Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Structural and functional characterization of the PNT and SAM domains : diverse roles in protein-protein… Mackereth, Cameron David 2003

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

Item Metadata


831-ubc_2003-859975.pdf [ 22.4MB ]
JSON: 831-1.0091208.json
JSON-LD: 831-1.0091208-ld.json
RDF/XML (Pretty): 831-1.0091208-rdf.xml
RDF/JSON: 831-1.0091208-rdf.json
Turtle: 831-1.0091208-turtle.txt
N-Triples: 831-1.0091208-rdf-ntriples.txt
Original Record: 831-1.0091208-source.json
Full Text

Full Text

STRUCTURAL AND FUNCTIONAL CHARACTERIZATION OF THE PNT AND SAM DOMAINS: DIVERSE ROLES IN PROTEIN-PROTEIN INTERACTIONS by CAMERON DAVID MACKERETH  B.Sc, University of Waterloo, 1996 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Biochemistry and Molecular Biology)  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA July 2003 © Cameron David Mackereth, 2003  In  presenting  degree freely  this  at the  thesis  in  University of  partial  fulfilment  of  British Columbia, I agree  available for reference and study. I further  copying of  this  department  or  publication of  the  his  or  her  representatives.  that the  of  It  is  granted  DE-6 (2/88)  ZS  advanced  by the  understood  that  extensive  head of copying  my or  this thesis for financial gain shall not be allowed without my written  6 i P C H 6 M l S T E y fMb  The University of British Columbia Vancouver, Canada  Date  an  Library shall make it  permission.  Department  for  agree that permission for  thesis for scholarly purposes may be by  requirements  CTuLV  2001  (VI0LEQAlAfcloU>&y  Abstract The PNT (or Pointed) domain, and highly related SAM (sterile alpha motif) domain, are a conserved region of 70 to 85 amino acids that is found in a diverse range of eukaryotic proteins including the Ets family of transcription factors and components of the yeast pheromone-response pathway. These domains, which can be expressed and purified in isolation, are structurally characterized by a core bundle of four a-helices. Additional helices in specific variants highlight plasticity in the architecture of these domains. The PNT domain structure from the Ets protein Erg, determined by using nuclear magnetic resonance (NMR) spectroscopy, reveals that only the four core helices are present; in contrast, the Ets proteins Ets-1 and G A B P a have PNT domains that display an intimately associated N-terminal helix. The generalized role of the PNT domain as a self-association module has been investigated. Biophysical characterization of various PNT domains reveals a diverse range of oligomeric states, with those from Ets members Ets-1, Ets-2, Pnt-P2, GABPa, Erg and Fli-1 being monomeric in solution, whereas Tel and Yan are polymeric. In contrast, Ste4 and Byr2 of the Schizosaccharomyces  pombe  mating pathway form an  unexpected 3:1 complex dependent on the interaction of their respective SAM domains coupled with trimerization of Ste4 via an adjacent coiled-coil region. For the Ets-1 and GABPa PNT domains, an alternative role in providing a mitogen activated protein (MAP) kinase docking site has been investigated. These docking sites increase the specificity of MAP kinase-mediated phosphorylation of phosphoacceptor sites N- and C-terminal to the PNT domains of Ets-1 and GABPa, respectively.  ii  The residues flanking the PNT domain in the Ets family transcription factors have additional importance with respect to post-translational modification and protein association. Limited proteolysis and NMR-based relaxation studies confirm that the peptide regions adjacent to the PNT domains of Ets-1, G A B P a and Erg are conformationally disordered in solution. In addition, phosphorylation of threonine 38 in Ets-1 or threonine 280 in GABPa does not significantly change the dynamic nature of these regions, nor does this modification perturb the structure of the adjoining PNT domains. For Ets-1, a consensus sumoylation site has also been identified on lysine 15 within its disordered N-terminus. Using chemical shift perturbation mapping, the binding interface on both Ets-1 and the small ubiquitin-like modifier (SUMO)-conjugating enzyme 9 (UBC9) were identified. This latter surface region coincides with that site of binding by previously characterized sumoylation targets. No direct interaction between Ets-1 and SUMO was detected, highlighting the role of UBC9 in mediating the chemical linkage of these two proteins. The biological consequence of this modification is still unclear, although sumoylation of the Ets family member Tel has been reported to differentially target the protein to sub-nuclear assemblies. However, in vitro analyses with Tel indicate that a previously proposed SUMO-acceptor lysine within its PNT domain does not appear to interact with UBC9. Rather, an N-terminal site, corresponding to the conserved sequence in Ets-1, may reflect a more prominent sumoylation site for Tel.  iii  TABLE OF CONTENTS Abstract  ii  Table of Contents  iv  List of Tables  vii  List of Figures  viii  Abbreviations  xi  Chapter 1 Introduction  1  1.1 Modular proteins  •  2  1.2 The PNT domain and SAM domain  3  1.3 PNT and SAM domain structure  8  1.4 Role of the PNT and SAM domains in protein association  11  1.5 Ets family of transcription factors  14  1.5.1 Ets sub-family  19  1.5.2 Erg/Fli-1 sub-family  20  1.5.3 GABP sub-family  21  1.5.4 Tel sub-family  23  1.5.5 Ets transcription factors and ERK2 phosphorylation  25  1.5.6 Ets proteins as targets for sumoylation  26  1.6 Yeast pheromone response pathway  30  1.6.1 Ste4  33  1.6.2 Byr2  34  1.7 Thesis overview  35  1.8 Publications  36  Chapter 2 Structural studies of the Erg PNT domain  39  2.1 Characterization of Erg constructs  40  2.2 Chemical shift assignment  42  2.3 E r g  46  ( 1 0 8 2 0 1 )  secondary structure iv  2.4 E r g  (108  "  201)  tertiary structure  51  2.5 Structural comparison to other PNT domain  54  2.6 Screen for in vitro association  58  2.7 Materials and methods  66  2.7.1 PNT domain cloning  66  2.7.2 Protein purification  68  2.7.3 Trypsin digest  70  2.7.4 NMR spectroscopy  70  2.7.5 Dipolar coupling measurement  71  2.7.6 Structure calculation  72  Chapter 3 Heteromeric complex of Ste4 and Byr2  74  3.1 Previous evidence of Byr2 and Ste4 interaction  75  3.2 Byr2 binding to trimeric Ste4  75  3.3 Coiled-coil of Ste4  88  3.4 Biological implications  96  3.5 Materials and methods  102  3.5.1 Cloning and purification  102  3.5.2 Circular dichroism spectroscopy  104  3.5.3 Native gel electrophoresis  104  3.5.4 Trypin digestion  104  3.5.5 NMR spectroscopy  105  Chapter 4 PNT domain as an ERK2 docking motif  106  4.1 Ets-1 PNT domain and MAPK docking  107  4.2 NMR titration of Ets-1 with ERK2  110  4.3 NMR titration of GABPa with ERK2  120  4.4 Role of PNT domain in ERK2 docking  130  4.5 Materials and methods  133  4.5.1 Cloning and purification...  133  4.5.2 NMR spectroscopy  134  4.5.3 Binding site determination  135 v  4.5.4 Relaxation analysis  135  4.5.5 In vitro phosphorylation  136  Chapter 5 Interactions of Ets proteins with UBC9 and SUMO  137  5.1 Evidence for UBC9-mediated sumoylation of PNT domains  138  5.2 Titration of TEL with UBC9  141  5.3 Titration of Ets-1 with UBC9 and SUMO  143  5.4 Binding surface on UBC9  149  5.5 Biological context  157  5.6 Materials and methods  163  5.6.1 Cloning and purification  163  5.6.2 Cross-linking  165  5.6.3 NMR spectroscopy  165  5.6.5 Determination of dissociation constant  165  5.6.5 Sumoylation assay  166  Chapter 6 Concluding remarks  167  6.1 Common and divergent aspects of the PNT and SAM domains  168  6.2 Surface areas utilized in association  169  6.3 Structured domains and flexible linkers  171  References  177  Appendix 1  199  Appendix 2  202  vi  List of Tables Table 1.1 Reported PNT domain interactions with non-Ets proteins  14  Table 2.1 NMR restraints and statistics for the ensemble often structures calculated for E r g  52  Table 3.1 Thermal denaturation values and secondary structure composition of the SAM domain fragments of Ste4 and Byr2  78  (108_201)  vii  List of Figures Figure 1.1 Multidomain composition of PNT and SAM domain-containing proteins  4  Figure 1.2 Sequence alignment of PNT and SAM domains  6  Figure 1.3 Tertiary structures of PNT and SAM domains  9  Figure 1.4 Domain composition for a subset of Ets family transcription factors  16  Figure 1.5 Phylogenetic alignment of Ets family PNT domains  17  Figure 1;6 Generic models of transcriptional activation and repression  27  Figure 1.7 Sumoylation pathway  28  Figure 1.8 Signaling proteins from yeast  31  Figure 2.1 Proteolysis of Erg " domain  41  (1  201)  to determine minimal-sized folded PNT  Figure 2.2 Superimposition of [ N]Erg 15  (1  201)  and [ N]Erg 15  (108  "  201) 1  H- N 15  HSQC spectra  43  Figure 2.3 Annotated H - N HSQC spectrum of [ N]Erg 1  15  15  (108  "  201)  45  Figure 2.4 NMR-derived secondary structure of E r g " 48 Figure 2.5 Conservation of C secondary chemical shift for Ets family PNT domains 50 (108  1 3  Figure 2.6 E r g  (108  "  201)  201)  a  consists of a bundle of four a-helices  53  Figure 2.7 Erg PNT domain lacks the integral N-terminal helix (H1) found in Ets-1 and GABPa 55 Figure 2.8 Structure similarity between several PNT and SAM domains  57  Figure 2.9 PNT domain constructs used in the study of Erg and the in vitro tests of association 59 Figure 2.10 1D H-NMR of selected PNT domain constructs 1  60  Figure 2.11 Test for heteromeric complex formation by using glutaraldehyde cross-linking 63 viii  Figure 2.12 Erg PNT domain lacks the oligomerization interfaces determined for Tel  65  Figure 3.1 Constructs used to study the SAM domain-mediated association between Byr2 and Ste4, and the trimeric coiled-coil of Ste4 76 Figure 3.2 Byr2 " is partially folded in solution  79  Figure 3.3 Ste4 " is folded in solution, but exists as an oligomer or aggregate  80  Figure 3.4 Coiled-coil predictions for full-length Ste4  82  Figure 3.5 Ste4 "  forms a stable trimer in solution  84  Figure 3.6 Native gel electrophoresis assay to determine stoichiometry of the Ste4/Byr2 complex  86  Figure 3.7 Structural characterization of the binding of Byr2 " to trimeric Ste4 "  89  Figure 3.8 S t e 4 "  157)  and Ste4 "  90  Figure 3.9 S t e 4 "  157)  is trimeric and flexible in solution  (1  70)  (1  (1  77)  157)  (1  (1  70)  157)  (66  (66  (52  157)  form large folded complexes  92  Figure 3.10 Sequence analysis of the Ste4 coiled-coil region  93  Figure 3.11 Screen to optimize Ste4 "  97  (52  157)  for NMR spectroscopy  Figure 3.12 Schematic representation of the Ste4:Byr2 complex  98  Figure 3.13 Potential biological role for Ste4  101  Figure 4.1 Schematic of a MAPK docking domain  108  Figure 4.2 H - N HSQC monitored titration of [ N]Ets-1 unlabeled active ERK2 1  15  15  (29  138)  with 111  Figure 4.3 H - C HNCO monitored titration of [ C, N]Ets-1 unlabeled active ERK2 1  13  13  Figure 4.4 ERK2 docking site on Ets-1 " (29  138)  15  (29  138)  with 114  structure  115  Figure 4.5 Cartoon model of Ets-1 PNT domain docking to ERK2 Figure 4.6 H - N HSQC monitored titration of [ N ] G A B P a unlabeled active ERK2 1  15  15  ix  (168  "  254)  119 with 121  Figure 4.7 Additional C-terminal residues in G A B P a the structure of the PNT domain Figure 4.8 Additional C-terminal residues in G A B P a Figure 4.9 Phosphorylation of G A B P a domain structure  (168  "  290)  v  ( 1 6 8  1  d o not perturb 124  "  2 9 0 )  are flexible  127  at Thr280 does not affect PNT 129  Figure 4.10 MAP kinase phosphorylation sites adjacent to Ets family PNT domains 132 Figure 5.1 Glutaraldehyde crosslinking of UBC9 to SUMO and Ets PNT domains  139  Figure 5.2 SUMO consensus motifs in the Ets family of transcription factors  144  Figure 5.3 Titration of [ N]Ets-1 with UBC9  146  15  Figure 5.4 Determination of K between UBC9 and Ets-1 " (1  138)  d  Figure 5.5 Structure of UBC9  147 150  Figure 5.6 Titration of [ N]UBC9 with unlabeled Ets-1 " Figure 5.7 Titration of [ N]UBC9- unlabeled E t s - 1 with unlabeled SUMO-1  152  Figure 5.8 Titration of [ N]UBC9 with unlabeled RanGAPI  158  15  15  (1  138)  (1_138)  15  Figure 5.9 In vitro sumoylation of Ets family proteins  155  161  Figure 6.1 Diverse modes of protein association exhibited by the PNT and SAM domains 170 Figure 6.2 Model of Ets-1 involving structured domains and flexible linkers 173  Abbreviations 1D  one-dimensional  2D  two-dimensional  3D  three-dimensional  AMP-PNP  5'-adenylylimidodiphosphate  ATP  adenosine 5'-triphosphate  BS  bis(sulfosuccinimidyl) suberate  3  CBD  catalytic binding domain  CD  circular dichroism  COSY  correlation spectroscopy  CSI  chemical shift index  CT  constant time  D 0  deuterium oxide  Da  Dalton  DQF  double quantum filtered  DTT  dithiothreitol  E1  SUMO/ubiquitin-activating enzyme  E2  SUMO/ubiquitin-conjugating enzyme  E3  SUMO/ubiquitin-protein ligase  EDTA  ethylenediaminetetraacetic acid  ERK  extracellular signal-regulated kinase  ESI-MS  electrospray ionization mass spectrometry  FPLC  fast protein liquid chromatography (Amersham Biosciences)  2  HAT  histone acetyltransferase  HDAC  histone deacetylase complex  HEPES  N-2-hydroxyethylpiperazine-N'-2-ethanesulphonic acid  HPLC  high pressure liquid chromatography  HMBC  heteronuclear multiple bond correlation  HMQC  heteronuclear multiple quantum coherence  HSQC  heteronuclear single quantum coherence  IPTG K  d  kDa  isopropyl-(3D-thiogalactopyranoside dissociation constant kiloDalton  MALDI-TOF matrix assisted laser desorption ionization time of flight MAP  mitogen activated protein  MAPK  mitogen activated protein kinase  MAPKK  mitogen activated protein kinase kinase  MAPKKK  mitogen activated protein kinase kinase kinase  MW  molecular weight  NMR  nuclear magnetic resonance  NOE  nuclear Overhauser effect  NOESY  nuclear Overhauser effect spectroscopy  PAGE  polyacrylamide gel electrophoresis  PCR  polymerase chain reaction  PDB  protein data bank (  PMSF  phenylmethanesulfonyl fluoride  PNT  pointed  xii  pi  isoelectric point  RA  Ras-associating  RBD  Ras binding domain  rmsd  root mean square deviation  SAM  sterile alpha motif  SDS  sodium dodecyl sulphate  TEMED  N,N,N',N'-tetramethylethylenediamine  TFA  trifluoroacetic acid  T  midpoint unfolding temperature  m  TOCSY  total correlation spectroscopy  Tris  tris(hydroxymethyl)aminomethane  TROSY  transverse-relaxation optimized spectroscopy  xiii  Chapter 1 Introduction  The PNT (pointed) domain, and closely related SAM (sterile alpha motif) domain, is characterized by a core bundle of four a-helices that may be appended by additional helices present in specific variants. In terms of function, an exclusive role in homotypic and heterotypic PNT-PNT (or SAM-SAM) association was initially described for this domain, as exemplified by the oligomerization of the Ets protein Tel, the interaction between Ste4 and Byr2 in the fission yeast mating pathway, and dimerization of the Ephrin receptor SAM domains. However, it is now clear that PNT and SAM domains may also serve alternative roles such as that of a protein-docking site for the ERK2 kinase. In addition, phosphorylation and sumoylation of residues adjacent to the PNT domain in several Ets transcription factors may modify their biological functions.  1  Chapter 1 - Introduction  1.1 Modular proteins Within the eukaryotic proteome, it is common to find large multifunctional proteins composed of a series of structurally and functionally independent domains. In the genesis of these proteins, a co-translational mechanism may exist to allow for sequential folding of discrete modules as they are synthesized by the ribosome (Netzer and Hartl, 1997). Each domain is, to the first approximation, structurally distinct and tethered by flexible linkers. A consequence of this modular composition is the efficient coordination of multiple functions within a single polypeptide. In addition, this combinatorial assembly of domains allows for the facile evolution of new functionality within a given protein through genetic shuffling and chromosomal duplication (Doolittle and Bork, 1993; Doolittle, 1995). Fortunately for the structural biologist, many protein domains can be expressed in isolation of the full-length gene product as soluble, stably folded polypeptides that retain their specific biological functions. It is by using such a reductionist technique that our laboratory has been able to investigate, at the molecular level, several domains of diverse cellular function within the Ets family of transcription factors. The benefit of this approach is three-fold. First, the small size of an isolated protein module typically resides within the ideal size limits of <15 kDa amenable to our chosen investigative technique of nuclear magnetic resonance (NMR) spectroscopy. Second, the function of the domain can be dissected from any additional functions that may be present on the full-length host proteins. Finally, domains from additional proteins that share sequence or structural identity to a given module can be easily compared, in isolation from the full-length protein function. In this vein, I have chosen to investigate the structure and function of the 2  Chapter 1 - Introduction conserved PNT (Pointed) and SAM (sterile alpha motif) domains found within a diverse collection of regulatory proteins involved in eukaryotic development (Figure 1.1). This introduction will outline the initial discovery and characterization of this domain, followed by a structural description and a brief summary of the proteins from which the various PNT and SAM domains have been isolated. At the end of the introduction the overall goals and organization of the thesis will be presented.  1.2 The PNT domain and SAM domain  The initial characterization of the Pointed (or PNT) domain coincided with the cloning of the Pointed gene products in Drosophila melanogaster (Klambt, 1993). Through the use of alternative promoter regions, Pointed codes for two Ets transcription factors that contain the hallmark ETS DNA-binding domain (detailed in section 1.5). However, only the Pointed-P2 isoform includes a second region of sequence conservation within its N-terminal region that is shared by a subset of Ets proteins including Ets-1, Ets-2, GABPa, Fli-1 and Erg. In Ets-1, this conserved PNT domain coincided with a previously characterized regulatory region (Schneikert et al., 1992), suggesting a possible role in transactivation of gene expression. Two years later, a bioinformatics approach was used to identify a region of sequence similarity within several yeast and Drosophila proteins involved in development (Ponting, 1995). Four of the identified yeast proteins (Ste4, Ste50, Byr2/Ste8 and STE11) elicited a sterile phenotype when disrupted. Since secondary structural predictions indicated a predominantly a-helical character, the newly discovered conserved region was named the sterile alpha motif (SAM). The SAM domain was subsequently identified in additional proteins involved in a wide range of 3  Chapter 1 - Introduction  Ets family of transcription factors EtS-1  N—  ( E T S ) — C  Yeast pheromone response pathway  kinase  N^KjjJXj^C  Ste4  p53 family of transcription factors p73  N  C  ^  ^  D  B  D  ^  ^  >  (  T  E  T  )  -  ^  ^  C  Polycomb group of transcriptional repressors polyhomeotic N  •//•  Zn  Ephrin receptor tyrosine kinases EphB2 receptor E  p  h  B  2  N  N  (  R  ^  R  ^  -  Q  Figure1.1. Multidomain composition of PNT and SAM domain-containing proteins. All of the protein families discussed in this thesis display a clear linear arrangement of independent structural domains. This property is highlighted above using representative family members: murine Ets-1 (SwissProt accession number P27577), S. pombe Byr2 (P28829), S. pombe Ste4 (P36622), human p73 (015350), Drosophila polyhomeotic (P39769) and human EphB2 receptor (P29323). The PNT and SAM domains (shown in black) are located at various positions within the proteins. Additional domains are shown in white, and have been identified using SMART and Pfam databases within SwissProt ( ETS, ETS DNA-binding domain; RBD, Ras-binding domain; CDB, cataytic-binding domain; kinase, kinase catalytic domain; coil, coiled-coil region; RA, Ras-association domain; DBD, DNA-binding domain; TET, tetramerization domain; Zn, Zinc finger; Fn, fibronectin type-Ill domain; TM, transmembrane region.  4  Chapter 1 - Introduction eukaryotic development and signal transduction pathways, with the important inclusion of members of the Ephrin tyrosine receptor kinase and p53 transcription factor families (Schultz et al., 1997). A suggested role in protein-protein interaction was assigned to the SAM domain based on the inability of the Schizosaccharomyces pombe Ste4 and Byr2 proteins to associate upon deletion or mutation of the SAM region in either protein (Barr et al., 1996). In addition, this domain appeared to mediate homo- and heterotypic association of Polycomb group proteins (Kyba and Brock, 1998; Peterson et al., 1997). Additional comparisons in the literature and through research presented in this thesis have revealed a strong conservation of sequence and structure between the PNT domain of Ets transcription factors and the SAM domain found in more diverse developmental proteins. An alignment of structurally and functionally important PNT and SAM domains is presented in Figure 1.2. Although very similar in their C-terminal portions, it is interesting to note that there is a distinction between PNT and SAM domains that is clearest at the molecular level of specific secondary structure (Section 1.3), particularly at their N-termini. However, there is also a growing amount of evidence that indicates variability in function that extends not merely between the PNT and SAM domains, but within even highly homologous families of these domains. The above distinction noted for PNT and SAM domains will be preserved within this thesis following  the  suggested nomenclature  of the  SMART database  (; Schultz et al., 1998). That is, only members of the Ets family of transcription factors will be designated as belonging to the PNT domain group in recognition of their founding member Pointed-P2. All other representatives will be classified as SAM domains. 5  Chapter 1 - Introduction  Figure 1.2. Sequence alignment of PNT and SAM domains. Alignment was  constructed using ClustalW (, and by visual inspection coupled with structural analysis. Black, dark grey and light grey shading indicates amino acid similarity in 80%, 60% or 40%, respectively, of the sequences shown. Also illustrated are example secondary structures of a PNT domain (mEts-1; PDB 1bqv) and SAM domain (EphB2 receptor; 1b4f). Note that helix H1 is absent in some PNT domains, and in all SAM domains characterized to date. Sequences include PNT domains from murine Ets-1 (mEts-1; SwissProt accession number P27577), murine Ets-2 (mEts-2; P15037), Drosophila Pointed-P2 (dPntP2; P51023), human Erg (hErg; unpublished, based on P11308), murine Fli-1 (mFli-1; P26323), Tel (P97360), Drosophila Yan (dYan; Q01842), and SAM domains from human EphB2 receptor tyrosine kinase (hEphB2; P29323), human EphA4 receptor tyrosine kinase (hEphA4; Q03137), human p73 (hP73; 015350), Drosophila Polyhomeotic (dPh; P39769), and Drosophila Sex combs on midleg (dScm; Q9VHA0).  6  Chapter 1 - Introduction  LO  o  -—- o5 * * + CM .—. — -  -—.. CD ID N ID CD 1*- CD •«cp co C N l C n c O ^ V C N ? ^ ? 0  -r-LOCDODCMi-t-CO L O C O - r - T - T - T - T t C O  c o c n c o V h - a j L o o o a  * * *  T—  r*-  co co  >  cr -Q  -Q  LO X  «  a  a  u  ft H | g QMS|^ j i f r a ,3 H w w O £ ^ a co u |H a co 1  S  C  'co  E o  X  CO X  "O  CM  X «o  i N a o  5  « o  IS»H 0. > 0. u CO l-MO O O P E B 9 N  J  COETTTCO  D Di f ISB5B7  CM  O H O  X  «s « Q  K  D D Q Q  b d S*S >< ft, <t,d ft f i ft a iT5 a 3 H M a J u J > > Ui Ci  n; »; a « Pi « o « o a S « o o > a B pa  B a§ I B ~" H H Bi Oft IK S X O O CO o co co rt. K  ft  gaa *-  N  f  J° J° ? LU UJ £  E  m  s  CD . C  E % E  3  ^ L U | L U CO = E  LU  g  t  T3  CO c/3 W «  w  c/3  CO f-  -P  °  w . c "a "a  lit  °-  e x :  c 'co  E o  T3  <  C7)  Chapter 1 - Introduction  1.3 PNT AND SAM domain structure There are now several structures reported in the literature for PNT and SAM domains, with two additional PNT domain structures appearing within this thesis. Figure 1.3 illustrates several examples of the secondary structures and folded architecture of PNT and SAM domains. This figure details common elements that group all of these structures together, as well as differences that reveal diversity in their tertiary folds. Beginning with the PNT domain (Figure 1.3A), examples already include the NMR-derived solution structures of those from the Ets-1 (Slupsky et al., 1998a) and GABPa (Scharpf et al., in preparation) PNT domain, and the X-ray crystallographicallyderived structure of the domain from Tel (Kim et al., 2001a; Tran et al., 2002). A core bundle of four a-helices is shared between Tel, Ets-1 and GABPa. An additional Nterminal helix present in Ets-1 and GABPa is an integral component of these latter two PNT domains, as evident by NMR relaxation studies and the required presence of this helix for proper protein expression and folding. Whether the lack of helix H1 in Tel and other Ets PNT domains (e.g. Erg) represents true biological diversity, or merely reflects artifacts of crystallization or inappropriate construct size, is one of the questions addressed later in the thesis. The related SAM domain is represented in Figure 1.3B with examples from the ephrin receptor (human EphB2; Thanos et al., 1999b), p53 (human p73; Chi et al., 1999) and Drosophila Polycomb group (polyhomeotic; Kim et al., 2002a) families. A significant conservation in architecture is present amongst these structures with a bundle of a-helices again comprising the core of the protein domain. Although clearly related to the PNT domain, differences include the presence of a small a-helical region  8  Chapter 1 - Introduction  Figure 1.3. Tertiary structures of PNT and SAM domains.  (A) PNT  domains from Tel (PDB 11ky), Ets-1 (1bqv) and G A B P a (unpublished). (B) SAM domains from EphB2 receptor tyrosine kinase (1b4f), p73 (1cok) and polyhomeotic (1kw4). Helices are numbered from H1 to H5 following the architecture of Ets-1, and using boundaries specified in the literature. Note that the N-terminal helix H1 is not present in the Tel PNT domain or in SAM domains. A small extra helix present in SAM domains is designated in this thesis as helix H2'. Figure made using Molscript v2.1.2 (Kraulis, 1991) and Raster3d v2.4b (Merritt and Bacon, 1997).  9  Chapter 1 - Introduction  Chapter 1 - Introduction (helix H2') between helix H3 and the distinctly shortened helix H2 (using the nomenclature derived for Ets-1). Again there is no indication of the additional Nterminal helix H1 that sets Ets1 and GABPa apart from the other PNT and SAM domains.  1.4 Role of the PNT and SAM domain in protein association Nearly coincident with the identification of the PNT and SAM module came evidence suggestive of a protein-protein association function for these domains within signaling pathways and in the repression of transcription. It has already been mentioned that mutation or deletion of the SAM domain in Byr2 and Ste4 prevented interaction of these two components of the sexual differentiation kinase cascade in S. pombe (Barr et al., 1996; Tu et al., 1997). Further investigation of this interaction in Section 3 of this thesis reinforces the validity of this association in vitro, albeit with an unexpected three-to-one stoichiometry between Ste4 and Byr2. The Drosophila Polycomb group of transcriptional repressors provides an additional example of SAM domain interactions. Several protein-protein association regions are involved in maintaining the integrity of large multi-component repression complexes and, in particular, intact SAM domains are required for the interactions of Polyhomeotic and Scm (Peterson et al., 1997; Kyba and Brock, 1998). During the structural determination of the Polyhomeotic SAM domain, an extended homopolymeric structure was observed by electron microscopy (Kim et al., 2002a). This oligomerization  was  preserved following  crystallization, again  confirming  the  association properties of this protein and providing a molecular description of polymerization process: two distinct surface regions of exposed hydrophobic residues 11  Chapter 1 - Introduction join in a head-to-tail manner to propagate the growing chain of SAM domains. A similar mode of polymerization was initially observed for the PNT domain of the Tel transcription factor. Previous biochemical studies had defined the potential of the Tel PNT domain to oligomerize in the context of the native transcriptional repressor, as well as for oncogenic chimeric proteins genetically fused with the catalytic kinase domains from NTRK3, ABL, Jak2 and PDGFRp (Papadopoulos et al., 1995; Golub et al., 1996; Jousset et al., 1997; Lacronique et al., 1997; Ho et al., 1999; Wai et al., 2000; Kim et al., 2001a; Tognon et al., in preparation), and the nearly intact transcription factor AML1 (McLean et al., 1996). In addition, the PNT domain from the Tel paralogue, Tel2, also mediates oligomerization with itself and in a heteropolymer with Tel (Poirel et al., 2000; Potter et al., 2000). Analysis of the crystal structure again detailed a head-to-tail tethering of exposed hydrophobic patches (Kim et al., 2001a; Tran et al., 2002). A K of 1.7 nM for Tel self-association was determined using a d  dimeric model of the Tel oligomer, created using a pair of solubilizing mutants that selectively destroy, in turn, one of the two association regions thus preserving a single wild-type interface. This high affinity is comparable to that of the Ste4/Byr2 complex (Kd of 19 nM; Ramachander et al., 2002) and, to a lesser degree, Polyhomeotic (K of d  190 nM; Kim etal., 2002a). In contrast to the above examples of oligomerization, the role of the ephrin receptor SAM domains in mediating biologically relevant association is far from clear. Several ephrin receptor SAM domain structures have been described in the literature with varying oligomeric states. Initial examples of dimeric (Stapleton et al., 1999) and polymeric (Thanos et al., 1999b; Smalla et al., 1999) SAM domains from EphA4 and EphB2 receptors hinted at a generalized role in protein self-association. However, the 12  Chapter 1 - Introduction biological relevance of this finding came into question with the observation of a monomeric form of the EphB2 receptor SAM domain within a different crystal space group (Thanos et al., 1999a), and the consistent observation of a very weak affinity (in the millimolar range) for the self-association of Ephrin receptor SAM domains. Although it is possible that ligand mediated clustering of the Ephrin receptors could reduce the entropic component of subsequent SAM domain binding and allow for the otherwise  unfavoured interaction, the EphA4 receptor SAM domain appears  dispensable in mice knock-in studies (Kullander et al., 2001). Even less clear is the generalized function of oligomerization for the PNT domain from the Ets family of transcription factors other than Tel. The literature is peppered with contrasting findings involving the ability these domains to form homooligomers and heteromeric complexes. Yeast two-hybrid and in vitro binding assays have suggested the ability of various PNT domain-containing constructs to interact including Erg/Erg, Erg/Ets-1, Erg/Fli-1, Erg/Ets-2 and Tel/Fli-1 (Carrere et al., 1998; Basuyaux et al., 1997; Kwiatkowski et al., 1998). However, investigation of the PNT domain from Ets-1 at millimolar concentrations (Slupsky et al., 1998a) failed to demonstrate even weak affinity for self-association, and the Tel/Fli-1 interaction was not observed during structural studies of the Tel PNT domain (Kim et al., 2001a). In addition, examples of distinctly monomeric SAM domains are now present, including the p73 SAM domain (Chi et al., 1999). It is therefore likely that a function different than self-association may have evolved for at least a subset of PNT and SAM domains. Indeed, there are several examples of PNT domains having a role in mediating association with other cellular proteins (Table 1.1). Unfortunately for many of these interactions, the actual biological function has also not been established. 13  Chapter 1 - Introduction  Table 1.1.  