential de-stabilization of ADM I1-3 under hypoxia.
ADM I1-3 Transcripts are Primarily Located in the Nucleus
Intron-retaining transcripts frequently contain premature termination codons and are either targeted for nonsense mediated decay or their presence is restricted to the nucleus (299, 300). Next, the sub-cellular localization of ADM FL and I1-3 transcripts were determined in normoxia or hypoxia Hep3B cells by examining the levels of ADM FL and I1-3 transcripts in the nucleus or cytoplasm. 61
First, we measured the expression of beta actin RNA (B-ACT) as a positive control for mRNA splicing and nuclear export, and as a negative control for hypoxia induction. As expected, B-ACT expression was not induced by hypoxia and a majority of the RNA was localized in the cytoplasm, with cytoplasmic to nuclear ratios of 3.05 and 3.87 under normoxia and hypoxia respectively (Fig.
2.2B). As expected of a mature transcript, ADM FL was found both in the nucleus and in the cytoplasmic fractions (Fig. 2.2C). However, more ADM FL
transcript was found in the cytoplasmic fraction, with the cytoplasmic to nuclear
ratios of ADM FL being 1.75 and 2.7 under normoxia and hypoxia respectively.
In contrast, a majority of the ADM I1-3 transcript accumulated in the nucleus
compared to the cytoplasm, with the cytoplasmic to nuclear ratios being 0.029
and 0.15 under normoxia and hypoxia respectively. This data suggested that
ADM I1-3 is an unspliced ADM transcript that is restricted primarily to the nucleus
(Fig. 2.2D). Additionally, we measured the total levels of ADM transcript (using
primers located in exon 4) and found that more ADM transcripts were localized to
the cytoplasm, with the cytoplasmic to nuclear ratio of total ADM being 1.33
under normoxia and 1.9 under hypoxia (Fig. 2.2E). In addition, we did not detect
a signal by qPCR in our nuclear or cytoplasmic fractions in which we excluded
reverse transcriptase from our cDNA synthesis (Fig. 2.2B-E, no reverse
transcriptase), suggesting that all of our transcripts including ADM I1-3 are in fact
mRNAs and not DNA contamination.
62
Figure 2.2
Figure 2.2: ADM FL and I1-3 Transcript Stability Does Not Account for the Hypoxia-mediated Increase in the ADM FL/I1-3 Ratio, and ADM FL and I1-3 Transcripts are Differentially Localized in Hep3B cells. A) qRT-PCR analysis ADM FL and I1-3 transcripts in normoxic and hypoxic Hep3B cells treated with actinomycin D for 2 to 8 hours. qRT-PCR analysis of b-actin (B), ADM FL (C), ADM I1-3 (D), and ADM total (E) transcripts in nuclear and cytoplasmic fractions from normoxic and hypoxic Hep3B cells.
63
HIF Activity is Required for Increased Splicing of ADM Pre-mRNA
ADM is a HIF regulated gene; therefore we wanted to determine if HIF
activity was necessary for the hypoxia induced ADM FL/I1-3 ratio increase. To test this, the levels of ARNT, HIF1α, or HIF2α mRNAs in Hep3B cells were reduced by 75-87% using siRNAs (Fig. 2.3A). ARNT knockdown inhibited hypoxic induction of HIF1α target genes, LDHA and PGK1, and HIF2α target
genes, EPO and PAI1 (Fig. 2.3B). On the other hand, HIF1α knockdown only
inhibited hypoxic induction of HIF1α target genes, whereas HIF2α knockdown
only inhibited hypoxic induction of HIF2α target genes (Fig. 2.3B). Hypoxia
preferentially increased the levels of the ADM FL transcript in Hep3B/control and
HIF2α siRNA cells (Fig. 2.3C) while knockdown of ARNT and HIF1α dramatically
inhibited the hypoxic induction of ADM FL mRNA (Fig. 2.3C). Using qRT-PCR ,
we found that hypoxia induced the expression of both the ADM FL and I1-3
transcripts but favored the induction of ADM FL transcript in Hep3B/control
siRNA cells (Fig. 2.3D). Interestingly, ARNT knockdown reduced the hypoxic
induction of both the ADM FL and I1-3 transcripts and also reduced the ADM
FL/I1-3 ratio to 7.4 versus 11.4 observed in hypoxic Hep3B/control siRNA cells.
In addition, HIF1α and HIF2α knockdown also decreased the hypoxia induction
of ADM FL and I1-3 transcripts; however, individual knockdown of either HIF1 or
HIF2 did not significantly alter the ADM FL/I1-3 ratio. For instance, HIF1
knockdown reduced the ADM FL/I1-3 ratio to 9.9 whereas HIF2 knockdown
increased the ADM FL/I1-3 ratio to 12.9. This is likely because ADM is a
64
Figure 2.3
Figure 2.3: HIF Activity is Required for Hypoxia-induced Splicing of ADM Pre-mRNA. A) qRT-PCR analysis of ARNT, HIF1α, and HIF2α mRNA levels in normoxic and hypoxic Hep3B cells targeted with control, ARNT, HIF1α, or HIF2α siRNAs. B) qRT-PCR analysis of the levels of HIF-1 target genes, LDHA and PGK1 and HIF2 target genes, EPO and PAI1, in normoxic and hypoxic Hep3B cells targeted with control, ARNT, HIF1α, or HIF2α siRNAs. C) RT-PCR analysis of ADM transcripts in normoxic and hypoxic Hep3B cells targeted with control, ARNT, HIF1α, or HIF2α siRNAs. D) qRT-PCR analysis of ADM FL, I1-3, and total transcripts in normoxic and hypoxic Hep3B cells targeted with control, ARNT, HIF1α, or HIF2α siRNAs.
65
common HIF1 and HIF2 target gene. This data suggested that HIF activity, but not hypoxia per se, is necessary for hypoxia-induced pre-mRNA splicing of ADM.
HIF Activity is Sufficient for Increased Splicing of ADM Pre-mRNA
Next, we determined if HIF activity is sufficient for increased splicing of
ADM pre-mRNA. To test this, Hep3B cells were transduced with lentiviruses expressing normoxia active, flag-tagged HIF1α, HIF2α, or GFP. Western blot analysis using an anti-flag antibody detected expression of HIF1α, HIF2α, and
GFP proteins in virus infected Hep3B cells under normoxia (Fig. 2.4A). In addition, HIF1α and HIF2α proteins were functional under normoxia since HIF1α target genes, LDHA and PGK1, and HIF2α target genes, EPO and PAI1, were induced in HIF1α or HIF2α transduced cells compared to the GFP transduced cells (Fig. 2.4B). Importantly, both HIF1α and HIF2α increased the levels of ADM
FL, but not I1-3 as determined by RT-PCR (Fig. 2.4C). qRT-PCR also determined that both HIF1α and HIF2α induced the expression of ADM FL 4.8 and 6.2 fold, respectively, and also induced the expression of ADM I1-3 2.3 and
2.35 fold, respectively (Fig. 2.4C left). Thus, HIF1α or HIF2α transduction increased the ADM FL/I1-3 ratio to 19.36 or 25.19 versus 4.19 for the GFP transduced cells.
To further validate that HIFs are sufficient to promote splicing of ADM pre- mRNA, we used RCC4 cells, a renal cell carcinoma cell-line that expresses constitutively active HIF1α and HIF2α protein even under normoxia due to mutation of the VHL gene. In addition, we used RCC4T cells in which functional 66 pVHL is re-introduced into RCC4 cells and therefore HIF proteins are only active under hypoxia (302). RCC4T cells expressed both ADM FL and I1-3 transcripts
(Fig. 2.4D); moreover, hypoxia preferentially induced the levels of the ADM FL transcript (Fig. 2.4D). However, RCC4 cells mainly expressed the ADM FL transcript under both normoxia and hypoxia (Fig. 2.4D). Using qRT-PCR, hypoxia was found to induce the expression of ADM FL and I1-3, 10 and 1.4 fold above normoxic RCC4T cells respectively and also increased the ADM FL/I1-3 ratio to 54.16 versus 8.72 in normoxic RCC4T cells (Fig. 2.4E). However, hypoxia only weakly increased the ADM FL/I1-3 ratio in RCC4 cells since ADM
FL was already favored even under normoxia. For instance, the ADM FL/I1-3 ratio was 53.41 under normoxia and 67.11 under hypoxia. These findings supported the idea that HIFs are sufficient to increase splicing of ADM pre- mRNA.
ADM Splicing Reporters Recapitulate Splicing Changes Observed for the
Endogenous ADM Gene
After determining the relationship between ADM gene activation and increased intron-removal using the endogenous ADM gene, we wanted to see if
HIF-mediated transcription activation of an ADM splicing reporter would result in increased ADM splicing. To test this idea, the full-length ADM gene (exons1-4 including introns) was cloned and placed downstream of the CA9 promoter, a
HIF1 target gene promoter, or the PAI1 promoter, a HIF2 target gene promoter,or
67
Figure 2.4
Figure 2.4: HIF Activity is Sufficient to Promote Splicing of ADM Pre- mRNA. A) Western blot analysis of Flag-tagged HIF1α, HIF2α, and GFP proteins in Hep3B cells transduced with lentivirus expressing normoxia active Flag-tagged HIF1αTM, or normoxia active HIF2αTM or GFP proteins under normoxia. B) qRT-PCR analysis of HIF-1 target genes, LDHA andPGK1, and HIF2 target genes, EPO and PAI1in the above described cells. C) RT-PCR (left) and qRT-PCR (right) analysis of ADM FL and I1-3 transcripts in the above described cells. The ADM FL/I1-3 ratio is indicated next to the graph legend. D and E, RT-PCR (D) and qRT-PCR (E) analysis of ADM transcripts in normoxic and hypoxic RCC4 or RCC4T cells with ADM FL/I1-3 ratio indicated next to the graph legend. * Indicates non-specific band.
68 an artificial hypoxia/HIF induced promoter containing two hypoxia-responsive- elements (2HRE) and the SV40 minimal promoter (Fig. 2.5A). Additionally, splicing reporters in which the HREs were mutated (mHRE) or deleted (ΔHRE) were generated. These ADM splicing reporters were then co-transfected into
Hep3B cells with empty vector expressing a His tag (His) or with normoxia active
HIF1α or HIF2α expression plasmids. All of the splicing reporters expressed two major transcripts, a transcript in which introns 1-3 were removed (intron skipping or IS, to distinguish from the endogenous FL transcript) and a second transcript in which introns 1-3 were retained (intron retaining, IR) (Fig. 2.5B-C). The ADM
IS and IR transcripts corresponded to endogenous ADM FL and ADM I1-3 respectively, demonstrating that the splicing reporters recapitulated the expression patterns of the endogenous ADM gene.
Using reporter specific primers and RT-PCR, we found that co-transfected
HIF1α activated the CA9P/ADM reporter and increased the levels of the ADM IS transcript but not the ADM IR transcript (Fig. 2.5B). Co-transfected HIF2α also increased the levels of the ADM IS transcript although to a lesser extent than
HIF1 (Fig. 2.5B). In contrast, the CA9Pm2HRE/ADM reporter was not activated by HIF1 or HIF2 nor were the ADM IS and IR transcripts induced (Fig. 2.5B). qRT-PCR confirmed that HIF1α induced the expression of ADM IS and IR by 16 and 2.6 fold respectively. Similarly, HIF2α induced the expression of the ADM IS and IR transcripts by 4 and 1.7 fold respectively. Moreover, HIF1α or HIF2α increased the ADM IS/IR ratio to 4.22 or 2.23 respectively versus 0.93 for the empty vector (His) (Fig. 2.5B, bar graph). In contrast, the CA9Pm2HRE/ADM 69 reporter was not induced by HIF1α or HIF2α, nor did HIF1α or HIF2α alter the
ADM IS/IR ratio.
Next, similar experiments were performed on ADM splicing reporters driven by the PAI1 promoter (PAI1P/ADM and PAI1PmHRE/ADM). Using reporter specific primers and RT-PCR, we found that HIF1α activated the
PAI1P/ADM reporter and increased the levels of the ADM IS, but not IR transcript
(Fig. 2.5C top). Similarly, HIF2α also increased the levels of ADM IS but not IR
(Fig. 2.5C top). In contrast, the PAI1PmHRE/ADM reporter was not activated by
HIF1α or HIF2α and nor did HIF change the levels of ADM IS and IR transcripts
(Fig. 2.5C top). Using qRT-PCR, we found that HIF1α induced the levels of ADM
IS and IR by 2.8 and 1.3 fold respectively (Fig. 2.5C, middle) while HIF2α increased the levels of the ADM IS and IR transcripts by 18 and 3.3 fold respectively. As a result, HIF1α or HIF2α increased the ADM IS/IR ratio to 9.85 or 19.73 respectively versus 3.31 for the His vector (Fig. 2.5C, middle). In contrast, the PAI1PmHRE/ADM reporter was not activated by HIF1α or HIF2α, nor did HIF1α or HIF2α significantly alter the ADM IS/IR expression ratio (Fig.
2.5C, middle left).
Finally, we utilized ADM splicing reporters driven by the SV40 minimal promoter in which two HREs from the PAI1 promoter were added to produce an artificial HIF/hypoxia responsive promoter (2HRE/ADM and ΔHRE/ADM). Using reporter specific primers and RT-PCR, we found that both HIF1α and HIF2α activated the 2HRE/ADM reporter and increased the expression of the ADM IS 70
Figure 2.5
Figure 2.5: HIFs Regulate Splicing of ADM Splicing Reporters. A) Diagram of ADM splicing reporters. Arrows represent primers used for RT-PCR and qRT- PCR. Forward primer in the promoter region (CA9P, PAI1P, or 2HRE) and ADM RT-PCR primer (in E4) were used in RT-PCR. CA9P, PAI1P, or 2HRE primers in conjunction with ADMIS, ADMIR, or ADM total were used in qRT-PCR. B) RT- PCR (top) and qRT-PCR (bottom) detection of ADM transcripts in normoxic Hep3B cells cotransfected with CA9P/ADM or CA9Pm2HRE/ADM and with His tagged empty vector control (His), HIF1αTM or HIF2αTM expression plasmids. RT-PCR (top) and qRT-PCR (middle and bottom) analysis of ADM transcripts in normoxic Hep3B cells transfected with PAI1P/ADM, PAI1PmHRE/ADM, 2HRE/ADM, or DHRE/ADM splicing reporters and empty vector (His), HIF1αTM, or HIF2αTM expression plasmids.
71
transcript (Fig. 2.5C, top left). In contrast, the ΔHRE/ADM reporter was not
activated by HIF1α or HIF2α and therefore the ADM IS and IR transcripts were
not induced (Fig. 2.5C, top left). Again, RT-qPCR confirmed that HIF1α induced the expression of ADM IS 4 fold but reduced the expression of ADM IR 1.4 fold
(Fig. 2.5C, bottom right). Moreover, HIF2α induced the expression of the ADM IS
and IR transcripts 17 and 1.2 fold respectively (Fig. 2.5C, bottom right). HIF1α
and HIF2α increased the ADM IS/IR ratio to 4.55 and 11.28 respectively versus
0.5 for the His vector (Fig. 2.5C, bottom right). In contrast, the ΔHRE/ADM
reporter was not activated by HIF1α or HIF2α, nor did HIF1α or HIF2α significantly alter the ADM IS/IR ratio (Fig. 2.5C, bottom left). Taken together,
this data showed that splicing of the ADM splicing reporters can recapitulate the
HIF dependent splicing changes observed for the endogenous ADM gene. This data further supported our conclusion that HIF activity is sufficient to promote splicing of ADM pre-mRNA. In addition, this data suggested that transcription activation of ADM is necessary to promote splicing of ADM pre-mRNA since mutation or deletion of HREs in the reporters prevented transcription activation and also the increase in the ADM IS/IR ratio.
The Transactivation Domain of HIFα Protein is Not Required for Increased
RNA Splicing of the ADM Splicing Reporter
Some transcription factors have dual roles in RNA splicing and gene
transcription (303-305). To determine if the activation domain of HIFα protein is
important for ADM pre-mRNA splicing, we constructed hybrid proteins that 72 contained the HIF1α DNA binding domain and ARNT dimerization domain (DBD), fused to the transactivation domains from the E2F1 or VP16 transcription factors
(Fig. 2.6A). To facilitate the detection of these hybrid proteins, 2xFlag tag was added at the C-terminuses of these constructs (Fig.6A). As determined by anti-
Flag western blot, the HIF1αDBD, HIF1αDBD/E2F1TAD, and HIF1αDBD/
VP16TAD expression plasmids expressed higher levels of proteins than the normoxia active HIF1α and HIF2α expression plasmids even though the same amount of plasmid DNAs were transfected (Fig. 2.6B). However, the CA9P/ADM reporter was activated similarly by HIF1α (2.2 fold) and HIF1αDBD/E2F1 (2.1 fold) while the HIF2α and HIF1αDBD/VP16 weakly activated the CA9P/ADM reporter (Fig. 2.6C, ADM total). Interestingly, HIF1α was more potent in increasing the ADM IS/IR ratio than the HIF1αDBD/E2F1, increasing the ADM
IS/IR ratio to 4.37 whereas HIF1αDBD/E2F1 increased the ADM IS/IR ratio to
2.47 versus 0.94 for the His control, although the constructs were able to activate the reporter similarly (Fig. 2.6C). In addition, HIF2α and HIF1αDBD/VP16 only increased the ADM IS/IR ratio slightly (Fig. 2.6C).
In addition, the above-described fusion constructs were able to activate the PAI1P/ADM reporter in Hep3B cells (Fig. 2.6D, ADM total). HIF1α activated the reporter 1.96 fold whereas HIF2α activated the reporter more than any other expression plasmid, 8.84 fold. Interestingly, even the HIF1αDBD slightly activated this reporter, 1.67 fold. In addition, the HIF1αDBD/E2F1 and the
HIF1αDBD/VP16 activated the reporter 2.17 and 2.35 fold respectively (Fig.
2.6D, ADM total). In addition, HIF1α and HIF2α were able to increase the ADM 73
Figure 2.6
Figure 2.6: Activation of the ADM Splicing Reporter is Sufficient to Increase Splicing of the ADM Reporter. A) schematic of HIF1α TM, HIF2α TM, and HIF-1a DNA binding and PAS domains (HIF1α DBD) fused with the activation domain of VP16 or E2F1. All of these constructs are tagged with a 2X Flag epitope at their C-terminus. B) Western blot analysis of Flag-tagged HIF1αTM, HIF2αTM, HIF1αDBD, HIF1αDBD/VP16TAD, and HIF1αDBD/E2F1TAD expression constructs. qRT-PCR analysis of ADM transcript levels in normoxic Hep3B cells cotransfected with the CA9P/ADM reporter (C) or the PAI1P/ADM reporter (D) and the above described expression constructs. E) qRT-PCR analysis of HIF-2aand USF2 mRNA levels (top) and splicing reporter-specific ADM transcripts (bottom) in Hep3B cells cotransfected with the PAI1P/ADM reporter and control (His), HIF2αTM or USF2 expression plasmids.
74
IS/IR ratio to 15.09 and 23.4 versus 7.5 for the His control (Fig. 2.6D). Although
HIF1αDBD activated the reporter slightly, it did not significantly change the ADM
IS/IR splicing ratio (7.7) compared to the His vector (7.5). On the other hand,
HIF1αDBD/E2F1 and HIF1αDBD/VP16 increased the ADM IS/IR splicing ratio to
13.3 and 9.85 respectively versus 7.5 for the His control (Fig. 2.6D).
Furthermore, the fusion constructs were also able to activate and promote splicing of endogenous ADM even under normoxia (data not shown). In summary, this data suggested that other transcription factors are also able to promote splicing of the ADM pre-mRNA and the transactivation domain of the
HIFα protein is replaceable for increased RNA splicing of the ADM splicing reporter.
Activation of Endogenous HIF Target Genes is Not Absolutely Required for
Increased Splicing of the ADM Splicing Reporter
As stated above, transcription activation by HIF1 or HIFDBD hybrid constructs is sufficient to promote splicing of the ADM pre-mRNA. However, both HIF and the fusion constructs can activate endogenous HIF target genes; therefore it is possible that HIF target genes may be responsible for promoting splicing of ADM pre-mRNA. To rule out or confirm this possibility, we utilized an
ADM splicing reporter driven by the PAI1 promoter (PAI1P/ADM), which contains an HRE and also binding sites for the upstream stimulatory factor 2 (USF2)
(306). In contrast to HIFs and the HIF fusion constructs, USF2 does not activate endogenous HIF target genes. To test this, normoxic Hep3B cells were co- 75
transfected with the PAI1/ADM splicing reporter and either HIF2α or USF2
expression plasmid. As determined by qRT-PCR using primers that detect endogenous and transfected HIF2α or USF2 RNA, we found that HIF2α transfection increased HIF2α mRNAs levels 18.31 fold and USF2 transfection increased USF2 levels 105.94 fold above the His vector (Fig. 2.6E, top). As
determined by qRT-PCR using reporter specific primers that detect total reporter
specific ADM transcripts, we found that HIF2α activated the PAI1P/ADM reporter
3 fold whereas USF2 activated the reporter 2 fold compared to the His vector
(Fig. 2.6E, bottom, ADM total). Also, HIF2α induced the expression of ADM IS
and IR 11.4 and 1.7 fold respectively while USF2 induced the expression of ADM
IS and IR 5.2 and 1.9 fold respectively. Thus, HIF2α or USF2 increased the
ADM IS/IR ratio to 9.13 or 3.71 versus 1.42 for the His control (Fig. 2.6E, lower
panel).
To further validate that transcription activation of ADM was sufficient to
promote splicing of ADM pre-mRNA, ADM splicing reporter was placed under the
control of a promoter containing 5 copies of the Gal4 DNA binding element (Fig.
2.7A). In addition, fusion constructs containing the Gal4 DNA binding domain
fused to the transactivation domains of normoxia active HIF1α, normoxia active
HIF2α, VP16, or E2F1 were generated (Fig. 2.7A). Therefore, the Gal4 DBD
fusion constructs are expected to activate the G5P/ADM splicing reporter but not
endogenous HIF target genes. Next, we co-transfected Hep3B cells with the
G5P/ADM splicing reporter and the Gal4 DBD fusion constructs. An anti-gal4 western blot indicated that the fusion constructs expressed proteins as expected 76
(Fig. 2.7B). Using qRT-PCR to assess activation of the G5P/ADM reporter, we found that the Gal4 DBD did not significantly activate the reporter above the His control (Fig. 2.7C, ADM total). In contrast, Gal4/HIF1αTAD, HIF2αTAD,
VP16TAD, and E2F1TAD were able to activate the reporter 16.4, 4.05, 71.4 and
66.75 fold respectively (Fig. 2.7C, ADM total). Moreover, the Gal4 DBD expression plasmid did not alter the expression of ADM IS and IR transcripts and did not change the ADM IS/IR ratio (Fig. 2.7C). However, the Gal4/HIF1αTAD expression plasmid increased the expression of ADM IS and IR by 80.95 and 3 fold and also increased the ADM IS/IR ratio to 45.98 from 2.26 for the His vector
(Fig. 2.7C). Similarly, the Gal4/HIF2αTAD expression plasmid increased the expression of ADM IS and IR by 12.3 and 2.35 fold and increased the ADM IS/IR expression ratio to 8.58 (Fig. 2.7C). Also, the Gal4/VP16TAD increased the expression of ADM IS and IR by 284.8 and 14.1 fold and thus increased the ADM
IS/IR ratio to 39.4 (Fig. 2.7C). Moreover, the Gal4/E2F1TAD expression plasmid increased the expression of ADM IS and IR by 255.13 and 8.03 fold and increased the ADM IS/IR ratio to 52.12 (Fig. 2.7C). Taken together, this data further confirmed that transcriptional activation of ADM is sufficient to enhance intron removal in the absence of endogenous HIF target gene activation.
Discussion
Martinez et al reported that the ADM gene expresses FL and I3-containing isoforms in normoxic cell lines (298). By analyzing ADM RNA transcripts using
RT-PCR, we found that normoxic cancer cells express fully-spliced (FL) and 77
Figure 2.7
Figure 2.7: Activation of Endogenous HIF Target Genes is Not Required for Increased Splicing of an ADM Splicing Reporter. A) Schematic of the G5P/ADM splicing reporter and hybrid constructs containing the Gal4 DBD (Gal4) fussed to the transactivation domain of HIF1aTM, HIF2aTM, VP16, or E2F1. B) Western blot analysis of the Gal4 DBD expression constructs using antibody specific for the Gal4 DBD. C) qRT-PCR analysis of ADM transcript levels in normoxic Hep3B cells cotransfected with the G5P/ADM reporter and the above described expression constructs.
78
several intron-retaining RNA transcripts including I1, I2, I3, and I1-3 isoforms
(Fig. 2.1). Furthermore, the FL and I1-3 isoforms, but not the I3 isoform, are the predominate isoforms expressed in normoxic cells (Figs 2.1 and 2.4). To our knowledge, this study is the first to report that the ADM gene expresses I1-3, I1, and I2 isoforms in normoxic cells. Importantly, in addition to cancer cells, we found that intron-retaining ADM isoforms are expressed in normal human tissues
(data not shown) and normal HUVEC cells, suggesting that expression of intron- retaining isoforms is not a consequence of abnormal splicing, but a normal control mechanism of gene expression.
ADM is a well-established HIF target gene (302, 307). However, it was
reported that the I3 but not the FL isoform was preferentially induced by hypoxia
(298). Because the ADM FL RNA codes for a more functionally important AM
peptide than the ADM I3 RNA, it is puzzling that I3 is preferentially induced by
hypoxia. Thus, we investigated ADM isoform expression in hypoxic cells.
Interestingly, we found that hypoxia preferentially induces the levels of ADM FL
over intron-containing ADM transcripts. It is not clear if this discrepancy is due to
different cell lines used or other differences.
To address the molecular mechanism concerning preferential induction of
ADM FL RNA under hypoxia, we first tested the hypothesis that ADM I1-3
transcripts are selectively degraded during hypoxia. However, we found that
ADM FL is even less stable than the ADM I1-3 isoform, suggesting that
differential transcript stability does not account for the hypoxia induced ADM
FL/I1-3 ratio increase. Furthermore, we found that the both ADM I1-3 and FL 79
transcripts are very unstable both under normoxia and hypoxia (Fig. 2.2A), a result consistent with a previous report that demonstrates similar patterns of
ADM mRNA degradation (301). Since there are significant amounts of ADM FL and I1-3 in the nucleus, one possibility is that the ADM FL and ADM I1-3 transcripts undergo Traf4/Trf5-Air1/Air2-Mtr4 (TRAMP) mediated polyadenylation and RNA degradation by the nuclear exosome complex (308-310). In the future, it will be interesting to address why ADM transcripts are so unstable.
After establishing that increased splicing of ADM pre-mRNA is responsible for the increased ADM FL/I1-3 ratio under hypoxia, using the endogenous ADM gene (Figs. 2.3 and 2.4) as well as ADM splicing reporters (Fig. 2.5), we determined that HIF activity, but not hypoxia per se, is necessary and sufficient to regulate hypoxia induced splicing of ADM pre-mRNA. Furthermore, we determined that transcription activation of ADM is necessary and sufficient to regulate splicing of ADM pre-mRNA independent of endogenous HIF target gene activation (Fig 2.6E and 2.7). Thus, our data demonstrated that ADM gene transcription and pre-mRNA splicing is coupled. Although it is well established that transcription and splicing are functionally coupled (257, 311, 312), these conclusions are mainly derived from studies using chimeric splicing reporters in which an alternatively spliced exon is placed between two constitutive exons of a different gene. To our knowledge, this is the first report that utilizes endogenous and cloned genes in the same study as models to demonstrate functional coupling of transcription and RNA splicing. 80
Additionally, our data indicated that transcription activation strength
contributes to splicing of ADM pre-mRNA. For instance, HUVEC cells exhibit an
ADM FL/I1-3 splicing ratio of 5.17 (Fig. 2.1C) under normoxia. In contrast, Hep3B
cells exhibit an ADM FL/I1-3 ratio of 3.28 under normoxia (Fig. 2.1A).
Interestingly, basal level expression of the ADM FL and the ADM I1-3 isoforms are 2.62 and 2.89 fold higher in HUVECs than in Hep3B cells (data not shown).
In addition, the HIF2α virus is a stronger inducer of ADM gene transcription and increases the ADM FL/I1-3 ratio more than the HIF1α virus (Fig. 2.4C). Further
support is garnered from our ADM splicing reporter studies. For instance the
ADM IS/IR ratio is 0.5 under basal conditions (His) for the 2HRE/ADM reporter
while the ADM IS/IR ratio is 3.31 for the PAI1P/ADM reporter under basal
conditions (His) (Fig. 2.2C). Interestingly, the basal expression of PAI1P/ADM
reporter is 2.8 fold higher than the basal expression of the 2HRE/ADM reporter
(data not shown). Moreover, higher ADM IS/IR expression ratios correlate with
higher levels of reporter activation. For instance, if HIF1α activates the reporter
more than HIF2α, HIF1α also increases the ADM IS/IR ratio more than HIF2α
(Fig. 2.5B) or if HIF2α activates the reporter more than HIF1α, HIF2α also
increases the ADM IS/IR ratio more than HIF1α (Fig. 2.5C). In summary, these
findings indicate that transcription activation strength regulates pre-mRNA
splicing efficiency of ADM independent of promoter differences. These findings
are novel and in contrast to what have been observed for alternatively spliced
genes. For instance, previous studies suggest that differences in promoter
structure lead to differences in alternative splicing of particular transcripts, 81 however transcription activation strength is not responsible for this effect (257,
313, 314). It is not clear why our results differ from previous studies. We speculate that these differences may be due to the fact that we studied intron- removal whereas previous studies assessed alternative exon inclusion.
Moreover, while previous studies compared different transcription factors in regulating alternative splicing we compared basal level expression of ADM transcripts in different cell lines. In fact, if we compared HIF with other transcription factors in regulating ADM splicing (Fig. 2.6C-D and Fig. 2.7C), our data is in agreement with previous studies and suggests that activation strength does not determine the efficiency of ADM splicing (see below).
Although transcription activation and activation strength influence ADM pre-mRNA splicing, HIFs are more potent than other transcriptional activators in promoting intron removal under condition in which both endogenous HIF target genes are activated or not activated. For example, HIF1α is more potent than
HIF1αDBD/E2F1 in inducing ADM IS from the CA9P/ADM splicing reporter despite similar activation of the reporter by the two activators (Fig. 2.6C).
Furthermore, Gal4/HIF1αTAD is also a stronger inducer of ADM IS than
Gal4/VP16TAD or Gal4/E2F1TAD although Gal4/VP16TAD or E2F1TAD activate the G5/ADM reporter more than Gal4/HIF1αTAD (Fig. 2.7C). It is not clear why
HIFs are more efficient than other transcription factors in increasing ADM pre- mRNA splicing. Based on the recruitment model of RNA splicing (315-317), we speculate that HIF may have stronger activity than other transcription factors in recruiting splicing factors to HIF target gene promoters. So far, HIFs have not 82 been reported to interact with splicing factors such as serine-arginine rich (SR) proteins or heterogeneous nuclear ribonucleoproteins (hnRNPs) or with components of the spliceosome. However, HIFs have been shown to recruit co- transcription factors that can promote splicing. For example, studies indicated that BRM, a component of the SWI/SNF chromatin remodeling complex, can act as a co-activator for HIF target genes (318). Moreover, a separate study showed that BRM could regulate alternative splicing by slowing the rate of RNA Pol II elongation (259). Therefore, it is possible that HIFs are more efficient than other transcription factors in recruiting BRM to the ADM promoter and regulate splicing by altering the elongation rate of RNA Pol II (315-317, 319).
In summary, we demonstrate for the first time that HIFs enhance pre- mRNA splicing of ADM in addition to regulating ADM transcription. In the future, it will be interesting to assess whether other HIF target genes also express intron- retaining or exon-skipping isoforms under normoxia and whether HIF promotes
RNA splicing of additional HIF target genes. In addition, it will be interesting to know if activation of other HIF target genes will lead to increased RNA splicing independent of endogenous HIF target gene activation.
Acknowledgments
This work was supported by grants from the National Cancer Institute
(RO1CA134687, Hu) and Cancer League of Colorado (Hu). Johnny Sena was supported by “Research Supplemental to Promote Diversity” (NCI) from June 1, 83
2010 to May 30, 2012. Dr. David Bentley generously provided the
Gal4/E2F1TAD plasmid used in this study.
84
CHAPTER III
HYPOXIA REGULATES ALTERNATIVE SPLICING OF HIF AND NON-HIF
TARGET GENES 5
Abstract
Hypoxia is a common characteristic of many solid tumors. The hypoxic
microenvironment stabilizes hypoxia-inducible transcription factor 1α (HIF1α) and
2α (HIF2α) to activate gene transcription, which promote tumor cell survival. 95%
of human genes are alternatively spliced, producing RNA isoforms that code
functionally distinct proteins. Thus, effective hypoxia response requires
regulation of gene transcription as well as RNA splicing. However, it is unclear
how hypoxia regulates RNA splicing. This study used exon arrays to determine
the genome-wide effects of hypoxia on alternative splicing in Hep3B cells. The
results indicated that hypoxia regulates RNA splicing of a significant number of
genes including genes that are hypoxia-induced, hypoxia-reduced, and not
changed by hypoxia. Moreover, hypoxia promoted exon inclusion for hypoxia- inducible genes, but reduced exon inclusion for hypoxia-reduced genes. RT-
PCR, qRT-PCR , and western blot confirmed hypoxia-mediated alternative splicing of several HIF target genes. Mechanistically, alternative splicing of
carbonic anhydrase 9, Angiopoietin-like 4, and pyruvate dehydrogenase kinase 1
(PDK1) is dependent on HIF activity since inhibition of HIF activity reduces gene
5 This chapter contains material from a manuscript under review at the journal of Molecular Cancer Research
Sena JA, Wang L, Heasley LE, Hu CJ. Hypoxia regulates alternative splicing of HIF and non-HIF target genes. Mol Cancer Res. Under Review 2014. 85 expression and alternative splicing while normoxia-active HIF induces gene expression and alternative splicing. Importantly, using PDK1 minigenes, we found that increased expression of endogenous HIF target gene(s) contributes to
RNA splicing of the PDK1 minigene. Thus, this study demonstrates a novel function of HIF in the regulation of RNA splicing of HIF-target and non-target genes.
Introduction
Hypoxia is a common characteristic of many solid tumors. The hypoxic microenvironment stabilizes hypoxia-inducible transcription factor 1α (HIF1α) and
2α (HIF2α) that are normally degraded under normoxia. The stabilized HIF1α and HIF2α proteins translocate to the nucleus, where they dimerize with aryl hydrocarbon receptor nuclear translocator (ARNT) to form HIF1α /ARNT (HIF1) and HIF2 α/ARNT (HIF2) complexes. HIF1 and HIF2 activate gene expression to promote tumor progression and metastasis (320-323). As HIF is thought to promote tumor cell survival by increasing gene transcription, almost all of the previous studies including ours investigated the hypoxia inducible genes using traditional microarrays (302, 324-327), in which probe sets are typically located in the 3’ UTR of the transcripts.
Recently, it was found that 95% of human genes are alternatively spliced and RNA isoforms produced from a single gene code proteins with distinct or opposing functions (328). Thus, it becomes clear that assessing isoforms expression will be more informative than quantifying total transcripts to 86
appreciate the functional consequence of gene regulation.
