UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Structural foundation for transcriptional regulation by Ets1 and CBP Kang, Hyun-Seo 2010

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2010_fall_kang_hyunseo.pdf [ 44.65MB ]
Metadata
JSON: 24-1.0071341.json
JSON-LD: 24-1.0071341-ld.json
RDF/XML (Pretty): 24-1.0071341-rdf.xml
RDF/JSON: 24-1.0071341-rdf.json
Turtle: 24-1.0071341-turtle.txt
N-Triples: 24-1.0071341-rdf-ntriples.txt
Original Record: 24-1.0071341-source.json
Full Text
24-1.0071341-fulltext.txt
Citation
24-1.0071341.ris

Full Text

  STRUCTURAL FOUNDATION FOR TRANSCRIPTIONAL REGULATION BY ETS1 AND CBP    by  HYUN-SEO KANG   B.S., The University of Oregon, 2002 M.Sc., The University of British Columbia, 2005      A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY   in   THE FACULTY OF GRADUATE STUDIES  (Biochemistry and Molecular Biology)      THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)     September 2010    © Hyun-Seo Kang, 2010   ii	
   Abstract Using cell-based assays and biophysical measurements, we have defined the mechanism by which Ras/MAP kinase signaling enhances Ets1 regulated gene expression via phosphorylation-enhanced recruitment of the transcriptional co-activator CBP. As confirmed by 31P/13C-NMR experiments, the MAP kinase ERK2 phosphorylates Thr38 and Ser41 within the unstructured region of Ets1, immediately N- terminal to the PNT domain. The NMR-derived structure of residues 29-138 of Ets1 revealed that the PNT domain is composed of a core four-helix bundle (H2-H5), also known as the SAM fold, appended with two additional helices (H0-H1). Most importantly, helix H0 is only marginally stable as shown by various NMR methods, including chemical shift, amide hydrogen exchange, and 15N relaxation analyses.  Dual phosphorylation of Ets1 perturbs a "closed-open" conformational equilibrium of the PNT domain, displacing the dynamic helix H0 from the core bundle. These modifications increase the affinity of Ets1 for the TAZ1 (or CH1) domain of CBP by ~30 fold as measured with isothermal titration calorimetry (Kd ~ 60 to 2 µM). NMR-monitored titration experiments mapped the interaction surfaces of the TAZ1 domain and Ets1, the latter encompassing both the phosphoacceptors and PNT domain, also showing sensitivity to ionic strength. Charge complementarity of these surfaces indicates that electrostatic forces act in concert with the conformational equilibrium to mediate phosphorylation effects.  We conclude that the dynamic helical elements of Ets1, appended to a conserved structural core, constitute a "phospho-switch" to direct Ras/MAPK signaling to  iii	
   downstream changes in gene expression. This detailed structural and mechanistic information illustrates an evolutionary development within a gene family to increase the capacity for biological regulation.  We also discovered that the CBP TAZ1 domain associates intramolecularly with residues 28-82 in its N-terminal nuclear receptor interacting domain (NRID). NMR studies indicated that the NRID undergoes a coil-helix conformational transition upon binding the same interface on TAZ1 as recognized by many transcription factor partners. This led us to hypothesize that CBP is regulated by an auto-inhibitory mechanism. In support of this model, affinity of the hypoxia inducible factor HIF-1α for TAZ1 is reduced competitively by the presence of the NRID.  iv	
   Table of Contents Abstract.......................................................................................................ii	
   Table of Contents......................................................................................iv	
   List of Tables ...........................................................................................viii	
   List of Figures ...........................................................................................ix	
   List of Symbols ........................................................................................xii	
   Acknowledgements ................................................................................xiv	
   Chapter 1 - Introduction..........................................................................1	
   1.1 Eukaryotic gene expression................................................................................ 1	
   1.1.1 Components of transcriptional regulation in eukaryotes.................................. 1	
   1.1.2 Levels of transcriptional regulation in eukaryotes............................................ 3	
   1.2 Phosphorylation of transcription factors......................................................... 11	
   1.2.1 Mechanistic view of protein phosphorylation ................................................. 12	
   1.2.2 Functions of transcription factor phosphorylation .......................................... 16	
   1.3 ETS Family transcription factors...................................................................... 20	
   1.4 Ets1...................................................................................................................... 23	
   1.4.1 ETS domain in Ets1....................................................................................... 25	
   1.4.2 PNT domain in Ets1....................................................................................... 27	
   1.5 CBP/p300............................................................................................................. 29	
   1.5.1 Domain structure of CBP............................................................................... 30	
   1.6 Thesis overview ................................................................................................. 36	
   1.7 Other publications.............................................................................................. 38	
   Chapter 2 - Phosphorylation and the Ets1 PNT Domain....................40	
   2.1 Introduction ........................................................................................................ 41	
    v	
   2.2 Results ................................................................................................................ 45	
   2.2.1 Identification of the ERK2 phosphoacceptors in Ets1.................................... 45	
   2.2.2 Structure determination of apo- and 2P-Ets129-138 by NMR........................... 51	
   2.3 Discussion .......................................................................................................... 77	
   2.3.1 Structural analysis of Ets129-138 and 2P-Ets129-138 ......................................... 77	
   2.3.2 Helix H0 exists in a phosphorylation-dependent conformational equilibrium 78	
   2.3.3 Biological specificity from regulatory elements appended to conserved domains .................................................................................................................. 79	
   2.4 Materials and methods ...................................................................................... 83	
   2.4.1 In vivo transcription assays and immunoprecipitation ................................... 83	
   2.4.2 Protein purification......................................................................................... 83	
   2.4.5 NMR spectroscopy ........................................................................................ 84	
   2.4.6 Structure calculations .................................................................................... 85	
   2.4.7 Amide hydrogen exchange............................................................................ 87	
   Chapter 3 - Phosphorylation-dependent Ets1-CBP interaction ........88	
   3.1 Introduction ........................................................................................................ 89	
   3.2 Results ................................................................................................................ 91	
   3.2.1 Identification of phosphorylation-dependent binding of Ets1 PNT with the TAZ1 domain of CBP.............................................................................................. 91	
   3.2.2 NMR-based analyses of TAZ1/PNT binding interface................................... 95	
   3.2.3 Role of newly identified structural components in MAPK signaling ............. 116	
   3.3 Discussion ........................................................................................................ 120	
   3.3.1 Roles of phosphorylation in regulating Ets1-CBP interaction ...................... 120	
   3.3.2 Distinct features of Ets/CBP binding............................................................ 124	
    vi	
   3.3.3 ETS family PNT domains and cellular signaling.......................................... 126	
   3.4 Conclusions...................................................................................................... 129	
   3.5 Materials and methods .................................................................................... 130	
   3.5.1 Protein purification....................................................................................... 130	
   3.5.2 Isothermal titration calorimetry .................................................................... 130	
   3.5.3 NMR spectroscopy ...................................................................................... 131	
   Chapter 4 - Discovery and characterization of a TAZ1 autoinhibitory region in CBP .........................................................................................132	
   4.1 Introduction ...................................................................................................... 133	
   4.1.1 The N-terminal region of CBP ..................................................................... 133	
   4.1.2 Biological and structural features of HIF-1α ................................................ 137	
   4.1.2 Discovery of the intramolecular NRID-TAZ1 domain-interaction in CBP..... 139	
   4.1.3 Significance of the study.............................................................................. 142	
   4.2 Results .............................................................................................................. 143	
   4.2.1 Discovery of the intra-molecular interaction in CBP .................................... 143	
   4.2.2 Characterization of the TAZ1-binding component in the N-terminal region of CBP ...................................................................................................................... 143	
   4.2.3 Structural and dynamic characterization of the N-terminal region of CBP .. 147	
   4.2.4 Identification of the TAZ1-binding sequence in NRID.................................. 151	
   4.2.5 Characterization of NRID-TAZ1 interaction by surface plasmon resonance160	
   4.3 Discussion ........................................................................................................ 162	
   4.4 Conclusions...................................................................................................... 173	
   4.5 Materials and methods .................................................................................... 173	
   4.5.1 Cloning ........................................................................................................ 173	
    vii	
   4.5.2 Protein expression and purification ............................................................. 174	
   4.5.3 NMR spectral assignment ........................................................................... 175	
   4.5.4 HSQC-monitored titrations .......................................................................... 176	
   Chapter 5 - Concluding remarks ........................................................177	
   5.1 Future directions.............................................................................................. 180	
   References..............................................................................................183	
     viii	
   List of Tables Table 3.1 Thermodynamic parameters of Ets11-138/TAZ1  domain binding ................... 94	
   Table 3.2 Residues of Ets1 functional in transcriptional superactivation..................... 117	
    ix	
   List of Figures Figure 1.1 Overview of eukaryotic transcription............................................................... 2	
   Figure 1.2 Phosphorylation of transcription factors ....................................................... 13	
   Figure 1.3 Regulation of transcriptional activity of Ets1................................................. 24	
   Figure 1.4 Regulation of DNA-binding of Ets1 ETS by phosphorylation........................ 26	
   Figure 1.5 Domain structure of murine CBP.................................................................. 31	
   Figure 2.1 Diversity of the N-terminal region of Ets family PNT domain........................ 43	
   Figure 2.2 Identification of Thr38 and Ser41 as phospho-acceptors by chemical shift changes .................................................................................................................. 48	
   Figure 2.3 In vivo studies of phosphorylation-dependent transactivation of Ets1.......... 50	
   Figure 2.4 Solution structures of Ets129-138 and 2P-Ets129-138 ........................................ 53	
   Figure 2.5 Ets2 PNT domain closely resembles that of Ets1......................................... 55	
   Figure 2.6 Chemical shift-based secondary structure analysis of Ets129-138 and 2P- Ets129-138 ................................................................................................................. 57	
   Figure 2.7 15N – relaxation and amide hydrogen exchange analyses of Ets129-138 and 2P-Ets129-138 demonstrate that helix H0 is dynamic................................................ 58	
   Figure 2.8 Helix H0 of Ets11-138 is conformationally dynamic and sensitive to trypsin proteolysis. ............................................................................................................. 61	
   Figure 2.9 Chemical shift perturbations upon phosphorylation of Ets129-138 .................. 64	
   Figure 2.10 Methyl-methyl NOE comparison between Ets129-138 and 2P-Ets129-138....... 65	
   Figure 2.11 The pKa values of pThr38 and pSer41 in 2P-Ets129-138.............................. 67	
   Figure 2.12 Dynamic helix H0 is in a conformational equilibrium .................................. 70	
    x	
   Figure 2.13 Co-linear amide chemical shift changes in the Ets1 PNT domain upon deletions or phosphorylation suggest that helix H0 exists in an equilibrium between "open" and "closed" states...................................................................................... 71	
   Figure 2.14 Identification of the minimal PNT domain using deletion mutants .............. 74	
   Figure 2.15 Overlaid 15N-HSQC spectra of various Ets1 deletion mutants ................... 76	
   Figure 2.16 Roles of helix H0 in MAPK docking and in disrupting self-association....... 81	
   Figure 3.1 Affinity column pull-down assays for the phosphorylation-dependent interaction of CBP and Ets1/2 ................................................................................ 92	
   Figure 3.2 TAZ1 domain-binding interface on Ets1 identified by 15N-HSQC monitored titrations .................................................................................................................. 96	
   Figure 3.3 TAZ1 domain-binding to Ets129-138 confirmed by 13C-HSQC monitored titrations .................................................................................................................. 99	
   Figure 3.4 TAZ1 domain-binding interface on Ets129-138 confirmed by mainchain and sidechain spectral changes .................................................................................. 101	
   Figure 3.5 Helix H0 and the phosphoacceptors are key to TAZ1 binding ................... 104	
   Figure 3.6 Helix H0 and the phosphoacceptors are key to TAZ1 binding ................... 105	
   Figure 3.7 The PNT domain is also required for TAZ1 binding ................................... 107	
   Figure 3.8 Ets1-binding interface on the TAZ1 domain ............................................... 109	
   Figure 3.9 Identification of the CBP TAZ1 domain binding interface on Ets129-138 ...... 111	
   Figure 3.10 Identification of the Ets1 binding interface on the CBP TAZ1 domain...... 113	
   Figure 3.11 Binding of Ets11-138 and the TAZ1 domain is disrupted by high ionic strength .............................................................................................................................. 114	
   Figure 3.12 In vitro mutational analysis on the Ets1-CBP binding............................... 119	
   Figure 3.13 Model of the phosphorylation-enhanced interaction of Ets1 and CBP ..... 121	
    xi	
   Figure 3.14 Position of H0 regulates the function of other PNT/SAM domains ........... 128	
   Figure 4.1 Charge clamps for binding LXXLL motif ..................................................... 135	
   Figure 4.2 Regulation of hypoxia inducible factors ...................................................... 138	
   Figure 4.3 Detection of the intramolecular NRID-TAZ1 domain-interaction in CBP .... 141	
   Figure 4.4 NMR spectroscopic confirmation that CBP1-356 binds the TAZ1 domain. ... 145	
   Figure 4.5 N-terminal CBP constructs used to map the TAZ1-binding site ................. 146	
   Figure 4.6 NMR characterizations of the N-terminal CBP constructs.......................... 148	
   Figure 4.7 HSQC-monitored titrations of TAZ1 with CBP1-101, CBP28-82, and CBP28-57 153	
   Figure 4.8 Identification of the TAZ1 interfaces for binding CBP1-356, CBP1-101 and CBP28-82 ................................................................................................................ 155	
   Figure 4.9 Titrations of various NRID constructs with TAZ1........................................ 158	
   Figure 4.10 Mapping of the TAZ1 domain-binding sequence of CBP and their conformational changes. ...................................................................................... 159	
   Figure 4.11 Free and bound structures of TAZ1 and TAZ2......................................... 164	
   Figure 4.12 Investigating the common binding interface on TAZ1 .............................. 167	
   Figure 4.13 Model of the putative autoregulatory mechanism for binding TAZ1 ......... 172	
     xii	
   List of Symbols ATP  adenosine 5’-triphosphate BLAST basic local alignment search tool D2O  deuterium oxide Da  Dalton DNA  deoxyribonucleic acid DTT  dithiothreitol EDTA  ethylenediaminetetraacetic acid ERK  extracellular signal-regulated kinase GST  glutathione S-transferase HSQC  heteronuclear single quantum coherence HX  hydrogen exchange IPTG  isopropyl-βD-thiogalactopyranoside ITC  isothermal titration calorimetry Kd  dissociation constant kDa  kiloDalton MALDI-TOF matrix-assisted laser desorption/ionization-time of flight MAPK  mitogen activated protein kinase MS-MS tandem mass spectrometers NMR  nuclear magnetic resonance NOE  nuclear Overhauser effect NOESY nuclear Overhauser effect spectroscopy PCR  polymerase chain reaction PDB  protein database (http://www.rcsb.org/pdb/)  xiii	
   pI  isoelectric point RDC  residual dipolar coupling RNA  ribonucleic acid S2  order parameter SDS PAGE sodium dodecyl sulfate polyacrylamide gel electrophoresis SSP  secondary structure propensity TOCSY total correlation spectroscopy   xiv	
   Acknowledgements The work presented in this thesis would not have been achieved without the help of many people in many different ways.  First of all, I am indefinitely indebted to my supervisor Lawrence for believing in me and giving me his full support to get me through this long journey of PhD study. The opportunity to work for Lawrence, especially with this amazing environment in Vancouver, undoubtedly improved my scientific education and, more importantly, also helped me to learn how to enjoy life outside of the lab. I am grateful to Manuela and Cameron for introducing me to the exciting PNT domain project, and for help in getting it off the ground. I thank Ira and Desmond for their help on many aspects of this project and their motivation and exceptional scientific skills in the lab. I also enjoyed working with Greg on an earlier and very challenging project regarding ETS autoinhibition. Without Mark, NMR experiments would have never been run so smoothly. I appreciated his presence in the NMR lab to help troubleshooting and to be available for many helpful discussions. I would also like to acknowledge all the current and former lab members (Eric, David, Gen, Markus, Patrick, Simon, Jerome, Adrienne, Helen, Shaheen, Martin, Gary, Chris, Mario, Matt, Wes, Meena, and Daigo) for their help in the lab. The exciting results of Ets1 PNT would not have been completed without the collaboration with Mary in Barbara Graves’ lab. I also thank my committee members, George and Natalie, for their guidance through my PhD study. This holds particularly for George, who was always available for my committee meetings, even during the time of his busy administrative work for the university. His careful comments on my thesis were invaluable.  xv	
   Rock climbing became my new passion since Manuela took me to the climbing gym and Lawrence taught me how to build an anchor for multi-pitch climbs. Being out and climbing with my climbing buddies were certainly one major way to keep me sane from the occasional frustration coming from science. I would like to thank all my climbing friends (Alex, Eric, Hans, Wayne, Allan, Rotem, Gen, Chris, Karena, Markus) for getting out, mostly in Squamish, with me and having great time together.  For my family, I would like to thank my parents for their endless moral support,despite the distance from South Korea, and for understanding the somewhat different lifestyle of mine. My sister, Myung-Seo, was always reachable and available to hear the goods and bads in my life. I would like to thank Siggi for her tremendous support in helping me through the toughest stage of my PhD study and always being there as a great partner in science, climbing, skiing, mountian biking, paddling, drinking, parenting, travelling and many more. I am also grateful for her helpful comments on this thesis. Lastly, I thank my little boy, Timo, for giving me a big smile everyday when I come home.	
  	