Reported PNT domain interactions with n o n - E t s proteins  PNT domain  Binding partner  Putative function  Reference  Ets-1  Daxx  repression  Li e f a / . , 2 0 0 0  Ets-1  ERK2  phosphorylation  Seidel and Graves, 2002  Ets-1  Sp100  transactivation  Wasylyk et al., 2002  Ets-1  Ubc9  sumoylation  Hahn etal., 1997  Ets-2  Brg1  Baker et al., 2003  Ets-2  Cdk10  chromatin remodeling cell cycle  Tel  mSin3A  repression  Tel  Ubc9  sumoylation  Kasten and Giordano, 2001 Chakrabarti and Nucifora, 1999 Wang and Hiebert, 2001 Chakrabarti etal., 1999  Yan  Mae  phosphorylation  Baker etal., 2001  1.5 Ets family of transcription factors  First identified within the tripartite E26 avian virus genome, Ets transcription factors have now been found in all major metazoan species, serving to regulate processes including development, apoptosis, and hematopoiesis (Dittmer and Nordheim, 1998; Sharrocks, 2001; Raouf and Seth, 2000; Mavrothalassitis and Ghysdael, 2000; Sementchenko and Watson, 2000; Maroulakou and Bowe, 2000; Remy and Baltzinger, 2000; Bartel et al., 2000; Gustin et al., 1998; Oikawa and Yamada, 2003; Lelievre et al., 2001). As with most transcription factor families, the mode and specificity of DNA-binding provides the pivotal characteristic that ties the member proteins together. This hallmark of the Ets family is a highly conserved ETS DNA binding domain that is capable of recognizing a conserved core 5-GGA-3' sequence 14  Chapter 1 - Introduction motif. DNA bases outside of the core trinucleotide-binding site add to specificity of the individual members of the Ets family (Graves and Petersen, 1998). At the molecular level, the ETS domain has a winged helix-turn-helix fold that mediates specific recognition of DNA via its major groove (Liang et al., 1994; Donaldson et al., 1996; Kodandapani et al., 1996; Werner et al., 1997; Batchelor et al., 1998; Mo et al., 1998; Mo et al., 2000). In addition, there are regions flanking the ETS domain in members such as v-Ets, Ets-1 and Ets-2 that serve to auto-inhibit DNA-binding. This autoinhibition is enforced via phosphorylation of adjacent sites by Ca -dependent 2+  calmodulin kinase II (Petersen et al., 1995; Cowley and Graves, 2000; Wang et al., 2002) and reduced via protein-protein partnerships (Goetz et al., 2000). As introduced previously, a subset of the Ets family includes a second region of homology now known as the PNT domain, found within the N-terminal regions of Ets1, Ets-2, Pnt-P2, Erg, Fli-1, GABPa, Tel and Yan (Figure 1.4). Despite an overall high level of identity amongst all of the PNT domain sequences, a phylogenetic-type analysis results in a clear clustering into four major sub-families (Figure 1.5). Members of each of these four sub-families will appear.throughout the thesis during analyses of the PNT domain fold and function. Therefore, a brief background into the major components of each will allow for a greater understanding of the diversity exhibited by the Ets family. Also included is a short overview of the enzymes involved in posttranslational modification of Ets transcription factors through phosphorylation and sumoylation, since these two modifications will be the focus of Chapters 4 and 5 in the thesis.  15  Chapter 1 - Introduction  Ets sub-family Ets-1  N  Ets-2  N  PNT  ETS  PNT  ETS PNT  Pnt-P2 NH  ETS  Erg/FH sub-family Erg(p55) N Fli-1  N  ETS PNT  c  ETS  c  GABP sub-family GABPa  N{  ETS  Tel sub-family ETS  Tel N-  ETS  Yan N-  Figure 1.4. Domain transcription factors.  composition  for  a  subset  of  Ets  family  Primary structure of full-length Ets-1, Ets-2, Pnt-P2, Erg, Fli-1, GABPa, Tel and Yan with PNT and ETS domain boundaries indicated. Proteins have been grouped into sub-families based on Figure 1.5.  16  Chapter 1 - Introduction  Figure 1.5. Phylogenetic alignment of Ets family PNT domains. Thirty-  three PNT domain sequences were clustered according to sequence similarity using the TRACE server ( -jiye/evoltrace/ evoltrace.html). The sequences were grouped into subfamilies based on phylogenetic similarity at the level of the dashed grey line, with the exception of Yan which was included with the Tel sub-family since there are no other Tel orthologues in Drosophila. The protein dEIg may be an additional member in the GABP sub-family, but the PNT domain-containing protein dMae is clearly an outlier from the Ets family proteins. The following sequences were used in the alignment (with accession numbers): dvYan (096416), dYan (Q01842), bTel (Q90ZS9), fTel (Q90ZS8), mTel (P97360), hTel (P41212), hTel2 (Q9Y603), hGABPa (Q06546), mGABPa (Q00422), xET2A (P19102), xET2B (Q91712), hEts-2 (P15036), mEts-2 (P15037), cEts-2 (P10157), cEtsB (P15062), cEtsA (P13474), rEts-1 (P41156), hEts-1 (P14921), mEts-1 (P27577), xET1A (P18755), dPnt-P2 (P51023), dEIg (Q04688), mFli-1 (P26323), hFli-1 (AAH01670), qFli-1 (093425), xFli-1 (P41157), zFli-1 (Q9PU61), xErgB (Q9W6Z9), xErgA (Q9W700), cErg (Q90837), hErg (P11308), rErg (BAB62744), dMae (NP_523786). The organism abbreviations are: b, bovine (Bos taurus); c, chicken (Gallus gallus); d, Drosophila melanogaster, dv, Drosophila virilis; f, Fugu rubripes; h, human (Homo sapiens); m, murine (Mus musculus); q, quail (Coturnix coturnix); r, rat (Rattus norvegicus); x, Xenopus laevis; z, zebrafish (Danio rerio).  17  Chapter  Q. CQ <  0)  33 03 K - in e£ 1 . 4 M M  a  K K bS c£ a; a: „ 0 0 0 0 0 « « K '  §S §§  1 i >H {>. w  w  w  Cr3  W  W  UJ  W  >* >-  UJ  M  UJ  M  CE;  El  S3  i-i t—i i—i o a  UJ  UJ  UJ UJ  C  c a >1 >1 o 33 & *i -j  o a w  >H  w  >H  UJ  w  W  O  C to w a UJ t«4  J  UJ  UJ  «  i-t U4  UJ  U  Q a  UJ  hS O O C i—i i—i 1-H  O  UJ  J  UJ  J I to tO  _3 J to tO  J tO to tO  to HQ to tO  to tO to tO  « k O  O « V5  h< lb k  > > ce, u.  Q  Q  «  -=;-=•!-=<: j j J  lh  >•  >J  ^ «J »j • J UJ  Q  Q  > U>H > UH  UH  «  •J  UJ  UH  PfW -=<  -*!  •J  w UJ K-3 i-J  >-J  E-  UH  CI  Q  -J  •J  UJ  UJ 1-3  J  B BBBB _3 j nb> EO H tO tO tO H K H E- E-H " Hj - t-^ H H tE-»o ^ j - a ^ j j BS a a P; o ;  i i i i i 0 0 0 6 6 O S S tT! in w H W i  3 w  O O O 3 3  D_ ffi[ g <  3 3 O O O 0 0  0  •D - a J 3  i cu I i-3 i PH SEOH  aa c O  cu J cu HP  UJ  O O o  > > > > £> UH > > UH  tTJ 02 ifl 33 PC 33 33  SogS S  :-> no w  > Q UJ  UJ  n  C w  UJ  -3 > UJ UJ  LQ to txJ faj w U J U4 UJ UJ UJ UJ UJ UJ UJ 33 3 33 33 33 33 33 33 33 33 33 [ 3 33  o o o o o o  «=! O O  <*•»  C  c  u  je  CD OJ «S  <  << p i£*  CN Or  & -M www w « - h c ° ) = ^ ^ OyWI±]W^EEffimUJ^LiJ0_Uj£Ll_U^  sz c x x sz  h  0 0  o  l  r  b  18  x  TJ TJ  t  r  i  i  c r x  PpPcPoiro NX  -  1-  Introduction  Chapter 1 - Introduction  1.5.1 Ets sub-family As the prototypic member of the Ets family of transcription factors, Ets-1 has been extensively studied with respect to expression, function and post-translational modification. Already alluded to is the complex regulation of DNA-binding exhibited by Ets-1, which in part dictates the level of transactivation towards several genes involved in development and function of lymphoid-derived cells. Several functions were also deduced from mouse studies in which disrupted Ets-1 genes led to problems in developing vasculature and lymphoid tissue, improper angiogenesis and reduced differentiation for a variety of lymphoid-derived cells (Bories et al., 1995; Muthusamy et al., 1995; Barton etai, 1998; Walunas et al., 2000). In addition to inhibition of DNA-binding by Ca -dependent phosphorylation of 2+  Ets-1, phosphorylation of Thr38 by the MAP kinase ERK2 leads to an increase in Rasdependent transactivation of target genes. This latter site is positioned in a disordered or conformationally flexible region N-terminal to the PNT domain, and is present in the other members of this sub-family that includes the paralogue Ets-2 and the orthologue from Drosophila, Pnt-P2 (Brunner et al., 1994; O'Neill et al., 1994; Yang et al., 1996). In fact, the importance of this site of protein modification was first established with studies involving the Pnt-P2 protein. Signaling by the sevenless receptor tyrosine kinase pathway via the Rolled MAP kinase leads to activation of the Pnt-P2 protein, and proper expression of genes necessary for development of the Drosophila R7 photoreceptor cell. Mutation of the single phosphoacceptor threonine adjacent to the PNT domain leads to severe defects in eye development (Brunner et al., 1994; O'Neill etai, 1994). For the last member in this group, phosphorylation of Ets-2 by ERK2 also leads 19  Chapter  1-  Introduction  to clear activation of the transcription factor (Yang et al., 1996). Although sharing a high level of similarity to Ets-1, the expression pattern of Ets-2 is more ubiquitous in nature. Ets-2 is shown to play a role in the cell cycle, is involved in apoptosis, and mediates the response to oxidative stress (Beier et al., 1999; Albanese et al., 1995; Wolvetang et al., 2003; Sanij et al., 2001). Overexpression of Ets-2 in mice mimics the skeletal abnormalities associated with Down syndrome, which is in keeping with its location in a region of chromosome 21 known to be a key determinant in this disease (Sumarsono et al., 1996). Compared to Ets-1, Ets-2 frequently plays a converse role in biological responses. This interplay is typified by the activation of T-cells, in which the level of Ets-2 transcript and protein increases whereas Ets-1 levels show a marked decrease (Bhat et al., 1990).  1.5.2 Erg/Fli sub-family The founding member of this subclass, Erg, was identified in a screen for Ets family members in a colon cancer cell line (Rao et al., 1987; Reddy et al., 1987). Transcription of the human erg gene leads to at least five possible splicing isoforms, each of which functions as a transcriptional activator but differs in the length of the Nterminal region and selective inclusion of exons between the central PNT domain and C-terminal ETS DNA binding domain (Rao et al., 1987; Reddy and Rao, 1991; Duterque-Coquillard et al., 1993; Prasad et al., 1994). As with other Ets proteins, Erg is also a target of signaling pathways with phosphorylation of one or more serine residues occurring upon activation of a MAP kinase cascade (Murakami etai., 1993). Within frog and mammalian cells, Erg has been associated with the gene regulation of a variety of proteins implicated in endothelial cell differentiation and 20  Chapter 1 - Introduction neural crest development (Baltzinger et al., 1999; Vlaeminck-Guillem et al., 2000; Hewett et al., 2001; McLaughlin et al., 2001). Specifically, a role in constitutive expression has been identified for the ICAM-2 gene (McLaughlin et al., 1999), in keeping with a long half-life for Erg of 21 hours (Murakami et al., 1993). Upon addition of TNF-a, Erg is down-regulated and is the direct cause of decreased ICAM-2 expression (McLaughlin et al., 1999). In chicken, different Erg isoforms play opposing roles in bone morphogenesis (Iwamoto etai., 2000). The other member of this sub-family, Fli-1 (also known as S i d ) , was initially identified through independent mapping of the common site of integration by the Friend leukemia virus (Ben-David et al., 1990; Baud et al., 1991; Bergeron etai., 1991) and as the breakpoint between Fli-1 and the Ewing sarcoma gene (EWS) resulting from chromosomal translocation (Delattre et al., 1992). Fli-1 shares a very high sequence identity with Erg and plays a complementary, but not quite identical, role in preventing erythroid differentiation (Starck et al., 1999; Athanasiou et al., 2000). Fli-1 is also involved in angiogenesis and is a specific determinant of megakaryocyte development (reviewed in Truong and Ben-David, 2000). In mouse models, disruption of Fli-1 leads to a loss in vascular integrity, problems with early vessel formation and defects in hematopoiesis (Melet et al., 1996; Brown etai., 2000; Hart et al., 2000; Spyropoulos et al., 2000).  1.5.3 GABP sub-family Unlike other members of the Ets family, GA-binding protein (GABP) has a heteromeric composition with requisite tandem DNA binding sites. Initially discovered as a binding complex within the promoters of the herpes simplex virus-1 (HSV-1) immediate early 21  Chapter  1-  Introduction  genes (Triezenberg et al., 1988), expression profiling has indicated a ubiquitous presence in mammalian tissues (LeMarco et al., 1991). In these diverse cell types, GABP  moderates  processes such  as T-cell  activation  and post-synaptic  neuromuscular differentiation (Avots etai., 1997; Fromm and Burden, 1998; Hoffmeyer etai.,  1998; Schaeffer et al., 1998; Fromm and Burden, 2001). The well-characterized murine and human proteins consist of an a and p  subunit which can form a dimeric a/p complex in solution (Chinenov et al., 2000) and bind with high affinity as a a/p /a tetramer to tandem pairs of 5'-GGA-3' motifs within 2  GABP-responsive promoter elements (Thompson et al., 1991). The G A B P a subunit is a true member of the Ets family and contains a central PNT domain and a C-terminal ETS DNA binding domain that directly contacts the GGA trinucleotide (Batchelor et al., 1998). In isolation, GABPa binds weakly to single DNA binding sites (Thompson et al., 1991). It is only upon addition of the non-Ets GABPp subunit that a high affinity protein-DNA complex is formed. This enhanced affinity is mediated through the association of two GABPa/p dimers with stabilized binding to two adjacent cognate DNA binding sites. At the molecular level, the interaction between the two subunits occurs via the ETS DNA binding domain of GABPa and the four N-terminal ankyrin repeats of GABPp (Thompson et al., 1991; Batchelor et al., 1998). Tetramerization is achieved  through  association  of  the  C-terminal  coiled-coil  motif  within  GABPp (Thompson et al., 1991). The transactivation domain is also provided by the P subunit. Of key importance to this thesis is the recent identification of a MAPK phosphoacceptor site in GABPa that is C-terminal to the PNT domain. Thr280 of 22  Chapter  1-  Introduction  human GABPa has been shown to be phosphorylated in response to neuregulin-1 (NRG-1) stimulation of muscle cells through the well-known ERK and JNK signaling cascades (Fromm and Burden, 2001). In addition, GABPa is known to be a target of ERK and JNK in the activation of interleukin-2 in T-cells (Hoffmeyer et al., 1998). Phosphorylation of an as yet unmapped threonine residue in GABPa (as well as serine and threonine residues in GABPp) is required for full ERK-mediated transactivation of the human immunodeficiency virus 1 (HIV-1) promoter (Flory et al., 1996). The molecular consequence of GABPa phosphorylation and the role of the PNT domain in GABP function have not yet been characterized.  1.5.4 Tel sub-family The Tel (Translocation Ets leukemia) protein gained notoriety through the discovery that a number of different genes encoding kinases or transcription factors form oncogenic fusions with its N-terminal region via chromosomal translocation events. Characterization of one such fusion event with the C-terminal kinase domain of platelet-derived growth factor receptor p (PDGFRp) coincided with the original isolation of Tel  (Golub et al., 1994). Since then, several types of leukemia have been traced to  translocations involving fragments of the tel gene with genes encoding the catalytic domains of tyrosine kinases ABL, JAK2 and Cyd, as well as transcription factors such as AML1 and MN1 (reviewed in Rubnitz et al., 1999). In addition, a solid tumor has been recently linked to the fusion protein formed from the N-terminal region of Tel with the C-terminal catalytic domain of neurotrophin tyrosine receptor kinase 3 (NTRK3) (Knezevich et al., 1998). In each case, the chimeric protein retains the entire region  23  Chapter  1-  Introduction  surrounding the PNT domain. Accordingly, it is possible that the behaviour of the chimeric protein results from its unregulated self-association as mediated by the Tel PNT domain. Such a case seems clearest for the class involving the tyrosine kinase domain,  since  ligand-independent  kinase  phosphorylation  brought  about by  oligomerization could lead to constitutive downstream signaling and deregulated growth in affected cells (Jousset et al., 1997). It is also possible that the fused protein merely leads to a hemizygous loss-of-function in at least a subset of childhood leukemia, thereby eliminating the potential tumor suppression capabilities of Tel (Rubnitzef al., 1999). Despite a wealth of genetic and biochemical research into the Tel oncogenic fusions, less is known about the native protein. Tel is ubiquitously expressed and appears to be most critical in embryonic angiogenesis and in adult hematopoesis (Bartel et al., 2000; Maroulakou and Bowe, 2000; Mavrothalassitis and Ghysdael, 2000). Coupled to a role in artery and tissue remodeling, Tel has been shown to be a bona  fide  regulator of the extracellular matrix metalloproteinase stromelysin-1  gene  (Fenrick et al., 2000). As for the mechanism of action, Tel appears to recruit mSin3A and N-CoR/SMRT via a region that includes the PNT domain, as well as histone deacetylase-3 through a central repression motif (Chakrabarti and Nucifora, 1999; Fenrick etai., 1999; Guidez etai., 2000; Wang and Hiebert, 2001). Also present in humans is the probable Tel paralogue known as Tel2 or TelB, which retains repressor activity and has been reported to be either ubiquitously expressed or limited to hematopoietic cells (Poirel et al., 2000; Potter et al., 2000). In Drosophila,  the orthologue Yan/Aop is best known as a second Ets target of the Ras-  dependent Sevenless pathway in development (Lai and Rubin, 1992; O'Neill et al., 24  Chapter 1 - Introduction 1994). Phosphorylation by the Rolled MAPK leads to translocation from the nucleus and subsequent degradation of the modified Yan protein (O'Neill et al., 1994). The presence of Yan serves as a general mechanism to prevent cellular differentiation, and this role has been specifically studied for the fate of the photoreceptor precursor cells in which it acts in opposition to Pnt-P2 (Rogge et al., 1995; Price and Lai, 1999). For gene promoters such as that for Lozenge (Behan et al., 2002), it is thought that vacation of the Yan-repressed 5'-GGA-3' motifs might allow for binding by Ets family activators such as the fly protein Pointed-P2 discussed previously (Brunner et al., 1994). It is also possible that the PNT-domain protein, Mae, aids in the phosphorylation  of Yan through  an unknown  mechanism that may include  conformational change within Yan or direct interaction with Rolled (Baker et al., 2001).  1.5.5 Ets transcription factors and ERK2 phosphorylation  It is now clear that the biological function of Ets transcription factors can be modified by phosphorylation via a variety of cell-signaling events (Yordy and Muise-Helmericks, 2000). For Ets-1, Ets-2, Pnt-P2, GABPa and Yan, there is a specific involvement of extracellular-regulated kinase 2 (ERK2) MAP kinase cascade in the regulation of activity and nuclear localization. First established for the transduction of the mating response in S. pombe and S. cerevisiae, the MAPK pathway has at its heart a trio of protein kinases that transduce an extracellular signal from a cell surface receptor (e.g. tyrosine receptor kinase or G-protein coupled receptor) to phosphorylation of a specific set of protein substrates (reviewed in Chen et al., 2001). The terminal kinase in this signaling module, the MAPK, is activated upon dual phosphorylation on a tyrosine and threonine  by a MAPK  kinase (MAPKK), 25  which  in turn was activated via  Chapter  1-  Introduction  phosphorylation by an upstream MAPKK kinase (MAPKKK). In mammalian cells, three major MAPK pathways exist that lead to activation of either the ERK1/ERK2, JNK or p38 MAP kinases. Each MAPK targets specific Ser/Thr-Pro sites in protein regions that, as shown in this thesis, typically display conformational flexibility. Additional specificity is gained by docking regions such as the PNT domain (Chapter 4) that also contact the MAPK, but are situated distal to the phosphoacceptor serine of threonine. The MAP kinase, ERK2, is found in the cytoplasm in a predominantly inactive state, but upon phosphorylation by the MAPKK MEK1 there is an activating conformation change in ERK2 (Boulton et al., 1991; Canagarajah et al., 1997). Although not unambiguously established, this may induce dimerization and a translocation to the nucleus (Khokhlatchev et al., 1998). Phosphorylation of protein substrates (e.g. Ets family transcription factors) occurs upon entry of ERK2 into the nucleus, leading to modulation of the expression of genes that are regulated by these proteins. The mechanism by which phosphorylation can affect these changes in gene expression can include altered transcription factor localization or stability, RNA polymerase association, DNA-binding affinity, or interaction with components of the chromatin remodeling complexes (Figure 1.6). It is likely that each of these mechanisms is utilized by at least one member of the Ets family following phosphorylation by ERK2 as a result of specific extracellular signaling.  1.5.6 Ets proteins as targets for sumoylation  Less well characterized is the possible post-translational modification of Ets proteins such as Tel and Ets-1 by the small ubiquitin-related modifier, SUMO. Analogous to the well-defined  ubiquitination  pathway  (Figure 26  1.7A), sumoylation ,  requires initial  Chapter 1 - Introduction  Figure 1.6. Generic models of transcriptional activation and repression.  Involvement of proteins in gene expression include both activator (grey circles labeled A) and repressor (circles labeled R) transcription factors, along with chromatin remodeling complexes (for example histone acetyl transferase, HAT, and the histone deacetylase complex, HDAC) and additional components that have been omitted for clarity (reviewed in Goodrich et al., 1996; Struhl, 1996; Kadonaga, 1998; Kornberg and Lorch, 1999). Activity of these many protein factors can be modulated by posttranslational modification such as phosphorylation and sumoylation. During a basal level of transcription (A), both activator and repressor proteins may be present. During specific induction of gene expression (B), additional activators may bind, with potential recruitment of chromatin remodeling proteins that increase the accessibility to the RNA polymerase machinery. Repressor proteins may be degraded or display reduced DNA binding ability whereas transcriptional activators may show increased activity due to posttranslation modification such as phosphorylation (indicated by the asterisk). During specific gene repression (C), additional repressor proteins may bind or display increased activity (again indicated with an asterisk). Some of these transcription factors may interact with proteins involved in the condensation of chromatin (e.g. by deacetylation of histones). In addition, transcriptional activators may be degraded or display reduced affinity for DNA binding. 27  Chapter 1 - Introduction  Figure 1.7. Sumoylation pathway. Comparison of (A) ubiquitination and (B) sumoylation mechanisms. For the ubiquitin pathway only the HECTfamily E3 enzymes are shown. For E3 ligation enzymes containing a RING finger, the transfer of ubiquitin from the E2 conjugating enzymes to the target protein substrate is facilitated without intermediate ubiquitination of the E3. (C) Detail of the thioester linkage present between ubiquitin/SUMO and the E1, E2 and E3-type enzymes. (D) Isopeptide bond formed between substrate and ubiquitin/SUMO.  28  Chapter 1 - Introduction proteolysis of SUMO to expose a C-terminal Gly-Gly motif, which is then attached via a thioester linkage to a cysteine residue in the E1 activating SAE1/SAE2 heterodimer (reviewed in Kim et al., 2002b). This reaction is followed by transfer of SUMO to a cysteine residue on the E2 conjugating enzyme UBC9 (Figure 1.7B.C). In many cases, UBC9 can directly interact with and transfer SUMO to a lysine s-amino group from the final substrate, resulting in a stable isopeptide linkage (Figure 1.7D). It is also possible that, similar to the ubiquitin system, an additional E3 ligase protein is required for efficient substrate modification (Hochstrasser, 2001). Unlike ubiquitination where multiple E2 and E3 set specificity, for sumoylation this appears to be directed by a consensus target sequence *FKxE (where Y is a large hydrophobic residue, usually isoleucine, valine or leucine and x is any amino acid; Rodriguez et al., 2001). The consequence of sumoylation is still unresolved for almost all of the identified targets, but it is evident that nuclear localization, enhanced protein stability (i.e. prevention of ubiquitin-directed proteolysis) and regulation of protein activity are observed for several proteins upon modification by SUMO-1 (reviewed in Muller et al., 2001, Kim et al., 2002b; Pichler and Melchior, 2002;). For example, RanGAPI appears to use this modification as a transportation signal to the nuclear pore complexes (Matunis et al., 1996; Mahajan et al., 1997). Similarly, sumoylated PML is targeted to sub-nuclear punctate bodies (Boddy et al., 1996; Sternsdorf et al., 1997; Muller et al., 1997; Zhong et al., 2000). An example of enhanced stabilization includes k B a , for which the sumoylated form is resistant to ubiquitination (Desterro et al., 1998). Finally, the involvement of sumoylation in p53 activity has been frequently reported, even though the molecular basis of this modulation is not known (Gostissa et al., 1999; Rodriguez et al., 1999, Muller et al., 29  Chapter 1 - Introduction 2000). For the Ets family of transcription factors, evidence of sumoylation is limited to yeast two-hybrid analysis for Ets-1, with a suggested role for UBC9 in modulating the level of transactivation (Hahn et al., 1997). In contrast, sumoylated Tel has been identified  in vivo, and additional evidence exists for SUMO-mediated nuclear  localization of Tel in distinct nuclear bodies (further discussed in Chapter 5; Chakrabarti etai., 1999; Chakrabarti etai., 2000).  1.6 Yeast pheromone response pathway  Apart from the PNT domains found within the Ets family, I have also investigated the association of the Ste4 and Byr2 proteins of the mating pathway in fission yeast. To understand the biological role of these two proteins, it is first necessary to describe the general mechanisms involved in sexual differentiation of the fission yeast S. pombe. For this process, the analogous pathway in the budding yeast S. cerevisiae has been the target of more extensive genetic and biochemical investigation and it is from this system that most of our understanding has developed (reviewed in Gustin et al., 1998; Elion, 2000; Breitkreutz and Tyers, 2002). Haploid S. cerevisiae exist in two mating types (a and a) that secrete type-specific peptide factors to stimulate receptors expressed on the opposing a or a cell. Upon activation, the receptor triggers an internal signal mediated in part by the trio of protein kinases STE11, STE7 and FUS3/KSS1 that make up the pheromone response MAP kinase cascade (central region in Figure 1.8A). FUS3 and KSS1, thus activated, subsequently induce matingspecific genes through phosphorylation of protein targets such as FAR1, DIG1, DIG2 and STE12, leading to cell-cycle arrest, polarized growth towards the mating partner and eventual fusion of the two haploid yeast to form a diploid cell. 30  Chapter 1 - Introduction  S. pombe  S.  nitrogen starvation and pheromone MAPKKKK  pheromone (or nitrogen starvation)  high osmolarity  STE20  STE20  Shk1 Ste4 I  MAPKKK  ?  Ras1  cerevisiae  STE50  STE50  [ Byr2 ]  STE11  \  I  J MAPKK  Byr1  PBS2  STE7  I  MAPK  STE11  I HOG1  Spk1  FUS3/KSS1  i  i  mating meiosis  I  mating /  osmotolerance  p s e u d o h y p h a l growth  B Ste4  Byr2  5 S H  STE50  RA  ^ K R B D K C B D H  kinase ~ l  STE11  RA  ^ C F W  kinase  Figure 1.8. Signaling proteins from yeast. (A) Kinase cascades from S.  pombe and S. cerevisiae. The scaffold protein STE5 involved in the S. cerevisae pheromone response pathway is not shown. (B) Domain structure of Ste4 and STE50 detailing the N-terminal SAM domain and Cterminal Ras-associating domain (RA). Ste4 also contains a central coiledcoil region (coil) responsible for Ste4 trimerization. (C) Domain structure of Byr2 and STE11 detailing the N-terminal SAM domain (SAM), Ras-binding domain (RBD), catalytic binding domain (CBD) and the C-terminal kinase domain.  31  Chapter 1 - Introduction Additional components add to the complexity of this signaling pathway. In brief, the scaffolding protein STE5 binds to STE11, STE7 and FUS3 to potentially sequester this particular cascade of kinases and prevent cross-talk to other pathways in yeast. In addition, both STE5 and STE20 (an upstream kinase of STE11) integrate the signal presented by the G-protein p subunit of the seven transmembrane helix pheromone receptor. Finally, the association of the STE50 protein to STE11 also appears necessary for full transduction of the mating response. In fission yeast, there is a similar but distinct pathway leading to sexual differentiation (Figure 1.8A). The two mating types P and M (or + and -) use the same approach of secreted peptides to activate cell surface receptors expressed by the opposing mating type. However, unlike S. cerevisiae, S. pombe requires a second signal of nitrogen starvation in order to progress through the steps of cellular morphogenesis. In contrast, it is a separate but highly similar pathway that links nitrogen limitation to filamentous growth in budding yeast. Although many components are shared between these two pathways in S. cerevisiae, FUS3 and KSS1 phosphorylate a distinct set of target regulatory genes in response to nitrogen starvation as compared to pheromone stimulation (Figure 1.7A). The two yeast genera also differ in detail with respect to the upstream protein components of the mating pathway. Instead of the G-protein p subunit in budding yeast, intracellular signaling in S. pombe is initiated through the G-protein a subunit, with the potential action of the MAPKKKK Shk1 (orthologous to STE20) and additional input from an activated Ras1 protein that has no parallel in S. cerevisiae. The subsequent trio of kinases is similar to that in budding yeast, with Byr2, Byr1 and Spk1 serving the respective roles of MAPKKK, MAPKK and MAPK. However, no scaffolding 32  Chapter 1 - Introduction equivalent to STE5 has yet been identified in S. pombe. In addition, there is an absolute requirement for Ste4 interaction with Byr2 for sexual differentiation to occur. As mentioned previously, the Ste4 orthologue in S. cerevisiae, STE50, has an important but not essential role in budding yeast mating. Indeed, STE50 is also not absolutely required for the filamentous response in this yeast, with deletion mutants displaying only a reduced ability to differentiate. Instead, it is in the pathway responsive to high osmolarity that STE50 plays a pivotal role in S. cerevisiae. The socalled HOG pathway (Figure 1.7A), is another MAP kinase cascade in S. cerevisiae that shares many components with the mating and nitrogen response pathways including input from STE20 and the initial MAPKKK STE11, which is the binding partner to STE50.  1.6.1 Ste4 The first member of the heteromer studied in Chapter 3 of this thesis is derived from the Ste4 gene of S. pombe, initially isolated in yeast screens for mutants that result in a sterile phenotype. Subsequent characterization of the cloned Ste4 protein indicated a central region predicted to form a coiled-coil (Figure 1.7B) (Okazaki et al., 1991). Yeast lacking the ste4 gene have normal growth and cell shape, but demonstrate a defect in transmission of the sexual differentiation signal (Barr et al., 1996). Further characterization of the protein using two-hybrid and co-purification techniques revealed a critical role for the N-terminal SAM domain in mediating the interaction with Byr2 (Barr et al., 1996; Tu et al., 1997). In addition, Ste4 was found to interact with itself (Barr etai., 1996). In comparison, the S. cerevisiae Ste50 gene was first identified during the yeast 33  Chapter  1-  Introduction  genome sequencing project and demonstrated a role in mating pathway activation similar to Ste4 (Rad et al., 1992). Although STE50 does not contain a central coiledcoil comparable to Ste4, the two demonstrate significant sequence identity in a Cterminal homology region (Figure 1.7B). Deletion of this region in either Ste4 or STE50 results in varying degrees of mating response deregulation (Okazaki et al., 1991; Rad et al., 1992). However, a role for this conserved region has not yet been defined.  1.6.2 Byr2 The immediate downstream target of Ste4 is the Byr2 MAPKK kinase (also known as Ste8), isolated from a screen of genes capable of bypassing ras1 inactivation (Wang et al., 1991). As with Ste4, disruption of byr2 resulted in viable cells with normal growth and shape, but an inability to sporulate or conjugate in response to pheromone and nitrogen starvation (Wang et al., 1991). Apart from the N-terminal SAM domain that mediates interaction with the SAM domain of Ste4, there are additional motifs present in Byr2 (Figure 1.7C). The Cterminal kinase region displays significant homology to other kinase domains, with retention of residues important for catalytic activity (Hanks et al., 1988). This activity is inhibited in non-stimulated yeast though steric blockage by the catalytic binding domain (CDB) (Tu et al.,  1997). Relief from inhibition  may occur through  phosphorylation of residues within the repression domain by an upstream kinase (potentially by the STE20 orthologue Shk1) or by binding of regulatory proteins (Tu et al., 1997). Finally, a Ras-binding domain (RBD) helps initiate the external mating signal Via binding of Ras1 and concurrent translocation to the cell membrane (Bauman etai.,  1998). 