Exon arrays contain probe sets interrogating every exon of the
transcriptome, thereby permitting the distinguishing of different isoforms of a
gene as well as measure gene expression levels. So far, two studies have
examined the role of hypoxia in regulating RNA splicing using exon arrays (281,
329). However, both studies used endothelia cells, a normal cell type. So far,
hypoxia-mediated RNA splicing changes have not been measured in cancer
cells. Additionally, these previous studies in endothelial cells focused on non-HIF
target genes and reported that hypoxia, like other stresses, inhibits RNA splicing
of non-HIF target genes by promoting exon skipping and/or intron-inclusion (281,
329). Thus, the role of hypoxia in regulating RNA splicing of HIF-induced genes is still unknown. In this study, we used exon arrays to assess global hypoxia- mediated changes in RNA splicing of HIF and non-HIF target genes in Hep3B cells, a cancer cell line. These findings have important implications for our understanding of how HIFs promote tumorigenesis in response to hypoxia.
Results
Genome-wide Exon Array Analysis Determines that Hypoxia Alters RNA
Splicing of HIF and Non-HIF Target Genes in Hep3B Cells
To assess the global effects of hypoxia on RNA splicing, Affymetrix
GeneChip Human Exon 1.0 ST arrays were used to probe total RNAs isolated from normoxic or hypoxic Hep3B cells. Hep3B cells were chosen because of 87 their high expression levels of HIFα proteins and high hypoxic induction of HIF target genes (48, 306, 330). Altanalyze (FIRMA) and EasyExon software predicted that hypoxia regulated alternative splicing (AS) of about 2005 (Fig.
3.1A, yellow and grey) and 3919 genes respectively (Supplemental File 1, AS by
Firma and AS by EasyExon). Venn diagrams using Venny indicated that 1012 genes were predicated to undergo AS by both programs while an additional 957 and 2634 genes were predicated to undergo AS by FIRMA and EasyExon respectively. FIRMA predictions of AS seemed to be more stringent than
EasyExon, therefore we chose FIRMA to characterize hypoxia mediated splicing events. FIRMA analysis indicated that hypoxia produced 3059 AS events in 2005 genes, averaging 1.53 splicing events per gene (Fig. 3.1A, All Genes). As expected, alternative cassette-exons were the most abundant type of splicing events, making up 51% of all splicing events. Other AS events included alternative 5’ splice-sites (16%), alternative 3’ splice-sites (14%), intron retention
(11%), mutually exclusive exon (4%), and others (4%).
Additionally, Altanalyze determined that hypoxia induced 2439 genes, for at least 1.4 fold, when compared to normoxic Hep3B controls (Fig. 3.1A, blue and grey; also see Supplemental File 1, Gene Expression). Venny analysis of hypoxia inducible genes and AS genes indicated that 134 genes were hypoxia induced and alternatively spliced and these 134 genes were called Hx AS genes
(Fig. 3.1A, in grey). Interesting, these 134 Hx AS genes had 238 splicing events, averaging 1.78 AS events per gene (Fig. 3.1A, Hx Induced Genes), suggesting that AS is enriched in hypoxia induced genes, compared to the total population of 88
genes that undergo AS (1.53 AS events per gene). Similar to the total population
of AS genes, alternative cassette-exons were the most abundant type of splicing events (62%) in Hx AS genes. Other AS events for these Hx AS genes included alternative 5’ splice-sites (16%), alternative 3’ splice-sites (13%),and intron
retention (9%). These data indicated that there is a 1.2 fold enrichment in
alternative cassette-exon inclusion and a 1.22 fold reduction in intron-retention in
Hx As genes in comparison with the total population of AS genes.
Since cassette-exon AS is the most abundant splicing event, hypoxia-
mediated cassette-exon regulation was further analyzed. In the total population
of cassette-exon genes (All Cassette, Fig. 3.1B), hypoxia promoted exon
inclusion in 454 genes out of 1022 (51% of 2005 genes) genes (44%) and
promoted exon skipping in 568 genes out of 1022 gene (56%), suggesting that
hypoxia generally promotes exon skipping in Hep3B cells. However, Hx AS
genes primarily underwent exon inclusion, in which 63 out of 83 genes (62% of
134 genes) underwent exon inclusion (76%), whereas 24 out of 83 genes
exhibited exon skipping (24%) (Fig. 3.1B, Hx induced AS). In contrast, hypoxia
promoted exon skipping of 75% of hypoxia reduced genes (Hx reduced, Fig.1B)
whereas the number of exon inclusion to exon skipping events was equal (50%)
for genes whose expression was not changed by hypoxia (No Change, Fig.
3.1B). In summary, these data indicated that hypoxia regulates AS of genes
whose expression levels are increased, decreased, or not changed during
hypoxia. Importantly, hypoxia promotes exon inclusion of Hx AS genes and
decreases exon inclusion of hypoxia-reduced AS genes. 89
Figure 3.1
Figure 3.1: Hypoxia Regulates Gene Expression and RNA Splicing in Hep3B Cells. 90
Figure 3.1: Hypoxia Regulates Gene Expression and RNA Splicing in Hep3B Cells. A) Venn diagrams of genes that are hypoxia induced (blue and grey), genes that are alternatively spliced under hypoxia (yellow and grey), and genes whose expression are induced by hypoxia and alternatively spliced under hypoxia (grey). The left column indicates the top 10 enrichment pathways whose genes are induced at least 1.4 fold by hypoxia. The top row indicates the top 10 enrichment pathways whose genes that are hypoxia-induced and undergo hypoxia-induced alternative splicing. The right column represents the top 10 enrichment pathways whose genes undergo hypoxia-induced alternative splicing. Numbers next to each box represent the DAVID enrichment score (ES), the higher the more enriched, followed by the Fisher Exact P-Value (P), the smaller the more enriched. P-Values equal to or less than 0.05 are considered strongly enriched. Boxes of the same color, excluding white, represent overlapping or functionally related gene pathways. White boxes are unique pathways for each group. Below the Venn diagram in Fig. 3.1A are the splicing events per genes for all alternatively spliced genes (All Genes, left) and hypoxia induced genes (Hx Induced Genes, right). B) Summary of exon skipping or inclusion events for all genes that undergo alternative cassette exon splicing (all cassette AS), genes whose expression are not changed during hypoxia (No Change AS), genes induced by hypoxia (Hx induced AS), and genes reduced by hypoxia (Hx reduced AS). 91
Functional Clustering of Genes that Undergo Hypoxia-Mediated AS Reveals Novel
Pathways Regulated By Hypoxia
To determine if hypoxia-induced AS was enriched in specific cellular pathways,
DAVID (331) was used to create functional annotation clusters of genes that were
hypoxia induced (Fig. 3.1A, left column), the entire population of AS genes (Fig. 3.1A,
right column), and the Hx AS genes (Fig. 3.1A, top row). As expected, hypoxia induced
the expression of genes in pathways allowing cancer cells to adapt to a hypoxic
microenvironment such as vascular development, glycolysis, glucose metabolism, L-
ascorbic acid binding, leucine-zipper transcription factors, G-protein couple receptors,
gluconeogenesis, transmembrane proteins, oxidoreductase activity, and cell
adhesion/cadherins (Fig. 3.1A, left column). Correspondingly, Hx AS genes were largely
overlapped with gene induction pathways including hexose metabolic processes (which
are related to glycolysis and gluconeogenesis), glycolysis, oxidoreductase activity, L-
ascorbic acid binding, blood vessel morphogenesis (related to vascular development),
gluconeogenesis, leucine-zipper transcription factor, and basement membrane pathways
(related to oxidoreductase activity) (Fig. 3.1A, top row). Interestingly, Hx AS genes were
also involved in membrane-bound vesicles and cell death pathways, the pathways that
were not significantly enriched by hypoxia when analyzing gene expression alone (Fig.
3.1A, left column). On the other hand, when analyzing all AS genes, gene clusters were
enriched for cellular pathways involving ATP-binding/protein kinase activity, pleckstrin
homology (related to rho protein signal transduction and cytoskeleton organization),
oxidoreductase activity, rho protein signal transduction, cytoskeleton organization,
zinc fingers, glycolysis, positive regulation of glucose transport, cell death and
nuclear lumen genes (Fig. 1A, right column). In general, gene pathways that
regulate glycolysis and glucose metabolism or oxidoreducase activity were 92
enriched in all three groups analyzed. This was not surprising since hypoxia shifts ATP metabolism from oxidative phosphorylation to glycolysis and since many genes involved in oxidoreducase activity act as oxygen-sensors. However, hypoxia induces gene expression in glycolytic pathways to generate ATP, but alters AS of genes involved in ATP-binding/protein kinase activity to reduce ATP usage, both of which act to maintain cellular ATP levels. In addition, hypoxia regulates other novel pathways such as rho protein signal transduction by regulating AS, but not gene transcription.
RT-PCR and qRT-PCR Validation of Alternative Splicing for Select HIF
Target Genes identified in Exon Array Analysis
Next, we sought to validate the exon array data using RT-PCR and qRT-
PCR. We focused on Hx AS genes since genes that are not hypoxia induced have been studied in endothelial cells (281, 329). First, we selected Hx AS genes as determined by both Altanalyze Firma and easyExon. Next, we used
EasyExon to identify regions of potential AS. Then we used RT-PCR to identify the RNA isoforms and qRT-PCR to quantify the levels of these isoforms in normoxic and hypoxic Hep3B cells or other cancer cells. For example, carbonic anhydrase 9 (CA9), a well-known HIF target gene (302, 332), was predicted to undergo AS in exon 2 and in a region spanning exons 6-10 as determined by the splicing index (Fig. 3.2A). AS was considered significant if the splicing index was greater than +1.5 or less than -1.5 (negative SI indicates exon inclusion by hypoxia) using EasyExon. RT-PCR was not able to detect an AS event in exon 2. 93
RT-PCR, using primers that amplify exons 4-10 (Fig. 3.2B), followed by Topo TA cloning and sequencing confirmed that Hep3B cells expressed two CA9 isoforms, a full-length isoform (FL), and an isoform in which exons 8-9 were skipped (ΔE89) (Fig. 3.2C left). qRT-PCR determined that hypoxia induced the expression of both isoforms but favored FL in Hep3B cells, thus hypoxia increased the FL/ΔE89 ratio by 9.3 fold (Fig. 3.2D). Similar CA9 isoform expression patterns and increased CA9 FL/ΔE89 ratio by hypoxia were observed in a neuroblastoma cell-line, SKNMC, suggesting that this splicing event is not cell-type specific (Fig. 3.2E-F). In addition, western blot of CA9 in Hep3B cells confirmed that both FL and ΔE89 protein isoforms were expressed and FL CA9 protein was preferentailly induced by hypoxia (Fig. 3.2C right). The larger CA9
FL band in hypoxic Hep3B cells was likely the result of post-translation modifications and was also observed previously (333).
Angiopoietin-like 4 (ANGPTL4) was predicted to undergo AS in exon 4
(Fig. 3.3A). RT-PCR, Topo TA cloning and sequencing confirmed that ANGPTL4 expressed a full-length isoform (FL), and an exon 4 skipping isoform (ΔE4) in
Hep3B cells (Fig. 3.3C left). qRT-PCR determined that hypoxia induced the expression of both isoforms but FL was favored; thus hypoxia increased the
FL/ΔE4 ratio by 4.9 fold (Fig. 3.3D). Additionally, ANGPTL4 FL and ΔE4 were expressed in a head and neck squamous cell carcinoma cell-line, UM-SCC-22B.
Also, hypoxia increased the ANGPTL4 FL/ΔE4 ratio by 3.5 fold in UM-SCC-22B
(Fig. 3.3E-F). The expression of ANGPTL4 FL and ΔE4 protein isoforms and
94
Figure 3.2
Figure 3.2: Hypoxia regulates AS of the HIF target gene, CA9 in Hep3B and SKNMC Cells. A) EasyExon analysis of differentially expressed probes for the CA9 gene in Hep3B cells. The red or blue line indicates probe intensities for normoxia or hypoxic samples. Differentially expressed probes were considered positive for alternative splicing if the splicing index was greater than 1.5 or less than -1.5 (negative SI indicates exon inclusion is increased by hypoxia). Vertical numbers in red below the splicing index indicate affymetric probe IDs. B) Diagram of the confirmed CA9 gene isoforms. Grey boxes represent constitutive exons and black boxes represent alternatively spliced exons. Solid arrows indicate primer used for RT-PCR (E4//E10) and qRT-PCR (E7//E10/7 for CA9∆E89, and E7//E8/7 for CA9 FL). C) RT-PCR and western blot analysis of CA9 mRNA and protein in normoxic and hypoxic Hep3B cells. CA9 FL proteins exhibit two bands with different sizes. D) qRT-PCR analysis of CA9FL and ∆E89 transcripts in normoxic and hypoxic Hep3B cells. E) RT-PCR analysis of CA9 mRNAs in normoxic and hypoxic SKNMC cells. F) qRT-PCR analysis of CA9FL and ∆E89 transcripts in normoxic and hypoxic SKNMC cells. Numbers in the panels D and E here and other qRT-PCR panels represent the ratio of the full- length isoform versus the exon skipping isform ± the standard deviation. Fold of ratio change is also indicated. One-way analysis of variance was performed for this and other studies in this paper unless otherwise stated. *, P= 0.05; **, P =0.01. Controls for statistical analysis are specified in each Figure. 95
preferential induction of ANGPTL4 FL protein by hypoxia were also observed in
Hep3B cells (Fig. 3.3C right).
Expression of multiple isoforms and hypoxia regulating RNA splicing of additional HIF target genes, pyruvate dehydrogenase kinase 1 (PDK1), WNK
lysine deficient protein kinase 1 (WNK1) and prolyl 4-hydroxylase alpha
polypeptide II (P4HA2), were also confirmed by RT-PCR, Topo-TA cloning and qRT-PCR (Fig. 3.4A-C). For example, the PDK1 gene expressed two isoforms, a
full-length isoform (PDK1FL, Nm_001278549) and an exon 4 skipping isoform
(PDK1ΔE4). Moreover, hypoxia enhanced the FL/ΔE4 ratio by 1.44 fold (Fig.
3.4A). The WNK1 gene expressed at least 3 isoforms, a full-length isoform
(WNK1FL, Nm_0018979), an exon 11 skipping isoform, and an exon 11-12
skipping isoform (WNK1ΔE11-12). Moreover, hypoxia enhanced the FL/ΔE11-12
ratio by 2.36 fold (Fig. 3.4B). The P4HA2 gene expressed 2 isoforms, a full-
length isoform (P4HA2FL, Nm_001142599) and an exon 2 skipping isoform
(P4HA2ΔE2). Again, hypoxia increased the FL/ΔE2 ratio by 2.5 fold (Fig. 3.4C).
CA9, ANGPTL4, PDK1, WNK1, and P4HA2 exhibited preferential induction of the
FL isoform. However, AltAnalyze predicts that hypoxia promotes exon skipping
of some HIF target genes including procollagen-lysine, 2-oxoglutarate 5-
dioxygenase 2 (PLOD2), and enolase 2 (ENO2). Therefore we selected PLOD2
and ENO2 for confirmation. Indeed, the PLOD2 gene expressed 2 isoforms, a
full-length isoform (PLOD2FL, Nm_182943) and an exon 14 skipping isoform
(PLOD2ΔE14). As predicted by exon array, hypoxia favored the expression of
the ΔE14 isoform and reduced the FL/ΔE14 ratio by 1.6 fold (Fig. 4D). The 96
Figure 3.3
Figure 3.3: Hypoxia Regulates AS of the HIF target Gene, ANGPTL4 in Hep3B and UM-SCC-22B Cells. A) EasyExon analysis of differentially expressed probes for the ANGPTL4 gene in normoxic and hypoxic Hep3B cells. B) Diagram of the confirmed ANGPTL4 gene isoforms. Solid arrows indicate primer used for RT-PCR (E1//E6) and qRT-PCR (E3//E5/3 for ∆E4 and E4//E5/4 for FL). C) RT-PCR and western blot analysis of ANGPTL4 mRNAs and proteins in normoxic and hypoxic Hep3B cells. D) qRT-PCR analysis of ANGPTL4 FL and ∆E4 transcripts in normoxic and hypoxic Hep3B cells. E) RT-PCR analysis of ANGPTL4 mRNAs in normoxic and hypoxic UM-SCC-22B cells. F) qRT-PCR analysis of ANGPTL4 FL and ∆E4 transcripts in normoxic and hypoxic UM-SCC- 22B. Also, qRT-PCR data was normalized to normoxia samples for each individual gene isoform.
97
ENO2 gene expressed a full-length isoform (ENO2FL, Nm_001975) and an exon
8 skipping isoforms (ENO2ΔE8). Moreover, hypoxia preferentially induced the expression of the ΔE8 isoform and reduced the FL/ΔE8 ratio by 6.7 fold (Fig. 4E).
In summary, we selected seven genes that are induced by hypoxia and undergo
hypoxia-mediated AS (Hx cassette-exon genes) and confirmed the exon array
analysis using RT-PCR, cloning, sequencing, qRT-PCR, and western blot.
These data demonstrated that hypoxia promotes exon inclusion for most HIF
target genes; however, hypoxia promotes exon skipping of some HIF target
genes such as PLOD2 and ENO2.
Differential Expression of HIF Target Genes during Hypoxia is Due to
Alternative Splicing
After confirming AS of several HIF target genes, we addressed the
molecular mechanism on how hypoxia regulates AS of HIF target genes.
Although AS was proposed to explain the observed FL to skipping isoform ratio
changes in the above sections, the hypoxia-induced isoform ratio variations could
be caused by RNA splicing or other mechanisms. For instance, preferential
degradation of exon skipping isoforms under hypoxia might explain the hypoxia-
induced isoform ratio changes since exon skipping RNA isoforms may incur
premature termination codons and such transcripts are often targeted for
nonsense-mediated decay (NMD) (299). In the seven genes analyzed, only RNA
of CA9 ΔE89 (E8-9, 172bp), but not ANGPTL4 (E4, 114bp), PDK1 (E4, 138bp),
WNK1 (E11-12, 609bp), P4HA2 (E2, 136bp in 5’UTR), PLOD2 (E13, 114bp), and 98
Figure 3.4
Figure 3.4: Hypoxia Regulates AS of Additional HIF Target Genes in Hep3B Cells. RT-PCR (left) and qRT-PCR (right) analysis of mRNAs of PDK1 (A), WNK1 (B), P4HA2 (C), PLOD2 (D), and ENO2 (E) in normoxic and hypoxic Hep3B cells. qRT-PCR data was normalized to normoxia samples for each individual isoform.
99
ENO2 (E8, 198bp) had premature stop codons and thereby predicted to be subject to NMD. To test the possibility that preferential instability of CA9 ΔE89 contributes to the increased CA9 FL/ΔE89 ratio under hypoxia, Hep3B cells were placed under hypoxia for 16 hrs to increase the levels of both CA9 FL and ΔE89 transcripts, followed by treatment with actinomycin D to inhibit de novo transcription. Cells were then placed back under normoxia or hypoxia for 0, 2, 4, or 8 hours to allow RNA decay. Using qRT-PCR, both CA9 FL and ΔE89 transcripts were found to be very stable since 90% of the FL and ΔE89 transcripts were detected even after 8 hrs. Moreover, CA9 FL and ΔE89 transcripts exhibited similar stability under normoxia and hypoxia at every time point. ANGPTL4 FL and ΔE4 transcripts were far less stable than CA9 transcripts since only 40%, 28%, and 20% of the transcripts were remained after
2, 4, and 8 hrs. However, ANGPTL4 FL and ΔE4 transcripts also exhibited similar stability under normoxia and hypoxia. Furthermore, actinomycin D treatment in Hep3B cells blocked hypoxic induction of HIF target genes and blocked splicing change of HIF target genes, indicating that the hypoxia- mediated isoform shift required active transcription. These data supported the idea that alternative splicing, not post-transcriptional regulation is responsible for the hypoxia-induced increased CA9 and ANGPTL4 FL/exon-skipping ratio.
100
Figure 3.5
Figure 3.5: HIF Activity, but Not Hypoxia Per Se is Necessary to Promote AS of HIF Target Genes. A) RT-PCR analysis of CA9, ANGPTL4, and PDK1 RNA transcripts in normoxic or hypoxic Hep3B cells targeted with control (Ctrl), ARNT, HIF1α, or HIF2α siRNAs. B-D) qRT-PCR analysis of FL and exon skipping isoforms of CA9 (B), ANGPTL4 (C), and PDK1 (D) in the normoxic or hypoxic Hep3B cells targeted with control, ARNT, HIF1α, or HIF2α siRNAs.
101
HIF Activity, Not hypoxia Per Se, is Necessary to Regulate AS of HIF Target
Genes
To test whether hypoxic stress or HIF activity is responsible for regulating
the splicing changes observed in HIF target genes, ARNT, HIF1α or HIF2α
mRNA levels were reduced by 80% using siRNAs in normoxic or hypoxic Hep3B
cells (data not shown). ARNT and HIF1α, but not HIF2α knockdown dramatically
reduced the hypoxic induction of CA9, ANGPTL4, and PDK1, consistent with the
idea that CA9, ANGPTL4, and PDK1 are primarily regulated by HIF1 in Hep3B
cells (Fig. 3.5A). qRT-PCR confirmed that ARNT and HIF1α knockdown
significantly reduced the levels of both FL and exon skipping isoforms of CA9,
ANGPTL4, and PDK1 and prevented the splicing changes of these genes (Fig.
3.5B-D). In contrast, HIF2α knockdown only mildly reduced hypoxic induction of
CA9FL (1.44 fold), but did not significantly alter the expression of CA9ΔE89, thus
HIF2α knockdown only reduced the CA9FL/ΔE89 ratio by 1.33 fold (Fig. 3.5B).
HIF2α knockdown reduced hypoxic induction of ANGPTL4FL and ΔE4 by 1.6 fold and 1.8 fold, maintaining the ANGPTL4FL/ΔE4 ratio (Fig. 3.5C). Additionally,
hypoxic induction of PDK1FL was not significantly changed by HIF2α
knockdown, whereas PDK1ΔE4 was reduced 1.6 fold, enhancing the
PDK1FL/ΔE4 ratio 1.5 fold (Fig. 3.5D). Knockdown of ARNT and HIF1α also
inhibited hypoxic induction of the FL and exon skipping isoforms of WNK1,
PLOD2, ENO2, and P4HA2 in Hep3B cells and prevented splicing ratio changes
for these genes. These data suggested that HIF activity, but not hypoxia per se
102
Figure 3.6
Figure 3.6: HIF Activity is Sufficient to Promote AS of HIF Target Genes in Hep3B Cells. A) Western blot analysis of Flag-tagged proteins in normoxic Hep3B cells transduced with lenti-viruses expressing GFP, or Flag-tagged, normoxia-active HIF1α or HIF2α proteins. B) RT-PCR analysis of CA9, ANGPTL4 and PDK1 transcripts in normoxic Hep3B cells transduced with GFP, or normoxia-active HIF1α or HIF2α proteins. C-E) qRT-PCR analysis of FL and exon-skipping RNA isoforms of CA9 (C), ANGPTL4 (D), and PDK1 (E) in the above described cells.
103
Figure 3.7
Figure 3.7: HIF Activity is Sufficient to Promote AS of HIF Target Genes in Normoxic RCC4 Cells. A) RT-PCR analysis of CA9, ANGPTL4 and PDK1 transcripts in normoxic and hypoxic RCC4T and RCC4 cells. B-D) qRT-PCR analysis of FL and indicated exon-skipping isoforms of CA9 (B), ANGPTL4 (C), and PDK1 (D) in the normoxic or hypoxic RCC4T and RCC4 cells.
104
is necessary for increased gene expression as well as hypoxia-mediated splicing changes of these HIF target genes.
HIF Activity is Sufficient to Regulate AS of HIF Target Genes
Next, we wanted to determine if HIF activity is sufficient for hypoxia regulated AS of HIF target genes. To test this, normoxic Hep3B cells were transduced with lentiviruses expressing normoxia active, flag-tagged, HIF1α or
HIF2α, or GFP as a negative control. As expected, HIF1αTM-Flag or HIF2αTM-
Flag transduced Hep3B cells exhibited HIF1α or HIF2α protein expression under
normoxia (Fig. 3.6A). HIF1α and HIF2α transduction induced the expression of
CA9, ANGPTL4, and PDK1 as determined by RT-PCR (Fig. 3.6B). More
importantly, qRT-PCR determined that HIF1α and HIF2α increased both FL and
exon skipping isoforms of CA9 (Fig. 3.6C), ANGPTL4 (Fig. 3.6D), and PDK1
(Fig. 3.6E). However, FL transcripts of CA9, ANGPTL4, and PDK1 were preferentially induced (Fig. 3.6C-E). Thus, HIF1 or HIF2 increased the FL/exon skipping ratio for CA9 by 14.5 or 6.0 fold (Fig. 3.6C), ANGPTL4 by 3.41 or 5.7 fold (Fig. 3.6D), and PDK1 by 1.43 or -1.2 fold (Fig. 3.6E).
To further validate these results, the RNA splicing of CA9, ANGPTL4, and
PDK1 were examined in RCC4 cells, a renal cell carcinoma cell-line that expresses constitutively active HIF1α and HIF2α proteins due to loss of functional pVHL (63, 302). In addition, the RNA splicing of the above three genes were also assessed in RCC4T cells in which functional pVHL is reintroduced into RCC4 cells and therefore HIF proteins are only active under 105
Figure 3.8
Figure 3.8: Transcription Activation is Not Sufficient to Regulate Splicing of a PDK1 Minigene. A) Schematic of a PDK1 splicing reporter driven by CA9 or PAI1 promoter. Arrows represent primers used for RT-PCR (E3//vector R) and qRT-PCR (E3/5, E4/5, or total with Vector R for ΔE4, FL, or total transcript respectively). B) RT-PCR analysis of PDK1 transcripts expressed from CA9P/PDK1 in 293T and Hep3B cells. C) Western blot analysis of Flag-tagged HIF1α, HIF1αDBD/VP16TAD, and HIF1αDBD/E2F1TAD proteins in 293T cells for experiments whose results are presented in Fig. 8D. HIF1αDBD/E2F1TAD construct produces two proteins. D) qRT-PCR analysis of minigene specific PDK1 isoforms in 293T cells co-transfected with CA9P/PDK1 minigene and HIF1α or HIF1αDBD expression constructs. E). qRT-PCR analysis of minigene specific PDK1 isoforms in 293T cells co-transfected with PAI1P/PDK1 minigene and HIF2α or USF2 expression constructs. F). qRT-PCR analysis of minigene specific PDK1 isoforms in 293T cells co-transfected with PAI1PmHRE/PDK1 minigene and HIF2α or USF2 expression constructs. 106 hypoxia (63, 302). RT-PCR confirmed that FL and exon skipping isoforms of
CA9, ANGPTL4, and PDK1 were expressed in both RCC4T and RCC4 cells (Fig.
3.7A). Additionally, RT-PCR analysis confirmed that these genes were hypoxia induced in RCC4T cells, and were constitutively expressed in RCC4 cells (Fig.
3.7A). qRT-PCR analysis determined that hypoxia increased the levels of
CA9FL and ΔE89 by 41.8 and 7 fold respectively in RCC4T cells, increasing the
FL/ΔE89 ratio by 6.2 fold (Fig. 3.7B). In contrast, hypoxia did not induce the levels of CA9FL and ΔE89 isoforms nor did hypoxia enhance the FL/ΔE89 ratio in RCC4 cells (Fig. 3.7B). However, the CA9 FL/ΔE89 ratio in RCC4 cells was already high even under normoxia (Fig. 3.7B). Similar findings were observed for
AS of ANGPTL4; however, hypoxia was still able to enhance the ANGPTL4
FL/ΔE4 ratio in RCC4 (Fig. 3.7C). Although, hypoxia was able to increase the expression of the PDK1FL and ΔE4 isoforms and to enhance the PDK1FL/ΔE4 ratio in RCC4T cells, PDK1 FL/ΔE4 ratio was not elevated in normoxic or hypoxic
RCC4 cells (Fig. 3.7D). Taken together, these data suggest that HIF activity is necessary and sufficient to regulate alternative splicing of a subset of HIF target genes independent of hypoxia.
PDK1 Splicing Reporters Recapitulate Splicing Changes Observed for the
Endogenous PDK1 Gene when Activated by HIF under Normoxia
Next, we wanted to see if similar splicing changes could be observed in splicing reporter gene. We selected the PDK1 gene for our splicing reporter model as both the FL and ΔE4 isoforms were highly expressed in 293T cells 107
(data not shown). Exons 3-5 including introns 3 and 4 of the PDK1 gene were
PCR amplified from human genomic DNA and placed downstream of the CA9 promoter, a HIF1 target gene promoter, or the PAI1 promoter, a HIF2 target gene promoter (Fig. 3.8A). RT-PCR using minigene specific primers determined that the PDK1 minigene expressed both FL and ΔE4 isoforms in 293T and Hep3B cells (Fig. 3.8B). In addition, the PDK1ΔE4 isoform was the dominantly expressed isoform in both 293T and Hep3B cells (Fig. 3.8B). Furthermore, PDK
FL isoforms were also highly expressed in 293T cells (Fig. 3.8B). These data demonstrated that the PDK1 splicing reporter recapitulated the expression patterns of the endogenous PDK1 gene in 293T and Hep3B cells (Fig. 3.4A and data not shown).
To assess AS of the PDK splicing reporter in response to HIF activation,
293T cells were transfected with the PDK1 minigene and normoxia active HIF1α expression plasmids (Fig. 3.8C). HIF1α induced the levels of total PDK1 minigene transcripts, FL, and ΔE4 isoform by 3.65, 24.9, and 10.6 fold respectively and enhanced the FL/ΔE4 ratio by 2.34 fold (Fig. 3.8D). Next, similar experiments were performed using the PAI1P/PDK1 splicing reporter (Fig.
3.8A). Normoxia active HIF2α induced the expression of PDK1FL and ΔE4 isoforms by 4.22 and 2.36 fold respectively, thus enhancing the FL/ΔE4 ratio
2.08 fold (Fig. 3.8E). These results indicated that the PDK1 splicing reporters recapitulate splicing changes observed for the endogenous PDK1 gene when activated by HIF under normoxia. 108
Some transcription factors have dual roles in RNA splicing and gene transcription by recruiting splicing factors and transcription co-factors using their transactivation domains (303-305). To determine if the activation domain of
HIF1α protein is important for PDK1 pre-mRNA splicing, fusion constructs containing the HIF1α DNA binding domain fused to the transactivation domains from the VP16 or E2F1 transcription factors were used to activate CA9P/PDK1 splicing reporter (Fig. 3.8C). HIF1αDBD/VP16 induced the expression of
PDK1FL (12.2 fold) and PDK1ΔE4 (6.4 fold), and enhanced the FL/ΔE4 ratio by
1.9 fold (Fig. 3.8D). Similarly, HIF1αDBD/E2F1 induced PDK1FL (21.7 fold) and
PDK1ΔE4 (9.4 fold), and increased FL/ΔE4 expression ratio by 2.3 fold (Fig.
3.8D). These data indicated that HIF transactivation domain is not required for
AS of the PDK1 minigene. In summary, these data indicated that HIF1 or HIF2- mediated activation of the PDK1 minigene is sufficient to increase PDK1 FL/ΔE4 ratio.
Activation of Endogenous HIF Target Genes Contributes to the Increased
FL/ΔE4 ratio of the PDK1 Splicing Reporter
As stated above, transcription activation of PDK1 by HIF or HIFDBD hybrid constructs is sufficient to increase the PDK1 minigene FL/ΔE4 ratio.
However, HIF and the fusion constructs can activate endogenous HIF target genes; therefore it is possible that activation of endogenous HIF target genes may increase FL/ΔE4 ratio. To rule out or confirm this possibility, we utilized upstream stimulatory factor 2 (USF2) and the PAI1P/PDK1 minigene since USF2 109
Figure 3.9
Figure 3.9. Activation of Endogenous HIF Target Genes Promotes AS of a PDK1 Minigene. A) Diagram of a PDK1 splicing reporter under control of 5 copies of Gal4 binding element. B) Western blot to monitor the protein expression of Flag-tagged HIF1α and HIF2α (top panel, using anti-Flag) and Gal4 DBD, Gal4DBD/HIF1αTAD, Gal4DBD/VP16TAD, Gal4DBD/E2F1TAD (bottom panel, using anti-Gal4 DBD) in 293T cells whose results are presented in Fig. 9C. Gal4DBD/E2F1TAD construct produces two proteins. C). qRT-PCR analysis of minigene specific PDK1 isoforms in 293T cells co-transfected with G5/PDK1 minigene and Gal4 DBD (Gal4), Gal4DBD/HIF1αTAD (Gal4/HIF1α), Gal4DBD/VP16TAD (Gal4/VP16), or Gal4DBD/E2F1TAD (Gal4/E2F1) with or without normoxia-active HIF1α and HIF2α (HIF1/2α). 110
can activate the PAI1 promoter (306), but not endogenous HIF target genes.
USF2 induced the expression of the PDK1FL and ΔE4 isoforms by 2.95 and 2.96
fold respectively but was not able to enhance the FL/ΔE4 ratio even though
USF2 was able to activate the minigene to a similar extent as HIF2α (USF2=1.97
and HIF2α=1.52 fold induction of total transcripts) (Fig. 3.8E). To test if
activation of endogenous HIF target gene could regulate AS of the PDK1
minigene, HIF-binding site on the PAI1 promoter was mutated to produce the
PAI1PmHRE/ PDK1 minigene (Fig. 3.8F). Interestingly, although HIF2α was not
able to activate the expression of PAI1PmHRE/PDK1 as assessed by total as
well as FL transcripts, HIF2α increased the FL/ΔE4 ratio by reducing the levels of
the PDK1 ΔE4 transcript (Fig. 3.8F).
To further validate that transcription activation of endogenous HIF target
genes increases the PDK1 FL/ΔE4 ratio, the PDK1 splicing reporter was placed
under the control of a promoter containing 5 copies of the Gal4 DNA binding
element (Fig. 3.9A, 5xUAS). Fusion constructs containing the Gal4 DNA binding
domain fused to the transactivation domains of normoxia active HIF1α, VP16, or
E2F1 were used to activate the G5P/PDK1 minigene in the presence or absence
of normoxia-active HIF1α and HIF2α that activate endogenous HIF target genes
(Fig. 3.9B). Interestingly, although Gal4/HIF1αTAD, Gal4/VP16TAD, and
Gal4/E2F1TAD increased the PDK1 FL/ΔE4 ratio by 1.27, 2.44, and 1.47 fold,
activation of endogenous HIF target genes further increased the PDK1 FL/ΔE4
ratio to 2.04, 3.28, and 3.29 fold. Importantly, co-tranfected HIF1α and HIF2α
increased the expression of endogenous HIF target genes including CA9 and 111
LOX (data not shown), but not the expression of G5/PDK1 (Fig. 3.9C, same
amount of PDK1 total transcripts with or without HIF). These data demonstrated
that activation of endogenous HIF target genes contributes to AS of the PDK1
minigene.
Discussion
Most analyses of hypoxia-mediated gene expression changes are
conducted at the gene level using traditional microarrays (302, 324-327).
However, due to significant functional difference between RNA isoforms and the
fact that a majority of genes express multiple RNA isoforms (328), gene
expression analysis by exon-array or RNA deep sequencing is necessary to fully understand gene expression programs. Our exon-array analysis in normoxic and
hypoxic Hep3B cells not only confirms hypoxic induction of previously identified
hypoxia-inducible genes (Fig. 3.1A) but also reveals significant changes in RNA
splicing induced by hypoxia.