   Chapter 1 – Introduction 	
    1	
   Chapter 1 - Introduction 1.1 Eukaryotic gene expression Regulation of gene expression is a central and indispensible process in any living organism. Eukaryotic gene expression is controlled primarily during transcription, employing various cis- and trans-regulatory elements in a temporal and spatial manner (Panning and Taatjes, 2008). Therefore, key to the eukaryotic transcriptional regulation is managing the complex interplay of these regulatory elements, from the initial recognition of DNA sequences to the recruitment and activation of the RNA polymerase complex (Figure 1.1). Many levels of transcriptional regulation are imposed to prevent aberrant gene expression that can subsequently cause malfunctions or even lethality to the host.  1.1.1 Components of transcriptional regulation in eukaryotes The cis- and trans-regulatory elements in transcription largely refer to specific DNA sequences/regions and regulatory proteins, respectively. In eukaryotes, promoter regions, such as the TATA-box and the distally-located enhancer sequences, are the common cis-regulatory elements (Figure 1.1). The trans-regulatory elements, including specific transcription factors, bind these DNA sites and/or other proteins to facilitate multiple layers of regulation.  Transcriptional co-regulators, especially co-activators, often serve as scaffolds or bridges to mediate the direct or indirect interactions among regulatory elements and/or enzymatically alter their interactions by catalyzing post- translational modifications, such as acetylation. Various common or general Chapter 1 – Introduction 	
    2	
          Figure 1.1 Overview of eukaryotic transcription Activation of the eukaryotic transcription occurs as (i) specific regulatory proteins (transcription factors, blue) bind the upstream enhancer region, (ii) RNA polymerase forms a preinitiation complex (yellow) with general transcription factors (TF) around the core promoter region, (iii) transcriptional co-regulators, mediators, and chromatin remodelers (green) serve as scaffolds or bridges for binding, or provide acetyltransferase activity (HAT) for histone modification (Ac, acetylation). Repression of gene expression is driven as the co-repressor complex recruits components, such as histone deacetylase (HDAC) or a silencing mediator, for switching the chromatin structure from the active to inactive state. The deacetylated histone tails support the tightly-packed and inactive heterochromatin structure. See text for more details.  !"#$%&'( )(%*+,($-&$'*%./ ,'%,&$0%&'( 12) 3'./44 )5441 )5446 )5447 )83 )5442 )5448 )5445 "*9%*," ( "*9%*,"( )2)2 :9(';%&$* (";'#".."( 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, )(%*+,($-&$'* 5%,&'( 172: 172: 12) 172: ,'("-("++'( Chapter 1 – Introduction 	
    3	
   transcription factors, such as the TATA-box binding protein (TBP), are involved in the assembly of the RNA polymerase complex to the core promoter region.  1.1.2 Levels of transcriptional regulation in eukaryotes Post-translational modifications play a pivotal role in managing temporal and spatial regulation of the three major players in transcription, namely, transcription factors, transcriptional co-regulators, and the pre-initiation complex (PIC), by influencing their localization, activity, and stability. Although many types of post-translational modifications exist, this chapter introduces only those most relevant to this thesis.  Post-translational modifications Phosphorylation. Phosphorylation is one of the best characterized and the most widespread on the list of post-translational modifications	
   (Tarrant and Cole, 2009). Addition of a phosphate on a serine, threonine, or tyrosine hydroxyl group of a target protein by a specific kinase directly impacts on numerous aspects of its function, including protein-protein/-DNA interactions, cellular localization, folding and stability, and catalytic activity. The reverse process, dephosphorylation, is mediated by various phosphatases that are generally classified by their substrate specificity (Ser/Thr, Tyr, or dual specificity).  Acetylation and methylation. Acetylation and methylation play key roles in transcriptional regulation, especially by modification of histones and many transcription related proteins (Kouzarides, 2002; Yang and Seto, 2008). Acetyltransferases catalyze the transfer of the acetyl group from acetyl-CoA to the ε-amino group of a lysine. This Chapter 1 – Introduction 	
    4	
   serves to both neutralize the positive charge of a lysine, and to provide a binding moiety for proteins with bromo domains. Acetylation of histones generally provokes changes in chromatin structure from transcriptionally-inactive heterochromatin to the more lightly- packed, active euchromatin by disrupting the electrostatic interaction with DNA. Similarly, histone lysine methyltransferases (HKMT) or protein arginine methyltransferases (PRMT) transfer a methyl group from S-adenosylmethionine to the ε-amine of lysine or the guanido group of arginine, respectively. In contrast to acetylation, methylation can occur to different levels, such as mono-, di-, and trimethylation for lysine and mono- and dimethylation for arginine. This provides extra complexity and specificity for transcription regulation, such as through binding proteins containing chromo- or PHD domains (Cavalli and Paro, 1998; Shi et al., 2006). Furthermore, methylation appends hydrophobic groups to otherwise polar sidechains for further intra- or inter-molecular interactions. Both modifications are reversible via different types of deacetylases (HDAC and SIRT proteins)	
   (Yang and Seto, 2007) and demethylases (LSD1, or deiminase PADI4)	
  (Bedford and Clarke, 2009).  Sumoylation. Sumoylation refers to the process of conjugating a small ubiquitin-related modifier (SUMO) to a target protein (Geiss-Friedlander and Melchior, 2007). Although the structure of SUMO and its conjugation mechanism are strikingly similar to those of ubiquitin, SUMO and ubiquitin share less than 20% sequence identity and have distinct physicochemical properties. Most importantly, unlike ubiquitination, which generally targets proteins for proteosomal degradation, the majority of sumoylation events alters the localization, activity, and stability of target proteins. SUMO also provides a docking- site for protein partners containing a SUMO interacting motif (SIM). The reversible Chapter 1 – Introduction 	
    5	
   sumoylation pathway requires, at least, three enzymes, namely, the E1 activating, the E2 conjugating, and one of several possible E3 ligases in a sequential manner, to attach the Gly-Gly C-terminal end of SUMO to the lysine side chain of the target protein via an isopeptide bond. Often, the lysine lies within the sumoylation concensus sequence ΨKXE (Ψ: aliphatic branched amino acid, X: any amino acid). This post- translational modification is also reversible through the activity of SUMO-specific proteases.  Regulation by transcription factors The term transcription factor usually refers to a sequence-specific DNA-binding protein that recognizes enhancer or promoter regions of DNA to activate or repress RNA polymerase activity. In eukaryotes, and especially for vertebrates, several transcription factors are invariably involved in binding the enhancer region in a cooperative manner to form an enhanceosome complex (Merika and Thanos, 2001). This complex synergistically activates gene expression through recruitment of other transcriptional components, such as co-regulators and general transcription factors, and eventually the RNA polymerase complex. Combinatorial assembly of such a high-order three- dimensional enhanceosome complex is advantageous for promoting specificity in gene transcription as the correct number of transcription factors is required in a spatially and temporally controlled manner.  The human genome encodes a remarkably high number of specific transcription factors (~2600). These are ~10 times more abundant than other transcription-related proteins (~200-300), such as coactivators and the transcriptional machinery proteins (Brivanlou Chapter 1 – Introduction 	
    6	
   and Darnell, 2002). Such a large pool of transcription factors creates a more diversified pool of enhanceosomes to provide higher specificity in regulating the complex processes of myriad gene expression.  Most eukaryotic transcription factors are modular, bearing independent structural domains that mediate functions including DNA-binding, transactivation, and protein- protein interactions. Indeed, based on the sequence and structural conservation of these domains, combined with common features of their DNA target sites, transcription factors fall within well-defined families. For example, the basic region leucine zipper (bZIP) and basic region helix-loop-helix (bHLH) families are some of the best-known transcription factor families	
   (Ellenberger, 1994), which are classified by their DNA- binding domain structures. The members of these two families form either homotypic or heterotypic dimers through the C-terminally located leucine zipper or helix-loop-helix to correctly position a basic region into the major grooves of two distinct DNA sequences. As an example, the cellular proto-oncogene c-Fos forms a heterodimer with c-Jun through their bZIP domains to bind a specific target DNA sequence	
   (O'Shea et al., 1992). The cell proliferation-related Myc family bears a basic region with both a leucine zipper and helix-loop-helix. Myc dimerizes with other members of the family (Dang et al., 1991). Yet another Myc family member, Max, form only homodimers	
   (Amati et al., 1993).  Combinatorial association of domains within a single polypeptide chain is also a common strategy for the evolution of eukaryotic transcription factors to gain specificity and affinity for discriminating among DNA sequences. For example, the ETS Chapter 1 – Introduction 	
    7	
   transcription factor family is classified by its DNA-binding ETS domain, which contains a winged helix-turn-helix motif	
   (Sharrocks, 2001). Members of this family are discriminated further by the presence of additional domains, such as a protein-protein interaction domains (PNT, OST) and a transactivation domain (TAD). For example, GABPα, GA-binding protein, is yet another member of the ETS transcription family that is structurally very similar to the prototypical members Ets1/2. However, in addition to the PNT and ETS domains, GABPα bears an additional protein interaction domain OST, that is localized N-terminal to the PNT domain and is unique in the ETS family, as described further in my MSc thesis (Kang, 2005; Kang et al., 2008).  Common mechanisms for regulating the activities of transcription factors include controlling of their expression, subcellular localization, stability, and interactions. A well- known tumor suppressor, p53, is comprised of four modular domains, namely, a proline- rich region, the core DNA-binding domain, the tetramerization domain, and the basic domain (Kruse and Gu, 2009). Numerous post-translational modifications, such as acetylation, phosphorylation and ubiquitination, are involved in regulating the activities of these individual domains to control the overall function of p53 (Kruse and Gu, 2009). In the ETS transcription family, phosphorylation of Ets1/2 by two signal-transduction pathways is critical for transcriptional regulation (see below in this chapter and in chapter 2).  Regulation of transcriptional co-regulators Transcriptional co-regulators play the central role in mediating activation and repression of transcription as they provide the structural framework, enzymatic activities, and Chapter 1 – Introduction 	
    8	
   external signal relays during transcriptional assembly. Most commonly, transcriptional co-regulators act as bridges or scaffolds to facilitate the interactions among transcription factors with the RNA polymerase complex. As described later in this chapter, the co- activator CBP interacts with ~300 transcription factors and many others via one or more of its constituent modules (Vo and Goodman, 2001). Protein-protein interactions are also found among the co-regulator scaffolds with domains containing enzymatic activities (e.g. kinases, methyltransferases, acetyltransferases) resulting in the formation of a multi-subunit co-regulator complex, like the steroid receptor coactivator (SRC)-3	
   (Lonard et al., 2009), to efficiently mediate gene expression in an orderly manner	
   (Lonard and O'Malley, 2005). In fact, many of these catalytic domains are integrated within the co-regulators, as exemplified by acetyltransferase domains in SRC and CBP/p300 (Sterner and Berger, 2000) and methyltransferase domains in CARM1 and PRMT1 (Bannister and Kouzarides, 2005). As enzymes, the co-regulators modify various targets, including the well-known nucleosomal histones, transcription factors, or even the co-regulators themselves.  The post-translational modification of histone tails can affect transcription ‘directly’ by triggering structural changes in chromatin, commonly through charge neutralization by acetylation	
   (Jenuwein and Allis, 2001). However, more interestingly, post-translational modifications of histone tails also ‘indirectly’ regulate transcription, as the distinct modification patterns, commonly referred to as the ‘histone code’, provide binding sites for various effector proteins, such as chromatin remodelers, general/site-specific transcription factors, and other co-regulators/mediators. For example, methylation of Lys4, Lys36, and Lys79 in histone H3 promotes transcriptional activation, whereas Chapter 1 – Introduction 	
    9	
   methylation of Lys9 or Lys27 in histone H3 or Lys20 in histone H4 is related to transcriptional repression	
   (Bannister and Kouzarides, 2005). Acetylation, phosphorylation, and arginine methylation, mostly in histones H3 and H4, are generally considered as favoring transcriptional activation. In contrast, sumoylation results in transcriptional repression as it often competes with acetylation for the same site or region or recruits other repressors  (Berger, 2007).  Commonly, the enzymatic domains of co-regulators coexist in proximity with domains such as the chromo-, PHD, or bromo-domain, which can interact with post- translationally modified residues within the same co-regulator to further enhance their activities. For example, bromodomains are found near the histone acetyltransferase domains in CBP/p300	
   (Yang, 2004) and GCN5	
   (Syntichaki et al., 2000), whereas the neighboring member of GCN5 in the yeast SAGA complex, chromatin remodeler CHD1, carries a chromodomain, to help stabilize the resulting complexes (Cavalli and Paro, 1998).  Last, co-regulators, especially co-activators, act as integrators of external signaling and transcription factors. In contrast to histone code modification that is specific for the surrounding gene, post-translational modifications of co-activators influence a broad range of gene expression. That is, by interacting with a variety of transcription factors, they regulate multiple sets of genes. For example, varying levels of phosphorylation of SRC-3 by diverse phosphorylation-signaling pathways are critical in determining the range of transcription factors that the co-activator can interact with	
   (Lonard and O'Malley B, 2007). In a moderate phosphorylation state, SRC-3 binds only certain Chapter 1 – Introduction 	
    10	
   transcription factors, such as nuclear factor-κB. However, once extensively phosphorylated, SRC-3 is capable of recruiting a larger pool of transcription factors	
   (Lonard and O'Malley B, 2007).  Regulation of PIC assembly Recruitment of the RNA polymerase II complex requires roughly 44 polypeptides involved in the ordered assembly of the PIC (Roeder, 2003). This process initiates as the monomeric TBP (TATA-binding protein) in TFIID, or by itself, binds to the core promoter region, often containing a TATA box, in the nucleosome free region (NFR). The additional interactions of TBP with both TFIIA and TFIIB are important in forming a stable complex of TBP with the TATA box. Subsequently, other RNA polymerase- specific general transcription factors are assembled on the promoter region, including TFIIF/Pol II, TFIIE, and TFIIH (Figure 1.1) (Hahn, 2004). Furthermore, transcriptional co-regulators and mediator complexes are involved in direct interactions with these general transcription factors and RNA polymerase (Roeder, 2005). In a model of activated transcription via such co-regulators, the large mediator complex binds the enhanceosome complex in the upstream DNA sequence, bridging it to the preinitiation RNA polymerase complex to direct subsequent gene expression. Chapter 1 – Introduction 	
    11	
   1.2 Phosphorylation of transcription factors Phosphorylation is one of the most prevalent and versatile cellular strategies for reversibly regulating the activity of proteins. About one thousand kinases and several hundred phosphatases are predicted to be encoded by the human genome. These catalyze the phosphorylation and dephosphorylation, respectively, of ~30% of all cellular proteins (Cohen, 2000). Therefore, phosphorylation is indispensible in regulating a vast majority of cellular processes, including most metabolic pathways, signaling cascades, and gene expression. Abnormality in phosphorylation can result in major diseases, such as cancer.  Such a great prevalence of phosphorylation originates from at least two reasons. First, phosphorylation exploits the universal energy source, ATP (adenosine 5’-triphosphate). Second, by changing the physicochemical properties (including electrostatic charge) of several possible acceptor amino acid side chains, phosphorylation targets the most fundamental molecular behaviors of proteins, namely their structures and intermolecular interactions.  Phosphorylation is involved in almost every step of the signal cascade to relay an external signal from a membrane receptor to its downstream (nuclear) components. Phosphorylation also contributes to the majority of protein-protein and protein-DNA interactions occurring in the enhanceosome and in chromatin structure, as required for transcription regulation. Not surprisingly, phosphorylation of transcription factors, as is the focus of this thesis, also functions virtually in any conceivable way for regulating their activities.  Chapter 1 – Introduction 	
    12	
   1.2.1 Mechanistic view of protein phosphorylation Phosphorylation-induced structural changes regulate the interaction of transcription factors with other components of the transcriptional machinery, including DNA. The types of structural changes in phosphorylated proteins range from as little as the simple, localized addition of a negatively-charged phosphate moiety to more long-range allosteric effects, including folding-unfolding transitions (Figure 1.2 (A)).  Such detailed mechanistic insights can only be gained from high resolution structural and dynamic studies of proteins.  Allosteric mechanism An allosteric mechanism generally refers to the case when a change at one site of a protein, such as binding other molecules or post-translational modifications, affects the function of a ‘distally’-localized region of the protein by propagated structural or dynamic changes. Phosphorylation-induced allosteric effects broadly include any conformational change spanning the spectrum from relatively localized effects to, in the most pronounced case, global folding or unfolding (Figure 1.2 (A) and (B)). For example, in the case of the MAPK (mitogen-activated protein kinase), ERK (extra-cellular signal- regulated kinases), double phosphorylation at the lip of the catalytic loop re-orients its local conformation as required for catalytic activity (Canagarajah et al., 1997). Phosphorylation of STAT (signal transducers and activator of transcription) proteins has been shown to rearrange the conformation of their dimeric forms from anti-parallel to parallel (Whitmarsh and Davis, 2000). In the case of DNA-binding, phosphorylation of Chapter 1 – Introduction 	
    13	
                     Figure 1.2 Phosphorylation of transcription factors (A) Phosphorylation affects the conformation of the target proteins from as little as simply adding a negatively charged phosphate group (steric) up to regulating folding- unfolding transitions to allosterically to activate or repress their activities. (B) In the allosteric mechanism, phosphorylation on transcription factor (TF) triggers a local conformational change to provide a better binding moiety for protein B. In the steric mechanism, a phosphorylated residue on protein A can attract other proteins, protein D, without major structural changes. In reverse, phosphorylation can also block an interaction sterically, as seen between protein A and C, due to the addition of negative charges in the binding site. (C) Phosphorylation of transcription factors (TF) can regulate their transcriptional activities through modulating their stability, subcellular localization, DNA-binding, and interaction with other proteins, such as co-regulators (C) or mediators (M). Chapter 1 – Introduction 	
    14	
   !" # $ %!" !" &''()*+,-. /*+,-. 0%123+.456-)*-.27-+82(92:4():4(,;'5*-(6 0#12<4():4(,;'5*-(62(92*,56).,-:*-(6295.*(,) )*5=-'-*; !" '(.5'->5*-(6 !" !" $?&@=-6A-6B :,(*+-6 -6*+,5.*-(6 3 # .(69(,C5*-(65'2.456B+ 9('A-6B@D69('A-6B 05''()*+,-.1 )C5'' '5,B+ 5AA-*-(65'20@12.45,B+ 0)*+,-.1 :4():4(,;'5*-(6 0&12E56B+2(92:4():4(,;'5*-(62+99+.*) F-65)+ :4 () :4 5* 5) + :4 () :4 5* 5) + F-65)+ Chapter 1 – Introduction 	
    15	
   IRF3 (interferon regulated transcription factor) triggers unfolding of a structured helix to expose its DNA-binding interface (Qin et al., 2003). The phosphorylation-induced folding-unfolding transition is also closely linked to changes in dynamics. Phosphorylation of the SRR (serine-rich region) adjacent to the ETS domain of Ets1 represses its DNA-binding activity by allosterically shifting the protein from a flexible, active state to a more rigid, inactive state (Pufall et al., 2005) (described in section 1.4.1).  Steric mechanism In a steric regulatory mechanism, conformational changes are minimal upon phosphorylation (Figure 1.2 (A) and (B)). The covalent attachment of a negatively- charged phosphate at a particular phosphoacceptor can sterically (i.e. via van der Waals interactions, including hydrogen bonding) or electrostatically shape a binding interface either to attract or repel other partners, including proteins and DNA. The interaction of a phosphorylated tyrosine kinase with a SH2 (src homology 2) domain is a typical example of local, phosphorylation-dependent docking mechanism (Pawson and Scott, 2005). The interaction of the CREB (cAMP response element binding protein) KID domain with the CBP (CREB binding protein) KIX domain is another example by which phosphorylation of KID triggers the binding of KIX via two induced-helices in KID (Radhakrishnan et al., 1997). However, upon phosphorylation per se, KID remains unstructured, as hinted by the presence of the phosphoacceptor site located only at the N-terminal end of the second binding-induced helix in KID (Radhakrishnan et al., 1997). In the NF-AT (nuclear factor of activated T-cells) transcription factor family, Chapter 1 – Introduction 	
    16	
   phosphorylation plays a key role in regulating the accessibility of localization sequences by sterically hindrance (Macian et al., 2001).  Phosphorylation is also frequently found in the N-terminal end of an α-helix on an N-cap residue, which stabilizes the unsatisfied hydrogen bonding in the N-terminal region of an α-helix and interacts with the helix dipole. NMR and CD spectroscopy-based studies of histidine-containing protein (HPr) of the phosphoenolpyruvate:sugar phosphotransferase system have shown that phosphorylation of Ser46 located at the N- terminal end of the α-helix B stabilizes the protein to solvent and thermal denaturation with only small and local changes in structure (Pullen et al., 1995). A similar case is seen with the phosphorylation of Ser113 in isocitrate dehydrogenase (Hurley et al., 1990). In these examples, no conformational change was observed upon phosphorylation, thereby exhibiting features of a steric mechanism. However, phosphorylation at the N-cap of a helix does influence the cooperative stability of the entire helix, thereby exhibiting allosteric properties as well. This illustrates that post- translational modifications are often not purely "steric" or "allosteric", but rather have effects lying along a continuum between these two endpoints (Figure 1.2 (A)).  1.2.2 Functions of transcription factor phosphorylation The functional diversity of phosphorylation can be easily found in many aspects of transcription factors, including modulation of their stability, subcellular localization, DNA- binding activity, and interactions with other components of the transcriptional machinery (Figure 1.2 (c)). Due to its central role in regulation, abnormalities in phosphorylation Chapter 1 – Introduction 	
    17	
   can result in aberrant gene expression, and consequently a wide range of diseases in development and differentiation.  Regulation of in vivo protein stability Phosphorylation can both prevent and trigger protein degradation in vivo. Under normal conditions, the p53 tumor suppressor turns over quickly through the ubiquitination- proteosome pathway. However, upon external stresses, multiple phosphorylation events occur in the N-terminal region to block the interaction of p53 with the ubiquitination ligase E3, Mdm2, and therefore prevent its degradation (Appella and Anderson, 2001). Alternatively, dephosphorylation of a C-terminal phosphoserine also prevents p53 degradation as a 14-3-3 protein binds the dephosphorylated C-terminal region of p53 and thereby blocks its interaction with Mdm2 (Waterman et al., 1998; Yang et al., 2003). Phosphorylation-dependent proteolysis is also seen for IκB (anchoring protein inhibitor) in complex with NF-κB (nuclear factor-κB) (Ghosh et al., 1998). In contrast to p53, phosphorylation triggers the degradation of the NES (nuclear exclusion sequence)-containing IκB, thus exposing the NLS (nuclear localization sequence) of NF-κB (Zandi and Karin, 1999).  Regulation of subcellular localization Phosphorylation-induced localization of transcription factors occurs in at least two ways: (i) directly masking or unmasking the localization sequences (NES, NLS), or (ii) regulating the interaction with proteins that mask or unmask these sequences. Phosphorylating the DNA-binding domain of FOXO1 (forkhead transcription factor) by protein kinase B blocks the adjacent NLS nuclear import signal, thereby promoting the Chapter 1 – Introduction 	
    18	
   nuclear exclusion of the protein (Yaffe, 2002). Phosphorylated FOXO1 also binds 14-3- 3 proteins, increasing cytoplasmic retention (Yaffe, 2002). As described in the previous section on protein stability, phosphorylation-dependent degradation of IκB results in the exposure of the NLS of NF-κB permitting the transcription factor to shuttle into the nucleus (Zandi and Karin, 1999).  Regulation of DNA-binding Phosphorylation can regulate the DNA-binding activity of a transcription factor both directly and indirectly. Introduction of negative charges in the DNA-binding domain can interrupt the electrostatic interaction with DNA, as seen for the repression of Wilms tumor gene product, WT1, upon phosphorylation in its zinc-finger region (Sakamoto et al., 1997). Phosphorylation adjacent to, but not in the DNA-binding domain, can often regulate the DNA-binding activity either through steric or allosteric mechanisms, such as those involving STAT dimerization (Darnell, 1997), the autoinhibitory elements of IRF3 (Qin et al., 2003), and the serine rich region (SRR) of Ets1 (Pufall et al., 2005). Phosphorylation-dependent binding of p53 to 14-3-3 and phosphorylation of C-terminal residues in c-Jun were also shown to inhibit DNA-binding (Appella and Anderson, 2001).  Regulation of interactions with other transcription related proteins Transcription factors interact with numerous transcriptional co-regulators, mediators and components of the basal transcription complex. As an example, the transcriptional co- activator CBP interacts with many transcription factors especially via its Cys/His-rich- and KIX domains. As described earlier, phosphorylation of CREB KID triggers the Chapter 1 – Introduction 	
    19	
   interaction with CBP KIX (Radhakrishnan et al., 1997). Different regions of the tumor suppressor p53 were also shown to interact with various domains in CBP and p300 in a phosphorylation-dependent manner (Teufel et al., 2007; Ferreon et al., 2009). Phosphorylation of the tumor suppressor RB (retinoblastoma) protein disrupts its interaction with the E2F transcription factors, leading to de-repression (Mittnacht, 1998). In the basal transcriptional complex, various factors bear kinase domains to phosphorylate other members of the complex. For example, TFIIH plays a key role in promoting the capping of the nascent RNA sequence to protect from degradation by phosphorylating the CTD region of RNA polymerase (Svejstrup et al., 1996). TAFs (TATA-binding protein associated factors) were also shown to phosphorylate other transcription factors, such as a subunit of TFIIF (Orphanides et al., 1996). Chapter 1 – Introduction 	
    20	
   1.3 ETS Family transcription factors The ETS transcription factor family plays key roles in regulating gene expression for various biological processes, including cellular differentiation, development, transformation, and proliferation (Graves and Petersen, 1998). Since the discovery of the founding member, Ets1, as the oncogene of avian E26 transformation specific retrovirus, more than 30 transcription factors from diverse metazoan species have been identified in this family.  ETS proteins are identified by the conserved ~ 85 amino acid DNA-binding ETS domain. The ETS domain contains a winged helix-turn-helix DNA-binding motif, which recognizes a consensus core DNA sequence, 5’-GGA(A/T)-3’. High-resolution structures of ten ETS domains, in isolation or in complex with DNA, are available in the RCSB Protein Data Bank, namely Ets1, SAP-1, GABPα, SPDEF, FLI-1, PU.1, ELK-1, ETV6, Elf3 and Elf5.  Autoinhibition is a common strategy for regulating DNA-binding activity for the ETS transcription factors (Pufall and Graves, 2002). However, the molecular mechanisms of autoinhibition differ among family members. In the cases of Ets1/2, a discontinuous inhibitory module, located both N- and C-terminal to the core ETS domain, allosterically changes the affinity for DNA-binding as it stabilizes the inhibited conformation by dampening the dynamics of an internal core of hydrophobic residues (Pufall et al., 2005). Upon DNA binding, a conformational change occurs, highlighted by the unfolding of a helix in the inhibitory module. The energetic coupling of this unfolding transition thus attenuates DNA binding. In the case of Tel (or ETV6), a C-terminal inhibitory Chapter 1 – Introduction 	
    21	
   domain (CID), sterically blocks the DNA-binding interface of the adjacent ETS domain (Green et al., 2010). Interestingly, a linker inhibitory damper (LID), located N-terminal to the ETS domain, may compete for binding to the CID, thus de-repressing autoinhibition (Green et al., 2010). In contrast, members of the ETS ternary complex factor (TCF) subfamily bind DNA cooperatively by interacting with the serum response factor (SRF) through the B-box located adjacent to their ETS domains (Buchwalter et al., 2004). As seen for two of these members, SAP-1 and ELK-1, the DNA-binding activity becomes de-repressed as phosphorylation of the C-terminally located transactivation domain (TAD) triggers conformational changes to disrupt the interaction between the TAD and ETS domains (Yang et al., 1999). Antagonistic, sumoylation of the inhibitory domain, called the R motif, is critical in repressing the transcription activity of ELK-1, possibly due to its role in recruiting histone deacetylases (Yang et al., 2003).  The second common feature of approximately one-third of the ETS family members is the presence of a protein interaction domain, PNT (pointed) (Slupsky et al., 1998), which contains a SAM (sterile alpha motif) fold (Ponting, 1995). Although the nomenclature of this domain is rather ambiguous as PNT and SAM are used interconvertibly in the literature, in the context of this thesis, the PNT domain only refers to those present in ETS family members (Mackereth et al., 2004). The PNT domain is involved in regulating gene expression through protein interactions either to itself, to other PNT domains, or to other transcription related proteins.  Polymerization of the PNT domain of Tel and Yan leads to transcriptional repression. In the transcriptional repressor Tel, and as well as its Drosophila ortholog Yan, head-to-tail Chapter 1 – Introduction 	
    22	
   self-association of the PNT domain occurs through their so-called ML (Mid Loop; Met57, Ala61, Leu64, Leu65) and EH (End Helix; Phe44, Leu46, Val80, Leu84) surfaces allowing the formation of an extended, helical polymer along the DNA (Kim et al., 2001). This is hypothesized to mediate formation of a higher order chromatin structure for transcriptional repression (Kim et al., 2001). In Drosophila eye development, the transcriptional repression by self-polymerization of Yan PNT is regulated by at least two other proteins also with a PNT domain, Mae and Pnt-P2 (Qiao et al., 2004; Song et al., 2005; Qiao et al., 2006). Mae PNT binds Yan PNT with high affinity, resulting in depolymerization of Yan (Qiao et al., 2004). Furthermore, Mae facilitates the phosphorylation of Yan by the Rolled MAP kinase, leading to CRM1-mediated export of Yan from the nucleus, and de-repression of gene expression (Song et al., 2005). Rolled also phosphorylates Pnt-P2, the Drosophila ortholog of Ets1, enhancing its ability to act as a transcriptional activator, presumably by recruiting the co-activator CBP. Interestingly, this leads to increased expression of Mae PNT, which in turn also binds the PNT domain of Pnt-P2, This inhibits the phosphorylation of Pnt-P2 by competing for the same docking site as MAPK Rolled, and thereby attenuates transcription via a negative feedback loop (Qiao et al., 2004; Qiao et al., 2006). In the case of Ets1/2, the PNT domain first provides a docking site for MAPK ERK2 for the phosphorylation of the adjacent phosphoacceptor sites, and second, facilitates a binding interface for transcriptional coactivator CBP and possibly with other transcription components (Seidel and Graves, 2002). Interestingly, in the case of the ETS family mamber GABPα, CBP interacts with the N-terminally located OST domain rather than the centrally located PNT domain (Kang et al., 2008). Chapter 1 – Introduction 	
    23	
   1.4 Ets1 Ets1, the prototypical member of the ETS transcription factor family, is composed of three modular domains; the N-terminal protein interaction PNT domain, the central transactivation domain (TAD), and the C-terminal DNA-binding ETS domain (Figure 1.3). The activity of Ets1 (and closely related Ets2) as a transcription factor is regulated by the complex interplay of post-translational modifications and protein-protein interactions. More specifically, phosphorylation of Ets1 is regulated by two signal transduction pathways (Rabault and Ghysdael, 1994; Graves and Petersen, 1998; Slupsky et al., 1998) (Figure 1.3). Multi-site phosphorylation of an unstructured region immediately N-terminal to the PNT domain is controlled by a receptor tyrosine kinase/Ras/MAPK pathway. The PNT domain initially provides a docking site for the MAPK ERK2 to enhance the phosphorylation of the adjacent phosphoacceptor sites. This likely results from a simple proximity mechanism in which docking increases the local concentration of the phosphoacceptors Thr38 and Ser41 near the active site of the kinase. Upon phosphorylation, the PNT domain gains higher affinity for the transcriptional co-activator CBP, and therefore promotes enhanced gene expression (Yang et al., 1998a; Foulds et al., 2004). In contrast, phosphorylation in the SRR domain by the calcium/calmodulin-dependent protein kinase II inhibits the DNA-binding activity of the ETS domain as the phosphorylated, but predominantly unstructured, SRR stabilizes the inhibitory module of the ETS domain (Pufall et al., 2005). In addition, sumoylation of Lys15 near the PNT domain or Lys227 in the TAD domain represses the transcriptional activity of Ets1, possibly by recruitment of histone deacetylases (HDAC) or death-associated protein (DAXX)  (Figure 1.3) (Hahn et al., 1997; Li et al., 2000; Macauley et al., 2006; Ji et al., 2007) Chapter 1 – Introduction 	
    24	
      Figure 1.3 Regulation of transcriptional activity of Ets1 The activities of the PNT and ETS domains of Ets1 are regulated by the Ras-dependent MAPK and Calcium/calmodulin-dependent protein kinase II signal transduction pathways, respectively. Phosphorylation of the N-terminal PNT domain and the Serine Rich Region (SRR) adjacent to the ETS domain cause transcriptional activation and repression of the subsequent genes, respectively. Sumoylation in both the unstructured N-terminal region and the transactivation domain (TAD) lead into transcriptional repression, possibly by recruiting HDACs. Co-operative protein partnerships involving the ETS domain lead into transcriptional activation. The transactivation domain (TAD) and auto-inhibitory regions (I) of the ETS domain are also indicated. !"#$ $%&" "'( #))!*+, % % % %% % % )"- ).+ "/) /.0123454.+4 /.6-$$ 7484249:34++;<8 6'%- #= 6 > :.3*843# = 6 > , ??, @('/ /A% Chapter 1 – Introduction 	
    25	
   1.4.1 ETS domain in Ets1 Ets1 and Ets2 contain an ETS domain and a flanking autoinhibitory module (Figure 1.4) (Pufall and Graves, 2002). The core ETS domain, encompassing residues 331-415, consists of three α-helices (H1, H2, H3) and four β-strands (S1, S2, S3, S4) forming a winged-helix-turn-helix DNA-binding motif (Figure 1.4) (Lee et al., 2005). Unlike other ETS transcription factors, the Ets1/2 ETS domains incorporate a distinct layer of regulation involving an additional four flanking helices (HI-1, HI-2, H4, H5) These form an autoinhibitory module, modulated by the adjacent unstructured phosphoacceptor SRR (Figure 1.3 and 1.4) (Pufall et al., 2005). The autoinhibitory module packs on the ETS domain from opposite side of the DNA-binding helices (H2, H3) and thus attentuates DNA-binding predominantly via an allosteric mechanism. The most dramatic feature of this mechanism is the unfolding of helix HI-1 upon DNA-binding. Multi-site phosphorylation of the serine-rich region by CaM kinase II progressively reinforces the autoinhibitory module, thereby leading to a graded, rather than all-or-none, control of DNA binding (Figure 1.4). Our lab has shown that this ‘rheostat’-like inhibitory mechanism of ETS domain occurs allosterically as the phosphorylated SRR stabilizes the structure and dynamics of the autoinhibitory module through transient interactions (Pufall et al., 2005). Chapter 1 – Introduction 	
    26	
    Figure 1.4 Regulation of DNA-binding of Ets1 ETS by phosphorylation The helical autoinhibitory module (cyan) of Ets1 inhibits ETS domain DNA-binding through an allosteric mechanism. In this mechanism, Ets1 exists in an equilibrium between a flexible high affinity open state and a rigid low affinity state. Upon DNA- binding, an energetic penalty is paid by the unfolding of HI-1. Multi-site phosphorylation of the unstructured SRR (dashed lines) reinforces inhibition via transient interactions to stabilize the closed auto-inhibited state (cyan). The core ETS domain is shown as red helices and orange strands. (PDB for unbound ETS: 1R36, and bound: 1MDM) ! ! ! ! ! "#$%"#&'&%( )*+&%( #*,#-&..*+*'/ 0$1-&..*+*'/ 2345 234627 28 25 26 29 :9 :7 :6 :5 :;; :;; 2345 2346 27 2825 26 29 :9 :7:6:5 6<= 77>9>5 995 758 ?@:-A$B&*+ 2346 27 28 25 26 29 :9 :7 :6 :5 CD+.$0E(E-2345F Chapter 1 – Introduction 	
    27	
   1.4.2 PNT domain in Ets1 The first structure of a PNT domain was solved for that of Ets1, residues 29-138, using NMR spectroscopy (Slupsky et al., 1998). The Ets1 PNT domain is comprised of six α- helices H0-H5 preceded by a consensus MAPK phosphoacceptor site, 38Thr-Pro39. In addition to a core helical bundle (helices H2-H5) shared with the SAM domain, a distinguishing feature of the Ets1 PNT domain is the presence of two additional N- terminal helices H0 and H1 that are also predicted to occur in the PNT domains of Ets2 and its Drosophila ortholog Pnt-P2 by sequence comparisons. Helix H1, but not H0, is also present in GABPα. Both H0 and H1 are absent in most other PNT or SAM domains ((Mackereth, 2003); this thesis). The location of helix H1 in the Ets1 PNT domain was not well defined in the initial structural ensemble of this protein, determined in 1998 by NMR methods using a 500 MHz spectrometer (Slupsky et al., 1998). Furthermore, helix H0 was not initially identified due to rapid amide hydrogen exchange at the experimental pH. The refined structure of the Ets1 PNT domain is described in Chapter 2.  One function of the Ets1 PNT domain was uncovered as Yang et al. identified the interaction of the N-terminal region of Ets1, including the PNT domain, with the cysteine-histidine rich regions CH1 and CH3 of the transcriptional co-activator CBP/p300 (Yang et al., 1998a). Subsequently, Foulds et al. discovered that the affinity of Ets1 for CBP/p300 is enhanced upon phosphorylation by MAPK ERK2, in vitro and in vivo (Foulds et al., 2004). Both the phosphoacceptor region and the core PNT are indispensible for the interaction with CBP/p300. In the report of Seidel et al., the Ets1 PNT domain was shown to be critical for phosphorylating the N-terminal Chapter 1 – Introduction 	
    28	
   phosphoacceptor site Thr38, by providing a docking site for MAPK ERK2 (Seidel and Graves, 2002). This involves residues clustered near helix H5 of the PNT domain (Leu114, Leu116, Phe120). These docking site residues, which are especially conserved among Ets1, Ets2 (Leu148, Leu150, Phe154) and Drosophila Pnt-P2 (Leu228, Ile230, Phe234), differ from other ERK2 docking sites, such as the LXL and FXF motifs (L, Leu; F, Phe; X, any residue) (Seidel and Graves, 2002). Mutation of these residues or deletion of the PNT domain (residues 53-138) leads to significant increases in the Km value of the Ets1 PNT domain for ERK2 (Seidel and Graves, 2002). Although Ets1 was predicted to bind to a novel site on ERK2 due to the unique sequence of the docking site motif in Ets1, several recent reports suggested that the docking site of Ets1 PNT binds one of the known binding interfaces of ERK2, the F- recruitment site (Callaway et al., 2010). Model building suggests that, in order for the phosphoacceptor region to reach the catalytic site of ERK2 while the PNT domain is docked at this site, the adjacent helix H0 must unwind partially (Callaway et al., 2010). This is consistent with the dynamic nature of helix H0, as discussed in chapters 2 and 3. Chapter 1 – Introduction 	
    29	
   1.5 CBP/p300 CBP/p300 is a key transcriptional co-activator, regulating gene expression of many biological pathways in various metazoan species (Kalkhoven, 2004). Among the commonly studied orthologs of CBP, mouse CBP (2441 a.a.) shows the highest similarity to human CBP (2440 a.a.) with 95% sequence identity (McManus and Hendzel, 2001). Genetic alterations or damage of CBP were found to be closely related to Rubinstein-Taybi syndrome (RTS) and possibly RTS-associated cancer in humans (Blobel, 2000). In mouse, CBP is essential, especially during embryonic development (Goodman and Smolik, 2000). Mutations in CBP also lead into several hematological malignancies, such as acute myeloid leukemia and chronic myeloid leukemia, as supported by its hematopoietic-related binding partners (E1A, c-Myb, GATA-1, Ets1, etc.)	
  (Blobel, 2000).  p300 is a paralog of CBP, commonly found in mammals. Although often referred to as CBP/p300 due to their sequence and functional similarities, numerous lines of evidence indicate that they interact with different partners or differentially interact with some of the common partners (Blobel, 2000; Vo and Goodman, 2001). Nevertheless, high sequential similarity (82-98%) for certain regions, such as cysteine-histidine rich regions and the bromodomain, and several common phenotypes upon mutation clearly indicate that the functions of CBP and p300 overlap to a large extent (McManus and Hendzel, 2001).  Chapter 1 – Introduction 	
    30	
   1.5.1 Domain structure of CBP CBP is a large, modular protein composed of at least 8 domains that are conserved among its orthologs (Figure 1.5). These domains carry out two major functions: protein interactions and lysine acetyltransferase catalytic activity. Protein interactions coordinated by ~7 of these domains reflect CBP’s central function as a scaffold to facilitate local clustering of transcription-related proteins or as a bridge between transcription factors and the basal transcription machinery to form a stable enhanceosome. As a member of the mammalian histone acetyltransferase (HAT) family, the CBP acetyltransferase domain acetylates both histone tails and transcription factors.  Non-acetylation-related domains CBP contains three zinc-binding cysteine-histidine rich regions, located in the N- terminal (CH1), central (CH2), and the C-terminal (CH3) portions of the protein. Two of these (CH1, CH3) are responsible for the majority of interactions with transcription factors (Figure 1.5). CH1, also referred as TAZ1 (transcriptional adaptor zinc-binding domain), and the C-terminal part of the CH3, TAZ2, are highly related structurally. Both are comprised of four α-helices with three zinc ions held in each loop (H1/H2, H2/H3, H3/H4) by three cysteines and one histidine in each domain. A BLAST search comparison of the TAZ1 and TAZ2 domains shows only 35% sequence identity, including the key residues for coordinating zinc-binding. Despite their striking structural resemblances, the fourth helices are packed differently against helix 1 and 3, being parallel and antiparallel to Chapter 1 – Introduction 	
    31	
      Figure 1.5 Domain structure of murine CBP CBP, as a transcriptional co-activator, exploits its multidomain structure to serve both as a scaffold (or bridge) and an acetyltransferase (HAT) for transcriptional regulation. Coordinates are denoted for murine CBP domains. Transcription factor and nuclear receptor binding domains are shown in orange and red. Acetylation-related domains are shown in green. See text for the details.  !"" #""" #!"" $""" $%%#" &'() *(+ ,' -.) /.$ .01 /.2 33 103$ 4()/.# # #"# 2%" %25 !67 777 #76" #6!" $"!6 $#72 #$26 #2"5 #8""#8!" #87% #6!" #2$% #8"% #"%8 ##58 #$26 #!72 9:;<=>?/,- 103# 2%" %25 Chapter 1 – Introduction 	
    32	
   helix 1 in TAZ1 and TAZ2, respectively, thereby allowing the two domains to recognize distinct sets of transcription factors. In addition, differences are also observed in the length of the loop regions (H1/H2) between the two domains (De Guzman et al., 2005). Curiously, Ets1 is somewhat unusual as it binds both the TAZ1 and TAZ2 domains. In addition to TAZ2, CH3 bears another small zinc-finger domain called the ZZ domain. Although its NMR-derived structure is also available (Legge et al., 2004), the function of the ZZ domain remains undefined.  KIX, located adjacent to the CH1 domain, also interacts with a number of different transcription factors, including CREB (Radhakrishnan et al., 1997). Unlike the TAZ domains, KIX can form a ternary complex by simultaneously binding two different transcription factors through its two different binding interfaces. This is exemplified by the cooperative interactions of KIX with both the hematopoietic transcription factor, c- Myb, and the mixed lineage leukemia transcription factor, MLL, to synergistically enhance the subsequent transcriptional activation (Goto et al., 2002).  CBP also contains several copies of a nuclear receptor interacting motif, called the NR box (L, leucine; X, any residue). The nuclear receptor (NR) box contains a short amino acid sequence termed the LXXLL motif (L, leucine; X, any residue), which is commonly found in co-activators (SRC-1, RIP140, CBP/p300) for mediating interactions with nuclear receptors (Heery et al., 1997; Plevin et al., 2005). It is also noteworthy that this motif is typically found in a helical conformation. CBP contains three copies of the LXXLL motif, spread widely across its sequence: N-terminal (69LSELL73), central (357LVLLL361), and C-terminal (2068LQDLL2072) (Heery et al., 2001). Previous studies Chapter 1 – Introduction 	
    33	
   have shown that the N-terminal NR box, also known as the NRID (nuclear receptor interacting domain), strongly interacts with the ligand binding domain (LBD) of various nuclear receptors, such as RAR (retinoic acid receptor), RXR (retinoid X receptor), TR (thyroid hormone receptor), and PPAR (peroxisome proliferation-activated receptor). This reflects the role of CBP as an integrator in signal transduction pathways within nucleus. In contrast, the central and C-terminal NR boxes have very low affinity for nuclear receptors (Chakravarti et al., 1996). Interestingly, the central NR box is located within the long helix 1 of TAZ1 which is heavily involved in binding many transcription factors as described below. The C-terminal NR box, often called the SID (SRC-1 interacting domain), is located in the glutamine rich region and interacts with other co- activators, such as steroid receptor coactivator 1 (SRC-1) and activator for thyroid hormone and retinoid receptors (ACTR) (Sheppard et al., 2001). It has also been shown that CBP interacts with basal transcription factors, TATA-binding protein, TFIIB, and RNA helicase A, through its N- and C-terminal regions (Figure 1.1) (Kwok et al., 1994; Dallas et al., 1997).  Acetylation-related domains The lysine acetyltransferase domain lies in the central region of CBP, near the bromodomain and PHD finger domain (Figure 1.5). Although it is a member of the mammalian histone acetyltransferase (HAT) family, CBP/p300 is categorized separately from other HAT families, such as the GNAT superfamily or the MYST family, especially due to its ubiquitous expression and its involvement in diverse cellular activities (Sterner and Berger, 2000). Unlike other HAT proteins, CBP/p300 shows specificity for all the histones (H2A, H2B, H3, H4) (McManus and Hendzel, 2001). Furthermore, it is more Chapter 1 – Introduction 	
    34	
   often referred to as a HAT domain, yet it is also capable of acetylating lysines on non- histone proteins including p53, E1A, TFIIE/F, and some high-mobility-group (HMG) chromatin-associated proteins. Hence it is also termed a factor acetyltransferase domain (FAT) or a lysine acetyltransferase domain (LAT or KAT; K = lysine) (McManus and Hendzel, 2001).  The CH2 region bears the centrally located zinc-binding domain with two zinc ions being held by 4 Cys – 1 His – 3 Cys, and thereby also known as a PHD finger. Especially for CBP, CH2 is considered to be an integral part of the enzymatic core of the CBP acetyltransferase domain as previous mutational studies indicate the essential role of CH2 in the HAT activity (Kalkhoven, 2004). The PHD finger is frequently found in nuclear proteins related to chromatin remodeling and tends to function with other chromatin-related domains, such as the bromodomain, chromodomain, acetyltransferase domain, or methyltransferase domain (Bienz, 2006). Its main functions were identified as histone binding, protein-protein interactions, and possibly DNA/RNA binding. The recognition and binding of methyl-lysine is critical in transcription regulation as the PHD domain in transcriptional co-activators like CBP can first, bind the methyl-lysine, then acetylate the histone tails for transcriptional activation. For both CBP and p300, the PHD finger is located adjacent to the bromodomain. Interestingly, the presence of the PHD finger is only essential for HAT activity in CBP, but not for that of p300, indicative of possible structural differences in the HATs of these two co-activators (Bordoli et al., 2001).  Chapter 1 – Introduction 	
    35	
   The acetyl-lysine binding bromodomain (residues ~1082-1197), located C-terminal of the HAT domain in CBP, is often found in transcription-related proteins, such as chromatin remodeler, acetyltransferase domain-containing proteins, and transcription factors (Marmorstein and Berger, 2001). The bromodomain is comprised of four helices. Two of the loop regions create a binding site for acetylated lysines. In relation to this biochemical property, biological functions of bromodomain are found in several categories including chromatin acetylation by HATs, nucleosome assembly/remodeling, chromosome organization, and interaction with acetylated non-histone substrates. Chapter 1 – Introduction 	
    36	
   1.6 Thesis overview This thesis focuses on investigating the structural and dynamic bases for the regulation of Ets1. Specifically, I focused on the interaction between the PNT domain and its binding partner, the TAZ1 domain of CBP. First, in chapter 2, the effects of phosphorylation at Thr38 as well as the previously unknown second phosphoacceptor site, Ser41, on the PNT domain were studied. NMR spectroscopy and several other biochemical methods were employed to monitor the changes in structure and dynamics of Ets1 PNT. In the course of this study, a new pulse sequence for assigning the NMR signals of phosphoserine and phosphothreonine was developed and published (section 1.7). However, the latter work is not included in this thesis as it was mainly carried out by Dr. McIntosh during his sabbatical in Grenoble. Second, in chapter 3, the phosphorylation-dependent interaction of Ets1 PNT domain with the CBP TAZ1 domain was investigated using NMR-based structural approaches to understand the mechanism of the increased affinity upon phosphorylation. In chapter 4, the serendipitous discovery of an interaction between the N-terminal region of CBP and TAZ1 is described. This interaction suggested a possible auto-inhibitory mechanism of TAZ1 for binding many other transcription factors. Using NMR spectroscopy, the minimal N-terminal region was delineated to residues 28-82. The key components of chapters 2 and 3 are condensed and rewritten into the recently published paper, entitled "Ras signaling requires dynamic properties of Ets1 for phosphorylation-enhanced binding to co-activator CBP" (PNAS, vol. 107, no. 22, pg. 10026-10031, 2010), which also includes the further in vivo and in vitro studies carried out by our collaborators, Dr. Graves lab in Utah. However, this thesis primarily focuses on the structural characterization of these two domains carried out at UBC, including 14 out of 16 figures Chapter 1 – Introduction 	
    37	
   (all but figures 2.3 and 2.8) in chapter 2 and 12 out of 13 figures (all but figure 3.1 and tables 3.1 and 3.2) in chapter 3.  Chapter 1 – Introduction 	
    38	
   1.7 Other publications This section lists the additional publications that I was involved in over the course of my PhD studies, which are not directly related to the focus of this thesis. During the early years of my graduate studies, my research was mostly focused on understanding the autoinhibitory mechanism of the DNA-binding ETS domain of Ets1 (Lee et al., 2005; Pufall et al., 2005; Lee et al., 2008). Then, for my MSc thesis, I identified and solved the structure of the OST domain in GABPα using NMR spectroscopy (Kang et al., 2008). Unlike Ets1/2, GABPα interacts with CBP TAZ1 through this newly-identified OST domain instead of its PNT domain. Lastly, I was also involved in developing methods for the detection and assignment of signals from phosphoserine and phosphothreonine using NMR spectroscopy (McIntosh et al., 2009).  Variable control of Ets-1 DNA binding by multiple phosphates in an unstructured region. Pufall, M.A., Lee, G.M., Nelson, M.L., Kang, H.-S., Velyvis, A., Kay, L.E., McIntosh, L.P., and Graves, B.J. Science, 309: 142-145 (2005)  The structural and dynamic basis of Ets-1 DNA binding autoinhibition. Lee, G.M., Donaldson, L.W., Pufall, M.A., Kang, H.-S., Pot, I., Graves, B.J., and McIntosh, L.P. Journal of Biological Chemistry, 280: 7088-7099 (2005)  The affinity of Ets-1 for DNA is modulated by phosphorylation through transient interactions of an unstructured region. Lee, G.M., Pufall, M.A., Meeker, C.A, Kang, H.S., Graves, B.J., and McIntosh, L.P. Journal of Molecular Biology, 382: 1014-1030 (2008) Chapter 1 – Introduction 	
    39	
    Identification and structural characterization of a CBP/p300-binding domain from the ETS family transcription factor GABPα. Kang, H.-S., Nelson, M.L. Mackereth, C.D., Schärpf, M., Graves B.J., and McIntosh, L.P. Journal of Molecular Biology, 377: 636- 646 (2008)  Detection and assignment of phosphoserine and phosphothreonine residues by 13C-31P spin-echo difference NMR spectroscopy. McIntosh, L.P., Kang, H.-S., Okon, M., Nelson, M.L., Graves, B.J., and Brutscher, B. Journal of Biomolecular NMR, 43: 31- 37 (2009)    Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   40	
   Chapter 2 - Phosphorylation and the Ets1 PNT domain  To understand the effects of phosphorylation on the structure and dynamics of the Ets1 PNT domain, a battery of NMR techniques was used in combination with complementary biophysical and molecular biological methods. First, two phosphoacceptor residues (Thr38 and Ser41) were identified and characterized both in vitro and in vivo. NMR analyses revealed that these residues lie within the unstructured N-terminal region of Ets1, immediately adjacent to the structured PNT domain. Phosphorylation shifted a conformational equilibrium of the PNT domain, displacing the dynamic helix H0 from the more stable core bundle (H1- H5). These data lead to a model in which the dynamic helical elements of Ets1, appended to a more conserved structural core, constitute a phospho-switch that directs Ras/MAPK signaling to downstream changes in gene expression. These changes are mediated through phosphorylation-dependent binding to the co-activator CBP, as presented in chapter 3. 	
   	
    Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   41	
   2.1 Introduction Post-translational modifications greatly expand the functional proteome, thus facilitating responsiveness of biological systems to developmental and environmental cues (Walsh et al., 2005). Proteins critical to the intra- and extracellular signaling response and control of gene expression display a full repertoire of post-translational modifications, including phosphorylation, acetylation, methylation, hydroxylation, and sumoylation. Although these modifications have been shown to influence DNA binding, nuclear trafficking, and association with co-activators and co-repressors, there is often only a superficial understanding of the molecular mechanisms underlying these effects. Post- translational modifications occur frequently outside of conserved structured regions, thus challenging the use of established biochemical approaches for defining macromolecular interfaces. In this study, we investigated the mechanism by which phosphorylation enhances the binding of the transcription factor Ets1 to the transcriptional co-activator CBP (or lysine acetyltransferase KAT3a). Based on earlier findings by Dr. Graves (Foulds et al., 2004), we anticipated that Ets1 would provide a model biochemical system to explore the integration of dynamic unstructured and structured regions in a regulated macromolecular interface.  The Ras/MAP kinase signaling pathway, a key player in regulation of cell proliferation, impacts gene expression through the ETS family of transcription factors	
   (Tootle and Rebay, 2005). Well-studied examples include the serum-response-element, SRE, of the c-Fos promoter, at which the ETS protein, Elk1, plays a central role (Li et al., 2003). Other Ras-responsive elements are characterized by tandem ETS or ETS/AP1 composite sites associated with Ets1 and Ets2 (Bassuk and Leiden, 1995; Stacey et al., Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   42	
   1995; Yang et al., 1996). In Drosophila, a genetic link between the ETS family and Ras/MAPK signaling implicates Pnt-P2, the ortholog of vertebrate Ets1 and Ets2, and Yan, the apparent ortholog of vertebrate Tel (O'Neill et al., 1994). Phosphorylation affects nuclear export (Yan) (Tootle et al., 2003) and CBP/p300 recruitment (Elk1, Ets1, and Ets2) (Li et al., 2003; Foulds et al., 2004; Qiao et al., 2006). Key observations from Dr. Graves' previous analysis of CBP recruitment by Ets1 include augmentation of transactivation by addition of the co-activator in cell-based transcription assays, as well as detection of phosphorylation-enhanced Ets1-CBP complexes in vitro and in vivo (Foulds et al., 2004). However, the mechanism of this enhancement was not elucidated prior to the studies presented in this chapter.  Ets1 provides an excellent system to investigate phosphorylation-dependent transcriptional regulation due to the extensive biophysical and structural data that are available. The previously characterized MAP kinase ERK2 phosphoacceptor, Thr38, lies within a consensus motif (Ser/Thr-Pro) (Clark-Lewis et al., 1991) in the unstructured N-terminal region of Ets1, preceding the folded PNT domain (Slupsky et al., 1998). Both these unstructured and structured regions are critical for CBP binding and enhanced transcriptional regulation (Foulds et al., 2004). The helical bundle PNT domain, which is present in a subset of ETS proteins (Figure 2.1), displays close structural similarity to the SAM domain, a widespread protein-protein and protein-RNA interaction module (Slupsky et al., 1998; Kim et al., 2001; Mackereth et al., 2004; Qiao et al., 2004; Goroncy et al., 2006). However, Ets1, Ets2, GABPα, SPDEF and, by sequence   Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   43	
      Figure 2.1 Diversity of the N-terminal region of Ets family PNT domain ETS family PNT domain structural elements and sequence conservation. (A) Dendrogram of selected family members based on their PNT domain sequences using ClustalW (Larkin et al., 2007). (B) Helical elements of the PNT domain (color coded as in figure 2.4). ERG (1sxe.pdb), Tel (1lky.pdb), Yan (1sv4.pdb), and Fli1 (2ytu.pdb) share a core helical bundle (H2-H5) also common to SAM domains, whereas GABPα (1sxd.pdb) and SPDEF (2dkx.pdb) have an N-terminal helix H1. Those phosphorylated by MAPK have additional observed (Ets1, Ets2) or predicted (Pnt-P2) helices H0 and H1. (C) Portions of murine Ets1, Ets2, GABPα, Tel, and Fli1, human SPDEF and ERG, and Drosophila Pnt-P2 and Yan sequences aligned by ClustalW. Residues with similar physicochemical properties as those in Ets1, Ets2 or Pnt-P2 are shaded grey, and Ets1 phosphoacceptors, Thr38 and Ser41, are marked by asterisks. Numbers refer to amino acid positions in the full-length proteins and helical boundaries in Ets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hapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   44	
   comparison, Drosophila Pnt-P2, share an additional PNT-specific N-terminal helix H1 (Mackereth et al., 2004). This helix is preceded, with conserved spacing, by an ERK2 consensus site only in Ets1, Ets2, and Pnt-P2. The interplay between conserved core structured domains and adjacent unstructured phosphorylated regions portends an interesting mechanistic model for signaling-regulated CBP binding.  In this chapter, the effects of phosphorylation on the structure and dynamics of an N- terminal fragment of Ets1 containing the PNT domain were explored using various NMR techniques and biophysical methods. In addition to Thr38, a second non-consensus ERK2 phosphoacceptor, Ser41, was identified and verified for its in vivo relevance. The structural ensembles of apo- and 2P-Ets129-138 were determined using NMR spectroscopy, and examined for the changes in their structural and dynamic properties. Most importantly, a helix (H0), appended to the core helical bundle (H1-H5) of the PNT domain, was discovered. This helix was found to be dynamic by chemical shift, 15N relaxation, and amide hydrogen exchange (HX) measurements. Phosphorylation of Thr38 and Ser41 shifted a conformational equilibrium, displacing helix H0 from the core PNT domain. We hypothesize that this constitutes a phospho-switch, which enhances binding of Ets1 to CBP to direct downstream changes in gene expression. Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   45	
   2.2 Results 2.2.1 Identification of the ERK2 phosphoacceptors in Ets1 To investigate the mechanism by which phosphorylation regulates Ets1, we developed an in vitro system using ERK2 to covalently modify purified PNT domain-containing fragments of this protein. Unexpectedly, Ets11-138 was phosphorylated at 1, 2, or 3 sites, depending upon the experimental conditions. Phosphorylation was readily detected by an alteration in SDS-PAGE mobility, by mass spectrometry (80 Da per phosphate), and, even more unambiguously, by NMR spectroscopy. This section describes the confirmation of the previously known MAPK ERK2 phosphoacceptor site, Thr38, and the identification of an additional non-consensus phosphoacceptor, Ser41. An in vivo mutagenesis study was also carried out to test the biological relevance of these two modifications.  Mass spectrometry Three phosphorylation states (1P, 2P, 3P) were observed for both Ets11-138 and Ets129- 138, when treated with the MAP kinase, ERK2, for different reaction times. In the presence of relatively high concentration of ATP (~2 mM), the predominant products were 1P after ~ 1 hour, 2P after 1 day, and 3 P after 3 days. The phosphorylation states were identified using MALDI-TOF mass spectrometry (not shown).	
  	
   The phosphoacceptor sites were identified in 2P-Ets11-138 via tandem mass spectrometry sequencing (Nelson, 2007). In this experiment, 2P-Ets11-138 was digested with chymotrypsin, and then was fractionated using iron metal affinity chromatography (IMAC) to isolate the phosphorylated fragments. After further purification using a C18 Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   46	
   reverse-phase HPLC column, the resulting peptides were sequenced by tandem mass spectrometry. Thr38 and Ser41 were identified as the sites of phosphorylation. Thr38 corresponds to a consensus site for ERK2 (Thr/Ser-Pro), whereas Ser41 is followed by a lysine residue.	
  	
   The identification of the third phosphorylation site was also attempted using mass spectrometry. Ets129-138 was treated with ERK2 for more than 3 days. An increase of its mass by the equivalent of three phosphates (~240 Da) was confirmed using MALDI- TOF mass spectrometry. The same sample was digested with trypsin, and then was purified using titanium dioxide beads to isolate the phosphorylated fragments (Larsen et al., 2005). The sequences of the resulting peptides were analyzed by tandem mass spectrometry (MS-MS). All the species identified contained Thr38 and either Ser40 or Ser41 as the phosphoacceptor sites. Phosphorylation of Ser40 and Ser41 could not be distinguished, possibly due to the lack of fragmenting between these residues. No fragment was observed with all three phosphates present simultaneously, possibly due to dephosphorylation during purification or the mass spectrometric analysis. Although we cannot exclude other sites of modification, these data indicate that Ser40 is a third site of ERK2 phosphorylation in Ets129-138. However, this non-consensus site is only modified in vitro after prolonged exposure to the kinase and thus unlikely to have an in vivo biological function.	
  	
   NMR spectroscopy To confirm the phosphoacceptor sites in the N-terminal region of Ets1, the 15N-HSQC spectra of Ets129-138 and 2P-Ets129-138 were assigned using standard 1H/13C/15N-triple Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   47	
   resonance experiments (Kang, 2005). The anticipated phosphoacceptors (Thr38, Ser41), as well as the residues near these positions, showed significant changes in amide chemical shifts upon phosphorylation (Figure 2.2A). More importantly, the distinct chemical shift changes in the 13Cβ resonances of Thr38 and Ser41 upon phosphorylation of Ets129-138 confirmed that these residues were indeed modified (Figure 2.2B). Subsequently, these phosphoacceptors were also identified unambiguously by the direct detection of sidechain and amide signals of pThr38 and pSer41 using newly-developed 1H/31P/13C/15N NMR pulse sequences (McIntosh et al., 2009). Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   48	
     	
  	
     Figure 2.2 Identification of Thr38 and Ser41 as phospho-acceptors by chemical shift changes (A) Amide chemical shift differences between Ets129-138 and 2P-Ets129-138 were observed in the phosphoacceptor region, as well as within the structurally adjacent helices H2 and H5. A histogram of these chemical shift perturbations, calculated as {(10Δδ1HN)2 + (Δδ15N)2}1/2 is shown. The sensitivity of amide chemical shifts to structural and electric field effects precluded a direct identification of Thr38 and Ser41 as the phosphoacceptors. (B) In contrast, large changes in the 13Cβ chemical shifts (blue) and to a lesser extent, 13Cα (black) of 2P-Ets129-138 in comparison to those of random coil polypeptide clearly identifiesThr38 and Ser41 as the sites of modification (*). !"#"$"%" &""'"(")" &&" &*" &%" &$" +",# " ",# *," +&," %,# 6 -. /0 01 2 !"#6$_ !"#6$` !"#! !"$! !"%& ' ! # 6 ( ) *+, ,- . /012340 !"5! 6 #! !"#"$"%" &""'"(")" &&" &*" &%" &$" 3" 3& 3* 3*4 3% 3#3$ 5% ( 6$ & +7. +8. 9 9 Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   49	
   In vivo activity The in vivo relevance of phosphorylation at Thr38 and Ser41 was tested in a cell-based assay for Ras/MAPK enhancement ("superactivation") of Ets1 activity. Ets1-dependent transcription of a reporter gene was boosted 2 to 3 fold by the co-expression of a constitutively active form of MEK1 to mimic Ras/MAPK signaling (Figure 2.3). Mutation of Thr38 or Ser41 to alanine reduced superactivation to 1.15 or 1.42 fold, respectively. Mutation of both residues reduced the level of Ets1-dependent activation to the control level observed in the absence of Ets1. In vitro controls confirmed that the remaining site could still be phosphorylated in either mutant context (Nelson, 2007). However, only phosphorylated Thr38 could be confirmed in vivo with phospho-specific antibodies (Figure 2.3), and thus the low activation of the T38A mutant could also reflect indirect changes in the phosphorylation state of Ser41. Regardless, the full effect of the signaling required the two phosphoacceptors.  To further test the in vivo relevance of the Thr38 and Ser41 sites, aspartate or glutamate substitutions were used as phosphomimetics. As previously reported, the T38E and T38D mutations did not replace phosphorylation	
   (Yang et al., 1996; Seidel and Graves, 2002). In contrast, Ets1 with the S41E, but not S41D, substitution activated transcription ~2 fold in the presence of phosphorylated Thr38, which was again detected with phospho-specific antibodies (Nelson, 2007). Test of the double substitution, T38E/S41E, also indicated that S41E was an effective phospho-mimetic. In sum, these data demonstrate that Ser41 phosphorylation functions in a transcriptional Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   50	
      Figure 2.3 In vivo studies of phosphorylation-dependent transactivation of Ets1 ERK2 phosphoacceptors Thr38 and Ser41 contribute to Ets1 transactivation. NIH 3T3 cells were transfected with a luciferase reporter with two ETS binding sites, an expression vector for wild type or the indicated mutant FLAG-tagged Ets1, and either an expression vector for constitutively active CA-MEK1 (+) or an empty vector control (-). An expression vector for Renilla luciferase provided an internal control to calculate relative luciferase activity (RLA). (A) (Upper) Plot of RLA for a representative experiment with mean and standard deviation for triplicate transfections. (Lower) Ets1 species were immunoprecipitated from transfected cells with FLAG-specific antibodies, and then probed by immunoblotting with pThr38-Ets1 or Ets1 specific antibodies (dash, 50 kDa marker). (B) Superactivation (RLAWT or mut / RLAempty vector) in the presence of CA- MEK1 determined by mean of at least two independent experiments ± the standard error of the mean. Negative charges introduced at positions 38 and 41 are indicated (CO2-, mutation; PO4-2, phosphorylation status of Thr38 confirmed in vivo by a phospho- specific antibody; PO4-2, presumed phosphorylation status of Ser41). MEK1-dependent activation, which was observed in the absence of Ets1 and thus set as the control level, is possibly due to ERK2-dependent effects on other elements of the transcription machinery (e.g. CBP/p300, (Gusterson et al., 2002)). (Data provided by M. Nelson and A. Blaszczak.) _ ! "#$%&'() _ &'() * +, -" "#$, ./), "#$,0 ./), 1,%2&3) 4 4 4 4% % % %4 % % !"#$%&'($ "#$ ./) )& &'($ !"#$ &'(* !"#* &'(*%!"#* &'(+ !"#+ &'(+%!"#+ ,-,. / # , ' " . 0 #-#. #-'1 #-", #-/2 #-/. #-'. #-#" #-/2 #-11 ! /-#/ ! /-/( ! /-,# ! /-/1 ! /-/0 ! /-/# ! /-#. ! /-/" ! /-/# ! /-," 5678 .&2 &'() &'() (9!6:7;'<=7'<>8 ?,@ ?A@ !"#$%34,5 64"5, 34,5 64"5, 34,5 64"5, 34,5 64"5, 34,5 !"#$% 34,5 !"#$% 34,5 !"#$% 34,5 Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   51	
   context and works in concert with Thr38 to mediate the response of Ets1 to Ras/MAPK signaling. 	
   2.2.2 Structure determination of apo- and 2P-Ets129-138 by NMR To understand the effects of phosphorylation, NMR spectroscopy was employed extensively. First, the structure of the apo-Ets129-138 was refined to better define the region of the PNT domain close to the phosphoacceptor sites. The structural and dynamic properties of the apo-Ets129-138 were then examined through 15N-relaxation, amide hydrogen exchange, and secondary structure propensity analyses. Next, the three dimensional structure of the 2P-Ets129-138 was determined to unveil the conformational changes of the newly identified dynamic helix H0 upon phosphorylation. Finally, to confirm the conformational equilibrium of H0, structural and dynamics properties of the 2P-Ets129-138 were compared to those of a series of Ets1 mutants and deletion fragments using both NMR and other complementary methods.	
  	
   Refined Ets1 PNT domain structure couples dynamic phosphoacceptors and helix H0 to a core helical bundle The initial NMR spectroscopic analysis of Ets129-138 (Slupsky et al., 1998) established that the PNT domain folds as a core bundle of four α-helices (H2-H5) with an additional N-terminal helix H1 (Figure 2.1B). However, the position of helix H1 in the structural ensemble of Ets129-138 was not well-defined using limited manually assigned restraints from data recorded with a 500 MHz NMR spectrometer. Thus, the structural ensemble Ets129-138 was refined to a higher resolution structure via a semi-automated ARIA/CNS protocol (Rieping et al., 2007) using extensive dihedral angle, distance and orientational Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   52	
   restraints derived from data measured for a sample of slightly lower pH (see below) with a 600 MHz instrument (Figure 2.4A, Table 2.1). The refined model shows the predominantly polar helix H1 (residues 54-62) well positioned along helices H2 and H5 via packing of Phe53 and Phe56 at its N-terminal end with Trp80, His128, and Ile131. The position of H1 resembles that also found in the closely-related GABPα PNT domain (Mackereth et al., 2004).  Most importantly, the refinement of the Ets129-138 structure also led to the discovery of a new Ets1 PNT domain-specific helix, designated H0 (residues 42-52). This amphipathic helix packs against helix H5 and the end of helix H2 via hydrophobic interactions involving Met44, Met45 and Leu49 with Phe88, Phe120 and Ile124, as well as potential salt bridges between Lys42-Asp123 and Lys50-Glu127 (Figure 2.4A). The newly recognized helix H0, which is also present in the closely related Ets2 (Figure 2.5), is preceded by an unstructured region (residues 29-41) bearing Thr38 and Ser41. 	
   Although clearly defined in the refined structural ensemble of Ets129-138, helix H0 is marginally stable and conformationally dynamic by several criteria. Using the algorithm SSP to predict secondary structure from main chain chemical shifts (Marsh et al., 2006), H0 exhibits helical scores less than those of the core PNT domain (Figure 2.6). Similarly, 15N-relaxation measurements revealed that residues in H0 have 1H{15N}-NOE values ranging from 0.24 to 0.53 and fit model free order parameters, S2, ranging from 0.74 to 0.86 (Figure 2.7). These relaxation parameters are indicative of enhanced mobility on the nsec-psec timescale relative to the well-ordered helices H1-H5 (average 1H{15N}-NOE 0.63 ± 0.03 and S2 0.90 ± 0.04). Furthermore, the pattern of SSP and S2 Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   53	
            Figure 2.4 Solution structures of Ets129-138 and 2P-Ets129-138 Superimposed structural ensembles from NMR analyses of (A) Ets129-138 and (C) 2P- Ets129-138 showing the helical bundle PNT domain preceded by a flexible region containing the ERK2 phosphoacceptors Thr38 and Ser41. The α-helices (H0, residues 42-52; H1, 54-62; H2, 74-87; H3, 102-107; H4, 109-116; H5, 119-134) and a 310-helix (H2', 95-98) are identified by rainbow coloring, and Thr38 and Ser41 phosphates as balls in cyan and blue, respectively. Disordered residues 29-37 omitted for clarity. (B) Expanded view of the interface between the dynamic helix H0 and helices H2 and H5 of the core PNT domain with colors coordinated to helix identity. Residues examined by mutation (Table 2.2) are underlined. Potential salt bridges (K42-D123, K50-E127) are shown as dashed lines. Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   54	
   !"#$ %"#" &'$ !'' ("#) *'+ ,"#' -.' /0" !)1 /)0 /'2 (02 -)# ,$ ," ,# ,#3 ,. ,0 ,) ,$ ," ,# ,#3 ,. ,0 ,) ,$ ," ,# ,#3 ,. ,0 ,) -.' /0" ,$ ," ,# ,#3 ,. ,0 ,) -.' /0" Û Û 4567*89"#2:".' 4;67#<:*89"#2:".' =-.' =/0" =-.'=/0" !"#$ >"#. *"#+ ?"." ?"#0 @0# A0) @)$ (02 -.' /0" !). &'$ &"#1 /)0 >""2 4B67?C8DEFGHD7IF7,$J7,#J7GCK7,)7 LC7*89"#2:".' Û Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   55	
   	
          Figure 2.5 Ets2 PNT domain closely resembles that of Ets1 Similar to Ets1, the Ets2 PNT domain is composed of helices H0-H5, preceded by unstructured phosphoacceptors Thr72 and potentially Ser75. (A) 15N-HSQC spectrum of Ets263-172 in 20 mM phosphate pH 7.0, 20 mM KCl, 0.1 mM EDTA, and 2 mM DTT, 25 °C. The peaks were assigned by standard heteronuclear correlation experiments (not labeled for clarity; aliased peaks are in red). (B) SSP scores derived from 15N, 1HN, 13Cα, and 13Cβ chemical shifts confirm that Ets269-172 has the same helical secondary structure as Ets129-138. The helices forming the PNT domain have scores approaching 1, whereas the unstructured N-terminal region has scores near 0. Missing data correspond to prolines or amides lacking unambiguous spectral assignments. (C) Sequence alignment of the corresponding PNT domain-containing fragments of Ets1 and Ets2 with identical residues highlighted in green and phosphoacceptors indicated with an asterisk. Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   56	
    	
  	
  	
  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hapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   57	
   	
     Figure 2.6 Chemical shift-based secondary structure analysis of Ets129-138 and 2P- Ets129-138 Both secondary structure propensity (SSP) and Δδ(13Cα-13Cβ) analyses predicted α- helical confomation for helices H1-H5 of the core Ets1 PNT domain. In contrast, helix H0 exhibits shifts indicative of dynamic behaviour such as fraying. (A) Secondary structure propensity (SSP) scores, calculated from mainchain 1H, 13C, and 15N chemical shifts for Ets129-138 (red) and 2P-Ets129-138 (blue) increase from 0 to +1 with increasing α-helical propensity (Marsh et al., 2006). (B) Difference of the SSP score between Ets129-138 and 2P-Ets129-138 indicates that secondary structures remain unchanged in most part. However, three regions with noticeable small changes (phosphoacceptor region, H0/H1 loop, N-term of H2) coincide with the regions observed in the structural changes between Ets129-138 and 2P-Ets129-138. (C) Per-residue Δδ(13Cα-13Cβ) values  in comparison to those of corresponding random coil chemical shifts also confirm the secondary structure of the Ets1 fragment. Note that helical residues have positive values, and that data are not smoothed by nearest-neighbor averaging as done in the SSP algorithm.  !" !# !$ !$% !& !'!( )& * +( # ,"'"("&" #""-"*"." ##" #$" #&" ++ / $/0123#$-0#&* 123#$-0#&* "4$ "4, #4" 0"4$ 0"4, 5637896 :;< ,"'"("&" #""-"*"." ##" #$" #&" 6 ++ / :=< 0"4( 0"4* "4( "4* " :>< 6 b: #& > _ 0# & > ` <?: @@ A < ,"'"("&" #""-"*"." ##" #$" #&" "4$ "4, #4" 0"4$ 0"4, Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   58	
        Figure 2.7 15N – relaxation and amide hydrogen exchange analyses of Ets129-138 and 2P-Ets129-138 demonstrate that helix H0 is dynamic The phosphoacceptors and helix H0 are dynamic in both Ets129-138 (red) and 2P-Ets129- 138 (blue), as revealed by amide 15N T1 (A), T2 (B) and heteronuclear 1H{15N}-NOE (C) relaxation data, recorded at 30 oC with a 600 MHz NMR spectrometer. Also shown are the generalized order parameters, S2 (D), and conformational exchange broadening terms, Rex (E), derived from an anisotropic model free analysis of these data using TENSOR2 (Dosset et al., 2000). To extract these parameters describing internal mobility, low energy members of the structural ensembles of Ets129-138 and 2P-Ets129-138 were first used to determine their diffusion tensors for global rotation. These tensors corresponded to isotropic tumbling times of 7.05 ± 0.04 nsec and 6.27 ± 0.02 nsec for the two species, respectively. Note that elevated T2 lifetimes and reduced NOE and hence S2 values reflect increased mobility on a nsec-psec timescale, whereas conformational exchange on a msec-µsec timescale leads to anomalously short T2 lifetimes. (F) Protection factors (PF) derived from the ratio of predicted versus observed rates of amide HX. (G) Relative protection factors for 2P-Ets129-138 versus Ets129-138. Missing data corresponds to prolines, residues before prolines, residues with unresolved 1HN-15N signals, or in the case of (D) and (E), residues which do not show measurable HX (pH 6.0, 30 oC) by the CLEANEX method and thus have protection factors > 10. The dynamic nature of helix H0 is not a trivial result of deleting residues 1- 28, as the NMR spectra of corresponding residues in Ets11-138 overlap those of Ets129- 138 closely, confirming identical features (Macauley et al., 2006). Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   59	
  	
   !"#"$"%" &""'"(")" &&" &*" &%" !"#"$"%" &""'#'"(")" &&" &*" &%" !"#"$"%" &""'"(")" &&" &*" &%" !"#"$"%" &""'"(")" &&" &*" &%" "+" "+* "+$ "+! "+( &+" "+" "+* "+$ "+! "+( &+" "+" "+& "+* "+% "+$ "+# "+" "+! &+" ,&+* ,"+$ ,"+( -& ./0 12 3 -* ./0 12 3 & 4 5&# 67 , 6 89 :* 4" 4& 4* 4*; 4% 4#4$ < < < < < < < < < < < < < < < < < < < <= = = = = = = = -% ( :$ & *>,9?0&*',&%( 9?0&*',&%( !"#"$"%" &""'"(")" &&" &*" &%" @1 A. /0 12 ,& 3 "+" *+" $+" !+" < < < < <= = !"#$%&" '(' )(' *(' +(' ,(- )*(' ./ 0'-'1'+' )''2'3',' ))' )*' )+' '(' *(' 1(' !" 45 6$7 "8 ./ 0'-'1'+' )''2'3',' ))' )*' )+' 0(' 9:; 9<; 9=; 9>; 9?; 9/; 9@; Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   60	
   values is suggestive of fraying towards the N-terminal end of H0, and confirms that the phosphoacceptors Thr38 and Ser41 are disordered. In addition, residues in H0 undergo facile amide HX, exhibiting protection factors of only ~2 relative to a corresponding random coil polypeptide (Figure 2.7F). Indeed, this behavior, indicative of extensive local unfolding, precluded the initial detection of helix H0 using NMR data recorded with Ets129-138 at pH 6.5 (Slupsky et al., 1998) rather than pH 6.2. These measurements are also consistent with previous data that demonstrated residues 1-50 are not critical for the structural integrity of helices H1-H5 (Slupsky et al., 1998). Conversely, Ets11-52 exhibits the NMR spectra of a predominantly unstructured polypeptide (Macauley et al., 2006), indicating that helix H0 forms only in the context of the core PNT domain.  In an independent approach to measuring structural flexibility, the sensitivity of Ets11-138 to partial trypsin proteolysis was interrogated. Cleavage at Lys42, Lys50, and Arg62 within helices H0 and H1 revealed at least transient accessibility of these residues to the protease. In contrast, lysines and arginines within the core PNT domain were resistant to cleavage under the same experimental conditions (Figure 2.8B). 	
   Phosphorylation shifts a conformational equilibrium involving helix H0 To investigate how phosphorylation enhances CBP binding (chapter 3), potential conformational changes in Ets1 were monitored upon ERK2 modification. Initially, circular dichroism (CD) spectroscopy was used to test for changes in helical structure. As expected from the structural independence of the core PNT domain from helix H0, the global stability of Ets11-138 against thermal unfolding (Tm = 77 ± 1 oC) did not change Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   61	
         	
  	
  	
  	
   Figure 2.8 Helix H0 of Ets11-138 is conformationally dynamic and sensitive to trypsin proteolysis. Phosphorylation of Ets11-138 promoted local protease resistance and changes in CD spectra, but not global thermal stability. (A) Schematic of Ets11-138 helical secondary structure (colored boxes) and potential trypsin cleavage sites (ticks). The gray box denotes the N-terminal His6-FLAG-HMK-tag. Mass spectrometry identified cleavage sites are indicated by amino acid position (2, 15, 42, and 50 result from lysine-directed cleavages, whereas 62 is an arginine cleavage). (B) SDS-PAGE gels of partial digests of unmodified Ets11-138 (upper) and 2P-Ets11-138 (lower) as a function of increasing trypsin concentration. Cleaved fragments are denoted by position of N-terminal cleavage site, and all fragments terminated at position 138. ♦, position of trypsin. (C) Thermal unfolding curves of Ets11-138 (Tm = 77 ± 1 °C) and 2P-Ets11-138 (Tm = 76 ± 1 °C), as monitored by CD. Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   62	
      ** !" # $ %& !'()*+,-.+/0 %1 2134 !"# !$# !%# %4 24 #4 56444 52444 53444 5%444 5"444 78*&&5&"# 19578*&&5&"# :e ; 1 11 -< =/ -> ? 1 @ <? AB 7HPSHUDWXUHÛ& 1 &3 78*&&5&"# 19578*&&5&"# 0 0.05 0.1 0.2 0.5 1 2 2 15 42 50 62 2 15 42 50 62 Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   63	
   significantly upon phosphorylation or introduction of the S41E mutation. However, the mean residue molar ellipticity [θ]222 of Ets11-138 decreased by ~7% from -6500 to -7000 deg cm2/dmol upon dual phosphorylation of Thr38 and Ser41 (Figure 2.8C). This is suggestive of slightly increased helical secondary structure content and/or perturbations in relative helix orientations (Gagne et al., 1994). Furthermore, trypsin susceptibility of Ets129-138 at Lys42, Lys50, and Arg62 within helices H0 and H1 was reduced upon phosphorylation and with the S41E mutation, suggesting increased stability in the phosphorylated state (Figure 2.8B). Finally, a comparison of the 15N-HSQC spectra of Ets129-138 and 2P-Ets129-138 indicated that, in addition to residues within helix H0, those in helices H2 and H5 are perturbed upon phosphorylation (Figure 2.2A and 2.9). This is indicative of phosphorylation-dependent conformational changes in the PNT domain.  For higher resolution analyses of the changes upon phosphorylation of Ets1 PNT, the structural ensemble of 2P-Ets129-138 was also determined by NMR-derived distance, dihedral angle, and orientational restraints (Figure 2.4C, Table 2.1). The core PNT domain and helix H1 superimpose closely on those of the unmodified protein. In contrast, although residues 42-52 continue to form helix H0 as evidenced by secondary structure predictions (Figure 2.6) and sequential mainchain NOE interactions, the position of this helix is no longer restrained by long-range NOE interactions (Figure 2.10). Rather, helix H0 adopts a broad conformational distribution, extending roughly outward from helix H1 and away from the core PNT domain (Figure 2.4C). This creates an "open" conformation that contrasts to the more compact, "closed" conformation observed in the ensemble of unmodified Ets129-138. This distribution is not experimentally unrestricted, but rather limited by short-range restraints around the Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   64	
    Figure 2.9 Chemical shift perturbations upon phosphorylation of Ets129-138 Consistent with the displacement of helix H0 away from the core PNT domain in the structural ensemble of 2P-Ets129-138 (Figure 2.4), chemical shift perturbations upon phosphorylation localize to Thr38, Ser41, and the interface of helices H0, H2, and H5. (A) Superimposed 15N-HSQC spectra of Ets129-138 (black) and 2P-Ets129-138 (red). (B) Combined chemical shift perturbations mapped by brown sphere size on corresponding low energy members of the structural ensembles of (left) Ets129-138 and (right) 2P-Ets129- 138. The sidechains of Thr38 (cyan) and Ser41 (blue) are shown. See figure 2.2A for the corresponding histogram of chemical shift perturbations. !"#$%& '"#$%& (")*++"##, ("+*-+"##, ("+*)."##, /) /+ /0 /1 /. /02 /- /) /+ /0 /1 /. /02 /- 3*+ 4*. 4*+ 5*. ))+ )). )0+ )0. ). 6 "7# #, 8 )/"7##,8 798 7:8 ;<&)03')-4 0=';<&)03')-4 >.. ?1) @-4 A)0+ ?1+ A.- @.0 ?1B C.+ D)0- /)04 E)0. 64B F)-) G4+ E-5 C10 E-B 9.) 914 ;1- H11 I15 F)01 941 Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   65	
      Figure 2.10 Methyl-methyl NOE comparison between Ets129-138 and 2P-Ets129-138 Strip plots are shown from constant time methyl 13C-13C-1H NOESY spectra of (left strip of pair) Ets129-138 and (right strip) 2P-Ets129-138 at the 13C methyl shifts of Leu49 and Ile124. Key NOE's that defined the position of helix H0 in the unmodified protein are identified in red boxes. These are uniformly absent in 2P-Ets129-138. !" #$ !" $$ !" %$ !" &! !" #! !" $! !" '$ !" ($ !" &$ )" !! !" '! !" (! !" #$ #" %$ #&'()*%#+'%!, -.+/0b# 1%#./0b% 1%#./0a# !" &! !" #! !" $! !" #$ !" $$ !" %$ !" '$ !" ($ !" &$ )" !! !" '! !" (! I124 Cb1 I131 Cb1 I124 Cb1 I131 Cb1 L49 Cb2 L49 Cb1 A48 C` L49 Cb2 L49 Cb1 -.+/0b% %2/34456 %! 0 /34 45 6 ()*%#+'%!, #&'()*%#+'%!,()*%#+'%!, #&'()*%#+'%!,()*%#+'%!, #&'()*%#+'%!,()*%#+'%!, Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   66	
   H0/H1 junction and by 1HN-15N and 1Hα-13Cα residual dipolar couplings. Although an ensemble of conformations was detected, one possible state may exist in which helix H0 is dissociated from H2/H5 and continuous with helix H1. This hypothesis is supported by several complementary lines of evidence. First, whereas helix H0 has similar SSP scores in Ets129-138 and 2P-Ets129-138, residues at the H0/H1 junction have slightly higher scores in the phosphorylated species (Figure 2.6A and B). This may reflect increased average helical character of these residues detected by CD and the stabilization detected by protease sensitivity measurements. Second, model free analysis of 15N relaxation data indicated a change in local dynamics upon phosphorylation, leading to a loss of conformational exchange broadening at the H0/H1 junction (Figure 2.7D and E), which would be consistent with a transition to a more predominant, continuous structure. Third, the protection factors against HX increased slightly (~2 to 5 fold) for amides near pThr38 and pSer41, including those at the N- terminus of helix H0 (Figure 2.7F). This may reflect local stabilization of the helix, although electrostatic retardation of base-catalyzed HX may also contribute. In spite of this evidence for stabilization, small decreases in order parameters S2 for residues in helix H0 (Figure 2.7D) are consistent with the overall increased mobility of this helix as it is released from packing against the PNT domain.  The sequences preceding helix H0, including Thr38 and Ser41, remained disordered, as also revealed by chemical shift, HX, and 15N relaxation measurements (Figures 2.6 and 2.7). Based on pH-dependent amide chemical shifts, the pKa values of pThr38 and pSer41 in 2P-Ets129-138 are 6.5 ± 0.1 and 5.8 ± 0.1 (Figure 2.11). These are close to the reported random coil pKa values of 6.3 for pThr and 6.0 for pSer Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   67	
          Figure 2.11 The pKa values of pThr38 and pSer41 in 2P-Ets129-138 The electrostatic environments of pThr38 and pSer41 in 2P-Ets129-138 are not significantly perturbed by the PNT domain as revealed by the similarity of their pKa values to those expected for a random coil polypeptide (Bienkiewicz and Lumb, 1999). The pKa values of pThr38 and pSer41 in 2P-Ets129-138 were determined from the pH- dependence of their amide 1HN (red) and 15N (blue) chemical shifts (20 mM Tris, 20 mM NaCl, 2 mM DTT, and ~10% D2O at 30 oC). The data were fit simultaneously to the illustrated equation to obtain limiting chemical shifts for the conjugate acid and base forms and the associated pKa values. Errors are from a Monte Carlo routine assuming standard deviations of 0.05 for pH, 0.02 ppm for 1HN shift, and 0.2 ppm for 15N shift. !!" !!# !$% !$! !$$ !$& !$' & ' ( ) * " # +,+, !( - ./ 01 2 3/ 45 .6 03 78 !( - ./ 01 2 3/ 45 .6 03 78 +9&".3:.$;<=86!$#<!&" +>'!.3:.$;<=86!$#<!&" b!"#$% b&'( )*+$,$b"'()*-& '()*+$,$'()*-& +?4 )@(!%@! b& b" !(- !,- !!"@'!%@$ !$&@*!%@! "@'%!%@%& #@&)!%@%! +?4 (@"!%@! b& b" !(- !,- !!*@*!%@$ !$%@)!%@! "@*!!%@%& #@'$!%@%$ !% "@$ #@) "@' "@) "@" #@% #@$ #@' !, -./012 3/45.60378 !, -./012 3/45.60378 #@) "@) "@" #@% #@$ #@' & ' ( ) * " # !% !!*@( !!"@% !!"@( !!#@% !$%@% !$%@( !$!@% !!#@( o !(- 6!,- o !(- 6!,- Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   68	
   (Bienkiewicz and Lumb, 1999), indicating that the electrostatic environments of the two residues are not significantly perturbed by the PNT domain. However, the mobility of helix H0 impacts the positioning of these phosphoacceptor sites relative to the core PNT domain, becoming more distal in the open conformation.  Helix H0 exists in a phosphorylation-dependent equilibrium between "open" and "closed" states The NMR-derived ensembles of Ets129-138 and 2P-Ets129-138, as well as changes observed in chemical shifts, HX, protease sensitivity and CD molar ellipticity, indicated that a substantial conformational transition accompanies phosphorylation. However, in recognizing that helix H0 is dynamic even in the absence of phosphorylation, we propose that the unmodified Ets129-138 also exists in a conformational equilibrium between these closed and open states. In this model, phosphorylation shifts the population distribution to the latter state with helix H0 displaced from the core PNT domain, possibly creating a continuous helix with H1 (Figure 2.12).  To further investigate the effect of phosphorylation on the structural equilibrium of the Ets1 PNT domain, several truncated mutants were generated and compared with respect to their 1HN-15N amide chemical shifts. The truncated mutants included the full PNT domain and phosphoacceptors (Ets11-138 and Ets129-138), the PNT domain without the phosphoacceptors (Ets142-138), and the core PNT domain without helix H0 (Ets151- 138) (Figure 2.13). An overlay of the 15N-HSQC spectra of these four constructs and 2P- Ets11-138 revealed progressive chemical shift changes for several residues located at the interface of H0, H2, and H5 in the structure of the apo-Ets129-138 (Figure 2.13B). As an Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   69	
   example, the amide 1HN and 15N chemical shifts of Leu125 and His128 located in H5 and involved in the hydrophobic core of the H0/H2/H5 interface were the same for Ets11-138, Ets129-138, and Ets142-138. This is diagnostic of the same structural environments of these residues in the three fragments. However, upon phosphorylation (2P-Ets11-138), the chemical shifts of these residues were perturbed (Figure 2.13C). Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   70	
      	