34  Chapter  1-  Introduction  1.7 Thesis overview It is now established that both PNT and SAM domains display a variety of cellular roles instead of an exclusive ability to self-associate. It is in the investigation of these divergent roles that the following thesis research was undertaken. Two distinct model systems were examined to probe the range of structure and function of this conserved domain. In addition, the effects of post-translational modifications of the regions adjacent to and including the PNT domain were also investigated, with specific interest into the ERK2-mediated phosphorylation and the UBC9-mediated sumoylation of Ets1. Towards  the  overall thesis goals, Chapter 2  outlines  the  structural  characterization of the PNT domain from the Ets transcription factor Erg, with comparison to other PNT and SAM domain folds. Similar to Tel, the Erg PNT domain lacks the N-terminal helix found in Ets-1 and GABPa, indicating a significant degree of plasticity in the PNT domain architecture. Unlike Tel, and in contrast to previous literature reports, Erg does not demonstrate an ability to oligomerize or indeed form an association with any other PNT domain. This is a characteristic shared by Ets-1, Ets-2, Pnt-P2, GABPa and Fli-1. Chapter 3 uses the Ste4 and Byr2 model system in order to investigate the mechanism by which SAM domains may affect association. Using a biochemical approach, a surprising stoichiometry of one Byr2 molecule binding to three Ste4 molecules was demonstrated for the Ste4/Byr2 complex. This high affinity interaction is dependent upon trimerization of the Ste4 protein via a coiled-coil motif, hinting at an even more diverse function displayed by the PNT and SAM domains. It is also now evident that post-translation modification of Ets proteins may 35  Chapter  1-  Introduction  involve various surface regions of the PNT domain. This scenario is explored in Chapter 4 with the identification of MAP kinase docking sites on Ets-1 and G A B P a that serve to increase the affinity of ERK2 for phosphoacceptor threonine residues Nterminal and C-terminal to the PNT domain, respectively. The regions of Ets-1 and G A B P a in which phosphorylation occur are flexible in solution, and addition of a phosphate group does not alter the structure or dynamics of this region or of the PNT domain itself. Therefore it is probable that the outcome of MAP kinase phosphorylation is the altered ability to associate with or dissociate from unknown members of the transcriptional machinery in order to activate or repress protein function. Adjacent to the MAP kinase phosphorylation site in Ets-1 is a sequence motif now characterized as a putative site of sumoylation. Unlike the initial reports of SUMO modification within the structured Tel PNT domain, it now appears that significant interaction with the sumoylation enzyme UBC9 is only evident for the flexible Nterminal region of Ets-1. Although the binding mechanism is characterized at the molecular level in Chapter 5, the biological outcome of in vivo Ets-1 modification is not yet known.  1.8 Publications The work presented in this thesis has been reported in several papers, in addition to publications that are currently in preparation. In Slupsky et al. (1998), which reported the first structure of a PNT or SAM domain, I used CD and NMR spectroscopy to confirm that helix H1 is integral to the conformation and thermodynamic stability of the PNT domain. This potential variance in PNT domain architecture prompted investigation of the Erg PNT domain described 36  Chapter 1 - Introduction in Chapter 2. Using different Ets-1 constructs, I also demonstrated that residues 1 to 50 are conformationally flexible in solution, and that their presence does not perturb the structure or fold stability of the adjacent PNT domain. This region contains the MAP kinase and sumoylation sites discussed in Chapters 4 and 5. Initial chemical shift assignments of the Erg PNT domain, described in Chapter 2, were reported in Mackereth et al., 2002. This allowed for comparison of secondary chemical shifts between the Erg, Ets-1 and GABPa PNT domains, and confirmed a shared core structure of four a-helices for all three proteins, with inclusion of the additional N-terminal helix only in Ets-1 and GABPa. Investigation of the in vitro association between Ste4 and Byr2, in collaboration with the Bowie laboratory (University of California, Los Angeles), revealed a surprising 3:1 stoichiometry dependent upon trimerization of Ste4 via an adjacent coiled-coil (Ramachander et al., 2002). Detailed in Chapter 3, my specific contribution to this publication involved native gel electrophoresis to confirm the stoichiometry of Ste4 and Byr2 within the complex, along with gel filtration and CD spectroscopic analyses of several Ste4 and Byr2 constructs. This work added to the known modes of association demonstrated by the SAM domain. Currently in preparation are three papers dealing with separate aspects of PNT domain function, based on research presented within this thesis. First, a report on the tertiary structure of the Erg PNT domain (Chapter 2) will allow for comparison with the crystal structure of the Tel PNT domain (Kim et al., 2000a), an NMR-derived structure of the GABPa PNT domain (determined by a co-worker, M. Scharpf) and a newly refined structure of the Ets-1 PNT domain (a collaboration with Dr. Tom Alber at the  37  Chapter 1 - Introduction University of California, Berkeley). This will help define the conserved and divergent features of the PNT and SAM domains. A second paper will present the NMR investigation of the ERK2 docking region on both Ets-1 and G A B P a (Chapter 4). Third, the involvement of Ets proteins in the sumoylation pathway and the interactions with UBC9 (Chapter 5) will comprise an additional future publication. Finally, a current collaboration with Dr. Poul Sorensen at the University of British Columbia deals with aspects of Tel PNT domain oligomerization not contained within this thesis. My contribution focused on the bacterial production of a Tel-NTRK3 oncogenic fusion protein for gel filtration and electron microscopy work, to confirm that this chimera of a PNT domain with a tyrosine kinase forms a large oligomeric complex in solution. The formation of a polymer is crucial for the oncogenic properties of Tel-NTRK3. By preventing this oligomerization via mutation or over-expression of the minimal Tel PNT domain, the fusion protein is rendered benign. Also absent from this thesis was the separate study of the RNA polymerase II subunit RPB10 from Methanobacterium thermoautotrophicum (Mackereth et al., 2000). I determined the NMR-derived structure of this protein, which comprised a bundle of three a-helices stabilized by a zinc atom. This paper was important in that it was the first description of RPB10, and was used to help trace the electron density map during calculation of the crystal structure of the yeast RNA polymerase II complex (Cramer et al., 2000). In addition, the M. thermoautotrophicum RPB10 project was part of a Canada-wide structural proteomics initiative aimed at determining the tertiary folds of all proteins within this thermophilic archaeon (Christendat et al., 2000).  38  Chapter 2 Structural studies of the Erg PNT domain  The solution structure of the PNT domain from Erg has been determined, revealing an architecture composed of a four a-helical bundle. Not present is the N-terminal helix found in Ets-1 and GABPa. Such diversity in structural features is echoed by varying surface motifs that may explain the many roles of the PNT and SAM domains explored in subsequent chapters in the thesis. The domains from Ets-1, Ets-2, Pnt-P2, GABPa, Erg, and Fli-1 are monomeric in solution, whereas Tel and Yan form insoluble polymers. Contrary to published data, no evidence was found for heteromeric complexes in vitro, bringing into question the generalized role of the PNT domain as a self-association module.  Part of the research in this chapter has been published in the Journal of Biomolecular NMR, 24:71-72 (Mackereth et al., 2002). This paper reports my chemical shift assignments of Erg '  (108 201)  and the comparison of secondary chemical shifts appearing  as Figure 2.5 in this thesis.  39  Chapter 2 - Structural studies of the Erg PNT domain 2.1 Characterization of Erg constructs  In order to investigate aspects of the PNT domain in molecular detail, I determined the structure of the domain from the human Erg transcription factor. In a classification of PNT domain sequences based on sequence similarity, Erg and Fli-1 form a group distinct from the previously determined PNT domains (Figure 1.5). In fact, the known PNT domain structures from Ets-1, GABPa and Tel also fall into separate classification groups. Therefore, determination of the Erg PNT domain architecture is required to discriminate common elements of the PNT domain fold from aspects specific to each family member. One aspect of divergence, already suggested by analysis of the known structures, was the presence of the N-terminal helix in some PNT domains (i.e. Ets-1 and GABPa) but not others (i.e. Tel). To define the full boundaries of the structural PNT domain from Erg, an initial construct encoding the entire N-terminal half of Erg (residues 1 to 201) was cloned into an expression vector. From the purified E r g  ( 1 2 0 1 )  protein, limited proteolysis was used to define a minimal structural fragment. Figure 2.1 illustrates that the region of Erg " (1  201)  resistant to trypsin cleavage, corresponds to  the core four helices common to both SAM and PNT domains (H2 to H5). In contrast, proteolytic cleavage occurs in the region that is occupied by helix H1 in Ets-1 and GABPa, suggesting that this N-terminal helix is not present in the Erg PNT domain. However, some trypsin susceptibility was observed for the well-structured helix H1 in Ets-1 (Slupsky et al., 1998a). To prevent inadvertent clipping of a potentially unstable helix H1 in Erg, I chose to study a construct containing residues 108-201 in keeping with the length of the Ets-1 and GABPa PNT domains.  40  Chapter 2 - Structural studies of the Erg PNT domain  45 31  MW  obs  = 22806  Erg(l-201)  21.5 14.4  MW  obs  = 9176/9020  Erg(124-201)/Erg(125-201)  Std  0  2'  5'  10'  30'  B  Ets-1 < - ) secondary structure 29  H1  138  H2  H3 H4  H5  Erg (-201) 1  Arg/Lys Protected from trypsin cleavage .  t  t  1J  M  t  i  t  u u  1 n  ,—  1  20  1  1  1  1  1  1  1  1  40  60  80  100  120  140  160  180  Residue  Figure 2.1. Proteolysis of Erg " <1  201)  to determine minimal-sized folded  PNT domain. (A) Trypsin digest of Erg< - > separated by a 20% SDSPAGE PhastGel (Amersham) and visualized by Coomassie stain. In parallel, samples were characterized by ESI-MS. (B) Schematic of all potential trypsin cleavage sites in Erg< " (arrows) with sites not cleaved during the 30 minute incubation shown in white. Secondary structure relates to sequence aligned helices from Ets-1 " . The region resistant to cleavage, residues 125-201, represents the minimum folded region of the PNT domain and does not appear to include the N-terminal helix (H1). 1 201  1  201)  (29  41  138)  1 200  Chapter 2 - Structural studies of the Erg PNT domain The E r g  (108  "  201)  peptide was expressed in E. coli and purified to homogeneity.  Circular dichroism spectroscopy revealed that E r g  (108  "  201)  contains a well-folded core,  with significant helical content. Furthermore, the protein exhibited reversible two-state folding with a T of 64 °C (at pH 7.0 in 20 mM potassium phosphate and 50 mM NaCl; m  not shown). To confirm that this smaller E r g  (108  "  201)  construct contains the same folded  tertiary structure that is present in the full N-terminal fragment Erg " (1  201)  , both proteins  were N-labeled and characterized using H - N HSQC spectroscopy. Figure 2.2 15  1  indicates that the minimal E r g 1  (108  "  201)  15  construct produced an exact superimposition of  H - N amide cross-peaks with the Erg " 15  (1  201)  spectrum. Since the  1  H  N  and  15  N  resonances of the amide groups are very sensitive to changes in its surrounding chemical environment, superimposition of these chemical shifts indicates that there is no significant structural difference for the PNT domain of E r g the extra N-terminal residues present in Erg " (1  201>  (108  "  upon addition of  201)  . Furthermore, the fact that all of the  additional peaks specific to the larger construct fall within a narrow H chemical shift 1  range of 8.0-8.5 ppm suggests that N-terminal residues (1-107) are flexible and disordered in solution. This result is in keeping with the sensitivity of this region to trypsin proteolysis (Figure 2.1), and supports the choice of the smaller and more amenable E r g  (108  "  201)  construct for further structural characterization. In addition, the  above results also indicate that residues 1-124 in native Erg are unfolded, at least in isolation from protein partners..  2.2 Chemical shift assignment Assignment of the NMR signals from E r g  (108  three-dimensional spectra collected from  N - and C/ N-labelled protein (Appendix  1 5  42  "  201)  utilized a standard set of two- and 13  15  Chapter 2 - Structural studies of the Erg PNT domain  Figure 2.2. Superimposition of [ N]Erg< - > and [ N]Erg< - > H - N H S Q C spectra. Tertiary structure is preserved between the common residues 15  1  201  15  108  201  1  15  of the larger Erg PNT domain construct (in bluefl) with the smaller Erg' " construct (in red []). Note that the residues specific to [ N]Erg (residues 1 to 107) are clustered in the region of the spectra diagnostic of flexible and unfolded peptides. The single downfield peak specific to [ N]Erg( , (10.1 ppm, boxed), likely represents the additional indole amide from the Nterminal Trp56. (1_201)  108  201)  15  (1201)  15  201)  43  1_  Chapter 2 - Structural studies of the Erg PNT domain 1). Backbone assignment was performed using typical H - N HSQC, HNCACB, 1  15  CBCA(CO)NH and HNCO spectra (section 2.6.4; reviewed by Sattler et al., 1999). In brief, all backbone amide chemical shifts were first identified from the H - N HSQC 1  15  spectra and then connected sequentially to adjacent amides using the HNCACB and CBCA(CO)NH spectra. These also provided the  1 3  C (and C ) chemical shifts of the a  13  P  previous and same residue, respectively. The HNCO spectra provided a link to the  1 3  C  backbone carbonyl carbon, which is useful for secondary structural analyses. Once complete, the backbone resonance identities were confirmed using (H)CC(CO)NHand H(CC)(CO)NH-TOCSY spectra; this also provides the chemical shift of all the protons and carbons, respectively, in the aliphatic side chains. These data are characteristic of specific side chain spin systems (as are the values of  1 3  C and C ) . a  13  P  Due to the absence of amide protons for proline residues, a HACAN spectrum (Kanelis et al., 2000) was collected in order, to obtain  15  N values for residues preceding a  proline, and to allow for assignment through the Pro-Pro-Pro sequence at the Nterminus of E r g  (108  "  201)  . An annotated H - N HSQC of the protein is illustrated in Figure 1  15  2.3. Aliphatic side chain assignment used the (H)CC(CO)NH- and H(CC)(CO)NHTOCSY spectra described above, with the help of a HCCH-TOCSY spectrum to confirm direct H - C connectivities. In order to stereospecifically assign the valine and 1  13  leucine prochiral methyl groups for a better-defined structure, a partial non-randomly 10% C-labelled protein was used (Neri et al., 1989). Using C - H S Q C and C T - C - 13  13  13  HSQC spectra, resonances from the y1 and y2 methyl groups of valine and the 51 and 82 groups of leucine were distinguished by the pattern of nearest neighbour labeling  44  Chapter 2 - Structural studies of the Erg PNT domain  R140E  T132 0  108  N1565©W142E  -  o,^ft L159  Q141  C1Y15O0 690  T1360  N117S F176  0  Q  a  u  Q161E  jj«N1215 1/N1626  116  196  \  °® W 9  S  W 1 3 4  E  QQ_  0CN184 0K17O 0 K U 8  120  «8^0Q161 ft •M171 ©1187 H138  % 0°  6  E110 M113 1157 0 A129  1126 0  &  »  l 1  g  124  1 6 4  0L2OO  0N121  S135©  128  W145tC> 11.5  112  0D13O  9? cR178  0  E109 W142® J  N156C)  N1845  D 1 8 6  E122  W134el  1  w  ^§ # /  0  L143 0 L192$/  Ai46f  0Q141E  0 v  l  °T18o\ /  E1440 i r f Q1770 "  L200B  6  P'  V1 V147tf \>Jf F160  H191  1  T120 \\ S19O0 o  2 N1620  |  N1178ft^  T119 T1640  G 1  ^  Q 1 7 7 £  10.5  G165C  9.5 8.5 1H ppm  7.5  6.5  Figure 2.3. Annotated H- N HSQC spectrum of [ N]Erg< - >. Peaks 1  15  15  108 201  corresponding to G 1 6 5 , W 1 4 2 e and R 1 4 0 s have opposite sign due to aliasing  in the N dimension. 1 5  45  Chapter 2 - Structural studies of the Erg PNT domain via one or two carbon units (Neri et al., 1989; Senn et al., 1989). Stereospecific assignment of the sidechain amides of asparagine and glutamine were determined with an EZ-HMQC experiment, as described by Mcintosh et al. (1997). Aromatic resonance assignments were made using an approach described for Ets-1 and GABPa PNT domains (Slupsky et al., 1998b). This relies on use of H - N HSQC, CT1  1  H - C HSQC, C(3H5, CBHe, DQF-COSY, 13  15  15  N-NOESY (150 ms mixing time) and  HMBC spectra. Using the HMBC spectrum, all four histidine residues (His112, His138, His191 and His193) were determined to be protonated solely on the 5i nitrogen. Using the POP program (Schubert et al., 2002), all proline residues were confirmed to be in the trans conformation. The final list of chemical shift assignments for E r g included  in  Appendix  2,  and  was  deposited  in  the  (108  "  is  201)  BioMagResBank  ( as accession number 5399.  2.3Erg  (108 2 0 1 )  secondary structure  The chemical shifts of backbone H , H , 1  N  1  a  1 3  C , a  1 3  C , N and 15  1 3  C atoms provide P  secondary structure information of a protein. This information can be obtained through several approaches. Protection from amide hydrogen-exchange following a transfer to D2O buffer typically relates to residues participating in hydrogen-bonding or protected from the solvent due to tertiary constraints. The program TALOS uses a database of structures and chemical shifts to estimate the backbone § and \\i angles of each residue (Cornilescu etal., 1999). Alternatively, the deviation in H , C , 1  a  1 3  a  1 3  C and C 1 3  P  chemical shifts from random coil values can also be used to estimate region of a-helix and B-strand via the chemical shift index (CSI) protocol (Wishart and Sykes, 1994). In  46  Chapter 2 - Structural studies of the Erg PNT domain addition, measurement of the helical conformation ( JHN-Ha 3  ( JHN-H<X 3  3  JHN-HCX  coupling constant can indicate residues in a  < 6 Hz) as compared to residues in an extended p-strand  > 8 Hz) (Kuboniwa et al., 1994). Finally, elements of secondary structure have  distinctive patterns of atoms in close three-dimensional space, which can be determined from NOESY spectra. All four approaches have been combined in Figure 2.4 to derive a picture of the secondary structure of E r g  (108  "  201)  . Consistent prediction of  helical structure is demonstrated for the four core helices common to all PNT domains (helix H2, 136-149; helix H3, 165-169; helix H4, 173-177; helix H5, 182-196). In addition, there is evidence of a helical conformation for residues 157-159. This corresponds to the 'helical turn' present in Tel (Kim et al., 2001a; Tran et al., 2002), and to the more prominent helix H2' found within the SAM domain fold. Once again, there is no indication of any helical secondary structure in the region N-terminal to helix H2, supporting the absence of helix H1 in Erg. As described above, secondary structure is reflected in the deviation of the observed chemical shift from a reference value of the same residue in a random coil conformation. There should also be a good agreement with these 'secondary chemical shift' values when compared with all three PNT domains. Indeed, if the secondary shift values for E r g *  108-201  ', Ets-1  (29_138)  and G A B P a  (168  "  254)  are aligned based on the  corresponding sequences, it is evident that this deviation is conserved for all three proteins. As an example, Figure 2.5 details this conserved pattern for  1 3  C values for a  residues in helices H2 to H5. Residues in the N-terminal helix (H1) of Ets-1 and GABPa show a pattern divergent from Erg. Additional discrepancy is noted for 47  Chapter 2 - Structural studies of the Erg PNT domain  2.4. NMR-derived secondary structure of Erg " ). A combination of HX, TALOS, CSI, J NOE-derived parameters define the helical regions as indicated above the numbered sequence of Erg* *. For the HX experiment, amides remaining after a 30 min transfer to D 0 buffer are indicated with a black circle. TALOS prediction of helical conformation, designated by an 'H', exists for residues estimated to have vj/ and <>| angles of -57 ± 20° and -47 ± 20 °, respectively, in nine out of the ten closest database entries. For the CSI analysis using H , C and C chemical shifts, values of +1, 0 and -1 indicate predicted regions of p-strand, random coil, and helix, respectively. J NHa couplings are indicated as belonging to one of three groups: 6 ( J < 6), 7 (6 < J < 8) and 8, ( J > 8), with the small coupling indicative of a helical conformation. Bars connect atoms for which the indicated NOE crosspeaks have been identified. Figure  (108  3  201  a n c l  H N H a  108-201  2  1  a  1 3  a  13  P  3  3  H  HNHa  3  3  H N H a  HNHa  48  Chapter 2 - Structural studies of the Erg PNT domain  H2 110 120 130 140 150 gshMEEKHMPPPNMTTNERRVIVPADPTLWSTDHVRQWLEWAVKEYGLP  HX  H  TALOS CSI  HHHHHHHHHHHHH  minimum 7  3  ^HNHa  d  aN(M+2)  8  8 8 8  7 8  8 78686766  666 667 87  d p (M+3) a  d (7,/+3) aN  <WM+4)  HH2'  H4  H3  160 170 180 190 200 DVNILLFQNIDGKELCKMTKDDFQRLTPSYNADILLSHLHYLRETPLP • •• • •  HX  HHH  TALOS  HHHH  HHHHHH  +1.  0-  CSI  -1  3T J  H5  llll 86 6!8(  HNHa  d (M+2) AN  d p(M+3) a  daNOV+3) daN(^V'+ ) 4  49  HHHHHHHHHHHHHHH  Chapter 2 - Structural studies of the Erg PNT domain  : MPPPNMTTNE ATFSGFTKEQCjj GABPu: AALEGYRKEQEi  Erg  Ets-1:  HI  Figure 2.5. Conservation of C secondary chemical shift for ETS family 13  a  PNT domains. Secondary chemical shifts were calculated for E r g ~ (in black), E t s - 1 (grey) and G A B P a " (white) from random coil reference values (Wishart and Sykes, 1994). Illustrated is a sequence alignment with conserved residues boxed, along with horizontal bars indicating locations of helices defined by chemical shift, J N - H coupling and, for Ets-1 " , analysis of its NMR-derived tertiary structure (Slupsky et al., 1998). Previously published in Mackereth etai., 2002. (108  (29_138)  (168  254)  (29  3  H  a  50  138)  201)  Chapter 2 - Structural studies of the Erg PNT domain residues within the loop region between helices H2 and H3, perhaps reflecting varying degrees of helical conformation for the short turn corresponding to helix H2' in SAM domains. In addition to a single amino acid insertion in E r g acids for G A B P a  (168  "  254)  (108  "  201)  , a gap of two amino  in the loop between helices H4 and H5 contributes to areas of  secondary chemical shift divergence.  2.4 E r g  ( 1 0 8 2 0 1 )  tertiary structure  Calculation of the Erg  (108  ~  201)  NMR-derived structure employed a combination of  NOESY spectra to determine which atoms are close in space within the folded protein. Distance restraints were acquired from 3D N-HSQC-NOESY, simultaneous C - and 15  15  N-HSQC-NOESY,  1 3  aromatic NOESY and CT-methyl-methyl and amide-methyl-  NOESY spectra (detailed in Materials and Methods section 2.8.6). Added to this information were the inclusion of several  § and xi angles (section 2.8.6) and  residual dipolar coupling information (section 2.8.5). Incorporation of the above data into the iterative assignment protocol ARIA v1.2 (Nilges, 1995; Nilges and O'Donoghue, 1998; Linge and Nilges, 1999; Linge et al., 2001), led to a final set often water-refined structures (Table 2.1). As shown in Figure 2.6A, E r g  (108  "  201)  consists of the four core a-helices found in  other PNT domains (Figure 1.3A). However, in place of the N-terminal helix H1 in Ets1 and GABPa, the ensemble of structures (Figure 2.6B) clearly indicates a flexible and unstructured region. This results from the lack of any long range NOE restraints observed for these residues. Half of the structures in the ensemble also demonstrate a short helix between helices H2 and H3, ranging from three to four residues between  51  Chapter 2 - Structural studies of the Erg PNT domain  Table 2.1. NMR restraints and statistics for the ensemble of structures calculated for E r g - . (108  Summary of restraints NOEs" Intraresidue Sequential Medium-range (1 < \i-j] < 5) Long-range > 5) All Dihedral angles V (j) •Xi Hydrogen bonds Residual dipolar coupling Deviation from restraints N O E restraints, A Dihedral restraints, Residual dipolar coupling restraints, Hz Deviation from idealized geometry Bonds, A Angles, ° Improper angles, ° Mean energies , kcal mot E d Ebonds Eangle 0  b  V  684(154) 400 (206) 299 (216) 495 (360) 1878(936) 34 34 8 2x14 63 0.024 ± 0 . 0 1 2 0.45 ± 0.07 0.35 ± 0.03 0.003 + 0.000 0.46 ± 0.01 0.41 ± 0.02  1  W  ENOE  E  136 ± 8 13 ± 1 97 ± 6 62  ESANI  8 ± 1  E .j E i rmsd from average structure, A  -261 ± 9 -3760 ± 1 5 0  L  e c  structured region 0.76 ± 0.04 0.46 ± 0.04  All heavy atoms Backbone ( C , C , N) a  a  ±13  0.9 ± 0 . 3  c d i h  e  ten  201)  0  number of unambiguous restraints, with ambiguous restraints in parentheses.  final ARIA/CNS energies for van der Waals (vdw), bonds, angles, NOE restraints (NOE), dihedral restraints (cdih) and residual dipolar coupling restraints (SANI). The Lennard-Jones (L-J) and electrostatic potential energies (elec) were calculated using the CHARMM energy function, and were not included during structure refinement. b  c  residues 127-197 (i.e. H-{ N} NOE > 0.5) 1  15  52  Chapter 2 - Structural studies of the Erg PNT domain  Figure 2.6. Erg  ' consists of a bundle of four a-helices. (A) Ribbon representation of a low energy structure of Erg ' ). The core four helices H2 (orange), H3 (yellow), H4 (green) and H5 (cyan) are indicated. (B) Same as in A but rotated 90° around the vertical axis. (C) Superimposition of the ten water-refined structures with the helices indicated as above. This representation emphasizes the lack of structural constraints for the N-terminal region. This reflects true conformational flexibility as indicated by the trypsin digest susceptibility (Figure 2.1), random coil chemical shifts (Figures 2.2 and 2.4) and H-{ N} NOE values < 0.5 (Figure 2.7). Not shown are the residues that form the hydrophobic core (i.e. less than 15% solvent accessible) including: V127, P128, A129, P131, W134, V139, W142, L143, A146, V147, L152, V155, 1157, F160, 1163, D164, G165, L168, C169, F176, L179, T180, N184, A185, L188, L189, L192, L195 and R196. (108 201)  (108  1  15  53  201  Chapter 2 - Structural studies of the Erg PNT domain Ile157 to Gln161. This element of secondary structure corresponds to the more prominent helix H2' found in SAM domain structures. It is now clear that structural diversity exists within the otherwise conserved PNT domain at the level of protein architecture. However, the static representations of the PNT domains in Figures 1.3A and 2.6 fail to address the level to which the additional helix H1 contributes to the folded domains from Ets-1 and GAPBa. To this end, backbone amide dynamics was probed using H-{ N} NOE relaxation analysis of 1  Erg  (io8-2oi)_  ^^(29-138)  a  n  d  G A B  p 068-254) ( a  F  i  g  u  r  e  15  The H - N heteronuclear NOE 1  j).  2  15  is a highly sensitive indicator of motions on a sub-nanosecond timescale, with values < 0.5 indicative of significant backbone dynamics. For Erg  (108  ~  201)  , only the core four a-  helices display reduced mobility ( H-{ N} NOE > 0.5), with clear conformational 1  15  flexibility in the N-terminal region ( H-{ N} NOE < 0.5). In contrast, residues 1  corresponding to helix H1 in Ets-1  (29  "  138)  15  and G A B P a  (168  "  254)  have relaxation parameters  indistinguishable from the remainder of the PNT domain. Therefore, for Ets-1 and GABPa, helix H1 is a bona fide component of the structured PNT domain and not merely an adjunct component of secondary structure. This result is in keeping with the observation that a single-state unfolding mechanism exists for Ets-1 " (29  138)  , with  residues in helix H1 following the same equilibrium of urea-induced unfolding as observed for helices H2 to H5 (Slupsky et al., 1998a).  2.5 Structural comparison to other PNT AND SAM domains The structure of the Erg PNT domain provides an example from a different sub-family than the previously reported PNT domain structures of Ets-1, GABPa and Tel (Figure  54  Chapter 2 - Structural studies of the Erg PNT domain  H3 H4 H5  Erg  Ets-1  GABPa  Helix 1  Figure 2.7. Erg PNT domain lacks the integral N-terminal helix (H1) found in Ets-1 and GABPa. H-{ N} NOE relaxation for [ N]Erg< - ), [ N]Ets-1< 1  15  15  108  201  15  29  138) p5N]GABPa( - ) indicating regions of stable structure ( H-{ N} NOE > 0.5) and conformational flexibility ( H-{ N} NOE < 0.5). Residues corresponding to helix H1 in Ets-1 and GABPa form an intimate component of the PNT domain. 168  a  n  254  1  d  1  55  15  15  Chapter 2 - Structural studies of the Erg PNT domain 1.5). As a result, it is of interest to determine if there are any differences detectable at the level of fold architecture between these four PNT domain structures, or indeed from examples of several SAM domains. Superimposition of the C atoms from Erg a  2 0 8 )  (108  "  residues in helix H2, H3, H4 and H5 to sequence-related C " atoms in Tel and  GABPa revealed a very high degree of structure similarity, with rmsd values of only 1.1 and 1.4 A, respectively (Figure 2.8). A less favourable rmsd of 2.2 A results from comparison of the E r g  (108  -  201)  structure to that of Ets-1 " (29  138)  (1bqv). In addition, Ets-  1 (29-138) foes not share a high degree of structural similarity to the NMR structure of GABPa  (168 2 5 4 )  or to the crystal structure of Tel (3.0 and 2.2 A, respectively). This may  indicate a true difference in the helical angles present in these domains, or may reveal a need for an improvement in the structure quality for the PNT domain from Ets-1. To this end, residual dipolar coupling data have since been acquired for Ets-1 " (29  Ets-1 " (51  138)  138)  and  , and a crystal structure for this latter construct is also being solved in the  laboratory of Dr. Tom Alber (University of California, Berkeley). It is expected that upon further refinement, the rmsd between E r g  (108  "  201)  and Ets-1 " (29  138)  will drop below 2 A as  expected for two domains such as these with otherwise high sequence identity. Among the SAM domain structures listed in Figure 2.8, E r g  (108  "  201)  shows the  highest similarity to that of the polyhomeotic SAM domain (rmsd = 2.0 A), a trend shared by the remaining three PNT domains. The p73 and EphB2 SAM domains, although displaying a high level of similarity amongst themselves, seem to form a group that is further divergent from the tertiary structure of the PNT domain.  56  Chapter 2 - Structural studies of the Erg PNT domain  Erg  Erg Tel (1lky)  Tel (1lky)  Ets-1 (1bqv)  GABPa  ph (1kw4)  p73 (1cok)  EphB2 (1f0m)  EphB2  1.1  2.2  1.4  2.0  2.6  2.2  2.7  2.2  1.4  2.3  3.1  2.8  3.2  3.0  2.9  3.5  3.2  3.7  2.9  3.2  3.2  3.4  2.3  1.9  2.4  1.7  1.5  1.1  (isgg)  Ets-1  dbqv)  2.2  2.2  1.4  1.4  3.0  Ph (1kw4)  2.0  2.3  2.9  2.9  p73 (1cok)  2.6  3.1  3.5  3.2  2.3  EphB2 (1fOm)  2.2  2.8  3.2  3.2  1.9  1.7  EphB2 (1sgg)  2.7  3.2  3.7  3.4  2.4  1.5  GABPa  1.2  1.2  Figure 2.8. Structure similarity between several PNT and SAM domains. The program psbbfit (a least squares best fit based on Kabsch, 1976) was used to calculate pair-wise rmsd values for sequence aligned residues (from alignment in Figure 1.2), comprising helices H2, H3, H4 and H5 in four PNT domain and four SAM domains structures. The following residue ranges were used: Erg (140-152, 166-182, 188-199), Tel/1 Iky (65-77, 90-106, 111-122), Ets-1/1bqv (75-87, 100-116, 122-133), GABPa (192-204, 217-233, 237-248), ph/1kw4 (1476-1464, 1452-1436, 1428-1417), p73/1cok (491-503, 515-531, 538-549), EphB2/1f0m (915-927, 939-955, 963-974) and EphB2/1sgg (934946, 958-974, 982-993).  57  Chapter 2 - Structural studies of the Erg PNT domain  2.6 Screen for in vitro association The hypothesis of the PNT domain as a protein-protein oligomerization module had its origins in the clear association properties of the Tel protein in the context of both wildtype and oncogenic fusion proteins (Papadopoulos et al., 1995; Golub et al., 1996; McLean et al., 1996; Jousset et al., 1997; Lacronique et al., 1997; Ho et al., 1999; Wai et al., 2000; Kim et al., 2001a). In addition, several groups have reported association of other PNT domains in either a homotypic or heterotypic manner between Erg and Ets-1, Fli-1, Ets-2 and between Tel and Fli-1 (Basuyaux et al., 1997; Carrere et al., 1998; Kwiatkowski et al., 1998). Following the determination of the PNT domain from Erg, I decided to study these reported interactions in vitro using a matrix of PNT domains from Ets-1, Ets-2, Pnt-P2, GABPa, Fli-1, Tel and Yan (see Materials and Methods section 2.7.2 and Figure 2.9). Of these PNT domains, Tel and Yan were insoluble, in keeping with their predicted tight association and model of extended polymerization (Kim et al., 2001a). The lack of solubility precluded any investigation into their association in vitro with other PNT domains. The remaining PNT domains from Ets-1, Ets-2, Pnt-P2, GABPa, Erg and Fli-1 produced soluble folded proteins that exclusively displayed monomeric behaviour in gel filtration, native gel electrophoresis and glutaraldehyde cross-linking studies (Figure 2.10; Slupsky et al., 1998). The contrasting observation by others that Erg and Ets-2 (Basuyaux et al., 1997; Carrere et al., 1998) exhibit dimerizing potential in vitro may be explained by the finding that several PNT domains have the potential for forming disulphide-bonds in solution if a reducing environment is not maintained. Clear evidence of this artifactual binding can be seen in preparations of Erg in the presence versus the absence of DTT. Non-reducing SDS-PAGE of these latter 58  Chapter 2 - Structural studies of the Erg PNT domain 115 E r g (p55)  199  296  Erg  Ets-1  C  ETS  (1-201)  136  415 440  335  PNT  N  464  (105-201)  Erg  55  376  I  PNT  N  ETS  P27577  Ets-1 (29-138) 87 Ets-2  170 PNT  H  442 468  362  ETS Ets  P15037  _ (85-172) 2  690 166 Pnt-P2  |  N-  250  610  PNT |  ETS  718  -c  P51023  p t-P2(159-253) n  170 GABPa  251  320  PNT  N-  400 454  GABPa* Fli-1  114  N-  198  281  PNT  39  1 6 8  "  2 5 4  361  P00422  ) 452 -  ETS F  Tel  -  ETS  P26323  |j1 (106-200)  125  335  N  416  485 P97360  ETS T |(38-127) e  T e  Te  31 Yan  |(38-127) g A  4 D  |(38-127) 113E V  117  396  479  732 Q01842  ETS  N Yan(22-129)  —i 100  1 200  1 300  1 400  1 500  1 600  1 700  1 800  1 900  1  Residue Figure 2.9. PNT domain constructs used in the study of Erg and the in vitro tests of association. Primary structure of full-length Erg, Ets-1, Ets-2, Pnt-P2, GABPa, Fli-1, Tel and Yan with SwissProt accession numbers indicated. PNT, SAM and ETS domain boundaries are defined using the SMART database ( Construct lengths are indicated with a black bar, with the site of Tel solubilizing mutations shown with an asterisk.  