Here, we reported novel hypoxia-regulated genes/pathways that cannot
be identified using traditional microarrays since expression levels of these genes
are typically not changed by hypoxia or reduced by hypoxia. In addition,
consistent with what is reported in endothelial cells (281, 329), we also found that
hypoxia primarily promotes exon skipping of these non-HIF target genes in
Hep3B cells. For example, most genes involved in ATP binding and protein
kinase activity exhibits exon skipping in hypoxic Hep3B cells, presumably to
reduce ATP usage to maintain cellular ATP levels. In addition, most of these 16
genes involved in the activation of Rho protein signal transduction exhibit exon 112 skipping in hypoxic Hep3B cells. Rho proteins are required for HIFα protein accumulation during hypoxia (334). Although speculative, it is conceivable that
HIFs/hypoxia may act in a negative feedback loop to inhibit HIFα protein accumulation during hypoxia by reducing rho protein activities via splicing inhibition. AS of some of these non hypoxia-induced genes are not associated with enriched pathways identified by DAVID Bioinformatics, however, their splicing is also altered and expected to be functionally important. For example, the KRAS 4B isoform (exon 4a is skipped in KRAS 4B, but included in KRAS 4A) is induced in hypoxic Hep3B, thus reducing the KRAS 4A/4B ratio. Interestingly, a decreased KRAS 4A/4B ratio is often observed in human colorectal cancer, but not in normal colon (335, 336). Furthermore, decreased KRAS 4A/4B ratios through reduction of KARS 4A expression or increased KRAS 4B expression promotes DMH-induced colonic tumorigenesis in in vivo mouse models independent of KRAS mutations (337), demonstrating the functional consequence of reduced KRAS 4A/4B ratio in tumor biology.
More importantly, we reported here, for the first time that hypoxia promotes exon inclusion for 75% of Hx AS genes. RT-PCR and qRT-PCR confirmed that hypoxia promotes exon inclusion for the HIF target genes, CA9,
ANGPTL4, PDK1, WNK1, and P4HA2 in Hep3B, RCC4T, and RCC4 cells. In addition, we confirmed that FL proteins of CA9 and ANGPTL4 are preferentially induced in Hep3B cells. Thus, these data indicate for the first time that RNA splicing changes under hypoxia is significantly different between hypoxia inducible genes and genes not induced by hypoxia. 113
Although no functional differences have been examined for the isoforms of most HIF target genes studied in this report, FL CA9 protein was reported to be a membrane-associated, tumor-promoting protein by regulating intercellular pH by transporting protons out of the cell during hypoxia. In contrast, the CA9∆E89 protein is a soluble protein and is not able to transport protons out of the cell
(333). Thus, it makes sense that CA9 FL is preferentially induced by hypoxia.
Future studies will assess the functional difference between FL and exon- skipping isoforms of these HIF target genes and to characterize the functional consequences of altering RNA splicing of these HIF target genes.
While AS of several HIF target genes have been reported (276-278, 280,
333, 338), none of these studies have addressed the molecular mechanisms concerning how hypoxia regulates AS of HIF target genes. We demonstrated in this study that the increased FL/exon-skipping ratio of several HIF target genes including CA9 and ANGPTL4 is the result of alternative splicing, not preferential degradation of exon-skipping isoforms. Moreover, we determined that HIF activity is necessary and sufficient to regulate AS of CA9, PDK1, and ANGPTL4 pre- mRNAs. Furthermore, we found that activation of endogenous HIF target genes contributes to AS of PDK1 minigenes. In the future, it will be important to uncover the identities of and the mechanism by which HIF target gene(s) regulate alternative splicing of HIF target genes.
In summary, this study suggests that hypoxia regulates alternative splicing of hypoxia induced genes, hypoxia reduced genes, and genes whose transcription is not changed by hypoxia. Additionally, HIF activity, not hypoxic 114
stress is found to be necessary and sufficient to regulate AS of a subset of
hypoxia inducible genes. Moreover, activation of endogenous HIF target genes
contributes to AS of some HIF target genes. These findings significantly
increase our understanding of how cancer cells regulate gene transcription as
well as RNA splicing to adapt to a hypoxic microenvironment.
Acknowledgments
This work was supported by grants from the National Cancer Institute
(RO1CA134687, Hu) and Cancer League of Colorado (Hu). Johnny Sena was
supported by “Research Supplemental to Promote Diversity” (RO1CA134687-S3 and RO1CA134687-S4).
115
CHAPTER IV
HYPOXIC INDUCTION OF CDC-LIKE KINASES REGULATES
TRANCRIPTION AND ALTERNATIVE SPLICING OF HIF TARGET GENES
Abstract
The hypoxic microenvironment stabilizes hypoxia-inducible transcription factor 1α (HIF1α) and 2α (HIF2α) to activate gene transcription, which promote tumor cell survival. 95% of human genes are alternatively spliced, producing
RNA isoforms that code for functionally distinct proteins. Thus, effective hypoxia
responses require regulation of gene transcription and RNA splicing. Previously,
we determined that HIFs, through activation of HIF target genes, regulate
hypoxia-mediated splicing. In this study, we determined that HIFs regulate
alternative splicing and transcription of HIF target genes through induction of
CLK1 and 4, which enhance phosphorylation of SR proteins SRp75, SRp40 and
SC35 during hypoxia. Inhibition of CLKs using TG003 reduced SR protein
phosphorylation and hypoxia-induced alternative splicing and transcription of HIF
target genes. Using siRNA knockdown, we determined that SC35 and SRp40
regulated hypoxia-induced alternative splicing of HIF target genes and that SC35
cooperates with HIFs to regulate splicing of a PDK1 minigene. Using TG003, we
determined that CLKs, like HIF2α, are necessary for survival and migration of
clear cell renal cell carcinoma by regulating alternative splicing of HIF target
genes. Importantly, we determined that HIF2α and CLKs function in the same 116 tumor promoting pathway since inhibition of CLKs in conjunction with stable knockdown of HIF2α did not further inhibit cell survival or migration in comparison to HIF2α knockdown alone.
Introduction
Hypoxia is a common characteristic of many solid tumors. The hypoxic microenvironment stabilizes hypoxia-inducible transcription factor 1α (HIF1α) and
2α (HIF2α) that are normally degraded under normoxia. Stabilized HIF1α and
HIF2α proteins translocate to the nucleus, where they dimerize with aryl hydrocarbon receptor nuclear translocator (ARNT) to form HIF1α /ARNT (HIF1) and HIF2 α/ARNT (HIF2) complexes. HIF1 and HIF2 activate gene expression to promote tumor progression and metastasis (320-323).
Recently, it was found that 95% of human genes are alternatively spliced producing multiple isoforms from a single gene which encode proteins with distinct or opposing functions (328). Thus, it is clear that assessing isoform expression is more informative than quantifying gene expression alone, allowing one to fully appreciate the functional consequence of gene expression programs.
So far, two studies have examined the role of hypoxia in regulating RNA splicing using exon arrays (281, 329). However, both studies used endothelial cells, a normal cell type. Additionally, previous studies in endothelial cells focused on non-HIF target genes and reported that hypoxia, like other stresses, inhibits RNA splicing of non-HIF target genes by promoting exon skipping and/or intron-inclusion (281, 329). Thus, the role of hypoxia in regulating RNA splicing of 117
HIF-induced genes was still unclear.
Subsequently, using exon arrays to assess global hypoxia-regulated splicing, we demonstrated that hypoxia regulated splicing of HIF and non-HIF target genes in Hep3B cells, a hepatocellular carcinoma cell line.
Mechanistically, we determined that HIFs were necessary and sufficient to regulate splicing of a subset of HIF targets including ADM, CA9, ANGPTL4,
WNK1, P4HA2, PLOD2 and ENO2. Also, activation of HIF target genes greatly contributed to hypoxia-regulated splicing. However, we did not determine which
HIF targets were responsible for regulating splicing.
Therefore, a major goal of this study was to identify which HIF targets are responsible for regulating HIF-mediated splicing during hypoxia and to determine the molecular signaling pathways involved. Additionally, we sought to determine the role of HIF-mediated splicing in promoting tumorigenesis of a clear cell renal cell carcinoma cell line.
Results
HIFs Regulate Hypoxic Induction of CDC-Like Kinase 1, and 3 and Hypoxia
Induces Phosphorylation of SR Proteins
In our previous studies, we demonstrate that HIFs are necessary and sufficient to regulate constitutive and alternative splicing of HIF target genes through induction of HIF target genes. Therefore, the next logical question is:
How do HIFs regulate splicing? It is possible that HIFs may induce the 118 expression of splicing regulators. To test this idea, we mined gene expression data from our exon array to identify potential hypoxia induced splicing regulators.
Interestingly, we found that hypoxia induced the expression of 3 out of 4 members of the CDC-like kinase family, CLK1, CLK3 and CLK4 but not CLK2
(Table 5.1). CLKs have been shown to phosphorylate serine- and arginine-rich
(SR) proteins of the spliceosomal complex and to regulate RNA splicing.
Western blot analysis determined that CLK 1, 3, and 4 were induced by hypoxia but CLK2 was not (Fig. 4.1A). Additionally, qRT-PCR analysis demonstrated that
CLK1, 3 and 4 mRNAs were induced by hypoxia and that siRNA-mediated knockdown of ARNT inhibited hypoxic induction of CLK1 and 3 but not CLK4, suggesting that CLK1 and CLK3 are HIF targets (Fig. 4.1B).
Since CLKs regulate pre-mRNA splicing through phosphorylation of SR proteins (Fig. 1.8) we wanted to confirm if phosphorylation of SR proteins was increased by hypoxia in Hep3B cells since hypoxia induced the expression of
CLKs (Fig 4.1C). Interestingly and consistent with the hypoxic induction of CLKs, western blot analysis using phospho-specific anti- SR protein antibody revealed that phosphorylation of SRp75, SRp40 and SC35 was induced by hypoxia (Fig.
4.1C). To confirm that hypoxia-induced phoshorylation of SR proteins was due to CLKs, Hep3B cells were treated with the CLK inhibitor, TG003, or with DMSO
(control). Western blot analysis demonstrated that TG003 was able to inhibit SR protein phosphorylation, suggesting that hypoxia-induced phosphorylation of SR proteins is partially dependent on CLKs (Fig 4.1D).
119
Figure 4.1
Figure 4.1: Hypoxia Induces the Expression of CLK1 and 3 in a HIF- Dependent Manner and Induces Phoshorylation of SR proteins. A) Western blot of CLK1-4 proteins in normoxia (Nx) and hypoxia (Hx) treated Hep3B cells. B) qRT-PCR to measure the mRNA expression of CLK1,3 and 4 in Nx and Hx Hep3B cells. One-way analysis of variance was performed for this and other studies in this paper unless otherwise stated. *, P= 0.05; **, P =0.01. Controls for statistical analysis are specified in each Figure. C) Western blot using a phospho-specific anti-SR protein antibody (1H4) to detect phosphorylated SR proteins in Nx and Hx Hep3B cells. D) Western blot for phosphorylated SR- proteins in Nx and Hx Hep3B cells treated with 50 µM of the CLK inhibitor, TG003. Actin was used as a loading control 120
TG003 Inhibits Hypoxia-induced Transcription and Splicing of HIF Targets
Next, we determined if CLKs regulate AS of HIF target genes by treating normoxic and hypoxic Hep3B cells with TG003. Consistent with our previous studies, RT-PCR showed that hypoxia induced the expression of CA9, ANGPTL4 and PDK1 isoforms (Chapter III) and that TG003 inhibited the hypoxic induction of these isoforms (Fig. 4.2A). qRT-PCR demonstrated that hypoxia induces the expression of CA9, ANGPTL4 and PDK1 full-length (FL) and exon skipping isoforms (ΔE) but preferentially induces the expression of the FL isoforms; promoting exon inclusion. Importantly, TG003 was able to inhibit hypoxia- regulated AS, for instance TG003 increased the CA9FL:ΔE89 ratio by 1.8 fold, but reduced the ANGPTL4FL:ΔE4 and the PDK1FL: ΔE4 ratios by 2 and 1.3 fold respectively (Fig 4.2 B-D). In addition, TG003 inhibited the expression of the FL,
ΔE isoforms and total transcripts for CA9, ANGPTL4 and PDK1 (Fig 4.2 B-D),
suggesting that CLKs are necessary for AS and transcription of HIF target genes,
consistent with the idea that pre-mRNA splicing is required for efficient
transcription (260, 261).
Next we wanted to determine if CLKs were specifically required for
transcription of HIF targets or if they regulate transcription of non-HIF targets as
well. qRT-PCR analysis in hypoxic and normoxic Hep3B treated with TG003
revealed that TG003 inhibited the expression of additional HIF target genes
under normoxia and hypoxia, including ADM, PAI1, LDHA, PGK1 and PTPRB
(Fig. 4.2 E) but did not significantly inhibit the expression of non-HIF targets, with
the exception of HIF2α (Fig. 4.2F). Consistent with this result, siRNA-mediated 121
Figure 4.2
Figure 4.2: CLKs Regulate Hypoxia-induced Transcription and Alternative Splicing of HIF Targets but Not of Non-HIF Targets. A) RT-PCR of CA9, ANGPTL4 and PDK1 isoforms in Nx and Hx Hep3B cells treated with DMSO (control) or TG003. qRT-PCR of CA9 isoforms and total transcripts (total) (B), ANGPTL4 isoforms and total transcripts (C), and PDK1 isoforms and total transcripts (D) in Nx and Hx Hep3B cells treated with DMSO or TG003. qRT- PCR of total transcripts for HIF target gene (E) and non-HIF target genes (F). 122 knockdown of CLK3 but not CLK4 modestly reduced the expression of HIF targets by 1.3-1.5 fold but did not influence the expression of non-HIF targets.
The knockdown efficiency for CLK1 was low and therefore CLK1’s constribution to gene expression could not be determined. Conversely, over-expression of
CLK3, but not a CLK3 kinase dead mutant, enhanced the normoxic expression of
HIF targets in Hep3B cells. However, hypoxic induction of these genes was not affected by CLK3 over-expression (data not shown), probably because CLK3 is already highly expressed during hypoxia. Interestingly, this data suggested that
CLKs are required specifically for the expression of HIF targets but are not required for the expression of non-HIF targets. Taken together, our data suggested that HIFs regulated AS and gene expression of HIF target genes by up-regulating the expression of CLK 3 during hypoxia.
SC35 and SRp40 Regulate AS of HIF Target Genes
As mentioned above, hypoxia increased the phosphorylation of several
SR proteins, including SRp75, SRp55, SRp40, and SC35. Therefore, we wanted to determine if SR proteins were important for regulating AS splicing of HIF target genes. To identify which SR proteins were potentially involved in regulating splicing of HIF target, we use the Human Splicing Finder program to identify putative SR protein binding sites within alternative exons and their corresponding flanking exons. Interestingly, we discovered that all of our genes contained predicted SC35 and SRp40 binding sites in the alternative and flanking exons;
SRp55 binding sites were not as strong as SC35 and SRp40 binding sites. In 123 addition, the Human Splicing Finder program does not predict SRp75 binding sites. Therefore, we chose to examine if SC35 and SRp40 could regulate hypoxia/HIF mediated AS. Western blots using an anti-SC35 antibody that detects unphosphorylated SC35 protein suggested that SC35 was slightly reduced by hypoxia. However using a phospho-specific anti-SC35 antibody, western blots confirmed that hypoxia enhanced SC35 phosphorylation in Hep3B cells (Fig. 4.3A). ARNT was used as a loading control. Human SC35 and
SRp40 mRNAs were knocked down in normoxic and hypoxic Hep3B cells using siRNAs. qRT-PCR confirmed that hypoxia reduced SC35 mRNA levels by 40% and that SC35 siRNAs reduced SC35 mRNA levels by 90% in normoxic and hypoxic Hep3B cells (Fig. 4.3B). In addition, hypoxia reduced SRp40 mRNAs level by 20% and SRp40 siRNAs reduced SRp40 mRNA levels by 75%. qRT-
PCR determined that SC35 but not SRp40 could regulate AS of CA9 since SC35 siRNAs did not significantly change the expression levels of CA9FL but increased the CA9ΔE89 levels by 2 fold and thus reduced the CA9FL:ΔE89 expression ratio by 2 fold under hypoxia compared to the control (Fig. 4.3C). In contrast, SRp40 siRNAs did not change the induction levels of CA9 isoforms and did not regulate AS. SC35 and SRp40 regulated AS of ANGPTL4 since SC35 and SRp40 siRNAs did not significantly change the ANGPTL4FL expression levels but increased the expression of ANGPTL4ΔE4. As a resut, SC35 and
SRp40 siRNAs reduced the ANGPTL4FL:ΔE4 ratio by 1.9 and 1.3 fold respectively (Fig. 4.3D, right). SC35 and SRp40 also regulated hypoxia-
124
Figure 4.3
Figure 4.3: SC35 and SRp40 Regulate Hypoxia-mediated AS of HIF Targets and SC35 Cooperates with HIF1 to Regulate AS of a PDK1 Minigene. A) Western blot for un-phosphorylated and phosphorylated SC35 protein in Nx and Hx Hep3B cells. ARNT was used as a loading control. B) qRT-PCR of SC35 and SRp40 mRNAs in Nx and Hx Hep3B treated with control (Ctrl) or SC35 or SRp40 siRNAs. qRT-PCR of CA9 (C), ANGPTL4 (D), and PDK1 (E) mRNA isoforms in Nx and Hx Hep3B cells treated with Ctrl, SC35 or SRp40 siRNAs. ANGPTL4E4 levels 1.9 and 1.3 fold respectively (Fig. 4.3D). F) qRT-PCR of minigene-specific PDK1 isoforms and total transcripts in 293T cells co- expressing a CA9PPDK1 minigene, empty vector (Ctrl), normoxia active HIF1αTM expression vector, Ctrl and/or SC35 expression vector, and control (siCtrl) or SC35 (siSC35) siRNAs. 125 mediated AS of PDK1. For instance, SC35 and SRp40 siRNAs reduced the hypoxic induction of PDK1FL 3.8 fold and 1.5 fold respectively (Fig. 4.3E). Also,
SC35 siRNAs increased the PDK1ΔE4 levels 1.5 fold and thus reduced the
PDK1FL:ΔE4 expression ratio by 5.8 fold compared to the hypoxia control. On the other hand, SRp40 did not change the PDK1ΔE4 levels under hypoxia and as a result the PDK1FL:ΔE4 expression ratio was reduced by 1.5 fold compared to the hypoxia control (Fig. 4.3E). In summary, this data suggests that SC35 can regulate AS of all HIF target genes examined in this study; however, SRp40 regulates AS in a gene specific manner.
SC35 Cooperates with HIF1α to Regulate AS of a PDK1 Minigene
To further validate SC35’s functional interaction with HIF1α in regulating
AS, a rescue experiment was performed in which Hep3B cells were treated with control (siCtrl) or SC35 (siSC35) siRNAs followed by co-transfection with a PDK1 minigene driven by the CA9 promoter (CA9PPDK1) and HIF1α alone or in conjunction with SC35. qRT-PCR showed that SC35 in conjunction with HIF1α further enhanced AS of the PDK1 minigene compared to HIF1α alone. HIF1α alone induced the expression of the FL and ΔE4 isoforms by 22.8 and 15.45 fold
(light grey bar). However, co-expression of HIF1α with SC35 induced the
PDK1FL and PDK1ΔE4 isoforms by 41.6 and 9.82 fold (Fig. 4.3F solid black bar). Thus, the PDK1FL:ΔE4 splicing ratio was enhanced by 2.9 fold when
HIF1α and SC35 were co-expressed, although HIF1α alone activated the reporter slightly more than co-expression with HIF1α and SC35 as determined by 126
measuring the expression of total PDK1 minigene transcripts (Fig. 4.3F). In
addition, siRNA-mediated knockdown of SC35 completely inhibited HIF1α’s
ability to enhance AS (Fig. 4.3F, diamond pattern) compared to HIF1α alone
(light grey bar). Interestingly, add back of SC35 reversed the inhibitory effects of
SC35 siRNAs (Fig. 4.3F, diagonal stripe) and restored HIF1α- induced AS,
definitively confirming a functional cooperation between SC35 and HIF1α in the
regulation of AS of a PDK1 minigene.
CLKs Promote Tumorigenesis of PRC3 Cells by Regulating AS of HIF
Target Genes
We determined that CLKs regulate HIF-mediated transcription and AS of
HIF target genes in Hep3B cells, however the role that CLKs play in promoting tumorigenesis is not known. Since CLKs specifically inhibit gene expression and splicing of HIF targets and inhibit gene expression of the HIF2α gene; we
hypothesized that inhibition of CLKs would effectively reduce HIF2α-mediated
tumorigenesis. To test this hypothesis, PRC3 cells stably expressing non-
targeting or HIF2α-targeting shRNAs were utilized since HIF2α is a major driver
of tumorigenesis in this cell line. PRC3 cells are a clear cell renal cell carcinoma
cell line that contain a mutation in the Von Hippel-Lindau tumor suppressor
protein and therefore express constitively active HIF2α protein and HIF2α target
genes. PRC3 cells expressing non-targeting control shRNAs or HIF2 α shRNAs
were treated with TG003 and assessed for clonogenic survival and cell migration
using colony forming and wound healing migration assays. Clonogenic survival 127 assays revealed that HIF2α knockdown cells exhibited reduced survival compared to control cells, which is consistent with previous reports (Fig. 4.4A)
(306). Additionally, TG003 inhibited the survival of control PRC3 cells in a dose dependent manner in comparision to DMSO treated controls. In contrast, TG003 treatment did not further inhibit the survival of HIF2α knockdown cells, suggesting that CLKs and HIFs function in the same pathway to promote cell survival.
Similar effects were observed for cell migration in PRC3 cells as well.
HIF2α knockdown cells demonstrated a lower rate of migration than PRC3 control cells (Fig. 4.4B). Additionally, PRC3 control cells exhibited decreased migration when treated with TG003 (40 µM) in comparison to DMSO treated cells. Conversely, TG003 treatment did not further reduce the migration rate of
HIF2α knockdown cells compared to DMSO treated HIF2α knockdown cells; suggesting that CLKs and HIF2α function in the same pathway to promote cell migration as well. In summary this data suggested that CLKs and HIFs promote tumorigenesis of PRC3 cells through overlapping pathways.
As mentioned above, CLKs can regulate alternative splicing and transcription of HIF target genes in Hep3B cell. Therefore, it is likely that CLKs promote tumorigenesis in PRC3 cells by regulating either alternative splicing or transcription or both. To address this, PRC3 cells were treated with DMSO
(control) or with TG003 and qRT-PCR was performed for serval HIF and non-HIF target genes. qRT-PCR analysis showed that TG003 did not significantly alter the expression of HIF (Fig. 4.4C) and non-HIF target genes (Fig. 4.4D) with the exception of PTPRB, suggesting that CLKs regulate splicing and transcription of 128
Figure 4.4
Figure 4.4: CLKs Regulate HIF2α-mediated Tumorigenesis in PRC3 Cell by Regulating AS. Clonogenic survival (A) or wound healing migration assays (B) in PRC3 cells treated with DMSO (control) or with TG003. qRT-PCR of HIF target gene mRNAs (C) and non-HIF target genes (D) in PRC3 cells treated with DMSO or TG003 (40 µM). E) RT-PCR of ANGPTL4, PAI1, and PTPRB mRNA isoforms in PRC3 cells treated with DMSO or TG003. B-Actin was used as a negative control. 129
HIF targets in this cell line. In contrast, RT-PCR analysis revealed that TG003 treatment alters the splicing of HIF target genes (Fig. 4.4E). For instance, TG003 treatment preferentially promoted the expression of shorter ANGPTL4 transcripts and increased the number of truncated transcripts expressed by the PAI1 gene
(Fig. 4.4E). RT-PCR also confirmed that the expression of PTPRB was reduced by TG003 treatment. Taken together, this suggested that CLKs promote tumorigenesis in PRC3 cells by regulating AS and transcription of HIF target genes.
Discussion
Although a few examples of hypoxia-dependent regulation of AS have been reported, the underlying mechanism was not known (276-279). Our data showed that hypoxia induces the expression of CDC-like kinases, CLK1, 3 and 4 and that hypoxic induction of CLK1 and 3 is HIF-dependent. Additionally, hypoxia-induced phosphorylation of SR protein splicing factors, SC35, SRp40,
SRp55, and SRp75, correlated with the hypoxic induction of CLKs; which are upstream regulators of SR protein phosphorylation (Fig. 1.8). Interestingly, we found that CLKs were also important for regulating AS and gene expression of
HIF target genes specifically, since a CLK inhibitor did not affect the expression of non-HIF targets. We determined that SC35 and SRp40 regulate hypoxia- induced AS of endogenous HIF target genes using siRNA knockdown and found that SC35 cooperates with HIFs to regulate AS of a PDK1 minigene. Taken together, we suggest a model of hypoxia-mediated alternative splicing in which 130
hypoxia, through HIF proteins, induces the expression of CDC-like kinases
(CLK1, 3 and 4). In turn, CLKs phosphorylate SC35 and SRp40 SR proteins during hypoxia which regulate alternative splicing of HIF target genes.
More importantly, we demonstrated for the first time that CLKs regulate
HIF-dependent tumorigenesis in PRC3 cells by regulating alternative splicing of
HIF target genes. This finding is very exciting because it will potentially give
researchers and clinicians the ability to treat cancer by targeting the HIF pathway
which is notoriously difficult to target using current treatments.
Acknowledgments
This work was supported by grants from the National Cancer Institute
(RO1CA134687, Hu) and Cancer League of Colorado (Hu). Johnny Sena was
supported by “Research Supplemental to Promote Diversity” (RO1CA134687-S3 and RO1CA134687-S4).
131
Figure 4.5
Figure 4.5: Model of Hypoxia/HIF-mediated Alternative Splicing and Tumorigenesis. In this smiplified model we propose that hypoxia stabilizes hypoxia inducible factors (HIFs). HIFs activate a large number of target genes which promote tumor progression, including CLK1 and CLK3. CLKs increase the phoshorylation and activation of SR protein splicing factors. Subsequently, increased SR protein activation enhances splicing which consequently enhances transcription of HIF targets allowing for maximal expression of HIF target genes and promotion of tumorigenesis.
132
CHAPTER V
CONCLUSIONS AND PERSPECTIVES
Summary and Conclusions
Oxygen is a basic requirement for life necessary for the biochemical reactions that generate ATP within the human body which fuels the processes of embryogenesis, wound healing, and stem cell maintenance. Since oxygen is such an important physiological factor, the molecular pathways controlled by it are under tight constraint. When these control mechanisms do not function properly in response to hypoxia, or operate outside of the correct physiological context, they can promote an array of diseases including cancer.
The role of HIFs in promoting tumorigenesis is well established and is generally attributed to transcriptional activation of HIF target genes. Recently, it was found that 90% of human genes are alternatively spliced producing multiple isoforms from a single gene which encode proteins with distinct or opposing functions (328). Thus, it is clear that assessing isoform expression is more informative than quantifying gene expression alone, allowing one to fully appreciate the functional consequence of gene expression programs during cancer progression.
In this thesis, I provide evidence suggesting that HIFs also promote tumorigenesis by regulating constitutive and alternative splicing. By describing a molecular mechanism by which HIFs regulate alternative splicing of HIF target genes, it may be possible to develop inhibitors of splicing activity that allow for 133
targeting of the HIF pathway. Selective inhibition of tumorigenic HIF target gene
isoforms while maintaining the expression of anti-tumorigenic gene isoforms
through the modulation of splicing in a given tumor may provide more effective
inhibition of tumorigenesis with fewer undesirable off-target effects in cancer
patients.
In my research, I have 1) determined that hypoxia regulates pre-mRNA
splicing of a subset of HIF and non-HIF target genes (Chapter II and III) (339), 2) demonstrated that HIFs are necessary and sufficient to regulate hypoxia- mediated splicing of a subset of HIF targets (Chapter II and III), 3) described
molecular signaling pathways by which HIFs regulate alternative splicing of HIF
target genes (Chapter IV), and 4) determined that HIFs regulate tumorigenesis,
in part, by regulating alternative splicing of HIF targets (Chapter IV). I will
summarize the main findings of my research in this chapter and discuss their
significance in the field.
Hypoxia, HIFs and Pre-mRNA Splicing
We have reported that hypoxia regulates pre-mRNA splicing of
Adrenomedullin (ADM), a well established HIF target gene (302, 307). We found
that normoxic cancer cells express fully-spliced (FL) and several intron-retaining
RNA transcripts including I1, I2, I3, and I1-3 isoforms (Fig. 2.1). Furthermore, the
FL and I1-3 isoforms were the predominantly expressed isoforms in normoxic
cells (Figs 2.1 and 2.4). Importantly, we found that intron-retaining ADM isoforms
were expressed in normal human tissues and normal HUVEC cells, suggesting 134
that expression of intron-retaining ADM isoforms is not a consequence of
abnormal splicing, but a normal control mechanism of gene expression.
Martinez et al reported that the ADM gene expresses FL and I3-
containing isoforms in normoxic cell lines (298). It was reported that the I3
isoform was preferentially induced by hypoxia in some cancer cell lines (298).
However, their findings were puzzling because the ADM FL RNA codes for a
more functionally important tumor promoting AM peptide than the ADM I3 RNA
which codes for only the PAMP peptide. Thus, we investigated ADM isoform
expression in hypoxic cells. Interestingly, we found that hypoxia preferentially induces the levels of ADM FL over intron-containing ADM transcripts. It is not
clear if this discrepancy is due to different cell lines used or other differences.
However, our finding are in agreement with previous studies suggesting that the
AM peptide plays important roles in tumorigenesis by inducing angiogenesis,
proliferation and by reducing cellular apoptosis and we would expect that cancer
cells would preferentially express gene isoforms that benefit their growth and
survival.
Using the endogenous ADM gene (Figs. 2.3 and 2.4) as well as ADM
splicing reporters (Fig. 2.5), we determined that HIF activity, but not hypoxia per
se, is necessary and sufficient to regulate hypoxia induced splicing of ADM pre-
mRNA.
In addition, using exon arrays, we discovered that hypoxia also regulates
alternative splicing of HIF and non-HIF target genes on a global scale (Chapter
III). Generally speaking, we found that hypoxia promotes exon inclusion of 135 hypoxia induced genes but promotes exon skipping of hypoxia reduced genes, demonstrating a functional coupling of transcription and alternative splicing.
Further analysis of HIF target genes demonstrated that HIFs were necessary and sufficient to promote alternative splicing of endogenous HIF target genes. Using
PDK1 minigene splicing reporters we confirmed that HIFs were sufficient to promote splicing of the reporter even during normoxia. Importantly, we determined that activation of endogenous HIF target genes greatly contributed to alternative splicing of the PDK1 minigene.
These findings are particularly important because they illuminate a novel function of HIFs in the regulation of gene expression, suggesting that HIFs, in addition to regulating gene transcription, also regulate RNA splicing. Moreover, it is well known that targeting HIF transcription factors for disease treatment is very difficult. However, given our findings, we propose that the HIF pathway may be targeted by modulating HIF-mediated splicing. Targeting RNA splicing for disease treatment is a growing field and serval different strategies currently exist, including drugs that inhibit phosphorylation of splicing factor kinases and splice site blocking antisense oligonucleotides.
Hypoxic Induction of CLKs Regulates AS of HIF target Genes
Although several examples of hypoxia-mediated AS splicing have been reported, a molecular mechanism describing how hypoxia regulates AS had not been reported until now. To reiterate, we determined that HIFs regulate hypoxia induced splicing by regulating the expression of HIF target genes which in turn 136 regulate alternative splicing. In subsequent studies (Chapter IV), we identified several potential splicing regulators that were induced by hypoxia in our exon array. We became particularly interested in the CDC-like kinase family members since three out of the four members (CLK1, 3 and 4 but not 2) were induced by hypoxia and since their role in RNA splicing was well established.
Consistent with our exon array data, western blot analysis confirmed that
CLK1, 3 and 4 but not CLK2 proteins were induced by hypoxia. qRT-PCR analysis demonstrated that CLK1, 3 and 4 were induced by hypoxia; and siRNA- mediated knockdown of ARNT revealed that CLK1 and 3 were HIF target genes.
Although CLK1 and 3 are HIF target genes, we did not determine if they are direct or indirect HIF targets. However, promoter analysis of CLKs using
JASPAR (http://jaspar.genereg.net/) predicted potential HIF/ARNT binding motifs
(data not shown). Moreover, ChIP-seq. and promoter microarray experiments in
MCF-7 breast cancer cells treated with a hypoxia mimic (DMOG) demonstrated that HIF1α binds to the CLK3 promoter (340), suggesting that CLK3 is a direct
HIF target gene.
Consistent with hypoxic induction of CLKs, western blot analysis showed that phosphorylation of several SR protein were induced by hypoxia and that a
CLK inhibitor, TG003, reduced phosphorylation of SR proteins. Subsequent experiments using TG003 and CLK3 siRNAs demonstrated that CLK3 regulates alternative splicing and transcription of HIF target genes. Interestingly, transcription inhibition by TG003 or CLK3 siRNAs was specific to HIF target genes since TG003 or CLK3 siRNAs did not affect the expression of non-HIF 137 targets. SiRNA-mediated knockdown of SR proteins, SC35 and SRp40, revealed that these splicing factors regulated AS of HIF target genes during hypoxia.
Taken together, these data suggested that HIFs regulate alternative splicing and gene expression during hypoxia by inducing CLKs which phosphorylate and activate SR protein splicing factors which ultimately modulate splicing.
Importantly, we determined that CLKs promote HIF-dependent tumor survival and migration in PRC3 cells, a renal cell carcinoma, and determined that
CLKs regulate tumorigenesis in this cell line by regulating AS and transcription of
HIF target genes. To our knowledge this is the first report to suggest that CLKs possess pro-tumorigenic properties by regulating the HIF signaling pathway.
This is an exciting finding that identifies a potential molecular signaling pathway that can be exploited for treatment of HIF-dependent cancers.
Perspectives and Future Directions
The work presented here has provided an answer to a persisting question in the field of hypoxia-mediated alternative splicing regulation, suggesting that
HIFs and not hypoxic stress per se regulates RNA splicing of HIF target genes.
Although our work has identified one molecular signaling pathway that regulates hypoxia/ HIF-mediated splicing, many questions are still unanswered.
Future research in the field of hypoxia-mediated splicing will undoubtedly uncover additional components involved in the regulation of hypoxia/HIF- mediated splicing. These will likely include additional SR protein splicing factors and RNA binding proteins, RNA helicases, spliceosomal snRNAs, and 138 transcription factors and co-transcription factors. For example, we discovered that hypoxia induced the expression of members of the CDC-like kinase family,
CLK1, CLK3, and CLK4; small non-coding RNAs of the spliceosome, U1, U2, U4, and U6; RNA helicases, DDX41, DHX32, and DDX59; the SR protein, SRSF12;
SMAD3, a TGF-β responsive transcription factor; the RNA binding protein
RBPMS; and NOL3, an inhibitor of apoptosis (Table 5.1). However, with the exception of CLKs, it is not known if these factors play a role in the regulation of hypoxia-mediated splicing. Future studies will need to address this question to allow for a more complete understanding of hypoxia-mediated splicing.
Moreover, if these splicing regulators are confirmed to control hypoxia/HIF- mediated splicing they may provide additional targets by which to inhibit hypoxia induced gene expression programs during tumorigenesis or other hypoxia related diseases. I suspect that many of these factors will regulate hypoxia and HIF- mediated splicing and that these factors work together to ultimately determine hypoxia-mediated splicing outcomes, but that remains to be determined. Below,
I will briefly discuss the function of these factors (except CLKs) in regulating alternative splicing and speculate as to how they might regulate hypoxia- mediated splicing.