     Figure 2.12 Dynamic helix H0 is in a conformational equilibrium In our model, in the absence of phosphorylation, H0 packs against H2 and H5, the closed state. However, in this state, H0 is dynamic as evidenced by chemical shift, 15N relaxation, amide HX, and proteolysis measurements (see the text). Upon phosphorylation of Thr38 and Ser41, the conformational equilibrium shifts towards the open state in which H0 is displaced away from the core PNT domain. Within the broad distribution of conformations, H0 and H1 may exist as a continuous helix (black).   - -- -- !! !!! ! !! !! !! !! !! !! - -- -- "! #$! !%& '!#!%& () (* () (*('#(+('#(+ --- - -- - - - - --- - - !"#$%&'$()(% #*%+'$()(% Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   71	
        Figure 2.13 Co-linear amide chemical shift changes in the Ets1 PNT domain upon deletions or phosphorylation suggest that helix H0 exists in an equilibrium between "open" and "closed" states. (A) N-terminal deletion mutants of Ets1. Arrows denotes the starting residue of the construct. (B) Overlaid 15N-HSQC spectra of the deletion mutants including that of the phosphorylated Ets11-138 (magenta) with the corresponding color coding. (C) Progressive changes of amide chemical shifts are observed for some H0/H2/H5 interfacial residues, such as Gly55 (left), and His128 and Leu125 (right). !! !" # $ % & !'" !() !(" !!) !!" H0 H1 H2 H2’ H3 H4 H5 PP Et s1 29 -1 38 Et s1 42 -1 38 Et s1 51 -1 38 #*+ #*( #*" $*$ !!! !!" !"# !"$ !) ,- -./ /0 1 2)) 2!"( 3!($ 4!() !"#$%&'()** Et s1 1- 13 8 !3--.//01 $*" %*$ %*& %*+ !(" !!# !!$ !!% .51 .61 .71 2P -E ts 11 -1 38 Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   72	
   When helix H0 was removed as in Ets151-138, the chemical shifts were changed even further, but still in a co-linear manner. The pattern of progressive change suggests that H0 is in a conformational equilibrium between "open" and "closed" states. In the absence of phosphorylation, the conformational equilibrium of H0 is in the "closed" state and in proximity to helices H2 and H5, as seen in the structure of apo-Ets129-138 (Figure 2.4A). Upon phosphorylation, the conformational equilibrium shifts towards the "open" state with helix H0 displaced from the PNT domain, as seen in the structure of 2P- Ets129-138 (Figure 2.4A and B). The complete deletion of helix H0 to yield Ets151-138 defines the fully "open" state.  Gly55, located at the junction of helices H0 and H1, also showed similar pattern of shift changes, with its 1HN-15N amide chemical shifts in 2P-Ets11-138 located between those in Ets151-138 ("open") and Ets129-138/Ets11-138 ("closed"). However, in Ets142-138, the chemical shifts of Gly55 changed co-linearly but in the opposite direction from 2P- Ets129-138 and Ets151-138 (Figure 2.13C). This suggests that helix H0 actually packs better against H2 and H5 when residues N-terminal to this helix (including Thr38 and Ser41) are removed, thereby forming a more "closed" state. Such a behavior might arise as Gly55 is located at the H0/H1 junction, which could act as a hinge-like point tethering either H0 to the core PNT domain or forming a more continuous helix with H1. If so, its amide chemical shifts would be exquisitely sensitive to effects such as hydrogen bonding. In contrast, Leu125 and His128 are located in the middle of the H5 and are involved in the H0/H2/H5 interface.  Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   73	
   An overall examination of the 15N-HSQC spectra of the Ets1 deletion mutants revealed three additional features of the PNT domain. Firstly, the structural boundaries of the full- sized Ets1 PNT domain are delineated by residues 42-138, as Ets142-138 is the minimal construct whose amide signals are well superimposed to the corresponding signals on the 15N-HSQC spectra of both Ets11-138 and Ets129-138 (Figures 2.14A, B, and 2.15). Secondly, H0 interacts with the core PNT domain (H1-H5), as seen in the structural ensemble of apo-Ets129-138. Removing an additional 9 residues (42-50) at the N-terminal end of Ets142-138 resulted in significant perturbations of the amide chemical shifts in the C-terminus of helix H2 and its structurally neighboring region, the N-terminus of helix H5 (Figures 2.14B, C, and 2.15C).  Thirdly, residues 51-138 constitute the minimal-sized core PNT domain (Figure 2.1). Despite the spectra changes resulting from the absence of residues 42-50, the 15N-HSQC spectrum of Ets151-138 still exhibited dispersed peaks, indicative of a well-folded structure. This is consistent with 15N-relaxation analysis, demonstrating that residues 51-138 are relatively rigid, whereas the preceding residues are relatively flexible (Figure 2.7). Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   74	
           Figure 2.14 Identification of the minimal PNT domain using deletion mutants Residues 42-138 constitute the approximate boundaries of the full-sized structured PNT domain. (A) Various deletion fragments of Ets1: intact, native N-terminus (Ets11-138), deletion of residues preceding the phosphoacceptor sites (PP) (Ets129-138), deletion of the phosphoacceptor sites (Ets142-138), deletion of helix H0 (Ets151-138). (B) Amide chemical shift changes, calculated as {(10Δδ1HN)2 + (Δδ15N)2}1/2, of the constructs in (A) as compared to Ets11-138 (see figure 2.15). The dashed line denotes a 0.25 ppm cutoff, used in (C). (C) Amide chemical shifts > 0.25 ppm highlighted in red on the structure of the PNT domain. The arrow shows the starting residue of the next shorter construct in each comparison.  Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   75	
      ! !" #" $" %" &" '" (" )" *" !"" !!" !#" !$" ! !" #" $" %" &" '" (" )" *" !"" !!" !#" !$" " "+# "+% "+' "+) !+" " "+# "+% "+' "+) !+" !"#$$%$&'()#(!"#$*+%$&'( ,- ./ 0, 12 (# -0 3"( ,- 14 5. #( 67 7/ 8 ! !" #" $" %" &" '" (" )" *" !"" !!" !#" !$"" "+# "+% "+' "+) !+" H0 H1 H2 H2’ H3 H4 H5 H0 H1 H2 H2’ H3 H4 H5 H0 H1 H2 H2’ H3 H4 H5 H1 H2 H2’ H3 H4 H5 PP PP !"#$$%$&' !"#$*+%$&' !"#$9*%$&' !"#$:$%$&' ;.#0<=. 6>8 6?8 6@8 !"#$*+%$&'()#(!"#$9*%$&'( !"#$9*%$&'()#(!"#$:$%$&'( !"#$$%$&'()#(!"#$*+%$&'( !"#$*+%$&'()#(!"#$9*%$&' !"#$9*%$&'()#(!"#$:$%$&' Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   76	
    Figure 2.15 Overlaid 15N-HSQC spectra of various Ets1 deletion mutants The N-terminal deletion mutants of Ets1 (see Figure 2.14A) were compared using amide chemical shifts from 15N-HSQC spectra. In each panel, the 15N-HSQC spectrum of the longer construct is placed in the lower layer (black), while that of the shorter construct is in the upper layer (red). !"#$$%$&'((!"#$)*%$&' !! !" # $ % & !'" !(" !!" !'" !(" !!" !'" !(" !!" $+ ,( -. ./ 0 -10 $2(-../0 !"#$)*%$&'((!"#$3)%$&'-40 !"#$3)%$&'((!"#$+$%$&'-50 Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   77	
   2.3 Discussion In this chapter, the structure and dynamic properties of the PNT domain and adjacent phosphoacceptors were investigated to understand the role of the N-terminal region of the Ets1 in Ras/MAPK signaling. Using mass spectrometry and NMR spectroscopy, the two phosphoacceptor sites, Thr38 and Ser41, were discovered, respectively, and then confirmed for their functional relevance by mutagenesis studies in vivo.  2.3.1 Structural analysis of Ets129-138 and 2P-Ets129-138 Using improved sample conditions and state-of-the-art NMR methods, we refined the previously reported structure of Ets129-138 (Slupsky et al., 1998). The refined structure is composed of the disordered N-terminal region (residues 29-41) with the two phosphoacceptors (Thr38 and Ser41), the newly identified dynamic helix H0, and the core PNT domain (H1-H5). Helices H2-H5, also known as the SAM fold, formed a well- ordered core.  Helix H1 packs against H2 and H5, as also seen in the structure of the GABPα PNT domain (Mackereth et al., 2004). Previously, the position of this helix was not well defined (Slupsky et al., 1998). Most importantly, an additional, N-terminal helix H0 was identified in Ets1 and Ets2. The dynamic nature of this helix was demonstrated by its lower SSP scores, lower S2 order parameters, fast amide HX, and sensitivity to trypsin cleavage. Moreover, deletion of helix H0 did not perturb the folding of the core PNT domain (H1-H5).  Upon phosphorylation, no major structural change is observed in the core PNT domain. However, in the ensemble structure of 2P-Ets129-138, helix H0 is displaced from this core, adopting a broad range of conformations. This is evidenced by the absence of Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   78	
   long-range NOEs from residues in helix H0 to those in helices H2 and H5, by chemical shift perturbations in H0/H2/H5, and by residual dipolar coupling analyses. In the presence of phosphorylated Thr38 and Ser41, H0 is still dynamic, albeit with a slightly increased local stability by protection from hydrogen exchange and proteolysis. In contrast, 15N relaxation analysis suggests slightly increased mobility for residues at the N-terminal end of helix H0, yet decreased mobility for Gly55 at the H0/H1 junction. The latter observation suggests that H0 may form an extended helix with H1 upon phosphorylation. The slightly more ‘structured’ junction may result from H0, which no longer wraps around the PNT domain to contact H2 and H5.  2.3.2 Helix H0 exists in a phosphorylation-dependent conformational equilibrium The dynamic nature of the H0, both in apo- and 2P-Ets1 PNT, strongly suggests that H0 exists in a conformational equilibrium between a closed state, represented by Ets129-138, and a more open state, represented by 2P-Ets129-138 (Figure 2.12). Although the mechanism for how phosphorylation alters the dynamics of the H0 is still unclear, it is plausible that the displacement of the H0 from H2 and H5 is triggered by the proximity of the negative charges in pThr38 and pSer41 to the numerous aspartates and glutamates in helices H2 and H5. Thus, electrostatic repulsion of the two phosphorylated residues may shift the population distribution of the PNT domain towards the open state. In fact, pThr38 and pSer41 may also interact with nearby Lys42 and Lys50 to disrupt the salt bridges, Lys42-Asp123 and Lys50-Glu127, formed between helix H0 and H5. Such charge repulsion could enable the altered conformation, including the putative continuous helix linking H0 to H1. Alternatively, the Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   79	
   negative charges introduced by phosphorylation may also serve as an N-cap by interacting with the dipole moment of the helix to stabilize the continuous helix.  2.3.3 Biological specificity from regulatory elements appended to conserved domains Helices H2-H5 form a conserved core in all PNT and SAM domains. Elements appended to this core provide possible routes to biological specificity. For example, most SAM domains, such as in STE50, polyhomeotic, EphB2 receptor, and p73, bear a shortened H2 and a lengthened H5 in comparison to those in the PNT domains (Ets1/2, GABPα, Tel, Erg, Pnt-P2). The size of the H5 in the EphB2 receptor was shown to be critical for its self-association (Stapleton et al., 1999; Thanos et al., 1999). Among the PNT domains, helix H1 is present in only Ets1, Ets2, GABPα, and Pnt-P2 (Figure 2.1). The role of H1 was speculated to be critical for preventing self-association by blocking one of the two oligomerization surfaces found in the PNT domains of polymeric Tel and Yan, as well as several SAM domains (Stapleton et al., 1999; Thanos et al., 1999; Kim et al., 2001; Qiao et al., 2004). Furthermore, in Ets1, Ets2 and Pnt-P2, helices H0 and H1 may play a key role in MAP kinase docking to increase the local concentrations of the phosphoacceptor near the enzyme catalytic site. As described in chapter 1, one function of the Ets1 PNT domain is to provide a binding interface or docking site for MAP kinase, namely the LXLXXXF motif (Leu114, Leu116, Phe120) localized adjacent to the H0/H2/H5 interface, to enhance the phosphorylation in the N-terminal region (Seidel and Graves, 2002). A significant increase in the Km value of the Ets1 PNT for ERK2 was observed when the PNT domain was mutated or deleted (Seidel and Graves, 2002). Consistent with the results shown in this chapter, several other studies Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   80	
   have shown that the structural plasticity of H0 is necessary for the phosphoacceptor sites in the Ets1 PNT to reach out for the catalytic site of ERK2 that is distant from the binding interface, the F-recruitment site (Figure 2.16A) (Mackereth, 2003; Callaway et al., 2010).  Self-polymerization of the SAM domain is a key step in the transcriptional repression activity of vertebrate Tel and its Drosophila ortholog Yan. In the Tel repression complex model, Kim et al (Kim et al., 2001) have proposed that when Tel is bound to DNA through its C-terminal ETS domain, the polymerization of its PNT domain provides compactness to the packing of the nucleosome core particles to block the accessibility of other proteins (transcriptional machinery) to chromatin. In Ets1, one of the Tel- equivalent dimerization interfaces coincides with the H0/H2/H5 interface and is adjacent to the MAPK docking site Phe120. When the Ets1 PNT structure was overlaid onto the TEL PNT polymer structure, H0 in the Ets1 PNT domain directly blocks the hydrophobic patch of the Tel dimerization surface, thereby preventing self-association (Figure 2.16B). However, removal of H0 did not trigger polymerization of the deletion mutant Ets151-138, suggesting that H1 may also interfere with the complete exposure of one of the interfaces, the EH.  Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   81	
            Figure 2.16 Roles of helix H0 in MAPK docking and in disrupting self-association. (A) Ets1 PNT is a docking module for the MAP kinase ERK2 to enhance phosphorylation of the adjacent Thr38 and Ser41. The modeled complex structure of Ets1 PNT (red, tube) on ERK2 (grey, surface) shows Ets1 PNT binds ERK2 through contacts that are distant from the active site of ERK2. However, the flexibility of H0 provides structural plasticity to position the phosphoacceptor sites near the active site. (B) The Ets1 PNT domain structure (orange) is overlaid onto one of two units in the TEL PNT polymer structure (grey; 1ji7.pdb). Based on this alignment, helix H0 would prevent self-association of the Ets1 PNT domain.  Some helices are labeled for Ets1 PNT, and selected sidechains are shown in green. !"#$%&'(%)*%"+,%(,-%.)-/0,1 23 2$ 24 25&'( 6)078* .9"7"8:, 6);<8*=%;-,>" 7;"8:,% #8", !?@4 &+,$43 (+1AB C,1D$ EFG EHG Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   82	
   In the context of this chapter, a key role of helix H1 lies with the newly identified dynamic helix H0. The latter helix H0 is found only in Ets1, Ets2, and by sequence comparison, Pnt-P2 (Figure 2.1). As will be shown and discussed further in Chapter 3, helix H0 is critical for recruitment of the co-activator CBP. Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   83	
   2.4 Materials and methods 2.4.1 In vivo transcription assays and immunoprecipitation Transient expression assays were performed as previously described in NIH3T3 mouse fibroblasts (Foulds et al., 2004). In brief, transfected plasmids included 2.5 mg of firefly luciferase reporter vector driven by an artificial double ETS site, 0.1 mg of wild-type or mutant FLAG-Ets1 expression vector (or empty vector), 0.1 mg constitutively active CA- MEK1 (Seidel and Graves, 2002) or empty vector, and 1 mg Renilla luciferase expression vector pRL-null (Promega). After transfection, cells were serum-starved to reduce constitutive signaling. Relative luciferase activity was determined as the ratio of firefly / Renilla activity. For expression controls, ~1.5x106 cells were transfected with 3 mg of full-length FLAG-Ets1 expression plasmid, 3 mg constitutively active CA-MEK1, and 50 ml Lipofectamine (Invitrogen). As described previously (Foulds et al., 2004), antibodies against the FLAG tag were used for immunoprecipitation, followed by immunoblot detection with either anti-Ets1 antisera (UT2) (Gunther and Graves, 1994) or phospho-specific antisera against pThr38-Ets1 (Biosource).  2.4.2 Protein purification Samples of unlabeled and uniformly 15N- or 15N/13C-labeled murine Ets1/2 constructs (Ets11-138; S26A, Ets129-138, Ets151-138, Ets11-52, Ets21-172, and Ets263-172) were prepared as described previously (Slupsky et al., 1998; Seidel and Graves, 2002; Mackereth et al., 2004; McIntosh et al., 2009). The His6-tagged constructs were incubated with thrombin (Roche) in 20 mM Tris pH 8.4, 150 mM NaCl, and 2.5 mM CaCl2 overnight. Thrombin and the cleaved His6-tag were removed using ρ-aminobenzamidine beads (Sigma) and TALON metal affinity resin (BD Biosciences). Ets11-138 and Ets129-138 were Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   84	
   phosphorylated in vitro using 50 mM ATP and a 1:20 molar ratio of ERK2:Ets1 (Foulds et al., 2004). After incubation at 30 oC for 1.5 hr, the active bacterially expressed His6- tagged ERK2 was removed with a HisTrap FF column (GE Healthcare). Ets1 samples of differing phosphorylation states were separated on a Mono Q column (GE Healthcare) using a 0-400 mM KCl linear gradient in 20 mM Tris pH 7.5, 10% glycerol, and 2 mM DTT buffer, followed by passage through a Superdex 75 column (GE Healthcare) equilibrated in 50 mM Tris pH 7.5, 10% glycerol, 50 mM KCl, 0.1 mM EDTA, and 2 mM DTT. Phosphorylation states were verified by ESI-MS. Final sample concentrations were determined from predicted ε280 absorbance values (ProPram, Expasy Proteomics Server). 	
   2.4.5 NMR spectroscopy NMR spectra of the Ets1/2 and CBP constructs were recorded using Varian 500 MHz Unity and 600 MHz Inova spectrometers, and analyzed using NMRpipe (Delaglio et al., 1995) and Sparky (Goddard). 1H, 13C, and 15N resonance assignments were obtained via standard heteronuclear correlation experiments, initially using proteins in 20 mM sodium phosphate (pH 6.2 for Ets129-138 and pH 6.3 for 2P-Ets129-138), 20 mM NaCl, 2mM DTT, and ~10% D2O at 30 oC. Amide 15N relaxation parameters for 15N-labeled protein samples were acquired on a 600 MHz spectrometer at 30 oC, as described previously (Farrow et al., 1994), and analyzed using TENSOR2 (Dosset et al., 2000). The chemical shifts of Ets129-138 (#4205) and 2P-Ets129-138 (#16426) have been deposited in the Biological Magnetic Resonance data bank. 	
   Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   85	
   Structural restraints NOE restraints were determined from simultaneous 3D 1H-15N/13C-1H NOESY-HSQC (aliphatic side chains), 1H-13C-1H NOESY-HSQC (aromatic side chains), and simultaneous constant-time 3D 1H-15N/13C-1H NOESY-HSQC spectra (methyl groups), all recorded with τm = 140 ms on a uniformly 15N/13C-labeled protein sample (Zwahlen et al., 1998). 1HN-15N RDC orientation restraints were obtained from 1H-15N IPAP-HSQC spectra of 15N-labeled Ets129-138 and 2P-Ets129-138 recorded in the absence versus presence of ~15 mg/mL and ~12 mg/mL Pf1 phage, which yielded splittings of ~16.5 Hz and ~15 Hz in the 2H-NMR spectra of the 1HO2H lock solvent, respectively (Ottiger et al., 1998). 1Hα-13C RDC restraints were also measured for 2P-Ets129-138 under the same conditions using the HNCO-based pulse scheme as described previously (Yang et al., 1998b; de Alba and Tjandra, 2004).  2.4.6 Structure calculations Analyses of the following structures were performed with Procheck-NMR (Laskowski et al., 1996), Promotif (HUTCHINSON AND THORNTON, 1996), and MolMol (Koradi et al., 1996). The atomic coordinates of the Ets129-138 (2jv3.pdb) and 2P-Ets129-138 (2kmd.pdb) ensembles have been deposited in the Research Collaboratory for Structural Bioinformatics Protein Databank.  Ets129-138 Structure calculations for Ets129-138 were performed using ARIA/CNS v2.1 (Rieping et al., 2007) utilizing distance, dihedral angle, and orientation restraints. The majority of the NOESY cross-peaks were manually picked and assigned prior to intensity Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   86	
   calibration by ARIA. Backbone dihedral angles were determined from 13Cα, 13Cβ, 1HN, and 1Hα chemical shifts using TALOS (CORNILESCU ET AL., 1999) and SSP (Marsh et al., 2006). A limited set of hydrogen bond distance restraints was included for amides located in helices that showed protection from rapid exchange after transfer to 2H2O buffer (Slupsky et al., 1998). 1HN-15N RDC orientation restraints for amides with heteronuclear 1H{15N}-NOE ratios > 0.2 were then incorporated with the SANI algorithm at iteration 5 of the ARIA protocol, using default energy constants of 0.2 and 1 kcal mol−1 Hz−2 for the first and second simulated annealing cooling stages, respectively. Values of the alignment tensor (R = 0.6 and Da = +4.9 Hz) were estimated by fitting the measured RDC's against preliminary structures of Ets129-138 calculated without these restraints. The ARIA/CNS calculations of Ets129-138 were performed as a two-step protocol. In the initial ARIA round (it0 → it8), starting with an unfolded polypeptide, was completed with only dihedral bond restraints and uncalibrated NOE data. A second full ARIA calculation (it0 → it8) was then performed, starting with an unfolded polypeptide and using the final unambiguous and ambiguous restraints from the previous round, as well as the dihedral angle, hydrogen bond, and RDC restraints. A total of 100 structures per iteration were calculated, with the 25 lowest energy structures of the final iteration refined in a water box using Lennard-Jones potentials.  2P-Ets129-138 Structure calculations for 2P-Ets129-138 were performed initially using CYANA (Guntert, 2004) incorporating distance, dihedral angle, and orientation restraints, and then refined using SculptorCNS (Charavay et al.) (Table 2.1). NOE-based distance restraints were generated using the automated NOESY assignment routine in CYANA. Chemical shift Chapter 2 – Phosphorylation and the Ets1 PNT domain 	
   	