59  Chapter 2 - Structural studies of the Erg PNT domain  Figure 2.10.  1D H-NMR spectra of selected PNT domain constructs. 1  Amide regions of PNT domains from Ets-2, Pnt-P2, Fli-1 and Tel indicating that these peptides are folded in solution as indicated by disperse peaks in 20 mM potassium phosphate (pH 7.0) and 50 mM NaCl. For Tel, two separate solubilizing mutants, A94D and V113E, were used in order to prevent oligomerization. Gel filtration chromatography using S-100 sepharose, native gel electrophoresis and glutaraldehyde cross-linking confirms that these constructs are monomeric (not shown).  60  Chapter 2 - Structural studies of the Erg PNT domain  Chapter 2 - Structural studies of the Erg PNT domain samples revealed significant amounts of disulphide dimers, and H - N HSQC 1  spectroscopy  revealed chemical shift perturbations  15  in amides including, and  neighbouring, Cys169. A similar high propensity of cystine formation was observed for E t s 1  (29-138)  a n d  GABPa  addition, G A B P a  (168  "  (168  254)  "  254)  during their respective structural characterization. In  and G A B P a  (168  "  254)  C 2 2 5 A also seem to succumb to an  irreversible oxidation-independent aggregation accompanied by significant structural alterations of unknown origin (S. Mcintosh, M.Sc. thesis). The additional observations of self-association by two-hybrid analysis may indicate that a very weak binding affinity (>1 mM) still exists that is not observed in our experiments, or that unknown bridging partners mediate the interaction in the yeast system. The soluble PNT domains were then tested in a pair-wise manner for interactions using native gel electrophoresis and glutaraldehyde cross-linking. As before, no evidence of association was observed. Since Tel has been implicated in complex formation with some of these domains, soluble mutants of Tel were also tested with the panel of PNT domains. Either of these solubilizing mutations (A94D and V113E) inhibits oligomer formation due to incorporation of a negative charge into the otherwise hydrophobic association interface (Kim et al., 2001a). However, since one native interface is retained in each mutant, a soluble dimer is produced upon mixing of both mutants. Glutaraldehyde cross-linking revealed that only Tel ~ (A94D) 36  127  and Tel " (V113E) were able to interact in solution, with no observed association 36  127  with Fli-1 or Erg (Figure 2.11), in contrast to previous reports (Carrere et al., 1998; Kwiatkowski etal., 1998). From these results, it is evident that only the PNT domains from Tel and Yan show association potential, whereas the domains from Ets-1, Ets-2, Pnt-P2, GABPa, 62  Chapter 2 - Structural studies of the Erg PNT domain  Glutaraldehyde  MW 45-f* 29-H 20-W  none  15-W  Q— 345 29  20-W 15-W  0.005% (v/v)  63  j |(38-i27)A94D e  Te  |(38-127)V113E  +  -  +  +  +  Fli-1  GABPa  Erg  Ets-2  Pnt-P2  (106-200)  (168-254)  (108-201)  (85-172)  (159-253)  +  +  +  +  + +  +  +  +  +  Figure 2.11. Test for heteromeric complex formation  by using  glutaraldehyde cross-linking. Example experiment showing crosslinked  complex (*) only for the mixture of the Tel mutants, TelA A94D and Tel " V113E. Equivalent investigation for other PNT domains in isolation and in complex failed to reveal any additional homo- or hetero-oligomers. Experiments were conducted in 20-100 mM potassium phosphate (pH 7.0-7.5). (38_127)  127)  63  (38  Chapter 2 - Structural studies of the Erg PNT domain Erg and Fli-1 are monomeric in solution, alone or in combination. This lack of selfassociation can be understood for E r g  (108  "  201)  at the molecular level through  comparison to the oligomerizing interfaces defined for the Tel PNT domain. A surface representation of the Tel crystal structure (Figure 2.12A) demonstrates the two hydrophobic patches, designated ML and EH, that associate within the polymeric form of this domain (Kim et al., 2001a; Tran et al., 2002). It is the introduction of acidic residues within these domains that destroys oligomerization for the solubilizing mutations Tel " (36  127)  A94D and Tel (36  127)  V113E. In E r g  (108  -  201)  , the EH region retains a  predominantly hydrophobic character, whereas the ML region is distinctly hydrophilic, with the incorporation of Glu167 and Lys170 (Figure 2.12B). The presence of these charged residues likely mimics the prevents the head-to-tail association of the two oligomerization regions that is required for polymer formation. An explanation for the lack of pair-wise binding to Erg the other monomeric PNT domains is less obvious. E r g  (108  (108  "  ~  201)  201)  , or indeed to any of and Ets-1 " (29  138)  , for  example, still retain at least one solvent exposed hydrophobic patch that could participate in protein binding. Therefore, additional factors such as incompatible charge-pairing or exact van der Waals surface shape must otherwise inhibit complex formation. This lack of protein association prompted both an investigation into a protein system that does exhibit biologically relevant binding, and also a search for additional solubilizing mutations generated for Tel, and non-PNT domain proteins that may be the true partners for these otherwise monomeric PNT domains. To this end, Chapter 3 will look into aspects of association between the SAM domains of Byr2 and Ste4 of the S. pombe mating pathway. In addition, Chapter 4 will deal with the 'docking' interaction 64  Chapter 2 - Structural studies of the Erg PNT domain  Figure 2.12. Erg PNT domain lacks the oligomerization interfaces determined for Tel. (A) Charge surface representation of the Tel crystal  structure (1 Iky) with negative potential in red fl) and positive in blue Q . Highlighted in yellow and cyan are the ML and EH oligomerizing interfaces defined by Kim et al. (2000), each centred around a hydrophobic patch composed of the residues shown. (B) A similar charge surface representation of Erg ~ indicates a definite lack in the ML hydrophobic patch, with two introduced charged residues. In contrast, the EH region may be consistent with oligomerization, with only a tyrosine hydroxyl group on the perimeter of an otherwise hydrophobic region. Figure made using SwissPdbViewer ( and Pov-Ray v3.1 (http:// (108  201)  65  Chapter 2 - Structural studies of the Erg PNT domain  observed between the Ets-1 PNT domain and ERK2. As for the function of the PNT domain in Erg, further research will be required in order to identify putative binding partners. These approaches could include affinity purification of Erg complexes from a variety of human cell extracts, for example using the tandem affinity purification (TAP) tag approach coupled with mass spectrometric analysis (Rigaut et al., 1999). Similarly, a two-hybrid experiment using the minimal Erg PNT domain can be used to identify putative binding partners. This approach has already identified the chaperone Hsp90 as a potential Erg-associated protein in a study that used a large PNT and ETS domain-containing fragment of X. laevis Erg as the bait protein (Deramaudt et al., 2001). Finally, it is known that Erg becomes phosphorylated on undefined serine residues following treatment of KG1 cells with phorbal myristate acetate (Murakami et al., 1993). Since this phosphorylation may involve a MAP kinase, the Erg PNT domain may serve as a kinase-docking site similar to that described for Ets-1 or GABPa in Chapter 4. Using surface plasmon resonance, NMR spectroscopy, cross-linking or native gel electrophoresis, a role in MAP kinase docking can be tested using purified ERK, p38 and JNK MAP kinases in an in vitro binding study with Erg - . (108  201)  2.7 Materials and methods 2.7.1 PNT domain cloning  Figure 2.12 outlines the constructs used in Chapter 2. All Erg constructs were cloned from the p55 isoform of human Erg (cDNA generously provided by M. DuterqueCoquillard) into pET28a (Novagen) using PCR-generated restriction enzyme sites. Cloning of Erg " , Erg - , Erg " (1  107)  (1  201)  (1  479)  and Erg (108  66  201)  used appropriate combinations  Chapter 2 - Structural studies of the Erg PNT domain  of the following primers: Erg-1F, 5'-GGAATTCAGCTAGCACTATTAAGG-3': Erg-108F 5-GCGGACATATGGAGGAGAAGCACATGCC-3';  Erg-107R, 5'-CGATGCTCGAG-  Erg-201R, 5'-CCGTGAAGCTT7CATGGAAGAGGA-  7CAGTAGCTGCCGTAGTTC-3';  GTCTC-3'; Erg-479R, 5-CCTACGTCTCGAG7TAGTAGTAAGTGCCC-3'. Cloning of  Ets-1 (29  138)  and GABPa " (168  have been described (Slupsky et al., 1998; Mackereth  254)  et al., 2002). Ets-2 - , Fli-1  (106  mouse  library  (85  embryo  172)  cDNA  -  200)  and Tel " (38  by  using  were PCR-amplified from a 10.5 d the  primers:  Ets2-85F,  5'-  Ets2-172R, 5'-CCTGGGGATCC7CATTC-  GGCCGCCATGGCCACCTTCAGTGGC-3', TTGGTTCTCTTTG-3';  127)  FIM-106F, 5-GCGGACATATGGATGAGAAGAACGG-3', Fli1-  200R, 5'-CCGTGAAGCTT7CACAGCAGTGAACTTTCC-3',  Tel-38F, 5'-GGCCG-  CATATGC AC AC AGTGCCTCG AGC-3', Tel-127R, 5'-CCTGGGGATCC 7CAAG ATTTCCTCTGCTTC-3'.  Pnt-P2 " (159  253)  and Yan " (22  129)  were PCR-amplified from their full-  length coding sequences in plasmids provided by H. Ruohola-Baker (University of Washington) using the primers: PntP2-159F, 5-GCGGACATATGAACGAGGTACTGAAGG-3', PntP2-253R, 5'-CCGTGAAG_CTTrCATGGTTTCTCGCAATC-3', Yan-22F, 5-GCGGACATATGAGCGATGTGCTGTGG-3', CATGTGGGACTCTATG-3'.  Yan-129R, 5'-CCGTGAAGCTT7CA-  For each primer the introduced restriction sites are  underlined, stop codons are in italics and coding bases are shown in bold. Except for Ets-2, all clones were inserted into both pET22b and pET28a (Novagen) using A/del and Hind\\\ or SamHI restriction enzymes, to allow for both exclusion and inclusion, respectively, of an N-terminal His6 purification tag. Due to an internal A/del site, Ets-2 was cloned solely with Nco\ and SamHI restriction sites into pET28a, resulting in exclusion of the N-terminal HiS6-tag. In order to produce soluble Tel " , the A94D (38  127)  and V113E mutations were introduced into both the pET22b:Tel " (38  67  127)  and  pET28a:Tel " (38  127)  Chapter 2 - Structural studies of the Erg PNT domain  plasmids using the QuikChange kit (Stratagene). Mutagenic primers  included: Tel-A64D-F,  5'-CGAAATGAATGGCAAGGACCTCCTGCTGCTG-ACC-3',  TelA64F-R, 5'-GGTCAGCAGCAGGAGGTCCTTGCCATTCATTTCG-3', Tel-V113E-F, 5'-CTCCTCATTCAGGCGACGAGCTCTATGAACTCCTTC-3', Tel-V113E-R, 5'-GAAGGAGTTCATAGAGCTCGTCGCCTGAATGAGGAG-3' (with the introduced mutations underlined). All constructs were verified by DNA sequence analysis.  2.7.2 Protein  purification  All plasmids were transformed into E. coli BL21(A,DE3) and grown in Luria broth (LB) or minimal M9 medium supplemented with N-ammonium chloride (1 g/L) or C615  13  glucose (3 g/L) and the appropriate antibiotic. Protein expression was induced at OD600  ~ 0.6 using 1 mM IPTG for 3 h at 30 °C. Cells were collected by centrifugation at  3,000 x g, and resuspended in Binding Buffer (50 mM HEPES (pH 7.5), 500 mM NaCl, 10% glycerol, 5 mM imidazole) for HiS6-tagged protein, 50 mM Tris (pH 8.5) for pET22b:GABPa " , (168  254)  pET22b:Ets-1 " , pET28a:Ets-2 - , and 50 mM (29  138)  (85 172)  potassium phosphate (pH 7.0) for pET22b:Tel - , pET22b:Tel " A94D and (38 127)  (38 127)  pET22b:Tel " V113E. Cells were disrupted by two cycles through a French press at (38 127)  10,000 psi, followed by 1 min of sonication on ice. Lysates were cleared with centrifugation at 10,000 x g for 30 min followed by filtration through a 0.8 jam cut-off membrane. All Hise-tagged proteins (including all Erg constructs, pET28a:Fli-1 " , (106  pET28a:Yan - , pET28a:Pnt-P2 (22 129)  (159 253)  200)  and pET28a:GABPa ' ) were purified by (168 254)  using Ni -affinity chromatography. Following loading onto a 5 mL Hi-Trap column 2+  68  Chapter 2 - Structural studies of the Erg PNT domain  (Amersham Biosciences), non-specific bound proteins were removed by a 100-150 mL of Wash Buffer (50 mM HEPES (pH 7.5), 500 mM NaCl, 5% (v/v) glycerol and 60 mM  imidazole). Elution at 1 mL/min into 1.5 mL Eppendorf tubes was achieved with Elution Buffer (50 mM HEPES (pH 7.5), 500 mM NaCl, 5% (v/v) glycerol and 250 mM imidazole). Fractions were tested qualitatively for protein using a standard Bradford assay (BioRad), and protein-containing fractions were pooled and placed in dialysis tubing. The His6-tag was removed by incubation with thrombin (-0.5 mg) during overnight dialysis at 4 °C into NMR buffer containing 20 mM sodium phosphate (pH 7.0), 50 mM NaCl and 2 mM B-mercaptoethanol. The extent of proteolysis was monitored by a reduction in apparent molecular mass by using SDS-PAGE, and upon completion, was terminated with a 15 min incubation with 200 ul p-aminobenzamidine beads (Sigma). In addition, 500 ul of water-rinsed Talon (Clontech) was added to the samples to remove the cleaved HiS6-tag and any uncleaved full-length protein. The supernatant contained purified protein with a residual N-terminal Gly-Ser-His sequence. The samples were cleared with centrifugation at 1000 x g for 2 min, and the supernatant concentrated using a 1K filter (Pall Filtron) in an Amicon stir-cell concentrator. The proteins were stored at 4 °C, and kept in the reduced state with addition of fresh DTT. pET22b:GABPa ' , pET22b:Ets-1 (168 254)  (29_138)  , pET28a:Ets-2 -  (85 172)  were purified  by FPLC using a Q-sepharose column and 50 mM Tris (pH 8.5) with a gradient of 0 to 300 mM NaCl over 80 min, at a flow rate of 4 mL/min. For pET22b:Ets-1 " (29  138)  and  pET28a:Ets-2 - , the protein eluted with 120 mM NaCl. pET22b:GABPa (85 172)  (168  254)  eluted in two peaks (80 mM NaCl and 150 mM NaCl), with only the first peak retained  69  Chapter 2 - Structural studies of the Erg PNT domain  due to irreversible aggregation observed for protein eluting with the higher salt concentration. The pET22b Tel constructs were also purified by FPLC using an SPsepharose column with 50 mM potassium phosphate (pH 7.0), a gradient of 0 to 300 mM NaCl and a flow rate of 4 mL/min. Elution occurred at 50 to 100 mM NaCl. All protein samples were dialyzed into 20 mM potassium phosphate (pH 7.0) and 50 mM NaCl, then stored at 4 °C with fresh DTT added to keep the protein in the reduced state. The purity and identity of each sample was confirmed by using SDS-PAGE and ESI-MS (performed by Dr. Shouming He). Protein concentrations were determined using a value of  S280  calculated with the ProtParam program (  tools/protparam.html).  2.7.3 Limited trypsin digestion  Samples were digested with trypsin at 25 °C in 25 mM Tris (pH 7.9), 50 mM KCI, 0.1 mM EDTA and 1 mM DTT at a ratio of 1:250 (w/w) of enzyme to protein. Proteolysis was stopped at defined intervals with 5 mM PMSF coupled with heat inactivation at 95 °C for five minutes. Samples were analyzed both by SDS-PAGE and by ESI-MS.  2.7.4 NMR spectroscopy  Samples contained 0.5 to 2.5 mM protein in 20 mM potassium phosphate and 50 mM NaCl, with 10 % or 99 % D 0 added for the lock. All spectra were recorded at 30 °C 2  using a 500 MHz Varian Unity or 600 MHz Varian Inova NMR spectrometer equipped with a triple resonance gradient probe. All pulse sequences were provided by Dr. Lewis Kay (University of Toronto). Spectra analysis utilized Felix 2000 (Accelrys, Inc.)  70  Chapter 2 - Structural studies of the Erg PNT domain  and Sparky 3 (T. D. Goddard and D. G. Kneller, University of California, San Francisco). Backbone and aliphatic sidechain assignments were obtained from C13  and  15  N-HSQC,  HNCACB, CBCA(CO)NH, HNCO, HACAN, (H)CC(CO)NH-,  H(CC)(CO)NH- and HCCH-TOCSY spectra (Sattler et al., 1999; Kanelis et al., 2000). Stereospecific assignments were achieved using the method of Senn et al. (1989) for Val and Leu methyl groups and Mcintosh et al. (1997) for the sidechain amides of Gin and Asn. Assignment of the aromatic sidechains was as described previously (Slupsky et al., 1998b). H-{ N NOE} relaxation measurements used a control delay of 5 s for 1  15  the no-NOE spectrum, with a relaxation delay of 2 s preceding a 3 s proton presaturation period for the NOE spectrum (Farrow et al., 1994).  2.7.5 Dipolar coupling measurements  Phage (Pf1) were generated in Pseudomonas aeroginosa by using the protocol outlined by Hansen et al. (1998). Addition of phage was monitored via measured splitting of the H0 H signal using 1D H NMR spectroscopy, to a final concentration of 1  2  2  -17 mg/mL Pf1 and a corresponding H0 H splitting of 18 Hz. Amide H- N residual 1  dipolar couplings of [ N]Erg " 15  (108  201)  2  1  15  were measured using an IPAP-HSQC pulse  sequence (Ottiger et al., 1998). Measured dipolar couplings were analyzed using a histogram (as in Clore et al., 1998) with 2 Hz groupings, revealing an initial estimate of 7 Hz and 0.9 for the values of Da and R required for the SANI module in ARIA v1.2. Using an initial low-energy structure generated with these residual dipolar couplings, dihedral, hydrogen-bond, assigned methyl-methyl and methyl-amide NOE restraints (see section 2.7.6), a grid search of Da and R values was performed with variation from 6 to 10 Hz and 0.2 to 0.9, respectively. The lowest ARIA energies and minimal 71  Chapter 2 - Structural studies of the Erg PNT domain  SANI violations were obtained for a Da value of 8 Hz and an R value of 0.7 that were used for the final structure calculation described in section 2.7.6  2.7.6 Structure calculation  Distance restraints were acquired from a combination of 3D N-HSQC-NOESY, 3D 15  simultaneous C- and N-NOESY-HSQC, 2D H-NOESY and simultaneous constant 13  15  1  time methyl-methyl and amide-methyl NOESY spectra, all recorded with a mixing time of 150 ms (See Appendix 1). The methyl-methyl and amide-methyl NOESY spectrum were manually assigned, providing unambiguous 51 methyl-methyl and 54 methylamide proton distance restraints, all of which were constrained to a generous range of 1 - 6 A. The remaining NOE peaks were left unassigned, and included 1100, 1791 and 230 restraints from the 3D N-NOESY-HSQC, 3D simultaneous aliphatic C- and 15  15  13  N-NOESY-HSQC and 3D aromatic C-NOESY-HSQC spectra, respectively. 13  Dihedral angles of v|/ (57° ± 30°) and § (47° ± 30°) were included for residues in which a helical conformation was predicted in both TALOS and CSI analyses (Wishart and Sykes, 1994; Cornilescu et al., 1999) and confirmed by JHNHCX coupling constants less than 6 Hz (detailed in Figure 2.5; Kuboniwa et al., 1994). xi angles were included for threonine, isoleucine and valine based on the pattern of J N C and Jccy coupling 3  3  y  constants, determined from long-range C -N and C -C spectra (Grzesiek et al., 1993; Y  H  Y  Vuister et al., 1993). Two restraints per hydrogen bond were included for amide protons displaying protection from hydrogen-deuterium exchange, as measured by H1  15  N-HSQC spectroscopy 30 min and 1 h following transfer to D2O buffer. The  hydrogen-bond partners were assumed to be to the i-4 C residue for helical regions  72  Chapter 2 - Structural studies of the Erg PNT domain  only. Starting at iteration 4, 63 H - N residual dipolar coupling restraints were 1  N  15  included using the SANI protocol. Structure calculations were carried out using ARIA v1.2 (Nilges, 1995; Nilges and O'Donoghue, 1998; Linge and Nilges, 1999; Linge et al., 2001) with a starting extended structure of Erg " (108  201)  and the above set of  distance, angle and residual dipolar coupling restraints. After the full cycle of eight iterations, the initial number of unambiguous distance restraints increased to 1868 from an initial amount of 116. Conversely, the starting pool of 3002 unassigned ambiguous restraints were reduced to 945 valid cases in which the chemical shifts of the NOE peak correspond to two or more pairs of close proximity protons. In the final set of ten water-refined structures, there were no violations greater than 0.5 A and 5° for distance and angle restraints, respectively. The complete analysis of structure quality appears within the text (see Table 2.1).  73  Chapter 3 Heteromeric complex of Ste4 and Byr2  In contrast to the monomeric PNT domains from the Ets family of transcription factors, it is discovered that the Byr2 and Ste4 SAM domains indeed play a role in proteinprotein association. The two SAM domains interact in vitro in a manner that is dependent on the trimerization of Ste4 via a flanking coiled-coil domain. Surprisingly, only one Byr2 SAM domain binds to the Ste4 trimer to form a stabilized complex. This necessarily asymmetric interaction likely involves a high affinity site coupled with two lower affinity sites. In isolation, the Byr2 SAM domain is unfolded, but may fold upon binding to Ste4, where it displays increased helical content. The Ste4 coiled-coil domain, essential for Ste4 function, likely extends from the last helix in the Ste4 SAM domain and can form an isolated trimer in solution.  Part of the research in this chapter has been published in the Journal of Biological  Chemistry, 277:39585-39593 (Ramachander et al,  2002).  Specifically, my  contributions entailed independent study of the stoichiometry of the Ste4:Byr2 complex using native gel electrophoresis and gel filtration analyses, and investigation of Ste4 and Byr2 constructs using circular dichroism spectroscopy. I also provided significant input into the discussion of biological relevance presented at the end of the paper.  74  Chapter 3 - Heteromeric complex ofSte4 and Byr2 3.1 Previous evidence of Byr2 and Ste4 interaction  One of the goals of this thesis was to investigate the structural rules that govern the ability of PNT AND SAM domains to interact with each other. It is now clear that representative domains from the Ets family of transcription factors are limited to monomeric or self-associated polymeric oligomerization states, with no evidence of heteromeric complexes with other PNT domains. Amidst the search for other model systems that rely upon PNT and SAM domain interactions, it was noted that the most convincing description of a heteromeric SAM-SAM interaction was between the fission yeast Byr2 and Ste4 proteins that function in mating and nitrogen starvation signaling pathways. As was already discussed in Section 1.6, intact SAM domains were required for the function of these proteins (Figure 1.6A). A similar but not identical story has developed for the orthologous STE11 and STE50 proteins in S. cerevisiae that mediate mating, nitrogen starvation and osmotolerance responses. However, investigation of this second example at the molecular level is still in the initial stages. Structural studies on isolated STE11 and STE50 SAM domains have been conducted, but with conflicting in vitro binding data (L. Donaldson, York University and E. Laue, Cambridge, personal communication).  3.2 Byr2 binding to trimeric Ste4  DNA encoding minimal SAM domains were cloned from Byr2 (residues 1-70) and Ste4 (residues 1-77). These cloning boundaries were based on the Byr2 construct used in the yeast two-hybrid analyses by Tu et al. (1997), with the Ste4 construct continuing from the N-terminus to the C-terminus of the SAM domain as determined by sequence alignment to Byr2 (Figure 3.1 B; see also Material and Methods section 3.5). 75  Chapter 3 - Heteromeric complex of Ste4 and Byr2  A 1  71 SAM  1  Ste4  RBD 71  I  151 180 CBD  kinase catalytic domain  coiled-coil  •. •. -i  264  146  W//////M SAM  660  261  1 STE50 homology  B Construct  MW  Byr2< - )  H ^ H  8368  10810  8.01  Ste4( " )  I  9001  19630  4.63  1 70  1  77  H  s  (M- cnr ) 1  2 8 0  1  Pi  Stfi4  (1-157) |  W////////A  18445  22190  4.85  Stfi4  (52-157)  | W//////M  12386  2560  5.76  Ste4( " )  IW///////M  11008  2560  5.91  66  157  Figure 3.1. Constructs used to study the SAM domain mediated association between Byr2 and Ste4, and the trimeric coiled-coil of Ste4. All clones were expressed using a pET28a vector without any additional purification tags. Values for MW, molar absorptivity e and isoelectric point (pi) were predicted by the ProtParam Web-based service (http:// 280  76  Chapter 3 - Heteromeric complex of Ste4 and Byr2  The recombinant proteins were expressed in E. coli BL21(A,DE3) and purified to homogeneity. Subsequent characterization of the isolated domains revealed that although soluble and monomeric, Byr2 " displayed very limited a-helical content (1  70)  (estimated at 35% by CD spectroscopy), and lacked a disperse range of amide proton NMR chemical shifts that would otherwise indicate a fully-folded structure (Table 3.1; Figure 3.2). Consistent with a partially folded state, Byr2 " (1  70)  did not show any  cooperative change in CD ellipticity when subjected to thermal denaturation (Table 3.1). In contrast, Ste4 " had the hallmark CD spectrum of a protein containing the (1  77)  expected a-helical content (estimated at 57%) and a disperse 1D H-NMR spectrum 1  (Table 3.1, Figure 3.3). Although mainly monomeric when dilute, as evident by gel filtration chromatography, concentration of Ste4 " to levels required for NMR studies (1  77)  (>100 u.M) led to a degree of self-association. This association is indicated by the broadened amide peaks in a 1D H-NMR spectrum (due to slower molecular tumbling) 1  exhibited by larger multimeric complexes as compared to the narrow peaks observed for the monomeric Byr2 " (compare Figure 3.3B to Figure 3.2B). Alternatively, (1  Ste4 " (1  77)  70)  could exhibit conformational exchange broadening indicative of extensive  internal motions on a millisecond time-scale. Due to the unfolded nature of Byr2 " and the oligomerization or aggregation (1  70)  of Ste4 " , neither peptide was deemed suitable for detailed structural analysis by (1  77)  NMR spectroscopy. However, it was still possible that these two domains would exhibit the expected heterodimeric association and perhaps result in improved stability and solubility upon complex formation. Unfortunately, gel filtration failed to indicate the presence of any Byr2:Ste4 complexes when tested at concentrations that minimize  77  Chapter 3 - Heteromeric complex of Ste4 and Byr2  Table 3.1. Thermal denaturation values and secondary structure composition of the SAM domain fragments of Ste4 and Byr2.  Predicted a helix (%) structure" sequence  Construct  Tm (°C)  Byr2<-°)  unfolded  -8350  35  62  30  Ste4<- >  39  -15050  57  56  64  Ste4( - >  44  -20120  74  74  71  .  .  1 7  1 77  1 157  Ste4:Byr2 o.. „ '. 3:1 complex v  a  46  9  a 2 2 2  a helix (%)  c  . nd  e  nd  nd  nd  measured by C D spectroscopy as detailed in the Materials and Methods (section 3.1.2) in 20 mM potassium phosphate (pH 7.0) and 50 mM N a C l . All Ste4 constructs were measured at 15 L I M relative to the monomer units, with Byr2< - ) at 5 L I M . The value of T was taken as the inflection point in the transition from fully folded to fully unfolded protein. 3  1  70  m  Percentage a-helix is calculated using the equation (-0 1972)  b  222  + 2340) / 30300 (Chen et al.,  predicted using the average helical content from sequence aligned S A M domains from the P D B database: polyhomeotic (PDB ID 1kw4), chicken EphB2(1sgg), human EphB2 (1f0m, 1b4f), mEphA4(1b0x), and human p73 (1cok, 1dxs). For Ste4( " > it was assumed that all residues C-terminal to the S A M domain were helical up to residue 145 (the predicted end of the coiled-coil region from Figure 3.4). c  1  157  predicted using the P R O F algorithm as part of the PredictProtein web-based tool (, an improved version of the P H D neuralnetwork based secondary structure alorithm (Rost, 1996).  d  78  Chapter 3 - Heteromeric complex of Ste4 and Byr2  A  _  !  ,  ,  !  !  r  11  10  9  8  7  6  1  12  ppm  B  -25000 200  210 220 230 240 Wavelength[nm]  250  Downfield region of a 1D H-NMR spectrum illustrating limited chemical shift dispersion of the amide proton resonances. Note the sharp cluster of peaks between 8 to 8.5 ppm indicative of a random coil conformation, and only a few peaks downfield of 9 ppm that indicate amides or indoles in a structured domain. (B) CD spectroscopy reveals only a limited amount of helical character as evident from the small amount of ellipticity at 222 nm, and the pronounced minima at 207 nm. Spectra were recorded in 20 mM potassium phosphate (pH 7.0) and 50 mM NaCl. The H- N HSQC spectra for Byr2<- > is illustrated in Figure 3.7C. Figure 3.2. Byr2< - > is partially folded in solution. (A) 1 70  1  1  15  1 70  79  Chapter 3 - Heteromeric complex of Ste4 and Byr2  n 12  1  1  11  10  1  1  1  r  9  8  7  6  ppm  0  = UJ ^  -5000  -10000  -15000  -20000  -25000 200  210  220  230  240  250  Wavelength[nm]  Figure 3.3. Ste4 is folded in solution, but exists as an oligomer or 1 15 77 aggregate. (A) 1D H-NMR spectrum of 0.2 mM [ N]Ste4C- > illustrating (177)  large chemical shift dispersion of the resonances from amide and indole protons due to the presence of a stable folded core. However, the broad clump of peaks are indicative of either a slower tumbling, larger molecular system suggestive of a higher oligomeric state, or instead a significant degree of conformational exchange broadening. (B) CD spectroscopy confirms the presence of a-helices with a pronounced minimum at 222 nm in 20 mM potassium phosphate (pH 7.0) and 50 mM NaCl.  80  Chapter 3 - Heteromeric complex of Ste4 and Byr2  Ste4 association (-20 u.M). Parallel to our research, the laboratory of Dr. James Bowie at UCLA has also produced minimal SAM domains from Byr2 and Ste4 in order to analyze their interaction, and hopefully determine a crystal structure of the heteromer. Although there are subtle differences between their constructs and the ones used in this thesis, they also observed poor behaviour of the isolated SAM domains. In addition, Ranjini Ramachander and others in the Bowie lab have more recently used surface plasmon resonance to derive a modest K of 56 u.M for their Byr2:Ste4 SAMd  SAM dimer (Ramachander et al., 2002). This finding is consistent with the lack of association observed in the gel filtration assays detailed above, since the concentrations used fall below this measured dissociation constant. Inspection of the domain structure of Ste4 (Figure 3.1 A) reveals a putative coiled-coil region directly following the SAM domain. This could possibly exist as a helical extension from the expected final a-helix H5 (residues 56-77) of the predicted model of the Ste4 SAM domain structure (based on the reported structures of related SAM domains). Two prediction algorithms, COILS (Lupas et al., 1991) and MultiCoil (Wolf et al., 1997) were used to estimate the length of this coiled-coil region (Figure 3.4). Both methods defined residue 145 as its C-terminal end, with conservative estimates of residue 59 or 71 for the N-terminus of the coiled-coil, using COILS and MultiCoil respectively. Furthermore, the MultiCoil algorithm provided a mixed prediction of dimeric or trimeric coiled-coils. Due to the possible co-dependent importance of the coiled-coil region and SAM domains for optimal function and behaviour in vitro, a second construct of Ste4 was created (residues 1-157, Figure 3.1) that included both the SAM domain, the entire coiled-coil region and an extra twelve residues at the C-terminus to ensure inclusion of 81  Chapter 3 - Heteromeric complex of Ste4 and Byr2 win=14 -win=21 -win=28  0.8  I  0.6  n  <o o  a.  0.4  ttm  0.2 0  50  100  150  200  250 -Total  B  Dimer -Trimer  50  100  150  200  250  Residue  COILS MultiCoil  H2 ~H2H—i H3 i — i H4 i — r n r MGDSDDFYWNWNNEAVCNWIEQLGFPHKEAFEDYHILGKDIDLLSSNDLRDMGIESVGHR fq I H5 I IDILSAIQSMKKQQKDKLQQENKDQELKNIEESYKKLEEKTEHLSDDNVSLEKRVEYLET abcdefgabefgabcdefgabcdefgabcdefgabcdefgabcdefgabcdefgabcdef fgabcdefgabcdefgabcdefgabcdefgabcdefgabcdefgabcdef  COILS MultiCoil  ENTKLVKTLNSLNSEFLQLLRKIAINVKEGRQLTTENSSDTSSMTHPVQPSPSVLGSFDL gabcdefgabcdefgabcdefgabc gabcdefgabcdefgabcdefgabc  180  EVNDSLTNAEKNRKLNVNLTYNEVLCSMLQRYRIDPNTWMSYDLLINYDDKEHAIPMDVK  24 0  PLQLFRNLQKRGKSPSFVLSRRSC  264  COILS  60 12 0  Figure 3.4. Coiled-coil predictions for full-length Ste4. Calculations were made using (A) COILS 2.2 with MTIDK matrix and windows of 14, 21 and 28 residues ( software/COILS_form.html; Lupas etai., 1991), and (B) Multicoil, (; Wolf et al, 1997). (C) Primary sequence of Ste4 with predicted heptad repeat from COILS (window of 28) and MultiCoil (total) shown underneath. The predominantly hydrophobic positions a and d are shown in bold. Predicted secondary structure for residues from 1-77, derived from the EphB2 SAM domain (PDB 1b4f), is shown above the sequence. 