Small Nucleolar RNAs
As discussed in Chapter I, snRNAs are part of the core spliceosome machinery. snRNAs are found in snRNP complexes and are responsible for targeting and binding of snRNPs to nascent pre-mRNAs during spliceosomal 139
assembly and the relative abundance of these factors influence splicing
outcomes or efficiency (341). Additionally, a recent study has demonstrated that pre-mRNAs compete for the splicing machinery and that this competition drives
global regulation of splicing (321). Based on this finding, it has been suggested that competition for or titration of general splicing factors by different sets of pre-
mRNAs due to massive changes in transcriptional and epigenetic programs can
lead to changes in the splicing of other transcripts (342, 343).
Interestingly, our exon array data suggests that U1, U2, U4 and U6 snRNAs are induced by hypoxia (Table 5.1). Therefore, it is entirely possible that
hypoxia and/or HIFs induce these snRNAs during hypoxia to keep up with the
large number of genes that are induced by hypoxia (Fig. 3.1). Since RNA
splicing is necessary for efficient transcription and vise versa, an increase in
transcription might create a deficit of freely available splicing factors due to
competition between different mRNAs. However, hypoxia may prevent or
compensate for this deficit by up-regulating these factors, allowing hypoxia-
induced genes to be efficiently transcribed and spliced during this time. Although
plausible, it is not currently known if hypoxia induced transcription creates an
increased demand for splicing factors.
Interestingly, the splicing factor competition hypothesis, could potentially
explain why we primarily observed exon skipping in hypoxia-reduced genes (Fig
3.1). For instance, during hypoxia a large number of genes are induced by
hypoxia, the hypoxia-induced mRNAs of these genes likely compete with other
subsets of mRNAs for splicing factor components, reducing the availability of 140 splicing factors. As a result, due to a lack of splicing factors, hypoxia reduced genes might undergo exon skipping or intron retention. Additionally, since splicing is important for efficient transcription, a lack of available splicing factors may further inhibit the expression of hypoxia reduced genes.
Alternatively, the reason that hypoxia reduced gene may undergo exon skipping could be explained by the recruitment model of alternative splicing
(Fig.1.9). In this context, reduced transcription and presumably, less recruitment of RNA pol II to hypoxia reduced genes would result in less recruitment of splicing factors to the nascently transcribed pre-mRNAs. Since splicing factors such as SR proteins regulate exon inclusion, less recruitment of splicing factors might result in exon skipping by default, since the splice sites of alternative exons are weak.
RNA Helicases
At least eight RNA helicases, DDX46 (a.k.a, prp5), DDX39B (sub2),
DDX23 (prp28), DHX16 (prp2), DHX38 (prp16), DHX8 (prp22), DHX15 (prp43), and SNRNP200 ( Brr2) are required for proper function of the spliceosome and pre-mRNA splicing in all eukaryotes (344). Generally speaking, these helicases are thought to break RNA secondary structures through ATP hydrolysis, which allows U snRNPs to bind to nascent RNAs during each step of spliceosome assembly and catalysis or to disassociate snRNPs from RNA during spliceosome progression. Because of their protein similarity, other RNA helicase are also predicted to regulate RNA splicing as well. Gene expression analysis of our 141
exon array data identified that DDX41, DDX59, DHX32 are induced by hypoxia;
furthermore, gene ontology (GO) analysis suggested that these helicases
potentially regulate RNA splicing (Table 5.1). For instance, DDX41 has been found to associate with the spliceosome, suggesting a possible role in RNA
splicing (345). However, it has not been determined if DDX41 plays a direct role
in splicing. In yeast, DHX32 was found primarily in the nucleolus and was
necessary for viability; however, DHX32 depletion did not affect pre-mRNA
splicing but was necessary for 18S rRNA synthesis (346). Together with our
findings, this may suggests that hypoxia might also regulate translation through
upregulation of DHX32. Unfortunately, the biological role of DDX59 has not yet
been examined, therefore we cannot speculate about its role in the hypoxia
response.
Other SR Proteins
As reported in Chapter IV, we found that the SC35 and SRp40 SR
proteins regulate splicing of HIF target genes. We also determined that SRp55
and SRp75 were phoshorylated and activated during hypoxia, however it was not
determined if these SR proteins also regulate splicing of HIF target genes.
Future studies should address if SRp40 and SRp75 also regulate hypoxia-
mediated splicing, we predict that these SR proteins will also regulate hypoxia-
mediated splicing by promoting exon inclusion of HIF target genes (347)
Interestingly, gene expression analysis of our exon array data revealed
that SRSF12, a non-classical SR protein, was induced by hypoxia (Table 5.1). 142
SRSF12 was the only SR protein that was induced by hypoxia at the gene
expression level. Interestingly, SRSF12 was shown to antagonize the function of classical SR proteins, SC35 and SRp40, which generally promote exon inclusion
(347). Intriguingly, it is possible that hypoxic induction of SRSF12 might explain how hypoxia promotes exon skipping of a large number of genes (Fig. 2.1B). In other words, SRSF12 might antagonize the function of classical SR proteins on genes that undergo hypoxia-induced exon skipping.
SMAD3 and RBPMS
Transforming growth factor-β (TGF-β) is a ubiquitously expressed cytokine that regulates cell growth, differentiation, matrix production, apoptosis and development (348-350). TGF-β mediated signaling has been shown to phosphorylate Smad proteins, promoting translocation of Smads into the nucleus where they activate or repress transcription (350). TGF-β signaling, through
Smad3/Samd4, decreases the expression of the epithelial splicing regulatory proteins (ESRP1 and 2), which leads to a differential splicing program during epithelial to mesenchymal transition (EMT) (351, 352). Since hypoxia can regulate EMT (Chapter I), it is tempting to speculate that hypoxia may regulate splicing during EMT by inducing the expression of Smad3. Consistent with this idea, ESRP1 and ESRP2 expression is reduced by hypoxia in our exon array data (Table 5.1). Further, support for this idea stems from the observation that
Rbpms, an RNA binding protein, enhances nuclear import of Smads following phoshorylation by TGF-β. Interestingly, the RBPMS gene is also induced by 143 hypoxia in our data (Fig. 5.1). Taken together, one could imagine that hypoxic induction of SMAD3 and RBPMS would lead to hypoxia-mediated repression of
ESRP1 and ESRP2 leading to alterations in alternative splicing during hypoxia.
NOL3
Nucleolar protein 3 (apoptosis repressor with CARD domain) (NOL3) is a direct HIF1α target gene that encodes an anti-apoptotic protein that has been shown to down-regulate the enzyme activities of caspase 2, caspase 8 and p53 and regulates the intrinsic and extrinsic cell death pathways (353). Interestingly,
NOL3 was shown to interact with the SR protein SRp30c/SFRS9 and overexpression of NOL3 regulated alternative splicing of preprotachykinin and
SRp20 reporter genes (354). Consistent with previous reports, we found that
NOL3 was induced by hypoxia in our exon array data (Table 5.1). Since NOL3 has been shown to regulate splicing previously, it is very likely that NOL3, through its interaction with SRp30c, might regulate splicing during hypoxia in a
HIF-dependent manner.
Although the factors mentioned above are induced by hypoxia, to fully understand hypoxia-mediated splicing, these factors must be tested for their invovelment in hypoxia-mediated splicing. If future studies confirm their involvement in splicing, the next questions to answer are: What factors regulate their induction during hypoxia? Is hypoxic induction of these factors HIF dependent or dependent on other hypoxia activated transcription factors?
144
Table 5.1
P-Value Fold- Hypoxia Hypoxia vs Ensembl Gene ID Gene Name Symbol vs Normoxia Normoxia GO-Biological Process
ENSG00000013441 CDC-like kinase 1 CLK1 1.25 0.09 regulation of RNA splicing
ENSG00000176444 CDC-like kinase 2 CLK2 -1.01 0.90 regulation of RNA splicing
ENSG00000179335 CDC-like kinase 3 CLK3 1.59 0.02 regulation of RNA splicing
ENSG00000113240 CDC-like kinase 4 CLK4 1.43 0.06 regulation of RNA splicing
ENSG00000202125 U1 spliceosomal RNA U1 1.77 0.01 regulation of RNA splicing
ENSG00000252604 U2 spliceosomal RNA U2 1.57 0.02 regulation of RNA splicing
ENSG00000202181 U4 spliceosomal RNA U4 1.60 0.02 regulation of RNA splicing ENSG00000222431 U6 spliceosomal ( p RNA ) U6 2.68 9.4E-05 regulation of RNA splicing ENSG00000089876 box polypeptide 32 DHX32 1.46 0.01 helicase activity helicase activity// RNA DEAD (Asp-Glu-Ala-Asp) splicing // RNA ENSG00000183258 box polypeptide 41 DDX41 2.24 4.9E-05 processing DEAD (Asp-Glu-Ala-Asp) // helicase activity // RNA ENSG00000118197 box polypeptide 59 DDX59 1.45 0.01 binding // Serine/arginine-rich regulation of alternative ENSG00000154548 splicing factor 12 SRSF12 1.54 0.01 nuclear mRNA splicing positive regulation of SMAD protein import into nucleus // regulation of RNA binding protein with transcription // RNA ENSG00000157110 multiple splicing RBPMS 1.47 0.01 processing
regulation of transcription factor import into nucleus // transforming growth factor beta receptor ENSG00000166949 SMAD family member 3 SMAD3 2.01 1.4E-04 signaling pathway Epithelial splicing ENSG00000104413 regulatory protein 1 ESRP1 -1.41 0.01 regulation of RNA splicing Epithelial splicing ENSG00000103067 regulatory protein 2 ESRP2 -1.35 0.02 regulation of RNA splicing RNA binding // protein ENSG00000140939 Nucleolar protein 3 NOL3 1.50 1.9E-03 binding
Table 5.1: Gene Expression Analysis of Exon Array Data for Genes Implacated in Hypoxia-mediated Alternative Splicing Regulation. Ensembl gene IDs, gene names, and gene symbols are provided. Expression data (Fold Change Hx Vs Nx) and gene ontology analysis (GO-Biological Process) generated by Altanalyze are also provided.
145
Future studies should also address if CLKs also regulate tumorigenesis of other types of cancer using in vitro assays described in Chapter IV. Furthermore, the role of CLKs in promoting HIF-dependent tumorigenesis must be determined in vivo, perhaps using nude mice zenografts treated with CLK inhibitors.
However, the inhibitor used in our studies would not be recommended because the doses used are in the micromolar range and are much too high for use in vivo.
Splicing Targeted Therapies
Hypoxia is more severe in solid tumors than in most normal tissues.
Therefore, development of therapeutic compounds targeting the hypoxic response have been pursued with the hopes of developing therapies that specifically target hypoxic tumor cells. One widely utilized anti-angiogenic chemotherapeutic strategy is to inhibit the function of the VEGFA protein, a HIF target, to inhibit tumor angiogenesis and thereby restricting growth and metastasis of brain, breast, kidney, and lung tumors (355-358). For example,
Bevacizumab is a humanized monoclonal antibody that inhibits both the pro- angiogenic VEGFAxxxa and the anti-angiogenic VEGFAxxxb isoforms by blocking their common receptor binding site kinase domain (359). However,
Bevacizumab has a low response rate, high rate of resistance and adverse side effects (360), which may be caused by non-specific targeting, resulting from the non-specificity for the pro- and anti-angiogenic isoforms (361). Therefore, a more effective treatment for cancer might be to specifically reduce the pro-angiogenic 146
VEGFAxxxa isoforms, while maintaining or enhancing the expression of the anti-
angiogenic VEGFAxxxb isoforms. Anti-angiogenic, VEGFAxxxb, isoforms are expressed in most normal tissues and are down-regulated in cancers, especially
in renal cell carcinomas (362). Consistent with this, we only detected the pro- angiogenic isoforms in Hep3B cells, a hepatocellular carcinoma cell line (data not shown). Using drugs that inhibit SR protein phosphorylation such as TG003 or by using splice-switching oligonucleotides which modulate splicing by blocking
splice sites and/or SR or hnRNP protein binding sites, it may be possible to shift the splicing patterns of VEGFA, selectively reducing the pro-angiogenic isoforms while maintaining or enhancing the anti-angiogenic isoforms. Interestingly, findings reported in this thesis, provide a foundation for targeting the HIF/hypoxia response in cancer biology through modulation of alternative splicing, where it has proven difficult to target HIF transcription factors or HIF target genes for treatment of cancer.
Interestingly, both topoisomerase I (camptothecin and homocamptothecin derivatives) and II (VP16, amsacrine, doxorubicin, mitoxantrone) inhibitors have been shown to alter alternative splicing (363). AQ4N (a di-N-oxide analog of mitoxantrone), a topoisomerase II inhibitor, is an example of a drug designed to exhibit hypoxia-dependent cytotoxicity with the goal of specifically targeting hypoxic cells in tumors. AQ4N exists in an inactive state in the presence of
oxygen but is converted to its active state during oxygen deprivation. AQ4N is
converted to the AQ4 compound under hypoxia which inhibits the topoisomerase
II enzyme and results in DNA damage to hypoxic cells (364) and likely alters 147 alternative splicing. AQ4N has limited efficacy when used alone, but can greatly increase the effectiveness of radiation and other anticancer drug therapies including cisplatin and cyclophosphamide in treatment of solid tumors (365-368).
Interestingly, cisplatin and cyclophamide were shown to regulate alternative splicing of Bcl-x in Hela cells, ehancing the expression of the pro-apoptotic Bcl-xs isoform, suggesting that these drugs may lead to cancer cell death through modulaton of alternative splicing (369). Since topoisomerases, cisplatin, and cyclophamide regulate alternative splicing it is tempting to speculate that the cooperative affects observed with AQ4 and cisplatin and cyclophamide are because of increased modulation of alternative splicing.
In summary, my thesis research has provided a mechanism explaining how HIFs regulate hypoxia-mediated RNA splicing. Importantly, these findings have addressed a major question in the field concerning the mechanism of hypoxia regulated splicing. Through identification of splicing factors required for activation and splicing of HIF target genes, my work has provided additional mechanistic information regarding HIF target gene activation and splicing that may be used as a molecular basis for pharmacologic targeting of the HIF pathway to inhibit HIF- dependent tumorigenesis. 148
CHAPTER VI
MATERIALS AND METHODS6
Cell Culture
Hep3B cells were cultured in MEM/EBSS (Hyclone) containing 10% FBS,
2 mM L-glutamine, 1 mM sodium pyruvate, 100,000 units/L Penicillin/
Streptomycin, 1.5 g/L sodium bicarbonate, and 1X non-essential amino acids
(NEAA). Hela cells were grown in high-glucose DMEM (Hyclone) with 10% FBS,
2 mM L-glutamine, 100,000 units/L Penicillin/Streptomycin and 1X NEAA. UM-
SCC-22B, HEK293T, RCC4, RCC4T and PRC3 cells were grown in high-glucose
Dulbecco modified Eagle medium (DMEM: Hyclone) with 10% FBS, 2 mM L-
glutamine, 100,000 units/L Penicillin/Streptomycin and 1X NEAA. HK2 cells were
grown in keratinocyte serum free medium (K-SFM) (GIBCO) with 0.05 mg/ml bovine pituitary extract (BPE), 5 ng/ml recombinant epidermal growth factor
(EGF), 2 mM L-glutamine, 100,000 units/L Penicillin/Streptomycin and 1X NEAA.
HUVEC cells were grown in F-12K medium (ATCC) containing 10% FBS with 0.1 mg/ml heparin, 0.04 mg/ml endothelial cell growth supplement (ECGS), 2 mM L-
6 Portions of this chapter appear in the following manuscripts and have been reprinted with required permission:
Sena JA, Wang L, Pawlus MR, Hu CJ. HIFs Enhance the Transcriptional Activation and Splicing of Adrenomedullin. Mol Cancer Res. 2014 Feb 12. PubMed PMID: 24523299
Sena JA, Wang L, Pawlus MR, Hu CJ. Hypoxia regulates alternative splicing of HIF and non-HIF target genes. Mol Cancer Res. Under Review 2014.
149
glutamine, 100,000 units/L Penicillin/Streptomycin and 1X NEAA. SK-N-MC cells were grown in RPMI (1640: Hyclone) with 10% FBS, 2 mM L-glutamine, 100,000 units/L Penicillin/Streptomycin and 1X NEAA. Prior to hypoxia treatment, 25 mM
HEPES was added to growth media and cells were incubated under normoxia
(21% O2) or hypoxia (1.5% O2) for 12-16 hrs. TG003 Treatment: For inhibition of
CLKs, Hep3B cell were cultured as described above, however, 30 minutes prior
to hypoxia treatment Hep3B cells were treated with 5, 10, 20, 40, 50 or 100 μM
(final concentration) of TG003 (sigma T5575-5MG) in DMSO. After 30 minutes of
TG003 treatment, HEPES (25 mM final concentration) was added to growth
media containing TG003 and cells were incubated under normoxia (21% O2) or
hypoxia (1.5% O2) for 16 hrs.
All parental cell lines were purchased from ATCC. After completing the
experiments, the parental (Hep3B, HEK293T, RCC4, HK2, HUVEC and PRC3)
and modified cell lines (RCC4T) were authenticated by DNA profiling or
“fingerprinting” by the University of Colorado DNA Sequencing and Analysis
Core.
Knockdown of Endogenous mRNA Using Small Interfering RNAs (siRNAs)
Control (Qiagen,1027281) or siRNAs specific for human ARNT (Qiagen,
equal mix of SI00304220, SI00304234, and SI03020913), HIF1α (Qiagen,
SI02664053) or HIF2α (Qiagen, SI00380212 ), CLK1 (Thermo Scientific
Dharmacon, J-004800-08), CLK3 (Thermo Scientific Dharmacon, J-004802-06),
CLK4 (Thermo Scientific Dharmacon, J-004803-08), SC35 (SRSF2) (Thermo 150
Scientific Dharmacon, L-019711-00), or SRp40 (SRSF5) (Thermo Scientific
Dharmacon, L-007279-01) mRNAs were transfected into Hep3B cells at 50% confluency using HiPerFect Transfection Reagent (Qiagen) according to the manufacturer’s protocol. 32hr post-transfection, cells were cultured at 21% or
1.5% O2 for 12-16 hr and then collected to prepare mRNA or protein for analysis.
Plasmid Constructs and Viral Transduction
The CA9P/ADM, PAI1P/ADM, and 2HRE/ADM splicing reporters were in the pcDNA3.1 (+) vector. The CA9 and PAI1 promoters were cloned from
CA9P/Luc and PAI1P/Luc (306, 370) using Advantage GC cDNA polymerase
(Clontech) and then inserted into the pcDNA3.1 plasmid, replacing the CMV promoter. 2 copies of HRE (hypoxia response element/ HIF binding sites) from the PAI1 promoter were added upstream of SV40 minimal promoter in the pGL3/Luc vector to produced the 2HRE/Luc construct, from which the
2HRE/SV40 was PCR amplified and replaced the CMV promoter in the pcDNA3.1 vector. Next, the ADM gene (full-length gene) from exon 1 to exon 4 was amplified from human genomic DNA and inserted downstream of the CA9,
PA11 and 2HRE promoters to yield the CA9P/ADM, PAI1P/ADM or 2HRE/ADM splicing reporter constructs. The CA9Pm2HRE/ADM, PAI1PmHRE/ADM, and
ΔHRE/ADM constructs were synthesized from the CA9P/ADM, PAI1P/ADM, and the 2HRE/ADM constructs using Pfu Ultra II polymerase (Invitrogen)-mediated mutagenesis PCR, in which the HREs were mutated or deleted from the constructs. The G5P/ADM reporter was generated by replacing the CA9 151
promoter in CA9P/ADM with the 5 copies of Gal4 DNA binding elements from
G5/Luc. The HIF1αTM-Flag, HIF2αTM-Flag, HIF1αDBD-Flag,
HIF1αDBD/VP16TAD-Flag, and USF expression plasmids were described previously (48, 306). The HIF1αDBD/E2F1TAD-Flag construct was synthesized by adding the E2F transcription factor 1 transactivation domain from the
PHKgalE2F 380-437 plasmid (gift from David Bentley) (371) to the HIF1αDBD construct using PCR. The Gal4DBD fusion constructs were synthesized by replacing the HIF1αDBD with the Gal4 DNA binding domain. The SC35-Flag expression plasmid was generated by replacing the HIF2αTM gene, described previously (48, 306), with the SC35 coding region that was PCR amplified from human cDNA using Advantage GC cDNA polymerase (Clontech). The
CA9P/PDK1, PAI1P/PDK1, and PAI1PmHRE/PDK1 constructs were generated
by replacing the ADM gene with PDK1 minigene that was PCR-amplified from human genomic DNA and contained exon 3 to exon 5 including introns 3 and 4.
The HIF1αTM-Flag, HIF2αTM-Flag, HIF1αDBD-Flag, HIF1αDBD/VP16-Flag,
HIF1αDBD/E2F-Flag, and USF expression plasmids were described previously
(48, 306, 339). The G5/PDK1 minigene was generated by replacing the ADM gene with PDK1 minigene in the G5/ADM construct that was described previously (339).
The HIF1α and HIF2α lentiviral constructs were synthesized by PCR cloning the HIF1α TM (triple mutation) and the HIF2α TM CDS, from the plasmids described above, into the pLEX-MCS vector containing a C-terminal 2X
Flag tag (Thermo Scientific). The GFP lentiviral construct was synthesized by 152
cloning green fluorescent protein cDNA (CDS) into the pLEX-MCS vector containing a C-terminal 2X Flag tag. The GFP, HIF1αTM, or HIF2αTM plasmid
were co-transfected with a psPAX2 packaging vector (Addgene) and a pMD2.G
envelope vector (Addgene) into HEK293T cells, grown in 10 cm dishes (40%
confluent), using TransIT LT-1 transfection reagent (Mirus). 24 hours after
transfection, transfection media was replaced with complete Hep3B media. The
following day, viral media was filtered and viral particles were concentrated using
high-speed ultracentrifugation at 26,000 X G. The viral pellet was re-suspended
in 1mL of PBS and added to Hep3B cells (30% confluency) for 6 hours. The
following day, a second round of viral PBS was added to Hep3B cells. 24 hours
after the second viral transduction, Hep3B cells were harvested for RNA and
protein preparation.
RNA Stability Assays
Hep3B cells were grown to ~50% confluency in 6-well plates and placed
under hypoxia for 16 hours. After which, the cells were treated with 2 µg/ml of
actinomycin D to inhibit de novo transcription and placed back under normoxia or
hypoxia for 2, 4 or 8 hours. Following treatment, cells were collected for RNA
isolation using SurePrep Nuclear or Cytoplasmic RNA Purification Kit (Fisher
BioReagents, BP2805-25). cDNA synthesis was performed and full-length and
exon skipping isoform mRNA stability was measured for ADM, CA9 and
ANGPTL4 using qRT-PCR.
153
ADM and PDK1 Splicing Reporter Assays
ADM splicing reporter assays were performed in Hep3B cells using
Lipofectamine Reagent (invitrogen 18324-012) and PLUS Reagent (invitrogen
11514-015) to transfect plasmid DNA. Typically, 2 X105 cells/per well in 6-well
plates were co-transfected with 0.2 µg of splicing report construct and 1.8 µg of
transcription activator. 48 hours after transfection, cells were collected for mRNA
and protein analysis. PDK1 splicing reporter assays were performed in
HEK293T cells using Lipofectamine Reagent. HEK293T cells were grown to
~30% confluency in 6-well plates and co-transfected with 200 ng of splicing
reporters (CA9P/PDK1, PAI1P/PDK1, PAI1PmHRE/PDK1, or G5/PDK1) and 1.8
µg of His (empty vector control), HIF1αTM, HIF2αTM, USF2, HIF1αDBD,
HIF1αDBD/E2F1, or HIF1DBD/VP16, expression constructs or Gal4/HIF1,
Gal4/E2F1 or Gal4/VP16 expression constructs with G5/ADM and G5/PDK1
splicing reporters. 48 hours after transfection, cells were collected for RNA or
protein preparation. Results were the average of at least three experiments.
Protein Analysis
Whole cell lysates were prepared and quantified for protein concentration.
Western blot analysis was performed using standard protocols with the following
primary antibodies: anti-ADM (C20) pAb (SC-16496, Santa Cruz), anti-Flag mAb
(F3165, Sigma), anti-HIF1α mAb (610959, BD Bioscience), anti-HIF2α mAb
(D9E3, Cell Signaling), anti-ARNT mAb (NB 100-124, Novus Biological), anti-
Actin pAb (SC-1616, Santa Cruz), anti-Gal4DBD pAb (SC-577, Santa Cruz), anti- 154
ANGPTL4 pAb (H-200) (SC-66806, Santa Cruz), anti-CAIX pAb (H-120) (SC-
25599, Santa Cruz), anti-CLK1 pAb (H00001195-B01P, Novus Biological), anti-
CLK2 pAb (H00001196-D01P, Novus Biological), anti-CLK3 pAb (H00001198-
B01, Novus Biological), anti-CLK4 pAb (H00057396-B01P, Novus Biological), anti-SR (1H14) mAb (SC-13509, Santa Cruz) anti-SC35 (E16) pAb (SC-10252,
Santa Cruz), and anti-SC35 mAb (S4045, Sigma).
RNA Preparation and Reverse Transcription PCR or Quantitative PCR
RNA was isolated from cells using the RNeasy Plus mini kit (Qiagen) which removes DNA, then was reverse transcribed using the iSCRIPT Advanced cDNA synthesis kit (Bio-Rad) containing oligo-dT and random hexamer. mRNA levels were semi quantified by RT-PCR or quantified by RT-qPCR using iQ Sybr
Green supermix (Bio-Rad) in triplicate on the CFX384 Real-Time System
(BioRad). All primer sets for RT-qPCR designed to measure mRNA levels were validated for their specificity and amplification efficiency (85%-110%) using melt curve analysis, RT-qPCR product sequencing, and standard dilution analysis.
RT-qPCR results were normalized using the ∆∆CT method using18s rRNA and
β-Actin as reference genes and untreated normoxia samples, GFP lentivirus, or empty vectors (His), or DMSO treated samples as controls. At least three independent experiments were performed to generate the results presented in the figures.
155
Exon Array Analysis of Alternative Splicing
The genome-wide effect of hypoxia on alternative splicing was determined in Hep3B cells using the Affymetrix GeneChip Human Exon 1.0 ST array. Hep3B cells were cultured, as described above, under normoxic (21%) or hypoxic (1.2%) conditions for 12 hours. RNA was prepared using a Qiagen RNeasy Plus RNA extraction kit. RNA was reverse transcribed and assessed for hypoxic induction of HIF target genes using RT-qPCR. RNAs that passed quality control were submitted to the Gene Expression Core at the University of Colorado
Comprehensive Cancer Center for exon array analysis. Three independent replicate samples were analyzed for each treatment. The Core performed target labeling, hybridization, and chip scanning. To assess alternative splicing events, raw CEL files were analyzed using freely available Altanalyze software
(http://www.altanalyze.org) (372) with the following settings: Species – Human;
Gene and Exon set – both core; gene and exon level normalization FIRMA. Raw
CEL files were alternatively analyzed using easyExon
(http://microarray.ym.edu.tw:8080/tools/ module/ easyexon/ index. jsp?mode
=home), with default settings and using RMA filtering (373). Additionally,
Altanalyze software was used for gene expression analysis. Genes predicted to undergo alternative splicing by Altanalyze and easyExon were then compared to genes that were induced by hypoxia (Altanalyze) using Venny
(http://bioinfogp.cnb.csic.es/tools/venny/) to indentify genes that were predicted to undergo alternative splicing and hypoxia induction.
156
Functional Clustering of Hypoxia Inducible Genes and Alternatively Spliced
Genes Using DAVID Bioinformatics Resources
DAVID Bioinformatics (331) was used to create functional annotation
clusters of genes 1) that were hypoxia induced, 2) genes that were predicted to
undergo hypoxia-induced AS, and 3) genes that were hypoxia-induced and
exhibiting alternative splicing. Gene list generated by DAVID included information
on gene-ontology classification, enrichment score, and gene names involved in
each functional cluster.
In Vitro Tumorigenic Assays
Clonogenic survival assays: 1000 cells per well were plated in
duplicate in 6-well plates in normal growth media. Media was changed every two
days, then cells were washed and colonies stained with crystal violet dye 7 to 10
days after plating, pictures of each well were taken and manually counted using imageJ software. Clonogenic survival and wound healing scratch assays were performed for PRC3 cells stably expressing control or HIF2 shRNAs and treated with DMSO (control) or TG003. Wound healing migration assays: Cells were
plated at 80% confluency in normal growth media in duplicate in 6-well plates.
The next day, when cells had formed a monolayer, the monolayers were
scratched with a 200 µL pipette tip. Then the cells were washed with PBS twice
to remove the detached cells and serum and then incubated for the remainder of
the experiment in serum-free media. Photographs of the scratches at the same
location in the well were taken daily and the width of the scratch measured using 157 imageJ software. Duplicate measurements were averaged and graphed as percent closure of the scratch compared to the initial scratch (Day 0).
Statistical Analysis
One-way analysis of variance was performed unless otherwise stated.
Error bars in figures indicate standard deviation. Asterisks indicate statistical significance as follows: *, P< 0.05; **, P< 0.01. Controls for statistical analysis are specified in each figure. All experiments were performed at least three separate times. 158
REFERENCES
1. Semenza GL. Life with oxygen. Science. 2007;318:62-4.
2. Folkman J. What is the evidence that tumors are angiogenesis dependent? J Natl Cancer Inst. 1990;82:4-6.
3. Carreau A, El Hafny-Rahbi B, Matejuk A, Grillon C, Kieda C. Why is the partial oxygen pressure of human tissues a crucial parameter? Small molecules and hypoxia. J Cell Mol Med. 2011;15:1239-53.
4. Parmar K, Mauch P, Vergilio JA, Sackstein R, Down JD. Distribution of hematopoietic stem cells in the bone marrow according to regional hypoxia. Proc Natl Acad Sci U S A. 2007;104:5431-6.
5. Braun RD, Lanzen JL, Snyder SA, Dewhirst MW. Comparison of tumor and normal tissue oxygen tension measurements using OxyLite or microelectrodes in rodents. Am J Physiol Heart Circ Physiol. 2001;280:H2533- 44.
6. Stockwell RA. Morphometry of cytoplasmic components of mammalian articular chondrocytes and corneal keratocytes: species and zonal variations of mitochondria in relation to nutrition. J Anat. 1991;175:251-61.
7. Keith B, Simon MC. Hypoxia-inducible factors, stem cells, and cancer. Cell. 2007;129:465-72.
8. Knighton DR, Silver IA, Hunt TK. Regulation of wound-healing angiogenesis-effect of oxygen gradients and inspired oxygen concentration. Surgery. 1981;90:262-70.
9. Semenza GL. Vascular responses to hypoxia and ischemia. Arterioscler Thromb Vasc Biol. 2010;30:648-52.
10. Carmeliet P. Angiogenesis in life, disease and medicine. Nature. 2005;438:932-6.
11. Dunwoodie SL. Reprint of mutation of the fucose-specific beta1,3 N- acetylglucosaminyltransferase LFNG results in abnormal formation of the spine. Biochim Biophys Acta. 2009;1792:862-73.
12. Semenza GL. Hypoxia, clonal selection, and the role of HIF-1 in tumor progression. Crit Rev Biochem Mol Biol. 2000;35:71-103. 159
13. Wenger RH. Cellular adaptation to hypoxia: O2-sensing protein hydroxylases, hypoxia-inducible transcription factors, and O2-regulated gene expression. FASEB J. 2002;16:1151-62.
14. Seagroves TN, Ryan HE, Lu H, Wouters BG, Knapp M, Thibault P, et al. Transcription factor HIF-1 is a necessary mediator of the pasteur effect in mammalian cells. Mol Cell Biol. 2001;21:3436-44.
15. Minchenko A, Leshchinsky I, Opentanova I, Sang N, Srinivas V, Armstead V, et al. Hypoxia-inducible factor-1-mediated expression of the 6-phosphofructo- 2-kinase/fructose-2,6-bisphosphatase-3 (PFKFB3) gene. Its possible role in the Warburg effect. J Biol Chem. 2002;277:6183-7.
16. Brown JM, Giaccia AJ. The unique physiology of solid tumors: opportunities (and problems) for cancer therapy. Cancer Res. 1998;58:1408-16.
17. Hockel M, Schlenger K, Hockel S, Vaupel P. Hypoxic cervical cancers with low apoptotic index are highly aggressive. Cancer Res. 1999;59:4525-8.
18. Brown JM. Exploiting the hypoxic cancer cell: mechanisms and therapeutic strategies. Mol Med Today. 2000;6:157-62.
19. Hockel M, Vaupel P. Tumor hypoxia: definitions and current clinical, biologic, and molecular aspects. J Natl Cancer Inst. 2001;93:266-76.
20. Vaupel P, Thews O, Hoeckel M. Treatment resistance of solid tumors: role of hypoxia and anemia. Med Oncol. 2001;18:243-59.
21. Wang GL, Jiang BH, Rue EA, Semenza GL. Hypoxia-inducible factor 1 is a basic-helix-loop-helix-PAS heterodimer regulated by cellular O2 tension. Proc Natl Acad Sci U S A. 1995;92:5510-4.
22. Wang GL, Semenza GL. Purification and characterization of hypoxia- inducible factor 1. J Biol Chem. 1995;270:1230-7.
23. Srinivas V, Zhang LP, Zhu XH, Caro J. Characterization of an oxygen/redox-dependent degradation domain of hypoxia-inducible factor alpha (HIF-alpha) proteins. Biochem Biophys Res Commun. 1999;260:557-61.
24. Hara S, Hamada J, Kobayashi C, Kondo Y, Imura N. Expression and characterization of hypoxia-inducible factor (HIF)-3alpha in human kidney: suppression of HIF-mediated gene expression by HIF-3alpha. Biochem Biophys Res Commun. 2001;287:808-13.
25. Talks KL, Turley H, Gatter KC, Maxwell PH, Pugh CW, Ratcliffe PJ, et al. The expression and distribution of the hypoxia-inducible factors HIF-1alpha and HIF-2alpha in normal human tissues, cancers, and tumor-associated macrophages. Am J Pathol. 2000;157:411-21. 160
26. Lee JS, Kim EY, Iwata H, Tanabe S. Molecular characterization and tissue distribution of aryl hydrocarbon receptor nuclear translocator isoforms, ARNT1 and ARNT2, and identification of novel splice variants in common cormorant (Phalacrocorax carbo). Comp Biochem Physiol C Toxicol Pharmacol. 2007;145:379-93.
27. Freeburg PB, Abrahamson DR. Divergent expression patterns for hypoxia- inducible factor-1beta and aryl hydrocarbon receptor nuclear transporter-2 in developing kidney. J Am Soc Nephrol. 2004;15:2569-78.
28. Petersen SL, Curran MA, Marconi SA, Carpenter CD, Lubbers LS, McAbee MD. Distribution of mRNAs encoding the arylhydrocarbon receptor, arylhydrocarbon receptor nuclear translocator, and arylhydrocarbon receptor nuclear translocator-2 in the rat brain and brainstem. J Comp Neurol. 2000;427:428-39.
29. Sparrow DB, Boyle SC, Sams RS, Mazuruk B, Zhang L, Moeckel GW, et al. Placental insufficiency associated with loss of Cited1 causes renal medullary dysplasia. Journal of the American Society of Nephrology : JASN. 2009;20:777- 86.