   87	
   derived backbone dihedral angles (with a minimum range of ±40°) were included only for residues with high TALOS scores (good > 9) and in agreement with SSP. Hydrogen bond distance restraints were included only for amides in helices H1-H5 of the core PNT domain that showed protection from rapid hydrogen exchange in the unmodified protein. For subsequent refinement with SculptorCNS, 1HN-15N and 1Hα-13Cα RDC orientational restraints for residues with heteronuclear 1H{15N}-NOE ratios > 0.2 were included in addition to the dihedral angle and CYANA-generated distance restraints. Independent alignment tensors were used for the dynamic H0 and the core PNT domain (H1-H5). From total of 500 structures, the 20 lowest energy structures were selected for the final structural ensemble.  2.4.7 Amide hydrogen exchange Rapid amide proton-proton exchange rates at 30 oC and pH 6.0 were determined for 15N-labeled Ets129-138 and 2P-Ets129-138 by the CLEANEX-PM method (Hwang et al., 1998) using transfer periods ranging from 10 to 60 msec (Lee et al., 2005). Each exchange spectrum was recorded using a recycle delay of 1.5 sec, and the reference spectra with a recycle delay of 12.0 sec. Pseudo-first order rate constants for chemical exchange, kex, were obtained by least squares fitting of peak intensities using a Matlab module provided by J. Choy (Univ. Toronto) and assuming a negligible contribution from water relaxation. HX protection factors (kpred/kex) were derived using predicted kpred values for an unstructured polypeptide with the sequence of Ets129-138, calculated with the SPHERE server (Bai et al., 1993; Zhang) using poly-D,L-alanine reference data corrected for amino acid type, pH, temperature, and isotope effects, and asparate as surrogate for phosphoserine/threonine. Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   88	
   Chapter 3 - Phosphorylation-dependent Ets1-CBP interaction 	
   In this chapter, the molecular basis of the Ras/MAPK signaling input to transcription factor Ets1 via phosphorylation-enhanced binding of the co-activator CBP was investigated. Phosphorylation of Thr38 and Ser41 shifted a conformational equilibrium of the PNT domain involving the dynamic helix H0 and increased the affinity of Ets1 binding to the TAZ1 domain of CBP by 34 fold. NMR-monitored titration experiments mapped the interaction surfaces of the TAZ1 domain of CBP and Ets1, the latter encompassing both the phosphoacceptors and PNT domain. Charge complementarity of these surfaces indicated that electrostatic forces act in concert with conformational change to mediate the effects phosphorylation.         Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   89	
   3.1 Introduction One of the key functions of the Ets1 PNT domain is to provide a docking site for MAPK/Erk2 to enhance phosphorylation at the adjacent phosphoacceptor sites, Thr38 and Ser41 (Seidel and Graves, 2002; McIntosh et al., 2009). Subsequently, phosphorylated Ets1 binds the transcriptional co-activator CBP (or KAT3a) with higher affinity than unmodified Ets1, thereby resulting in increased expression of Ras- dependent target genes.  CBP and the closely related p300 are widely used co-activators that serve both as scaffolds for assembly of specific and general transcription factors and as acetylases to modify lysines of both histone and non-histone proteins (Wang et al., 2008). CBP and p300 are recruited by over 300 reported DNA binding transcription factors to their respective enhancer and promoter elements (Kasper et al., 2006). These co-activators bear several conserved domains, linked by intrinsically disordered sequences which mediate these interactions. However, our structural understanding of how any given domain can interchangeably bind so many different transcription factors is limited to a small number of examples. These include complexes of the isolated TAZ1 and TAZ2 domains of CBP with polypeptides corresponding to the transactivation domains of HIF1α (Dames et al., 2002; Freedman et al., 2002), CITED2 (Freedman et al., 2003; De Guzman et al., 2004), STAT1/2 (Wojciak et al., 2009), and p53 (Feng et al., 2009), as well as complexes of the KIX domain of CBP with segments of CREB (pKID) (Radhakrishnan et al., 1997), MLL 	
  (Goto et al., 2002), c-Myb (Zor et al., 2004), and p53 (Teufel et al., 2007; Lee et al., 2009). Thus, a molecular analysis of the Ets1-CBP interaction will address not just the role of phosphorylation in CBP recruitment, but also Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   90	
   help to shed light on the competitive binding of many factors by this general co- activator.  This chapter demonstrates the coupled roles of the structured PNT domain and an adjacent disordered region bearing the phosphoacceptors, Thr38 and Ser41, in the ERK2-enhanced binding of CBP by Ets1. Phosphorylation plays a dual role by altering a conformational equilibrium of a dynamic helix H0 that tethers the flexible phosphorylated region of Ets1 to the core PNT domain and by augmenting electrostatic forces that drive TAZ1 recognition. In this integrated model, structured elements with intrinsic flexibility provide a scaffold for responsiveness to a signaling pathway. The overall mechanism represents an evolutionary development within a gene family, whereby dynamic elements are appended to conserved core folds to increase the capacity for biological regulation. Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   91	
   3.2 Results 3.2.1 Identification of phosphorylation-dependent binding of Ets1 PNT with the TAZ1 domain of CBP Biophysical investigation of the mechanism by which phosphorylation of Ets1 enhances CBP binding required minimally-sized active fragments. Previous analyses, as well as studies presented in chapter 2, had shown that both the PNT domain and the preceding N-terminal residues, present in the deletion fragment Ets11-138, were required for in vivo signaling dependent effects and in vitro binding to full-length CBP (Foulds et al., 2004). A smaller fragment (residues 1 to 52), which lacks the PNT domain but retains the phosphoacceptors, was ineffective in binding. Analogous data were obtained with the highly-related Ets2 (Foulds et al., 2004).  The CBP interaction domain(s) for binding to Ets1 and Ets2 were initially mapped by pull-down assays with GST-tagged CBP fragments. An N-terminal fragment of CBP, which bears three transcription factor interaction domains, displayed binding and was subjected to further analysis by the Graves' laboratory (Figure 3.1). The isolated N- terminal domain, which binds nuclear receptors, and the KIX domain, best known for CREB binding, did not measurably associate with Ets1. In contrast, the TAZ1 domain, which had been previously shown to bind Ets1 in a constitutive manner (Yang et al., 1998a), bound full-length Ets1, truncated Ets11-138 and the analogous Ets21-172. Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   92	
              Figure 3.1 Affinity column pull-down assays for the phosphorylation-dependent interaction of CBP and Ets1/2 Pull-down assays demonstrate that the TAZ1 domain is necessary for the phosphorylation-dependent interaction of CBP with Ets1 and Ets2. FLAG-HMK-Ets11-440 (full-length), His6-FLAG-HMK-Ets11-138, His6-FLAG-HMK-Ets21-172, and the indicated GST-tagged CBP constructs (TAZ1 is CBP340-439) were expressed in E. coli BL21(λDE3) cells, purified, and in vitro ERK2 phosphorylated (+ ATP) or mock-treated (- ATP), as previously described (Seidel and Graves, 2002; Foulds et al., 2004). FLAG pull-down experiments were conducted using published methods, except that NP-40 was omitted from all reaction buffers (Foulds et al., 2004). Briefly, 240 pmol of the FLAG-tagged Ets1 and Ets2 constructs were captured on FLAG antibody beads (Sigma) and 400 nM of the GST-CBP proteins were tested for binding. Binding reactions were conducted between 4 hr (GST-CBP fragments) and 16 hr (full-length GST-CBP; not shown (Foulds et al., 2004)) at 4oC. Bound proteins were eluted from the FLAG beads with 3xSDS sample buffer and reaction components resolved on 4-20 % gels. Proteins were detected as indicated with α-GST (GE Healthcare), α-CBP (A22 Santa Cruz), α-His (His Probe Santa Cruz), or α-Ets1 (UT2) antibodies (Gunther and Graves, 1994). Lane 1 shows 4% of the GST-CBP input proteins tested for binding. Lanes 2-8 show GST-CBP constructs bound to beads or to the indicated Ets proteins. The immuno-blot loading controls confirmed that equal amounts of the Ets species ± phosphorylation were bound to FLAG antibody beads. Phosphorylation of Ets1 and Ets2 is apparent by the slight decrease in mobility on the immuno-blot. The positions of molecular mass markers (kDa) are indicated. Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   93	
     ! " ! " #$%&'()* +$(,-$./ #$%&'()* +$(,-$./ 012 #$%&'()* +$(,-$./ 3,/44!556 #$%&'()* +$(,-$./ ! " 3,/44!478 3,/94!4:9 58;6 7<;< :=;6 <9;5 :=;6 <9;5 :=;6 <9;5 5> *?( @A , BC %& / DE1F+B24!7G< DE1F+B24!5G9 DE1F+B25G4!:99 DE1F10H4 466 7G< HH!10H9 10H4 I?J B- +K9*"L?M K01 ?B?M 9554:995G44 Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   94	
   Calorimetric analysis of the phosphorylation-dependent binding of Ets1 PNT with the TAZ1 domain of CBP The interaction of Ets11-138 with the isolated TAZ1 domain was analyzed further by the Graves' laboratory using isothermal titration calorimetry (ITC) to measure the effects of phosphorylation. The affinity of unmodified Ets11-138 for TAZ1 (Kd = 58 ± 12 µM) was enhanced 34 ± 1 fold upon phosphorylation (Table 3.1). Association of the Ets1 and CBP fragments was endothermic (ΔH > 0) and thus entropically-driven (ΔS > 0). The unfavorable enthalpic cost of TAZ1 binding was greater for Ets11-138 than for 2P-Ets11- 138, accounting for the lower affinity of the non-phosphorylated species. This indicated a complex binding mechanism involving possible conformational changes, as well as likely electrostatic and hydrophobic interactions.   Table 3.1 Thermodynamic parameters of Ets11-138/TAZ1  domain binding  Ets11-138 a 2P-Ets11-138 a Change c stoichiometry  1 b 1.3 ± 0.3 Kd (µM) d 58 ± 12 1.7 ± 1.5  34 ± 0.9 ΔG (kcal/mol) d -5.6 ± 0.1 -7.8 ± 0.6 -2.2 ± 0.6 ΔH (kcal/mol) d 4.2 ± 0.5 1.2 ± 0.3 -3.0 ± 0.6 TΔS (kcal/mol) d 9.8 ± 0.4 9.0 ± 0.3  0.7 ± 0.4  a Iodoacetamide-modified Ets11-138, S26A titrated with the TAZ1 domain at pH 7.9 and 15oC. Mean and standard deviations of 4-5 measurements. b Fixed stoichiometry. c Fold change for Kd and direct difference for the remaining parameters. d Values derived from data fitting with Origin 7.0 MicroCal LLC software.  	
   Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   95	
   3.2.2 NMR-based analyses of TAZ1/PNT binding interface To dissect the molecular bases for the constitutive and phosphorylation-enhanced recruitment of CBP by Ets1, 15N-HSQC monitored titrations were used to detect spectral changes, reflective of structural perturbations, accompanying the interaction of TAZ1 with Ets11-138 and 2P-Ets11-138. Furthermore, four deletion mutants of the PNT domain were tested for interaction with TAZ1 by NMR-monitored titrations.  TAZ1-binding interface on Ets1 The binding interface of Ets1 with CBP was identified by monitoring the NMR signals of 15N-labeled Ets11-138 upon titration with the unlabeled TAZ1 domain (Figure 3.2). 15N- HSQC spectra were recorded to identify changes in the mainchain amide moieties of the labeled species during the titrations. Upon addition of unlabeled TAZ1, a subset of 1HN-15N groups in labeled Ets11-138 showed a progressive loss of peak intensity from the free protein signals, without the concomitant appearance of any new signals from the resulting complex, possibly due to conformational exchange broadening (Figure 3.2A).  The residues with the most pronounced signal loss at a 1:1 molar ratio of proteins were mapped on the Ets129-138 structure (Figure 3.2C). These residues were clustered in the region centered around helix H0, the H0/H1 junction, the C-terminal portion of helix H2, and the N-terminal portion of helix H5 of the PNT domain. This qualitatively defines the binding interface for TAZ1 on the Ets1 PNT domain.    Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   96	
           Figure 3.2 TAZ1 domain-binding interface on Ets1 identified by 15N-HSQC monitored titrations The TAZ1-binding interface on Ets11-138 involves primarily helices H0 and H5 of the PNT domain and the phospho-acceptors Thr38 and Ser41. (A) Overlaid 15N-HSQC spectra of ~0.1 mM 15N-labeled (left) Ets11-138 and (right) 2P-Ets11-138 titrated with unlabeled TAZ1 in molar ratios of 1:0 (light blue), 1:0.5 (dark blue), and 1:1 (red). Predicted saturation levels were calculated to be 47% and 88% for Ets11-138 and 2P-Ets11-138, respectively, at the addition of an equimolar TAZ1, using Kd 56 µM and 1.7 µM, respectively, from the ITC measurements. (B) Histograms of amide signal intensity at 50% and 100% equivalence relative to unbound Ets11-138 (top) and 2P-Ets11-138 (bottom). (C) Residues showing > 0.85 loss of signal intensity upon formation of a 1:1 complex are mapped onto corresponding low energy members of the structural ensembles of (left) Ets129-138 and (right) 2P-Ets129-128 (see Figure 2.15). Increasing cyan sphere size corresponds to greater signal loss upon complex formation (intensity ratios: 0 < large sphere < 0.05, 0.05 < medium < 0.10, 0.10 < small < 0.15). Missing data points correspond to prolines or residues with unassigned or overlapping NMR signals. Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   97	
  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hapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   98	
   Titrations were also performed for the shorter construct of the Ets1 PNT, Ets129-138, that was used for the structure calculation. Spectral changes in the 15N-HSQC during the titration of Ets129-138 with non-labeled TAZ1 closely mimic those seen in the titration of Ets11-138 with TAZ1, suggesting residues 1-28 are not critical in binding TAZ1 (Figure 3.3A). Additionally, 13C-HSQC spectra were recorded to monitor changes in signals from methyl and aromatic nuclei in Ets129-138 upon titration with TAZ1 (Figure 3.3). Similar to the mainchain amides, several of these sidechain nuclei also showed a progressive loss of intensity for peaks from the free protein, without the concomitant appearance of any new signals from the resulting complex. When mapped onto the structure of Ets129-138, the most perturbed amide, methyl, and aromatic nuclei all cluster to the same interface described above, namely involving H0, H2, and H5 (Figure 3.4). The amide signal of the first phosphoacceptor site, Thr38, still overlaps with that of another residue in the 15N-HSQC spectrum of Ets129-138, and thus the effect of TAZ1 domain binding could not be determined. Surprisingly, the resolved methyl Hγ2 signal for Thr38 was not perturbed in the 13C-HSQC spectrum, indicating Thr38 is unlikely to be involved in a hydrophobic interaction.  The absence of detectable 15N- and 13C-HSQC signal intensity from the TAZ1/Ets11-138 and TAZ1/Ets129-138 complexes is indicative of conformational exchange broadening within the higher molecular mass species. Even the addition of a 4 fold molar excess of the TAZ1 domain to ensure saturation did not produce signals from the Ets129-138/TAZ1 domain complex with the predicted saturation levels of 44% (equimolar TAZ1) and 82% (4X excess TAZ1) using a Kd value of 56 µM from the ITC measurements.  Similar behaviour was also observed upon titration of Ets21-172 with TAZ1 (not shown), and has Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   99	
            Figure 3.3 TAZ1 domain-binding to Ets129-138 confirmed by 13C-HSQC monitored titrations 15N- and 13C-HSQC spectra of the (A) amide, (B) methyl, and (C) aromatic moieties of 15N/13C-labeled Ets129-138 were recorded before (red) and after (black) the addition of the unlabeled TAZ1 domain to a molar ratio of 1:1 or 4:1 (right, B). The top 20% of residues, ranked by loss of signal intensity, are labeled.  C hapter 3 – P hosphorylation-dependent E ts1/C B P  interaction 	
   100	
    !! !" # $ % & !'" !(" !!" ! ) * + , - - . / !"" #$$ %&'' (") *&)" *&)+ ,&)$ -&)' .// */+ %/0 1/' 2&)& 3"& 2$& #&)4 #"' 3$/ 3/$ !&4+ !&)) $0) $0" %0) %0" &0) &0" )0) !') !'" !() !(" !!) (0" !0) !0" "0) "0" '" () (" !) (")!)5 *&)+-&5 *&)+-)5 *&)"-)5 6&)/!)5 6&)/-&5 2&)&!)5 3&&075 *&4"-&5 3"&75 */+-)5 3/$75 0% TAZ1 100%TAZ1 0% TAZ1 400%TAZ1 !1+,--./ ! ' 2 + , - - . / !1+,--./ (0" !0) !0" "0) "0" ! ' 2 + , - - . / !1+,--./ 8$49) 80)9) 80)1' 8$'-& ,&)$1& #$$15 #&)415 8$',) #$$-5 : : : : : : : 0% TAZ1 100%TAZ1 0% TAZ1 100%TAZ1 !"#$%&'()*+ !,#$,-./&+ !0#$,12-3'.4+ Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   101	
        Figure 3.4 TAZ1 domain-binding interface on Ets129-138 confirmed by mainchain and sidechain spectral changes The top 20% of amides (A), methyls (B) and aromatics (C), ranked by loss of signal intensity upon titration with the TAZ1 domain to 1:1 (figure 3.3), are mapped onto a low energy member of the Ets129-138 structural ensemble as red spheres. Also shown are all remaining methyl and aromatic groups as blue spheres. A comparison with Figure 3.2C confirms that Ets129-138 mimics the binding of Ets11-138 for TAZ1. Helices are labeled and colored in light blue. Û Û Û !"#$"%&'($)*(%&)+,$-*&./ !0#$1(/*2,$)*(%&)+,$-*&./ !3#$"45%+/&)$)*(%&)+,$-*&./ 67 68 69 6: 6; 6<= 3 67 68 69 6: 6; 6< = 3 Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   102	
   been reported for the interaction of various CBP and p53 fragments (Teufel et al., 2007). This exchange broadening may result from the interconversion between bound conformations with different chemical shifts on a milli- to micro-second times, thus precluding detection of any associated NMR signal. Although this behaviour enabled us to qualitatively map the TAZ1 binding interface on Ets1, it unfortunately prevented determination of the three dimensional structure of the resulting complex.  	
   Effect of phosphorylation on TAZ1 binding The effects of phosphorylation on the CBP/Ets1 binding interface were also investigated by NMR-monitored titrations of 15N-labeled 2P-Ets11-138 with the unlabeled TAZ1 domain. Again, a progressive loss of amide 1HN-15N signal intensity from the free protein was observed, without concomitant appearance of signals from the bound protein. The residues showing the most pronounced loss in intensity were mapped onto the structural ensemble of 2P-Ets11-138 (Figure 3.2C). Similar amides were perturbed in the phosphorylated versus unmodified Ets1 fragment, yet to a greater degree at the same molar ratio of TAZ1 (Figure 3.2). This could be due to the increased affinity of TAZ1 for the phosphorylated Ets1 fragment, possibly by additional contacts, with the predicted saturation levels of 47% and 88% for Ets11-138 and 2P-Ets11-138, respectively, at the equimolar amounts of TAZ1, using Kd 56 µM and 1.7 µM, respectively, from the ITC measurements. The altered residues cluster within a region of the PNT domain where the helices H0, H2, and H5 converge in the closed conformation, yet are exposed in the open state. It is plausible that this region functions in binding the TAZ1 domain and the affected residues may directly contribute to the intermolecular interface. Importantly, the Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   103	
   amides of Thr38 and Ser41, especially when phosphorylated, are also perturbed by the TAZ1 domain, indicating a role in binding.  Delineation of the TAZ1 binding interface using Ets1 deletion fragments The determinants for TAZ1 binding were investigated further with the deletion mutants Ets129-138, Ets142-138, and Ets151-138 (Figure 3.5A). Ets129-138, missing a part of the flexible N-terminal region (Figure 3.5A) and used for structure calculations in chapter 2, exhibited very similar 15N-HSQC spectral changes as did Ets11-138 when titrated with TAZ1 (Figure 3.5C and D). When equimolar TAZ1 was added, amide peaks of Ets129-138 mostly corresponding to the residues in H0/H2/H5 lost their signal intensities in a comparable pattern, yet to a slightly lesser extent, than seen for Ets11-138 (Figures 3.5 and 3.6).  To a first approximation, this indicates that the TAZ1-binding residues are contained within Ets129-138. However, the disordered residues 1-28 might enhance binding marginally, perhaps via non-specific electrostatic interactions. This is addressed further below.  The deletion mutant Ets142-138, missing the N-terminal region and the phosphoacceptors Thr38 and Ser41, also showed a loss of 15N-HSQC signal intensity for amides in H0, H2, and H5 upon titration with TAZ1. However, the effect was less dramatic than with the longer Ets1 constructs (Figures 3.5 and 3.6). This is qualitatively indicative of weaker binding, suggesting that the residues preceding H0 (including Thr38 and Ser41) contribute to TAZ1 recognition.    Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   104	
   Figure 3.5 Helix H0 and the phosphoacceptors are key to TAZ1 binding (A) Four deletion constructs of Ets1 were used to identify the TAZ1 binding interface. (B, C, D, E) Overlaid 15N-HSQC spectra from the titrations of the various 15N-labeled Ets1 fragments in the presence of three different molar ratios of TAZ1 to Ets1; 0:1 (red), 0.5:1 (green), and 1:1 (blue). !"# !"$ %"# %"$ &"# &"$ '($ ')# ')$ ''# ''$ '# * ++, -- . / !"$ %"# %"$ &"#!"# '0+,--./ '($ ')# ')$ ''# ''$ '# * +,- -. / !"#$%&' (!"#$%&' '!!"#$%&' !"#$%&' (!"#$%&' '!!"#$%&' !"$ %"# %"$ &"#!"# 123''4'(%+5+678' )-4123''4'(%+5+678' !"# !"$ %"# %"$ &"# &"$ !"#$%&' (!"#$%&' '!!"#$%&' !"#$%&' (!"#$%&' '!!"#$%&' 123'9)4'(%+5+678' 123')!4'(%+5+678' 0$ 0' 0) 0): 0( 09 0# 0$ 0' 0) 0): 0( 09 0# 0$ 0' 0) 0): 0( 09 0# ;;123''4'(% 123')!4'(% 123'9)4'(% 0' 0) 0): 0( 09 0#123'#'4'(% ,7/ '0+,--./ ,</ ,=/ ,>/ ,1/ Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   105	
    Figure 3.6 Helix H0 and the phosphoacceptors are key to TAZ1 binding Intensity ratios of amide 1HN-15N signals from various Ets1 deletion constructs in the presence versus absence of 1:1 TAZ1. Remaining amide signal intensity in the presence of TAZ1 is shown on the right, with dashed lines indicating rough estimates of the highest ratio point of the central region. ! !! "! #! $! %! &! '! (! )! !*! !!! !"! !#! ! !! "! #! $! %! &! '! (! )! !*! !!! !"! !#! * *+" *+$ *+& *+( !+* * *+" *+$ *+& *+( !+* ,-./01-. ,-./01-. ,2 3/4 ,2 3/4 5 555555 5 5 555555 5 6* 6! 6" 6"7 6# 6%6$ 8# ( 9$ ! ! !! "! #! $! %! &! '! (! )! !*! !!! !"! !#! ! !! "! #! $! %! &! '! (! )! !*! !!! !"! !#! * *+" *+$ *+& *+( !+* * *+" *+$ *+& *+( !+* ,-./01-. ,-./01-. ,2 3/4 ,2 3/4 5 555555 5 5 555555 5 ! !! "! #! $! %! &! '! (! )! !*! !!! !"! !#! ,-./01-. * *+" *+$ *+& *+( !+* ,2 3/4 :3.!%!;!#(<=<8>?! 5 555555 5 :3.!$";!#(<=<8>?! :3.!");!#(<=<8>?! :3.!!;!#(<=<8>?! "@;:3.!!;!#(<=<8>?! A>B ACB ADB AEB A:B F-G2/H/HI ./IH2J /H3-H./3K !* *L '* L %* L %* L "* L Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   106	
   In contrast, Ets151-138, containing only H1-H5 of the PNT domain, did not bind TAZ1 measurably by NMR  (Figure 3.6A). Thus the dynamic helix H0 is required for the TAZ interface.  The titrations of all the above Ets1 constructs in the presence of equimolar TAZ1 were compared in Figure 3.6. At the equimolar versus no TAZ1, the average 15N-HSQC peak intensities were ~ 100% for Ets151-138, ~ 70% for Ets142-138, ~ 50% for Ets129-138, ~ 50% for Ets11-138, and ~ 20% for 2P-Ets11-138. This indicated that TAZ1 recognition is dependent upon multiple determinants in Ets1, including primarily H0 and residues 29- 42, which include the phosphoacceptors. Importantly, similar patterns of amide intensity perturbations were observed, indicating common roles for helices H2 and H5 in all these constructs.  Conversely, Ets11-52, the N-terminal region missing most of the PNT domain, was tested for binding TAZ1. Not surprisingly, this construct showed predominantly random coil 1HN chemical shifts with a minor helical propensity for the H0 as indicated in their SSP values (Figure 3.7). Previous 15N relaxation measurements also demonstrated that Ets11-52 is disordered (Macauley et al., 2006). This indicates that helix H0 only forms in the context of the core PNT domain. Somewhat unexpectedly, when 15N-Ets11-52 was titrated with TAZ1, a weak interaction was observed via 15N-HSQC spectra as amide peaks showed small progressive chemical shift changes. This corresponds to binding in a fast exchange regime with an estimated Kd ~ 140 ± 18 µM from the average of 3 residues showing the largest chemical shift changes (Phe24, Leu37, Leu21) (Figure 3.7). Surprisingly, these residues are located in the center of the fragment, ~10-38, and Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   107	
    Figure 3.7 The PNT domain is also required for TAZ1 binding 15N-HSQC of Ets11-52 was monitored while titrating with TAZ1. (A) Secondary structure propensity (SSP) values demonstrate that this Ets1 fragment is disordered. Also shown are changes in amide chemical shifts from 0% TAZ1 to 800% TAZ1. (B) Overlaid HSQC spectra of Ets11-52 while adding TAZ1; 0% TAZ1 in red, 50% in orange, 100% in yellow, 200% in green, 400% in blue, and 800% in magenta. Phosphoacceptors, Thr38 and Ser41, and the position of helix H0 in Ets11-138are indicated at the top. ! " !#$ !% &! &" ##&$ #% '! '" '$ ()' !)& ()* +(, #* -' ! ! "& ./ 01 2. 34 56 /2 78 5. /3 9: 06 5;< <1 = *)" *)( %)" %)( !&" !&( !!" !+5;<<1= !" > 5;< <1 = ;?= ;@= ?1AB985,?C!5 D04382E058A5F86!!G"& (H "(H !((H &((H '((H *((H ( -- I ()# G()# G()" ( Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   108	
   are not the phosphoacceptors nor those that form helix H0 (Figure 3.7). The top 10 residues with most chemical shift changes include six hydrophobic (Leu11, Val19, Leu21, Ala32, Leu36, Leu37), two negatively charged (Glu17, Asp20), and one aromatic (Phe24) moiety. Based on these results we conclude that high affinity, phosphorylation-dependent binding of Ets1 to TAZ1 requires the presence of both the core PNT domain, helix H0, and Thr38/Ser41. However, in the absence of the core PNT domain, Ets11-52 weakly binds TAZ1 via an alternative interface involving residues ~10- 38. These interactions might also account for the qualitatively tighter binding of Ets11-138 versus Ets129-138 (Figure 3.5).  Ets1-binding interface on TAZ1 Complementary 15N-HSQC monitored titrations of the 15N-labeled TAZ1 domain were carried out with unlabeled Ets11-138 and 2P-Ets11-138 (Figure 3.8A). Again, a progressive loss of signal intensity from selected amides in the TAZ1 domain resulted upon addition of either Ets1 species, confirming specific binding. Also consistent with the higher affinity of TAZ1 for phosphorylated Ets1, similar, yet more pronounced, spectral changes occurred in the presence of an equimolar amount of 2P-Ets11-138 relative to Ets11-138. Mapping the residues showing the largest amide 15N-HSQC intensity changes on the structure of TAZ1 reveals that the binding interface encompasses an extended region, including primarily helices H1, H3, and H4 (Figures 3.8 and 3.9).       Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   109	
                   Figure 3.8 Ets1-binding interface on the TAZ1 domain (A) Overlaid 15N-HSQC spectra of ~0.1 mM 15N-labeled TAZ1 titrated with (left) Ets11-138 and (right) 2P-Ets11-138 in molar ratios of 1:0 (light blue), 1:0.66 (dark blue), and 1:1 (red). (B) Histograms of amide signal intensity at 66% and 100% equivalence relative to unbound TAZ1. (C) Residues showing >0.90 loss of signal intensity upon formation of a 1:1 complex with (left) Ets11-138 and (right) 2P-Ets11-138 are mapped onto a low energy member of the structural ensemble of TAZ1 (1u2n.pdb) (also see Figure 3.7). Increasing cyan sphere size corresponds to greater signal loss upon complex formation (intensity ratios: 0 < large sphere (labeled in (A)) < 0.05, 0.05 < small < 0.10). The Zn+2 ions of TAZ1 are shown as magenta balls. Missing data points correspond to prolines or residues with unassigned or overlapping NMR signals.  Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   110	
  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hapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   111	
       Figure 3.9 Identification of the CBP TAZ1 domain binding interface on Ets129-138 (Left) Ribbon diagrams of low energy members of the (A) Ets129-138 and (B) 2P-Ets129- 138 structural ensembles with residues showing < 0.15 relative 15N-HSQC intensity in the presence of 1:1 TAZ1 identified (see Figure 3.4). Increasing cyan sphere size corresponds to greater signal loss upon complex formation (intensity ratios: 0 < large sphere < 0.05, 0.05 < medium < 0.10, 0.1 < small < 0.15), and qualitatively defines the binding interface between the Ets1 and CBP fragments. (Center) Surface representations of Ets129-138 and 2P-Ets129-138 with residues showing > 0.85 loss of signal intensity upon titration (see Figure 3.4) identified by physicochemical properties (red, Asp, Glu; blue, Arg, Lys, His; yellow, neutral polar; green, hydrophobic). Residues examined by mutation (Table 2.1) are underlined. (Right) Electrostatic surfaces of Ets129-138 and 2P-Ets129-138, calculated with MolMol (default parameters with ‘simple charge’; red, negative; blue, positive). !"# $%" &'( )*+ ,'' -./ 0*# $*' !*1 2"+ 3##" 2** $'# -.( 4#+1 $%" $*' 5%/ !*1 ,'' -./ 6'. 3.. 7.' 8).% $'1 8$'# ,'* )*+ 9#+/ 2"+ !"# -"14%% :#+' 3#+. 3##" 4#+1 2** -#+" 7%# !"#$%&'()*+(,- ;# ;1 ;* ;+ ;+< ;' 5 = ;# ;1 ;* ;' ;+ ;+< ;. 5 = !.#$)/+%&'()*+(,- !"#$ !%&' "#$ %&' Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   112	
   Electrostatic Interactions and CBP/Ets1 binding The interaction surfaces identified by NMR titrations are net negatively-charged for Ets11-138 and net positively-charged for the TAZ1 domain (Figures 3.9 and 3.10), thus suggesting that binding is electrostatically-driven. To test this hypothesis, the interaction detected by NMR was challenged using buffers containing 20 to 500 mM NaCl. Consistent with a role for electrostatic forces, binding was disrupted with increasing ionic strength (Figure 3.11).      Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   113	
     Figure 3.10 Identification of the Ets1 binding interface on the CBP TAZ1 domain (A) Ribbon diagram of low energy member of the TAZ1 (1u2n.pdb) structural ensemble with residues showing < 0.10 relative 15N-HSQC intensity in the presence of 2P-Ets11- 138 identified (see Figure 3.8). Increasing cyan sphere size corresponds to greater signal loss upon complex formation (intensity ratios: 0 < large sphere < 0.05, 0.05 < small < 0.10), and qualitatively defines the binding interface between the Ets1 and CBP fragments. Zn+2 ions are shown as magenta balls. (B) Electrostatic surface of the TAZ1 domain, calculated with MolMol (default parameters with ‘simple charge’; red, negative; blue, positive). Surface representations of TAZ1 upon titration with (C) Ets11-138 and (D) 2P-Ets11-138 with residues showing > 0.90 loss of signal intensity (see Figure 3.8) identified by physicochemical properties (red, Asp, Glu; blue, Arg, Lys, His; yellow, neutral polar; green, hydrophobic). !"#" $#%& '#() *+ *, *# *" ! - !"# !$# !%#&'()*++,+-. !/#&'01,()*++,+-. .#(+ /#0+ 1#&, '#%0 '#%2 *#%, .#%( *#%" 3#2) ."+2 '"+, 4"+0 /"#, '#%0 *#%" *#%, '"+, .#(+ $#(( Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   114	
                  Figure 3.11 Binding of Ets11-138 and the TAZ1 domain is disrupted by high ionic strength Electrostatic interactions contribute to the Ets1/TAZ1 binding as higher salt concentration disrupts complex formation. (A) 15N-HSQC spectrum of free 15N-Ets11-138 (black) is overlaid with that of 15N-Ets11-138 in the presence of equimolar unlabeled TAZ1 (red), both in 50mM NaCl. Signals absent due to TAZ1 binding are labeled. (B) Same as (A) except with higher salt concentration (500mM; purple). Signals from the free protein are shown as green labels.  Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   115	
   !"# !$% !$# !!% !!# &'# ('% ('# )'% )'# !"# !$% !$# !!% !!# *%% *&$ +!$, -!$# ./# 0%$ -(( 1!$% .(& *!$$ *!#& 1") -%" 2/) 1", 1/& 3%! 4%# 5!$" 1!$& 5!!&6 7864&! *%% *&$ +!$, -!$# ./# 0%$ -(( 1!$% .(& *!$$ *!#& 1") -%" 2/) 1", 1/& 3%! 4%# 5!$" 1!$& 5!!&6 7864&! !96:;;<= !% > 6:; ;< = ?@A!!B!"(6CD@E6 6 6 6 F7603G!6H7C6AIH@6:%#<J6>IKH= 6 6 6 !##L603G!6H7C6AIH@6:%#<J6>IKH= ?@A!!B!"(6CD@E6 6 6 6 F7603G!6H7C6AIH@6:%#<J6>IKH= 6 6 6 !##L603G!6EDME6AIH@6:%##<J6>IKH= :3=6 :N=6 Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   116	
   3.2.3 Role of newly identified structural components in MAPK signaling The proposed conformational change of the PNT domain and the hypothesized role of electrostatic forces at the binding interface suggest that a large number of residues in Ets1 play a role in CBP binding. A subset of residues that showed NMR spectral changes due to phosphorylation (Figures 2.2 and 2.9) or TAZ1 association (Figure 3.2) were tested by the Graves' laboratory for a role in responsiveness to MAPK signaling in cell-based transcription assays (Table 3.2). Supporting the role of electrostatics in binding of Ets1 to CBP, the negatively-charged Asp119, Asp123, or Glu127, lying along helix H5 (Figures 2.4, 3.9, and 3.10), gave severe reductions when mutated to arginines, but not alanines. Asp123 and Glu127 also form salt bridges to Lys42 and Lys50, helping to position helix H0 in the closed conformation of unmodified Ets129-138 (Figure 2.4). Mutation of Leu49, which is buried in the H0-H5 interface of this closed state but exposed in the open conformation, to a glutamate or arginine abrogated MAPK induction. This suggests a role for Leu49 either in maintaining the closed conformation or as a contact residue in the open conformation. Finally, we interrogated Thr52, Phe53 and Ser54, which are located in the H0-H1 junction. Introduction of a positively-charged arginine at position 52 or removal of the aromatic sidechain of Phe53 eliminated superactivation. Again, mutant phenotypes could arise from disrupting the macromolecular interface in the open state or from a direct or indirect conformational perturbation of the closed state. In conclusion, these functional assays corroborate the role of key residues in the structural and mechanism model of the Ets1-CBP regulated interaction.   Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   117	
     Table 3.2 Residues of Ets1 functional in transcriptional superactivation Mutation a Structural environment b Superactivation c WT  2.25 ± 0.10 L49E H0 interior 0.96 ± 0.09 L49R " 1.00 ± 0.10 K50A H0, salt bridge to E127 1.71 ± 0.03 T52A C-terminus H0, exposed 2.51 ± 0.09 T52R " 1.06 ± 0.07 F53A H0/H1 junction, interior 1.02 ± 0.02 S54A N-terminus H1, exposed 1.54 ± 0.08 S89A H2/H2’ loop, exposed 1.79 ± 0.07 D119A H5, exposed 1.79 ± 0.09 D119R " 0.99 ± 0.08 D123A H5, salt bridge to K42 1.24 ± 0.15 D123R " 0.92 ± 0.05 D119R/D123R H5 0.73 ± 0.15 W126A H5, interior 2.90 ± 0.06 E127A H5, salt bridge to K50 2.26 ± 0.07 E127R " 1.05 ± 0.02  a Amino acid coordinates and helical assignments as in Figures 2.3 and 2.5 b Environment in the closed state of Ets129-138. c Superactivation = RLAWT or mut / RLAempty vector in the presence of CA-MEK1. Mutant phenotypes are reported as mean values for 3-6 independent experiments ± the standard error of the mean, except W126A and D123A, which had two replicates. Controls confirmed that mutant phenotypes did not results from changes in expression levels or phosphorylation state of Thr38 (not shown). Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   118	
   In vitro mutational studies of the TAZ1 recognition interface of the Ets1 PNT domain showed that TAZ1 binding could be disrupted without perturbing global folding, yet pushing the conformational equilibrium towards the open state. First,15N-HSQC spectra of 15N-labeled Ets129-138 with the point mutations L49R in helix H0 and D123R or E127R in helix H5 exhibited well-dispersed signals, suggesting that the protein remains stably folded. This demonstrates that disrupted binding is not an indirect consequence of the mutations perturbing the global fold of the PNT domain. The amide chemical shifts of Gly55 in these three mutants are co-linear with those in wildtype, other deletion mutants, and phosphorylated species discussed previously in chapter 2 (Figure 3.12D and E). This indicates that the mutations, which lie along the interface between helices H0 and H1, shift the population distribution of the PNT domain towards the open state. Although shifting the PNT domain towards the open state, each of the three mutants also weakens TAZ1 binding, presumably due to the introduction of an unfavorable positive charge at the complex interface. The weakening of binding is shown by the average reduced intensity of four amides (Gly55, Ser89, Gly122, and Leu125) upon titration to a 1:1 molar ratio with labeled TAZ1 (Figure 3.12E).  In sum, these disrupted phenotypes, in combination with controls demonstrating proper folding, in vivo expression and phosphorylation, as well as reduced TAZ1 binding in vitro, indicate that these residues play a role in complex formation, either directly at the interface or indirectly via helix H0 packing. Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   119	
    Figure 3.12 In vitro mutational analysis on the Ets1-CBP binding Mutation of the TAZ recognition interface of the Ets1 PNT domain disrupts TAZ1 binding without perturbing global folding. (A-C) 15N-HSQC spectra of 15N-labeled Ets129- 138 with the point mutations L49R (cyan) in helix H0 and D123R (dark orange) or E127R (light orange) in helix H5. (D) A small portion of the overlaid spectra of these three mutants, along with various deletion fragments and phosphorylated proteins, color coded according to the cartoon of panel (E). The amide chemical shifts of Gly55 in these three mutants are co-linear with the species discussed previously in Figure 2.13. (F) The weakening binding is shown by the average reduced intensity of four amides (Gly55, Ser89, Gly122, and Leu125) of various Ets1 constructs and mutants (E) upon titration to a 1:1 molar ratio with unlabeled TAZ1. Note that the intensity losses for 2P- Ets11-138 (magenta), Ets129-138 (green), and Ets142-138 (blue) reflect the relative affinities of these species for TAZ1. In contrast, Ets152-138 (red) does not measurably bind TAZ1 due to the deletion of all residues forming helix H0. !"# !$# !!# !## %# &# '# $# # () *+ ), (*- ./0 ,, .12 3 !4 5 .16 67 3 894 %94 :94 PP 29 -1 38 42 -1 38 51 -1 38 1- 13 8 2P  1 -1 38 ;# ;! ;$ ;" ;' ;4;$< ‡ ‡‡ 1=3 1>3 1?3 1@3 1A3 !;.16673 894 %94 :94 894 %94 :94 89' 89" 89$ 89! 89# %98 !!! !!# !#8 !#% !;.16673 !4 5 .16 67 3 1B3C/-!#$ C/-44 B*,!$8D!"%DE'8F B*,!$8D!"%D@!$"F B*,!$8D!"%DB!$:F $8 D!" % '$ D!" % 4! D!" % $G .!D !" % E' 8F @! $" F B! $: F E' 8 @ !$ " B! $: !"#$ %&!'#( C/-44 C/-44C/-44 Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   120	
   3.3 Discussion In chapters 2 and 3, the structure and dynamics of the Ets1 PNT domain were dissected to understand the molecular bases for the phosphorylation-enhanced binding of Ets1-CBP. In chapter 2, this mechanistic investigation was built on the refined structure of Ets129-138, in which two helices, H1 and the dynamic H0, tether a flexible region containing the phosphoacceptors Thr38 and Ser41 to the core PNT domain. Phosphorylation of these residues shifts the conformational equilibrium of the helix H0 to an ‘open’ state by displacing H0 from the core PNT domain, and thereby exposing the negatively charged and hydrophobic interface on helices H2 and H5. Subsequently, in this chapter, by a combination of pull-down binding assays, ITC measurements, and NMR-monitored titrations, the PNT domain and adjacent phosphoacceptors were shown to interact with the TAZ1 domain. The constitutive binding of Ets11-138 to TAZ1 is enhanced ~34 fold upon ERK2-dependent phosphorylation. Deletion and phosphorylation studies showed that helix H0 and the phosphoacceptors Thr38/Ser41 are key determinants of TAZ1 binding. However, neither these residues (Ets11-52) nor the core PNT domain alone (Ets151-138) was capable of reproducing the binding of the longer PNT domain constructs (Ets129-138 and Ets11-138) with TAZ1. Thus binding is dependent upon cooperative (or "multi-valent") interactions with both the phosphoaccetors and an interface involving helices H0, H2, and H5 of the PNT domain.  3.3.1 Roles of phosphorylation in regulating Ets1-CBP interaction Phosphorylation of Ets1 by MAPK acts in two ways to increase the binding affinity of Ets1 for CBP (Figure 3.13). First, addition of phosphate shifts a conformational Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   121	
           Figure 3.13 Model of the phosphorylation-enhanced interaction of Ets1 and CBP The PNT domain exists in a conformational equilibrium with the dynamic helix H0 in closed and open states (as well as being frayed or fully unfolded; not shown). Phosphorylation shifts the population distribution to the open state, which is favored for TAZ1 binding via complementary electrostatic interactions.  !! - -- -- + + + + + - -- -- !! !!! ! !! !! !! !! !! !! - -- -- "! #$! "$%&'( #$%&'( !)% *!#!)% !)% %&'(+( +, +( +, +( +, +*#+- +*#+- +*#+- ++ - - -- - -- - - - - --- - - - Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   122	
   equilibrium towards a proposed binding-competent open state. Second, phosphates add negative charge to electrostatic complementarity of the Ets1-CBP interface that includes both the core PNT domain and the appended regulatory helices. The general principles of this mechanism illustrate a route for the evolution of signaling-dependent regulation.  Phosphorylation-enhanced binding relies on two flexible structural components that contrast with the highly structured, conserved core PNT domain. First, the phosphoacceptors, Thr38 and Ser41, are conformationally disordered regardless of their phosphorylation states. Such flexibility may be necessary for accessibility to kinases and phosphatases, as well as CBP. The second major dynamic component is the marginally-stable helix H0 that undergoes substantial conformational fluctuation among locally frayed or unfolded state(s), as described in chapter 2. Phosphorylation of Ets129-138 leads to a loss of detectable long-range NOE restraints involving residues 42- 52 in H0 to H2 and H5, and thus helix H0 appears to adopt a broad distribution of positions displaced from the core PNT domain in the calculated structural ensemble of 2P-Ets129-138. The energetic cost of this conformational change could account in part for the unfavorable endothermic (ΔH > 0) association of Ets11-138 and the TAZ1 domain measured by ITC. The conformational equilibrium might also cause the absence of detectable 15N-HSQC signals in the Ets11-138/TAZ1 domain complex, as used in mapping the binding interface (Figures 3.2A and 3.8A). It is possible that the conformational exchange between free and bound states or within the bound state occurring on an unfavorable timescale relative to the chemical shift differences between these states can lead to severe line-width broadening and, therefore, the loss of Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   123	
   detectable signals, particularly in context of the increased molecular mass (Δmass ~ 11 kDa) of the complex.  The second role of phosphorylation is to add to the electrostatic complementarity of the Ets1 and CBP interface. Several lines of evidence support the predominant role of electrostatic forces in driving the binding event. Based on the patterns of intensity changes in NMR-monitored titrations, the complex interface is formed by a net negative- charged surface of the PNT domain of Ets1 (and Ets2; not shown) and a positive- charged region of TAZ1. Phosphorylation of Thr38 and Ser41 increases the negative charge of the PNT domain, thus reinforcing a complimentary electrostatic binding interface for TAZ1. This binding mechanism is supported by mutational analysis, in which substitution of arginines in Ets1 at the CBP interface disrupted both ERK2- dependent transcriptional activation and the ionic strength sensitivity of the Ets1-CBP interaction. We also envisage a possible electrostatic effect in which displacement of monovalent counterions from the highly charged surfaces of Ets11-138 and TAZ1 upon complex formation contributes to the net increase in entropy (ΔS > 0) detected by ITC measurements. It is noteworthy that displacement of helix H0 exposes several hydrophobic residues at the convergence of helices H1, H2, and H5 which also contribute towards the TAZ1 interface as defined by 15N-HSQC signal intensity changes (Figures 3.9 and 3.10). A hydrophobic component of binding would also partly explain the observed favorable entropy changes. In sum, we propose that MAPK phosphorylation promotes CBP binding to Ets1 via electrostatic effects combined with a conformational transition involving dynamic elements of the PNT domain. Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   124	
   3.3.2 Distinct features of Ets/CBP binding Aspects of the CBP/Ets1 interaction are reminiscent of documented CBP binding modes. Previous structural studies demonstrated that disordered polypeptides corresponding to the negatively-charged transactivation domains of several transcription factors, including HIF1α, CITED2, STAT1/2 and p53, fold as induced helices, wrapping around the extended binding clefts of the TAZ1 and TAZ2 domains (Dames et al., 2002; Freedman et al., 2002; Freedman et al., 2003; De Guzman et al., 2004; Feng et al., 2009; Wojciak et al., 2009). Likewise, the CREB KID domain, as well as the activation domains of c-Myb, MLL and p53, undergo a folding transition upon binding to the KIX domain of CBP (Radhakrishnan et al., 1997; De Guzman et al., 2006; Teufel et al., 2007; Lee et al., 2009). The flexibility of the Ets1 PNT domain and our proposed conformational equilibrium model exhibits a clear similarity to these examples. Furthermore, the TAZ1 residues mapped as potential contacts for Ets1 overlap with those involved in binding to these other transcriptions factors. The ability of a single domain, such as TAZ1, to function in the recognition of numerous partners may require this flexibility. However, the involvement of a pre-structured module (the core PNT domain) in combination with a flexible, yet structured region (helix H0) is distinctive to Ets1. Furthermore, the dynamic features of the Ets1 PNT domain are relatively complex, indicative of a multi-component mechanism for regulated phosphorylation- dependent binding to CBP.  Earlier binding assays with full-length CBP and full-length Ets1 suggested a higher affinity than observed for Ets11-138 and the isolated TAZ1 domain (Foulds et al., 2004). Signaling induced binding of phosphorylated Ets1 via TAZ1 could be enhanced by Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   125	
   additional co-operative interactions with CBP. For example, a region of Ets1 C-terminal to the PNT domain has also been mapped as an interface for CBP in constitutive binding assays (Yang et al., 1998a). In native biological contexts, Ets1 binding to Ras- responsive promoter elements is accompanied by a partner transcription factor such as Pit1 and AP1 (Stacey et al., 1995; Dwivedi et al., 2000; Duval et al., 2003) or a second molecule of Ets1 (Buttice and Kurkinen, 1993). Multiple transcription factors on a promoter or enhancer can often display cooperative binding that stabilizes a resulting complex. If more than one transcription factor binds CBP through different domains, these multi-protein complexes could also facilitate synergistic CBP binding. Thus, the PNT domain/TAZ1 interface may nucleate a network of other contributing interactions.  A focus of this study was the regulation of CBP/Ets1 binding by phosphorylation. There is a variety of other transcription factors whose binding to CBP is (or is not) regulated by phosphorylation. However, in only a few cases has the role of phosphates at the interface been investigated structurally or thermodynamically. The cAMP signaling pathway phosphorylates the KID domain of CREB to induce binding to the KIX domain of CBP (Radhakrishnan et al., 1997). Phosphorylation of p53 enhances the binding of the transactivation domain to TAZ2. The 1.7 µM dissociation constant of the phosphorylated Ets11-138 for TAZ1 is similar to KIX binding to phosphorylated KID (Kd = 10 µM) (Matsuno et al., 2004) and the phosphorylated p53 binding to TAZ2 (Kd = 17.6 µM) (Feng et al., 2009). However, these phosphorylation-regulated interactions are relatively weak in comparison with HIF1α binding to TAZ1 with Kd = 10 nM (De Guzman et al., 2004). It may be that macromolecular interfaces designed to be highly regulated Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   126	
   by phosphorylation function best through weaker interactions that enable responsiveness to pathways that down regulate activity.  3.3.3 ETS family PNT domains and cellular signaling Throughout chapters 2 and 3, the role of appended helices, H0 and H1, of Ets1 PNT was established. These helices are also found in the PNT domain of Ets2, and likely in the Drosophila ortholog Pnt-P2 (Figure 2.1). These structural elements are functionally linked to the phosphoacceptor sites that lie at a conserved distance N-terminal to the the core PNT domain. An ERK2 docking site is located on the Ets1 and Ets2 PNT domains, namely the exposed Phe120 and Phe154, respectively, in helix H5 (Figure 2.5), which facilitates phosphorylation through a reduction in the Km value for the phospho-acceptor-kinase interaction (Seidel and Graves, 2002; Waas and Dalby, 2002). The conformational flexibility of helix H0 may contribute to the accessibility of Thr38 and Ser41 to the catalytic site of ERK2 while the PNT domain is docked at an ancillary site on the enzyme. Helix H0 and, intriguingly, residues near the docking site also interact with CBP. Again, the flexibility of helix H0 appears critical for enhancement of TAZ1 binding via phosphorylation. We conclude that the conservation in spacing between the core PNT domain and phosphoacceptor sites is likely explained by the role of helix H0 in both CBP binding and potentially ERK2 docking.  Structural divergence of the PNT domain provides functional diversity to ETS family members. The PNT domain of GABPα, which binds TAZ2, but not TAZ1, contains only the analogous helix H1 and shows no regulation by phosphorylation (Kang et al., 2008). However, this ETS protein has a structured OST domain that also contributes to CBP Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   127	
   binding via TAZ1 and TAZ2 (Kang et al., 2008). In the cases of vertebrate Tel and Drosophila Yan, the core PNT domain displays homopolymerization, which is implicated in transcriptional repression (Kim et al., 2001; Qiao et al., 2004). The added elements of helix H0 and H1 block the potential for polymerization of Ets1 and other related ETS proteins with PNT domains (Mackereth et al., 2004), as described in chapter 2.  The Drosophila ortholog of Ets1, Pnt-P2 is a transcriptional activator for the genes related to eye development in Drosophila. Qiao et al have shown that the SAM domain of Mae competes with the MAP kinase Rolled (the ortholog of ERK2) for the docking site in Pnt-P2 to repress the subsequent gene expression (Qiao et al., 2004; Qiao et al., 2006). Although the biochemical basis of this interaction is not well established yet, the sequence similarity of Pnt-P2 and Ets1 in the PNT domains, including the N-terminal helices H0, H1, and the phosphoacceptor sites, suggest that phosphorylation of the Pnt- P2 by the Rolled kinase may also affect the dynamics of the H0 to regulate its interaction with other partners, possibly CBP/p300, as seen for Ets1 PNT. It is also notable how phosphorylation might affect the binding affinity of the Mae domain for tPnt- P2. When the modeled Pnt-P2 structure (generated by SWISS-MODEL using the Ets1 PNT domain as a template), was overlaid on the transcriptional repressor Yan, of the Yan/Mae complex structure, H0 of Pnt-P2 would clearly hinder binding of Mae (Figure 3.14). This suggests that a phosphorylation-induced conformational change of H0 may alter the binding of Mae. Thus, the broader ETS family illustrates additional ways in which extensions of the core PNT domain provide routes to biological regulation. Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   128	
           Figure 3.14 Position of H0 regulates the function of other PNT/SAM domains Modeled Pnt-P2 structure (red) based on the Ets1 PNT structure (2JV1.pdb) is overlaid on the Yan (green) structure of the Yan (green) /Mae (cyan) complex. Some helices are labeled for Pnt-P2.  !"#$%$#&'()*'+ ,- ,. ,+ ,/ 01(2!1$& 3"45%$6 Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   129	
   3.4 Conclusions By exploring the molecular detail of a regulated macromolecular interaction, we have uncovered the mechanism by which Ras/MAPK-dependent signaling impacts function of Ets1 as a transcriptional activator. In addition, we have shown that the dynamic character of proteins such as Ets1 is critical in mediating biological regulation, especially via post-translational modifications. Finally, the divergence of the PNT domain within the ETS family via appended dynamic elements is noted as a route to develop regulatory potential and functional uniqueness within a family of related proteins. Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   130	
   3.5 Materials and methods 3.5.1 Protein purification See section 2.4.2 for Ets1/2 purification. The gene encoding TAZ1 (CBP340-439) was cloned from the murine CBP gene (Genbank 70995311) into the pET28b (Novagen) vector for expression as a His6-tagged protein in E. coli BL21(λDE3) cells. The protein expression was induced by adding 1 mM IPTG and incubated at 16 °C overnight with an additional ZnSO4 (150uM) in the M9T media. The cells were harvested by centrifugation for 15 minutes at 5K (rpm). Then, cell pellets were resuspended in buffer A (150 µM ZnSO4, 5 mM imidazole, 50 mM Tris, pH 7.5, 500 mM NaCl, and 5% glycerol) and lysed by passing through a French press. After centrifugation at 15K (rpm) for 60 minutes, TAZ1 was purified by running the lysate supernatant through a HisTrap HP column (GE Healthcare) with a imidazole gradient (5-500mM) using buffer A. The purified sample was treated with thrombin to remove the His6-Tag and concentrated using an Amicon in the final buffer 20 mM Tris, pH 6.9, 50 mM NaCl, 2 mM DTT  (for more details, see (Kang et al., 2008)).  3.5.2 Isothermal titration calorimetry ITC experiments were conducted on a VP-ITC Microcalorimeter, MicroCal LLC. To block reactive cysteines, Ets11-138 and 2P-Ets11-138 were treated with 50 mM iodoacetamide at pH 8.0 and 25 oC for 1 hr, followed by quenching with 150 mM DTT at 25 oC for 30 min. Mass spectrometry (ESI-MS Quatro-II) revealed three acetamide modifications, on four possible cysteines, in both species. The TAZ1 domain and the modified Ets11-138 samples were equilibrated into 25 mM Tris pH 7.9, 50 mM KCl by dialysis. After equilibration of the VP-ITC to 15.0 oC for each experiment, 7.5 µL aliquots Chapter 3 – Phosphorylation-dependent Ets1/CBP interaction 	
   131	
   of 500 µM TAZ1 were titrated into 1.4 mL of 10 µM non-phosphorylated or dual phosphorylated Ets11-138. Data were analyzed with Origin 7.0 MicroCal LLC software.  3.5.3 NMR spectroscopy NMR spectra of the Ets1/2 and CBP constructs were recorded using Varian 500 MHz Unity and 600 MHz Inova spectrometers, and analyzed using NMRpipe (Delaglio et al., 1995) and Sparky	
   (Goddard). 1H, 13C, and 15N resonance assignments were obtained via standard heteronuclear correlation experiments. NMR-monitored titrations were carried out by recording sensitivity enhanced 15N-HSQC spectra of ~0.1 mM 15N-labeled protein in 20 mM Tris pH 7.0, 20 mM NaCl, 2 mM DTT, and ~10 % D2O at 25 oC, to which the unlabeled protein partner (~ 1.2 - 2.8 mM stock solution in the same buffer) was added in small aliquots.  Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   132	
   Chapter 4 - Discovery and characterization of a TAZ1 autoinhibitory region in CBP  	
   In chapters 2 and 3, post-translational modification of the transcription factor Ets1 was shown to regulate its binding to the transcriptional co-activator CBP. The Ets1 PNT domain employs phosphorylation as a switch to increase its affinity for the CBP TAZ1 domain and thereby enhances expression of Ras-responsive genes. Conversely, this chapter describes the discovery and dissection of a potential autoregulatory module localized in CBP itself. The serendipitous discovery of the intramolecular interaction of the TAZ1 domain with the N-terminally located nuclear receptor interaction domain (NRID) of CBP suggested an auto-inhibitory mechanism in which the NRID regulates interactions with transcription factors by competing for a similar binding interface on the TAZ1 domain. This hypothesis was supported by measurements showing decreased binding affinity of the TAZ1 domain for hypoxia inducible factor (HIF) in the presence of the NRID. This chapter focuses on the identification of the minimal TAZ1 domain- binding NRID sequence and characterization of the interactions of these two CBP modules using NMR spectroscopy.          Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   133	
   4.1 Introduction 4.1.1 The N-terminal region of CBP CBP and p300 have been reported to interact with numerous nuclear receptors through the most N-terminally located nuclear receptor interaction domain (NRID). The exact boundaries of the NRID are loosely defined as they vary depending on the specific binding partner. However, several recent studies have demonstrated that residues 1- 101 of CBP mediate the interaction with most nuclear receptors, including peroxisome proliferator-activated receptor (PPAR), retinoic acid receptor (RAR), retinoid X receptor (RXR), estrogen receptor (ER) and thyroid-hormone receptor (TR)  (Chakravarti et al., 1996; Kamei et al., 1996; Heery et al., 1997; Klein et al., 2005).  As commonly observed for many nuclear receptor binding proteins (see chapter 1), the NRID region contains one of the three nuclear receptor (NR) boxes, 69LSELL73, of CBP. The NR box (LXXLL motif) was first identified in the course of studying several transcriptional co-activators (RIP-140, SRC-1, CBP) binding the LBD (ligand binding domain) of the ER and RAR nuclear receptors (Heery et al., 1997). Mutation of the core sequence, especially the leucines in the LXXLL motif, was reported to disrupt ligand- induced nuclear receptor binding (Heery et al., 1997). More interestingly, the N- and C- terminal flanking residues (Heery et al., 2001) or residues even further away from the motif (Klein et al., 2005) are critical for providing specificity in selecting binding partners. This further defines the core NR box as 8 amino acids, including two N-terminal residues and one C-terminal residue flanking the LXXLL motif, respectively (Heery et al., 2001). For example, the presence of a hydrophobic residue, especially Ile or Leu, preceding the first conserved Leu (LXXLL, underlined) in the NR box is critical for Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   134	
   binding the LBD of the steroid receptor coactivator and 140 kDa receptor interacting protein with high affinity (Heery et al., 2001), thereby possibly explaining a relatively weaker binding of CBP with the nuclear receptors due to the presence of Gln at this position. Furthermore, structural studies have shown that the LXXLL motif forms an induced α-helical conformation upon binding the hydrophobic groove in the LBD of the nuclear receptor by a ‘charge clamp’ mechanism using the charged residues at the ends of the motif (Figure 4.1) (Nolte et al., 1998; Shiau et al., 1998).  The LXXLL motif has also been reported to mediate interactions with non-nuclear receptor proteins. For example, a LXXLL motif in the C-terminal transactivation domain of signal transducer and activator of transcription factor 6 (STAT6) binds a PAS-B domain of the co-activator NCoA-1 (Litterst and Pfitzner, 2002). The KIX domain of CBP was also shown to interact with a LXXLL motif in the transactivation domain of c-Myb. However, none of the interaction of LXXLL motifs with non-nuclear receptor proteins utilized the ‘charge clamp’. Instead, the polar residue preceding the first conserved leucine in the motif was involved in binding non-nuclear receptor proteins.  In addition, the NRID domain of CBP also contains two other characteristic sequences, 40LPDEL44 and 50LSLL53. The 40LPDEL44 resembles the LPQ/EL motif found in TAZ1- binding domains of HIF-1α (hypoxia inducible factor 1 alpha) and CITED2 (CBP/p300- interacting transactivator with ED-rich tail 2), except for the additional aspartate at position 42. Using NMR spectroscopy, Klein et al. have shown that the nuclear receptors PPARγ and RXRα interact with both the NR box and the 40LPDEL44 Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   135	
       Figure 4.1 Charge clamps for binding LXXLL motif In the crystal structure of the nuclear receptor PPAR-γ (white), Lys301 (blue) and Glu471 (red) act as a charge clamp to interact with the LXXLL motif (green sidechain) of the co-activator SRC-1 fragment (green). (PDB ID: 2prg)  !"#$%& '()*+& ,,-.a /.01& Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   136	
   (Klein et al., 2005). The 50LSLL53 resembles the LXXLL motif except for missing an amino acid between the first and second leucines. Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   137	
   4.1.2 Biological and structural features of HIF-1α Hypoxia-inducible factor (HIF) is a heterodimeric transcription factor, initially discovered for binding a hypoxia response element in the erythropoietin gene (reviewed by (Kaelin, 2005)).  The complex is composed of a proteolytically labile α-subunit and a stable β- subunit, which together form the DNA-binding bHLH domain and the signal-sensing PAS domain (Figure 4.2A).  Additionally, the α-subunit contains two C-terminally located transactivational domains, named NTAD and CTAD.  Two key determinants for regulating the expression of hypoxia-inducible genes are mainly involved with the α-subunit (Figure 4.2B). In the presence of oxygen, HIFα becomes hydroxylated at one of the two potential prolines in the NTAD domain by the egg-laying-defective nine (EGLN) family protein. The recognition of the hydroxylated prolines by the von Hippel-Lindau tumor suppressor leads HIFα to pVHL-dependent polyubiquitination and proteosomal degradation. Secondly, Asn803 in the CTAD is also subject to hydroxylation by factor-inhibiting HIF (FIH).  However, in this case, the hydroxylation interrupts recruitment of the transcriptional co-activator CBP/p300, resulting in attenuated gene activation. In contrast, under hypoxic condition, the absence of hydroxylation leads to the accumulation of HIFα and its uninhibited interaction with co-activators, thus resulting in the full activation of hypoxia-related genes. Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   138	
     Figure 4.2 Regulation of hypoxia inducible factors (A) Cartoons of the domain structures of HIFα and HIFβ. Hydroxylation sites (Pro, Asn) in the N-terminal transactivation (NTAD) and C-terminal transactivation (CTAD) domains are indicated. Both subunits contain basic helix-loop-helix (bHLH) and Per/Arnt/Sim(PAS) domains. (B) Mechanism for regulation of the HIF stability and HIF- mediated transcription. See text for details. !"#` $!%! &'( )*'+ ,*'+ $!%! &'( !"#_ !"#` &-. &-. '/0 !"#_&-. '/0 !"#_!12&-. !12'/0 !"#_&-. !12'/0 !"#_!12&-. !12'/0 !"#_ !"#`!"#_ ,3&45677 58!% 9$ 9$ 9$ 9$ 9$ +:;-<=<>?.0 #"! @A%)B .C D; :0 E:// F.-: G'H G3H Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   139	
   Several previous studies have documented the interaction of HIFα and CBP (Arany et al., 1996; Kallio et al., 1998).  Biochemical investigations have unveiled that the CTAD of HIF-1α interacts specifically with the TAZ1 domain of CBP/p300 to enhance the expression of hypoxia-related genes (Arany et al., 1996; Kallio et al., 1998; Dames et al., 2002; Freedman et al., 2002). Numerous transcription factors, including Ets1/2 as described in chapter 3, also interact with TAZ1 domain to recruit CBP to their respective promoter regions. Structural details of the HIF-1α/TAZ1 complex have been reported for CBP and p300 (Dames et al., 2002; Freedman et al., 2002). In both cases, the CTAD domain of HIF-1α undergoes a disorder-to-order transition, wrapping around the TAZ1 domain through mostly hydrophobic contacts. Similar interactions are seen with the transactivation domains of CITED2 and STAT2 bound to the TAZ1 and TAZ2 domains (Freedman et al., 2003; De Guzman et al., 2004; Wojciak et al., 2009).  4.1.2 Discovery of the intramolecular NRID-TAZ1 domain-interaction in CBP. In parallel with the study of the NRID domain in CBP described in this chapter, Dr. Mary Nelson in the Dr. Graves’ lab detected a NRID-TAZ1 interaction using GST-tag pulldown and limited proteolysis measurements. The intramolecular interaction of NRID- TAZ1 was recognized initially by identifying a prominent ~50 kDa N-terminal degradation product of CBP that co-purified with full-length CBP expressed in Sf9 cells (Figure 4.3A). Various GST-tagged CBP fragments were introduced to map the intramolecular interaction more precisely. In Figure 4.3B, the lack of binding to CBPSf9 1- 675 is attributed to a more favorable intramolecular NRID-TAZ1 domain interaction in this species. CBPTAZ1 only interacts with the NRID-containing CBPSf9 1-675. The constructs containing the NRID and TAZ1 regions (CBP1-452 and CBP1-675) can interact with CBPSf9 Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   140	
   1-675 also containing both of these regions. This approach localized the interaction of the NRID region, including residues 1-356, with the TAZ1 domain, residues 340-439 (Figure 4.2B). Analytical ultracentrifugation analysis of the CBP N-terminal construct demonstrated that this region behaves as a monomer (Nelson, 2007). Last, limited proteolysis was performed on two CBP constructs, one containing both the NRID and TAZ1 domains (CBP1-675) and the other missing the NRID (CBP102-675). Increased resistance of the TAZ1 domain towards proteolytic degradation was observed with CBP1-675, suggesting that the NRID interacted with and thereby stabilized the TAZ1 domain. Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   141	
   Figure 4.3 Detection of the intramolecular NRID-TAZ1 domain-interaction in CBP (A) An N-terminal fragment of CBP co-purifies with full-length protein. Sequential purification of Sf9-expressed CBP  (N-terminal His-tagged and C-terminal HA-tagged) using IMAC and anti-HA antibody affinity methods reveals the presence of a co- purifying 50 kDa N-terminal degradation product (arrow). The proteins are visualized on a silver stain gel (left panel) and by western blot using the N-terminal CBP-specific (A22) antibody (right panel). His-purified proteins were loaded on HA-antibody beads (Load). Unbound proteins were removed from the beads (FT) and washed (W) with buffer containing 0.5, 1, and 2M NaCl. Proteins were eluted (Elute) from the HA beads using 0.1M Glycine pH 2.5. (B) The interaction of NRID and TAZ1 is shown by GST pull-down assays. 200 pmol of the indicated GST-CBP fragments were incubated with a mixture of 5 µM dual tagged His-CBPSf9 1-675-HA protein and two degradation products (F1 and F2) as seen in (A). Approximate size of F1 and F2 are depicted with solid and dashed lines. After 4-20 hours, GST-beads were used to pull-down the bound species. The resulting proteins were separated on 4-15% SDS PAGE (BioRad) and visualized by western blotting using an anti-HA antibody to detect species containing the C-terminal HA-tag. CBP1-356 only binds fragments containing the TAZ1 domains (F1 and F2). !" #" $" %# &# '( )* +, '- ". #/ '0 123 45 '- '6 / '- '% / '( )* +, 23 45 %%" 6$" !" #" 7" 8" 9: ;* + <= >' 6? &# $ =, 45 @ AB 1 <= >' 1C D6 <= >' 6? E# % <= >' 6? $7 # <= >' F9 G 6 1CD6 F9G $7# ?HC 06 0% HI@? E&8&E" <=>'BJ8'6?$7# 06 0% KCL K=L MN9O Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   142	
   4.1.3 Significance of the study The aim of this study was to characterize the newly identified intramolecular interaction of the NRID and TAZ1 domains of CBP. We hypothesized that this constitutes an auto- regulatory mechanism by which the NRID domain competitively inhibits TAZ1 domain binding of numerous transcription factors. Despite extensive studies, it is still unclear how the promiscuous interaction of CBP/p300 with over 300 different transcription related partners is regulated.  To date, most studies have focused on mechanisms, such as post-translational modifications (i.e. phosphorylation or acetylation) of the partner proteins of CBP, as exemplified in chapter 3 for Ets1. Although detailed studies have been carried out on different parts of CBP/p300, including many with the TAZ1 and TAZ2 domains, no structure has been reported for the NRID region.  Identification of a cis-regulatory element in CBP/p300 and elucidation of the structural details of intramolecular complex formation would provide an alternative perspective on how multi-domain transcriptional co-activators help control gene expression. Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   143	
   4.2 Results 4.2.1 Discovery of the intra-molecular interaction in CBP NMR-monitored titrations were use to confirm the co-purification and pulldown experiments that identified the NRID-TAZ1 domain interaction. During this titration, the backbone amide resonances of uniformly 15N-labeled TAZ1 were monitored with 1H-15N HSQC spectra while adding unlabeled CBP1-356, which contains the NRID domain. The amide resonances of TAZ1 progressively shifted upon the gradual addition of CBP1-356, indicative of binding with relatively weak affinity (Kd > µM; fast exchange limit) (Figure 4.4A). Fitting of the resulting titration curves for the amide of Lys351 and the indole of Trp418 in the TAZ1 domain versus added CBP1-356 yielded an estimated Kd of ~ 30 ± 20 µM (not shown). This value is only an estimate due to the absence of any other aromatic residues in CBP1-356, thus precluding accurate measurement of its concentration. When the residues with the largest chemical shift changes were mapped on the structure of TAZ1, an extended binding interface involving mainly helices H1 and H3 was identified (Figure 4.4B). This interface overlaps that seen with other TAZ1- binding proteins. The reverse titration of 15N-labeled CBP1-356 with unlabeled TAZ1 was not performed as CBP1-356 is large and likely predominantly unstructured. Thus it would certainly yield poor quality NMR spectra.  4.2.2 Characterization of the TAZ1-binding component in the N-terminal region of CBP To identify the TAZ1-binding sequence in the NRID, three deletion fragments (CBP1-101, CBP28-82, CBP28-57) were characterized by NMR spectroscopy (Figure 4.5A). CBP1-101 was chosen as this region had previously been reported to interact with various nuclear Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   144	
   receptors (Kamei et al., 1996). The two other constructs, CBP28-82 and CBP28-57, were generated based on the results of titrations of CBP1-101 with the TAZ1 domain (see below).  The 1H-15N HSQC spectra of each fragment were carefully assigned, compared, and analyzed for secondary structure prediction and backbone dynamics, then further used for analyzing HSQC-monitored titrations with the TAZ1 domain. Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   145	
    Figure 4.4 NMR spectroscopic confirmation that CBP1-356 binds the TAZ1 domain. (A) The 1H-15N HSQC spectrum of the 15N-labeled TAZ1 domain was monitored while adding aliquots of non-labeled CBP1-356. The 10 residues showing the largest amide or indole shift changes are labeled on the spectrum. Estimates of the relative CBP1-356:TAZ domain molar ratios at each point are 0 (red) (0), 0.25 (yellow), 0.50 (green), 1.0 (blue), and 1.34 (magenta). (B) The 20 residues of TAZ1 showing the largest spectral perturbations upon titration with CBP1-356 are identified in red sticks on the cartoon of the unbound TAZ1 structure (PDB ID:1u2n). !" # $ % !&" !'( !'" !!( !!" )*!" +*!# +&(! ,&-& .*!$/0¡!1 2*'# 3*!% 4&#( 2&(# 3&-' !( 5 /06 67 1 !3/06671 0,1 8/9:;!<&(- !"#$ 0:1 ,&%' =&-$ )*!" >&*- +&(!?&(* @*'$ @&($ 3&-' 2&(# 2&-! ?&%" 9*'# ,&%' 4&#( =&-$ +*!# .*!$ ?&%" 2&-! 9*'! %! %& %' %! %" %' %& Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   146	
   Figure 4.5 N-terminal CBP constructs used to map the TAZ1-binding site (A) TAZ1 and four N-terminal constructs of CBP used for the identification of the minimal TAZ1-binding region. (B) Sequences of CBP1-101 and CBP28-82 with the LPQ/EL- like (green), the LXXLL-like (blue) and the LXXLL (red) motifs. Signals from the underlined residues were not assigned in the 1H-15N HSQC spectra of the constructs. Residues with large chemical shift changes (Fig 4.8A) upon binding TAZ1 are italicized.  !"# $%& %$' ( $)* ( (&( +, ,+ +, )- ./0(1$)* ./0(1(&( ./0+,1,+ ./0+,1)- 2"3( 4 4 4 4 4 4 4 24 !" # $4 54 64 #" !4 64 %" &" !" '" 04 74 %4 64 (4 04 84 94 :4 6" 54 64 '4 84 54 94 84 64 ;4 04 74 "4 "4 54 <4 =4 <4 >4 64 54 %4 64 '" )4 94 94 54 94 54 54 ?4 84 +, $( %( )( *( -( ,( $& %& )& *& -& ,& ,+ @! "4 :4 84 64 64 74 94 04 04 84 04 <4 A4 "4 <4 64 5! 54 04 94 B4 54 "4 84 74 84 24 74 B4 $" *" '" #" !4 64 %" &" !4 6" 04 74 :4 64 (4 04 84 94 :4 6" 54 64 '" &4 54 94 84 64 ;4 04 74 "4 "4 54 <4 =4 <4 >4 64 54 %4 64 '" )4 94 94 54 94 54 54 ?4 84 04 94 ?4 94 84 ;4 54 "4 54 54 04 ;4 >4 >4 94 64 94 94 > ( (( +( $( %( )( *( -( ,( '( (&( (& +& $& %& )& *& -& ,& '& (&& !/# ./0(1(&( ./0+,1,+ Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   147	
   Backbone amide chemical shifts of the N-terminal constructs of CBP The 1H-15N HSQC spectra of CBP28-82 and CBP1-101 were assigned using standard triple resonance methods, such as the HNCACB, CBCA(CO)NNH, and 13C-TOCSY experiments (Sattler et al., 1999).  Approximately 98% and 93% of the expected amide 1HN-15N signals were identified in the spectra of CBP28-82 and CBP1-101, respectively (Figure 4.5B). Most 13Cα and 13Cβ signals were also assigned by this approach.  The 1H-15N HSQC spectrum of CBP28-57 was assigned by comparison with those of CBP1-101 and CBP28-82. However, a majority of the amides in CBP28-57 exhibited more perturbed chemical shifts relative to the corresponding peaks in CBP1-101 and CBP28-82 (Figure 4.6A and B, orange bars). This suggests that structural perturbations result upon deleting residues 58-82. In contrast, with the exception of residues near the new termini, most peaks from corresponding amides in the HSQC of CBP28-82 were superimposable on those in CBP1-101 (Figure 4.6A and B). This suggests that CBP28-82 mimics CBP1-101, whereas CBP28-57 may differ in the middle (residues 41-44), possibly due to the loss of a weak or transient intramolecular interaction with residues 58-82.  4.2.3 Structural and dynamic characterization of the N-terminal region of CBP Most amide 1HN nuclei in the 3 CBP fragments had random coil chemical shifts (i.e. poorly dispersed within the limited range of ~7.5 - 8.5 ppm), suggesting that the N- terminal region of the co-activator is predominantly unstructured in isolation (Figure 4.6A). However, this is contrasted by the small spectral perturbations observed when the residues 58-82 were removed from CBP28-82, indicative of a weak structural element  Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   148	
         Figure 4.6 NMR characterizations of the N-terminal CBP constructs (A) 1H-15N HSQC spectra of CBP1-101 (red) superimposed upon that of CBP28-82 (left, blue) and CBP28-57 (right, green). (B) Changes in corresponding amide chemical shifts (calculated as {(10Δδ1HN)2 + (Δδ15N)2}1/2) for CBP1-101 versus CBP28-82 (upper panel) and CBP28-57 (lower panel). Orange bars denote residues showing larger changes in CBP28- 57 than CBP28-82 in comparison to CBP1-101. (C, D) Secondary structure elements predicted from the SSP algorithm using 1HN, 15N, 13Cα, and 13Cβ chemical shifts (upper panels). Values approach +1 for α-helices, -1 for β-strands, and zero for random coil conformations. Backbone dynamics were measured by a heteronuclear 15N NOE experiment (lower panels). Decreasing values indicate increasing flexibility on the sub- nsec timescale. (E) CD spectrum of CBP1-101 recorded at 10 °C. Although mostly unstructured, small peak at 222 nm (arrow) suggests the existence of a weak helical conformation. Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   149	
       !"# !"$ %% & !"' (#"! ! ()"! ($"! $ $$ )$ #$ *$ '$ +$ ,$ -$ .$ $!$ / 0 (0 1 2 34567849:8;<43 (!"' (!"# (!"$ !"' !"# !"$ %% & !"' (#"! ! ()"! ($"! $ $$ )$ #$ *$ '$ +$ ,$ -$ .$ $!$ / 0 (0 1 2 34567849:8;<43 (!"' (!"# (!"$ !"' =>? =@? @A&$($!$ @A&)-(-) -"* -") -"! $)! $$' $$! -"* -") -"! $' 0 9=B B; ? $/9=BB;? =C? 6 b DE 4; 6D FG 95 E6 HI9 =B B; ? @A&$($!$9J59@A&)-(-) @A&$($!$9J59@A&)-(', =A? $ $$ )$ #$ *$ '$ +$ ,$ -$ .$ $!$ ! $"! )"! ! $"! )"! 34567849:8;<43 @A&$($!$9J59@A&)-(-) @A&$($!$99J59@A&)-(', ; 74 K‡ D; ) L 7; MG ($* ($) ($! (- (+ (* () ! NFJ4G4:KIE9=:;? $.' )$' )'')#' ),' ).' #$' =2? Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   150	
   that is lost upon truncation of CBP1-101 and CBP28-82. To examine this further, more detailed chemical shift analyses and NMR relaxation measurements were undertaken.  Secondary structure prediction and backbone dynamics The SSP (secondary structure propensity) algorithm was utilized to predict the secondary structure components of CBP1-101 and CBP28-82 based on the chemical shifts of the assigned 1HN, 15N, 13Cα, and 13Cβ  nuclei. In both constructs, the SSP scores were small, with values indeed indicative of predominant random coil conformations (Figures 4.6C and D). However, in CBP1-101, weak but clear SSP-based secondary structure propensities were observed (Figure 4.6C). Residues 53-76 exhibited positive SSP scores indicative of partial α-helical character, whereas residues 10-25 and 79-99 had negative scores suggestive of some β-strand character. Consistently, the helical pattern was also observed in the same region in CBP28-82 (Figure 4.6D). Thus the central portion of the isolated NRID appears to transiently adopt a helical conformation.  The presence of structural components in the N-terminal region was further dissected for CBP1-101 and CBP28-82 by 15N-relaxation experiments (Figures 4.6C and D). The heteronuclear NOE is a sensitive measure of motion, with values decreasing from ~ 0.82 to ~ -3.6 with increasing mobility of the 1HN-15N bond vector on the sub-nsec timescale (Kay et al., 1989). No rigid components were identified for both constructs as all heteronuclear NOE values were < 0.3. However, unlike typical random coils with heteronuclear NOE ~ -0.5 (Cho et al., 1996; Lee et al., 2000; Seo et al., 2007), residues 32-54 and 62-74 in CBP1-101 and residues 35-73 in CBP28-82 were rather close to or even above zero (Figures 4.6C and D). Thus, consistent with SSP scores, this suggests Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   151	
   the presence of transiently ordered secondary structure elements in this region of the NRID. In contrast, the C-terminal and N-terminal ends of the fragments exhibited large negative values indicative of a random coil-like behavior (Figures 4.6C and D).  The secondary structure of CBP1-101 was also assessed using circular dichroism spectroscopy. No indication of any predominant secondary structure was observed in the CD spectrum of this fragment recorded at 10 oC (Figure 4.6E). However, slightly negative ellipticity was observed at 222 nm, indicative of some residual helicity in the construct. These features were diminished in the spectrum recorded for the sample with 4M guanidine hydrochloride denaturant (not shown). Overall, these measurements indicated that the CBP1-101 and CBP28-82 fragments are predominantly unstructured in isolation, yet exhibit behavior indicative of transient helical conformation for residues 61- 74.  4.2.4 Identification of the TAZ1-binding sequence in NRID HSQC-monitored titrations were used extensively to delineate the minimal N-terminal region of CBP for TAZ1 binding and to identify the binding interface on TAZ1. Titrations were performed for all the combinations of the labeled and unlabeled TAZ1 domain and various N-terminal constructs, except for 15N-CBP1-356 with TAZ1 due to the expected overcrowded spectrum of CBP1-356.  Monitoring TAZ1 To identify the minimal N-terminal TAZ1-binding fragment of CBP, 15N-TAZ1 was titrated with CBP1-356, CBP1-101, CBP28-82, and CBP28-57. HSQC-based titrations Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   152	
   demonstrated that the NRID construct CBP1-101 (Kamei et al., 1996) binds the TAZ1 domain in a manner very similar to that of the larger CBP1-356 (Figure 4.4A and 4.7A). The spectral changes of TAZ1 in the presence of CBP28-82 also match those of the TAZ1 domain with CBP1-356 (Figure 4.4A and 4.7B). When the top 20 residues with the most chemical shift perturbations were mapped on the structure of the TAZ1 domain, the titrations with all three constructs (CBP1-356, CBP1-101, and CBP28-82) revealed a very similar binding region (Figure 4.8).  Further deletion of residues 58-82 in CBP28-82 significantly disrupted the interaction with TAZ1 as evidenced by the reduced spectral changes during the titration of TAZ1 with CBP28-57 compared to other constructs (Figure 4.7C and D). This is also consistent with the HSQC comparison analyses in the previous section by which deleting residues 58- 82 in CBP28-82 causes a structural perturbation relative to CBP1-101 and CBP28-82 (Figure 4.6A and B). Therefore, residues 28-82 of CBP contain the minimal N-terminal fragment that binds the TAZ1 domain. Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   153	
          Figure 4.7 HSQC-monitored titrations of TAZ1 with CBP1-101, CBP28-82, and CBP28-57 The 1H-15N HSQC spectra of 15N-labeled TAZ1 domain were monitored while adding aliquots of non-labeled (A) CBP1-356, (B) CBP1-101, (C) CBP28-82, and (D) CBP28-57. The 10 residues showing the largest amide or indole shift changes are labeled on each spectrum except for CBP28-57. The volume of the concentrated CBP1-101 (exact concentration unknown) gradually added to an initial volume of 300 µL TAZ1 domain until saturation shown as red (0 µL), orange (75 µL), yellow (175 µL), green (275 µL), and blue (350 µL). CBP28-82 concentration at each point relative to that of TAZ1 domain (initially 0.2 mM) is shown as red (0%), orange (25%), yellow (50%), green (100%), and blue (200%). Volume of the concentrated CBP28-57 (exact concentration unknown) gradually added to an initial volume of 300 µL TAZ1 domain until saturation shown as red (0 µL), orange (5 µL), yellow (25 µL), green (50 µL), and blue (100 µL).  C hapter 4 – D iscovery and characterization of a TA Z1 autoinhibitory region in C B P  	
  	