82  Chapter 3 - Heteromeric complex of Ste4 and Byr2  any additional folded residues. The expressed and purified Ste4 " (1  157)  construct  displayed by CD spectroscopy the high a-helical content expected from the addition of a helical coiled-coil region (74%), with increased thermal stability reflected by a T of m  44 °C compared to that of 39 °C for the isolated Ste4 SAM domain (Table 3.1, Figure 3.5A). Surprisingly, this larger Ste4 construct formed a stable trimer in solution with no significant evidence of any dimer or monomer species during the gel filtration stage of purification. A similar construct generated in the Bowie laboratory, Ste4 " -HiS6 also (1 152)  displayed a trimeric oligomerization state confirmed by an equilibrium sedimentation molecular mass of 57.5 kDa (Ramachander et al., 2002). Previously characterized structures of trimeric coiled-coil proteins provide examples of both symmetric and asymmetric architectures (Harbury et al., 1993; Lovejoy et al., 1993; Harbury et al., 1994; Betz et al., 1995). From the 2D H- N 1  15  TROSY-HSQC spectrum of Ste4 " , there are only three H- N cross-peaks in the (1  157)  1  15  region specific for tryptophan side chains (boxed in Figure 3.5B). These likely correspond to the three tryptophan residues in the monomeric Ste4 " (1  157)  sequence (i.e.  Trp9, Trp11 and Trp19). Since there is no evidence for three peaks from each Ste4 monomer (i.e. nine peaks in total) it is likely that each Ste4 " (1  157)  is in the same  chemical environment and therefore symmetric. This scenario is also supported by the observation of only -150 H- N amide resonances in the H- N TROSY-HSQC 1  15  1  15  spectrum (Figure 3.5B), equivalent to the number of residues in a single Ste4 " (1  molecule. Distinct chemical environments for two or three of the Ste4 " (1  157)  157)  molecules  would produce -300 or -450 peaks, respectively. Thus each monomer is in the same head-to-tail orientation around a long central axis of symmetry. From the same spectrum, a selective view of the most intense peaks (Figure 3.5C) reveals only 10-12 83  Chapter 3 - Heteromeric complex of Ste4 and Byr2  200  210  220  230  240  250  Wavelength[nm]  c E o.  o  flexible backbone amides  Q.  (1157)  0 •  <H (ppm)  H (ppm)  Figure 3.5. Ste4  "•"Hi  °«  forms a stable trimer in solution. (A) CD  spectroscopy reveals a large amount of a-helical content in the trimer as expected from the addition of a presumably all helical coiled-coil region. (B) 2D H - N TROSY-HSQC spectrum of the large 54 kDa trimeric complex has both dispersed but weak peaks that may originate from the SAM domain, plus a large overlapping central area between proton chemical shifts of 8 and 8.5 ppm. This overlapping region likely results from the extended coiledcoil that is sparse in aromatic side chains and therefore displays only limited amide resonance dispersion. In addition, possible flexibility or detrimental relaxation properties may contribute to the poor quality of the spectrum thus precluding facile interpretation. (C) Same as B but plotted with a 10-fold higher contour, with sidechain asparagine and glutamine amides at the top right of the spectrum and the flexible backbone amides boxed. 1  15  84  Chapter 3 - Heteromeric complex of Ste4 and Byr2  backbone amides that display a significant degree of conformational flexibility. In the context of the large trimer, these likely correlate to unstructured termini, possibly the same C-terminal residues sensitive to trypsin proteolysis in Figure 3.9B.C. Thus it appears that most of the construct indeed corresponds to an otherwise completelystructured trimer. Indeed, if the coiled-coil region is assumed to continue from the last helix of the SAM domain until residue 145, the measured helical content corresponds exactly to the predicted helicity from simple structural modeling (Table 3.1). Further characterization of Ste4 " (1  157)  by NMR spectroscopy was generally  confounded by the large size of the trimeric peptide (54 kDa). In addition, spectral quality was poor due to either a high degree of conformational exchange broadening or by unfortunate relaxation parameters. This latter scenario could arise from the significantly anisotropic nature of the Ste4 trimer, which is predicted to be rod-like in solution (Ramachander et al., 2002). The series of amide N- H bonds in the coiled15  1  coil should be predominantly parallel to the long central axis of symmetry. Compared to the more rapid rotation expected to occur around this long axis, the slower endover-end tumbling would produce an effective x longer than expected for a globular c  protein of the same size. The increased x would lead to anomalously short T2 values c  and decreased signal intensity. When combined in solution, Byr2 " and Ste4 " (1  70)  (1  157)  formed a stable and well-  defined complex as judged by native gel electrophoresis and gel filtration binding studies. Surprisingly, further characterization of the resulting Ste4:Byr2 complex revealed that in the final heteromeric ensemble, only one molecule of Byr2 " binds (1  70)  to one trimeric Ste4 " . Figure 3.6 illustrates a native gel electrophoresis binding (1  157)  assay used to derive the stoichiometry of Byr2 and Ste4 in the complex. At a Byr2 " (1  85  70)  Chapter 3 - Heteromeric complex of Ste4 and Byr2  I  in  11  *H  Ste4:Byr2 complex  $-4* Ste4 trimer  1:3  1:2  1:1  2:1  3:1  4:1  6:1  8:1  6  7  Ste4:Byr2 molar ratio  B  to c o  T3 C  o  2  3  4  5  Ste4:Byr2 molar ratio  Figure  3.6. Native gel  electrophoresis assay to  determine  stoichiometry of the Ste4/Byr2 complex. (A) Coomassie stained native  gel illustrating the titration of Byr2 ~ > with increasing amounts of Ste4 ~ > followed by separation of the bound and free Ste4< - ) by native gel electrophoresis. Under these conditions, no free Byr2 " was observed. (B) Using the integrated band intensity of the complex as a function of the protein monomer molar ratios, the stoichiometry was estimated at 2.8 moles S t e 4 per mole Byr2 . The error bars represent the range of values from three separate experiments, each performed in duplicate. (1  70  (1  1 157  (1  (1_157)  (1_70)  86  70)  157  Chapter 3 - Heteromeric complex of Ste4 and Byr2  concentration of 20 u.M, Ste4 " was added at various molar ratios with subsequent (1  157)  separation of the bound and free forms of Ste4 " (1  157)  upon electrophoresis. Since the  assay was performed at a pH of 7, Byr2 " was excluded from the gel due to its (1  70)  positive charge at this pH (predicted pi equals 8.0, see Figure 3.1 B) causing it to migrate in the opposite direction. Based on the intensity of the stained band for the complex, the titration was complete at a 3:1 ratio of Ste4 " (1  157)  to Byr2 " monomer (1  70)  units. In order to verify this unusual finding, a collaboration was initiated with James Bowie to further characterize the complex using the above Ste4 " (1  their HiS6-tagged Byr2 " (1  72)  157)  construct and  protein. Equilibrium sedimentation yielded a molecular  mass of 67,300 Da for the Ste4:Byr2 hetero-oligomer, which is comparable to the predicted mass of 67,600 Da for a complex containing one Byr2 " and three Ste4 " (1  157)  72)  (1  molecules. In addition, his research group measured a remarkably low K (19 nM) d  for the binding of the Byr2 SAM domain to the Ste4 " (1  157)  trimer, as determined using  surface plasmon resonance (Ramachander et al., 2002). Therefore, Ste4 " (1  indeed bind Byr2 " (1  70)  157)  does  in a 3:1 molar ratio, and with an enhanced affinity of  approximately 2000-fold compared to the 1:1 complex of their isolated SAM domains. The discovery of a tight heteromeric complex lent promise to a more suitable target for NMR spectroscopic analysis, with the intriguing goal of explaining how a single non-symmetric Byr2 SAM domain could bind with such high and specific affinity to a cluster of the three Ste4 SAM domains. Characterization of the complex with CD spectroscopy indicated that binding of the Byr2 SAM domain to the Ste4 trimer resulted in an increased thermal stability of the complex as evident by a small elevation in the global T from 44 °C for Ste4 m  Byr2 " (1  70)  (157)  alone to 46 °C (Table 3.1). Note that  did not display any cooperative unfolding transition when alone. This 87  Chapter 3 - Heteromeric complex of Ste4 and Byr2  elevated T is expected due to the thermodynamic linkage of binding and folding. In m  addition, a careful comparison of the CD spectra of the Byr2 " , Ste4 (1  70)  (157)  and  Ste4:Byr2 complex indicated a small amount of induced helical content that resulted upon complex formation (Figure 3.7A.B). Although it is possible that either Byr2 " or (1  Ste4 " (1  157)  70)  was the source of this increased a-helix, it is most probable that this results  from stabilization of the structure of the Byr2 " SAM domain upon binding. Indeed, (1  70)  the 2D H- N HSQC spectra of N-labeled Byr2 " displayed a noticeable change of 1  15  15  (1  70)  the amide resonances upon association with trimeric, unlabeled Ste4 " (1  157)  (Figure  3.7C,D). Unfortunately, the large size, molecular flexibility and perhaps poor relaxation properties of the Ste4:Byr2 complex again prohibited the acquisition of NMR spectra suitable for structural studies, even when TROSY-based pulse sequences were used to help offset the deleterious relaxation effects normally effected by the slower tumbling of a large molecule. Therefore, further analysis of the complex by NMR was not performed.  3.3 Coiled-coil of Ste4 A key finding from the previous study of Ste4 " (1  157)  was the identification of a trimeric  coiled-coil region that had dramatic effects on the affinity of Ste4 towards the Byr2 SAM domain. Since there is minimal literature on the structural details concerning natural trimeric coiled-coils, it was decided to further characterize this functional domain from Ste4. Based on the predicted limits of the coiled-coil region from both the COILS and MultiCoil methods (Figure 3.4), two constructs were expressed and purified (residues 52-157 and 66-157; Figure 3.1 B). Both Ste4 (52  157)  and Ste4 (66  157)  formed soluble proteins with modestly dispersed 1D H-NMR amide spectra expected 1  88  Chapter 3 - Heteromeric complex of Ste4 and Byr2  Figure 3.7. Structural characterization of the binding of Byr2 to trimeric Ste4 " . (A) Circular dichroism spectroscopy of isolated 5 u.M Byr2< - °), 1 5 Ste4< - > (i.e. 5 u M of the trimer), and the resulting 5 u M Ste4:Byr2 3:1 complex. (B) The induced ellipticity following subtraction of the uncomplexed Byr2<- > and Ste4< - > spectra from that of the Ste4:Byr2 complex. The significant increase in helical content is presumably due to the folding of Byr2<- > upon binding. (C,D) The change in the 2D H- N TROSYHSQC spectrum of [ N]Byr2< - > from the free (C) to complexed form (D). Such changes are in keeping with an altered chemical environment resulting from the bound Ste4 S A M domains, possible change in tertiary structure, elevated stability in fold and an overall increase in molecular weight. (170)  (1  157)  1 7  1 157  1 70  1 157  1 70  1  15  1 70  89  15  Chapter 3 - Heteromeric complex of Ste4 and Byr2  1  0  a  ''  ^ ......... r .  .^  ^  Figure 3.8. Ste4< ) and Ste4< ) form large folded complexes. 1D HNMR spectra of Ste4< - ) (A) and Ste4< - > (B). The small dispersion of amide and methyl resonances confirm the absence of a highly structured globular core and instead reflect a combination of limited aromatic residue content, and a repeat sequence of residues within a helical conformation. The broadened peaks reflect the large size (33 or 37 kDa, respectively) of the trimeric coiled-coils and possibly unusual relaxation properties from anisotropic tumbling. Representative H- N HSQC spectra for these two proteins can be found in Figure 3.11 66157  52157  1  66 157  52 157  1  90  15  Chapter 3 - Heteromeric complex of Ste4 and Byr2  from a trio of three lengthy a-helices in a parallel symmetric trimer (Figure 3.8). Chemical cross-linking confirmed that even the smaller Ste4 " (66  157)  construct formed an  exclusive trimer, with no evidence of monomeric or dimeric sub-species (Figure 3.9A). Further analysis with trypsin digestion revealed that only the final C-terminal residues were susceptible to facile proteolytic cleavage, possibly indicating residues that are excluded from the preceding coiled-coil (Figure 3.9B.C). Note that these C-terminal residues were previously postulated to be highly disordered in the Ste4 " (1  157)  trimer  based on the presence of a small number of sharp peaks in the H- N HSQC of this 1  15  species. This hypothesis could be further investigated with proteolysis of a N-labeled 15  sample of Ste4 " followed by characterization of the tryptic fragments by NMR (1  66)  spectroscopy. Using Figure 3.9C as a guide, it would be expected that these sharp peaks would be eliminated following the first cleavage event after a 10' incubation with trypsin. Eventual cleavage at the N-terminal residues and in discrete regions within the peptide indicate less structured or conformationally flexible regions within the coiledcoil. Consistent with this flexibility is the lack of protection of amide hydrogen exchange following a 30 minute buffer conversion to D2O (not shown), as well as possible line broadening seen in the NMR spectrum of the protein. A model of the prototypic coiled-coil trimer (Figure 3.1 OA) can be used to explain some of the features of Ste4 " (52  157)  and Ste4 " . Based primarily on (66  157)  engineered trimeric and tetrameric coiled-coils, it has been established that residues in the a and d position of the repeating heptad sequence motif define the interior of the coiled-coil, much as is the case with the more common dimeric coiled-coils and leucine zippers (Figure 3.1 OB; Betz et al., 1995; Harbury et al., 1993; Harbury et al., 1994).  91  Chapter 3 - Heteromeric complex of Ste4 and Byr2  Std  UJM  0'  ft  2'  5'  10' 30'  ft  ftlft  60' 90' 120' 150' 0'  UJ  M  Jl  U  AIQSMKKQQKDKLQQENKDQELKNIEESYKKLEEKTEHLSDDNVSLEKRVEYLETENTKLVKTLNSLNSEFLQLLRKIAINVKEGRQLTTEN  Figure 3.9. Ste4 is trimeric and flexible in solution. (A) Cross-linking of lysine residues in Ste4 ~ with BS revealed only trimeric complexes, with no evidence of dimeric species. (B) Digestion by trypsin followed by SDSPAGE and Coomassie staining indicated a highly sensitive site cleaved within minutes of enzyme addition, with additional sensitive sites appearing from 30' onwards. (C) Sequence of Ste4 " with all potential trypsin cleavage sites indicated by an arrow, with cleavage sites observed after 10' and 90' indicated in black or grey colouring, respectively. ESI-MS was used to identify the cleavage products (data not shown). (66_157)  (66  157)  3  (66  157)  92  Chapter 3 - Heteromeric complex of Ste4 and Byr2  Figure 3.10. Sequence analysis of the Ste4 coiled-coil region. Schematic of a prototypic parallel coiled-coil trimer (A) and dimer (B) illustrating the central hydrophobic core created by residues in the a and d position, and potential ionic interactions between side chains occupying positions e and g. (C) Sequence of Ste4 " with labels a to g corresponding to the heptad repeat. (D) Alternative representation of the Ste4 sequence as a helical wheel to emphasize the preponderance of leucine at position d, and the mainly hydrophobic but highly varying content of residues at position a. Also evident are the predominantly hydrophilic and charged residues around the remainder of the amphipathic helix. (66  157)  (66_157)  93  Chapter 3 - Heteromeric complex of Ste4 and Byr2  66  70 80 90 100 110 AIQSMKKQQKDKLQQENKDQELKNIEESYKKLEEKTEHLSDDNVSL abcdefgabcdefgabcdefgabcdefgabcdefgabcdefgabcd 120 130 140 150 EKRVEYLETENTKLVKTLNSLNSEFLQLLRKIAINVKEGRQLTTEN efgabcdefgabcdefgabcdefgabcdefgabcdefgabcdefga  AQQLYTNVNL FIGN MKDEESEEVNLVT (?)  /  k  \  gfc S D K I L L L L L L L N L  IQEKKEVETNLAR (JJXA KLQEEDKTKSRKT 0 /  KQESKDRETEKEE ^ © QKNNKHSYKSQIQ  94  157  Chapter 3 - Heteromeric complex of Ste4 and Byr2  Applying the heptad repeat to the sequence of the smaller Ste4 " (66  157)  construct (Figure  3.10C), emphasizes the hydrophobic residues at positions a and d of the coiled-coil motif, a feature that is more evident in the helical wheel representation of the same peptide (Figure 3.10D). From this latter picture of the coiled-coil, the high leucine content at position d is evident, in keeping with structural studies of other coiled-coil leucine zipper domains (first identified by Landschulz et al., 1988). Although mainly hydrophobic, position a contains a contrasting variance in amino acids with the inclusion of a tyrosine, valine, phenylalanine, isoleucine, glutamine and threonine in addition to only two leucine residues. The presence of isoleucine and valine residues at position a has been proposed to define the ability of an amphipathic helix to adopt a trimeric structure. There is also at least one described coiled-coil trimer that incorporates a phenylalanine and tyrosine (Kovacs et al., 2002). However the surprisingly wide variance in amino acid types appears to be a distinct property of Ste4. The presence of the hydrophilic threonine, glutamine and asparagine residues within the otherwise hydrophobic core of the coiled-coil may contribute to the relatively low stability and the conformational flexibility observed for the Ste4 trimer, and explain the intra-coil cleavage during the trypsin digest of Ste4 " . However,, a comparable (66  157)  buried asparagine located in the c-Jun leucine zipper (Asn22) does not contribute to increased backbone dynamics as measured using N NMR relaxation (MacKay et al., 15  1996). Instead, Asn22 sets the register of the coiled-coil through hydrogen-bonding to the symmetry related Asn22 on the second peptide. Alternatively, the observed instability in the Ste4 " (66  157)  protein complex could derive from the mixed preference of  a trimer for the first half of the coiled-coil and a dimer for the second half (as predicted by MultiCoil in Figure 3.4B). In addition, the helical wheel representation of Ste4 " (66  95  157)  Chapter 3 - Heteromeric complex ofSte4 and Byr2  (Figure 3.10D) indicates that the charge-pairing between the e and g positions is not always optimal, with the possible presence of adjacent acidic residue pairs as well as adjacent basic residue pairs. The conformational flexibility or unfavourable anisotropic relaxation properties of both Ste4 trimeric coiled-coils, similar to the characteristics of the larger Ste4 " (1  157)  construct, once again precluded detailed structural studies by NMR. Figure 3.11 details the matrix of parameters that were scanned in an attempt to optimize conditions for higher quality NMR spectra. Varying the sample temperature, pH and salt concentration did not result in conditions providing NMR spectra amenable for structural studies. In the case of a 2D H- N HSQC spectrum, these criteria would 1  15  include the clear observation of all amides present in the construct (i.e. one would expect to see 121 peaks for the backbone and sidechain amides of Ste4 " ) with (52  157)  minimal overlap of these peaks. Instead, the spectrum exhibits limited dispersion with peaks clustered between proton chemical shifts of 8.0-8.5 ppm, and with far fewer signals than expected for Ste4 " . Therefore, it appears that the Ste4 coiled-coil, (52  157)  although capable of forming an elongate trimer in solution, contains significant dynamic behaviour or unfortunate relaxation properties that prevents NMR characterization. Nevertheless, the possible flexibility may play a critical role in the natural biology of S. pombe.  3.4 Biological implications From the data presented in the previous sections of this chapter, it is evident that the tight binding between Ste4 and Byr2 is only accomplished through the trimerization of the Ste4 protein (Figure 3.12). It is possible that this improved affinity by Ste4 " (1  96  157)  is  Chapter 3 - Heteromeric complex of Ste4 and Byr2  90  B5  10  76  70  |Q  fl 0  B*  75  :  TO  IV5  }0  80  7S  o  ° s St  f  A  o 0  40 °C  30 °C  20 °C  B ;  c  — cr. ^  IT~  ,  ...  ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^  o  pH 7.0  pH6.0  pH8.0  D  c °  o  O  o  500 mM  0 mM NaCl 1 0  '  5  1 0  raj  )0  85  co, -'M (ppm)  .  NaCl  BO 73 w.-'Hlppml  7.0  S t e 4  0.0  85  80 u, •'H (ppmj  (66-157) 7.S  7.0  Figure 3.11. Screen to optimize Ste4< > for NMR spectroscopy. Despite changes in temperature (A), pH (B) or salt concentration (C) the 2D H- N HSQC spectra of [ N]Ste4 ' did not change significantly towards an optimal dispersed pattern of strong amide H- N chemical shifts in a 1:1 ratio of peaks to amino acid residues. A representative spectrum of [ N]Ste4 ~ is included (D) to show similar behaviour for this construct, and the presence of several similar chemical shifts that confirm that the same tertiary structure is present. 52_157  1  15  15  (52 157)  1  15  15  97  (66  157)  Chapter 3 - Heteromeric complex of Ste4 and Byr2  Figure 3.12. Schematic representation of the Ste4:Byr2 complex. (A) A single molecule of Byr2 - °) binds to a parallel and symmetric trimer of St 4(i-157) Detail of the SAM-SAM interactions that incorporate a single high affinity site (B) and two low affinity sites (C), as described for association between the isolated SAM domain monomers. Note that NMR spectroscopy (e.g. Figure 3.7D) indicates symmetry averaging, and therefore it is possible that Byr2 rotates rapidly amongst the three Ste4 SAM domains in order to create only one set of amide and indole resonances. (1 7  e  98  Chapter 3 - Heteromeric complex of Ste4 and Byr2  due to some indirect influence by the coiled-coil domain apart from trimerization, such as a change in Ste4 SAM domain structure, or that the coiled-coil is somewhat involved in the Byr2 binding site. However, the isolated Ste4 SAM domain had characteristics of a well-folded protein, which makes this hypothesis less likely. Furthermore, the members of the Bowie lab were unable to detect binding of the Byr2 SAM domain to the isolated Ste4 coiled-coil (Ramachander et al., 2002). The importance of the coiled-coil domain has also been shown in vivo with a significant decrease in sporulation resulting from its deletion from Ste4 (Okazaki et al., 1991). Removal of the coiled-coil presumably results in decreased affinity of Byr2 and Ste4 thus leading to a defective pheromone response. At the molecular level, the binding of a single Byr2 SAM domain to the three Ste4 SAM domains is easiest to explain if one considers three distinct binding sites on Byr2 - one high affinity site and two additional low affinity regions (Figure 3.12B.C). The high affinity site reflects the interface detected by the Bowie group in the study of isolated SAM domain. A modest K of 56 u.M was established for this interaction d  (Ramachander et al., 2002). The remaining low affinity sites presumably have weaker binding constants when one considers isolated monomers. However, within the context of the trimer, they can contribute to increased affinity since loss of entropy encountered upon binding has already been satisfied with the high affinity site. Via this classic chelate effect, the additional two binding regions enhance the specificity of the interaction and create the tight complex observed between Ste4 " (1  157)  and Byr2 " . (1  70)  Apart from the clear requirement for the interaction of Ste4 with Byr2 for the mating process in S. pombe, the actual mechanism by which this complex mediates the pheromone response still remains unresolved. An earlier model of Byr2 activation 99  Chapter 3 - Heteromeric complex of Ste4 and Byr2  invoked a role for Ste4 in the oligomerization of uninhibited Byr2 at the plasma membrane, following association with active Ras-1 via the Ras-binding domain of Byr2 (Tu et al., 1997). Bringing kinase domains together in space is a prominent theme of other signaling systems such as the receptor tyrosine kinase pathways (Schlessinger, 2000), in which ligand-mediated dimerization results in autoactivation and phosphorylation of the kinase followed by downstream signal transduction with additional regulatory proteins (Figure 3.13A). However, this model is not consistent with the data presented in this thesis indicating that Ste4 does not result in the oligomerization of Byr2. That is, only one Byr2 is bound by the Ste4 trimer. Instead, separate regions of Byr2 or Ste4 must be important in the context of binding additional proteins involved in the signaling process. The C-terminal homology region of Ste4, which displays a high sequence identity to the orthologous STE50 protein in S. cerevisiae (Section 1.6.1, Figure 1.6B), has also been shown to result in a sterile phenotype in yeast upon site-directed deletion (Okazaki era/., 1991). Therefore, this region could provide the key link in coupling the Ste4:Byr2 complex to additional components of the mating pathway. Further analysis of the Ste4 C-terminal sequence reveals a modest relation to the Ras-associating (RA) domain as defined by the SMART database (Schultz et al., 1998), suggesting that Ste4 could bind to Ras-1 distinct from the Ras-1 association mediated by the Ras-binding domain (RBD) on Byr2. However, crystal structures of the RBD and RA binding sites on Ras-1 are highly overlapping, which would preclude this two-fold binding (Nassar et al., 1995; Gronwald er al., 2001; Scheffzek et al., 2001; Ramachander et al., 2002). Ste4 could still bind to a separate molecule of activated Ras (Figure 3.13B). It has also been found in vivo that Ste4 is not required for the translocation of Byr2 to the yeast membrane upon 100  Chapter 3 - Heteromeric complex of Ste4 and Byr2  Figure 3.13. Potential biological role for Ste4. (A) Early model where dimerization of Ste4 brings together the catalytic domains of two Byr2 proteins leading to autophosphorylation and downstream signaling (Tu et al., 1997). Byr2 is localized to the membrane following activation of Ras (indicated by an asterisk). This model does not match the data presented in this thesis (i.e. Ste4 is a trimer, not dimer, and binds to only one molecule of Byr2). (B) Possible interaction between the RA domain of Ste4 with activated Ras. Note that based on structural studies of Ras complexes, Byr2 and Ste4 cannot simultaneously bind to the same Ras protein due to overlapping binding surfaces (Nassar et al., 1995; Gronwald et al., 2001; Scheffzek et al., 2001). However, a mutational study has shown that Ste4 is not required for the membrane translocation of Byr2 (Bauman et al., 1998). (C) It is most likely that the RA domain of Ste4 associates with as yet unknown components of the fission yeast mating pathway. Domains present in Ste4 and Byr2 have been detailed in B (see Figure 1.7). 101  Chapter 3 - Heteromeric complex of Ste4 and Byr2  stimulation by pheromones (Bauman et al., 1998). Therefore, the ancestral function of the Ste4 RA domain is likely to have given way to a new binding module, whose partner has not yet been found (Figure 3.13C).  3.5 Materials and methods 3.5.7 Cloning and purification  All Byr2 and Ste4 constructs were cloned using PCR amplification from a Schizosaccharomyces pombe cDNA library generously provided by Dr. Sean Hemmingsen (NRC in Saskatoon, Canada). Restriction sites Nco\ and Hin6\\\ were introduced into the 5' and 3' PCR primers, respectively, to allow for directional cloning into pET28a so as to exclude the N-terminal tag from the vector. Cloning of Byr2 " , (1  Ste4 " , Ste4 " , Ste4 " (1  77)  (1  157)  (52  157)  and Ste4 (66  157)  70)  used appropriate pairs of the following  primers: Byr2-1F, 5-GGCGAATTCCCATGGAATATTATACC-TCG-3': Byr2-70R, 5'CCGC-AAGCJTC/\TGGACGTGGAAACTCTCG-3'; Ste4-1F, 5'-GGCGAATTCCAT-  Ste4-52F, 5-GGCGAATTCCATGGGAATTGAGAGTGTG-  GGGAGATTCTGACG-3';  3';  Ste4-66F, 5'-GGCGAATTCCATGGCAATACAGTCAATG-3'; Ste4-77R, 5'-  CCGCAAGCTTC/4TTTATCCTTCTGCTGC-3'; Ste4-157R, 5'-CCGCAAGCTTC>AATTTTCGGTAGTAAGC-3'  (with introduced restriction sites underlined, stop codons in  italics, and coding bases shown in bold). Accuracy of cloning was confirmed by DNA sequence analysis using both T7 forward and reverse primers targeted to regions on the plasmid flanking the inserts. Plasmids were transformed into E. coli BL21(XDE3) for Ste4 constructs or HMS174(A,DE3) for Byr2 " . Proteins were expressed in bacteria grown in LB media (1  70)  or minimal M9 media supplemented with 1 g/L NH CI, with induction at 15  4  102  OD oo 6  ~ 0.6  Chapter 3 - Heteromeric complex of Ste4 and Byr2  using 1 mM IPTG. After growth for 3 h at 30 °C, cells were collected by centrifugation at 3,000 x g, and resuspended in 50 mM Tris, pH 7.5 for Ste4 or 50 mM potassium phosphate, pH 6.5 for Byr2 " , with added complete protease inhibitor (Boehringer (1  70)  Mannheim). Cells were disrupted by two cycles through a French press at 10,000 psi, followed by 1 min of sonication. Lysates were cleared with centrifugation at 10,000 x g for 30 min followed by filtration through a 0.8 u.m cut-off membrane. All Ste4 proteins were purified by FPLC in 50 mM Tris buffer and a salt gradient of 0 to 500 mM NaCl using Q-sepharose anion exchange (Pharmacia). The first purification was at pH 7.5, with the protein eluting at -300 mM NaCl for Ste4 ' , and -200 mM NaCl for Ste4 " (1 157)  77)  , Ste4 (52  157)  (1  and Ste4 ' . Purification of Ste4 " was followed by a second (66 157)  (1  77)  column at pH 8.5, with elution at -250 mM NaCl. For Ste4 " , a final purification step (1  157)  by Sephacryl S-100 size exclusion chromatography (Pharmacia) in 20 mM potassium phosphate (pH 7.0) and 50 mM NaCl was used to isolate the trimeric complex from small molecular weight protein contaminants. Ste4 " (52  157)  and Ste4 " (66  157)  were further  purified by reverse-phase HPLC in 0.1% trifluoroacetic acid (TFA), with a gradient of 0100% acetonitrile and elution at -20% acetonitrile. Byr2 " was purified by FPLC in (1  70)  50 mM potassium phosphate (pH 6.5) and a salt gradient of 0-300 mM NaCl with a SP-sepharose column, with the protein eluting at -100 mM NaCl. Protein samples were stored at 4 °C in 20 mM potassium phosphate (pH 7.0) and 50 mM NaCl. For each protein, purity of > 90% was ascertained by using standard SDS-PAGE and Coomassie staining, with protein identity confirmed by using ESI-MS. Protein concentrations were determined by absorption at 280 nm, using the molar absorptivity (s28o)  as predicted by ProtParam ( tools/protparam.html; see Table  3.1). 103  Chapter 3 - Heteromeric complex of Ste4 and Byr2  3.5.2 Circular dichroism spectroscopy  Spectra were measured using a Jasco-810 spectrometer equipped with a PFD-452S Pelletier thermal control. Samples were measured in 20 mM potassium phosphate (pH 7.0) and 50 mM NaCl at 5 or 15 u.M using a 2 mm quartz cell. Wavelength scans were performed at 10 °C with a scanning speed of 100 nm- min" and data acquisition for 2 s 1  every 0.1 nm. Six scans were averaged and the buffer signal subtracted for the final spectra. Thermal denaturation studies at 222 nm used a heating rate of 1 °C- min" , 1  with data sampled for 4 s every 0.2 °C with a bandwidth 1 nm. The a-helical content was estimated using the formula (-6222 + 2340) / 30300 (Chen et al., 1972).  3.5.3 Native gel electrophoresis S t e 4  d-157)  10 uM solution of Byr2 " in a range of 3.3 to 80 (1  w  a  s  t  j  t  r  a  t  e  d  i  n  t  0  a  70)  LIM,  resulting in Ste4:Byr2 molar ratios of 0.33:1 to 8:1. After adding loading buffer to a final concentration of 125 mM Tris (pH 7.0) and 10% glycerol, samples were separated using native gels consisting of 125 mM Tris (pH 7.0), 15% (w/v) acrylamide (29:1), 0.05 % (v/v) ammonium persulphate and 0.005 % (v/v) TEMED. Separation occurred at a constant 30 mA in a Tris-Glycine buffer (25 mM Tris, 19 mM glycine, pH 7.5). Data points represent an average of three experiments, each performed in duplicate.  3.5.4 Trypsin digestion  Samples were digested with trypsin (Sigma) at 25 °C in 25 mM Tris (pH 7.9), 50 mM KCI, 0.1 mM EDTA and 1 mM DTT at a ratio of 1:250 (w/w) of trypsin to protein. Proteolysis was stopped at defined intervals with 5 mM PMSF coupled with heat 104  Chapter 3 - Heteromeric complex of Ste4 and Byr2  inactivation at 95 °C for five minutes. One-half of the sample was analyzed by 20% PhastGel SDS-PAGE (Amersham Biosciences) and standard Coomassie staining, whereas the second half was sent for analysis by ESI-MS.  3.5.5 NMR spectroscopy  All spectra were recorded on a 500 MHz Varian Unity or a 600 MHz Varian Inova NMR spectrometer equipped with a triple resonance probe and a pulsed field gradient accessory. Standard 1D and 2D pulse sequences were provided by Dr. Lewis Kay at the University of Toronto and collected data were analyzed using Felix 2000 (MSI) and Sparky. Except where indicated in the text, samples contained from 0.5 to 2.5 mM protein in 20 mM potassium phosphate (pH 7.0), 50 mM NaCl, 1 mM DTT with 10% D2O added for the lock.  105  Chapter 4 PNT domain as an ERK2 docking motif  It has recently become apparent that the specificity of MAP kinases for target proteins is set not only by a phosphoacceptor consensus sequence, but also by an ancillary docking event between the kinase and its substrate. A possible role in this MAPK docking is investigated for a subset of PNT domains, using the association of Ets-1 and ERK2 as a model. It is found that the ERK2 binding site on the PNT domain of Ets-1 is centred on an exposed phenylalanine side chain, with contributions from adjacent aromatic and acidic residues. This docking site appears distinct from those described previously for MAP kinases. Based on titrations monitored by NMR spectroscopy, Ets-1 and ERK2 bind with a K ~ 30 \iM. Interestingly, GABPa also d  binds ERK2 but through a different site on its PNT domain. This variance may reflect the alternative placement of the phosphoacceptor sites N- and C-terminal to the PNT domains of Ets-1 and GABPa, respectively. Knowledge of the regions important for kinase specificity will allow for a better understanding of the process by which an external signal is ultimately transmitted to the activity of a defined set of regulatory proteins. Future characterization of this kinase-substrate interaction, especially determination of the corresponding docking regions on ERK2, could lead to in vivo modulation of the signaling response with either investigative or therapeutic implications.  