30. Semenza GL, Wang GL. A nuclear factor induced by hypoxia via de novo protein synthesis binds to the human erythropoietin gene enhancer at a site required for transcriptional activation. Mol Cell Biol. 1992;12:5447-54.
31. Wiener CM, Booth G, Semenza GL. In vivo expression of mRNAs encoding hypoxia-inducible factor 1. Biochem Biophys Res Commun. 1996;225:485-8.
32. Manalo DJ, Rowan A, Lavoie T, Natarajan L, Kelly BD, Ye SQ, et al. Transcriptional regulation of vascular endothelial cell responses to hypoxia by HIF-1. Blood. 2005;105:659-69.
33. Imamura T, Kikuchi H, Herraiz MT, Park DY, Mizukami Y, Mino-Kenduson M, et al. HIF-1alpha and HIF-2alpha have divergent roles in colon cancer. Int J Cancer. 2009;124:763-71.
34. Mole DR, Blancher C, Copley RR, Pollard PJ, Gleadle JM, Ragoussis J, et al. Genome-wide association of hypoxia-inducible factor (HIF)-1{alpha} and HIF- 2{alpha} DNA binding with expression profiling of hypoxia-inducible transcripts. J Biol Chem. 2009;284:16767-75.
35. Xia X, Lemieux ME, Li W, Carroll JS, Brown M, Liu XS, et al. Integrative analysis of HIF binding and transactivation reveals its role in maintaining histone methylation homeostasis. Proc Natl Acad Sci U S A. 2009;106:4260-5.
161
36. Ema M, Taya S, Yokotani N, Sogawa K, Matsuda Y, Fujii-Kuriyama Y. A novel bHLH-PAS factor with close sequence similarity to hypoxia-inducible factor 1alpha regulates the VEGF expression and is potentially involved in lung and vascular development. Proc Natl Acad Sci U S A. 1997;94:4273-8.
37. Flamme I, Frohlich T, von Reutern M, Kappel A, Damert A, Risau W. HRF, a putative basic helix-loop-helix-PAS-domain transcription factor is closely related to hypoxia-inducible factor-1 alpha and developmentally expressed in blood vessels. Mech Dev. 1997;63:51-60.
38. Hogenesch JB, Chan WK, Jackiw VH, Brown RC, Gu YZ, Pray-Grant M, et al. Characterization of a subset of the basic-helix-loop-helix-PAS superfamily that interacts with components of the dioxin signaling pathway. J Biol Chem. 1997;272:8581-93.
39. Tian H, McKnight SL, Russell DW. Endothelial PAS domain protein 1 (EPAS1), a transcription factor selectively expressed in endothelial cells. Genes Dev. 1997;11:72-82.
40. Elvert G, Lanz S, Kappel A, Flamme I. mRNA cloning and expression studies of the quail homologue of HIF-2alpha. Mech Dev. 1999;87:193-7.
41. Wenger RH, Gassmann M. Oxygen(es) and the hypoxia-inducible factor- 1. Biol Chem. 1997;378:609-16.
42. Wiesener MS, Turley H, Allen WE, Willam C, Eckardt KU, Talks KL, et al. Induction of endothelial PAS domain protein-1 by hypoxia: characterization and comparison with hypoxia-inducible factor-1alpha. Blood. 1998;92:2260-8.
43. Kappel A, Schlaeger TM, Flamme I, Orkin SH, Risau W, Breier G. Role of SCL/Tal-1, GATA, and ets transcription factor binding sites for the regulation of flk-1 expression during murine vascular development. Blood. 2000;96:3078-85.
44. Jain S, Maltepe E, Lu MM, Simon C, Bradfield CA. Expression of ARNT, ARNT2, HIF1 alpha, HIF2 alpha and Ah receptor mRNAs in the developing mouse. Mech Dev. 1998;73:117-23.
45. Krock BL, Skuli N, Simon MC. Hypoxia-induced angiogenesis: good and evil. Genes Cancer. 2011;2:1117-33.
46. Jiang BH, Zheng JZ, Leung SW, Roe R, Semenza GL. Transactivation and inhibitory domains of hypoxia-inducible factor 1alpha. Modulation of transcriptional activity by oxygen tension. J Biol Chem. 1997;272:19253-60.
47. Pugh CW, O'Rourke JF, Nagao M, Gleadle JM, Ratcliffe PJ. Activation of hypoxia-inducible factor-1; definition of regulatory domains within the alpha subunit. J Biol Chem. 1997;272:11205-14. 162
48. Hu CJ, Sataur A, Wang L, Chen H, Simon MC. The N-terminal transactivation domain confers target gene specificity of hypoxia-inducible factors HIF-1alpha and HIF-2alpha. Molecular biology of the cell. 2007;18:4528-42.
49. Lau KW, Tian YM, Raval RR, Ratcliffe PJ, Pugh CW. Target gene selectivity of hypoxia-inducible factor-alpha in renal cancer cells is conveyed by post-DNA-binding mechanisms. Br J Cancer. 2007;96:1284-92.
50. Pawlus MR, Wang L, Murakami A, Dai G, Hu CJ. STAT3 or USF2 contributes to HIF target gene specificity. PLoS One. 2013;8:e72358.
51. Ke Q, Costa M. Hypoxia-inducible factor-1 (HIF-1). Mol Pharmacol. 2006;70:1469-80.
52. Huang LE, Gu J, Schau M, Bunn HF. Regulation of hypoxia-inducible factor 1alpha is mediated by an O2-dependent degradation domain via the ubiquitin-proteasome pathway. Proc Natl Acad Sci U S A. 1998;95:7987-92.
53. Stolze IP, Mole DR, Ratcliffe PJ. Regulation of HIF: prolyl hydroxylases. Novartis Found Symp. 2006;272:15-25; discussion -36.
54. Wenger RH. Mammalian oxygen sensing, signalling and gene regulation. J Exp Biol. 2000;203:1253-63.
55. Jaakkola P, Mole DR, Tian YM, Wilson MI, Gielbert J, Gaskell SJ, et al. Targeting of HIF-alpha to the von Hippel-Lindau ubiquitylation complex by O2- regulated prolyl hydroxylation. Science. 2001;292:468-72.
56. Epstein AC, Gleadle JM, McNeill LA, Hewitson KS, O'Rourke J, Mole DR, et al. C. elegans EGL-9 and mammalian homologs define a family of dioxygenases that regulate HIF by prolyl hydroxylation. Cell. 2001;107:43-54.
57. Masson N, Willam C, Maxwell PH, Pugh CW, Ratcliffe PJ. Independent function of two destruction domains in hypoxia-inducible factor-alpha chains activated by prolyl hydroxylation. EMBO J. 2001;20:5197-206.
58. Ivan M, Kondo K, Yang H, Kim W, Valiando J, Ohh M, et al. HIFalpha targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing. Science. 2001;292:464-8.
59. Bruick RK, McKnight SL. A conserved family of prolyl-4-hydroxylases that modify HIF. Science. 2001;294:1337-40.
60. Cockman ME, Masson N, Mole DR, Jaakkola P, Chang GW, Clifford SC, et al. Hypoxia inducible factor-alpha binding and ubiquitylation by the von Hippel- Lindau tumor suppressor protein. J Biol Chem. 2000;275:25733-41. 163
61. Mole DR, Maxwell PH, Pugh CW, Ratcliffe PJ. Regulation of HIF by the von Hippel-Lindau tumour suppressor: implications for cellular oxygen sensing. IUBMB Life. 2001;52:43-7.
62. Kamura T, Sato S, Iwai K, Czyzyk-Krzeska M, Conaway RC, Conaway JW. Activation of HIF1alpha ubiquitination by a reconstituted von Hippel-Lindau (VHL) tumor suppressor complex. Proc Natl Acad Sci U S A. 2000;97:10430-5.
63. Maxwell PH, Wiesener MS, Chang GW, Clifford SC, Vaux EC, Cockman ME, et al. The tumour suppressor protein VHL targets hypoxia-inducible factors for oxygen-dependent proteolysis. Nature. 1999;399:271-5.
64. Ohh M, Park CW, Ivan M, Hoffman MA, Kim TY, Huang LE, et al. Ubiquitination of hypoxia-inducible factor requires direct binding to the beta- domain of the von Hippel-Lindau protein. Nat Cell Biol. 2000;2:423-7.
65. Tanimoto K, Makino Y, Pereira T, Poellinger L. Mechanism of regulation of the hypoxia-inducible factor-1 alpha by the von Hippel-Lindau tumor suppressor protein. EMBO J. 2000;19:4298-309.
66. Kaelin WG, Jr. The von Hippel-Lindau protein, HIF hydroxylation, and oxygen sensing. Biochem Biophys Res Commun. 2005;338:627-38.
67. Kaluz S, Kaluzova M, Stanbridge EJ. Regulation of gene expression by hypoxia: integration of the HIF-transduced hypoxic signal at the hypoxia- responsive element. Clin Chim Acta. 2008;395:6-13.
68. Wenger RH, Stiehl DP, Camenisch G. Integration of oxygen signaling at the consensus HRE. Sci STKE. 2005;2005:re12.
69. Chachami G, Paraskeva E, Mingot JM, Braliou GG, Gorlich D, Simos G. Transport of hypoxia-inducible factor HIF-1alpha into the nucleus involves importins 4 and 7. Biochemical and biophysical research communications. 2009;390:235-40.
70. Mahon PC, Hirota K, Semenza GL. FIH-1: a novel protein that interacts with HIF-1alpha and VHL to mediate repression of HIF-1 transcriptional activity. Genes Dev. 2001;15:2675-86.
71. Sodhi A, Montaner S, Patel V, Zohar M, Bais C, Mesri EA, et al. The Kaposi's sarcoma-associated herpes virus G protein-coupled receptor up- regulates vascular endothelial growth factor expression and secretion through mitogen-activated protein kinase and p38 pathways acting on hypoxia-inducible factor 1alpha. Cancer Res. 2000;60:4873-80.
72. Minet E, Arnould T, Michel G, Roland I, Mottet D, Raes M, et al. ERK activation upon hypoxia: involvement in HIF-1 activation. FEBS Lett. 2000;468:53-8. 164
73. Conrad PW, Freeman TL, Beitner-Johnson D, Millhorn DE. EPAS1 trans- activation during hypoxia requires p42/p44 MAPK. J Biol Chem. 1999;274:33709- 13.
74. Richard DE, Berra E, Gothie E, Roux D, Pouyssegur J. p42/p44 mitogen- activated protein kinases phosphorylate hypoxia-inducible factor 1alpha (HIF- 1alpha) and enhance the transcriptional activity of HIF-1. J Biol Chem. 1999;274:32631-7.
75. Laughner E, Taghavi P, Chiles K, Mahon PC, Semenza GL. HER2 (neu) signaling increases the rate of hypoxia-inducible factor 1alpha (HIF-1alpha) synthesis: novel mechanism for HIF-1-mediated vascular endothelial growth factor expression. Mol Cell Biol. 2001;21:3995-4004.
76. Treins C, Giorgetti-Peraldi S, Murdaca J, Semenza GL, Van Obberghen E. Insulin stimulates hypoxia-inducible factor 1 through a phosphatidylinositol 3- kinase/target of rapamycin-dependent signaling pathway. J Biol Chem. 2002;277:27975-81.
77. Fukuda R, Hirota K, Fan F, Jung YD, Ellis LM, Semenza GL. Insulin-like growth factor 1 induces hypoxia-inducible factor 1-mediated vascular endothelial growth factor expression, which is dependent on MAP kinase and phosphatidylinositol 3-kinase signaling in colon cancer cells. J Biol Chem. 2002;277:38205-11.
78. Gingras AC, Raught B, Sonenberg N. Regulation of translation initiation by FRAP/mTOR. Genes Dev. 2001;15:807-26.
79. Dunwoodie SL. The role of hypoxia in development of the Mammalian embryo. Dev Cell. 2009;17:755-73.
80. Ryan HE, Lo J, Johnson RS. HIF-1 alpha is required for solid tumor formation and embryonic vascularization. EMBO J. 1998;17:3005-15.
81. Compernolle V, Brusselmans K, Acker T, Hoet P, Tjwa M, Beck H, et al. Loss of HIF-2alpha and inhibition of VEGF impair fetal lung maturation, whereas treatment with VEGF prevents fatal respiratory distress in premature mice. Nat Med. 2002;8:702-10.
82. Peng J, Zhang L, Drysdale L, Fong GH. The transcription factor EPAS- 1/hypoxia-inducible factor 2alpha plays an important role in vascular remodeling. Proc Natl Acad Sci U S A. 2000;97:8386-91.
83. Tian H, Hammer RE, Matsumoto AM, Russell DW, McKnight SL. The hypoxia-responsive transcription factor EPAS1 is essential for catecholamine homeostasis and protection against heart failure during embryonic development. Genes Dev. 1998;12:3320-4. 165
84. Maltepe E, Schmidt JV, Baunoch D, Bradfield CA, Simon MC. Abnormal angiogenesis and responses to glucose and oxygen deprivation in mice lacking the protein ARNT. Nature. 1997;386:403-7.
85. Schipani E, Ryan HE, Didrickson S, Kobayashi T, Knight M, Johnson RS. Hypoxia in cartilage: HIF-1alpha is essential for chondrocyte growth arrest and survival. Genes & development. 2001;15:2865-76.
86. Yun Z, Maecker HL, Johnson RS, Giaccia AJ. Inhibition of PPAR gamma 2 gene expression by the HIF-1-regulated gene DEC1/Stra13: a mechanism for regulation of adipogenesis by hypoxia. Dev Cell. 2002;2:331-41.
87. Wang Y, Wan C, Deng L, Liu X, Cao X, Gilbert SR, et al. The hypoxia- inducible factor alpha pathway couples angiogenesis to osteogenesis during skeletal development. J Clin Invest. 2007;117:1616-26.
88. Takubo K, Goda N, Yamada W, Iriuchishima H, Ikeda E, Kubota Y, et al. Regulation of the HIF-1alpha level is essential for hematopoietic stem cells. Cell Stem Cell. 2010;7:391-402.
89. Dang EV, Barbi J, Yang HY, Jinasena D, Yu H, Zheng Y, et al. Control of T(H)17/T(reg) balance by hypoxia-inducible factor 1. Cell. 2011;146:772-84.
90. Zinkernagel AS, Johnson RS, Nizet V. Hypoxia inducible factor (HIF) function in innate immunity and infection. J Mol Med (Berl). 2007;85:1339-46.
91. Scortegagna M, Ding K, Oktay Y, Gaur A, Thurmond F, Yan LJ, et al. Multiple organ pathology, metabolic abnormalities and impaired homeostasis of reactive oxygen species in Epas1-/- mice. Nat Genet. 2003;35:331-40.
92. Tandara AA, Mustoe TA. Oxygen in wound healing--more than a nutrient. World J Surg. 2004;28:294-300.
93. Liu L, Marti GP, Wei X, Zhang X, Zhang H, Liu YV, et al. Age-dependent impairment of HIF-1alpha expression in diabetic mice: Correction with electroporation-facilitated gene therapy increases wound healing, angiogenesis, and circulating angiogenic cells. J Cell Physiol. 2008;217:319-27.
94. Lerman OZ, Galiano RD, Armour M, Levine JP, Gurtner GC. Cellular dysfunction in the diabetic fibroblast: impairment in migration, vascular endothelial growth factor production, and response to hypoxia. Am J Pathol. 2003;162:303-12.
95. Brem H, Tomic-Canic M, Entero H, Hanflik AM, Wang VM, Fallon JT, et al. The synergism of age and db/db genotype impairs wound healing. Exp Gerontol. 2007;42:523-31. 166
96. Catrina SB, Okamoto K, Pereira T, Brismar K, Poellinger L. Hyperglycemia regulates hypoxia-inducible factor-1alpha protein stability and function. Diabetes. 2004;53:3226-32.
97. Studer L, Csete M, Lee SH, Kabbani N, Walikonis J, Wold B, et al. Enhanced proliferation, survival, and dopaminergic differentiation of CNS precursors in lowered oxygen. J Neurosci. 2000;20:7377-83.
98. Morrison SJ, Csete M, Groves AK, Melega W, Wold B, Anderson DJ. Culture in reduced levels of oxygen promotes clonogenic sympathoadrenal differentiation by isolated neural crest stem cells. J Neurosci. 2000;20:7370-6.
99. Danet GH, Pan Y, Luongo JL, Bonnet DA, Simon MC. Expansion of human SCID-repopulating cells under hypoxic conditions. J Clin Invest. 2003;112:126-35.
100. Gustafsson MV, Zheng X, Pereira T, Gradin K, Jin S, Lundkvist J, et al. Hypoxia requires notch signaling to maintain the undifferentiated cell state. Dev Cell. 2005;9:617-28.
101. Gustafsson MV, Zheng X, Pereira T, Gradin K, Jin S, Lundkvist J, et al. Hypoxia requires notch signaling to maintain the undifferentiated cell state. Dev Cell. 2005;9:617-28.
102. Coussens M, Davy P, Brown L, Foster C, Andrews WH, Nagata M, et al. RNAi screen for telomerase reverse transcriptase transcriptional regulators identifies HIF1alpha as critical for telomerase function in murine embryonic stem cells. Proceedings of the National Academy of Sciences of the United States of America. 2010;107:13842-7.
103. Covello KL, Kehler J, Yu H, Gordan JD, Arsham AM, Hu CJ, et al. HIF- 2alpha regulates Oct-4: effects of hypoxia on stem cell function, embryonic development, and tumor growth. Genes Dev. 2006;20:557-70.
104. Lee TI, Jenner RG, Boyer LA, Guenther MG, Levine SS, Kumar RM, et al. Control of developmental regulators by Polycomb in human embryonic stem cells. Cell. 2006;125:301-13.
105. Loh YH, Wu Q, Chew JL, Vega VB, Zhang W, Chen X, et al. The Oct4 and Nanog transcription network regulates pluripotency in mouse embryonic stem cells. Nat Genet. 2006;38:431-40.
106. Kehler J, Tolkunova E, Koschorz B, Pesce M, Gentile L, Boiani M, et al. Oct4 is required for primordial germ cell survival. EMBO Rep. 2004;5:1078-83.
107. Kim Y, Lin Q, Glazer PM, Yun Z. Hypoxic tumor microenvironment and cancer cell differentiation. Curr Mol Med. 2009;9:425-34. 167
108. Cardenas-Navia LI, Mace D, Richardson RA, Wilson DF, Shan S, Dewhirst MW. The pervasive presence of fluctuating oxygenation in tumors. Cancer Res. 2008;68:5812-9.
109. Vaupel P. Prognostic potential of the pre-therapeutic tumor oxygenation status. Adv Exp Med Biol. 2009;645:241-6.
110. Ruan K, Song G, Ouyang G. Role of hypoxia in the hallmarks of human cancer. J Cell Biochem. 2009;107:1053-62.
111. Cojocariu OM, Huguet F, Lefevre M, Perie S. [Prognosis and predictive factors in head-and-neck cancers]. Bull Cancer. 2009;96:369-78.
112. Kalliomaki TM, McCallum G, Wells PG, Hill RP. Progression and metastasis in a transgenic mouse breast cancer model: effects of exposure to in vivo hypoxia. Cancer Lett. 2009;282:98-108.
113. Shweiki D, Itin A, Soffer D, Keshet E. Vascular endothelial growth factor induced by hypoxia may mediate hypoxia-initiated angiogenesis. Nature. 1992;359:843-5.
114. Mukhopadhyay D, Tsiokas L, Zhou XM, Foster D, Brugge JS, Sukhatme VP. Hypoxic induction of human vascular endothelial growth factor expression through c-Src activation. Nature. 1995;375:577-81.
115. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell. 2000;100:57-70.
116. Kizaka-Kondoh S, Tanaka S, Harada H, Hiraoka M. The HIF-1-active microenvironment: an environmental target for cancer therapy. Adv Drug Deliv Rev. 2009;61:623-32.
117. Kaelin WG, Jr. Molecular basis of the VHL hereditary cancer syndrome. Nat Rev Cancer. 2002;2:673-82.
118. Maher ER, Kaelin WG, Jr. von Hippel-Lindau disease. Medicine (Baltimore). 1997;76:381-91.
119. Maddock IR, Moran A, Maher ER, Teare MD, Norman A, Payne SJ, et al. A genetic register for von Hippel-Lindau disease. J Med Genet. 1996;33:120-7.
120. Richard S, Campello C, Taillandier L, Parker F, Resche F. Haemangioblastoma of the central nervous system in von Hippel-Lindau disease. French VHL Study Group. J Intern Med. 1998;243:547-53.
121. Lamiell JM, Salazar FG, Hsia YE. von Hippel-Lindau disease affecting 43 members of a single kindred. Medicine (Baltimore). 1989;68:1-29. 168
122. Lonser RR, Weil RJ, Wanebo JE, DeVroom HL, Oldfield EH. Surgical management of spinal cord hemangioblastomas in patients with von Hippel- Lindau disease. J Neurosurg. 2003;98:106-16.
123. Wanebo JE, Lonser RR, Glenn GM, Oldfield EH. The natural history of hemangioblastomas of the central nervous system in patients with von Hippel- Lindau disease. J Neurosurg. 2003;98:82-94.
124. Weil RJ, Lonser RR, DeVroom HL, Wanebo JE, Oldfield EH. Surgical management of brainstem hemangioblastomas in patients with von Hippel- Lindau disease. J Neurosurg. 2003;98:95-105.
125. Dollfus H, Massin P, Taupin P, Nemeth C, Amara S, Giraud S, et al. Retinal hemangioblastoma in von Hippel-Lindau disease: a clinical and molecular study. Invest Ophthalmol Vis Sci. 2002;43:3067-74.
126. Webster AR, Maher ER, Moore AT. Clinical characteristics of ocular angiomatosis in von Hippel-Lindau disease and correlation with germline mutation. Arch Ophthalmol. 1999;117:371-8.
127. Hammel PR, Vilgrain V, Terris B, Penfornis A, Sauvanet A, Correas JM, et al. Pancreatic involvement in von Hippel-Lindau disease. The Groupe Francophone d'Etude de la Maladie de von Hippel-Lindau. Gastroenterology. 2000;119:1087-95.
128. Richard S, David P, Marsot-Dupuch K, Giraud S, Beroud C, Resche F. Central nervous system hemangioblastomas, endolymphatic sac tumors, and von Hippel-Lindau disease. Neurosurg Rev. 2000;23:1-22; discussion 3-4.
129. Friedrich CA. Von Hippel-Lindau syndrome. A pleomorphic condition. Cancer. 1999;86:2478-82.
130. Frantzen C, Links TP, Giles RH. Von Hippel-Lindau Disease. In: Pagon RA, Adam MP, Bird TD, Dolan CR, Fong CT, Stephens K, editors. GeneReviews. Seattle (WA)1993.
131. Gnarra JR, Tory K, Weng Y, Schmidt L, Wei MH, Li H, et al. Mutations of the VHL tumour suppressor gene in renal carcinoma. Nat Genet. 1994;7:85-90.
132. Brieger J, Weidt EJ, Schirmacher P, Storkel S, Huber C, Decker HJ. Inverse regulation of vascular endothelial growth factor and VHL tumor suppressor gene in sporadic renal cell carcinomas is correlated with vascular growth: an in vivo study on 29 tumors. J Mol Med (Berl). 1999;77:505-10.
133. Gallou C, Joly D, Mejean A, Staroz F, Martin N, Tarlet G, et al. Mutations of the VHL gene in sporadic renal cell carcinoma: definition of a risk factor for VHL patients to develop an RCC. Hum Mutat. 1999;13:464-75. 169
134. Kondo K, Yao M, Yoshida M, Kishida T, Shuin T, Miura T, et al. Comprehensive mutational analysis of the VHL gene in sporadic renal cell carcinoma: relationship to clinicopathological parameters. Genes Chromosomes Cancer. 2002;34:58-68.
135. Huntly BJ, Gilliland DG. Leukaemia stem cells and the evolution of cancer- stem-cell research. Nat Rev Cancer. 2005;5:311-21.
136. Reya T, Morrison SJ, Clarke MF, Weissman IL. Stem cells, cancer, and cancer stem cells. Nature. 2001;414:105-11.
137. Crea F, Mathews LA, Farrar WL, Hurt EM. Targeting prostate cancer stem cells. Anticancer Agents Med Chem. 2009;9:1105-13.
138. Maitland NJ, Collins AT. Prostate cancer stem cells: a new target for therapy. J Clin Oncol. 2008;26:2862-70.
139. Vescovi AL, Galli R, Reynolds BA. Brain tumour stem cells. Nat Rev Cancer. 2006;6:425-36.
140. Ma Y, Liang D, Liu J, Axcrona K, Kvalheim G, Stokke T, et al. Prostate cancer cell lines under hypoxia exhibit greater stem-like properties. PLoS One. 2011;6:e29170.
141. Li Z, Rich JN. Hypoxia and hypoxia inducible factors in cancer stem cell maintenance. Curr Top Microbiol Immunol. 2010;345:21-30.
142. Gordan JD, Bertout JA, Hu CJ, Diehl JA, Simon MC. HIF-2alpha promotes hypoxic cell proliferation by enhancing c-myc transcriptional activity. Cancer Cell. 2007;11:335-47.
143. Plouet J, Moukadiri H. Characterization of the receptor to vasculotropin on bovine adrenal cortex-derived capillary endothelial cells. J Biol Chem. 1990;265:22071-4.
144. Vaisman N, Gospodarowicz D, Neufeld G. Characterization of the receptors for vascular endothelial growth factor. J Biol Chem. 1990;265:19461-6.
145. Jakeman LB, Winer J, Bennett GL, Altar CA, Ferrara N. Binding sites for vascular endothelial growth factor are localized on endothelial cells in adult rat tissues. J Clin Invest. 1992;89:244-53.
146. Jakeman LB, Armanini M, Phillips HS, Ferrara N. Developmental expression of binding sites and messenger ribonucleic acid for vascular endothelial growth factor suggests a role for this protein in vasculogenesis and angiogenesis. Endocrinology. 1993;133:848-59. 170
147. Gerber HP, McMurtrey A, Kowalski J, Yan M, Keyt BA, Dixit V, et al. Vascular endothelial growth factor regulates endothelial cell survival through the phosphatidylinositol 3'-kinase/Akt signal transduction pathway. Requirement for Flk-1/KDR activation. J Biol Chem. 1998;273:30336-43.
148. Eliceiri BP, Paul R, Schwartzberg PL, Hood JD, Leng J, Cheresh DA. Selective requirement for Src kinases during VEGF-induced angiogenesis and vascular permeability. Mol Cell. 1999;4:915-24.
149. Guo D, Jia Q, Song HY, Warren RS, Donner DB. Vascular endothelial cell growth factor promotes tyrosine phosphorylation of mediators of signal transduction that contain SH2 domains. Association with endothelial cell proliferation. J Biol Chem. 1995;270:6729-33.
150. Byzova TV, Goldman CK, Pampori N, Thomas KA, Bett A, Shattil SJ, et al. A mechanism for modulation of cellular responses to VEGF: activation of the integrins. Mol Cell. 2000;6:851-60.
151. Takahashi T, Ueno H, Shibuya M. VEGF activates protein kinase C- dependent, but Ras-independent Raf-MEK-MAP kinase pathway for DNA synthesis in primary endothelial cells. Oncogene. 1999;18:2221-30.
152. Wu LW, Mayo LD, Dunbar JD, Kessler KM, Baerwald MR, Jaffe EA, et al. Utilization of distinct signaling pathways by receptors for vascular endothelial cell growth factor and other mitogens in the induction of endothelial cell proliferation. J Biol Chem. 2000;275:5096-103.
153. Gille H, Kowalski J, Li B, LeCouter J, Moffat B, Zioncheck TF, et al. Analysis of biological effects and signaling properties of Flt-1 (VEGFR-1) and KDR (VEGFR-2). A reassessment using novel receptor-specific vascular endothelial growth factor mutants. J Biol Chem. 2001;276:3222-30.
154. Tang N, Wang L, Esko J, Giordano FJ, Huang Y, Gerber HP, et al. Loss of HIF-1alpha in endothelial cells disrupts a hypoxia-driven VEGF autocrine loop necessary for tumorigenesis. Cancer Cell. 2004;6:485-95.
155. Ben-Yosef Y, Miller A, Shapiro S, Lahat N. Hypoxia of endothelial cells leads to MMP-2-dependent survival and death. Am J Physiol Cell Physiol. 2005;289:C1321-31.
156. Skuli N, Liu L, Runge A, Wang T, Yuan L, Patel S, et al. Endothelial deletion of hypoxia-inducible factor-2alpha (HIF-2alpha) alters vascular function and tumor angiogenesis. Blood. 2009;114:469-77.
157. Patel NS, Li JL, Generali D, Poulsom R, Cranston DW, Harris AL. Up- regulation of delta-like 4 ligand in human tumor vasculature and the role of basal expression in endothelial cell function. Cancer Res. 2005;65:8690-7. 171
158. Hellstrom M, Phng LK, Hofmann JJ, Wallgard E, Coultas L, Lindblom P, et al. Dll4 signalling through Notch1 regulates formation of tip cells during angiogenesis. Nature. 2007;445:776-80.
159. Lobov IB, Renard RA, Papadopoulos N, Gale NW, Thurston G, Yancopoulos GD, et al. Delta-like ligand 4 (Dll4) is induced by VEGF as a negative regulator of angiogenic sprouting. Proc Natl Acad Sci U S A. 2007;104:3219-24.
160. Suchting S, Freitas C, le Noble F, Benedito R, Breant C, Duarte A, et al. The Notch ligand Delta-like 4 negatively regulates endothelial tip cell formation and vessel branching. Proc Natl Acad Sci U S A. 2007;104:3225-30.
161. Kelly BD, Hackett SF, Hirota K, Oshima Y, Cai Z, Berg-Dixon S, et al. Cell type-specific regulation of angiogenic growth factor gene expression and induction of angiogenesis in nonischemic tissue by a constitutively active form of hypoxia-inducible factor 1. Circ Res. 2003;93:1074-81.
162. Ceradini DJ, Kulkarni AR, Callaghan MJ, Tepper OM, Bastidas N, Kleinman ME, et al. Progenitor cell trafficking is regulated by hypoxic gradients through HIF-1 induction of SDF-1. Nat Med. 2004;10:858-64.
163. Bosch-Marce M, Okuyama H, Wesley JB, Sarkar K, Kimura H, Liu YV, et al. Effects of aging and hypoxia-inducible factor-1 activity on angiogenic cell mobilization and recovery of perfusion after limb ischemia. Circ Res. 2007;101:1310-8.
164. Simon MP, Tournaire R, Pouyssegur J. The angiopoietin-2 gene of endothelial cells is up-regulated in hypoxia by a HIF binding site located in its first intron and by the central factors GATA-2 and Ets-1. J Cell Physiol. 2008;217:809-18.
165. Lu X, Kang Y. Hypoxia and hypoxia-inducible factors: master regulators of metastasis. Clin Cancer Res. 2010;16:5928-35.
166. Hung JJ, Yang MH, Hsu HS, Hsu WH, Liu JS, Wu KJ. Prognostic significance of hypoxia-inducible factor-1alpha, TWIST1 and Snail expression in resectable non-small cell lung cancer. Thorax. 2009;64:1082-9.
167. Polyak K, Weinberg RA. Transitions between epithelial and mesenchymal states: acquisition of malignant and stem cell traits. Nat Rev Cancer. 2009;9:265- 73.
168. Yang J, Mani SA, Donaher JL, Ramaswamy S, Itzykson RA, Come C, et al. Twist, a master regulator of morphogenesis, plays an essential role in tumor metastasis. Cell. 2004;117:927-39. 172
169. Thiery JP, Acloque H, Huang RY, Nieto MA. Epithelial-mesenchymal transitions in development and disease. Cell. 2009;139:871-90.
170. Turley EA, Veiseh M, Radisky DC, Bissell MJ. Mechanisms of disease: epithelial-mesenchymal transition--does cellular plasticity fuel neoplastic progression? Nat Clin Pract Oncol. 2008;5:280-90.
171. Luo D, Wang J, Li J, Post M. Mouse snail is a target gene for HIF. Mol Cancer Res. 2011;9:234-45.
172. Yang MH, Wu MZ, Chiou SH, Chen PM, Chang SY, Liu CJ, et al. Direct regulation of TWIST by HIF-1alpha promotes metastasis. Nat Cell Biol. 2008;10:295-305.
173. Cannito S, Novo E, Compagnone A, Valfre di Bonzo L, Busletta C, Zamara E, et al. Redox mechanisms switch on hypoxia-dependent epithelial- mesenchymal transition in cancer cells. Carcinogenesis. 2008;29:2267-78.
174. Krishnamachary B, Zagzag D, Nagasawa H, Rainey K, Okuyama H, Baek JH, et al. Hypoxia-inducible factor-1-dependent repression of E-cadherin in von Hippel-Lindau tumor suppressor-null renal cell carcinoma mediated by TCF3, ZFHX1A, and ZFHX1B. Cancer Res. 2006;66:2725-31.
175. Semenza GL, Roth PH, Fang HM, Wang GL. Transcriptional regulation of genes encoding glycolytic enzymes by hypoxia-inducible factor 1. J Biol Chem. 1994;269:23757-63.
176. Warburg O. On the origin of cancer cells. Science. 1956;123:309-14.
177. Pollard PJ, Wortham NC, Tomlinson IP. The TCA cycle and tumorigenesis: the examples of fumarate hydratase and succinate dehydrogenase. Ann Med. 2003;35:632-9.
178. King A, Selak MA, Gottlieb E. Succinate dehydrogenase and fumarate hydratase: linking mitochondrial dysfunction and cancer. Oncogene. 2006;25:4675-82.
179. Gray LH, Conger AD, Ebert M, Hornsey S, Scott OC. The concentration of oxygen dissolved in tissues at the time of irradiation as a factor in radiotherapy. Br J Radiol. 1953;26:638-48.
180. Aebersold DM, Burri P, Beer KT, Laissue J, Djonov V, Greiner RH, et al. Expression of hypoxia-inducible factor-1alpha: a novel predictive and prognostic parameter in the radiotherapy of oropharyngeal cancer. Cancer Res. 2001;61:2911-6.
173
181. Koukourakis MI, Giatromanolaki A, Sivridis E, Simopoulos C, Turley H, Talks K, et al. Hypoxia-inducible factor (HIF1A and HIF2A), angiogenesis, and chemoradiotherapy outcome of squamous cell head-and-neck cancer. Int J Radiat Oncol Biol Phys. 2002;53:1192-202.
182. Moeller BJ, Cao Y, Li CY, Dewhirst MW. Radiation activates HIF-1 to regulate vascular radiosensitivity in tumors: role of reoxygenation, free radicals, and stress granules. Cancer Cell. 2004;5:429-41.
183. Aebersold DM, Kollar A, Beer KT, Laissue J, Greiner RH, Djonov V. Involvement of the hepatocyte growth factor/scatter factor receptor c-met and of Bcl-xL in the resistance of oropharyngeal cancer to ionizing radiation. Int J Cancer. 2001;96:41-54.
184. Unruh A, Ressel A, Mohamed HG, Johnson RS, Nadrowitz R, Richter E, et al. The hypoxia-inducible factor-1 alpha is a negative factor for tumor therapy. Oncogene. 2003;22:3213-20.
185. Semenza GL. Targeting HIF-1 for cancer therapy. Nat Rev Cancer. 2003;3:721-32.
186. Garcia-Barros M, Paris F, Cordon-Cardo C, Lyden D, Rafii S, Haimovitz- Friedman A, et al. Tumor response to radiotherapy regulated by endothelial cell apoptosis. Science. 2003;300:1155-9.