   154	
    !" # $ % !&" !'( !'" !!( !!" )&(# *+!" ,+!# -&%' .&%" ,&(! -&/& 0+!$12¡!3 4&($ 5&/' !" # $ % !&" !'( !'" !!( !!" *+!" ,+!# ,&(! -&/& 0+!$12¡!3 )+'# 5+!% 6&#( )&(# 5&/' ! ( 7 1 2 8 8 9 3 !5128893 2-3 2:3 ;1<:=!>&(/ !&" !'( !'" !!( !!" !" # $ % 2<3 )&(# *+!" ,+!# ,&(! -&/& 0+!$12¡!3 5+!% 6&#( 4&($ 5&/' !" # $ % !&" !'( !'" !!( !!" 2?3 ;1<:='$>$' ;1<:=!>!"! ;1<:='$>(% ! ( 7 1 2 8 8 9 3 ! ( 7 1 2 8 8 9 3 ! ( 7 1 2 8 8 9 3 !5128893 !5128893 !5128893 Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   155	
                  Figure 4.8 Identification of the TAZ1 interfaces for binding CBP1-356, CBP1-101 and CBP28-82 (A) Histogram showing changes of amide chemical shifts (calculated as {(10Δδ1HN)2 + (Δδ15N)2}1/2) in the TAZ1 domain resulting from the addition of approximately equimolar CBP1-356 (top), CBP1-101 (middle), and CBP28-82 (bottom). (B) Signals from the 20 residues of TAZ1 showing the largest spectral perturbations upon titration with each CBP fragment are highlighted in red on the surface of the unbound TAZ1 structure (PDB ID:1u2n). Black labels denote the common changes among three constructs and blue labels show minor differences; however this is largely due to the arbitrary cut-off criteria used to generate these figures.    C hapter 4 – D iscovery and characterization of a TA Z1 autoinhibitory region in C B P  	
  	