106  Chapter 4 - PNT domain as an ERK2 docking motif  4.1 Ets-1 PNT domain and MAPK docking Unlike the case of Ste4 and Byr2, valid protein partners to most Ets family PNT domains still remain largely unknown. This lack of insight into the function of these domains prompted several laboratories, including ours, to search for putative binding proteins. Using an affinity column constructed from Ets-1 " , our collaborator (1  138)  Barbara Graves (University of Utah) discovered several putative binding partners from an extract of calf thymus, one of which was identified as the MAP kinase ERK2 (Seidel and Graves, 2002). Further enzymatic characterization of this interaction revealed that the Ets-1 PNT domain serves as a 'docking' module, resulting in a decreased K for m  phosphorylation of Thr38 by ERK2 (Figure 4.1). The identification of an ancillary MAP kinase-docking region is not unique to Ets-1. Members of the bZIP, Ets and MAD family of transcription factors have been found to contain a consensus docking motif situated within 100 residues N-terminal to a MAP kinase phosphoacceptor site. This motif, composed of basic amino acids closely followed by a <bx<b sequence (where <>| is any hydrophobic amino acid, but usually leucine, valine or isoleucine), is known as a DEJL- or D-domain (Sharrocks et al., 2000). Recognized by the ERK, JNK and p38 family of MAP kinases, this docking region is proposed to bind the MAPK in a conserved acidic patch identified as the common docking (CD) domain (Tanoue et al., 2000). An additional motif recognized solely by ERK2 involves a short Phe-x-Phe-Pro tetrapeptide called the DEF domain (Jacobs et al., 1999). In the Ets transcription factor Elk-1, it is believed that flanking Nterminal  DEF- and C-terminal DEJL/D-domain differentially  target specific  serine/threonine residues for ERK2-mediated phosphorylation (Fantz et al., 2001). Apart from the above docking motifs, additional regions have been identified within 107  Chapter 4 - PNT domain as an ERK2 docking motif  Figure 4.1. Schematic of a MAPK docking domain, in the case of Ets-1, the well-structured PNT domain serves as a docking module to increase the effective concentration of the conformationally flexible phosphoacceptor threonine to the ERK2 catalytic site.  108  Chapter 4 - PNT domain as an ERK2 docking motif  cytoplasmic targets of phosphorylation, and a crystal structure of two such motifs bound to the MAPK p38 has been determined (Chang et al., 2002). The salient point in common with each of the above docking motifs, is the physical separation of MAPK association from the site of phosphorylation (Figure 4.1). For Ets-1, this model would allow for independent contact of Ets-1 to the docking and catalytic interfaces of ERK2. The possible importance of a second site of contact, apart from the actual phosphoacceptor site, becomes clear upon consideration of the minimal elements required for recognition by the MAPK enzyme. The MAPK consensus motif, namely a serine or threonine followed by a proline, can be expected to appear numerous times within a typical metazoan proteome; however, only a small number of such sites are phosphorylated during activation of a given MAPK. Therefore, additional binding regions distant from the phosphorylation site likely increase specificity via a decrease in the K or an increase in the k t of the m  ca  phosphorylation reaction. In particular, the PNT domain of Ets-1 aids in phosphorylation efficiency of Thr38 by lowering the K from 190 uM (i.e. for the m  phosphoacceptor peptide Ets-1 " ) to 6.8 pM (for Ets-1 " ), without significant (1  52)  (1  138)  changes in the k t value (Seidel and Graves, 2002). In addition, there is also a ca  necessity for the phosphoacceptor site to be accessible by the kinase; in Ets-1, the region surrounding Thr38 is disordered and flexible (Slupsky etal., 1998a). Jeff Seidel and Barbara Graves postulated that a potentially novel docking motif found in Ets-1, LxLxxxF, was responsible for binding ERK2. With site-directed mutagenesis they demonstrated an absolute requirement for Phe120 within this motif. However, mutation of the remaining leucine residues in the motif to arginine residues led to only a small increase in K , thus raising a question as to their role in docking. m  109  Chapter 4 - PNT domain as an ERK2 docking motif  Replacement of the entire Ets-1 PNT domain with the corresponding domain from GABPa resulted in an increased K towards phosphorylation of Thr38 (Seidel and m  Graves, 2002). Studies using native gel electrophoresis and surface plasmon resonance (BIAcore) also confirmed a specific interaction between Ets-1 PNT and ERK2 (B. Graves, personal communication). In order to gain further insight into the molecular determinants of association, the binding of Ets-1 and ERK2 was investigated using NMR spectroscopy.  4.2 NMR titration of Ets-1 with ERK2 To determine the specific site on Ets-1 responsible for binding ERK2, a titration was performed in which increasing amounts of unlabelled, active (phosphorylated) ERK2 was added to [ N]Ets-1 " . For this type of experiment, a fully assigned H- N 15  (29  138)  1  15  HSQC spectrum of a target protein is required in order to follow the site-specific effects of the added ligand. During the titration, binding of the unlabeled protein ligand elicits progressive changes in chemical shift or peak intensity for specific amides within the N-labeled target. Since amide chemical shifts are exquisitely sensitive to even small  15  changes in local chemical environment such as through altered hydrogen-bonding, residues that exhibit such perturbations either directly occupy the interfacial region between the associated proteins, or are indirectly sensitive to intramolecular changes in conformation that occur upon binding (Wittekind et al., 1989; Gbrlach et al., 1992). The titration was performed with 400 u.M [ N]Ets-1 " , with introduction of 15  (29  138)  phosphorylated, active ERK2 to a final level of 1.2 molar equivalents. Analysis of the resultant H- N HSQC spectra revealed only limited changes in most amide chemical 1  15  110  Chapter 4 - PNT domain as an ERK2 docking motif  9.0  8.8  8.6  8.4  8.2  HI  8.0 (ppm)  7.8  7.6  Figure 4.2. H- N HSQC monitored titration of [ N]Ets-1< - > with unlabeled acive ERK2. An expanded region of the HSQC spectra of [ N]Ets-1( - >(A) without, and (B) with 1.2 molar equivalents of ERK2. Residues F120, V121, L125, W126, H128 and L129 show specific decreases in peak intensity upon addition of ERK2, whereas L114 and L116 do not. D119, the only other amide which also demonstrates specific reduction in intensity, lies outside of the region shown. Spectrum B was acquired with an 8-fold increase in the number of transients and plotted with a 4.6-fold higher contour level than A to reflect the 2.4-fold dilution of [ N]Ets-1( - >. Titrations were performed at 30 °C in 25 mM Tris (pH 7.5), 150 mM KCI, 1.5 mM MgCI 250 uM AMP-PNP (to stabilize active ERK2) and 1 mM DTT. 1  15  29 138  15  29 138  15  15  2  Ill  29 138  Chapter 4 - PNT domain as an ERK2 clocking motif  shifts, indicating no significant global perturbations in structure. More importantly, after consideration of dilution effects, a specific decrease in intensity was observed for a small subset of peaks (Figure 4.2). These residues, which included Asp119, Phe120, Val121, Leu125, Trp126, His128 and Leu129 indeed clustered on the surface of the Ets-1 PNT domain and centred on Phe120 (Figure 4.4). This result not only demonstrated clear binding between Ets-1 " (29  138)  and ERK2 in vitro, but also localized  the site of interaction to a small hydrophobic region on the surface of the PNT domain. The central residue in this region, Phe120, is the same residue that led to an increase in K upon mutation to alanine (Seidel and Graves, 2002). However, there was no m  indication of Leu114 or Leu116 (i.e. within the putative LxLxxxF binding motif) being involved in contacting ERK2. Therefore, it appears that the docking site on the Ets-1 PNT domain may be a novel motif with no clear similarity to the known ERK2 docking motifs within the DEF domain (Phe-x-Phe-Pro motif) or DEJL/D-domain (basic residues followed by Leu/lle-x-Leu/lle) (Fantz etai., 2001; Jacobs etai., 1999). The selective perturbation in the intensity, rather than chemical shift, of binding site amides warrants further comment. Typically, with the fast exchange limit, amides affected by binding are tracked by a progressive change in chemical shift as a result of the population weighted average of the bound and unbound correlation peaks. The reason for this weighted average is that binding and release of the ligand takes place at a rate that exceeds the chemical shift difference between the bound and unbound state of each amide (e.g. /c > Aw). This scenario generally corresponds to K > 100 ex  d  u.M. At the other extreme, a slow exchange regime occurs with the formation of high affinity complexes (i.e. typically K < 10 u.M), in which the extended lifetimes of the free d  and bound states (/c < Aw) results in the observation of separate signals from each. ex  112  Chapter 4 - PNT domain as an ERK2 docking motif  For this latter situation, H- N HSQC spectra recorded as a function of added ligand 1  15  demonstrate a gradual disappearance of peaks corresponding to the unbound state, with a concomitant and balanced appearance of peaks corresponding to bound form of the protein. However, for the titration of Ets-1 with ERK2, the observation of decreased intensity, small chemical shift changes and a lack of newly appearing peaks reflects a level of binding that is likely in the intermediate exchange regime. In this case, the interconversion between the bound and free states occurs at a rate comparable to the chemical shift difference (/c  ex  « Aw). This selective conformational exchange  broadening has been observed for several kinase-related and other protein complexes (Kim er al., 2001b; Shekhtman et al., 2001; Amezcua er al., 2002; Hill et al., 2002). Closer inspection of Figure 4.2 indicates some broadening of all peaks, presumably due to time-averaged formation of a higher molecular weight complex with ERK2, but even at an equimolar ratio the linewidths are more like the 15 kDa free Ets1 " (29  138)  peptide than the 85 kDa mass expected upon the binding of dimeric ERK2. This discrepancy may be caused by a degree of local mobility within the complex, and will be further investigated using relaxation and chemical exchange measurements. Due to H- N HSQC peak overlap in the above titration, it was unfortunately 1  15  impossible to observe the effects of ERK2 binding upon the amide of the phosphoacceptor Thr38. Therefore, a second titration was performed in which the peak intensities were tracked using a two-dimensional version of the HNCO experiment with correlation peaks between the backbone amide proton of one residue with the carbonyl carbon of the preceding residue. After recording an initial spectrum of isolated [ C, N]Ets-1 " , unlabeled phosphorylated ERK2 was added as before 13  15  (29  138)  (Figure 4.3). Again, residue specific decreases in intensity (with very small changes in 113  Chapter 4 - PNT domain as an ERK2 docking motif  87s  eTl  s!~6  HI  8^2  (ppm)  eTo  71)  Figure 4.3. H- C HNCO monitored titration of [ C N]Ets-1< - > with unlabeled active ERK2. An expanded region of the HNCO spectrum of [ C, N]Ets-1 (A) without, and (B) with 1 molar equivalent of active ERK2. The spectra show correlations between the H of residue i and the C of residue M (i.e. within a peptide group). Residues T38, D119, F120, V121, L125, W126, and L129 (annotated on the figure by the H resonance) show specific decreases in peak intensity upon addition of ERK2, whereas L116 does not. H128 and L114 are not annotated due to spectral overlap. Spectrum B was acquired with an 3-fold increase in the number of transients and plotted with a 2-fold lower contour level than A to reflect the 2.5-fold dilution of [ C, N]Ets-1< - >. Titrations were performed at 30 °C in 25 mM Tris (pH 7.5), 150 mM KCI, 1.5 mM MgCI 250 uM AMP-PNP (to stabilize active ERK2) and 1 mM DTT. 1  13  15  13  13  15  29 138  (29_138)  N  N  13  15  29 138  2  114  Chapter 4 - PNT domain as an ERK2 docking motif  Figure 4.4. ERK2 docking site on Ets-1 " structure. Docking site residues identified from H- N HSQC- (A) and HNCO- (B) monitored titrations. Height ratios were normalized to Lys138. Black lines indicate residues that fall below 75% or 50% of the average peak ratio (grey dashed line). (C) Residues involved in binding ERK2 are highlighted in a backbone model of Ets-1 " . Residues that displayed a 75% or 50% decrease in peak intensity from A are coloured yellow and green, respectively. (D) Molecular surface representation of Ets-1 " coloured as in C. (E) Molecular surface coloured red and blue for calculated negative and positive charge, respectively. (F) Sequence of of Ets-1 (residues 110 to 130) with the position of the putative LxLxxxF motif and the NMR-derived ERK2 docking site labeled. Arrows indicate residues that show a significant specific decrease in intensity upon titration with ERK2. (29  1  138)  15  (29  138)  (29  115  138)  Chapter 4 - PNT domain as an ERK2 docking motif  D119  W126  L x L x x x F motif 11  °KECFLELAPDFVGDIL W E H L E  1 3 0  NMR-derived docking site  116  Chapter 4 - PNT domain as an ERK2 docking motif  chemical shift) were observed for Asp119, Phe120, Val121, Leu125, Trp126 and Leu129, with no apparent perturbation of Leu116. His128 and Leu114 are overlapped with other peaks in this spectrum. In addition, an unambiguous decrease in HNCO peak intensity was also observed for the phosphoacceptor Thr38. This site-specific decrease in intensity presumably indicated a direct contact to the catalytic site of ERK2. An estimate of the binding constant between Ets-1 " (29  138)  and ERK2 was  obtained from the residue-specific decrease in peak intensity for Phe120 and Thr38 as a function of added kinase. This estimate revealed a similar K of -30 L I M for both the d  PNT domain docking site (i.e. Phe120), as well as for the phosphoacceptor Thr38. This estimate of binding affinity is in keeping with the observed intermediate exchange regime, which predicts a K < 50 uM. In addition, the dissociation constant is d  comparable to the K of 6.8 m  Ets-1 " (1  138)  LIM  established using an in vitro ERK2 kinase assay with  (Seidel and Graves, 2002). A more accurate determination of the  dissociation constant (including /c / ff) is currently underway in the laboratory of Dr. on 0  Graves using surface plasmon resonance (i.e. BIAcore). Quantifying the specific change in affinity displayed by Ets-1 " (29  138)  harbouring site-specific mutations within the  docking site (e.g. Asp119, Phe120 and Trp126) are also in progress. In addition, it will be of interest to determine if inactive, unphosphorylated ERK2 will bind with the same affinity as the active form used in the NMR-based binding studies. Even though Ets-1 phosphorylation likely takes place exclusively within the nucleus (where active ERK2 predominates), it is possible that there may be some biological role for a phosphorylation-independent competition between ERK2-binding by Ets-1, and association with other Ets-1 protein partners. 117  Chapter 4 - PNT domain as an ERK2 clocking motif  Further insight into the association between Ets-1 " (29  138)  and ERK2 will rely on  additional experiments currently in progress in our laboratory and in collaboration. It is of primary interest to determine if there are changes in Ets-1 " (29  138)  dynamics upon  binding by ERK2. It is expected that the region around Thr38 will experience decreased flexibility when bound to ERK2, evident by a change in N relaxation 15  parameters of [ N]Ets-1 " 15  (29  138)  following addition of a molar equivalent of unlabeled,  active ERK2. Such a decrease in flexibility for this region would add evidence to a model in which binding of ERK2 by the PNT domain leads to a more stable interaction of the catalytic site with a well-placed phosphoacceptor site, with an increased chance of a productive phosphorylation event. In addition, calculation of global x for molecular c  tumbling will aid in understanding the discrepancy observed between the formation of the high molecular weight Ets-1 " - ERK2 complex and the observed minimal (29  138)  increase in line-broadening. Also of interest will be the identification of the Ets-1 binding site on ERK2. Due to the complexity in spectra expected for a protein the size of ERK2 (i.e. 358 amino acids), NMR-based methods are not amenable to this issue. Instead, crystallization of the Ets-1- ERK2 complex is currently in progress in the laboratory of Dr. Tom Alber (University of California, Berkeley). In addition, it may be possible to probe the Ets-1 binding surface on ERK2 through the use of site-directed mutations, chemical crosslinking or peptide competition experiments. In lieu of such data, it is still possible to create a model of the docked Ets-1 " (29  138)  on the surface of ERK2 based on the crystal  structure of the p38 MAP kinase in complex with peptides from the protein substrate MEF2A, and an upstream activating kinase, MKK3b (Figure 4.5; Chang et al., 2002). From this picture it is evident that there is ample peptide length separating the distal 118  Chapter 4 - PNT domain as an ERK2 docking motif  ERK2  Figure 4.5. Cartoon model of Ets-1 PNT domain docking to ERK2. A schematic model of the Ets-1 " construct positioned on ERK2 such that the PNT domain is adjacent to a region identified as the CD docking site on the p38 MAP kinase, and the phosphoacceptor Thr38 is positioned within the catalytic site. Although providing no molecular clues as to the nature of this association, the model illustrates that the N-terminus of Ets-1 has sufficient length to reach the active site as required for a biologically relevant docking interaction. (29  138)  119  Chapter 4 - PNT domain as an ERK2 docking motif  docked PNT domain from the catalytic site occupied by the phosphoacceptor Thr38. However, it is important to remember that the PNT domain is well-folded, and more likely interacts with ERK2 via a different site and in a different manner than the previously studied docking motif peptides.  4.3 NMR titration of GABPa-PNT with ERK2 As a control for the preceding study of the MAP kinase-docking site on Ets-1, it was decided that a titration with the GABPa PNT domain (for which an NMR-derived structure is known) would provide an interesting contrast. This domain failed to reduce the K value for phosphorylation of Thr38 when fused as a chimera to the N-terminal m  region of Ets-1 (Seidel and Graves, 2002). Furthermore, GABPa clearly lacks the essential phenylalanine residue critical for ERK2 binding by Ets-1 (see alignment in Figure 1.5). In contrast to these expectations, a construct containing only the PNT domain of GABPa (residues 168-254) also bound to ERK2 in the intermediate exchange regime (i.e. K ~ 50 p.M). Residue-specific decreases in peak intensity were d  observed for Arg174, Asp185, Arg248, Tyr250, Leu252 and Ala253 upon binding to ERK2 (Figure 4.6). These residues mapped to region on the GABPa PNT domain Cterminal to the ERK2 docking site determined for the Ets-1 PNT domain (Figure 4.6C). Indeed, there were no residue-specific decreases in signal intensity for any or the residues from 236-244 that correspond to the docking site residues in Ets-1. These results suggest that GABPa may possess an ERK2 docking motif equally potent to the one described for Ets-1, yet composed of a distinct motif that includes Arg248, Tyr250, Leu252 and Ala253. The presence of this MAP kinase120  Chapter 4 - PNT domain as an ERK2 docking motif  Figure 4.6. H - N HSQC monitored titration of [ N]GABPa< - > with unlabelled active ERK2. An expanded region of the HSQC spectra 1  15  15  168  254  of [ N]GABPa( - > (A) without, and (B) with 1 molar equivalent of active ERK2. Residues D185, R248, Y250, L252 and A253 show specific decreases in peak intensity upon addition of ERK2 (indicated by arrows in C). Residues within the LxLxxxF motif (e.g. F231) and the Ets-1 MAPK kinase docking site (e.g. W241) do not exhibit any significant change upon binding ERK2. Spectrum B was acquired with 4-fold more transients, and is plotted at a 1.7-fold higher contour level than A to reflect the 1.7-fold dilution of [ N]GABPa( " ). (C) Ratio of the H- N amide peak intensities for GABPa " with one and zero molar equivalents of active ERK2, normalized to produce an average ratio of one. Also indicated for comparison are the residues in Ets-1 that display a specific decrease in peak intensity upon titration with ERK2. The involvement of D185 likely arises from its close proximity in the tertiary structure of GABPa " (see inset of Figure 4.7B). Titrations were performed at 30 °C in 25 mM Tris (pH 7.5), 150 mM KCI, 1.5 mM MgCI 250 uM AMP-PNP (to stabilize active ERK2) and 1 mM DTT. 15  168 254  15  168  (168  (168  254  1  15  254)  254)  2  121  Chapter 4 - PNT domain as an ERK2 docking motif  0 [ N]GABPa( 15  168  -  254  8 9  0  254  I  w  [ N]GABPa< 15  168  -  254  o 1'  @  E  Y250  fl  '  n  OQ  0  £  a  ft in  U  e  c  B  ^  0 00 0  gQ  10 0 . 0  A  0  >  168  0 R248  6  F2310  6  0 L252  0 1  0 A253  )  + ERK2  168 /  ERK2  254  .6  c  8.2 7.8 HI (ppm)  7.4  1.4 -  1.2  o  1  0.8  cn CD  0.6  I  0.4  0.2  0  Residue  QEDFFQRVPR- -GEILWSHLELLRKYVLAS  227  (DFVGDILWEHL)  MAPK perturbed region in Ets-1  122  254  Chapter 4 - PNT domain as an ERK2 docking motif  docking region in GABPa does not, however, lack a putative role: literature reports have identified a bona fide ERK2 phosphorylation site on GABPa that is responsive to neuregulin-1 stimulation of muscle cells and comprises part of the Raf-responsive element within the HIV promoter (Flory et al., 1996; Fromm and Burden, 2001). The phosphoacceptor residue Thr280 is C-terminal to the GABPa PNT domain, and it is plausible that a docking motif would be required for this phosphorylation site. In addition, the residues that display the largest reduction in intensity upon titration with ERK2 conform roughly to the DEJL/D-domain in which a basic region is followed by a d)x<> j sequence (where § is any hydrophobic amino acid, but usually leucine, valine or isoleucine). Significantly, the DEJL/D-domain is invariably found N-terminal to its targeted phosphoacceptor site, similar to the positioning of the PNT domain relative to Thr280 in GABPa. The surprising identification of a MAPK docking site on GABPa prompted the creation of a second construct to include both the PNT domain and phosphoacceptor Thr280. This construct, GABPa " , contained an additional 36 residues C-terminal (168  290)  to the PNT domain and was stable and soluble upon expression and purification. Backbone amide assignment of [ C, N]GABPa " 13  15  (168  290)  (by M. Scharpf in our  laboratory) using H- N HSQC, HNCACB, HBCBCACONH and (H)CC(CO)NH1  15  TOCSY spectra allowed for comparison of the H and N chemical shifts between 1  GABPa (168  290)  and the shorter GABPa '  (168 254)  N  15  (Figure 4.7A). As expected, the largest  difference in amide resonances were observed for residues 246-254, which directly precede the additional residues in GABPa " (168  290)  (Figure 4.7B). Chemical shift  perturbation was also noted for amides within helix H1 and in the following loop region. 123  Chapter 4 - PNT domain as an ERK2 docking motif  Figure 4.7. Additional C-terminal residues in G A B P a perturb the structure of the PNT domain. (A) Comparison of  ( 1 6 8 2 9 0 )  do not  the H- N HSQC spectra from [ N]GABPa< - > (red) and [ N]GABPa( - > (black) reveal superimposition of peaks for most PNT domain residues. This confirms a similar three-dimensional PNT domain structure with and without addition of a C-terminal extension in GABPa " . Residues 255-290, unique to GABPa " , have been annotated in black. Annotations in grey indicate the few residues that experience a chemical shift difference between the GABPa constructs. Spectra were acquired at 30 °C in 20 mM potassium phosphate (pH 7.2), 20 mM potassium chloride and 5 mM DTT. (B) Residue specific changes in chemical shift between GABPa ' and GABPa " for amino acids 168-254. Residues showing a combined A5 > 0.5 ppm (where A5 = [(8 *10) +(5 ) ] ) are highlighted in red on the backbone representation of GABPa ~ shown in the inset, and are limited to the final residues of GABPa " (L246 to S254), and those residues in helix H1 and adjacent loop (E171, R174, E178, Y184, D185 and W189) that contact the C-terminus of helix H5. 15  168 254  15  (168  (168  (168-254  (168  1  15  168 290  290)  290)  290)  2  2 05  HN  (168  (168  124  N 254)  254)  Chapter 4 - PNT domain as an ERK2 docking motif  10.0  9.0  8.5  8.0  1  H2  7.0  6.5  6.0  (ppm)  Hi  CHD  7.5  I  Hl3lH~H4~l  125  1 H5  |—  Chapter 4 - PNT domain as an ERK2 docking motif  This second region directly contacts resides 246-254 in the tertiary structure of GABPa " (168  254)  (Figure 4.7B inset). The remaining PNT domain amide chemical shifts  were unchanged in the two GABPa constructs, indicating that the C-terminal extension did not affect the overall structure of the PNT domain. Analysis of N relaxation further 15  confirms that residues 255 to 290 are unstructured and display significant conformational flexibility (Figure 4.8). Upon addition of active ERK2, GABPa " (168  290)  was efficiently phosphorylated at a single site. This modification produced a significant retardation in SDS-PAGE mobility and an increase in mass by 80 Da as measured by mass spectrometry (Figure 4.9A). Similar to Ets-1 " , addition of a phosphate group (29  to GABPa " (168  290)  138)  failed to elicit any change in the chemical shifts (and by inference,  the structure) of the PNT domain (Figure 4.9B). Changes in chemical shift were instead localized around Thr280, reflecting the local changes in chemical environment from the introduction of a phosphate group on this residue. This in vitro site of modification (i.e. Thr280) correlates with the in vivo identified phosphoacceptor site (Flory et al., 1996; Fromm and Burden, 2001).  15  N relaxation studies also  demonstrated that, upon phosphorylation, there was no change in the dynamic property of the flexible region in which Thr280 is located (not shown). It is possible, however, that phosphorylation of a full-length clone of GABPa may be coupled with structural effects. The construct chosen for the above study aimed at a compromise between residues necessary for in vitro phosphorylation (i.e. ten residues C-terminal to the phosphoacceptor threonine), but minimizing spectral overlap by limiting the size of the C-terminal unstructured region. Secondary structure prediction of GABPa using the PROF algorithm (accessible through Predict Protein;  126  Chapter 4 - PNT domain as an ERK2 docking motif  4.8. Additional C-terminal residues in GABPa ' are flexible. N relaxation analysis (T1, T2 and H-{ N} NOE) of [ N]GABPa< - > using curvefit and 7M_F77 programs, as detailed in section 4.5.4. Note that while residues corresponding to the PNT domain display restricted mobility (model-free S > 0.8), the additional C-terminal residues demonstrate conformational disorder in solution (S < 0.4). Similar to N relaxation analysis of GABPa " (Scharpf et al., in preparation), residues in the loop between helices H2 and H3 display increased conformational flexibility as compared to the remainder of the PNT domain. (168 290)  Figure  15  1  15  15  168 290  2  2  (168  254)  127  15  Chapter 4 - PNT domain as an ERK2 docking motif  Chapter  4 - PNT domain as an ERK2 docking  motif  MW  monophosphorylated [ N]GABPa< 68-290) 15  MW  o b s  1  = 14857  [ N]GABPa( MW = 14776 15  168  290  >  o b s  ERK2  Oh  16 h  B  O O M  Tj- 00 CN CO O 5* O O T - CM CM CM CM CN CN CM N C N C \ I C N C N C N C N C \ I C N C N C \ I C \ I C \ I C N C N C \ I C N  Residue  Figure 4.9. Phosphorylation of GABPa at Thr280 does not affect PNT domain structure. (A) Addition of a single phosphate group results from the incubation of [ N]GABPa( " ) with active ERK2 and ATP, as indicated by gel retardation in a Coomassie stained 20% SDSPAGE gel. ESI-MS confirmed the identity of the bands, using the calculated MW of 14777 and 14857 for native and monophosphorylated [ N]GABPa (ProtParam; (B) Amide chemical shift changes between phosphorylated and unphosphorylated [ N]GABPa< - > (A5 = [(A8 -10) +(A5 ) ] ). Note that shift perturbations are clustered near Thr280, confirming this residue as the site of phosphorylation, and indicating only local changes in the chemical environment of surrounding residues. (168  15  15  168  290)  290  (168_290)  15  168 290  2  HN  129  2  N  05  Chapter 4 - PNT domain as an ERK2 docking motif predicts a random coil conformation for residues from 250 to 320 (i.e. between the PNT domain and ETS domain boundaries). Therefore, it may be necessary to test a longer GABPa construct (residues 168-320 instead of 168-290) to fully conclude that there are no phosphorylation-induced changes in structure or dynamics. Nevertheless, a model emerges in which ERK2 phosphorylation of GABPa occurs within a flexible region and does not affect the protein architecture of the adjacent PNT domain (which serves as the docking site). This model is in keeping with phosphorylation of Ets-1 at Thr38, which lies in the flexible N-terminal region of the protein, preceding its structured PNT domain.  4.3 Role of PNT domain in ERK2 docking An inspection of several Ets transcription factor sequences illustrates the conspicuous presence of MAP kinase phosphorylation sites adjacent to PNT domains (Figure 4.10). It is now clear that this placement partly reflects the emerging function of several PNT domains as docking modules, used to increase the efficiency and likely specificity of MAP kinases to these phosphorylation sites. One interesting outcome of the research is the identification of different docking motifs present on the Ets-1 and GABPa PNT domains. It is likely that these regions are positioned for optimal phosphorylation of their respective N-terminal or C-terminal phosphoacceptor sites. Such constraints could include the distance between the docking site and the phosphoacceptor threonine, or perhaps orientation of the. docked PNT domain and thus relative orientation of flanking substrate residues towards the MAPK catalytic site. It will be of interest to investigate further the ability of the Yan PNT domain to serve as a MAP kinase-docking module. The presence of multiple Rolled (the ERK2 130  Chapter 4 - PNT domain as an ERK2 docking motif  enzyme in Drosophila) phosphorylation target sites on Yan prompts one to hypothesize the involvement of its PNT domain in recruiting activated kinase. Since these phosphoacceptor sites lie C-terminal to the Yan PNT domain (Figure 4.10), the involvement of a DEJL/D-domain-type motif (such as discovered for GABPa) would be expected. However, such a sequence is not present within this region of Yan. In contrast, it appears that for Yan the role of MAP kinase-docking resides in a separate PNT domain-containing protein. This protein, Mae, is thought to associate with the PNT domain of Yan, and aid in the subsequent phosphorylation of Yan by Rolled without itself becoming modified (Baker et al., 2001). This alternative mechanism may reflect the fact that Yan is a transcriptional repressor, which upon phosphorylation is no longer bound to DNA and is expelled from the nucleus. In the absence of Mae, Rolled is capable of phosphorylating Yan in vitro, on six of the seven target sites. The seventh site, Ser127, is the closest to the PNT domain and the functionally most critical phosphoacceptor. It is only with addition of Mae that Rolled is able to phosphorylate this key residue. Therefore, it is possible that a Mae-dependent conformational change in the PNT domain is required for Rolled to gain access to Ser127. Experiments designed to look into the interaction of Mae with Yan have been initiated in our lab, but have met with limited success due to significant problems with the solubility of these two proteins. Therefore the mechanism by which Mae aids in Yan phosphorylation remains unknown. Also poorly understood is the functional outcome of phosphorylation for proteins such as Ets-1 and GABPa. Although it is clear that activated ERK2 leads to an increase in transactivation by these proteins, the mechanisms by which this change 131  Chapter 4 - PNT domain as an ERK2 docking motif  55  Ets1  136  N  ETS  87  I hC  170  PNT  ETS  PLLT P 72  166  PntP2  250  ETS  NH  PPLT P 151  170  GABPa  251  N  ETS  KC  PPAT P 280  39  125  Tel N-l I ld3hM ~  ETS  he  PPES P 22  31  Yan  H\  117  PNT HIIIII PNS PVTPSRYPLS PHSHPPT P.. 127  ~\00  137  2O0  300^  144  40o"  &00  600  700  800  Residue  Figure 4.10. MAP kinase phosphorylation sites adjacent to Ets family PNT domains. Primary structures of Ets family transcription factors shown, with verified MAPK phosphorylation sites indicated as vertical grey bars. Sequence information is included below each protein schematic, with the phosphoacceptor sites in bold. Only the first three of seven MAPK phosphorylation sites are detailed for Yan.  132  Chapter 4 - PNT domain as an ERK2 docking motif  occurs are mainly unknown. Research described in this thesis highlights the fact that GABPa and Ets-1 have phosphoacceptor site in flexible regions that do not change in structure, dynamics or oligomerization state upon modification by ERK2. Instead, there must be a phosphorylation-dependent change in the ability to associate with as yet unknown components of the transcriptional machinery. A screen for proteins that interact with Ets-2 has recently provided a clue to the identity of one of these factors (Baker et al., 2003). It was found that Brg-1 of the chromatin remodeling SWI/SNF complex was able to interact with unmodified Ets-2, leading to a reduction in the level of transcription for the Ets-2 target gene BRCA1. Interestingly, ras-dependent phosphorylation of Ets-2 at Thr72 (i.e. by ERK2) abrogated the association with Brg-1, causing a relief from repression, and producing an observed increased in transactivation. It is likely that a similar mechanism also exists for Ets-1 and GABPa. The molecular details of the interaction of unmodified Ets-2 remains to be established.  