187. Knowles HJ, Raval RR, Harris AL, Ratcliffe PJ. Effect of ascorbate on the activity of hypoxia-inducible factor in cancer cells. Cancer Res. 2003;63:1764-8.
188. Comerford KM, Wallace TJ, Karhausen J, Louis NA, Montalto MC, Colgan SP. Hypoxia-inducible factor-1-dependent regulation of the multidrug resistance (MDR1) gene. Cancer Res. 2002;62:3387-94.
189. Berget SM, Moore C, Sharp PA. Spliced segments at the 5' terminus of adenovirus 2 late mRNA. Proceedings of the National Academy of Sciences of the United States of America. 1977;74:3171-5.
190. Chow LT, Gelinas RE, Broker TR, Roberts RJ. An amazing sequence arrangement at the 5' ends of adenovirus 2 messenger RNA. Cell. 1977;12:1-8.
191. Darnell JE, Jr. Implications of RNA-RNA splicing in evolution of eukaryotic cells. Science. 1978;202:1257-60.
192. Knapp G, Beckmann JS, Johnson PF, Fuhrman SA, Abelson J. Transcription and processing of intervening sequences in yeast tRNA genes. Cell. 1978;14:221-36.
174
193. Will CL, Luhrmann R. Spliceosomal UsnRNP biogenesis, structure and function. Curr Opin Cell Biol. 2001;13:290-301.
194. Nilsen TW. The spliceosome: no assembly required? Molecular cell. 2002;9:8-9.
195. Staley JP, Guthrie C. Mechanical devices of the spliceosome: motors, clocks, springs, and things. Cell. 1998;92:315-26.
196. Black DL. Mechanisms of alternative pre-messenger RNA splicing. Annu Rev Biochem. 2003;72:291-336.
197. Veretnik S, Wills C, Youkharibache P, Valas RE, Bourne PE. Sm/Lsm genes provide a glimpse into the early evolution of the spliceosome. PLoS Comput Biol. 2009;5:e1000315.
198. Das R, Zhou Z, Reed R. Functional association of U2 snRNP with the ATP-independent spliceosomal complex E. Molecular cell. 2000;5:779-87.
199. Kim E, Goren A, Ast G. Alternative splicing: current perspectives. Bioessays. 2008;30:38-47.
200. Modrek B, Lee C. A genomic view of alternative splicing. Nat Genet. 2002;30:13-9.
201. Jumaa H, Wei G, Nielsen PJ. Blastocyst formation is blocked in mouse embryos lacking the splicing factor SRp20. Curr Biol. 1999;9:899-902.
202. Xu X, Fu XD. Conditional knockout mice to study alternative splicing in vivo. Methods. 2005;37:387-92.
203. Wang HY, Xu X, Ding JH, Bermingham JR, Jr., Fu XD. SC35 plays a role in T cell development and alternative splicing of CD45. Molecular cell. 2001;7:331-42.
204. Roshon MJ, Ruley HE. Hypomorphic mutation in hnRNP U results in post- implantation lethality. Transgenic Res. 2005;14:179-92.
205. Williamson DJ, Banik-Maiti S, DeGregori J, Ruley HE. hnRNP C is required for postimplantation mouse development but Is dispensable for cell viability. Mol Cell Biol. 2000;20:4094-105.
206. Jensen KB, Dredge BK, Stefani G, Zhong R, Buckanovich RJ, Okano HJ, et al. Nova-1 regulates neuron-specific alternative splicing and is essential for neuronal viability. Neuron. 2000;25:359-71. 175
207. Cartegni L, Krainer AR. Disruption of an SF2/ASF-dependent exonic splicing enhancer in SMN2 causes spinal muscular atrophy in the absence of SMN1. Nat Genet. 2002;30:377-84.
208. Kashima T, Rao N, David CJ, Manley JL. hnRNP A1 functions with specificity in repression of SMN2 exon 7 splicing. Hum Mol Genet. 2007;16:3149- 59.
209. Liu F, Gong CX. Tau exon 10 alternative splicing and tauopathies. Mol Neurodegener. 2008;3:8.
210. Buratti E, Brindisi A, Pagani F, Baralle FE. Nuclear factor TDP-43 binds to the polymorphic TG repeats in CFTR intron 8 and causes skipping of exon 9: a functional link with disease penetrance. Am J Hum Genet. 2004;74:1322-5.
211. Kalnina Z, Zayakin P, Silina K, Line A. Alterations of pre-mRNA splicing in cancer. Genes, chromosomes & cancer. 2005;42:342-57.
212. Krawczak M, Reiss J, Cooper DN. The mutational spectrum of single base-pair substitutions in mRNA splice junctions of human genes: causes and consequences. Hum Genet. 1992;90:41-54.
213. Lopez-Bigas N, Audit B, Ouzounis C, Parra G, Guigo R. Are splicing mutations the most frequent cause of hereditary disease? FEBS letters. 2005;579:1900-3.
214. Karni R, de Stanchina E, Lowe SW, Sinha R, Mu D, Krainer AR. The gene encoding the splicing factor SF2/ASF is a proto-oncogene. Nat Struct Mol Biol. 2007;14:185-93.
215. Boise LH, Gonzalez-Garcia M, Postema CE, Ding L, Lindsten T, Turka LA, et al. bcl-x, a bcl-2-related gene that functions as a dominant regulator of apoptotic cell death. Cell. 1993;74:597-608.
216. Olopade OI, Adeyanju MO, Safa AR, Hagos F, Mick R, Thompson CB, et al. Overexpression of BCL-x protein in primary breast cancer is associated with high tumor grade and nodal metastases. Cancer J Sci Am. 1997;3:230-7.
217. Takehara T, Liu X, Fujimoto J, Friedman SL, Takahashi H. Expression and role of Bcl-xL in human hepatocellular carcinomas. Hepatology. 2001;34:55- 61.
218. Xerri L, Parc P, Brousset P, Schlaifer D, Hassoun J, Reed JC, et al. Predominant expression of the long isoform of Bcl-x (Bcl-xL) in human lymphomas. Br J Haematol. 1996;92:900-6.
219. Mazurek S, Boschek CB, Hugo F, Eigenbrodt E. Pyruvate kinase type M2 and its role in tumor growth and spreading. Semin Cancer Biol. 2005;15:300-8. 176
220. Christofk HR, Vander Heiden MG, Harris MH, Ramanathan A, Gerszten RE, Wei R, et al. The M2 splice isoform of pyruvate kinase is important for cancer metabolism and tumour growth. Nature. 2008;452:230-3.
221. Paronetto MP, Cappellari M, Busa R, Pedrotti S, Vitali R, Comstock C, et al. Alternative splicing of the cyclin D1 proto-oncogene is regulated by the RNA- binding protein Sam68. Cancer research. 2010;70:229-39.
222. David CJ, Manley JL. Alternative pre-mRNA splicing regulation in cancer: pathways and programs unhinged. Genes & development. 2010;24:2343-64.
223. Warzecha CC, Sato TK, Nabet B, Hogenesch JB, Carstens RP. ESRP1 and ESRP2 are epithelial cell-type-specific regulators of FGFR2 splicing. Molecular cell. 2009;33:591-601.
224. Bates DO, Cui TG, Doughty JM, Winkler M, Sugiono M, Shields JD, et al. VEGF165b, an inhibitory splice variant of vascular endothelial growth factor, is down-regulated in renal cell carcinoma. Cancer research. 2002;62:4123-31.
225. Oltean S, Bates DO. Hallmarks of alternative splicing in cancer. Oncogene. 2013, In Press.
226. Long JC, Caceres JF. The SR protein family of splicing factors: master regulators of gene expression. Biochem J. 2009;417:15-27.
227. Dreyfuss G, Kim VN, Kataoka N. Messenger-RNA-binding proteins and the messages they carry. Nat Rev Mol Cell Biol. 2002;3:195-205.
228. Edmond V, Moysan E, Khochbin S, Matthias P, Brambilla C, Brambilla E, et al. Acetylation and phosphorylation of SRSF2 control cell fate decision in response to cisplatin. The EMBO journal. 2011;30:510-23.
229. Sinha R, Allemand E, Zhang Z, Karni R, Myers MP, Krainer AR. Arginine methylation controls the subcellular localization and functions of the oncoprotein splicing factor SF2/ASF. Mol Cell Biol. 2010;30:2762-74.
230. Wagner SA, Beli P, Weinert BT, Nielsen ML, Cox J, Mann M, et al. A proteome-wide, quantitative survey of in vivo ubiquitylation sites reveals widespread regulatory roles. Mol Cell Proteomics. 2011;10:M111 013284.
231. Vertegaal AC, Ogg SC, Jaffray E, Rodriguez MS, Hay RT, Andersen JS, et al. A proteomic study of SUMO-2 target proteins. The Journal of biological chemistry. 2004;279:33791-8.
232. Mermoud JE, Cohen PT, Lamond AI. Regulation of mammalian spliceosome assembly by a protein phosphorylation mechanism. The EMBO journal. 1994;13:5679-88. 177
233. Colwill K, Pawson T, Andrews B, Prasad J, Manley JL, Bell JC, et al. The Clk/Sty protein kinase phosphorylates SR splicing factors and regulates their intranuclear distribution. The EMBO journal. 1996;15:265-75.
234. Blaustein M, Pelisch F, Tanos T, Munoz MJ, Wengier D, Quadrana L, et al. Concerted regulation of nuclear and cytoplasmic activities of SR proteins by AKT. Nat Struct Mol Biol. 2005;12:1037-44.
235. Patel NA, Kaneko S, Apostolatos HS, Bae SS, Watson JE, Davidowitz K, et al. Molecular and genetic studies imply Akt-mediated signaling promotes protein kinase CbetaII alternative splicing via phosphorylation of serine/arginine- rich splicing factor SRp40. The Journal of biological chemistry. 2005;280:14302- 9.
236. Rossi F, Labourier E, Forne T, Divita G, Derancourt J, Riou JF, et al. Specific phosphorylation of SR proteins by mammalian DNA topoisomerase I. Nature. 1996;381:80-2.
237. Aranda S, Laguna A, de la Luna S. DYRK family of protein kinases: evolutionary relationships, biochemical properties, and functional roles. FASEB journal : official publication of the Federation of American Societies for Experimental Biology. 2011;25:449-62.
238. de Graaf K, Hekerman P, Spelten O, Herrmann A, Packman LC, Bussow K, et al. Characterization of cyclin L2, a novel cyclin with an arginine/serine-rich domain: phosphorylation by DYRK1A and colocalization with splicing factors. The Journal of biological chemistry. 2004;279:4612-24.
239. Ding JH, Zhong XY, Hagopian JC, Cruz MM, Ghosh G, Feramisco J, et al. Regulated cellular partitioning of SR protein-specific kinases in mammalian cells. Molecular biology of the cell. 2006;17:876-85.
240. Wang HY, Arden KC, Bermingham JR, Jr., Viars CS, Lin W, Boyer AD, et al. Localization of serine kinases, SRPK1 (SFRSK1) and SRPK2 (SFRSK2), specific for the SR family of splicing factors in mouse and human chromosomes. Genomics. 1999;57:310-5.
241. Gui JF, Tronchere H, Chandler SD, Fu XD. Purification and characterization of a kinase specific for the serine- and arginine-rich pre-mRNA splicing factors. Proceedings of the National Academy of Sciences of the United States of America. 1994;91:10824-8.
242. Yun CY, Fu XD. Conserved SR protein kinase functions in nuclear import and its action is counteracted by arginine methylation in Saccharomyces cerevisiae. J Cell Biol. 2000;150:707-18. 178
243. Yun CY, Velazquez-Dones AL, Lyman SK, Fu XD. Phosphorylation- dependent and -independent nuclear import of RS domain-containing splicing factors and regulators. The Journal of biological chemistry. 2003;278:18050-5.
244. Mermoud JE, Cohen P, Lamond AI. Ser/Thr-specific protein phosphatases are required for both catalytic steps of pre-mRNA splicing. Nucleic Acids Res. 1992;20:5263-9.
245. Naro C, Sette C. Phosphorylation-Mediated Regulation of Alternative Splicing in Cancer. Int J Cell Biol. 2013;2013:151839.
246. Huang Y, Gattoni R, Stevenin J, Steitz JA. SR splicing factors serve as adapter proteins for TAP-dependent mRNA export. Molecular cell. 2003;11:837- 43.
247. Huang Y, Steitz JA. SRprises along a messenger's journey. Molecular cell. 2005;17:613-5.
248. Sanford JR, Gray NK, Beckmann K, Caceres JF. A novel role for shuttling SR proteins in mRNA translation. Genes & development. 2004;18:755-68.
249. Padgett RA, Hardy SF, Sharp PA. Splicing of adenovirus RNA in a cell- free transcription system. Proceedings of the National Academy of Sciences of the United States of America. 1983;80:5230-4.
250. Krainer AR, Maniatis T, Ruskin B, Green MR. Normal and mutant human beta-globin pre-mRNAs are faithfully and efficiently spliced in vitro. Cell. 1984;36:993-1005.
251. Hernandez N, Keller W. Splicing of in vitro synthesized messenger RNA precursors in HeLa cell extracts. Cell. 1983;35:89-99.
252. Green MR, Maniatis T, Melton DA. Human beta-globin pre-mRNA synthesized in vitro is accurately spliced in Xenopus oocyte nuclei. Cell. 1983;32:681-94.
253. Osheim YN, Miller OL, Jr., Beyer AL. RNP particles at splice junction sequences on Drosophila chorion transcripts. Cell. 1985;43:143-51.
254. Pandya-Jones A, Black DL. Co-transcriptional splicing of constitutive and alternative exons. Rna. 2009;15:1896-908.
255. Listerman I, Sapra AK, Neugebauer KM. Cotranscriptional coupling of splicing factor recruitment and precursor messenger RNA splicing in mammalian cells. Nat Struct Mol Biol. 2006;13:815-22. 179
256. McCracken S, Fong N, Yankulov K, Ballantyne S, Pan G, Greenblatt J, et al. The C-terminal domain of RNA polymerase II couples mRNA processing to transcription. Nature. 1997;385:357-61.
257. Cramer P, Pesce CG, Baralle FE, Kornblihtt AR. Functional association between promoter structure and transcript alternative splicing. Proceedings of the National Academy of Sciences of the United States of America. 1997;94:11456- 60.
258. Dowhan DH, Hong EP, Auboeuf D, Dennis AP, Wilson MM, Berget SM, et al. Steroid hormone receptor coactivation and alternative RNA splicing by U2AF65-related proteins CAPERalpha and CAPERbeta. Molecular cell. 2005;17:429-39.
259. Batsche E, Yaniv M, Muchardt C. The human SWI/SNF subunit Brm is a regulator of alternative splicing. Nat Struct Mol Biol. 2006;13:22-9.
260. Fong YW, Zhou Q. Stimulatory effect of splicing factors on transcriptional elongation. Nature. 2001;414:929-33.
261. Furger A, O'Sullivan JM, Binnie A, Lee BA, Proudfoot NJ. Promoter proximal splice sites enhance transcription. Genes & development. 2002;16:2792-9.
262. Misteli T, Caceres JF, Spector DL. The dynamics of a pre-mRNA splicing factor in living cells. Nature. 1997;387:523-7.
263. Mortillaro MJ, Blencowe BJ, Wei X, Nakayasu H, Du L, Warren SL, et al. A hyperphosphorylated form of the large subunit of RNA polymerase II is associated with splicing complexes and the nuclear matrix. Proceedings of the National Academy of Sciences of the United States of America. 1996;93:8253-7.
264. Robert F, Blanchette M, Maes O, Chabot B, Coulombe B. A human RNA polymerase II-containing complex associated with factors necessary for spliceosome assembly. The Journal of biological chemistry. 2002;277:9302-6.
265. Caceres JF, Kornblihtt AR. Alternative splicing: multiple control mechanisms and involvement in human disease. Trends Genet. 2002;18:186-93.
266. Sanchez-Alvarez M, Goldstrohm AC, Garcia-Blanco MA, Sune C. Human transcription elongation factor CA150 localizes to splicing factor-rich nuclear speckles and assembles transcription and splicing components into complexes through its amino and carboxyl regions. Mol Cell Biol. 2006;26:4998-5014.
267. Davies RC, Calvio C, Bratt E, Larsson SH, Lamond AI, Hastie ND. WT1 interacts with the splicing factor U2AF65 in an isoform-dependent manner and can be incorporated into spliceosomes. Genes & development. 1998;12:3217-25. 180
268. Price DH. P-TEFb, a cyclin-dependent kinase controlling elongation by RNA polymerase II. Mol Cell Biol. 2000;20:2629-34.
269. Ni Z, Schwartz BE, Werner J, Suarez JR, Lis JT. Coordination of transcription, RNA processing, and surveillance by P-TEFb kinase on heat shock genes. Molecular cell. 2004;13:55-65.
270. Mott BT, Tanega C, Shen M, Maloney DJ, Shinn P, Leister W, et al. Evaluation of substituted 6-arylquinazolin-4-amines as potent and selective inhibitors of cdc2-like kinases (Clk). Bioorg Med Chem Lett. 2009;19:6700-5.
271. de la Mata M, Alonso CR, Kadener S, Fededa JP, Blaustein M, Pelisch F, et al. A slow RNA polymerase II affects alternative splicing in vivo. Molecular cell. 2003;12:525-32.
272. Guil S, Long JC, Caceres JF. hnRNP A1 relocalization to the stress granules reflects a role in the stress response. Mol Cell Biol. 2006;26:5744-58.
273. Shi Y, Manley JL. A complex signaling pathway regulates SRp38 phosphorylation and pre-mRNA splicing in response to heat shock. Mol Cell. 2007;28:79-90.
274. Zhong XY, Ding JH, Adams JA, Ghosh G, Fu XD. Regulation of SR protein phosphorylation and alternative splicing by modulating kinetic interactions of SRPK1 with molecular chaperones. Genes Dev. 2009;23:482-95.
275. Munoz MJ, Perez Santangelo MS, Paronetto MP, de la Mata M, Pelisch F, Boireau S, et al. DNA damage regulates alternative splicing through inhibition of RNA polymerase II elongation. Cell. 2009;137:708-20.
276. Tacconelli A, Farina AR, Cappabianca L, Desantis G, Tessitore A, Vetuschi A, et al. TrkA alternative splicing: a regulated tumor-promoting switch in human neuroblastoma. Cancer Cell. 2004;6:347-60.
277. Elias AP, Dias S. Microenvironment Changes (in pH) Affect VEGF Alternative Splicing. Cancer Microenviron. 2008;1:131-9.
278. Hirschfeld M, zur Hausen A, Bettendorf H, Jager M, Stickeler E. Alternative splicing of Cyr61 is regulated by hypoxia and significantly changed in breast cancer. Cancer research. 2009;69:2082-90.
279. Tacconelli A, Farina AR, Cappabianca L, Gulino A, Mackay AR. TrkAIII. A novel hypoxia-regulated alternative TrkA splice variant of potential physiological and pathological importance. Cell Cycle. 2005;4:8-9.
280. Gang H, Hai Y, Dhingra R, Gordon JW, Yurkova N, Aviv Y, et al. A novel hypoxia-inducible spliced variant of mitochondrial death gene Bnip3 promotes survival of ventricular myocytes. Circ Res. 2011;108:1084-92. 181
281. Weigand JE, Boeckel JN, Gellert P, Dimmeler S. Hypoxia-induced alternative splicing in endothelial cells. PLoS One. 2012;7:e42697.
282. Bibel M, Barde YA. Neurotrophins: key regulators of cell fate and cell shape in the vertebrate nervous system. Genes & development. 2000;14:2919- 37.
283. Oppenheim RW. Cell death during development of the nervous system. Annu Rev Neurosci. 1991;14:453-501.
284. Hempstead BL, Martin-Zanca D, Kaplan DR, Parada LF, Chao MV. High- affinity NGF binding requires coexpression of the trk proto-oncogene and the low-affinity NGF receptor. Nature. 1991;350:678-83.
285. Perbal B. CCN proteins: multifunctional signalling regulators. Lancet. 2004;363:62-4.
286. Park SJ, Kim JG, Son TG, Yi JM, Kim ND, Yang K, et al. The histone demethylase JMJD1A regulates adrenomedullin-mediated cell proliferation in hepatocellular carcinoma under hypoxia. Biochem Biophys Res Commun. 2013;434:722-7.
287. Ouafik L, Sauze S, Boudouresque F, Chinot O, Delfino C, Fina F, et al. Neutralization of adrenomedullin inhibits the growth of human glioblastoma cell lines in vitro and suppresses tumor xenograft growth in vivo. Am J Pathol. 2002;160:1279-92.
288. Martinez A. Biology of adrenomedullin. Introduction. Microsc Res Tech. 2002;57:1-2.
289. Kawakami Y. [Adrenomedullin antagonist suppresses in vivo proliferation of cancer cells in SCID mice via angiogenesis inhibition]. Hokkaido Igaku Zasshi. 2005;80:575-83.
290. Tsuchiya K, Hida K, Hida Y, Muraki C, Ohga N, Akino T, et al. Adrenomedullin antagonist suppresses tumor formation in renal cell carcinoma through inhibitory effects on tumor endothelial cells and endothelial progenitor mobilization. Int J Oncol. 2010;36:1379-86.
291. Deville JL, Bartoli C, Berenguer C, Fernandez-Sauze S, Kaafarani I, Delfino C, et al. Expression and role of adrenomedullin in renal tumors and value of its mRNA levels as prognostic factor in clear-cell renal carcinoma. Int J Cancer. 2009;125:2307-15.
292. Oehler MK, Norbury C, Hague S, Rees MC, Bicknell R. Adrenomedullin inhibits hypoxic cell death by upregulation of Bcl-2 in endometrial cancer cells: a possible promotion mechanism for tumour growth. Oncogene. 2001;20:2937-45. 182
293. Oehler MK, Hague S, Rees MC, Bicknell R. Adrenomedullin promotes formation of xenografted endometrial tumors by stimulation of autocrine growth and angiogenesis. Oncogene. 2002;21:2815-21.
294. Nikitenko LL, MacKenzie IZ, Rees MC, Bicknell R. Adrenomedullin is an autocrine regulator of endothelial growth in human endometrium. Mol Hum Reprod. 2000;6:811-9.
295. Eto T, Samson WK. Adrenomedullin and proadrenomedullin N-terminal 20 peptide: vasodilatory peptides with multiple cardiovascular and endocrine actions. Trends Endocrinol Metab. 2001;12:91-3.
296. Nakamura M, Yoshida H, Hiramori K. Comparison of vasodilator potency of adrenomedulling and proadrenomedullin N-terminal 20 peptide in human. Life sciences. 1999;65:2151-6.
297. Martinez A, Zudaire E, Portal-Nunez S, Guedez L, Libutti SK, Stetler- Stevenson WG, et al. Proadrenomedullin NH2-terminal 20 peptide is a potent angiogenic factor, and its inhibition results in reduction of tumor growth. Cancer Res. 2004;64:6489-94.
298. Martinez A, Hodge DL, Garayoa M, Young HA, Cuttitta F. Alternative splicing of the proadrenomedullin gene results in differential expression of gene products. J Mol Endocrinol. 2001;27:31-41.
299. Lejeune F, Maquat LE. Mechanistic links between nonsense-mediated mRNA decay and pre-mRNA splicing in mammalian cells. Curr Opin Cell Biol. 2005;17:309-15.
300. Wong JJ, Ritchie W, Ebner OA, Selbach M, Wong JW, Huang Y, et al. Orchestrated intron retention regulates normal granulocyte differentiation. Cell. 2013;154:583-95.
301. Kitamuro T, Takahashi K, Nakayama M, Murakami O, Hida W, Shirato K, et al. Induction of adrenomedullin during hypoxia in cultured human glioblastoma cells. J Neurochem. 2000;75:1826-33.
302. Hu CJ, Wang LY, Chodosh LA, Keith B, Simon MC. Differential roles of hypoxia-inducible factor 1alpha (HIF-1alpha) and HIF-2alpha in hypoxic gene regulation. Mol Cell Biol. 2003;23:9361-74.
303. Auboeuf D, Dowhan DH, Kang YK, Larkin K, Lee JW, Berget SM, et al. Differential recruitment of nuclear receptor coactivators may determine alternative RNA splice site choice in target genes. Proceedings of the National Academy of Sciences of the United States of America. 2004;101:2270-4. 183
304. Auboeuf D, Dowhan DH, Li X, Larkin K, Ko L, Berget SM, et al. CoAA, a nuclear receptor coactivator protein at the interface of transcriptional coactivation and RNA splicing. Mol Cell Biol. 2004;24:442-53.
305. Cheng D, Cote J, Shaaban S, Bedford MT. The arginine methyltransferase CARM1 regulates the coupling of transcription and mRNA processing. Molecular cell. 2007;25:71-83.
306. Pawlus MR, Wang L, Ware K, Hu CJ. Upstream stimulatory factor 2 and hypoxia-inducible factor 2alpha (HIF2alpha) cooperatively activate HIF2 target genes during hypoxia. Mol Cell Biol. 2012;32:4595-610.
307. Yang BC, Lippton H, Gumusel B, Hyman A, Mehta JL. Adrenomedullin dilates rat pulmonary artery rings during hypoxia: role of nitric oxide and vasodilator prostaglandins. J Cardiovasc Pharmacol. 1996;28:458-62.
308. LaCava J, Houseley J, Saveanu C, Petfalski E, Thompson E, Jacquier A, et al. RNA degradation by the exosome is promoted by a nuclear polyadenylation complex. Cell. 2005;121:713-24.
309. Vanacova S, Wolf J, Martin G, Blank D, Dettwiler S, Friedlein A, et al. A new yeast poly(A) polymerase complex involved in RNA quality control. PLoS biology. 2005;3:e189.
310. Wyers F, Rougemaille M, Badis G, Rousselle JC, Dufour ME, Boulay J, et al. Cryptic pol II transcripts are degraded by a nuclear quality control pathway involving a new poly(A) polymerase. Cell. 2005;121:725-37.
311. Rosonina E, Ip JY, Calarco JA, Bakowski MA, Emili A, McCracken S, et al. Role for PSF in mediating transcriptional activator-dependent stimulation of pre- mRNA processing in vivo. Mol Cell Biol. 2005;25:6734-46.
312. Nogues G, Kadener S, Cramer P, Bentley D, Kornblihtt AR. Transcriptional activators differ in their abilities to control alternative splicing. J Biol Chem. 2002;277:43110-4.
313. Cramer P, Caceres JF, Cazalla D, Kadener S, Muro AF, Baralle FE, et al. Coupling of transcription with alternative splicing: RNA pol II promoters modulate SF2/ASF and 9G8 effects on an exonic splicing enhancer. Mol Cell. 1999;4:251- 8.
314. Auboeuf D, Honig A, Berget SM, O'Malley BW. Coordinate regulation of transcription and splicing by steroid receptor coregulators. Science. 2002;298:416-9.
315. Kornblihtt AR. Coupling transcription and alternative splicing. Adv Exp Med Biol. 2007;623:175-89. 184
316. Kornblihtt AR, de la Mata M, Fededa JP, Munoz MJ, Nogues G. Multiple links between transcription and splicing. Rna. 2004;10:1489-98.
317. Perales R, Bentley D. "Cotranscriptionality": the transcription elongation complex as a nexus for nuclear transactions. Mol Cell. 2009;36:178-91.
318. Sena JA, Wang L, Hu CJ. BRG1 and BRM chromatin-remodeling complexes regulate the hypoxia response by acting as coactivators for a subset of hypoxia-inducible transcription factor target genes. Mol Cell Biol. 2013;33:3849-63.
319. Braunschweig U, Gueroussov S, Plocik AM, Graveley BR, Blencowe BJ. Dynamic integration of splicing within gene regulatory pathways. Cell. 2013;152:1252-69.
320. Finger EC, Giaccia AJ. Hypoxia, inflammation, and the tumor microenvironment in metastatic disease. Cancer metastasis reviews. 2010;29:285-93.
321. Kaelin WG, Jr. Cancer and altered metabolism: potential importance of hypoxia-inducible factor and 2-oxoglutarate-dependent dioxygenases. Cold Spring Harbor symposia on quantitative biology. 2011;76:335-45.
322. Keith B, Johnson RS, Simon MC. HIF1alpha and HIF2alpha: sibling rivalry in hypoxic tumour growth and progression. Nat Rev Cancer. 2012;12:9-22.
323. Semenza GL. Hypoxia-inducible factors in physiology and medicine. Cell. 2012;148:399-408.
324. Hu CJ, Iyer S, Sataur A, Covello KL, Chodosh LA, Simon MC. Differential regulation of the transcriptional activities of hypoxia-inducible factor 1 alpha (HIF- 1alpha) and HIF-2alpha in stem cells. Mol Cell Biol. 2006;26:3514-26.
325. Wang V, Davis DA, Haque M, Huang LE, Yarchoan R. Differential gene up-regulation by hypoxia-inducible factor-1alpha and hypoxia-inducible factor- 2alpha in HEK293T cells. Cancer research. 2005;65:3299-306.
326. Sowter HM, Raval RR, Moore JW, Ratcliffe PJ, Harris AL. Predominant role of hypoxia-inducible transcription factor (Hif)-1alpha versus Hif-2alpha in regulation of the transcriptional response to hypoxia. Cancer research. 2003;63:6130-4.
327. Elvidge GP, Glenny L, Appelhoff RJ, Ratcliffe PJ, Ragoussis J, Gleadle JM. Concordant regulation of gene expression by hypoxia and 2-oxoglutarate- dependent dioxygenase inhibition: the role of HIF-1alpha, HIF-2alpha, and other pathways. The Journal of biological chemistry. 2006;281:15215-26. 185
328. Wang ET, Sandberg R, Luo S, Khrebtukova I, Zhang L, Mayr C, et al. Alternative isoform regulation in human tissue transcriptomes. Nature. 2008;456:470-6.
329. Hang X, Li P, Li Z, Qu W, Yu Y, Li H, et al. Transcription and splicing regulation in human umbilical vein endothelial cells under hypoxic stress conditions by exon array. BMC genomics. 2009;10:126.
330. Warnecke C, Zaborowska Z, Kurreck J, Erdmann VA, Frei U, Wiesener M, et al. Differentiating the functional role of hypoxia-inducible factor (HIF)-1alpha and HIF-2alpha (EPAS-1) by the use of RNA interference: erythropoietin is a HIF-2alpha target gene in Hep3B and Kelly cells. FASEB journal : official publication of the Federation of American Societies for Experimental Biology. 2004;18:1462-4.
331. Huang da W, Sherman BT, Lempicki RA. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res. 2009;37:1-13.
332. Pawlus MR, Wang L, Hu CJ. STAT3 and HIF1alpha cooperatively activate HIF1 target genes in MDA-MB-231 and RCC4 cells. Oncogene. 2014:1670-9.
333. Barathova M, Takacova M, Holotnakova T, Gibadulinova A, Ohradanova A, Zatovicova M, et al. Alternative splicing variant of the hypoxia marker carbonic anhydrase IX expressed independently of hypoxia and tumour phenotype. Br J Cancer. 2008;98:129-36.
334. Turcotte S, Desrosiers RR, Beliveau R. HIF-1alpha mRNA and protein upregulation involves Rho GTPase expression during hypoxia in renal cell carcinoma. Journal of cell science. 2003;116:2247-60.
335. Plowman SJ, Berry RL, Bader SA, Luo F, Arends MJ, Harrison DJ, et al. K-ras 4A and 4B are co-expressed widely in human tissues, and their ratio is altered in sporadic colorectal cancer. J Exp Clin Cancer Res. 2006;25:259-67.
336. Abubaker J, Bavi P, Al-Haqawi W, Sultana M, Al-Harbi S, Al-Sanea N, et al. Prognostic significance of alterations in KRAS isoforms KRAS-4A/4B and KRAS mutations in colorectal carcinoma. The Journal of pathology. 2009;219:435-45.
337. Luo F, Ye H, Hamoudi R, Dong G, Zhang W, Patek CE, et al. K-ras exon 4A has a tumour suppressor effect on carcinogen-induced murine colonic adenoma formation. The Journal of pathology. 2010;220:542-50.
338. Kikuchi Y, Kakeya T, Nakajima O, Sakai A, Ikeda K, Yamaguchi N, et al. Hypoxia induces expression of a GPI-anchorless splice variant of the prion protein. The FEBS journal. 2008;275:2965-76. 186
339. Sena JA, Wang L, Pawlus MR, Hu CJ. HIFs Enhance the Transcriptional Activation and Splicing of Adrenomedullin. Mol Cancer Res. 2014.
340. Schodel J, Oikonomopoulos S, Ragoussis J, Pugh CW, Ratcliffe PJ, Mole DR. High-resolution genome-wide mapping of HIF-binding sites by ChIP-seq. Blood. 2011;117:e207-17.
341. Jodelka FM, Ebert AD, Duelli DM, Hastings ML. A feedback loop regulates splicing of the spinal muscular atrophy-modifying gene, SMN2. Hum Mol Genet. 2010;19:4906-17.
342. Munding EM, Shiue L, Katzman S, Donohue JP, Ares M, Jr. Competition between pre-mRNAs for the splicing machinery drives global regulation of splicing. Molecular cell. 2013;51:338-48.
343. Bechara E, Valcarcel J. Competition by the masses. Molecular cell. 2013;51:279-80.
344. Cordin O, Hahn D, Beggs JD. Structure, function and regulation of spliceosomal RNA helicases. Curr Opin Cell Biol. 2012;24:431-8.
345. Jurica MS, Moore MJ. Pre-mRNA splicing: awash in a sea of proteins. Molecular cell. 2003;12:5-14.
346. Colley A, Beggs JD, Tollervey D, Lafontaine DL. Dhr1p, a putative DEAH- box RNA helicase, is associated with the box C+D snoRNP U3. Mol Cell Biol. 2000;20:7238-46.
347. Cowper AE, Caceres JF, Mayeda A, Screaton GR. Serine-arginine (SR) protein-like factors that antagonize authentic SR proteins and regulate alternative splicing. The Journal of biological chemistry. 2001;276:48908-14.
348. Itoh S, Itoh F, Goumans MJ, Ten Dijke P. Signaling of transforming growth factor-beta family members through Smad proteins. Eur J Biochem. 2000;267:6954-67.
349. Moustakas A, Souchelnytskyi S, Heldin CH. Smad regulation in TGF-beta signal transduction. Journal of cell science. 2001;114:4359-69.
350. Sun Y, Ding L, Zhang H, Han J, Yang X, Yan J, et al. Potentiation of Smad-mediated transcriptional activation by the RNA-binding protein RBPMS. Nucleic Acids Res. 2006;34:6314-26.
351. Warzecha CC, Jiang P, Amirikian K, Dittmar KA, Lu H, Shen S, et al. An ESRP-regulated splicing programme is abrogated during the epithelial- mesenchymal transition. The EMBO journal. 2010;29:3286-300. 187
352. Lamouille S, Xu J, Derynck R. Molecular mechanisms of epithelial- mesenchymal transition. Nat Rev Mol Cell Biol. 2014;15:178-96.
353. Ao JE, Kuang LH, Zhou Y, Zhao R, Yang CM. Hypoxia-inducible factor 1 regulated ARC expression mediated hypoxia induced inactivation of the intrinsic death pathway in p53 deficient human colon cancer cells. Biochemical and biophysical research communications. 2012;420:913-7.
354. Stoss O, Schwaiger FW, Cooper TA, Stamm S. Alternative splicing determines the intracellular localization of the novel nuclear protein Nop30 and its interaction with the splicing factor SRp30c. The Journal of biological chemistry. 1999;274:10951-62.