   156	
    !"#$ !"#$ !"#$!"#$ %&'( %&'( )( )' )" )& "&* "+* "#* ",* "-* "$* &** &(* &'* &"* &(-.¡!/ 6 b  0 1 2 3 4 0 5 6 7 8 1 4 9 : 7 . ; ; 3 / " "&* "+* "#* ",* "-* "$* &** &(* &'* &"* &(-.¡!/ "&* "+* "#* ",* "-* "$* &** &(* &'* &"* &(-.¡!/ * ' ( " * ' ( + * " ( ' & !"#$ <",' !"#- !"-+ =&(* >"&# ?"+(@"+& A&'- A"+- )"#' B"+$B"#( )&(, <",' C"$+ !"-+ !"#- ?&($ D&(- E7%FG'-H-' E7%FG(H"+# E7%FG(H(*( .</ .F/ E7%FG'-H-' E7%FG(H"+# E7%FG(H(*( Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   157	
   Monitoring N-terminal constructs A complementary titration of 15N-labeled CBP1-101 with unlabeled TAZ1 was used as an initial point for localizing the minimal TAZ1-binding fragment. (As mentioned earlier, CBP1-356 was not titrated with TAZ1 because a poorly resolved spectrum was expected.) The progressive changes in the 1HN and 15N chemical shifts of several amides in this construct with increasing concentrations of the TAZ1 domain confirmed binding in the fast exchange limit (Figure 4.9), with estimated Kd of ~ 9 ± 4 µM (average values of Phe34 and Asn38). Furthermore, these amides were clustered between residues 30 and 75, indicating that this is the TAZ1-binding sequence (Figure 4.10A).  Subsequently, two other 15N-labeled deletion constructs, CBP28-82 and CBP28-57, were also tested for binding to the TAZ1 domain. Comparable changes were observed for corresponding residues in the HSQC spectra of CBP28-82 as seen with CBP1-101, with estimated Kd of ~ 10 ± 2 µM (average values of Phe34 and Asn38), confirming that amino acids 1-28 and 83-101 do not contribute measurably to binding the TAZ1 domain (Figure 4.9 and 4.10A). Corresponding residues in CBP28-57 also showed chemical shift changes, indicating binding of this small fragment to TAZ1, yet with a significantly reduced affinity of Kd ~ 50 ± 25 µM (average values of Phe34 and Asn38). Consistent with the results seen for the reverse titrations of the labeled TAZ1 domain with the N- terminal constructs of CBP, this supports the conclusion that residues 58-82 are important for full affinity. However, residues 28-57 still interact with the TAZ1 domain, suggestive of an extended binding interface. Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   158	
          Figure 4.9 Titrations of various NRID constructs with TAZ1 The 1H-15N HSQC spectra of 15N-labeled N-terminal CBP fragments were monitored upon titration with the unlabeled TAZ1 domain. Signals from the 15 residues of CBP1-101 (left), CBP28-82 (middle), and CBP28-57 (right) showing the largest chemical shift changes upon binding TAZ1 are labeled on each spectrum.  !"# !"$ !"% &$% &&' &&% (#' )*# +*! ,*& -'* -.* /.# 0*' 1*. !"# !"$ !"% &$% &&' &&% &' + 234 45 6 789&:&%&2;2<=>& )*# +*! ?*$ ,*& +'# 1#* 1*. (#' -.* /.# -'*0*' 1.& 789$!:!$2;2<=>& &@234456 0*A -B 1#* )*% 0*A 1.& -#% 0$A !"B !"# !"$ !"% ."! &$' &$% &&' &&% ,*& 0#$ )*# +*! ?*$ 1*. 0*' (#' -'* 789$!:'.2;2<=>& Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   159	
    Figure 4.10 Mapping of the TAZ1 domain-binding sequence of CBP and their conformational changes. (A) Histogram of changes in amide chemical shift changes (calculated as {(10Δδ1HN)2 + (Δδ15N)2}1/2) upon the addition of approximately equimolar TAZ1 domain were measured for CBP1-101 (top), CBP28-82 (middle), and CBP28-57 (bottom). The LPQ/EL-like (green), the LXXLL-like (blue) and the LXXLL (red) motifs are highlighted. (B) SSP scores for each titration point of CBP28-82 with the TAZ1 domain. The lower panel shows the difference of SSP scores between fully bound and free CBP28-82. !6 " # $% &! '( &# $' ") !* ($ +,! -. .# / 0&*$%1&!21#3&0 -4/ 56 77 76 87 86 97 96 :7 :6 ;7 ;6 56 77 76 87 86 97 96 :7 :6 ;7 ;6 <=6 < <=: <=5 <=8 ><=5 <=6 < <=: <=5 <=8 ><=5 ?? @ 6 ?? @ +1))!A4BC!>!2D!A4BC 0&*$%1&!21#3&0 -E/ C CC 5C 7C 8C 9C :C ;C 6C FC C<C C CC 5C 7C 8C 9C :C ;C 6C FC C<C C CC 5C 7C 8C 9C :C ;C 6C FC C<C 7 5 C < 7 5 C < 7 5 C < GE@C>C<C GE@56>9; GE@56>65 2D!A4BC 59H!A4BC ;9H!A4BC C<<H!A4BC Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   160	
   Secondary structure prediction for the N-terminal construct bound to the TAZ1 domain For the minimal N-terminal construct CBP28-82, 13Cα and 13Cβ resonances were also assigned for each TAZ1 titration point using HNCACB and CBCA(CO)NH spectra. This enabled use of the SSP algorithm to predict the secondary structure of the CBP fragment when bound to TAZ1. As shown in Figure 4.10B, progressive increases in SSP scores with added TAZ1 were observed for 65-74, reaching values up to 0.6 (Figure 4.10B). This confirms further that these residues interact with the TAZ1 domain and suggests that a helical conformation is induced upon binding.  4.2.5 Characterization of NRID-TAZ1 interaction by surface plasmon resonance The functional implications of the NRID-TAZ1 domain interaction were characterized using SPR. The "in trans" interactions of the isolated TAZ1 domain with three N-terminal CBP constructs (CBP1-356, CBP1-101, CBP28-57) were measured to be in low µM range (~50-70 µM). This is at least ~1000 fold weaker than the interaction with a HIF-1α (Kd ~ 3 nM) (Nelson, 2007). Thus the N-terminal sequence of CBP is not optimal for TAZ1 domain binding and would not effectively compete against HIF-1α.  However, the HIF-1α affinity of CBP1-452, which contains both the NRID and TAZ1 domains, was ~19 fold weaker than that of the isolated TAZ1 domain (Nelson, 2007). Thus the "in cis" presence of the NRID does reduce the overall affinity for targets such as HIF-1α. This difference likely results from the reduced entropic penalty of the intra- versus inter-molecular binding equilibrium. More importantly, these measurements Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   161	
   suggest that, in the context of native CBP, the NRID can indeed compete for the interaction of the TAZ1 domain with HIF-1α and possibly with other binding partners. Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   162	
   4.3 Discussion Extensive NMR-based analyses have revealed that the N-terminal region of CBP, contained within residues 28-82, interacts with the TAZ1 domain. Based on a common binding site on the TAZ1 domain and the reduced affinity of CBP1-452 versus that of the TAZ1 domain alone for HIF-1α, we hypothesize that this provides a route for the auto- regulation of TAZ1 domain binding with numerous transcription factors.  Residues 28-82 form the minimal N-terminal binding fragment for the TAZ1 domain. This conclusion is based on complementary titrations with four N-terminal fragments of CBP, combined with secondary structure propensity and protein backbone dynamic measurements. Although mostly unstructured and flexible in isolation, residues 38-74 showed noticeably higher SSP and heteronuclear 15N-NOE values than expected for a completely random coil polypeptide, suggestive of transient helical conformations. NMR-based titrations localized the TAZ1-binding sequence within residues 31-74, thus defining CBP28-82 as the minimal TAZ1-binding N-terminal fragment. Increases in SSP values were observed for residues 65-71 of CBP28-82 when interacting with TAZ1, suggesting binding-induced helix formation. Furthermore, chemical shift perturbation mapping indicated that the N-terminal CBP fragments wrap around TAZ1 in a similar manner as seen with other TAZ1 binding proteins, employing multiple anchoring points that collectively contribute to binding. That is, unlike the binding of a folded domain that requires its structural integrity, the extended binding sequence of the NRID still interacts even in the case of sequential disruption, such as deletion, albeit with reduced affinity.  Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   163	
   TAZ1-binding sequence motifs within the N-terminal region of CBP The minimal N-terminal TAZ1 domain binding fragment CBP28-82 bears one LPQ/EL-like motif (40LPDEL44), one LXXLL-like motif (50LSLL53) and a LXXLL motif (69LSELL73). As described earlier, the LPQ/EL and LXXLL motifs have shown to interact with some of the CBP binding partners (Klein et al., 2005). Intriguingly, both 50LSLL53 and 69LSELL73 are located in the region of CBP with relatively higher helical propensity and reduced flexibility. Upon binding the TAZ1 domain, chemical shift changes were observed for the 40LPDEL44 and 69LSELL73 regions, and less so for 50LSLL53. In addition, significantly increased helical propensities were observed for 69LSELL73 and its adjacent residues during the titration with TAZ1 (Figure 4.10B), as frequently seen for LXXLL motifs adopting an amphipathic helix upon binding their partners, such as nuclear receptors.  Comparison to other TAZ1 complex structures Many examples of intrinsically unstructured transactivation domains have been reported to bind TAZ1/2 domains by wrapping around their surfaces as induced helices. Currently, there are eight NMR-derived TAZ1 or TAZ2 domain-related structures (Figure 4.11) available in the Protein Data Bank. These include six TAZ1 domains complexed with one of the transactivation domains of HIF-1α (Dames et al., 2002; Freedman et al., 2002), CITED2 (Freedman et al., 2003; De Guzman et al., 2004) and STAT1/2 (Wojciak et al., 2009) from both CBP and p300.  Structural comparisons of these TAZ1 domain complexes unveil a common binding interface for its partners, as well as some variations (Figure 4.11). At first, this  Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   164	
            Figure 4.11 Free and bound structures of TAZ1 and TAZ2 Structural studies of TAZ1 and TAZ2, either from CBP or p300, reveal that peptide models of most transcription factors (red) bind by wrapping around a common extended interface. The directionalities of wrapping of transcription factors around TAZ1 are compared from the C-terminal end of H1 (arrow, see text). For TAZ1, four common interfaces (I(H1), I(LPQ/EL), I(H1/H3), I(H1/H3/H4)) were observed and highlighted on the free in green, yellow, blue, and orange, respectively (see text). Three zinc sites are shown in magenta spheres.    C hapter 4 – D iscovery and characterization of a TA Z1 autoinhibitory region in C B P  	
  	
   165	
   Û Û !"#$%"&'$()*+$,-.$!./-0$'123# !-#$%"&2$()*+$,-.$!./-0'45'# ‡7$=IURP&%3 ‡7$=IURPS ‡7$=IURP&%3 +LI_ 3'%/& +LI_ 3'%/( &,7(' 3'%34 &,7(' 3'%58 67$7 3'%.$ 67$7 3'%.$ 6' 62 6768 6' 62 67 68 6' 62 67 68 6' 62 67 68 3 , 3 3 , 3 , 3 , 3 , , Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   166	
   conclusion seemed unclear due to the differences in peptide size and sequence and entering/exit points for binding. However, when the TAZ1 domains in five complex structures are superimposed to the unbound structure, several interesting features are apparent (Figure 4.12). First, no major structural change occurs in TAZ1 due to the peptide binding. Second, all the transactivation domain peptides of HIF-1α, CITED2, and STAT2 follow very similar paths along an extended interface wrapping around the surface of TAZ1 (Figure 4.11A and 4.12B). Third, differences do exist in the precise details of binding. Most notably, when the TAZ1 domain is viewed from the C-terminal end of H1 (Figure 4.11A), HIF-1α and CITED2 spiral counter-clockwise whereas STAT2 wraps clockwise (Figure 4.12A).  The common binding interface on TAZ1 can be divided roughly into four regions (Figure 4.11A). The first region, I(H1), is located on helix H1 between the N-terminal portion of H2 and H4. The second region, I(LPQ/EL), is surrounded by H1/H2(N-term)/H3, and interacts with the LPQ/EL motifs in HIF-1α and CITED2	
   (De Guzman et al., 2004). Interestingly, STAT2 also binds this region, not via a LPXL motif but with its second induced-helix, 802PMEIFR807	
   (Wojciak et al., 2009). Thus the LPXL motif is not an indispensible feature of a TAZ1-binding sequence. The third region, I(H1/H3), is located between H1 (N-term) and H3. Both HIF-1α and STAT2 form an induced helix at this interface (Dames et al., 2002; Freedman et al., 2002; Wojciak et al., 2009). Last, the fourth region, I(H1/H3/H4) spans the junction of H1 (C-term), H3, and H4. Again, HIF-1α and STAT2 form another induced-helix at this interface. In the structure of CITED2 with the CBP TAZ1 domain, CITED2 does not interact through I(H1/H3/H4)	
    Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   167	
                    Figure 4.12 Investigating the common binding interface on TAZ1 (A) The aligned CBP TAZ1 domain structures in the unbound state (grey, PDB: 1u2n) and  bound to peptide models of HIF-1α (red, PDB: 1l8c), CITED2 (blue, PDB: 1r8u), and STAT2 (green, PDB: 2ka4). Peptides are removed for clarity. (B) Localization of the peptides when TAZ1 structures in the complexes are aligned to the unbound TAZ1 structure. (Left) TAZ1 removed for clarity. (Right) Backbone of the TAZ1 domain in surface representation. Same color codes as (A) except HIF-1α (orange, PDB: 1l3e) and CITED2 (cyan, PDB: 1p4q) from TAZ1 of p300. (C) Consensus of the top 20 TAZ1 domain residues (in red) showing spectral changes (Figure 4.7B) upon CBP28-82 binding highlighted on the structure of TAZ1/HIF-1α complex. Backbone of the HIF-1α peptide is shown in blue with the LPQL highlighted in green.   Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   168	
   !" !# !$!% Ý Ý &'( &)( &*( !"!# !$ !% Ý Ý !" !# !$ !% Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   169	
   (De Guzman et al., 2004). However, in the structure of p300 TAZ1 with a longer CITED2 fragment, the construct does interact with I(H1/H3/H4)	
  (Freedman et al., 2003).	
   N-terminal fragment of CBP also binds the common interface on TAZ1 The interaction of the N-terminal fragment of CBP with the TAZ1 domain resembles that of the transactivation domains. This conclusion is based on the size and physicochemical properties of CBP28-82, followed by the binding interface mapping from chemical shift perturbations. The minimal N-terminal fragment CBP28-82 includes up to 55 residues, which is similar to those found in the transactivation domains, CITED2 (44- 50 residues), HIF1α (41-51 residues), and STAT2 (53 residues), for wrapping around the TAZ1 domain. Previous structural studies also demonstrated that all the TAZ1- binding transactivation domains are negatively charged with pI values ranging ~ 3.6 – 4.5 to complement the positively charged TAZ1 domain (pI ~ 9.5). Similarly, the N- terminal region of CBP is also negatively charged at neutral pH, (pI ~4.4 for CBP28-82).  The electrostatic contribution for the CBP28-82/TAZ1 domain binding is evident from the large chemical shift changes observed for the charged residues in both proteins during the titration experiments. First, in CBP28-82, six negatively charged residues (Asp29, Asp35, Glu37, Asp39, Glu43, Glu71) were found among the top 15 residues with the largest amide chemical shift changes upon binding the TAZ1 domain. The complementary charges were identified in TAZ1 during the titration with CBP28-82, including 6 positively charged residues (Lys351, His362, Arg368, Arg385, His417, Lys419), localized in the I(H1/H3/H4) region (Figure 4.11A).  Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   170	
   In parallel with the electrostatic interactions, hydrophobic contributions were also implicated by large chemical shift changes of non-polar residues in both the TAZ1 domain and CBP28-82 during the titrations. For example, Ile45 in CBP28-82 exhibits the largest chemical shift change upon addition of the TAZ1 domain (Figure 4.9A and B). Importantly, three of the four hydrophobic residues with large chemical shift changes in CBP28-82 were located in the 40LPDEL44, 50LSLL53, and 69LSELL73 motifs (underline: large change), with the fourth located adjacent to the first sequence, Ile45. Clusters of hydrophobic residues with large chemical shift changes were identified along helix H1 in the TAZ1 domain, namely Val358, Leu359, Leu361 and Leu363, when this domain was titrated with CBP28-82. Last, when the CBP28-82 binding residues were mapped on the TAZ1 domain of the TAZ1-HIF-1α complex structure, these residues are clearly localized along the binding interface of HIF-1α including all four regions on TAZ1 (Figure 4.12).  Despite all these similarities, strikingly weaker TAZ1 domain-binding was observed for the CBP N-terminal fragments (Kd ~ 9 µM) than for peptide models of the transactivation domains of HIF-1α, CITED2, and STAT2 (Kd ~ 13 – 58 nM) (Dames et al., 2002; De Guzman et al., 2004; Wojciak et al., 2009). Although sharing common physicochemical features, no obvious sequence similarities were observed among the three transactivation domains other than the LPQ/EL motif, which is present in both HIF-1α and CITED2, but not in STAT2. Thus the exact reason for the ~1000-fold difference in affinity is unclear.  Nevertheless, it is important to recognize that the interaction of the N-terminal fragment with TAZ1 was tested in trans, and not in its natural context. Therefore, it is very likely that the binding affinity between the TAZ1 domain and the N- Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   171	
   terminal region of CBP increases when they are localized in a same molecule. Post- translational modifications, such as phosphorylation, may also be required in the NRID region for higher affinity binding with TAZ1. Alternatively, the relatively weaker binding may be necessary if such an interaction is involved in a competitive binding mechanism in the regulation of binding TAZ1 partners.  Autoinhibitory mechanism of CBP NRID and TAZ CBP, as a transcriptional co-activator, is a large scaffold protein, known to interact with numerous transcription factors to enhance subsequent gene activation. Although the TAZ domains are involved in many of these interactions, regulation of binding affinity has only been shown through their partners (i.e. via post-translational modifications), and not through CBP itself. The intramolecular interaction of TAZ1 and NRID suggests a possible model for an autoinhibitory mechanism by which NRID competes for TAZ1 domain binding to attenuate interactions with partner transcription factors. For example, the interaction of NRID with its binding partner, such as nuclear hormone receptors, could activate gene expression by exposing the binding interface of the TAZ1 domain to recruit other transcription factors (Figure 4.13). Conversely, binding of these transcription factors to the TAZ1 domain would free the NRID to interact with nuclear hormone receptors. Further in vitro and in vivo studies are necessary to test these hypotheses. Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   172	
            Figure 4.13 Model of the putative autoregulatory mechanism for binding TAZ1 The intramolecular interaction of NRID and TAZ1 competitively inhibits the binding of transcription factors. Upon external stimuli, the activation function (AF) region of nuclear receptor binds the LXXLL motif in NRID of CBP, thereby exposing the common binding interface on TAZ1 for interacting with transcription factors.  !"# $ !"# $ !"#$ !"#$ !%& '"(!)* +,-./01 1/-/2341 Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   173	
   4.4 Conclusions The newly identified intramolecular interaction between the TAZ1 and NRID domains suggests a putative autoinhibitory mechanism for regulating the interaction of CBP with its myriad of binding partners. From the NMR-based analyses, the minimal TAZ1 domain-binding NRID fragment was delineated as spanning residues 28-82. Chemical shift perturbation mapping indicated that these residues adopt at least one region of induced helical structure upon binding the common interface on the TAZ1 domain via electrostatic and hydrophobic interactions.  4.5 Materials and methods 4.5.1 Cloning Constructs encoding three CBP fragments (CBP1-356, CBP1-101, CBP28-57) were generated by PCR-amplification of the CBP gene, followed by ligation into pGEX vector for expression as GST-fusion proteins. In the case of CBP28-82, a pET28a vector was modified to encode an N-terminal His6 tag, followed by GABPα35-121, new thrombin cleavage site as a linker (LVPRGS), CBP28-82, and an additional C-terminal tryptophan for concentration measurements. The original thrombin site between the His6 tag and GABPα35-121 was destroyed by site-directed mutagenesis (LVPRGS -> LVPGGS). GABPα35-121 is a highly expressed OST domain, characterized previously during my M.Sc. studies (Kang, 2005). This fusion was used to increase expression of the short CBP28-82 polypeptide. The protocol for cloning the gene encoding TAZ1 was described previously (Kang et al., 2008). All the final plasmids were transformed into E. coli BL21(λDE3) cells.  Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   174	
   4.5.2 Protein expression and purification All the bacterial cultures were incubated in rich LB or minimal M9T media with shaking at 37 °C until the cell density reached OD600 ~ 0.8. Then protein expression was induced with the addition of IPTG (1 mM, final), followed by growth overnight at 30 °C, The final cell cultures were harvested with centrifugation at 5K for 10 min in a Sorval GSA rotor.  Cell pellets were resuspended in the binding buffer for GSTrap- or HisTrap- affinity column (GE Healthcare) and lysed with both a French press (2X at 15000 psi) and by sonication (5 min).  Each lysate was cleared by centrifugation at 15K for 60 min, and then the supernatant loaded onto GSTrap- or HisTrap-affinity column, pre- equilibrated with its binding buffer. After washing, tagged proteins were eluted with a gradient from the binding buffer to the elution buffer using an AKTA purifier (Amersham Biosciences). Desired fractions were identified by SDS-PAGE electrophoresis, and pooled for overnight dialysis (4L, 20mM Tris-HCl, pH 8.4, 150 mM NaCl, 2.5 mM CaCl2) in the presence of 1µL thrombin (Roche). Thrombin was removed by centrifugation after incubation with p-aminobenzamidine beads (Sigma) for 10 min. Any uncleaved protein and the cleaved tag were removed by re-running each sample through the appropriate column. Finally, all N-terminal constructs except CBP1-356 were purified by HPLC using a C18 reverse phase column with a gradient from 0 to 100 (acetonitrile:water, 0.1 mM trifluoroacetic acid). Fractions with the desired proteins were identified by mass spectrometry. After lyophilization, the final samples were resuspended, dialyzed in 20 mM Tris, pH 6.9, 50 mM NaCl, 2 mM DTT overnight, and concentrated using an Amicon Ultra filtration device (Millipore). Sample concentrations were determined by uv absorbance using predicted ε280 values (Pace et al., 1995).  Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   175	
   GSTrap column Constructs: CBP1-356, CBP1-101, CBP28-57 Binding buffer: 140 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, pH 7.3, freshly added 5 mM DTT Elution buffer (freshly prepared): 50 mM Tris-HCl, pH 8, 10 mM glutathione (Bioshop), 5 mM DTT  HisTrap column Construct: CBP28-82 Binding buffer: 20 mM Na2HPO4, pH 7.4, 500 mM NaCl, 5 mM Imidazole Elution buffer: 20 mM Na2HPO4, pH 7.4, 500 mM NaCl, 500 mM Imidazole  4.5.3 NMR spectral assignment All NMR spectra were recorded at 25 °C using 500 MHz Varian Unity and 600 MHz Varian Inova NMR spectrometers equipped with triple resonance gradient probes. The pulse sequences were provided by Dr. Lewis Kay (University of Toronto). Spectra were processed and analyzed using NMRpipe (Delaglio et al., 1995), nmrDraw (Delaglio et al., 1995), and Sparky (Goddard). 1H, 13C, and 15N resonances of CBP1-101 were assigned using standard triple resonance correlation experiments.  The protocols for assigning protein backbone resonances were described in my M.Sc (Kang, 2005). The resonances of the backbone amides of CBP28-57 were assigned by comparison with the spectra of CBP28-82 and CBP1-101.  Chapter 4 – Discovery and characterization of a TAZ1 autoinhibitory region in CBP 	
   	
   176	
   4.5.4 HSQC-monitored titrations Uniformly 15N-labeled samples (~ 0.1 - 0.5 mM) in titration buffer (20 mM Tris, pH 6.9, 50 mM NaCl, 2 mM DTT, 10% D2O) were titrated with small aliquots of concentrated (~ 1.0 -1.5 mM) non-labeled binding partners to minimize dilution. Accurate quantitative analyses and comparisons could not be achieved easily for most titrations other than CBP28-82 (with an added C-terminal Trp) due to the absence of any aromatic residues as required to measure reliable protein concentrations by absorbance spectroscopy. In these cases, concentrations were only estimated based on the SDS-PAGE electrophoresis bands sizes and relative NMR peak intensities.  Chapter 5 – Concluding remarks 	
   	
   177	
   Chapter 5 - Concluding remarks Eukaryotic transcription factors are comprised of modular domains joined by flexible linkers that collectively are crucial in providing various activities related to transcriptional regulation. Targeting the activities of these domains and linkers, commonly by post- translational modifications in response to signal transduction pathways, provides multiple layers of transcriptional regulation. As described throughout this thesis, the prototypical ETS family member, Ets1, includes three segments with protein-interaction, trans-activation, and DNA-binding activities that are governed by, at least, three different signal transduction pathways involving phosphorylation and sumoylation. The extensively studied DNA-binding C-terminal region of Ets1 is composed of the core ETS domain interfaced with a dynamic helical inhibitory module and a flexible serine rich region (SRR). Increasing levels of multi-site CaMKII-dependent phosphorylation of the SRR allosterically inhibits DNA-binding through transient interactions that, like a rheostat, progressively shift the inhibitory module and ETS domain from a flexible, active state to a rigid, inactive one. Similarly, as focused upon in this thesis, the N- terminal protein-interaction PNT domain is also comprised of regulatory components, including the core SAM domain-like helical bundle, the dynamic helix H0, and the adjacent phosphoacceptors, Thr38 and Ser41. MAPK-dependent phosphorylation of Thr38 and Ser41 affects the conformational equilibrium of helix H0, shifting it towards the open state. This further exposes the binding interface for the transcriptional co- activator CBP, thereby enhancing Ets1-dependent gene expression at Ras-responsive promoters.  Chapter 5 – Concluding remarks 	
   	
   178	
   In chapter 2 of this thesis, the structural and dynamic effects of phosphorylation on the Ets1 PNT domain were characterized using NMR spectroscopy. Unlike the long range allosteric linkage between the SRR and auto-inhibited ETS domain, phosphorylation of Thr38 and Ser41 primarily perturbed helix H0 and the H0/H2/H5 interface. Most strikingly, the NMR-derived structural ensemble of modified 2P-Ets129-138 demonstrated that phosphorylation triggers the displacement of the flexible helix H0 away from the helices H2 and H5, thus exposing a negatively-charged patch in the PNT domain. More interestingly, amide hydrogen exchange and 15N-relaxation analyses, along with mutagenesis studies, strongly suggested that even in the unmodified Ets1 fragment, helix H0 exists in a conformational equilibrium between ‘closed’ and ‘open’ states. The additional negative charges introduced upon phosphorylation near the N-terminus of helix H0 shifts this equilibrium towards the ‘open’ state.  The conformational equilibrium of helix H0 is functionally significant in regulating the interaction of Ets1 with the TAZ1 domain of the transcriptional co-activator CBP. In chapter 3, we showed that this interaction is enhanced ~ 30 fold upon phosphorylation. In vivo mutational studies with a transcription reporter system and in vitro deletion mapping studies combined with NMR analyses revealed that TAZ1 binding requires the core PNT domain, the appended helix H0, and the phosphoacceptors. Furthermore, the NMR perturbation mapping revealed that the negatively-charged TAZ1-binding interface of the PNT domain interacts with the positively-charged TAZ1 domain along the common interface found in the complex structures of the TAZ1 domain with several TADs. Thus, phosphorylation acts to both increase the charge complementarity of Ets1 Chapter 5 – Concluding remarks 	
   	
   179	
   and CBP, and to shift a conformational equilibrium of the PNT domain towards a more active, binding-competent state.  In chapter 4, we identified a putative autoinhibitory mechanism by which the interaction of the TAZ1 domain and the NRID in CBP lowers their affinities for various binding partners, possibly by competing for the same binding interfaces. The serendipitous discovery of this intramolecular interaction was followed by defining the boundaries of the minimal TAZ1 domain-binding NRID sequence, localizing it to within residues 28-82 by NMR methods. Further characterization of the NRID region revealed that it is predominantly flexible and random coil-like, yet may exhibit some weak intra-molecular interactions and a helical propensity. Similar to Ets1 and other TAZ1-binding partners, chemical shift perturbation mapping demonstrated that this minimal NRID region binds the common interface on the TAZ1 domain. Furthermore, an induced helical conformation was observed for residues 65-74 in the NRID region upon binding the TAZ1 domain. This behavior is analogous to those reported for other transactivation domain sequences, such as HIF-1α and CITED2, associated with the TAZ1 domain. Competitive in vitro binding studies confirmed that presence of the NRID sequence lowers the affinity of the TAZ1 domain HIF-1α. However, the role of this putative auto- inhibitory mechanism in vivo remains to be established.  In sum, this thesis focused on dissecting the structural and dynamic mechanisms involved in the regulation of Ets1 and CBP. In addition to the well-structured PNT and TAZ1 domains, in both mechanisms intrinsically disordered regions play a key role. These include the disordered phosphoacceptors and the dynamic helix H0 of the Ets1 Chapter 5 – Concluding remarks 	
   	
   180	
   PNT domain and the unstructured NRID region. The importance of intrinsically disordered regions, including entire intrinsically disordered proteins, has been reported in a wide range of examples, especially for numerous components of transcription, such as histone tails in nucleosomes, transactivation domains in transcription factors, various post-translational modification sites in co-activators, and the C-terminal domain in RNA polymerase II. The ‘conformational malleability’ of these disordered sequences allows diverse binding modes to other molecules resulting in various levels of molecular recognition and functional versatility (Fuxreiter et al., 2008). The two proteins described in this thesis especially exemplify the properties of intrinsically disordered regions. Such properties include the evolution of function by being appended to conserved core domains, bearing post-translational modification sites (Ets1 phosphoacceptors), undergoing structural rearrangements (conformational change of helix H0), providing short molecular recognition motifs (LXXLL), and stabilizing transient structural elements upon binding (induced folding) as seen for the NRID-TAZ1 interaction. Furthermore, these features of intrinsically disordered regions are often observed together in one region to accommodate more than one binding partner (promiscuity). For example, NRID of CBP can both bind intramolecularly to the TAZ1 domain and intermolecularily to nuclear hormone receptors.  5.1 Future directions The key discoveries presented in this thesis involve dissecting the phosphorylation- dependent conformational equilibrium of the Ets1 PNT domain for regulating CBP interactions (chapters 2 and 3) and a putative autoinhibitory mechanism of CBP arising Chapter 5 – Concluding remarks 	
   	
   181	
   from an intramolecular NRID/TAZ domain interaction (chapter 4). Despite this progress, fundamental questions regarding these regulatory mechanisms remain to be answered.  Phosphorylation of Thr38 and Ser41 shifts the conformational equilibrium of helix H0 to the ‘open’ state. However, the ensemble distribution of this helix is still poorly defined, and it is unclear whether H0 adopts a broad range of conformations or a more restricted space (i.e., an extension helix of H1) in the ‘open’ state. Using the ARIA protocol, the conformation of helix H0 was only weakly restricted by the15N-1HN and 13Cα-13C' RDCs, introduced during the structure calculations. Therefore, additional RDCs (13Cα-1Hα, 15N- 13C') measured in a range of alignment media (polyacrylamide gel, bicelle) (Bax, 2003) may help better define the position of helix H0 in both the ‘open’ and 'closed' states of the PNT domain. In parallel, the ensemble distribution of helix H0 could also be determined by introducing paramagnetic centers at various positions of the Ets1 PNT domain, including the N-terminal end of helix H0 and at exposed sites near the H0/H2/H5 interface. Paramagnetic relaxation enhancements (PRE) are readily used to extract longer range distance restraints (10 – 24 Å) possibly with interproton NOEs (< 6 Å) for structure calculations. Since chemical shifts are highly sensitive to conformation, direct refinement against this fundamental NMR parameter should better define the 'open' and 'closed' states of the PNT domain. Finally, in collaboration with Professor Joerg Gsponer, we are combining these experimental data with a computational molecular dynamics approach to better understand the motions of helix H0 upon phosphorylation.  Chapter 5 – Concluding remarks 	
   	
   182	
   High-resolution structural information of the NRID-TAZ1 domain complex still remains to be obtained, as the current study was limited to qualitatively identifying the residues involved in binding and their resulting conformation changes. However, attaining a stable complex could be difficult due to the relatively weak binding of the two isolated CBP fragments (Kd in µM range). Screening for better binding conditions may help stabilize complex formation. For example, the electrostatic contributions of binding, as expected from the oppositely charged NRID and TAZ1 domain, can be exploited by performing the experiments with lower ionic strength buffer. Joining the two CBP sequences covalently with a short flexible linker may help stabilize the resulting complex through entropic effects. However, problems could arise with increased NMR spectral overlap as both portions of the resulting chimera would be isotopically labeled. Similarly, this approach could facilitate or interfere with crystallization. Alternatively, docking procedures using limited NMR restraints may provide a low resolution model of the NRID bound to the TAZ1 domain. The insights from such models may help direct studies, such as mutational analyses, necessary to test in vivo the biological relevance of this potentially important, yet hitherto unrecognized, regulatory mechanism of CBP. References 	
   	
   183	
   References Amati, B., Brooks, M. W., Levy, N., Littlewood, T. D., Evan, G. I. and Land, H. (1993). Oncogenic activity of the c-Myc protein requires dimerization with Max. Cell 72(2): 233-45.   Appella, E. and Anderson, C. W. (2001). Post-translational modifications and activation of p53 by genotoxic stresses. Eur J Biochem 268(10): 2764-72.   Arany, Z., Huang, L. E., Eckner, R., Bhattacharya, S., Jiang, C., Goldberg, M. A., Bunn, H. F. and Livingston, D. M. (1996). An essential role for p300/CBP in the cellular response to hypoxia. Proc Natl Acad Sci U S A 93(23): 12969-73.   Bai, Y., Milne, J. S., Mayne, L. and Englander, S. W. (1993). Primary structure effects on peptide group hydrogen exchange. Proteins 17(1): 75-86.   Bannister, A. J. and Kouzarides, T. (2005). Reversing histone methylation. Nature 436(7054): 1103-6.   Bassuk, A. G. and Leiden, J. M. (1995). A direct physical association between ETS and AP-1 transcription factors in normal human T cells. Immunity 3(2): 223-37.   Bax, A. (2003). Weak alignment offers new NMR opportunities to study protein structure and dynamics. Protein Sci 12(1): 1-16.   Bedford, M. T. and Clarke, S. G. (2009). Protein arginine methylation in mammals: who, what, and why. Mol Cell 33(1): 1-13.   Berger, S. L. (2007). The complex language of chromatin regulation during transcription. Nature 447(7143): 407-12.   Bienkiewicz, E. A. and Lumb, K. J. (1999). Random-coil chemical shifts of phosphorylated amino acids. J Biomol NMR 15(3): 203-6.   Bienz, M. (2006). The PHD finger, a nuclear protein-interaction domain. Trends Biochem Sci 31(1): 35-40.   References 	
   	