4.5 Materials and methods 4.5.1 Cloning and purification  Ets-1 " (29  138)  and GABPa " (168  254)  were cloned and purified as previously described  (Slupsky et al., 1998a; Sections 2.7.1 and 2.7.2). GABPa " (168  from a full-length  290)  was PCR-amplified  murine GABPa cDNA using the primers GABPs, 5'-  CCCGCGGGATCCATATGGCTGCACTGGAAGGCTAC-3',  and GABP-290R, 5'-  GGAACCGCTCGAGrC>AACTGCTGTTTATAAC-3' (with restriction sites underlined, introduced stop codon in italics and coding bases in bold) then cloned into pET28a vector using A/del and Xhol restriction enzyme sites to retain the N-terminal His tag. 6  133  Chapter 4 - PNT domain as an ERK2 docking motif  Expression in BL21(^DE3) cells using LB or minimal M9T media supplemented with 1 g/L NH CI or 3 g/L [ C]-glucose, was followed by the standard Ni -column 15  13  2+  4  purification as described for previously for Erg and GABPa. Phosphorylated ERK2 was prepared using the plasmid and procedure from B. Graves as initially provided by the laboratory of Dr. Melanie Cobb (Wilsbacher and Cobb, 2001) and stored at +4 °C or20 °C in 25 mM Tris (pH 7.5), 200 mM KCI, 10% (v/v) glycerol, 0.1 mM EDTA and 1 mM DTT until ready for dialysis into appropriate buffer.  4.5.2 NMR spectroscopy  All spectra were recorded at 30 °C on a 500 MHz Varian Unity or a 600 MHz Varian Inova NMR spectrometer equipped with a triple resonance probe and a pulsed field gradient accessory, and were analyzed using Felix 2000 (MSI) and Sparky. Assignment of GABPa " (168  254)  chemical shifts was performed by L. Gentile as  described by Mackereth et al. (2002). Backbone H and N amide resonances for 1  GABPa " (168  290)  1 5  H  were assigned by M. Scharpf in our laboratory, using standard H- N 1  15  HSQC, HNCO, HNCACB and HBCBCACONH spectra (Scharpf et al., unpublished). Titration experiments were performed in 25 mM Tris (pH 7.5), 150 mM KCI, 1.5 mM MgCI , 1 mM DTT, 0.1 mM EDTA with 10% D 0 added for the lock. The non2  2  hydrolyzable ATP analogue, AMP-PNP, was added at a final concentration 250 \xM to help stabilize ERK2 during protein concentration and subsequent titration steps. Unlabeled ERK2 was added at 0.25, 0.50, 0.75, 1.0 and 1.2 molar equivalents to a 400 nM sample of [ N]Ets-1 15  (29_138)  or 200 ^M [ N]GABPa - , or at 0.05, 0.10, 15  (168  254)  0.15, 0.20, 0.25, 0.30, 0.40, 0.50, 0.75 and 1.0 molar equivalent to a sample of 200 [iM 134  Chapter 4 - PNT domain as an ERK2 docking motif  [ C, N]Ets-1 " . At each titration point a sensitivity-enhanced H- N HSQC 13  15  (29  138)  1  15  spectrum was acquired, with an additional 2D C- H plane of an HNCO spectrum 13  collected for [ C, N]Ets-1 13  15  (29  138)  1  .  4.5.3 Binding site determination  Residues that contribute to binding were identified by using the peak intensity ratio of the initial and final titration points. In order to compensate for the effects of dilution, peak heights were normalized to the intensity of the flexible C-terminal Lys138 in Ets1 (29-138)  t h a t  d o e s  n o t  a  residue in GABPa " (168  pp 254)  e a r  t contribute to ERK2 binding. Since the C-terminal 0  may be involved in ERK2 binding, peak heights were  normalized to the average peak height ratio.  4.5.4 Relaxation analysis  Calculation of relaxation parameters utilized a combination of N T1, N T2 and H15  15  1  { N} NOE experiments (Farrow et al., 1994). T1 values were calculated using delay 15  times of 10.0, 40.1, 70.2, 130, 200, 301, 452, 602, 803 and 1003 ms. T2 values were calculated with delay times of 16.5, 33.1, 49.6, 66.2, 82.7, 99.3, 115.8, 132.4 and 148.9 ms. Two repeat points were recorded within each experimental series for the purpose of error analysis. Data were analyzed using a Levenberg-Marquardt nonlinear least squares fitting algorithm of the curvefit program (Dr. Art Palmer, Columbia University). For measurement of H-{ N} NOE relaxation, a control delay of 5.017 s 1  15  was used for the no-NOE spectrum, with a relaxation delay of 2 s preceding a 3.017 s proton presaturation period for the NOE spectrum. The order parameter S based on 2  the optimal overall x (6.9 ns) calculated using the TM_F77 and J_F77 programs from c  135  Chapter 4 - PNT domain as an ERK2 docking motif  Neils Farrow (formerly in the laboratory of Dr. Lewis Kay, University of Toronto) based on the Lipari and Szabo model-free formalism (Lipari and Szabo, 1982).  4.5.5 In vitro phosphorylation GABPa " (168  290)  at a concentration of 30  was phosphorylated in 25 mM Tris (pH 7.5),  150 mM KCI, 1.5 mM MgCI , 0.1 mM EDTA, 1 mM DTT, 0.5 mM ATP with 150 nM 2  activated ERK2. Following incubation for 16 h at 37 °C, the reaction was checked for completion using 20% SDS-PAGE analysis for phosphate-mediated gel retardation (phosphorylated GABPa " (168  290)  migrates at an apparent 17 kDa, versus 14 kDa for the  native form). Monophosphorylated [ N]GABPa " 15  (168  254)  was repurified by FPLC, by  using an anion exchange Q-sepharose column and a buffer of 50 mM Tris (pH 8.5). The protein was eluted at -125 mM NaCl using a gradient of zero to 300 mM NaCl. Following purification, the sample was dialyzed into buffer containing 25 mM Tris (pH 7.5), 150 mM KCI, 1 mM DTT and 0.1 mM EDTA.  136  Chapter 5 Interactions with UBC9 and SUMO-1 Recent reports have suggested the involvement of sumoylation in the biology of Ets family transcription factors. In particular, Tel appears to be localized into punctate nuclear bodies via a SUMO-dependent mechanism. Initial cross-linking experiments indicate that the SUMO-conjugating enzyme UBC9 may directly contact several PNT domain-containing constructs including those from Tel and Ets-1. However, the reported interaction between UBC9 and Tel PNT domain was not observed using NMR spectroscopy. In contrast, a consensus sumoylation site N-terminal to the PNT domain in Ets-1 was found to associate with a substrate-binding site on UBC9 that is coincident with that of other bona fide protein targets such as p53, c-Jun and RanGAPL  This interaction has a measured dissociation constant of -240 \xM, which  reflects tighter binding than that reported for the association of UBC9 with p53 or c-Jun peptides, but weaker than RanGAPL  In addition, this newly identified target site in  Ets-1 is efficiently sumoylated in an in vitro assay, providing further support of Ets-1 involvement in the SUMO pathway. Although the functional significance of this posttranslational modification remains to be established, it is intriguing that the sites of sumoylation and phosphorylation in Ets-1 are immediately adjacent within the flexible N-terminal region of this transcription factor.  137  Chapter 5 - Interactions with UBC9 and SUMO-1 5.1 Evidence for UBC9-mediated sumoylation of PNT domains  The covalent attachment of SUMO to target proteins is dependent upon the activity of a series of enzymes similar to those of the ubiquitin pathway. At the heart of sumoylation lies the E2 SUMO-conjugating enzyme UBC9. This protein directly contacts appropriate substrates (possibly aided by an adapter E3 SUMO-protein ligase), and results in the transfer of the thioester-linked SUMO molecule to a target eamino group. This acceptor lysine residue typically resides within the consensus sumoylation motif ^ K x E (where W is a large hydrophobic residue that is usually isoleucine, valine or leucine, and x is any amino acid). Previous studies have demonstrated that UBC9 directly interacts with several ETS family transcription factors. For many of these interactions, the key association region appeared to include, or fall adjacent to, the PNT domain (Hahn et al., 1997; Chakrabarti er al., 1999; Chakrabarti et al., 2000). Tel is the best-characterized sumoylation substrate within the Ets members, and such modification has been shown to regulate nuclear localization to punctate nuclear bodies similar to the sub-nuclear assemblies known as PODs, PML bodies or nuclear dots (ND10) (Chakrabarti et al., 2000). These regions, which have been implicated in areas of large-scale transcriptional repression, are cell-cycle or stimulus-dependent in composition, and are enriched in SUMO-1 (Everett et al., 1999; Li and Chen, 2000; Maul et al., 2000). Association of the remaining Ets family members with UBC9 suggests a similar role for sumoylation in regulating their level of transactivation or repression (Hahn et al., 1997). To further verify these interactions in vitro, several PNT domains were purified and tested for binding to human UBC9 by using glutaraldehyde cross-linking (Figure 138  Chapter 5 - Interactions with UBC9 and SUMO-1  Figure 5.1. Glutaraldehyde crosslinking of UBC9 to SUMO and ETS PNT domains. (A) Identification of complex formation with samples of  UBC9 (lanes 1,2), SUMO-GG (lanes 3,4) and a combination of UBC9 and SUMO-GG (lanes 5,6) with and without 0.001% (v/v) glutaraldehyde with a 30 min incubation at 25 °C. UBC9 and SUMO-GG form a heteromeric dimer (lane 6), and UBC9 is shown to form a homodimer (lanes 2 and 6) under these conditions. (B,C) Crosslinking assay with PNT domains from TelV80E, Ets-1, Ets-2, Erg and GABPa without and with UBC9 (B) or SUMO-GG (C). Glutaraldehyde (0.001% v/v) is present for all samples. Note that a heterodimer is formed with UBC9 for all five tested constructs (lanes 2,4,6,8,10 in B) as indicated by an asterisk. The UBC9 homodimer is indicated by a black circle. Ets1(1-138) oxidation-induced dimers are indicated with a white circle. No heterodimers are observed between the PNT domain samples and SUMO-GG.  139  Chapter 5 - Interactions with UBC9 and SUMO-1 UBC9 dimer UBC9/SUMO-GG heterodimer  UBC9 SUMO-GG glutaraldehyde  B  45-  + +  + +  1  2  3  4  5  6  7  8  + + +  9  10  UBC9 PNT  domains  TelvsoE  Ets-1 (1-138)  UBC9  -  +  Ets-2  Erg  GABPa  (85-172)  (108-201)  (168-254)  . + - + -  +  -  +  -SUMO-GG PNT  domains Tel V 8 0 E  Ets-1  Ets-2  Erg  GABPa  (1-138)  (85-172)  (108-201)  (168-254)  SUMO-GG  140  Chapter 5 - Interactions with UBC9 and SUMO-1  5.1). As a positive control, purified human SUMO-1 was first shown to bind UBC9 as previously established in other in vitro assays (Figure 5.1 A; Liu et al., 1999a). Also observed was the expected UBC9 homodimer resulting from a dimer-monomer dissociation constant of 1.2 mM (Liu et al., 1999b). Additional cross-linked heterodimeric species involved UBC9 and purified Tel, Ets-1, Ets-2, Erg and GABPa PNT domains (Figure 5.1B). These latter results suggest that Ets proteins may indeed form a complex with the conjugating enzyme, consistent with the reported yeast-twohybrid screen (Hahn et al., 1997). Interestingly, there was no indication of complex formation between the above five PNT domains and SUMO-1 (Figure 5.1 C). As a step towards understanding this overall association between Ets proteins, UBC9 and SUMO-1 at the molecular level, we used NMR spectroscopy to monitor the titrations of various combinations of N-labeled and unlabeled proteins. 15  5.2 Titration of Tel with UBC9 In a process similar to the identification of the ERK2 binding site on Ets-1, N15  isotopically labeled Tel " (38  127)  was studied for possible association with UBC9. As  mentioned in Chapter 2, two single mutants of Tel are available that render its PNT domain monomeric due to the introduction of an acidic residue within one of two hydrophobic oligomerization interfaces. Therefore, both [ N]Tel " A94D and 15  (38  127)  [ N]Tel " V113E were used in a titration assay with UBC9. Unexpectedly, there 15  (38  127)  were no significant changes in the amide chemical shift or peak intensity of these PNT domains as a result of adding two molar equivalents of UBC9, up to a final concentration of 420 |iM (not shown). Thus neither Tel PNT domain mutant binds UBC9 with any appreciable affinity (i.e. conservatively, Kd > 1 mM). Given the previous 141  Chapter 5 - Interactions with UBC9 and SUMO-1  cell-based assays suggesting Lys99 within the PNT domain of Tel is sumoylated in vivo, this observation was surprising. One possible explanation is the presence of the  oligomerizing mutations. It is conceivable that one of the introduced single mutations may adversely affect the association between Tel and UBC9. However, the individual mutations are on opposite sides of the Tel PNT domain, and both are at some distance from the putative sumoylation target Lys99 (from the structure by Kim et al., 2001a). Therefore, it is hard to imagine both single mutations functioning to abrogate an interaction with UBC9. Another possible explanation is the size of the construct tested. This thesis has so far focused on the structure and function of the PNT domain, and therefore the Tel protein used in the above titration contained only residues 38 to 127. It may turn out that additional flanking residues are required to improve the affinity of Tel for UBC9. This possibility is currently being tested. In addition, it is always possible that the Tel PNT domain binds UBC9 or sumoylated-UBC9 in a transient manner that is sufficient for SUMO-1 transfer, or that an E3 is required for efficient UBC9 binding and subsequent sumoylation. This latter case is exemplified by the enhanced modification of p53 and c-Jun in the presence of PIAS proteins that serve as putative E3-protein ligases (Hochstrasser, 2001; Kotaja et al., 2002; Schmidt and Muller, 2002). The reported K for the direct UBC9 binding to target sumoylation d  peptides derived from p53 or c-Jun is only 3-6 mM (Lin et al., 2002), and it is possible that the interaction between Tel and UBC9 may also be of low affinity. Upon addition of PIAS proteins, in vitro assays have indicated an increased sumoylation of p53 and c-Jun (Kotaja et al., 2002; Schmidt and Muller, 2002). Therefore, a higher affinity complex between UBC9 and Tel may be dependent upon the addition of the appropriate PIAS E3 protein ligase. Other putative E3 proteins could also serve this 142  Chapter 5 - Interactions with UBC9 and SUMO-1  role. These additional examples include RanBP2, which may help to couple sumoylation of protein targets with their nuclear import (Kirsh et al., 2002; Pichler et al., 2002), or a homolog of the Drosophila polycomb group E3 protein Pc2 (Kagey et al., 2003). A final explanation becomes apparent upon re-examination of the Tel amino acid sequence in a search for additional regions that contain the optimal *FKxE sumoylation motif (Rodriguez et al., 2001). Indeed, a sequence N-terminal to the PNT domain in Tel ( I K - M Q E ) conforms to a typical sumoylation consensus site (Figure 5.2), while the original Lys99 site (TK ED) within the PNT domain does not. A re99  examination of the COS7 cell in vivo data of Tel sumoylation presented by Chakrabarti et al. (1999) reveals a significant amount of residual sumoylation (about 50% of wildtype levels) for a Tel construct in which Lys99 has been mutated to an arginine. This Tel N-terminal peptide, however, does include the above-mentioned consensus IKuQE sumoylation motif, and it is possible that Lys11 is in fact the true site of sumoylation. We are currently pursuing this alternative, by using NMR-monitored titration to investigate the in vitro binding of UBC9 to this region in Tel. In addition, mutation of Lys11 in Tel constructs for in vivo and in vitro assays will also ascertain the importance of this site in sumoylation.  5.3 Titration of Ets-1 with UBC9 and SUMO At this point, sequence analyses revealed a similar sumoylation consensus site at the N-terminus of Ets-1  (IK15TE;  Figure 5.2). Since a PNT-domain-containing fragment of  Ets-1 was found to interact with UBC9 via yeast two-hybrid and in vitro association assays, and given the previous work summarized in this thesis on various Ets-1 143  Chapter 5 - Interactions with UBC9 and SUMO-1  IK^QE Tel  TK ED 99  N  ETS  IK QE  IK TE  227  15  Ets-1  m  ETS  ERK2 Ets-2  hC  CKII  N  ETS  PNT  ERK2  IK EE  LK  118  Pnt-P2  N  428  EE  LK711IE  \  PNT  ETS  Rolled  100  Figure 5.2.  SUMO  200  300  400  consensus  500  700  600  motifs in the  800  Ets family of  transcription factors. Tetrapeptide regions corresponding to the [IA//L]-K-x-E sumoylation motif are indicated for the murine Tel, Ets-1, Ets-2 and Drosophila Pnt-P2 proteins. Also shown is the motif corresponding to the K99 putative sumoylation site in Tel, as well as sites of phosphorylation by ERK2/Rolled and Ca -dependent calmodulin kinase II (CKII). The sequences are drawn to scale with the bottom bar indicating residue number. Of significance is the presence of sumoylation sites N-terminal to several regions of phosphorylation, and the complete lack of any putative sumoylation motifs in Ets-2 2+  144  Chapter 5 - Interactions with UBC9 and SUMO-1  constructs, it was decided to switch to Ets-1 as a model for sumoylation studies. This approach first necessitated chemical shift assignment of the backbone amide H and 1  15  N resonances of Ets-1 " . This longer construct comprises all amino acids of the (1  138)  Lys15 sumoylation motif as well as the MAPK phosphorylation site and the entire PNT domain. The additional residues not contained within Ets-1 " (29  138)  (i.e. the N-terminal  residue 1 to 28, assigned by Manuela Scharpf and Matthew Macauley in the laboratory) display random coil chemical shifts and N relaxation behaviour consistent 15  with a high degree of conformational flexibility. Following assignment, 400 ^M [ N]Ets-1 " 15  (1  138)  was titrated with unlabeled UBC9  and monitored for binding-induced changes in progressive H- N HSQC spectra. At 1  15  an Ets-1 " :UBC9 molar ratio of 1:4, significant chemical shift perturbations were (1  138)  observed for residues 13 to 21 (Figure 5.3B). This region includes the entire  IK15TE  sumoylation motif. Although these chemical shift changes are relatively modest and do not exceed a combined A8 of 1.4 ppm, the values are in keeping with the titrations of UBC9 with p53 and c-Jun also monitored by NMR spectroscopy (combined A8 values less than 0.4 and 0.8 ppm, respectively; Lin et al., 2002). In contrast to the intermediate exchange regime described for Ets-1 and ERK2 (Chapter 4), the progressive change in amide HSQC peaks monitored during the above titration between Ets-1 and UBC9 was indicative of fast exchange. Using the UBC9-binding induced perturbation in the N chemical shifts of Ile13, Ile4, Lys15 and 15  H  Lys18, an estimate K of 240 ± 50 \iM was obtained (Figure 5.4). This value is in d  keeping with a fast exchange regime in which Kd values are typically higher than 100 |j,M. A literature search for reported Kd values for other UBC9/substrate interactions  145  Chapter 5 - Interactions with UBC9 and SUMO-1  [ N]Ets-1< - > 15  29  138  Lit  l---j- r  [ N] Ets-1 < - > 15  1 138  0.4 0.2  0  -U•H  11 nilJ.t ill J Ju,.L.jj d..iliiu .i .i.lnlll .L  .nil  Residue  29  _J_  H1  138  H2  LH3J"iH4j  H5  GK EC  IK TE  110  15  Figure 5.3. Titration of [ N]Ets-1 with UBC9. Unlabeled UBC9 was added to [ N]Ets-1< - > (A) or [ N]Ets-1< - > (B) and the resulting chemical shift changes from zero to one molar equivalent UBC9 were determined. The amount of change has been calculated using A 5 = [ ( 5 * 1 0 ) + ( 8 ) ] . Note that significant perturbation only occurs in the longer Ets-1 " construct within the region of Lys15. Combined digital resolution is 0.1 ppm (dashed line), with significant chemical shift change determined to be at least five times the digital resolution (black line). 15  15  29 138  2  2  HN  15  05  N  (1  138)  146  1 138  Chapter 5 - Interactions with UBC9 and SUMO-1  Ile13 K = 168 ± 5 6 (iM d  Ile14 K = 264 ± 40 [iM d  Lys15 K = 199 + 27 d  Lys18 K = 320 ± 63 |aM d  Volume U B C 9 added / nL  Figure 5.4. Determination of K between UBC9 and Ets-1 " . The titration between [ N]Ets-1 " and unlabeled UBC9 was used to estimate binding affinity with the approach detailed in section 5.4.4. Four of the residues in the region surrounding Lys15 (A to D; Ile13, Ile14, Lys15 and Lys18) displayed clear N chemical shift perturbation upon addition of UBC9, and are well fit by a binding curve for protein-ligand association in a fast exchange regime. Using all four calculated values, an average K of 240 ± 50 \xM was obtained. (1  d  15  (1  138)  1 5  H  d  147  138)  Chapter 5 - Interactions with UBC9 and SUMO-1  revealed a dearth of such information, except for an equivalent NMR titration study of UBC9 with peptides corresponding to the sumoylation sites of p53 and c-Jun (Lin et al., 2002). Performed at a slightly lower ionic strength and pH (100 mM potassium phosphate, pH 6.5), an estimate K of 3-6 mM was obtained for these systems. As d  mentioned previously, these proteins likely require the function of E3 SUMO-protein ligases for a maximal level of sumoylation. In section 5.4, a Kd of <50 ^M is estimated for the interaction between UBC9 and RanGAPL Unlike the case for p53 and c-Jun, RanGAPI does not appear to require association with an E3 protein, which is in keeping with the observed increase in affinity. The finding of a moderate binding affinity between Ets-1 and UBC9 (K of -240 p.M) supports the idea that this N-terminal d  sumoylation site is a true site of interaction within the sumoylation system, but does not rule out the possibility that increased affinity may be observed upon the identification of an appropriate E3 SUMO-protein ligase. To confirm the lack of interaction with Lys110 (the residue in Ets-1 corresponding to Lys99 in Tel), a second titration was performed with N-labeled Ets15  ^ (29-138)  T h j s  c o n s  t  r u c  t fj  oes  no  t contain the N-terminal sumoylation consensus site  and does not specifically interact with UBC9 in an NMR-monitored titration study performed using the same method as described above for Ets-1 " (1  138)  (Figure 5.3A).  Although small shifts were noted near Lys110, these were comparable to those for many residues across the PNT domain and may reflect experimental uncertainty in shifts (» 0.2 ppm), as well as non-specific effects such as subtle changes in buffer composition, pH, or a slight increase in Ets-1 " (29  138)  oxidation (involving the adjacent  Cys99, Cys106 and Cys112). Also noted were likely non-specific chemical shift changes for residues within helix H1. An estimation of affinity for the small changes 148  Chapter 5 - Interactions with UBC9 and SUMO-1  observed for residue Gly109 and Glu111, which neighbour Lys110, reveals a linear change in chemical shift upon addition of UBC9, with no evidence of a binding plateau. Therefore a K » 1 mM is predicted for binding to the Lys110 site on the PNT domain, d  as compared to the tighter binding (K of -240 yM) that UBC9 exhibits toward the Nd  terminal consensus motif. It therefore appears that Lys15 is an actual sumoylation site in Ets-1. From the NMR characterization of Ets-1 " , random coil H and N chemical shifts indicate (1  138)  1  N  15  that this region is unstructured. However, calculation of the CSI from the C and C 13  a  13  B  chemical shifts predicts a p-strand for Leu11-Lys15 and Asp6-Asp20, respectively. This region covers the  IK15TE  motif in Ets-1. It is interesting to note that the short  peptides of p53 and c-Jun used in defining the substrate-binding site on UBC9 were also extended (Rustandi et al., 2000; Lin et al., 2002). In addition, the crystal structure of the UBC9- RanGAPI complex indicates that the RanGAPI consensus sumoylation site exists within a large loop region that is possibly flexible (Bernier-Villamor, 2002). Finally, secondary structure predictions for other protein substrates, such as kBa, AdE1B and PML, indicate a coil conformation for the SUMO acceptor regions (Lin et al., 2002). It is therefore likely that a prerequisite for sumoylation may be the presentation of a consensus YKxE motif within an otherwise extended and flexible conformation.  5.4 Binding surface on UBC9 Several aspects of UBC9 biochemistry have already been described, including surfaces important for SUMO-1 binding, residues critical in catalysis and regions  149  Chapter 5 - Interactions with UBC9 and SUMO-1  Figure 5.5. Structure of UBC9. (A) Ribbon representation of UBC9 (PDB 1A3S) with the N- and C-termini indicated, as well as the location of the four a-helices and four p-sheet regions. (B) Same view as in A, but with a surface representation indicating regions of positive (blue) and negative (red) charge. The hydrophobic surface region shown to interact with substrate proteins is indicated, as is the location of the catalytic cysteine (Cys93). (C) Same representation as in B, but rotated 180° around the vertical axis. Shown is the large region found to bind SUMO-1 by NMR investigation both in our laboratory and by Liu et al. (1999). Catalytic importance has not yet been attributed to this latter region. Figures were made by using molscript v2.1.2 (Kraulis, 1991) and Raster3d v2.4b (Merritt and Bacon, 1997).  150  Chapter 5 - Interactions with UBC9 and SUMO-1  required for substrate binding. Figure 5.5 details the tertiary structure of human UBC9 with an accompanying representation of surface charge. Important to this thesis is the identification of UBC9 surface regions implicated in substrate binding. Fortunately, some previous work has already established residues important for several known sumoylation targets. NMR investigation of the substrate binding regions on UBC9 revealed a small site of interaction for p53 and c-Jun adjacent to, but not including, the catalytic Cys93 (Figure 5.6C.D; Lin et al., 2002). Similar to many other surface regions involved in protein association, the area of substrate interaction is predominantly hydrophobic with surrounding acidic and basic residues. The crystal structure of UBC9 bound to the substrate RanGAPI (Bernier-Villamor et al., 2002) indicates a similar binding surface, with some additional contacts relative to those defined by NMR shift perturbations. Validation of the interaction between Ets-1 " (1  138)  and UBC9 was therefore  possible by using NMR spectroscopy to determine which residues in UBC9 were in contact with Ets-1 " . The identified association region on UBC9 could then be (1  138)  compared with the known binding sites of p53, c-Jun and RanGAPI (Figure 5.5B). To this end, a titration of 400 |aM [ N]UBC9 was completed by using increasing amounts 15  of unlabeled Ets-1 - . Comparison of the initial H- N HSQC spectrum of [ N]UBC9 (1  138)  1  15  15  (fortunately assigned in the literature; Liu era/., 1999b) with one obtained with a molar equivalence of Ets-1 " (1  138)  reveals significant chemical shift perturbation for residues  including Ala5, Ala10, Gly23, Phe64, Val92, Leu94, Ala129, Gln130, Ala131, Tyr134 and Thr135. (Figure 5.6A). This pattern of chemical shift change compares well to the shifts obtained for UBC9 binding to the p53 and c-Jun peptides (Figure 5.6A). A surface representation of UBC9 highlighting the residues perturbed by addition of Ets151  Chapter 5 - Interactions with UBC9 and SUMO-1  Figure 5.6. Titration of N-labeled UBC9 with unlabeled Ets-1< - >. (A) Difference in the H- N amide chemical shifts for [15N]UBC9 with zero and one molar equivalence of Ets-1 " , calculated with A5=[(8 -10) +(5 ) ] - . Combined digital resolution of 0.1 ppm (dashed line), with significant chemical shift change determined to be at least three times the digital resolution (black line). Also indicated below the graph are residues that have been shown to experience chemical shift perturbation upon binding p53 () and c-Jun () peptides (Lin et al., 2002). (B) Surface representation of UBC9 displaying residues that shift by more than 0.3 ppm in panel A (cyan). (C) For comparison, the residues involved in p53 and c-Jun binding (green; Liia et al., 2002) and RanGapl binding (magenta; Bernier-Villamor et al., 2002) have been indicated on the surface of UBC9. For all three binding diagrams, the catalytic Cys93 is highlighted in yellow. Figures were generated by using GRASP (Nicholls etal., 1993). 15  1  1 138  15  (1  2  HN  2 0 5  N  152  138)  Chapter 5 - Interactions with UBC9 and SUMO-1  A  2  1  c-Jun binding site  153  Chapter 5 - Interactions with UBC9 and SUMO-1  1  (1-138)  (pjg  ure  5.6B), reveals a binding region consistent with the sites of previously  characterized targets of sumoylation (Figure 5.6C). A site on UBC9 has also been described as a SUMO-1 binding region, presumably reflecting the location occupied by SUMO-1 following thioester formation, but preceding transfer to the target substrate (Liu et al., 1999a). To confirm this interaction, and assure that SUMO-1 and Ets-1 do not compete for binding, unlabeled SUMO-1 was added to the final point of the above [ N]UBC9 and Ets-1 " 15  (1  138)  titration.  Following SUMO-1 addition there were residue specific decreases in the UBC9 H- N 1  15  HSQC amide peak intensity of Lys14, Ala15, Arg17, Lys18, Asp19, His20, Gly23 and Phe24 (Figure 5.7A). These residues map a SUMO-1 binding region on UBC9 (Figure 5.7B) that is consistent with the interface previously described using NMR-derived decreases in peak intensity (Figure 5.7C; Liu et al., 1999a). In addition, there were no significant changes in chemical shift for residues involved in Ets-1 " (1  indicating the potential for concurrent binding of Ets-1 " (1  138)  138)  binding,  and SUMO-1 by UBC9. As  a result, it is possible to think of UBC9 as a 'scaffolding' enzyme, using a catalysis mechanism based on mediating the interaction of the Gly-Gly tail of SUMO-1 with an appropriately placed acceptor lysine residue on the target protein (Figure 5.7D). This schematic description is in keeping with a model in which the minimal TKxE sumoylation motif is sufficient for productive modification of substrates by UBC9 and SUMO-1. Indeed, unlike the absolute requirement for E3 enzymes in the recognition and modification of ubiquitinylated targets, it appears that the E2 enzyme UBC9 is all that is required for sumoylation of targets that display the tetrapeptide YKxE sequence. However, it is still possible that many or all SUMO-1 substrates display 154  Chapter 5 - Interactions with UBC9 and SUMO-1  Figure 5.7. Titration of [ N]UBC9unlabeled Ets-1 < » with unlabeled SUMO-1. (A) Ratio in the H- N amide chemical shift intensity for [ N]UBC9 with one-half and zero molar equivalents of human SUMO-1 ~ (SUMO-GG). Height ratios were normalized to give an average change of 1.0 (grey dashed line) (B) Surface representation of UBC9 showing residues (in red) that decrease by more than 50% versus the average (indicated by the black line in A). (C) For comparison, the residues of UBC9 involved in SUMO-1 binding by Liu et al. (1999) are shown in black. (D) Cartoon illustrating the role of UBC9 to bridge together the C-terminal GlyGly (GG) of SUMO-1 (orange) with the Lys e-amino group (K) of the target protein sumoylation motif (Lys15 in Ets-1, shown in teal). UBC9 is represented as a molecular surface with the NMR-derived Ets-1 " (see Figure 5.6) and SUMO-GG binding sites shown in cyan and red, respectively. The catalytic cysteine, Cys93, is shown in yellow. Figures were generated by using GRASP (Nicholls etai., 1993). 15  1138  1  15  15  (1  (1  155  138)  97)  Chapter 5 - Interactions with UBC9 and SUMO-1  156  Chapter 5 - Interactions with UBC9 and SUMO-1  increased rates of sumoylation with the aid of an appropriate E3, especially for substrate proteins that do not contain an optimal consensus motif. Alternatively, some of these substrates may possess extra interacting surfaces that increase affinity to UBC9, and act somewhat as an intramolecular SUMO-protein ligase. An example of this latter case is observed for the binding of UBC9 with RanGAPI (Bernier-Villamor et al., 2002). In agreement with the crystal structure, NMR titration of [ N]UBC9 with 15  unlabeled RanGAPI ~ (422  587)  reveals a more extensive region of contact than that  observed for the interaction with Ets-1 " (1  138)  (Figure 5.8). In keeping with an increased  affinity, it is observed that an intermediate exchange regime is displayed by the interaction of [ N]UBC9 with RanGAPI " 15  (420  587)  , instead of the fast exchange observed  for the binding of UBC9 with Ets-1 " . As a result of the intermediate exchange (1  138)  regime, the K for the interaction between UBC9 and RanGAPI d  (420  "  587)  is expected to  be <50 uM. Unfortunately, a binding curve to determine the actual Kd (as for the fast exchange binding between Ets-1 " (1  138)  and UBC9) cannot be used for RanGAPI, due  to this slower exchange. The K that would be obtained by fitting the change in peak d  intensity for residues Gln126, Ala129, Gln130, Glu132 and Tyr134 produces a clearly too high estimate of 210 ± 50 |aM. A more accurate dissociation constant for this nonEts family protein would require alternative methods, such as surface plasmon resonance (BIAcore).  