355. Di Domenico M, Ricciardi C, Fusco A, Pierantoni GM. Anti-VEGF therapy in breast and lung mouse models of cancers. J Biomed Biotechnol. 2011;2011:947928.
356. Miyajima K, Takekoshi S, Itoh J, Kakimoto K, Miyakoshi T, Osamura RY. Inhibitory effects of anti-VEGF antibody on the growth and angiogenesis of estrogen-induced pituitary prolactinoma in Fischer 344 Rats: animal model of VEGF-targeted therapy for human endocrine tumors. Acta Histochem Cytochem. 2010;43:33-44.
357. Whitehurst B, Flister MJ, Bagaitkar J, Volk L, Bivens CM, Pickett B, et al. Anti-VEGF-A therapy reduces lymphatic vessel density and expression of VEGFR-3 in an orthotopic breast tumor model. Int J Cancer. 2007;121:2181-91.
358. Escudier B. Anti-VEGF therapy for renal cell carcinoma. Clin Adv Hematol Oncol. 2007;5:530-1.
359. Muller YA, Li B, Christinger HW, Wells JA, Cunningham BC, de Vos AM. Vascular endothelial growth factor: crystal structure and functional mapping of the kinase domain receptor binding site. Proc Natl Acad Sci U S A. 1997;94:7192-7.
360. Troiani T, Martinelli E, Orditura M, De Vita F, Ciardiello F, Morgillo F. Beyond bevacizumab: new anti-VEGF strategies in colorectal cancer. Expert opinion on investigational drugs. 2012;21:949-59.
361. Miura K, Fujibuchi W, Unno M. Splice isoforms as therapeutic targets for colorectal cancer. Carcinogenesis. 2012;33:2311-9.
362. Bates DO, Cui TG, Doughty JM, Winkler M, Sugiono M, Shields JD, et al. VEGF165b, an inhibitory splice variant of vascular endothelial growth factor, is down-regulated in renal cell carcinoma. Cancer research. 2002;62:4123-31. 188
363. Solier S, Lansiaux A, Logette E, Wu J, Soret J, Tazi J, et al. Topoisomerase I and II inhibitors control caspase-2 pre-messenger RNA splicing in human cells. Mol Cancer Res. 2004;2:53-61.
364. Patterson LH. Rationale for the use of aliphatic N-oxides of cytotoxic anthraquinones as prodrug DNA binding agents: a new class of bioreductive agent. Cancer metastasis reviews. 1993;12:119-34.
365. Dorie MJ, Brown JM. Tumor-specific, schedule-dependent interaction between tirapazamine (SR 4233) and cisplatin. Cancer Res. 1993;53:4633-6.
366. Brown JM, Lemmon MJ. Potentiation by the Hypoxic Cytotoxin Sr4233 of Cell Killing Produced by Fractionated-Irradiation of Mouse-Tumors. Cancer research. 1990;50:7745-9.
367. McKeown SR, Hejmadi MV, McIntyre IA, McAleer JJ, Patterson LH. AQ4N: an alkylaminoanthraquinone N-oxide showing bioreductive potential and positive interaction with radiation in vivo. Br J Cancer. 1995;72:76-81.
368. Friery OP, Gallagher R, Murray MM, Hughes CM, Galligan ES, McIntyre IA, et al. Enhancement of the anti-tumour effect of cyclophosphamide by the bioreductive drugs AQ4N and tirapazamine. Br J Cancer. 2000;82:1469-73.
369. Shkreta L, Froehlich U, Paquet ER, Toutant J, Elela SA, Chabot B. Anticancer drugs affect the alternative splicing of Bcl-x and other human apoptotic genes. Mol Cancer Ther. 2008;7:1398-409.
370. Pawlus MR, Wang L, Murakami A, Dai G, Hu CJ. STAT3 or USF2 contributes to HIF target gene specificity. PLoS One.8:e72358.
371. Blau J, Xiao H, McCracken S, O'Hare P, Greenblatt J, Bentley D. Three functional classes of transcriptional activation domain. Mol Cell Biol. 1996;16:2044-55.
372. Emig D, Salomonis N, Baumbach J, Lengauer T, Conklin BR, Albrecht M. AltAnalyze and DomainGraph: analyzing and visualizing exon expression data. Nucleic Acids Res. 2010;38:W755-62.
373. Chang TY, Li YY, Jen CH, Yang TP, Lin CH, Hsu MT, et al. easyExon--a Java-based GUI tool for processing and visualization of Affymetrix exon array data. BMC bioinformatics. 2008;9:432.
374. Mazumdar J, Hickey MM, Pant DK, Durham AC, Sweet-Cordero A, Vachani A, et al. HIF-2alpha deletion promotes Kras-driven lung tumor development. Proc Natl Acad Sci U S A. 2010;107:14182-7. 189
375. Kubasiak LA, Hernandez OM, Bishopric NH, Webster KA. Hypoxia and acidosis activate cardiac myocyte death through the Bcl-2 family protein BNIP3. Proc Natl Acad Sci U S A. 2002;99:12825-30.
376. Sowter HM, Ratcliffe PJ, Watson P, Greenberg AH, Harris AL. HIF-1- dependent regulation of hypoxic induction of the cell death factors BNIP3 and NIX in human tumors. Cancer Res. 2001;61:6669-73.
377. Bruick RK. Expression of the gene encoding the proapoptotic Nip3 protein is induced by hypoxia. Proc Natl Acad Sci U S A. 2000;97:9082-7.
378. Corn PG, Ricci MS, Scata KA, Arsham AM, Simon MC, Dicker DT, et al. Mxi1 is induced by hypoxia in a HIF-1-dependent manner and protects cells from c-Myc-induced apoptosis. Cancer biology & therapy. 2005;4:1285-94.
379. Mack FA, Patel JH, Biju MP, Haase VH, Simon MC. Decreased growth of Vhl-/- fibrosarcomas is associated with elevated levels of cyclin kinase inhibitors p21 and p27. Mol Cell Biol. 2005;25:4565-78.
380. Pawlus MR, Wang L, Ware K, Hu CJ. USF2 and HIF2alpha cooperatively activate HIF2 target genes during hypoxia. Mol Cell Biol. 2012;32:4595-610.
381. Ebert BL, Bunn HF. Regulation of transcription by hypoxia requires a multiprotein complex that includes hypoxia-inducible factor 1, an adjacent transcription factor, and p300/CREB binding protein. Mol Cell Biol. 1998;18:4089- 96.
382. Ruas JL, Berchner-Pfannschmidt U, Malik S, Gradin K, Fandrey J, Roeder RG, et al. Complex regulation of the transactivation function of hypoxia-inducible factor-1 alpha by direct interaction with two distinct domains of the CREB-binding protein/p300. J Biol Chem. 2010;285:2601-9.
383. Kallio PJ, Okamoto K, O'Brien S, Carrero P, Makino Y, Tanaka H, et al. Signal transduction in hypoxic cells: inducible nuclear translocation and recruitment of the CBP/p300 coactivator by the hypoxia-inducible factor-1alpha. EMBO J. 1998;17:6573-86.
384. Nie Z, Xue Y, Yang D, Zhou S, Deroo BJ, Archer TK, et al. A specificity and targeting subunit of a human SWI/SNF family-related chromatin-remodeling complex. Mol Cell Biol. 2000;20:8879-88.
385. Cairns BR. The logic of chromatin architecture and remodelling at promoters. Nature. 2009;461:193-8.
386. Hargreaves DC, Crabtree GR. ATP-dependent chromatin remodeling: genetics, genomics and mechanisms. Cell research. 2011;21:396-420. 190
387. Trotter KW, Archer TK. The BRG1 transcriptional coregulator. Nuclear receptor signaling. 2008;6:e004.
388. Khavari PA, Peterson CL, Tamkun JW, Mendel DB, Crabtree GR. BRG1 contains a conserved domain of the SWI2/SNF2 family necessary for normal mitotic growth and transcription. Nature. 1993;366:170-4.
389. Sif S, Saurin AJ, Imbalzano AN, Kingston RE. Purification and characterization of mSin3A-containing Brg1 and hBrm chromatin remodeling complexes. Genes Dev. 2001;15:603-18.
390. Wong AK, Shanahan F, Chen Y, Lian L, Ha P, Hendricks K, et al. BRG1, a component of the SWI-SNF complex, is mutated in multiple human tumor cell lines. Cancer Res. 2000;60:6171-7.
391. Ramirez-Carrozzi VR, Braas D, Bhatt DM, Cheng CS, Hong C, Doty KR, et al. A unifying model for the selective regulation of inducible transcription by CpG islands and nucleosome remodeling. Cell. 2009;138:114-28.
392. Ramirez-Carrozzi VR, Nazarian AA, Li CC, Gore SL, Sridharan R, Imbalzano AN, et al. Selective and antagonistic functions of SWI/SNF and Mi- 2beta nucleosome remodeling complexes during an inflammatory response. Genes Dev. 2006;20:282-96.
393. Grabmaier K, MC AdW, Verhaegh GW, Schalken JA, Oosterwijk E. Strict regulation of CAIX(G250/MN) by HIF-1alpha in clear cell renal cell carcinoma. Oncogene. 2004;23:5624-31.
394. Reisman D, Glaros S, Thompson EA. The SWI/SNF complex and cancer. Oncogene. 2009;28:1653-68.
395. Roberts CW, Orkin SH. The SWI/SNF complex--chromatin and cancer. Nat Rev Cancer. 2004;4:133-42.
396. Raval RR, Lau KW, Tran MG, Sowter HM, Mandriota SJ, Li JL, et al. Contrasting properties of hypoxia-inducible factor 1 (HIF-1) and HIF-2 in von Hippel-Lindau-associated renal cell carcinoma. Mol Cell Biol. 2005;25:5675-86.
397. Wang F, Zhang R, Beischlag TV, Muchardt C, Yaniv M, Hankinson O. Roles of Brahma and Brahma/SWI2-related gene 1 in hypoxic induction of the erythropoietin gene. J Biol Chem. 2004;279:46733-41.
398. Wang F, Zhang R, Wu X, Hankinson O. Roles of coactivators in hypoxic induction of the erythropoietin gene. PLoS One. 2010;5:e10002.
399. Kenneth NS, Mudie S, van Uden P, Rocha S. SWI/SNF regulates the cellular response to hypoxia. J Biol Chem. 2009;284:4123-31. 191
400. Brownlee PM, Chambers AL, Oliver AW, Downs JA. Cancer and the bromodomains of BAF180. Biochemical Society transactions. 2012;40:364-9.
401. Wu JN, Roberts CW. ARID1A Mutations in Cancer: Another Epigenetic Tumor Suppressor? Cancer discovery. 2013;1:35-43.
402. Shanahan F, Seghezzi W, Parry D, Mahony D, Lees E. Cyclin E associates with BAF155 and BRG1, components of the mammalian SWI-SNF complex, and alters the ability of BRG1 to induce growth arrest. Mol Cell Biol. 1999;19:1460-9.
192
APPENDIX A
BRG1 AND BRM CHROMATIN REMODELING COMPLEXES REGULATE THE
HYPOXIA RESPONSE BY ACTING AS A CO-ACTIVATOR FOR A SUBSET
OF HIF TARGET GENES 7
Abstract
Chromatin remodeling is an active process, which represses or enables
access of transcription machinery to genes in response to external stimulus,
including hypoxia. However, in hypoxia, the specific requirement as well as the
molecular mechanism via which the chromatin remodeling complexes regulate
gene expression remains unclear. In this study, we report that Brahma (BRM)
and the Brahma-related gene 1 (BRG1) ATPase containing SWI/SNF chromatin-
remodeling complexes promote expression of HIF1α and HIF2α genes and also
promote hypoxic induction of a subset of HIF1 and HIF2 target genes. We show
that BRG1 or BRM knockdown in Hep3B and RCC4T cells reduces hypoxic
induction of HIF target genes while BRG1 or BRM re-expression in BRG1/BRM
deficient SW13 cells increases HIF target gene activation. Mechanistically, HIF1
and HIF2 increase hypoxic induction of HIF target genes by recruiting BRG1
complexes to HIF target gene promoters, which promotes nucleosome
7 This chapter is reprinted with the permission of Molecular and Cellular Biology.
Sena JA, Wang L, Hu CJ. BRG1 and BRM chromatin-remodeling complexes regulate the hypoxia response by acting as coactivators for a subset of hypoxia- inducible transcription factor target genes. Mol Cell Biol. 2013 Oct;33(19):3849- 63. 193 remodeling of HIF target gene promoters in a BRG1 ATPase-dependent manner.
Importantly, we found that the function of BRG1 complexes in hypoxic SW13 and
RCC4T cells is dictated by the HIF-mediated hypoxia response and could be opposite from its function in normoxic SW13 and RCC4T cells.
Introduction
Hypoxia (Hx) is a common characteristic of many solid tumors. The Hx intratumoral microenvironment stabilizes hypoxia-inducible transcription factor 1α
(HIF1α) and 2α (HIF2α) that are normally degraded under normoxia. The stabilized HIF1α and HIF2α proteins translocate to the nucleus, where they dimerize with the constitutive nuclear protein, the aryl hydrocarbon receptor nuclear translocator (ARNT, also called HIF1β) to form HIF1α/ARNT (HIF1) and
HIF2α/ARNT (HIF2) heterodimers. HIF1 and HIF2 bind to HIF binding sites
(HBS) on HIF target gene promoters and/or enhancers and transactivate genes involved in neovascularization, glycolysis, cellular proliferation and metastasis.
Thus, the HIF-mediated Hx transcriptional response is critical for tumor progression by allowing cancer cells to adapt to a low oxygen environment (320-
323). However, recent reports indicate that HIF2 and particularly HIF1-mediated
Hx response can be tumor suppressive by activating tumor suppressive genes such as Scgb3a, Bnip3, Bnip3L, Nix, MYC inhibitor Mxi, p21 and p27 in a cell- type specific manner (374-379).
It is well established that multiple transcription factors (TFs) are required to achieve maximal activation of target genes in response to a specific stimulus 194
such as Hx. This multi-factorial transcription complex has been termed the
“enhanceosome”. Individual TFs in the enhanceosome complex may promote
transcription by recruiting RNA polymerase II/general TFs, and/or by recruiting
chromatin-modifying enzymes such as histone acetylase CBP and p300 and chromatin remodeling SWI/SNF complex. In the context of the enhanceosome
associated with the Hx response, two additional TFs, STAT3 and USF2 are
required for maximal Hx response (50, 332, 380). STAT3 and USF2 function in
the Hx response by recruiting CBP and p300 to the HIF1 and HIF2 target genes
respectively (332, 380). Despite several reports indicating that HIFs activate
their target genes by recruiting CBP and p300 (381-383), our recent studies,
using chromatin immunoprecipitation, did not find a significant role for HIF in
recruiting CBP and p300 to HIF target genes (332, 380). Therefore, our recent
findings lead to an important question in the field of Hx: how does HIF activate
target gene transcription in Hx cells?
SWI/SNF complexes regulate transcription through nucleosome disruption
and reconstruction in an ATP-dependent manner. SWI/SNF complexes can be
divided into the BAF complexes containing the BRG1 or BRM ATPase, and the
PBAF complexes containing the BRG1 ATPase (384-387). BAF and PBAF
complexes share nine protein subunits (B-actin, BAF53, BAF60, BAF57,
BAF155, BAF170, SNF5, BAF45 and BRG1), but can be distinguished by unique protein subunits in which the BAF complex is uniquely composed of, BAF250a or
BAF250b whereas PBAF contains BAF200, BAF180 and BRD7. While other subunits are important for the integrity and targeting of the SWI/SNF complex to 195
DNA, the BRG1 or BRM ATPase is thought to be absolutely required for the function of SWI/SNF; thus, a mutation or loss of expression of the BRG1 or BRM
ATPase inactivates the whole complex (384-387).
We hypothesized that HIF recruits SWI/SNF complexes to HIF target genes, increasing Hx induction of HIF targets. In this study, we have probed the interplay between HIF and BRM/BRG1 SWI/SNF complexes. Our studies showed there are SWI/SNF dependent and independent Hx inducible genes.
Functional studies indicated that the BRG1 complex amplifies HIF-mediated hypoxia responses. The role of the BRG1 complex in Hx cells is determined by the function of the HIF-mediated Hx response. These findings further our understanding of the role of BRG1 complexes in cancer cells under Hx.
Materials and Methods
Cell Culture.
Hep3B and Hep3B/ARNT shRNA cells were cultured in MEM/EBSS
(Hyclone) containing 10% FBS, 2 mM L-glutamine, 1 mM sodium
pyruvate, 100,000 units/L Penicillin/Streptomycin, 1.5 g/L sodium
bicarbonate and 1X non-essential amino acids (NEAA). SW13, RCC4,
RCC4T, U2OS and 293T cells were grown in high-glucose DMEM
(Hyclone) with 10% FBS, 2 mM L-glutamine, 100,000 units/L
Penicillin/Streptomycin and 1X NEAA. Prior to Hx treatment, 25 mM
HEPES was added to growth media and cells were incubated under Nx
(21% O2) or Hx (1.5% O2) for 12-16 hrs. 196
Knockdown of Endogenous mRNA Using Small Interfering RNAs
(siRNAs) or Short Hairpin RNA (shRNAs)
Control (Qiagen,1027281) or siRNAs specific for human BRM
(Dharmacon, L-017253-00) or BRG1 (Dharmacon, L-010431-00) mRNAs were
transfected into Hep3B cells at 50% confluency using HiPerFect Transfection
Reagent (Qiagen) according to the manufacturer’s protocol. 32hr post-
transfection, cells were cultured at 21% or 1.5% O2 for 12-16hr and then
collected to analyze mRNA or protein. To stably knock down BRM or BRG1
mRNA, Hep3B or RCC4T cells were transduced with pLKO.1 lentiviruses
expressing shRNAs targeting mRNA of BRM (TRCN0000020329 or
TRCN0000020332, Open Biosystems) or BRG1 (TRCN0000015549 or
TRCN0000015550, Open Biosystems) or both, and transduced cells were
selected by puromycin treatment. To stably knock down of HIF1α or HIF2α in
SW13 or RCC4T cells, were transduced with pLKO.1 lentiviruses expressing
shRNAs targeting mRNA of HIF1α (TRCN000003810, Open Biosystems) or
HIF2α (TRCN000003806, Open Biosystems), and transduced cells were selected by puromycin treatment. The specificity of HIF1α or HIF2α shRNA was
tested previously (332, 380).
Plasmid Constructs and Transient/Stable Transfection
The pBJ5 hBRG1 (Addgene plasmid 17873, deposited by Jerry
Crabtree)(388) and the pBABE hBRM (Addgene plasmid 1961, deposited by
Robert Kingston) (389) constructs were purchased from Addgene. The pBJ5 197 hBRG1 plasmid was used as template to generate an ATPase-dead BRG1 construct by PCR-mediated mutation of amino acids lysine 785 to an arginine
(K785R) and threonine 786 to a serine (T786S). These constructs were used for transient transfection experiments in SW13 cells using Lipofectamine Reagent
(invitrogen 18324-012) and PLUS Reagent (invitrogen 11514-015). Typically, 2
X105 cells/per well in 6-well plates were transfected with 1 µg of either pBABE hBRM, pBJ5 hBRG1 (WT or ATPase-dead) or with 500 ng each of pBABE hBRM and pBJ5 hBRG1. 32 hours after transfection, cells were placed under Nx or Hx for 16 hrs, and protein and mRNA were collected for analyses. The expression plasmids of mouse HIF1α or mouse HIF2α under control of the elongation factor
1 promoter have been described (324). Hep3B/mHIF1α or Hep3B/mHIF2α cells were generated by transfection of EF1/mHIF1α or EF1/mHIF2α DNA and hygromycin selection.
Protein Analysis
Whole cell lysates were prepared and quantified for protein concentration.
Western blot analysis was performed using standard protocols with the following primary antibodies: monoclonal anti-HIF1α (mAb) (610959, BD Bioscience, detecting human HIF1α protein), polyclonal anti-HIF1α (pAb) (NB 100-134,
Novus Biologicals, detecting both human and mouse HIF1α protein), anti-HIF2α mAb (D9E3, Cell Signaling, detecting both human and mouse HIF2α protein), anti-ARNT mAb (NB 100-124, Novus Biological), anti-BRM pAb (SC-6450, Santa
Cruz), anti-BRG1 pAb (SC-10768, Santa Cruz) and anti-Flag mAb (F3165, 198
Sigma), anti-Actin pAb (SC-1616, Santa Cruz). Densitometric analysis was
performed using ImageJ software (freely available at http://rsb.info.nih.gov/ij/).
RNA Preparation and Reverse Transcription Quantitative-PCR
RNA was isolated from cells using the RNeasy Plus mini kit (Qiagen), then
was reverse transcribed using the iSCRIPT Advanced cDNA synthesis kit (Bio-
Rad). mRNA levels were quantified by qRT-PCR using iQ Sybr Green supermix
(Bio-Rad) in triplicate on the CFX384 Real-Time System (BioRad). All primer
sets designed to measure mRNA levels or used in ChIP were validated for their
specificity and amplification efficiency (85%-110%) using melt curve analysis,
qRT-PCR product sequencing, and standard dilution analysis. qRT-PCR results
were normalized using the ∆∆CT method, 18s rRNA and β-Actin were used as
reference genes and untreated Nx samples were used as controls. At least three
independent experiments were performed to generate the results presented in
the figures.
ChIP Experiments
ChIP assays were performed as described previously (380). Anti-ARNT
(NB 100-110, Novus Biologicals) and anti-BRG1 pAb (SC-10768, Santa Cruz) antibodies were used for protein-DNA complex precipitation, rabbit pre-immune serum served as a control. DNA from input or immunoprecipitated samples was assayed using Sybr-Green based qPCR methods with specific primers designed 199 to amplify the CA9, EPO, HIF1α, or HIF2α promoters or in the exons/introns as non-specific controls as described previously (380).
Co-Immunoprecipitation
The BRG1 cDNA in pBJ5 hBRG1 was PCR amplified and inserted into pcDNA3.1 Flag vector, generating a BRG1 expression vector with a Flag tag at the c-terminus of the BRG1 protein. HEK293T cells were co-transfected with
BRG1Flag, mouse HIF1α, or mouse HIF2α constructs alone, or BRG1+mHIF1 or
BRG1+mHIF2α and placed under Hx for 16 hrs to analyze BRG1/HIFα interactions. Cleared cell lysates were incubated with M2-anti-Flag agarose beads (Sigma) to pull-down Flag-tagged BRG1. The lysate samples and eluents from the beads were then assayed for protein expression and precipitation or co- precipitation.
Nucleosome Scanning Assay (NUSA)
Hep3B cells or SW13 cells transfected with empty vector or vector expressing WT BRG1 or ATPase-dead BRG1 in 10 cm dishes (~60% confluency) were cultured under Nx or Hx for 16 hrs. Nuclei were isolated using the EZ Nucleosomal DNA Prep Kit (Zymo Research; cat. no. D5220) and treated with 0.5 units of micrococcal nuclease (MNase) for 10 minutes at room temperature. Following MNase incubation, digestion was stopped using 5X MN
Stop Buffer and nucleosomal DNA was purified and precipitated as described by the manufacturer’s instructions. Following purification, nucleosomal DNA was 200
used for Sybr Green–based qPCR to determine nucleosome positioning on the
CA9 and LDHA promoters. Overlapping qPCR primer sets were designed from
+407 to -836 relative to the transcription start site for the CA9 promoter and +36
to -1070 relative to the transcription start site for LDHA to generate amplicons of
150 bps, the size of DNA associated with one nucleosome. All primer sets
designed to measure nucleosome positioning were validated for their specificity
and amplification efficiency (80%-120%) using melt curve analysis and standard
dilution analysis. qPCR results were normalized using the ∆CT method. For
CA9 in Hep3B cells, a region spanning +273 to +407 was used as an internal
control since this region did not differ between Nx and Hx treated cells. For CA9
in SW13 cells, a region spanning -214 to -87 was used as an internal control
since this region did not differ between Nx and Hx treated cells or by transfection
with BRG1Flag or mBRG1Flag. For LDHA, a region spanning +20 to +117 was
used as an internal control. Three independent tests were performed in triplicate for each experiment and results indicate representative samples.
Luciferase Reporter Assay
Typically, HEK293T cells were grown to ~50% confluency in 6-well plates and transfected with 400 ng of a luciferase reporter driven by either the HIF1α or
HIF2α promoter, 200 ng β-galactosidase, and 300 ng HIF1αTM and 300 ng of
STAT3 or 300 ng HIF1αTM, 300 ng of STAT3 and 300 ng of BRG1 (pBJ5 hBRG1) expression vectors using Fugene transfection reagent. 36 hours after transfection, cells were collected into 400 ul 1x Reporter Lysis Buffer (Promega) 201 and assayed for β-gal activity and luciferase activity using a luminometer.
Promoter activation by HIF1, STAT3 and BRG1 was corrected for β-gal transfection efficiency and presented as fold of induction relative to promoter activities from an empty vector control. Results were the average of three experiments.
In Vitro Tumorigenic Assays.
Proliferation assays: 20,000-40,000 cells/well were plated in growth media in 24 well plates. The following day, the cell media was replaced with complete media containing 25 mM of HEPES and placed under Nx or Hx. Media was changed every 2 days, and every 24 hrs, 3 wells of cells per treatment were collected by trypsinization and counted by hemacytometer to yield an average number of cells/day. Scratch assays: Cells were plated at 80% confluency in normal growth media in duplicate in 6-well plates. The next day, when the cells formed a monolayer, the monolayers were scratched with a pipette tip. Then the cells were washed with PBS twice to remove detached cells and serum and then incubated under Nx or Hx for the remainder of the experiment in serum-free media containing HEPES (25mM). Photographs of the scratches at the same location in the well were taken daily and the width of the scratch measured using imageJ software. Duplicate measurements were averaged and graphed as percent closure of the scratch compared to the initial scratch (Day 0).
Clonogenic survival assays: A 1,000 cells/well were plated in duplicate in 6-well plates in normal growth media. Media was changed every two days, then 6 days 202
after plating, cells were washed and colonies stained with crystal violet dye and
counted.
Results
BRG1 or BRM Transient Knockdown Reduces the Hx Induction of Several
Known HIF1 and HIF2 Target Genes in Hep3B Cells.
Due to their high levels of HIF1α and HIF2α protein and high induction of
HIF target genes, Hep3B cells are often used in Hx studies (48, 302, 330). To
determine the role of SWI/SNF chromatin remodeling complexes in the Hx
response, Hep3B cells were transfected with non-targeting control siRNA or
siRNAs against BRM, BRG1, or both BRM and BRG1. BRM and BRG1 siRNAs
significantly reduced the mRNA (Fig. A1A) and protein (Fig. A1B) levels of BRM
and BRG1 in Nx and Hx Hep3B cells. Interestingly, knockdown of BRG1
consistently reduced the Hx induction of HIF1 target genes, CA9 and ANGPTL4
(Fig. A1C), and HIF2 target genes, EPO and PTPRB (Fig. A1D) while knockdown
of BRM reduced EPO induction, but not CA9, ANGPTL4 and PTPRB (Fig. A1C-
D). Simultaneous knockdown of BRG1 and BRM did not further reduce HIF
target gene induction. Importantly, the mRNA and protein levels of HIF1α, HIF2α
and ARNT were not affected by BRM, BRG1, or BRM+BRG1 transient
knockdown (Fig. 1A-B). These data suggested that both BRG1 and BRM
complexes are involved in the Hx response in Hep3B cells. However, BRG1
complex appears to have a broader and more important role than BRM complex
in the Hx response in Hep3B cells. 203
Stable Knockdown of BRG1 Reduces HIF1α and HIF2α Gene Transcription in Hep3B Cells.
Transient BRG1 or BRM knockdown using siRNA may be insufficient to remove the pre-made BRG1/BRM proteins in the SWI/SNF complexes that were already associated with their target gene promoters thus we stably reduced the levels of BRG1 or BRM mRNA (Fig. A2A) and protein (Fig. A2B) in Hep3B cells using shRNAs. Compared to Hep3B/GFP shRNA cells, Hep3B/BRG1 shRNA cells exhibited greatly reduced Hx induction of all HIF target genes analyzed including CA9, ANGPTL4, GLUT1, PGK1, EPO, PAI1, PTPRB, ADM, and VEGF
(Fig. A2C-D and data not shown) while Hep3B/BRM shRNA cells exhibited slightly reduced induction of CA9 and EPO, but not ANGPTL4 and PTPRB (Fig.
A2C-D). Reduced Hx induction of HIF target genes in Hep3B/BRG1 shRNA cells was likely due to reduced mRNA and protein levels of HIF1α and HIF2α (Fig.
A2A-B), suggesting an important role of BRG1 complex in maintaining HIF1α and
HIF2α gene expression in Hep3B cells.
The BRG1 Complex Promotes Hx Induction of HIF Target Genes in Hep3B
Cells.
Since the BRG1 complex is more important than the BRM complex in
HIFα and HIF target gene regulation in Hep3B cells, we focused 204
Figure A1
Figure A1. BRG1 is More Important than BRM Complexes in Regulating Hx- Induction of HIF Target Genes in Hep3B Cells. RNA or protein was prepared from Nx and Hx Hep3B cells transfected with control siRNA or siRNAs targeting human BRM, BRG1 or BRM+BRG1 mRNAs for the following assays. A) RT- qPCR analysis of BRM, BRG1, HIF1α, HIF2α and ARNT mRNA levels. B) Western blot analysis of BRM, BRG1, HIF1α, HIF2α and ARNT proteins. C) RT- qPCR analysis of HIF1α target gene levels, CA9 and ANGPTL4. D) RT-qPCR analysis of HIF2α target gene levels, EPO and PTPRB. qPCR data in this figure and other figures was normalized using the ∆∆CT method or otherwise specified; 18s rRNA and β-actin mRNAs were used as reference and wild-type Nx samples were used as a calibrator. All experiments were performed at least three separate times. Error bars are +/- 1 s.d. in this and other figures. A One Way Analysis Of Variance (ANOVA) was performed for this and other studies in this paper unless otherwise stated. * Indicates P<0.05 and ** indicates P<0.01. Controls for statistical analysis were specified in each figure.
205
Figure A2
Figure A2: BRG1, but not BRM Complexes are Important in HIF1α and HIF2α Gene Transcription in Hep3B Cells. Hep3B cells were stably transduced with lenti-virus expressing BRM or BRG1 shRNAs. RNA and protein were prepared from BRM or BRG1 knockdown Hep3B cells cultured under Nx or Hx for the following assays. A) RT-qPCR analysis of BRM, BRG1, ARNT, HIF1α and HIF2α mRNA levels. B) Western Blot analysis of BRM, BRG1, HIF1α, HIF2α and ARNT proteins. C) RT-qPCR analysis of the mRNA levels of HIF1- specific targets, CA9 and ANGPTL4. D) RT-qPCR analysis of the mRNA levels of HIF2-specific targets, EPO and PTPRB. Two different shRNA for BRG1 or BRM produced identical results. Showed here were results from BRG1 TRCN0000015549 and BRM TRCN0000020329 shRNA.
206 specifically on BRG1. To determine BRG1’s function in HIF target gene induction, we established Hep3B cells that have higher levels of HIFα protein, but reduced levels of BRG1 protein. First, we stably transfected Hep3B cells with wild-type mouse HIF1α or HIF2α cDNAs under control of the human elongation factor 1 promoter (EF1). We used mouse HIF1α or HIF2α cDNA, not human HIF cDNA, because mouse and human HIFα proteins are functionally interchangeable and there are antibodies that specifically recognize the human
HIF1α proteins. Compared to parental Hep3B cells, Hep3B/mHIF1α and
Hep3B/mHIF2α cells expressed 3-6 fold more (endogenous and transfected)
HIF1α or HIF2α mRNA in Nx cells, but 10 fold more in Hx cells (Fig. A3A).
Increased levels of mouse HIF1α or HIF2α mRNA under Hx was due to increased expression of the mHIFα gene which is enhanced by the PGK1 promoter (Hx induced promoter) located upstream of the EF1 promoter in the vector (Fig. A3A). Relative to the parental Hep3B cells, Hep3B/mHIF1α cells exhibited 2-3 fold higher induction of HIF1 target genes, but 20-30% reduced induction of HIF2 target genes (Fig. A3B) while Hep3B/mHIF2α cells exhibited 2-
3 fold higher induction of HIF2 target genes, but 20-30% reduced induction of
HIF1 target genes (Fig. A3C).
Then, BRG1 was stably knocked down in Hep3B/mHIF1α or
Hep3B/mHIF2α cells (Fig. A4A-B). Consistent with results in Fig A2A-B, compared to Hep3B/mHIF1α or Hep3B/mHIF2α cells, stable knockdown of
BRG1 significantly reduced endogenous HIF1α and HIF2α mRNAs as detected 207
Figure A3
Figure A3. BRG1 Knockdown Reduces Endogenous HIF1α and HIF2α mRNA Levels, but not Transfected Mouse HIF1α or HIF2α mRNA. Hep3B cells were stably transfected with wild-type mouse HIF1α or mouse HIF2α STATION vectors driven by the Elongation Factor 1 promoter. Hep3B/ STmHIF1α or Hep3B/ STmHIF2α cells were then targeted with shRNA to stably knock down BRG1 mRNA. The cells were cultured under Nx and Hx for the following assays. A) RT-qPCR detection of total (endogenous and transfected) HIF1α or HIF2α mRNA by primers detecting both mouse and human HIFα. Hx induction of mHIF1α or HIF2α was due to the PGK1 promoter (controlling the hygromycin gene) in the STATION vector that is located upstream of the EF1 promoter. B) RT-qPCR detection of mRNA levels of HIF1 target genes, ANGPTL4, CA9, ENO2 and KCD11. C) RT-qPCR detection of mRNA levels of HIF2 target genes, EPO, INHBB, PLAC8 and RAB42. D) RT-qPCR detection of the endogenous human HIF1α and HIF2α mRNA and transfected mouse HIF1α and HIF2α mRNA in the indicated cells.
208
by qPCR primers located in the 3’ UTR of human HIFα mRNA (Fig. A3D).
Interestingly, compared to Hep3B/mHIF1α or Hep3B/mHIF2α cells,
Hep3B/mHIF1α/BRG1 shRNA or Hep3B/mHIF2α/BRG1 shRNA cells exhibited
similar levels of mouse HIF1α or mouse HIF2α mRNA (Fig. A3D) as detected by qPCR primers located in the 3’ UTR (from the vector) of the mouse HIF mRNAs.
Consistent with mRNA analysis, endogenous human HIF1α protein in Hx
Hep3B/mHIF1α/BRG1 shRNA and Hep3B/HIF2α/BRG1 shRNA cells was significantly reduced, as detected by antibody specifically recognizing endogenous human HIF1α protein (Fig. A4B, anti-HIF1α human). However, the
Hx Hep3B/mHIF1α/BRG1 shRNA or Hep3B/HIF2α/BRG1 shRNA cells
expressed higher levels of total HIF1α or HIF2α protein respectively, as detected
by antibody that recognized total HIFα protein (Fig. A5B, anti-HIF1α
human/mouse or anti-HIF2α human/mouse). Since Hep3B/GFP shRNA cells
express lower levels of HIF1α and HIF2α protein to that of Hep3B/mHIF1α/BRG1
shRNA and Hep3B/mHIF2α/BRG1 shRNA cells respectively, the Hep3B/GFP
shRNA cells, not the Hep3B/mHIF1α or Hep3B/mHIF2α cells were used as
controls in the following HIF target gene studies. Interestingly, Hx induction of
HIF1 target genes, CA9 and ENO2, in Hep3B/mHIF1α/BRG1 shRNA cells (Fig.