   184	
   Blobel, G. A. (2000). CREB-binding protein and p300: molecular integrators of hematopoietic transcription. Blood 95(3): 745-55.   Bordoli, L., Husser, S., Luthi, U., Netsch, M., Osmani, H. and Eckner, R. (2001). Functional analysis of the p300 acetyltransferase domain: the PHD finger of p300 but not of CBP is dispensable for enzymatic activity. Nucleic Acids Res 29(21): 4462-71.   Brivanlou, A. H. and Darnell, J. E., Jr. (2002). Signal transduction and the control of gene expression. Science 295(5556): 813-8.   Buchwalter, G., Gross, C. and Wasylyk, B. (2004). Ets ternary complex transcription factors. Gene 324: 1-14.   Buttice, G. and Kurkinen, M. (1993). A polyomavirus enhancer A-binding protein-3 site and Ets-2 protein have a major role in the 12-O-tetradecanoylphorbol-13-acetate response of the human stromelysin gene. J Biol Chem 268(10): 7196-204.   Callaway, K., Waas, W. F., Rainey, M. A., Ren, P. and Dalby, K. N. (2010). Phosphorylation of the transcription factor Ets-1 by ERK2: rapid dissociation of ADP and phospho-Ets-1. Biochemistry 49(17): 3619-30.   Canagarajah, B. J., Khokhlatchev, A., Cobb, M. H. and Goldsmith, E. J. (1997). Activation mechanism of the MAP kinase ERK2 by dual phosphorylation. Cell 90(5): 859-69.   Cavalli, G. and Paro, R. (1998). Chromo-domain proteins: linking chromatin structure to epigenetic regulation. Curr Opin Cell Biol 10(3): 354-60.   Chakravarti, D., LaMorte, V. J., Nelson, M. C., Nakajima, T., Schulman, I. G., Juguilon, H., Montminy, M. and Evans, R. M. (1996). Role of CBP/P300 in nuclear receptor signalling. Nature 383(6595): 99-103.   Charavay, C., Eynard, J., Hus, J., Bouvignies, G. and Blackledge, M. SculptorCNS : Biomolecular structure determination and refinement in solution from NMR residual dipolar couplings, paramagnetic phenomena and spin relaxation data. (not published).   References 	
   	
   185	
   Cho, H. S., Liu, C. W., Damberger, F. F., Pelton, J. G., Nelson, H. C. and Wemmer, D. E. (1996). Yeast heat shock transcription factor N-terminal activation domains are unstructured as probed by heteronuclear NMR spectroscopy. Protein Sci 5(2): 262-9.   Clark-Lewis, I., Sanghera, J. S. and Pelech, S. L. (1991). Definition of a consensus sequence for peptide substrate recognition by p44mpk, the meiosis-activated myelin basic protein kinase. J Biol Chem 266(23): 15180-4.   Cohen, P. (2000). The regulation of protein function by multisite phosphorylation--a 25 year update. Trends Biochem Sci 25(12): 596-601.   Cornilescu, G., Delaglio, F. and Bax, A. (1999). Protein backbone angle restraints from searching a database for chemical shift and sequence homology. J Biomol NMR 13(3): 289-302.   Dallas, P. B., Yaciuk, P. and Moran, E. (1997). Characterization of monoclonal antibodies raised against p300: both p300 and CBP are present in intracellular TBP complexes. J Virol 71(2): 1726-31.   Dames, S. A., Martinez-Yamout, M., De Guzman, R. N., Dyson, H. J. and Wright, P. E. (2002). Structural basis for Hif-1 alpha /CBP recognition in the cellular hypoxic response. Proc Natl Acad Sci U S A 99(8): 5271-6.   Dang, C. V., Barrett, J., Villa-Garcia, M., Resar, L. M., Kato, G. J. and Fearon, E. R. (1991). Intracellular leucine zipper interactions suggest c-Myc hetero- oligomerization. Mol Cell Biol 11(2): 954-62.   Darnell, J. E., Jr. (1997). STATs and gene regulation. Science 277(5332): 1630-5.   de Alba, E. and Tjandra, N. (2004). Residual dipolar couplings in protein structure determination. Methods Mol Biol 278: 89-106.   De Guzman, R. N., Goto, N. K., Dyson, H. J. and Wright, P. E. (2006). Structural basis for cooperative transcription factor binding to the CBP coactivator. J Mol Biol 355(5): 1005-13.   References 	
   	
   186	
   De Guzman, R. N., Martinez-Yamout, M. A., Dyson, H. J. and Wright, P. E. (2004). Interaction of the TAZ1 domain of the CREB-binding protein with the activation domain of CITED2: regulation by competition between intrinsically unstructured ligands for non-identical binding sites. J Biol Chem 279(4): 3042-9.   De Guzman, R. N., Wojciak, J. M., Martinez-Yamout, M. A., Dyson, H. J. and Wright, P. E. (2005). CBP/p300 TAZ1 domain forms a structured scaffold for ligand binding. Biochemistry 44(2): 490-7.   Delaglio, F., Grzesiek, S., Vuister, G. W., Zhu, G., Pfeifer, J. and Bax, A. (1995). NMRPipe: a multidimensional spectral processing system based on UNIX pipes. J Biomol NMR 6(3): 277-93.   Dosset, P., Hus, J. C., Blackledge, M. and Marion, D. (2000). Efficient analysis of macromolecular rotational diffusion from heteronuclear relaxation data. J Biomol NMR 16(1): 23-8.   Duval, D. L., Jean, A. and Gutierrez-Hartmann, A. (2003). Ras signaling and transcriptional synergy at a flexible Ets-1/Pit-1 composite DNA element is defined by the assembly of selective activation domains. J Biol Chem 278(41): 39684-96.   Dwivedi, P. P., Omdahl, J. L., Kola, I., Hume, D. A. and May, B. K. (2000). Regulation of rat cytochrome P450C24 (CYP24) gene expression. Evidence for functional cooperation of Ras-activated Ets transcription factors with the vitamin D receptor in 1,25-dihydroxyvitamin D(3)-mediated induction. J Biol Chem 275(1): 47-55.   Ellenberger, T. (1994). Getting a grip on DNA recognition - structures of the basic region leucine-zipper, and the basic region helix-loop-helix DNA-binding domains. Current Opinion in Structural Biology 4(1): 12-21.   Farrow, N. A., Zhang, O., Forman-Kay, J. D. and Kay, L. E. (1994). A heteronuclear correlation experiment for simultaneous determination of 15N longitudinal decay and chemical exchange rates of systems in slow equilibrium. J Biomol NMR 4(5): 727-34.   Feng, H., Jenkins, L. M., Durell, S. R., Hayashi, R., Mazur, S. J., Cherry, S., Tropea, J. E., Miller, M., Wlodawer, A., Appella, E. and Bai, Y. (2009). Structural basis for p300 Taz2-p53 TAD1 binding and modulation by phosphorylation. Structure 17(2): 202-10.  References 	
   	
   187	
    Ferreon, J. C., Lee, C. W., Arai, M., Martinez-Yamout, M. A., Dyson, H. J. and Wright, P. E. (2009). Cooperative regulation of p53 by modulation of ternary complex formation with CBP/p300 and HDM2. Proc Natl Acad Sci U S A 106(16): 6591-6.   Foulds, C. E., Nelson, M. L., Blaszczak, A. G. and Graves, B. J. (2004). Ras/mitogen- activated protein kinase signaling activates Ets-1 and Ets-2 by CBP/p300 recruitment. Mol Cell Biol 24(24): 10954-64.   Freedman, S. J., Sun, Z. Y., Kung, A. L., France, D. S., Wagner, G. and Eck, M. J. (2003). Structural basis for negative regulation of hypoxia-inducible factor-1alpha by CITED2. Nat Struct Biol 10(7): 504-12.   Freedman, S. J., Sun, Z. Y., Poy, F., Kung, A. L., Livingston, D. M., Wagner, G. and Eck, M. J. (2002). Structural basis for recruitment of CBP/p300 by hypoxia- inducible factor-1 alpha. Proc Natl Acad Sci U S A 99(8): 5367-72.   Fuxreiter, M., Tompa, P., Simon, I., Uversky, V. N., Hansen, J. C. and Asturias, F. J. (2008). Malleable machines take shape in eukaryotic transcriptional regulation. Nat Chem Biol 4(12): 728-37.   Gagne, S. M., Tsuda, S., Li, M. X., Chandra, M., Smillie, L. B. and Sykes, B. D. (1994). Quantification of the calcium-induced secondary structural changes in the regulatory domain of troponin-C. Protein Sci 3(11): 1961-74.   Geiss-Friedlander, R. and Melchior, F. (2007). Concepts in sumoylation: a decade on. Nat Rev Mol Cell Biol 8(12): 947-56.   Ghosh, S., May, M. J. and Kopp, E. B. (1998). NF-kappa B and Rel proteins: evolutionarily conserved mediators of immune responses. Annu Rev Immunol 16: 225-60.   Goddard, T. D., Kneller, D.G. Sparky 3. University of California, San Francisco.   Goodman, R. H. and Smolik, S. (2000). CBP/p300 in cell growth, transformation, and development. Genes Dev 14(13): 1553-77.   References 	
   	
   188	
   Goroncy, A. K., Sato, M., Kosshiba, S., Inoue, M., Kigawa, T. and Yokoyama, S. (2006). Solution structure of the SAM_PNT-domain of ETS transcription factor PDEF (prostate ets). unpublished.   Goto, N. K., Zor, T., Martinez-Yamout, M., Dyson, H. J. and Wright, P. E. (2002). Cooperativity in transcription factor binding to the coactivator CREB-binding protein (CBP). The mixed lineage leukemia protein (MLL) activation domain binds to an allosteric site on the KIX domain. J Biol Chem 277(45): 43168-74.   Graves, B. J. and Petersen, J. M. (1998). Specificity within the ets family of transcription factors. Adv Cancer Res 75: 1-55.   Green, S. M., Coyne, H. J., 3rd, McIntosh, L. P. and Graves, B. J. (2010). DNA binding by the ETS protein TEL (ETV6) is regulated by autoinhibition and self- association. J Biol Chem 285(24): 18496-504.   Guntert, P. (2004). Automated NMR structure calculation with CYANA. Methods Mol Biol 278: 353-78.   Gunther, C. V. and Graves, B. J. (1994). Identification of ETS domain proteins in murine T lymphocytes that interact with the Moloney murine leukemia virus enhancer. Mol Cell Biol 14(11): 7569-80.   Gusterson, R., Brar, B., Faulkes, D., Giordano, A., Chrivia, J. and Latchman, D. (2002). The transcriptional co-activators CBP and p300 are activated via phenylephrine through the p42/p44 MAPK cascade. J Biol Chem 277(4): 2517-24.   Hahn, S. (2004). Structure and mechanism of the RNA polymerase II transcription machinery. Nat Struct Mol Biol 11(5): 394-403.   Hahn, S. L., Wasylyk, B. and Criqui-Filipe, P. (1997). Modulation of ETS-1 transcriptional activity by huUBC9, a ubiquitin-conjugating enzyme. Oncogene 15(12): 1489-95.   Heery, D. M., Hoare, S., Hussain, S., Parker, M. G. and Sheppard, H. (2001). Core LXXLL motif sequences in CREB-binding protein, SRC1, and RIP140 define affinity and selectivity for steroid and retinoid receptors. J Biol Chem 276(9): 6695-702.  References 	
   	
   189	
    Heery, D. M., Kalkhoven, E., Hoare, S. and Parker, M. G. (1997). A signature motif in transcriptional co-activators mediates binding to nuclear receptors. Nature 387(6634): 733-6.   Hurley, J. H., Dean, A. M., Thorsness, P. E., Koshland, D. E., Jr. and Stroud, R. M. (1990). Regulation of isocitrate dehydrogenase by phosphorylation involves no long-range conformational change in the free enzyme. J Biol Chem 265(7): 3599- 602.   Hutchinson, E. G. and Thornton, J. M. (1996). PROMOTIF--a program to identify and analyze structural motifs in proteins. Protein Sci 5(2): 212-20.   Hwang, T. L., van Zijl, P. C. and Mori, S. (1998). Accurate quantitation of water-amide proton exchange rates using the phase-modulated CLEAN chemical EXchange (CLEANEX-PM) approach with a Fast-HSQC (FHSQC) detection scheme. J Biomol NMR 11(2): 221-6.   Jenuwein, T. and Allis, C. D. (2001). Translating the histone code. Science 293(5532): 1074-80.   Ji, Z., Degerny, C., Vintonenko, N., Deheuninck, J., Foveau, B., Leroy, C., Coll, J., Tulasne, D., Baert, J. L. and Fafeur, V. (2007). Regulation of the Ets-1 transcription factor by sumoylation and ubiquitinylation. Oncogene 26(3): 395- 406.   Kaelin, W. G. (2005). Proline hydroxylation and gene expression. Annu Rev Biochem 74: 115-28.   Kalkhoven, E. (2004). CBP and p300: HATs for different occasions. Biochem Pharmacol 68(6): 1145-55.   Kallio, P. J., Okamoto, K., O'Brien, S., Carrero, P., Makino, Y., Tanaka, H. and Poellinger, L. (1998). Signal transduction in hypoxic cells: inducible nuclear translocation and recruitment of the CBP/p300 coactivator by the hypoxia- inducible factor-1alpha. EMBO J 17(22): 6573-86.   Kamei, Y., Xu, L., Heinzel, T., Torchia, J., Kurokawa, R., Gloss, B., Lin, S. C., Heyman, R. A., Rose, D. W., Glass, C. K. and Rosenfeld, M. G. (1996). A CBP integrator References 	
   	
   190	
   complex mediates transcriptional activation and AP-1 inhibition by nuclear receptors. Cell 85(3): 403-14.   Kang, H.-S. (2005). Structural and functional studies of the N-terminal domain of GABPalpha. MSC Dissertation (University of British Columbia).   Kang, H. S., Nelson, M. L., Mackereth, C. D., Scharpf, M., Graves, B. J. and McIntosh, L. P. (2008). Identification and structural characterization of a CBP/p300-binding domain from the ETS family transcription factor GABP alpha. J Mol Biol 377(3): 636-46.   Kasper, L. H., Fukuyama, T., Biesen, M. A., Boussouar, F., Tong, C., de Pauw, A., Murray, P. J., van Deursen, J. M. and Brindle, P. K. (2006). Conditional knockout mice reveal distinct functions for the global transcriptional coactivators CBP and p300 in T-cell development. Mol Cell Biol 26(3): 789-809.   Kay, L. E., Torchia, D. A. and Bax, A. (1989). Backbone dynamics of proteins as studied by 15N inverse detected heteronuclear NMR spectroscopy: application to staphylococcal nuclease. Biochemistry 28(23): 8972-9.   Kim, C. A., Phillips, M. L., Kim, W., Gingery, M., Tran, H. H., Robinson, M. A., Faham, S. and Bowie, J. U. (2001). Polymerization of the SAM domain of TEL in leukemogenesis and transcriptional repression. EMBO J 20(15): 4173-82.   Klein, F. A., Atkinson, R. A., Potier, N., Moras, D. and Cavarelli, J. (2005). Biochemical and NMR mapping of the interface between CREB-binding protein and ligand binding domains of nuclear receptor: beyond the LXXLL motif. J Biol Chem 280(7): 5682-92.   Koradi, R., Billeter, M. and Wuthrich, K. (1996). MOLMOL: a program for display and analysis of macromolecular structures. J Mol Graph 14(1): 51-5, 29-32.   Kouzarides, T. (2002). Histone methylation in transcriptional control. Curr Opin Genet Dev 12(2): 198-209.   Kruse, J. P. and Gu, W. (2009). Modes of p53 regulation. Cell 137(4): 609-22.   References 	
   	
   191	
   Kwok, R. P., Lundblad, J. R., Chrivia, J. C., Richards, J. P., Bachinger, H. P., Brennan, R. G., Roberts, S. G., Green, M. R. and Goodman, R. H. (1994). Nuclear protein CBP is a coactivator for the transcription factor CREB. Nature 370(6486): 223-6.   Larkin, M. A., Blackshields, G., Brown, N. P., Chenna, R., McGettigan, P. A., McWilliam, H., Valentin, F., Wallace, I. M., Wilm, A., Lopez, R., Thompson, J. D., Gibson, T. J. and Higgins, D. G. (2007). Clustal W and Clustal X version 2.0. Bioinformatics 23(21): 2947-8.   Larsen, M. R., Thingholm, T. E., Jensen, O. N., Roepstorff, P. and Jorgensen, T. J. (2005). Highly selective enrichment of phosphorylated peptides from peptide mixtures using titanium dioxide microcolumns. Mol Cell Proteomics 4(7): 873-86.   Laskowski, R. A., Rullmannn, J. A., MacArthur, M. W., Kaptein, R. and Thornton, J. M. (1996). AQUA and PROCHECK-NMR: programs for checking the quality of protein structures solved by NMR. J Biomol NMR 8(4): 477-86.   Lee, C. W., Arai, M., Martinez-Yamout, M. A., Dyson, H. J. and Wright, P. E. (2009). Mapping the interactions of the p53 transactivation domain with the KIX domain of CBP. Biochemistry 48(10): 2115-24.   Lee, G. M., Donaldson, L. W., Pufall, M. A., Kang, H. S., Pot, I., Graves, B. J. and McIntosh, L. P. (2005). The structural and dynamic basis of Ets-1 DNA binding autoinhibition. J Biol Chem 280(8): 7088-99.   Lee, G. M., Pufall, M. A., Meeker, C. A., Kang, H. S., Graves, B. J. and McIntosh, L. P. (2008). The affinity of Ets-1 for DNA is modulated by phosphorylation through transient interactions of an unstructured region. J Mol Biol 382(4): 1014-30.   Lee, H., Mok, K. H., Muhandiram, R., Park, K. H., Suk, J. E., Kim, D. H., Chang, J., Sung, Y. C., Choi, K. Y. and Han, K. H. (2000). Local structural elements in the mostly unstructured transcriptional activation domain of human p53. J Biol Chem 275(38): 29426-32.   Legge, G. B., Martinez-Yamout, M. A., Hambly, D. M., Trinh, T., Lee, B. M., Dyson, H. J. and Wright, P. E. (2004). ZZ domain of CBP: an unusual zinc finger fold in a protein interaction module. J Mol Biol 343(4): 1081-93.   References 	
   	
   192	
   Li, M., Pascual, G. and Glass, C. K. (2000). Peroxisome proliferator-activated receptor gamma-dependent repression of the inducible nitric oxide synthase gene. Mol Cell Biol 20(13): 4699-707.   Li, Q. J., Yang, S. H., Maeda, Y., Sladek, F. M., Sharrocks, A. D. and Martins-Green, M. (2003). MAP kinase phosphorylation-dependent activation of Elk-1 leads to activation of the co-activator p300. EMBO J 22(2): 281-91.   Litterst, C. M. and Pfitzner, E. (2002). An LXXLL motif in the transactivation domain of STAT6 mediates recruitment of NCoA-1/SRC-1. J Biol Chem 277(39): 36052-60.   Lonard, D. M., Kumar, R. and O'Malley, B. W. (2009). Minireview: the SRC family of coactivators: an entree to understanding a subset of polygenic diseases? Mol Endocrinol 24(2): 279-85.   Lonard, D. M. and O'Malley B, W. (2007). Nuclear receptor coregulators: judges, juries, and executioners of cellular regulation. Mol Cell 27(5): 691-700.   Lonard, D. M. and O'Malley, B. W. (2005). Expanding functional diversity of the coactivators. Trends Biochem Sci 30(3): 126-32.   Macauley, M. S., Errington, W. J., Scharpf, M., Mackereth, C. D., Blaszczak, A. G., Graves, B. J. and McIntosh, L. P. (2006). Beads-on-a-string, characterization of ETS-1 sumoylated within its flexible N-terminal sequence. J Biol Chem 281(7): 4164-72.   Macian, F., Lopez-Rodriguez, C. and Rao, A. (2001). Partners in transcription: NFAT and AP-1. Oncogene 20(19): 2476-89.   Mackereth, C. D. (2003). STRUCTURAL AND FUNCTIONAL CHARACTERIZATION OF THE PNT AND SAM DOMAINS: DIVERSE ROLES IN PROTEIN-PROTEIN INTERACTIONS. PHD Dissertation (University of British Columbia).   Mackereth, C. D., Scharpf, M., Gentile, L. N., MacIntosh, S. E., Slupsky, C. M. and McIntosh, L. P. (2004). Diversity in structure and function of the Ets family PNT domains. J Mol Biol 342(4): 1249-64.   References 	
   	
   193	
   Marmorstein, R. and Berger, S. L. (2001). Structure and function of bromodomains in chromatin-regulating complexes. Gene 272(1-2): 1-9.   Marsh, J. A., Singh, V. K., Jia, Z. and Forman-Kay, J. D. (2006). Sensitivity of secondary structure propensities to sequence differences between alpha- and gamma-synuclein: implications for fibrillation. Protein Sci 15(12): 2795-804.   Matsuno, H., Furusawa, H. and Okahata, Y. (2004). Kinetic study of phosphorylation- dependent complex formation between the kinase-inducible domain (KID) of CREB and the KIX domain of CBP on a quartz crystal microbalance. Chemistry 10(23): 6172-8.   McIntosh, L. P., Kang, H. S., Okon, M., Nelson, M. L., Graves, B. J. and Brutscher, B. (2009). Detection and assignment of phosphoserine and phosphothreonine residues by (13)C- (31)P spin-echo difference NMR spectroscopy. J Biomol NMR 43(1): 31-7.   McManus, K. J. and Hendzel, M. J. (2001). CBP, a transcriptional coactivator and acetyltransferase. Biochem Cell Biol 79(3): 253-66.   Merika, M. and Thanos, D. (2001). Enhanceosomes. Curr Opin Genet Dev 11(2): 205-8.   Mittnacht, S. (1998). Control of pRB phosphorylation. Curr Opin Genet Dev 8(1): 21-7.   Nelson, M. L. (2007). SIGNALING AND AUTOINHIBITION DEPENDENT REGULATION OF THE TRANSCRIPTIONAL COACTIVATOR CBP: LESSONS FROM ETS-1 AND ETS-2. PHD Dissertation (University of Utah).   Nolte, R. T., Wisely, G. B., Westin, S., Cobb, J. E., Lambert, M. H., Kurokawa, R., Rosenfeld, M. G., Willson, T. M., Glass, C. K. and Milburn, M. V. (1998). Ligand binding and co-activator assembly of the peroxisome proliferator-activated receptor-gamma. Nature 395(6698): 137-43.   O'Neill, E. M., Rebay, I., Tjian, R. and Rubin, G. M. (1994). The activities of two Ets- related transcription factors required for Drosophila eye development are modulated by the Ras/MAPK pathway. Cell 78(1): 137-47.   References 	
   	
   194	
   O'Shea, E. K., Rutkowski, R. and Kim, P. S. (1992). Mechanism of specificity in the Fos- Jun oncoprotein heterodimer. Cell 68(4): 699-708.   Orphanides, G., Lagrange, T. and Reinberg, D. (1996). The general transcription factors of RNA polymerase II. Genes Dev 10(21): 2657-83.   Ottiger, M., Delaglio, F. and Bax, A. (1998). Measurement of J and dipolar couplings from simplified two-dimensional NMR spectra. J Magn Reson 131(2): 373-8.   Pace, C. N., Vajdos, F., Fee, L., Grimsley, G. and Gray, T. (1995). How to measure and predict the molar absorption coefficient of a protein. Protein Sci 4(11): 2411-23.   Panning, B. and Taatjes, D. J. (2008). Transcriptional regulation: it takes a village. Mol Cell 31(5): 622-9.   Pawson, T. and Scott, J. D. (2005). Protein phosphorylation in signaling--50 years and counting. Trends Biochem Sci 30(6): 286-90.   Plevin, M. J., Mills, M. M. and Ikura, M. (2005). The LxxLL motif: a multifunctional binding sequence in transcriptional regulation. Trends Biochem Sci 30(2): 66-9.   Ponting, C. P. (1995). SAM: a novel motif in yeast sterile and Drosophila polyhomeotic proteins. Protein Sci 4(9): 1928-30.   Pufall, M. A. and Graves, B. J. (2002). Autoinhibitory domains: modular effectors of cellular regulation. Annu Rev Cell Dev Biol 18: 421-62.   Pufall, M. A., Lee, G. M., Nelson, M. L., Kang, H. S., Velyvis, A., Kay, L. E., McIntosh, L. P. and Graves, B. J. (2005). Variable control of Ets-1 DNA binding by multiple phosphates in an unstructured region. Science 309(5731): 142-5.   Pullen, K., Rajagopal, P., Branchini, B. R., Huffine, M. E., Reizer, J., Saier, M. H., Jr., Scholtz, J. M. and Klevit, R. E. (1995). Phosphorylation of serine-46 in HPr, a key regulatory protein in bacteria, results in stabilization of its solution structure. Protein Sci 4(12): 2478-86.   References 	
   	
   195	
   Qiao, F., Harada, B., Song, H., Whitelegge, J., Courey, A. J. and Bowie, J. U. (2006). Mae inhibits Pointed-P2 transcriptional activity by blocking its MAPK docking site. EMBO J 25(1): 70-9.   Qiao, F., Song, H., Kim, C. A., Sawaya, M. R., Hunter, J. B., Gingery, M., Rebay, I., Courey, A. J. and Bowie, J. U. (2004). Derepression by depolymerization; structural insights into the regulation of Yan by Mae. Cell 118(2): 163-73.   Qin, B. Y., Liu, C., Lam, S. S., Srinath, H., Delston, R., Correia, J. J., Derynck, R. and Lin, K. (2003). Crystal structure of IRF-3 reveals mechanism of autoinhibition and virus-induced phosphoactivation. Nat Struct Biol 10(11): 913-21.   Rabault, B. and Ghysdael, J. (1994). Calcium-induced phosphorylation of ETS1 inhibits its specific DNA binding activity. J Biol Chem 269(45): 28143-51.   Radhakrishnan, I., Perez-Alvarado, G. C., Parker, D., Dyson, H. J., Montminy, M. R. and Wright, P. E. (1997). Solution structure of the KIX domain of CBP bound to the transactivation domain of CREB: a model for activator:coactivator interactions. Cell 91(6): 741-52.   Rieping, W., Habeck, M., Bardiaux, B., Bernard, A., Malliavin, T. E. and Nilges, M. (2007). ARIA2: automated NOE assignment and data integration in NMR structure calculation. Bioinformatics 23(3): 381-2.   Roeder, R. G. (2003). Lasker Basic Medical Research Award. The eukaryotic transcriptional machinery: complexities and mechanisms unforeseen. Nat Med 9(10): 1239-44.   Roeder, R. G. (2005). Transcriptional regulation and the role of diverse coactivators in animal cells. FEBS Lett 579(4): 909-15.   Sakamoto, Y., Yoshida, M., Semba, K. and Hunter, T. (1997). Inhibition of the DNA- binding and transcriptional repression activity of the Wilms' tumor gene product, WT1, by cAMP-dependent protein kinase-mediated phosphorylation of Ser-365 and Ser-393 in the zinc finger domain. Oncogene 15(17): 2001-12.   Sattler, M., Schleucher, J. and Griesinger, C. (1999). Heteronuclear multidimensional NMR experiments for the structure determination of proteins in solution References 	
   	
   196	
   employing pulsed field gradients. Progress in Nuclear Magnetic Resonance Spectroscopy 34(2): 93-158.   Seidel, J. J. and Graves, B. J. (2002). An ERK2 docking site in the Pointed domain distinguishes a subset of ETS transcription factors. Genes Dev 16(1): 127-37.   Seo, M. D., Park, S. J., Kim, H. J. and Lee, B. J. (2007). Identification of the WW domain-interaction sites in the unstructured N-terminal domain of EBV LMP 2A. FEBS Lett 581(1): 65-70.   Sharrocks, A. D. (2001). The ETS-domain transcription factor family. Nat Rev Mol Cell Biol 2(11): 827-37.   Sheppard, H. M., Harries, J. C., Hussain, S., Bevan, C. and Heery, D. M. (2001). Analysis of the steroid receptor coactivator 1 (SRC1)-CREB binding protein interaction interface and its importance for the function of SRC1. Mol Cell Biol 21(1): 39-50.   Shi, X., Hong, T., Walter, K. L., Ewalt, M., Michishita, E., Hung, T., Carney, D., Pena, P., Lan, F., Kaadige, M. R., Lacoste, N., Cayrou, C., Davrazou, F., Saha, A., Cairns, B. R., Ayer, D. E., Kutateladze, T. G., Shi, Y., Cote, J., Chua, K. F. and Gozani, O. (2006). ING2 PHD domain links histone H3 lysine 4 methylation to active gene repression. Nature 442(7098): 96-9.   Shiau, A. K., Barstad, D., Loria, P. M., Cheng, L., Kushner, P. J., Agard, D. A. and Greene, G. L. (1998). The structural basis of estrogen receptor/coactivator recognition and the antagonism of this interaction by tamoxifen. Cell 95(7): 927- 37.   Slupsky, C. M., Gentile, L. N., Donaldson, L. W., Mackereth, C. D., Seidel, J. J., Graves, B. J. and McIntosh, L. P. (1998). Structure of the Ets-1 pointed domain and mitogen-activated protein kinase phosphorylation site. Proc Natl Acad Sci U S A 95(21): 12129-34.   Song, H., Nie, M., Qiao, F., Bowie, J. U. and Courey, A. J. (2005). Antagonistic regulation of Yan nuclear export by Mae and Crm1 may increase the stringency of the Ras response. Genes Dev 19(15): 1767-72.   References 	
   	
   197	
   Stacey, K. J., Fowles, L. F., Colman, M. S., Ostrowski, M. C. and Hume, D. A. (1995). Regulation of urokinase-type plasminogen activator gene transcription by macrophage colony-stimulating factor. Mol Cell Biol 15(6): 3430-41.   Stapleton, D., Balan, I., Pawson, T. and Sicheri, F. (1999). The crystal structure of an Eph receptor SAM domain reveals a mechanism for modular dimerization. Nat Struct Biol 6(1): 44-9.   Sterner, D. E. and Berger, S. L. (2000). Acetylation of histones and transcription-related factors. Microbiol Mol Biol Rev 64(2): 435-59.   Svejstrup, J. Q., Vichi, P. and Egly, J. M. (1996). The multiple roles of transcription/repair factor TFIIH. Trends Biochem Sci 21(9): 346-50.   Syntichaki, P., Topalidou, I. and Thireos, G. (2000). The Gcn5 bromodomain co- ordinates nucleosome remodelling. Nature 404(6776): 414-7.   Tarrant, M. K. and Cole, P. A. (2009). The chemical biology of protein phosphorylation. Annu Rev Biochem 78: 797-825.   Teufel, D. P., Freund, S. M., Bycroft, M. and Fersht, A. R. (2007). Four domains of p300 each bind tightly to a sequence spanning both transactivation subdomains of p53. Proc Natl Acad Sci U S A 104(17): 7009-14.   Thanos, C. D., Goodwill, K. E. and Bowie, J. U. (1999). Oligomeric structure of the human EphB2 receptor SAM domain. Science 283(5403): 833-6.   Tootle, T. L., Lee, P. S. and Rebay, I. (2003). CRM1-mediated nuclear export and regulated activity of the Receptor Tyrosine Kinase antagonist YAN require specific interactions with MAE. Development 130(5): 845-57.   Tootle, T. L. and Rebay, I. (2005). Post-translational modifications influence transcription factor activity: a view from the ETS superfamily. Bioessays 27(3): 285-98.   Vo, N. and Goodman, R. H. (2001). CREB-binding protein and p300 in transcriptional regulation. J Biol Chem 276(17): 13505-8.  References 	
   	
   198	
    Waas, W. F. and Dalby, K. N. (2002). Transient protein-protein interactions and a random-ordered kinetic mechanism for the phosphorylation of a transcription factor by extracellular-regulated protein kinase 2. J Biol Chem 277(15): 12532- 40.   Walsh, C. T., Garneau-Tsodikova, S. and Gatto, G. J., Jr. (2005). Protein posttranslational modifications: the chemistry of proteome diversifications. Angew Chem Int Ed Engl 44(45): 7342-72.   Wang, L., Tang, Y., Cole, P. A. and Marmorstein, R. (2008). Structure and chemistry of the p300/CBP and Rtt109 histone acetyltransferases: implications for histone acetyltransferase evolution and function. Curr Opin Struct Biol 18(6): 741-7.   Waterman, M. J., Stavridi, E. S., Waterman, J. L. and Halazonetis, T. D. (1998). ATM- dependent activation of p53 involves dephosphorylation and association with 14- 3-3 proteins. Nat Genet 19(2): 175-8.   Whitmarsh, A. J. and Davis, R. J. (2000). Regulation of transcription factor function by phosphorylation. Cell Mol Life Sci 57(8-9): 1172-83.   Wojciak, J. M., Martinez-Yamout, M. A., Dyson, H. J. and Wright, P. E. (2009). Structural basis for recruitment of CBP/p300 coactivators by STAT1 and STAT2 transactivation domains. EMBO J 28(7): 948-58.   Yaffe, M. B. (2002). How do 14-3-3 proteins work?-- Gatekeeper phosphorylation and the molecular anvil hypothesis. FEBS Lett 513(1): 53-7.   Yang, B. S., Hauser, C. A., Henkel, G., Colman, M. S., Van Beveren, C., Stacey, K. J., Hume, D. A., Maki, R. A. and Ostrowski, M. C. (1996). Ras-mediated phosphorylation of a conserved threonine residue enhances the transactivation activities of c-Ets1 and c-Ets2. Mol Cell Biol 16(2): 538-47.   Yang, C., Shapiro, L. H., Rivera, M., Kumar, A. and Brindle, P. K. (1998a). A role for CREB binding protein and p300 transcriptional coactivators in Ets-1 transactivation functions. Mol Cell Biol 18(4): 2218-29.   Yang, D. W., Tolman, J. R., Goto, N. K. and Kay, L. E. (1998b). An HNCO-based pulse scheme for the measurement of C-13(alpha)-H-1(alpha) one-bond dipolar References 	
   	
   199	
   couplings in N-15, C-13 labeled proteins. Journal of Biomolecular NMR 12(2): 325-332.   Yang, S. H., Jaffray, E., Hay, R. T. and Sharrocks, A. D. (2003). Dynamic interplay of the SUMO and ERK pathways in regulating Elk-1 transcriptional activity. Mol Cell 12(1): 63-74.   Yang, S. H., Shore, P., Willingham, N., Lakey, J. H. and Sharrocks, A. D. (1999). The mechanism of phosphorylation-inducible activation of the ETS-domain transcription factor Elk-1. EMBO J 18(20): 5666-74.   Yang, X. J. (2004). Lysine acetylation and the bromodomain: a new partnership for signaling. Bioessays 26(10): 1076-87.   Yang, X. J. and Seto, E. (2007). HATs and HDACs: from structure, function and regulation to novel strategies for therapy and prevention. Oncogene 26(37): 5310-8.   Yang, X. J. and Seto, E. (2008). Lysine acetylation: codified crosstalk with other posttranslational modifications. Mol Cell 31(4): 449-61.   Zandi, E. and Karin, M. (1999). Bridging the gap: composition, regulation, and physiological function of the IkappaB kinase complex. Mol Cell Biol 19(7): 4547- 51.   Zhang, Y. Z. (1995). Protein and peptide structure and interactions studied by hydrogen exchange and NMR. Ph.D. Thesis.   Zor, T., De Guzman, R. N., Dyson, H. J. and Wright, P. E. (2004). Solution structure of the KIX domain of CBP bound to the transactivation domain of c-Myb. J Mol Biol 337(3): 521-34.   Zwahlen, C., Vincent, S. J. and Kay, L. E. (1998). Analytical description of the effect of adiabatic pulses on IS, I2S, and I3S spin systems. J Magn Reson 130(2): 169- 75.    	
  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0071341/manifest

Comment

Related Items