A4C) and HIF2 target genes, EPO and PTPRB, in Hep3B/mHIF2α/BRG1 shRNA
cells (Fig. A4D) were still significantly reduced, demonstrating a functional
importance of the BRG1 complex in HIF target gene activation in Hep3B cells.
209
Figure A4
Figure A4: BRG1 Promotes Hx Induction of HIF Target Genes in Hep3B Cells. Hep3B cells were stably transfected with vectors expressing mouse HIF1α or mouse HIF2α protein. These Hep3B/mHIF1α or Hep3B/mHIF2α cells were used to stably knock down BRG1 mRNA. Protein and RNA were isolated from Nx and Hx Hep3B/GFP shRNA, Hep3B/mHIF1α/BRG1 shRNA or Hep3B/mHIF2α/BRG1 shRNA cells for the following assays. A) RT-qPCR analysis of BRM and BRG1 mRNA levels. B) Western blot analysis of BRM, BRG1, human HIF1α protein, human and mouse HIF1α protein, human and mouse HIF2α protein and ARNT proteins. The numbers are relative protein expression levels, using ARNT as a calibrator. C) RT-qPCR analysis of the mRNA levels of HIF1-specific targets, CA9 and ENO2 in indicated cells. D) RT- qPCR analysis of the levels of HIF2-specific targets, EPO and PTPRB in indicated cells.
210
Re-introduction of BRM or BRG1 into BRM/BRG1 Deficient SW13 Cells
Enhances Hx Induction of HIF1 Target Genes.
To further investigate the role of the BRG1 and BRM complexes in the Hx response, we transiently transfected BRM, BRG1, or BRM+BRG1 expression plasmids into SW13 human adrenal adenocarcinoma cells, a cell line that is deficient for BRM and BRG1 (390), as confirmed by our western blot analysis
(Fig. A5A). Re-expression of BRM and particularly BRG1 significantly increased
HIF2α mRNA levels (Fig. A5C) in Nx and Hx SW13 cells and HIF2α protein in Hx cells (Fig. A5B). However, BRG1 and BRM complexes did not regulate HIF1α and ARNT expression in the SW13 cells (Fig. A5B-C).
Although no HIFα protein was detected in Nx SW13/BRM or BRG1 cells, these cells exhibited increased expression of HIF1 target genes, CA9 and
ANGPTL4 (Fig. A5D) and the HIF2 target gene, PAI1 under Nx (Fig. A5E), demonstrating a significant function of BRG1 and BRM complexes in regulating the basal expression of HIF target genes. Furthermore, Hx SW13/BRG1 or BRM cells also exhibited significantly increased induction of HIF1 targets CA9 and
ANGPTL4 (Fig. A5D) despite normal mRNA and protein levels of HIF1α and
ARNT (Fig. A5B-C), supporting a role of BRG1 and BRM complexes in activating
HIF1 target genes. Due to increased expression of HIF2α protein in
SW13/BRG1 or BRM cells, we were not able to conclude whether increased expression of the HIF2 target gene, PAI1, under Hx was the function of increased
HIF2α protein levels or the function of the restored BRG1/BRM complex (Fig.
A5E). To confirm if BRG1’s ATPase activity is required for the expression of HIF 211
Figure A5
Figure A5: Re-expression of BRG1 or BRM in BRM/BRG1 Deficient SW13 Cells Increases Expression of HIF2α and HIF Target Genes. A) Western blot analysis of BRM and BRG1 proteins in Nx and Hx Hep3B, RCC4, SW13 and U2OS cells. SW13 cells were transiently transfected with empty vector (Ctrl) or vector expressing WT (BRG1). 32 hours post transfection, cells were cultured under Nx or Hx for 16 hours for the following assays. B) Western blot analysis of BRM, BRG1, HIF1α and HIF2α proteins. C) RT-qPCR analysis of ARNT, HIF1α and HIF2α mRNA levels. D) RT-qPCR analysis of the mRNA levels of HIF1- specific genes, CA9 and ANGPTL4. E) RT-qPCR analysis of the mRNA levels of the HIF2-specific gene, PAI1.
212
Figure A6
Figure A6: Re-expression of WT, not ATPase-dead BRG1 in BRM/BRG1 Deficient SW13 Cells Increases Expression of HIF2α and HIF Target Genes. SW13 cells were transiently transfected with empty vector (Ctrl) or vector expressing WT (BRG1) or ATPase-dead BRG1 (mBRG1). 32 hours post transfection, cells were cultured under Nx or Hx for 16 hours for the following assays. A) Western blot analysis of BRG1, HIF1α, HIF2α and ARNT proteins. B) RT-qPCR analysis of BRM and BRG1 mRNA levels. C) RT-qPCR analysis of ARNT, HIF1α and HIF2α mRNA levels. D) RT-qPCR analysis of the mRNA levels of HIF1-specific genes, CA9 and ANGPTL4. E) RT-qPCR analysis of the mRNA level of HIF2-specific gene, PAI1.
213
target genes, we transfected SW13 cells with an ATPase-dead BRG1 (mBRG1)
(Fig. A6). Although similar protein levels of WT BRG1 or mBRG1 were detected
(Fig. A6A), WT BRG1 but not ATPase-dead BRG1 was able to increase
expression of HIF2α (Fig. A6C) and HIF target genes (Fig. A6D-E). In summary, we determined that while the BRG1/BRM complex is not absolutely required for the HIF-mediated Hx response, BRG1 and BRM complexes promote HIF target gene transcription in Nx and Hx SW13 cells. Consistent with data from Hep3B
cells, BRG1 appears to be more potent than BRM in enhancing Hx response of
SW13 cells.
BRG1 Binds to HIF Target gene Promoters in and ARNT/HIF dependent
Manner in Hep3B Cells.
To determine if BRG1 binds to the promoters of HIF1α, HIF2α and HIF
target genes, chromatin immunoprecipitation (ChIP) was performed using BRG1
or ARNT (positive control) antibodies in Nx or Hx Hep3B cells. As expected,
ARNT binding on the promoters of HIF target genes, CA9 and EPO, was
significantly increased in Hx Hep3B cells (Fig. A7A-B) while little ARNT binding
was detected on the HIF1α gene in both Nx and Hx Hep3B cells. Surprisingly,
we detected significant ARNT binding on the HIF2α promoter in Hx Hep3B cells
(Fig. A7D) although HIF2α is not a HIF target gene in this cell type (Fig. A1A).
Interestingly, we detected BRG1 binding on the promoters of HIF target genes,
CA9 (Fig. A7A) and EPO (Fig. A7B), and on the promoters of HIF2α and HIF1α
(much less) in Nx Hep3B cells (Fig. A7C-D), consistent with fact that BRG1 214
Figure A7
Figure A7. Hx Increases BRG1 Binding on the Promoters of CA9 and EPO in an ARNT-dependent Manner. ChIP analysis of ARNT and BRG1 binding on the promoters of the HIF1-specific gene, CA9 (A), HIF2-specific target EPO (B), HIF1α gene (C) or HIF2α gene (D) in Nx and Hx Hep3B cells or Hep3B/ARNT shRNA cells. E) Co-IP of HIF1α or HIF2α with Flag-tagged BRG1 in hypoxic HEK293T cells. F) Activation of HIF1α promoter/Luciferase (left) or HIF2α promoter/Luciferase (right) reporters by the empty vector (His), or HIF1α+STAT3 or HIF1α+STAT3+BRG1 in 293T cells.
215
activity is required for the expression of these genes in Nx Hep3B cells (Fig. A2).
Interestingly, BRG1 binding on the promoters of CA9, EPO and HIF2α (not
HIF1α) was increased by Hx (Fig. A7A-D).
To assess if increased BRG1 binding in Hx cells is HIF/ARNT dependent,
similar ChIP assays were conducted in Nx and Hx Hep3B/ARNT shRNA cells.
As expected, ARNT knockdown significantly reduced ARNT association on all
four promoters in Hx Hep3B cells. Interestingly, BRG1 binding on HIF target
gene promoters, CA9 and EPO, was significantly decreased in the Hx
Hep3B/ARNT shRNA cells (Fig. A7A-B). However, ARNT knockdown did not
change BRG1 binding on HIF1α or HIF2α promoters (Fig. A7C-D). To provide further evidence that HIF1 and HIF2 proteins recruit the BRG1 complex to HIF target gene promoters, we performed co-immunoprecipitation (Co-IP) experiments in Hx HEK293T cells co-transfected with Flag-tagged BRG1 and wild-type mouse HIF1α or HIF2α expression plasmids. Pull-down of BRG1-Flag protein co-precipitated endogenous as well as transfected HIF1α protein (Fig.
A7E,). BRG1-Flag also co-precipitated transfected HIF2α protein (Fig. A7E, bottom panel). However, no endogenous HIF2α co-IP was observed, likely due to low levels of endogenous HIF2α expression in 293T cells. To further test the role of the BRG1 complex in activating HIF1α and HIF2α gene transcription, we performed HIF1α and HIF2α promoter reporter gene assays. Transfection of
BRG1 alone did not change HIF1α or 2α promoter activity (data not shown).
However, co-expression of BRG1 enhanced the ability of HIF1α+STAT3 to activate the HIF1α and HIF2α promoters (Fig. A7F). These data suggest that 216
while there is basal binding of BRG1 on HIF target genes, Hx increases BRG1
binding to HIF target genes in an ARNT/HIF dependent manner. Additionally,
BRG1 also binds to and activates HIF1α and HIF2α promoters in an ARNT-
independent manner.
BRM or BRG1 Complexes are not Required for Hx Induction of a Subset of
HIF Target Genes.
Previous studies suggest that the induction of primary response genes in
LPS stimulated macrophages is BRG1 independent (391, 392). These primary
response genes typically have CpG-rich promoters (391, 392). Thus, we
selected a subset of HIF target genes with CpG island promoters to see if their
induction requires BRG1/BRM activity. Since all these genes (GLUT1, LDHA,
PDK1 and PGK1) are HIF1-specific targets (48, 330), we addressed this question
in Hep3B/mHIF1α/BRG1 shRNA cells, a cell line that has higher levels of HIF1α
protein despite stable BRG1 knockdown (Fig. A4B). Although Hx induction of
CA9 and ENO2 was markedly reduced in Hep3B/mHIF1/BRG1 shRNA cells (Fig.
A4C), GLUT1, LDHA, PDK1 and PGK1 exhibited normal or higher levels of
induction (Fig. 8A). Additionally, Hx induction of these HIF1 target genes was not
significantly enhanced by re-expression of BRM, BRG1 or both BRM and BRG1
in SW13 cells (Fig. A8B). Also, reduction of BRG1 activity in Hep3B cells and
reconstitution of BRG1 or BRM in SW13 cells did not change the Nx levels of
these genes (Fig. A8A-B). Consistent with expression data, BRG1 protein levels
on the PGK1 promoter (0.3% and 0.5% of input in Nx and Hx Hep3B cells 217 respectively) was much lower compared BRG1 dependent genes, CA9 and EPO
(Fig. A7A-B). Taken together, our data supports a novel concept that HIF target genes can be divided into BRG1/BRM dependent or independent genes.
Hx Induces Nucleosome Remodeling of CA9 Promoter in a BRG1
Dependent Manner in Hep3B Cells.
To better understand how BRG1 promotes HIF target gene transcription in
Hx Hep3B cells and if there is a nucleosomal binding difference between BRG1 dependent vs. BRG1 independent HIF targets, we performed a qPCR based nucleosome scanning assay (NUSA) on the promoter of a BRG1 dependent
HIF1 target gene, CA9 (-836 to +345), and a BRG1 independent HIF1 target,
LDHA (-1070 to +36), in Nx and Hx Hep3B cells. We found in Nx Hep3B cells that there was significant nucleosome association on the CA9 promoter in an area of -337 to +63 as judged by less sensitivity to micrococcal nuclease
(MNase) digestion reflected by increased detection by qPCR, than in an area of -
836 to -337 which is more sensitive to MNase as reflected by decreased detection by qPCR (Fig. A9A). However, Hx reduced nucleosome association on the CA9 promoter in an area of -337 to +63, an area important for RNA pol II and
HIF/ARNT binding (HBS is at -16) (393), as judged by increased sensitivity to
MNase digestion reflected by reduced detection by qPCR (Fig. A9A). To test if the BRG1 complex is responsible for the increased sensitivity of the CA9 promoter to MNase in Hx Hep3B cells, similar experiments were performed in
Hep3B/mHIF1α/BRG1 shRNA cells. Interestingly, no MNase sensitivity 218
Figure A8
Figure A8: Hx Induction of a Subset of HIF Target Genes is BRM/BRG1 Independent. A) RT-qPCR analysis of mRNA levels of HIF1 target genes, GLUT1, LDHA, PDK1 and PGK1 in Nx and Hx Hep3B/GFP shRNA or Hep3B/mHIF1α/BRG1 shRNA cells. B) RT-qPCR analysis of mRNA levels of HIF1 target genes in Nx and Hx SW13 cells transiently transfected with vector expressing BRM, BRG1 or BRM+BRG1.
219
Figure A9
Figure A9: Hx Decreases Nucleosome Association on the CA9 Promoter in a BRG1-dependent Manner. A) Nucleosome scanning assays of the CA9 promoter in Nx or Hx Hep3B cells. B) Nucleosome scanning assays of the CA9 promoter in Nx or Hx Hep3B/mHIF1α/BRG1 shRNA cells. C) Nucleosome scanning assays of the LDHA promoter in Nx or Hx Hep3B cells. D) Nucleosome scanning assays of the LDHA promoter in Nx or Hx Hep3B/mHIF1α/BRG1 shRNA cells. The location of each primer set was given relative to the transcription start site designated +1. qPCR data was quantified using the ∆CT method; for CA9 in Hep3B cells, a region spanning +273 to +407 was used as an internal control since this region did not differ between Nx and Hx treated cells. For LDHA, a region spanning +20 to +117 was used as an internal control. Relative % nucleosome bound DNA in this figure and figure 10 represent the nucleosome bound DNA relative to internal controls. For this figure t-tests were performed for Nx versus Hx samples for each corresponding position. Asterisk ‘*’ indicate P<0.05 and ‘**’ indicate a P<0.01
220 difference was observed among different regions of the CA9 promoter in Nx cells
(Fig. A9B) or between Nx and Hx Hep3B/mHIF1α/BRG1 shRNA cells (Fig. A9B), suggesting that BRG1 is responsible for the Hx-induced nucleosome remodeling observed on the CA9 promoter. Although Hx increased the MNase sensitivity in -
31 to -214 regions of LDHA promoter (Fig. A9C), this change was BRG1 independent since Hx-induced nucleosome remodeling of the LDHA promoter was observed in Hep3B/mHIF1α/BRG1 shRNA cells (Fig. A9D). These results provide direct evidence that BRG1 regulates chromatin remodeling of a BRG1 dependent HIF target gene, CA9 but has no role in chromatin remodeling of a
BRG1 independent HIF target gene, LDHA
Nucleosome Remodeling of the CA9 Gene Requires BRG1’s ATPase
Activity.
To test if BRG1’s ATPase activity is required for nucleosome remodeling of the CA9 promoter, we performed NUSA experiments for the CA9 promoter in
Nx and Hx SW13 cells transfected with empty vector (Control), WT or ATPase- dead BRG1. Interestingly, significant nucleosome remodeling of the CA9 promoter was observed in the Hx SW13 cells expressing WT BRG1 (Fig. A10A), but not ATPase-dead BRG1 (Fig. A10B) in comparison with Hx SW13/control cells. This data is consistent with expression data showing that Hx induction of
CA9 is increased by re-expression of WT but not of ATPase-dead BRG1 (Fig.
A6).
221
Figure A10
Figure A10: Reintroduction of wild-type but not ATPase-dead BRG1 in SW13 cells decreases nucleosome association on the CA9 promoter. Nucleosome scanning assays of the CA9 promoter in Nx or Hx SW13 cells transiently transfected with empty vector or vector expressing wild-type BRG1 (BRG1) (A) or ATPase-dead BRG1 (mBRG1) (B). qPCR data in this figure was quantified using the ∆CT method; a region spanning -214 to -87 was used as an internal control since nucleosome association on this region did not differ between Nx and Hx or by transfection with BRG1Flag or mBRG1Flag. 222
Re-introduction of BRG1 into SW13 Cells Increases the HIF-mediated Hx
Response and HIF-mediated Inhibition of Proliferation.
To determine BRG1’s role in Nx and Hx cancer cells, SW13 cells were stably transduced with HIF1α, HIF2α or scrambled shRNAs and followed by transient transfection with empty or WT BRG1 expression plasmids. HIF1α protein was similarly expressed in Hx SW13 cells transfected with empty or
BRG1 expression vectors and significantly reduced by HIF1α shRNA (Fig.
A11A). HIF2α protein was only detected in Hx SW13 cells transfected with WT
BRG1 expression vector and reduced in Hx SW13/HIF2α shRNA+BRG1 cells
(Fig. A11A). In agreement with Fig. 5C, HIF2α mRNA levels were increased by
BRG1reintroduction, whereas HIF1α and ARNT mRNA levels were not changed in these cells (Fig. A11B). As expected, Hx induction of CA9 and ANGPTL4 was significantly reduced in SW13/HIF1α shRNA cells, but not in SW13/HIF2α shRNA cells (Fig. A11C) as these cells express little HIF2α mRNA (Fig. A11B).
Relative to SW13/Scrm+Ctrl cells, Nx and Hx SW13/scram shRNA+BRG1 cells exhibited significantly increased CA9 and ANGPTL4 expression. However, relative to SW13/scram shRNA+BRG1 cells, SW13/HIF1α shRNA+BRG1 cells exhibited reduced CA9 and ANGPTL4 expression under Hx, but not under Nx.
The levels of CA9 and ANGPTL4 in Nx and Hx SW13/HIF2α+BRG1 cells were also similar to that in SW13/scram+BRG1 cells, indicating that these genes are primarily activated by HIF1 but not HIF2.
We then performed a proliferation assay for these cells (Fig. A11D). Hx completely blocked SW13/Scrm shRNA+Ctrl cell proliferation compared to Nx 223
Figure A11
Figure A11: Re-introduction of BRG1 into SW13 Cells Increases the HIF- mediated Hx Response and HIF-mediated Growth Suppressive Activity. SW13 cells were targeted with shRNA against non-targeting scrambled RNA (Scrm), or HIF1α, HIF2α mRNA. These cells were then transiently transfected with an empty vector (Ctrl) or vector expressing BRG1 (BRG1). 32 hours post transfection, cells were cultured under Nx or Hx for 16 hours for the assays in A- C. A) Western blot analysis of BRG1, HIF1α, HIF2α and ARNT. B) RT-qPCR analysis of BRG1, ARNT, HIF1α, HIF2α mRNA. C) RT-qPCR analysis of HIF target genes CA9 and ANGPTL4. D) Cell proliferation assay. For proliferation assay, 32 hours post transfection, cells were cultured under Nx or Hx for additional 24, 48, or 72 hours. Nx SW13/HIF1α or HIF2α+Ctrl cells or Nx SW13/HIF1α or HIF2α+BRG1 cells exhibited similar patter to that of Nx SW13/Scrm+Ctrl cells or Nx SW13/Scrm+BRG1 cells respectively. These data was not included in Fig. 11D to make the figure less crowded.
224
SW13/Scrm shRNA+Ctrl cells, which grew steadily from Day 1-3. These data suggested that Hx not only stops SW13 cell proliferation also causes low levels of SW13 cell death. Stable knockdown of HIF1α significantly enhanced cell proliferation at Day 1 under Hx, however proliferation plateaued at Day 1 and there was no further growth during Day 2-3. Similar to SW13/Scrm shRNA+Ctrl cells, Hx SW13/HIF2α shRNA+Ctrl cells exhibited low levels of cell death, which is expected since these cells do not express HIF2α mRNA. Consistent with a tumor suppressive function of BRG1 (390, 394, 395), Nx SW13/Scrm
shRNA+BRG1 cells exhibited slower growth than Nx SW13/Scrm+ctrl cells.
However, these BRG1 re-expressed cells continued to grow under Nx.
Interestingly, under Hx, SW13/Scrm shRNA+ BRG1 cells exhibited increased cell
death at Day 2 and then recovered a little at Day 3, providing strong evidence for
increased cell death by BRG1 re-expression. SW13/HIF1α shRNA+BRG1 cells
proliferated similar to SW13/HIF1 shRNA+Ctrl cells under Hx at Day 1, likely
because of delayed manifestation of BRG1’s activity during Day 1, consistent
with the observation that re-expression of BRG1 did not reduce Nx cell number in
Day 1. Although SW13/HIF1α shRNA+BRG1 cell numbers began to decrease at
Day 2-3 compared to SW13/HIF1α shRNA+Ctrl cells, proliferation was still higher
than in SW13/Scrm shRNA+BRG1 cells, suggesting that HIF1α knockdown could
partially compensate for BRG1’s proliferation inhibitory/cell death function under
Hx, implying that BRG1’s inhibitory function under Hx is likely through regulation
the HIF1-mediated Hx response. In addition, SW13/HIF2α shRNA+BRG1 cells
displayed proliferation increase, compared to SW13/Scrm +BRG1 cells in Days 1 225 and 2, suggesting that HIF2α also plays proliferation inhibitory role in Hx SW13 cells. In summary, these data supports a proliferation suppressive role of BRG1 in Nx SW13 cells and also indicates that under Hx BRG1 re-expression causes cell death by enhancing the HIF-mediated Hx response.
The HIF-mediated Hx Response Dictates the Role of the BRG1 Complex in
Hypoxic RCC4T Cells During Proliferation and Migration.
To further explore BRG1’s role in HIF-independent and dependent tumorigenesis, we stably knocked down BRG1, HIF1α, HIF2α, BRG1+HIF1α or
BRG1+HIF2α in RCC4T cells, a renal cell carcinoma cell line, using shRNAs
(Fig. A12A). Unlike in Hep3B cells, BRG1 knockdown did not reduce HIF1α or
HIF2α protein levels (Fig. A12A). As expected, CA9 levels were greatly reduced in Hx RCC4T/HIF1α shRNA cells, but increased in Hx RCC4T/HIF2α shRNA cells, consistent with previous findings suggesting that CA9 is a HIF1 target gene in RCC4T cells (380, 396). ANGPTL4 was decreased in Hx RCC4T/HIF1α and
HIF2α shRNA cells, suggesting ANGPTL4 is a HIF1/HIF2 common target gene in
RCC4T cells (Fig. A12B). BRG1 shRNA reduced Hx induction of both genes
(Fig. A12B), demonstrating, in another cell line, that BRG1 is important for HIF target gene expression. Hx induction of CA9 was further decreased in
RCC4T/BRG1+HIF1α shRNA cells relative to Hx RCC4T/BRG1 shRNA or HIF1α shRNA cells. CA9 induction in RCC4T/BRG1+HIF2α shRNA cells was intermediate between RCC4T/HIF2α shRNA and RCC4T/BRG1 shRNA cells
(Fig. A12B). However, Hx induction of ANGPTL4, a HIF1/HIF2 target gene, was 226 similar in RCC4T/BRG1+HIF1α and RCC4T/BRG1+HIF2α to that of
RCC4T/HIF1α and RCC4T/HIF2α cells, respectively (Fig. A12B).
After confirming that BRG1 regulates HIF target genes in RCC4T cells, we performed several functional assays for each of the above cells cultured under
Nx or Hx (Fig. A12C-E). First, we performed clonogenic assays (Fig. A12C) and found that Hx did not affect clonogenic survival of control RCC4T/Scrm cells, likely due to the opposing effects of HIF1 and HIF2, in which HIF1 inhibits clonogenic survival whereas HIF2 promotes clonogenic survival of RCC4T cells under Hx (Fig. A12C). BRG1 knockdown significantly reduced cell survival under
Nx, indicating that BRG1 promotes clonogenic survival of Nx RCC4T cells (Fig.
A12C). However, there was no survival difference between Nx and Hx
RCC4T/BRG1 shRNA cells, consistent with the observation that hypoxia did not change cell survival. Knockdown of HIF1α or HIF2α on top of BRG1 knockdown also did not change the survival of these cells, likely because BRG1 knockdown itself has already reduced the activity of HIF1 and HIF2 to sufficiently low levels, which mask the effects of HIF1α or HIF2α knockdown.
We also performed proliferation assays and found that Hx increased cell proliferation of RCC4T/Scrm cells (Fig. A12D). In addition, knockdown of HIF1α or HIF2α increased or decreased cell proliferation, suggesting a proliferation inhibitory or promoting role for HIF1 or HIF2 respectively (Fig. A12D).
Interestingly, BRG1 knockdown increased cell proliferation, suggesting that
BRG1 inhibits cell proliferation in Nx RCC4T cells. However, under Hx, BRG1 knockdown reduced cell proliferation relative to Nx RCC4T/BRG1 shRNA cells 227
Figure A12
Figure A12. The HIF-mediated Hx Response Dictates the Role of the BRG1 Complex in Hypoxic RCC4T Cells for Proliferation and Migration. RCC4T cells were targeted with shRNA against non-targeting scrambled RNA (Scrm), HIF1α, HIF2α, BRG1, BRG1+HIF1α, or BRG1+HIF2α. These cells were cultured under Nx or Hx for the following assays. A) Western blot analysis of BRG1, HIF1α, HIF2α and ARNT. B) RT-qPCR analysis of HIF target genes CA9 and ANGPTL4. C) Clonogenic survival assay. D) Cell proliferation assay for 24 hr culture. F) Wound healing cell migration assay. Nx RCC4T/HIF1α or HIF2α shRNA cells or Nx RCC4T/BRG1+HIF1α or HIF2α shRNA cells exhibited similar patter to that of Nx RCC4T/Scrm or RCC4T/BRG1 shRNA cells respectively. These data was not included in Fig. 12C-E to make the figure less crowded.
228
and prevented the Hx induced increase in proliferation seen in RCC4T/Scrm
cells, suggesting that the BRG1 complex promotes RCC4T cell proliferation during Hx. Knockdown of HIF1α or HIF2α on top of BRG1 knockdown did not affect cell proliferation relative to BRG1 knockdown cells (Fig. A12D), likely
because BRG1 knockdown reduced the activity of HIF1 and HIF2 to sufficiently
low level which conceal the effects of further HIF1α or HIF2α knockdown.
Finally, we performed a wound healing migration assay (Fig. A12E). We
found that Hx slightly reduced cell migration of RCC4T/Scrm shRNA cells at 48
and 72, but not 24 hours. Interestingly, migration was increased in
RCC4T/HIF1α shRNA cells compared to control cells under Hx, indicating that
the HIF1-mediated Hx response inhibits migration in RCC4T cells. In contrast,
migration was not significantly changed by HIF2α knockdown although there was
a trend that HIF2α knockdown reduced cell migration. Nx RCC4T/BRG1 shRNA
cells exhibited reduced migration at all three time points, suggesting that the
BRG1 complex promotes RCC4T cell migration under Nx. However, under Hx,
BRG1 knockdown increased cell migration compared to Hx RCC4T/Scrm shRNA
cells at all tree time points, suggesting a migration suppressive function of BRG1
under Hx. The migration suppressive role of the BRG1 complex in Hx RCC4T
cells is consistent with the migration suppressive role of the hypoxia response in
this cell type. Again, knockdown of HIF1α on top of BRG1 knockdown did not
significantly affect cell migration relative to BRG1 knockdown cells (Fig. A12E).
However, knockdown of HIF2α in addition of BRG1 knockdown reduced cell
migration at 24 and 48, but not 72 hours. In conclusion, the BRG1 complex can 229
have opposing functions in Nx or Hx RCC4T cells and the function of the BRG1
complex in Hx RCC4T cells are in agreement with the function of the HIF-
mediated Hx response for RCC4T cell survival, proliferation and migration in
vitro.
Discussion
While Hx-mediated HIFα protein stability and the major players in Hx
signaling pathway are well established (320-323), how HIF activates its target
genes is much less understood. Since HIF is not the main recruiter of p300 and
CBP (332, 380) we investigated the role of HIF in recruiting the BRG1/BRM
complexes in the Hx response.
To date, the conclusions from three published papers concerning the role
of BRG1/BRM complexes in the Hx response (397-399) did not agree with each
other and are inaccurate. Kenneth et al showed that knockdown of BRG1 (but
not BRM) or other components of the BRG1 complex such as BAF57, BAF155,
BAF170 in U2OS cells significantly decreased the levels of HIF1α (not HIF2α)
mRNA and protein (399). Since they did not detect BRG1 protein binding on the
promoter(s) of HIF target gene(s) (399), they concluded that BRG1 has no co-
activator role in the HIF-mediated Hx response. In contrast, Wang et al. showed
that knockdown of BRM or BRM+BRG1, but not BRG1 in Hep3B cells reduced
Hx induction of EPO (397). In their second paper, they showed that knockdown
of both BRG1 and BRM in Hep3B cells reduced Hx induction of LDHA and PGK1
(398). Thus, Wang et al suggested that the BRM complex served as a co- 230 activator for the HIF-mediated Hx response including genes such as LDHA and
PGK1 that are BRG1/BRM independent HIF target genes as we reported here.
We think we can now reconcile the discrepancy between our, Kenneth and
Wang’s data. We believe that Wang’s data could be explained by BRG1/BRM’s ability to regulate HIF1α gene expression, thus BRG1/BRM knockdown in Hep3B cells reduced HIF1α levels and reduced the Hx induction of all HIF target genes including LDHA and PGK1 that are BRG1 independent. Kenneth et al, did not detect the binding of BRG1/BRM protein on the promoter(s) of HIF target gene(s) although they did not specify the genes they assessed (399). We speculate that they might have selected BRG1 independent HIF target genes such as GLUT-1 for their BRG1 binding assay since GLUT-1 was one of the two genes they used for functional readout in response to BRG1 knockdown (399). Thus, our data clearly addressed the inconsistencies between these published papers and provides strong evidence for the roles of BRG1 and BRM complexes in regulating HIF1α expression and in the HIF-mediated Hx response.
We reported here, for the first time, that the BRG1 complex also regulates
HIF2α gene expression in Hep3B (Fig. A2) and SW13 cells (Fig. A5), and that the BRM complex also activates HIF2α gene expression in SW13 cells (Fig. A5).
In addition, since we used multiple cell lines, we showed for the first time that the interplay between the BRG1/BRM complex and HIFα genes is cell-type specific.
For example, BRG1, but not the BRM complex activates both HIF1α and HIF2α in Hep3B cells (Fig. A2) while both BRG1 and BRM complexes activate HIF2α, but not HIF1α in SW13 cells (Fig. A5). However, the BRG1/BRM complex does 231
not regulate HIF1α and HIF2α gene expression in RCC4T cells (Fig. A12).
Currently, we are not able to determine the reasons for these phenomena. For
example, we found that the negligible role of BRM in regulating HIFα gene
expression in Hep3B cells is not due to low BRM expression since BRM is highly
expressed in Hep3B cells (data not shown) and over-expression of BRM in
Hep3B cells also does not induce HIFα gene expression (data not shown).
Additionally, the BRG1/BRM-independent expression of HIF1α gene expression
in SW13 cells is not due to high HIF1α expression since HIF1α levels are similar
in SW13 and Hep3B cells (data not shown).
We also established for the first time that the BRG1 complex acts as a co- activator of HIF target genes by showing that BRG1 binds to the promoters of
HIF target genes, CA9 and EPO (Fig. A7). Using a nucleosome scanning assay,
we showed that the BRG1 complex is involved in remodeling the CA9 promoter
in Hx cells (Fig. A9), providing direct evidence for a co-activator role of the BRG1
complex in HIF target gene activation. Moreover, we showed that Hx induced
nucleosome remodeling (Fig. A10) and transcription activation (Fig. A6) of CA9 is
dependent on BRG1’s ATPase activity. While Hx-mediated increases in BRG1
recruitment to the EPO promoter was reported previously (397), we furthered
these studies by showing that HIF/ARNT is responsible for increased BRG1
recruitment to HIF target genes during Hx (Fig. A7). Also, we showed that BRG1
interacts with HIF1α and HIF2α proteins as assessed by co-IP. Thus, our results
provide a novel mechanism, suggesting that HIF/ARNT activates its target genes
by recruiting BRG1 complexes to HIF target gene promoters. BRG1 recruitment 232
increases nucleosome remodeling and enhances transcription in an ATPase-
dependent manner.
BRG1, BRM and particularly the non-ATPase components of the
SWI/SNF complexes are frequently inactivated in multiple cancers (390, 394,
395, 400, 401). Thus, the SWI/SNF complex has long been regarded as a tumor suppressor. For example, reintroduction of BRG1 into cells lacking BRG1 expression was sufficient to reverse their transformed phenotype and induce growth arrest and a flattened morphology (390). This is consistent with our data in SW13 cells showing that re-introduction of BRG1 into Nx SW13 cells reduced cell proliferation (Fig. A11D). For the first time, we analyzed the role of the BRG1 complex in cell proliferation in Hx cells. Consistent with a proliferation suppressive role of the HIF1-mediated Hx response in SW13, we found that the
BRG1 re-expression amplified proliferation suppressive activity of HIF-mediated hypoxia response to cause more cell death which is rescued by HIF1α knockdown (Fig. A11D). Unexpectedly, we found that the BRG1’s effects on
SW13 cell proliferation/death differ between Day 0-1 and Day 2-3. This difference is likely due to a lag period in BRG1 function, therefore, the proliferation effects observed during Day 0-1 are BRG1 independent but HIF dependent since HIFs are stably reduced prior to BRG1 transfection.
Unfortunately, we were not able to perform clonogenic survival or wound healing migration assays in SW13 cells because stable transfection of BRG1 promotes cell senescence, as was reported previously (402). 233
The roles of the BRG1 complex in RCC4T cells are particularly interesting.
Under Hx, HIF1- or HIF2-mediated Hx responses have opposite roles in cell survival, proliferation and migration as reported previously (142). Despite the opposing effects of HIF1 and HIF2, and the fact that BRG1 knockdown reduced both HIF1 and HIF2 mediated Hx responses, we found that BRG1’s function in
Hx RCC4T cells is closely related with the net effect of the Hx response in cell survival, proliferation and migration. Another interesting observation is that the
BRG1 complex has opposite functions in cell proliferation and migration in Nx or
Hx RCC4T cells. For example, the BRG1 complex inhibits cell proliferation under
Nx, but increases cell proliferation under Hx. In addition, the BRG1 complex promotes cell migration under Nx, but decreases cell migration under Hx. These findings are particularly important, providing evidence, for the first time, that the
Hx tumor microenvironment could be a determinant for the function of the BRG1 complex in Hx cells.
In summary, this study confirms that the BRG1 complex regulates expression of the HIF1α gene and establishes a novel role of BRG1 and BRM complexes in HIF2α gene transcription. In addition, we separate BRG1/BRM’s function in regulating HIFα genes from their role in HIF target gene activation.
Furthermore, we reported, for the first time, that activation of a subset of HIF target genes is BRG1/BRM independent. Importantly, we found that increased
BRG1 recruitment to HIF target genes is an important way for HIF/ARNT to activate its target gene expression. In addition, our functional studies in SW13 and RCC4T cells indicated that the BRG1 complex in cell proliferation, migration 234 and survival in hypoxic cells is dictated by the role of HIF-mediated hypoxia responses.
Acknowledgments
This work was supported by grants from the National Cancer Institute
(RO1CA134687, Hu). Johnny Sena was supported by “Research Supplemental to Promote Diversity” (NCI) from June 1, 2010 to May 30, 2012. We thank Dr.
Trevor Williams for thoughtful discussions and reading of the manuscript.