5.5 Biological context The above set of experiments serves to describe in molecular detail the association of a consensus sumoylation site within Ets-1 with the SUMO-conjugating enzyme UBC9. However, apart from a two-hybrid result indicating interaction between a PNT-domain157  Chapter 5 - Interactions with UBC9 and SUMO-1  Figure 5.8. Titration of [ N]UBC9 with unlabeled RanGAPL (A) Ratio in the H- N amide chemical shift intensity for [ N]UBC9 with one and zero molar equivalents of human RanGAPI < " . Downward arrows below the graph indicate UBC9 residues that are situated within 6 A of the mouse RanGAPI protein in the crystal structure (BernierVillamor et al., 2002). (B) Surface representation of UBC9 showing residues (in cyan) that decrease by more than 80 % versus Ile4 in panel A (black line). (C) For comparison, the residues of UBC9 involved in RanGAPI binding from X-ray work (Bernier-Villamor et al., 2002) are shown in magenta. In both B and C the catalytic Cys93 is highlighted in yellow. Figures were generated using GRASP (Nicholls etai., 1993). 15  1  15  15  420  (420_589)  158  587)  Chapter 5 - Interactions with UBC9 and SUMO-1  1.2  1  0.8  H  i  0.6  as CD  0.4  0.2  **T  VT  O J C M < O C O ^ - ^ e r t « i / 5 c O C D r ^ f » - 0 0 0 0 0 ) C T i O O  ,  » —  i -  CM  CM  to  CO  mmnmi i  Crystal structure contacts  159  1  Ifl  Ifl  Chapter 5 - Interactions with UBC9 and SUMO-1  containing fragment of Ets-1 domain and UBC9 (Hahn et al., 1997), there is no data providing direct support for SUMO-1 modification of Ets-1. Therefore, it was decided to start with an in vitro study to determine if indeed Ets-1 could be sumoylated by a reconstituted system of purified proteins. In addition, large preparations of sumoylated proteins (e.g. SUMO- Ets-1 " ) would allow for characterization of the modified (1  138)  protein by NMR spectroscopy. Investigation at the molecular level would provide necessary information regarding the consequence of sumoylation, including any induced structural perturbations or changes in protein dynamics. In lieu of such changes, there would be an indication that modification by SUMO may simply result in the addition of a protein 'tag' that elicits specific nuclear targeting. Following the protocols from several different laboratories, sumoylation of test substrates was first attempted by using a cell extract containing catalytic amounts of the SUMO-activating enzymes, along with purified UBC9 and C-terminally truncated SUMO-1 (i.e. SUMO-GG). No significant amount of sumoylated protein substrate was observed. To increase rate of SUMO-GG activation, the E1 components SAE1 and SAE2 were co-expressed in E. coli BL21(A,DE3) cells, then purified as a single complex using Ni -affinity chromatography and gel filtration chromatography. 2+  Unfortunately, using the purified SUMO-activating enzyme produced only minimal amounts of sumoylated RanGAPI (even after a 24 h reaction). In addition, there was no indication of Ets-1 " (1  138)  sumoylation, or indeed sumoylation of any other PNT-  domain containing construct. Since attempts in our lab to achieve in vitro sumoylation have so far met with failure, collaboration was initiated with Dr. Barbara Schulman (St. Jude Children's Research Hospital). Dr. Schulman has used the yeast sumoylation system in previous 160  Chapter 5 - Interactions with UBC9 and SUMO-1  ySAE1/2  + +  hUBC9  + - +  ySUMO  - +  substrate  -  +  -  +  + + + + + + + + + +  +  + + + + + + + + + + +  +  _ _ __  -  +  -  Tel  +  -  +  Ets-1 Ets-1  V113E  1-138  29-138  -  +  Tel  Fli-1  A94D  106-200  66 —  -SAE2  45 —  -SUMO-SUMO "SAE1  3114—  — —  Jggr tm» mm mm mm mm. mm  •4—UBC9 -SUMO  12-  substrates  6— 1  2  3  4 5  6  7  8  9  10 11 12 13 14  Figure 5.9. In vitro sumoylation of Ets family proteins. Experiment and figure by Dr. Brenda Schulman (St. Jude Children's Research Hospital). In vitro sumolation was tested using Tel< - >V113E, TeK - >A94D, Ets-1 <" >, . - | (29-138) p|(106-200) | p jfj | SAE1/SAE2, yeast SUMO and human UBC9. Only Ets-1 <- > demonstrates the addition of SUMO (*) during the 60 min incubation at room temperature (compare lanes 7 and 8). 38 127  E  T  S  a  n  d  a  o n g  w j t h  38 127  Ur  1 138  161  ec  y e a s t  1 138  Chapter 5 - Interactions with UBC9 and SUMO-1  experiments (Benscath et al., 2002), and she was able to assay several of our purified PNT domains for their ability to be sumoylated with the yeast E1 and SUMO proteins and human UBC9 (Figure 5.9). As expected, only the larger Ets-1 " (1  138)  construct was  found to be sumoylated to any significant extent, especially in comparison to the Ets^ (29-138) c o n s  t  r u c  SUMO- Ets " (1  t that lacked Lys15 (compare lanes 8 and 10, Figure 5.9). The putative  138)  protein (asterisk in lane 8, Figure 5.9) was excised, subjected to  trypsin digest and analyzed by using MALDI-TOF mass spectrometry to confirm the presence of both Ets-1 " (1  138)  and yeast SUMO peptides (not shown). For the Tel PNT  domains, it was observed that once again the PNT domain was not susceptible to modification (lanes 6 and 12, Figure 5.9) despite the previous claim that Lys99 was a site of sumoylation (Chakrabarti et al., 2000). It will be of interest to clone larger Nterminal constructs of Tel and determine if Lys11 (corresponding to Lys15 in Ets-1) can be sumoylated in vitro. A fifth control protein, the PNT domain from Fli-1, was also resistant to SUMO attachment. Still lacking in the treatment of Ets-1 and UBC9 interaction is the demonstration of sumoylation in vivo - a key requirement to establishing a cellular role for this modification. Although outside the scope of this thesis, the presence of sumoylated Ets-1 will be sought in mouse- and human-derived cell lines using immunoprecipitation and Western blot identification, and it is anticipated that similar to Tel, a fraction of Ets1 will show an increased molecular mass that will be identified as covalently attached SUMO-1. However, there is already some indication that Ets-1 may be functionally tied to the same mechanism of repression exhibited by other proteins that interact with sumoylated members of the nuclear bodies. In particular, the protein Daxx has already been shown to be a true component of the nuclear bodies, and is responsible for 162  Chapter 5 - Interactions with UBC9 and SUMO-1  repressing the activity of Ets-1 via a further association with the histone deacetylase complex (HDAC). This interaction between compartmentalized Daxx and Ets-1 causes a specific repression for the Ets-1 regulated mmp-1 and bcl-2 genes (Li et al., 2000). Analysis of the primary sequence of Ets-1 allows for additional conjecture concerning the potential role of sumoylation (Figure 5.2). The proximity of the ERK2 phosphorylation site to Lys15 provides a scenario in which either sumoylation ability is modified by the phosphorylated state of Thr38, or the reverse case in which sumoylation of Lys15 modifies the ability of Ets-1 to become activated via ERK2. These hypotheses are made all the more interesting with the observation of additional sumoylation consensus sites near to the Ca -dependent calmodulin kinase II 2+  phosphorylation sites N-terminal to the ETS DNA binding domain. Involvement of sumoylation in this region could add a further regulatory step in the ability of Ets-1 to bind DNA apart from phosphorylation, truncation or protein association (Petersen et al., 1995; Cowley and Graves, 2000; Wang et al., 2002). Similar regulation may also be observed for the  orthologue Pnt-P2 due to conservation of putative  Drosophila  sumoylation sites N-terminal to both the MAPK phosphorylation site (which in Drosophila  is the Rolled kinase) and the region subjected to calmodulin kinase II  phosphorylation in Ets-1. Interestingly, Ets-2 is devoid of *¥KxE consensus motifs and may be resistant to any form of regulation by the sumoylation pathway.  5.4 Materials and methods 5.4.1 Cloning  Murine Tel " A94D, Tel - V113E, Ets-1 (38  127)  (38  127,  (1_138)  and Ets-1 " (29  138)  were cloned,  expressed and purified as previously described (Section 2.5 and Slupsky et al., 1998). 163  Chapter 5 - Interactions with UBC9 and SUMO-1  Human UBC9, SUMO (SUMO-1), SUMO-GG (SUMO - ), RanGAPI " (1  97)  (420  587)  and  SAE1 were PCR amplified from a HeLa cell QUICK-Clone cDNA library (Clontech Laboratories,  Inc.)  using  TCGGGGATCGCCCTC-3',  the  primers  UBC9-R,  UBC9-F,  5'-GGAATTCACATATG-  5'-GGAACGTCTCGAGGGTCGCTGC7TAT-  GAGG-3', SUMO-F, 5-GGAATTCACATATGTCTGACCAGGAGG-3', SUMO-R, 5'CCTACGTCTCGAGCTAAACTGTT-GAATG-3', SUMO-R2, 5'-CCTACGTCTCGAGRanGAPI-F, 5'-GATATCATATGGGGGAGCCAGC-  TCAACCCCCCGTTTGT-TCC-3', TCCCGTG-3',  RanGAPI-R,  5'-GGCCGCTCGAGCL4GACCTTGTACAGCG-3'.  SAE1-F, 5-GGAATTCAGCTAGCATGGTGGAGAAGGAGGAGG-3' and SAE1-R, 5'CCTACGTTCTCGAGAA4TCTTGAGTTCACTTGG-3'. SAE2 was PCR-cloned from a plasmid generously provided by Dr. Ronald Hay (St. Andrew's University; Desterro et al., 1999) using the primers, SAE2-F, 5'-GGAATTCAGCTAGCATGGCACTGTCGCGGGGGCT-3',  SAE2-R,  5'-CCTACGTCTCGAGCTGTTCAATCTAATGCTATG-3'.  Restriction sites in the primers have been underlined, introduced stop codons are in italics and coding bases in bold. PCR products were cloned into pET28a vector using Nde\/Nhe\ and Xho\ restriction enzyme sites to retain the N-terminal HiS6 tag, except  for SAE1, which was cloned into pET22b using Nhe\ and Xho'l without any purification tags. DNA sequencing was used to confirm the accuracy of cloning. Proteins were expressed in BL21(?JDE3) cells for 3 h at 37 °C with 1 mM IPTG, and purified by the standard Ni -column purification as described previously for Erg in section 3. The 2+  SAE1/SAE2 complex was further purified using an S-200 sepharose gel filtration column with 25 mM Tris (pH 7.5) and 150 mM NaCl. Following purification, samples were checked for identity and purity using SDS-PAGE and electrospray ionization mass spectrometry. 164  Chapter 5 - Interactions with UBC9 and SUMO-1  5.4.2 Cross-linking  PNT domain samples (40 |aM each) were variably mixed with UBC9 (40 [iM) or SUMOGG (40 \iM) in 20 mM potassium phosphate (pH 7.0), 50 mM NaCl and 10 mM DTT. Cross-linking utilized 0.001 % (v/v) glutaraldehyde for 30 min at 22 °C. Reactions were quenched by using SDS-PAGE loading buffer and analyzed with 15% SDS-PAGE and Coomassie staining.  5.4.3 NMR spectroscopy  All spectra were recorded on a 500 MHz Varian Unity or a 600 MHz Varian Inova NMR spectrometer equipped with a triple resonance probe and a pulsed field gradient accessory and were analyzed using Felix 2000 (MSI) and Sparky. Titration experiments were performed in 10 mM potassium phosphate (pH 6.5), 10 mM KCI and 5 mM DTT, with 10% D 0 added for the lock. 2  5.4.4 Determination of dissociation constant  Assuming fast exchange between the bound and unbound states, the following equation was used to calculate the dissociation constant (based on Johnson et al., 1996; Amezcua et al., 2002):  5N = 8N + (5N - 5N ) x {(L + P + K ) - [(L + P • + K ) - (4 x L x P)f } / (2 x P) 2  U  b  U  d  5  d  where 8N is the observed N amide chemical shift at the ligand concentration L, 5N is 15  U  the initial unbound N amide chemical shift and P is the concentration of protein. For 15  165  Chapter 5 - Interactions with UBC9 and SUMO-1  ease of calculation, the values of L and P were rewritten based on the added volume of ligand:  L  =  ([L]stock X  Vadded) / (Vdded Vj) +  a  P = ([PJstock X V|) / (Vadded + Vj)  Where V ed is the added volume, add  [L] t ck s 0  and  [P] tock s  are the initial concentration of  ligand and protein, respectively, and Vi is the initial volume. Values for the dissociation constant, Kd, and N chemical shift of the bound protein, 8N , were determined by 15  b  using a non-linear regression fit (Kaleidagraph) with the experimental values of 8N and Vadded-  5.4.5 Sumoylation assay  Purified Tel - A94D, Tel - V113E, Ets-1 " , Ets1 " (37  126)  (37  126)  (29  138)  (1  138)  and Fli-1 (106  200)  were  tested as previously described (Benscath et al., 2002) at an approximate concentration of 10 uM in a buffer containing 12 uM Smt3p (yeast SUMO), 4 |JVI human UBC9, 4 mM ATP, 30 mM Tris (pH 7.6), 190 mM NaCl, 10 mM MgCI and 0.2 mM DTT. 2  Sumoylation was initiated with the addition of 4.5 nM yeast E1 and allowed to react for 60 min at room temperature. Reactions were stopped by adding SDS-PAGE gel loading buffer coupled with boiling, and separated on 15% SDS-PAGE visualized by Coomassie blue staining.  166  Chapter 6 Concluding remarks  This thesis has dealt with the structural and functional characterization of several PNT and SAM domain-containing proteins, at the level of tertiary architecture, association and post-translational modification. Common to these domains is a core bundle of four a-helices. The additional N-terminal helix found in Ets-1 and GABPa illustrates plasticity in architecture. Functional characterization has reveled a diverse spectrum of possible protein interactions ranging from the 3:1 Ste4:Byr2 complex to a role for Ets-1 as a MAPK docking module. In addition, modification of N-terminal residues by phosphorylation and sumoylation has also been investigated, but so far the detailed consequence of these changes remain unknown. In the end, however, two overall themes are apparent from the work presented in this thesis. First is the surprising level of structural variance displayed by these otherwise conserved protein domains. By looking at conserved surface features, several of these functions can be understood and predicted at the molecular level. Secondly, there is an increased significance given to unstructured regions in Ets proteins. It is indeed within these flexible regions that post-translational modification occurs. For Ets-1 '  <1 138)  and GABPa '  (168 290)  it has  been shown further that these changes do not elicit induced structure, altered dynamics or a perturbation of the adjacent PNT domains. Instead, it is likely that these regions must retain a random coil conformation in order to bind to proteins involved in the transcriptional machinery. 167  Chapter  6 - Concluding  remarks  6.1 Common and divergent aspects of the PNT and SAM domains  An overall theme of this thesis is the repeated observation of features that are shared by PNT and SAM domains, as well as differences evident in both structure and function. Identification of common attributes is perhaps not surprising, since it is through sequence and structure similarity that these domains were initially classified. As a realistic corollary, it is expected that PNT AND SAM domain members would also share a core tertiary structure. As detailed in Chapter 2, all PNT and SAM domain structures currently available indeed demonstrate a core bundle of four a-helices with conservation in the position and identity of the residues comprising the hydrophobic core. More surprising was the finding of diversity in the structure and the clear variability in function demonstrated by this otherwise conserved domain. For example, analysis of the Erg PNT domain structure (as with Tel) reveals a lack of the N-terminal helix present in Ets-1 and GABPa (Chapter 2). This plasticity in the backbone structure hints at a more profound variance in potential biological roles. During the initial stages of this thesis work, it was believed that the PNT and SAM domains were modules of homotypic and heterotypic association with other PNT AND SAM domains. A second reading of the SAM acronym became 'self-association module' (Kyba and Brock, 1998). However, it should now be clear from the results of this thesis and due to an increasing number of newly found binding partners, that the role of the PNT and SAM domains is far more diverse than originally suspected. Only the domains from Tel, Yan and several polycomb group proteins demonstrate strong evidence of oligomerization. The Byr2 and Ste4 SAM domains illustrate a unique story of a 3:1 heteromeric complex (Chapter 3), and the ephrin receptor SAM domains may or may not contribute 168  Chapter 6 - Concluding remarks  •  to functional dimerization due to a very low binding affinity. The association of auxiliary proteins such as ERK2 (Chapter 4) adds additional functionality to the PNT domain. Several other members are clearly monomeric in solution and have as yet no identified binding partner or other defined role. This amazing variety in function is further detailed in Figure 6.1.  6.2 Surface areas utilized in association  A distinct benefit of structural investigation is the ultimate description of both the architecture and surface properties of the proteins under study. Although the overall fold of a domain can be important in terms of classification, relatively little functional information can be gleaned from the backbone architecture alone. It is instead the arrangement of solvent exposed residues that define the ability of one protein to interact with another protein or molecule. The surfaces of monomeric PNT domains illustrate a clear lack in the hydrophobic residues required for self-association by the Tel PNT domain and polyhomeotic SAM domain. Similarly, the residues deemed important for the interaction of Ets-1 with ERK2 allow for the identification of other PNT domains that may serve a comparable function (e.g. Ets-2 and Pnt-P2). It is only with a full list of potential functions and corresponding surface residues that realistic hypotheses can be generated for as yet uncharacterized PNT AND SAM domains. To this end further complexes must be studied through both NMR analyses and cocrystallization studies. Collaboration has also been initiated with Dr. Tom Alber (University of California, Berkeley) to further characterize the interaction of the Ets-1 PNT domain with ERK2 via X-ray crystallography. In addition, crystallization trials of the Byr2/Ste4 complex are still underway in the laboratory of Dr. James Bowie 169  Chapter 6 - Concluding remarks  Oligomeric state  Schematic  Examples Ets-1 GABPa Erg Fli-1 p73  monomer  Dimer (Kd ~ mM)  EphA4 EphB2  Byr2/Ste4  Heteromer (Kd nM - uM)  Ste50/Ste11 Mae/Yan  Tel Polyhomeotic  Oligomer (Kd ~ nM)  Yan  Ets-1  Docking (Kd ~ uM)  Ets-2 Pnt-P2  Figure 6.1. Diverse modes of protein association exhibited by the PNT and SAM domains. Examples in italics have not been verified at the molecular level.  170  Chapter 6 - Concluding remarks  (University of California, Los Angeles) following the inability to structurally describe these proteins using NMR-based techniques.  6.3 Structured domains and flexible linkers  Another important theme within this thesis was the investigation of protein dynamics within several different PNT AND SAM domain constructs. As expected for an isolatable peptide module, the PNT domain of Erg (as well as Ets-1 and GABPa; Slupsky et al., 1998a; Scharpf et al, unpublished) resisted proteolytic cleavage by trypsin and exhibited a classic two-state thermal unfolding mechanism. At the molecular level, NMR relaxation analyses have confirmed that most residues within the Erg and GABPa PNT domains exhibit restricted motions on the millisecondpicosecond timescale indicative of well-folded globular proteins. Addition of N-terminal or C-terminal flanking peptides did not affect stability or structure, nor did association with binding partners. Together these results argue for a structurally rigid PNT domain resistant to changes in association state or extraneous peptide sequence. In stark contrast to these well-folded domains, adjacent regions of Erg, GABPa and Ets-1 are highly sensitive to proteolysis and do not contribute to enhanced protein stability in these larger constructs. Using NMR spectroscopy, regions neighbouring PNT domains exhibited random coil chemical shifts with significant flexibility as indicated by N relaxation analysis and lack of amide hydrogen-exchange protection. 15  Addition of these regions did not affect the structure or stability of the adjacent PNT domains, and even phosphorylation of the proximal ERK2 phosphorylation sites in Ets1 and GABPa failed to elicit noticeable changes in backbone dynamics. Therefore it  171  Chapter 6 - Concluding remarks  appears that for Erg, Ets-1 and GABPa the entire N-terminal region is composed of a stretch of flexible and disordered residues - a trend broken only by the single structured PNT domain. It is still possible that these random coil regions are mere artifacts of the protein truncations or working in vitro with pure protein (e.g. the region may be complexed in vivo). However, trypsin digest of full-length native Ets-1 indicates rapid proteolysis of residues that lie outside of the PNT and ETS domains, thus confirming a significant degree of conformational flexibility for the N-terminal and central regions. It is likely that this scenario will also apply to the other members of the Ets family of transcription factors. Illustrated in Figure 6.2 is a model of the Ets-1 transcription factor that highlights its modular organization of folded domains separated by flexible random coil-like regions. A similar diagram can likely be drawn for other members of this subset of Ets proteins that include Ets-1, Ets-2, Pnt-P2, Erg, Fli-1, GABPa, Tel and Yan. In the molecular representation of the full-length protein (Figure 6.2B) there is a significant amount of disordered peptide in the N-terminal region preceding the PNT domain, and in the large stretch of residues comprising the previously characterized transactivation domain. The number of unstructured residues present in a protein such as Ets-1 is perhaps surprising, given the large amount of well-defined structures available within the Protein Data Bank. However, the very nature of structural research relies upon the isolation of a folded soluble protein. For X-ray crystallography, highly flexible proteins are notoriously difficult to crystallize, necessitating mutations, truncations or binding partners to stabilize the protein of interest prior to structure calculation. NMR spectroscopy has no such absolute need for protein fold stability, but is still hindered by the highly degenerate nature of chemical shifts displayed by random coil residues. 172  Chapter 6 - Concluding remarks Ca -dependent CKII phosphorylation 2+  sumoylation  inhibitory module  ERK2 phosphorylation  N-  mm  Figure 6.2. Model of Ets-1 involving structured domains and flexible linkers. (A) A simple diagram of Ets-1 primary sequence indicates regions containing the PNT domain and ETS DNA-binding domain. (B) Using NMRderived restraints of the Ets-1 PNT and ETS domains (Slupsky et al., 1998; Lee et al., unpublished) a ribbon representation of the entire Ets-1 protein was generated using ARIA v1.2. Regions adjacent to each domain appear to be flexible from N relaxation, and only the region from residue 139 to 243 has not yet been characterized at the molecular level. However, this latter region is sensitive to proteolysis and likely adopts a random coil conformation as depicted in the figure. Also indicated are functionally important residues including the ERK2 docking region and phosphoacceptor threonine, putative sumoylation site and region comprising several Ca -dependent CKII phosphoacceptor serines. The tertiary model was created by using Molscript v2.1.2 (Kraulis, 1991) and Raster3d v2.4b (Merritt and Bacon, 1997). 15  2+  173  .  Chapter 6 - Concluding remarks  It is therefore not surprising that relatively little structural research has been contributed to the study of these important protein regions sometimes relegated in literature to mere 'linkers' between structural domains.  The importance of this 'random coil protein module' has gained prominence in recent years (Wright and Dyson, 1999; Dunker et al., 2001; Dyson and Wright, 2002), and is supported by results described within this thesis. In relation to transcription factors, activation domains are frequently composed of a stretch of repetitive amino residues within a region potentially devoid of a-helical or p-strand conformation. The general activation domain of the yeast heat shock transcription factor provides one example of a completely random coil-like peptide (Cho et al., 1996). The acidic activation domain from p53 contains only a single a-helix and two p-turns in an extended conformation, and remains conformationally flexible in the full-length protein (Lee et al., 2000; Ayed et al., 2001; Bell et al., 2002). Phosphorylation of this region in p53 appears to further destabilize residual tertiary interactions and allow for binding by co-factors such as p300/CBP (Kar et al., 2002). For the unstructured KID domain of CREB, binding of the CBP KIX domain induces the formation of two a-helices that are further stabilized or destabilized by phosphorylation on Ser133 and Ser142, respectively (Parker et al., 1998). In other proteins, post-translational modification also seems to be mainly targeted to sequences that lie outside of folded domains. Specific to this thesis, Chapters 4 and 5 examined the disordered sites of Ets proteins associated with ERK2 phosphorylation and UBC9-mediated sumoylation, respectively. An obvious question therefore remains: what is the biological necessity for these large unstructured regions? Using the sumoylation tetrapeptide motif of RanGAPI as an example, it is evident that the four residues must be in an extended 174  Chapter  6 - Concluding  remarks  conformation in order to associate with UBC9 and become modified by SUMO-1. Similarly, protein domains such as SH2, SH3 and PDZ bind to short sequence motifs from a variety of protein targets (Pawson et al., 2001; Hung and Sheng, 2002). It is therefore possible that for many proteins, association occurs between a structured domain and an extended, flexible target sequence. From an evolutionary standpoint, it is clearly easier to introduce a specific linear sequence of amino acid residues within a disordered peptide versus a structured domain with many tertiary constraints. In addition, flexible regions would provide less steric resistance toward binding, but with a concurrent switch to a larger unfavourable entropic difference between the unbound and bound states. However, an unfortunate consequence of this simple motif is the inevitable appearance of inappropriately placed binding sequences with the flexible regions of additional non-specific proteins. An example of this scenario is the plentiful occurrence of the minimal MAP kinase phosphoacceptor motif (Ser/Thr-Pro) in proteins not likely to be true targets of MAP kinase phosphorylation. Some of these Ser/Thr-Pro dipeptides are likely to be intimately connected with a structured protein region and thus inaccessible to kinase contact. However, for exposed motifs this potential problem may have been overcome through the use of very specific docking sites that increase the affinity between an otherwise degenerate binding motif and its associated binding partner. The role of the Ets-1 and GABPa PNT domains as ERK2 docking modules (Chapter 4) is an excellent example of a mechanism by which specificity can be generated toward an otherwise common sequence motif. There are also several other described cases in which a second site increases the specificity between a substrate sequence and a protein kinase (reviewed in Biondi and Nebreda, 2003). 175  Chapter 6 - Concluding remarks  A final aspect to consider is the consequence of modification within otherwise disordered protein regions. Detailed in Chapter 4, phosphorylation of the ERK2 phosphoacceptor threonine residue in Ets-1 and GABPa, does not result in structural or dynamic changes but within the cell leads to an increase in transactivation. Therefore, it is most likely that the outcome of phosphorylation is either the increased association with elements of transcriptional activation (e.g. components of the RNA polymerase complex) or the reduced binding to repressor proteins. An example of this latter case has been discovered for Ets-2, for which ERK2 phosphorylation abrogates the interaction with the repressive SWI-SNF chromatin-remodeling complex (Baker et al., 2001). A similar case will likely develop for other transcription factors and lead to a broader scenario in which a change in protein association, and not altered structure or flexibility, is the outcome for post-translational modification.  176  References  References Albanese, C , Johnson, J., Watanabe, G., Eklund, N., Vu, D., Arnold, A., and Pestell, R. G. (1995). Transforming p21ras mutants and c-Ets-2 activate the cyclin D1 promoter through distinguishable regions. J. Biol. Chem. 270:23589-23597. Amezcua, C. A., Harper, S. M., Rutter, J., and Gardner, K. H. (2002). Structure and interactions of PAS kinase N-terminal PAS domain: model for intramolecular kinase regulation. Structure 10:1349-1361. Archer, S. J., Ikura, M., Torcia, D. A., and Bax, A. (1991). An alternative 3D-NMR technique for correlating backbone N-15 with side-chain H-beta resonances in larger proteins. J. Magn. Reson. 95:636-641. Athanasiou, M., Mavrothalassitis, G., Sun-Hoffman, L, and Blair, D. G. (2000). FLI-1 is a suppressor of erythroid differentiation in human hematopoietic cells. Leukemia 14:439-445. Avots, A., Hoffmeyer, A., Flory, E., Cimanis, A., Rapp, U. R., and Serfling, E. (1997). GABP factors bind to a distal interleukin 2 (IL-2) enhancer and contribute to cRaf-mediated increase in IL-2 induction. Moi. Cell. Biol. 17:4381-4389. Ayed, A., Mulder, F. A., Yi, G. S., Lu, Y., Kay, L. E., and Arrowsmith, C. H. (2001). Latent and active p53 are identical in conformation. Nat. Struct. Biol. 8:756-760. Baker, D. A., Mille-Baker, B., Wainwright, S. M., Ish-Horowicz, D., and Dibb, N. J. (2001). Mae mediates MAP kinase phosphorylation of Ets transcription factors in Drosophila. Nature 411:330-334. Baker, K. M., Wei, G., Schaffner, A. E., and Ostrowski, M. C. (2003). Ets-2 and components of mammalian SWI/SNF form a repressor complex that negatively regulates the BRCA1 promoter, J. Biol. Chem. epub. Baltzinger, M., Mager-Heckel, A. M., and Remy, P. (1999). XI erg: expression pattern and overexpression during development plead for a role in endothelial cell differentiation. Dev. Dyn. 216:420-433. Barr, M. M., Tu, H., Van Aelst, L., and Wigler, M. (1996). Identification of Ste4 as a potential regulator of Byr2 in the sexual response pathway of Schizosaccharomyces pombe. Moi. Cell. Biol. 16:5597-5603. Bartel, F. O., Higuchi, T., and Spyropoulos (2000). Mouse models in the study of the Ets family of transcription factors. Oncogene 19:6443-6454. Barton, K., Muthusamy, N., Fischer, C., Ting, C. N., Walunas, T. L., Lanier, L. L., and 177  References  Leiden, J. M. (1998). The Ets-1 transcription factor is required for the development of natural killer cells in mice. Immunity 9:555-563. Basuyaux, J. P., Ferreira, E., Stehelin, D., and Buttice, G. (1997). The Ets transcription factors interact with each other and with the c-Fos/c-Jun complex via distinct protein domains in a DNA-dependent and -independent manner. J. Biol. Chem. 272:26188-26195. Batchelor, A. H., Piper, D. E., de la Brousse, F. C., McKnight, S. L., and Wolberger, C. (1998). The structure of GABPa/(3: an ETS-domain-ankyrin repeat heterodimer. Science 279:1037-1041.  Baud, V., Lipinski, M., Rassart, E., Poliquin, L., and Bergeron, D. (1991). The human homolog of the mouse common viral integration region, FLU, maps to 11q23q24. Genomics 11:223-224.  Bauman, P., Cheng, Q. C., and Albright, C. F. (1998). The Byr2 kinase translocates to the plasma membrane in a Ras1 -dependent manner. Biochem. Biophys. Res. Comm. 244:468-474. Bax, A., Clore, M., and Gronenborn, A. M. (1990). H-1-H-1 correlation via isotropic mixing of C-13 magnetization, a new 3-dimensional approach for assigning H-1 and C-13 spectra of C-13-enriched proteins. J. Magn. Reson. 88:425-431. Behan, K. J., Nichols, C. D., Cheung, T. L., Farlow, A., Hogan, B. M., Batterham, P., and Pollock, J. A. (2002). Yan regulates Lozenge during Drosophila eye development. Dev. Genes. Evol. 212:267-276. Beier, F., Taylor, A. C , and LuValle, P. (1999). The Raf-1/MEK/ERK pathway regulates the expression of the p21(Cip1/Waf1) gene in chondrocytes. J. Biol. Chem. 274:30273-40279. Bell, S., Klein, C , Muller, L., Hansen, S., and Buchner, J. (2002). p53 contains large unstructured regions in its native state. J. Moi. Biol. 322:917-927. Ben-David, Y., Giddens, E. B., and Bernstein, A. (1990). Identification and mapping of a common proviral integration site Fli-1 in erythroleukemia cells induced by Friend murine leukemia virus. Proc. Natl. Acad. Sci. USA 87:1332-1336. Benscath, K., P., Podgorski, M. S., Pagala, V. R., Slaughter, C. A., and Schulman, B. A. (2002). Identification of a multifunctional binding site on UBC9p required for SMT3p conjugation. J. Biol. Chem. 277:47938-47945. Bergeron, D., Poliquin, L., Kozak, C. A., and Rassart, E. (1991). Identification of a common viral integration region in Cas-Br-E murine leukemia virus-induced nonT-, non-B-cell lymphomas. J. Virol. 65:7-15. 178  References  Bernier-Villamor, V., Sampson, D. A., Matunis, M. J., and Lima, C D . (2002). Structural basis for E2-mediated SUMO conjugation revealed by a complex between ubiquitin-conjugating enzyme Ubc9 and RanGAPI. Cell 108:345-56. Betz, S. F., Bryson, J. W., and DeGrado, W. F. (1995). Native-like and structurally characterized designed alpha-helical bundles. Curr. Opin. Struct. Biol. 5:457-463. Bhat, N. K., Thompson, C. B., Lindsten, T., June, C. H., Fujiwara, S., Koizumi, S., Fisher, R. J., and Papas, T. S. (1990). Reciprocal expression of human ETS1 and ETS2 genes during T-cell activation: regulatory role for the protooncogene ETS1. Proc. Natl. Acad. Sci. USA 87:3723-3727.  Biondi, R. M. and Nebreda, A. R. (2003) Signalling specificity of Ser/Thr protein kinases through docking site-mediated interactions. Biochem. J., 372:1-13. Boddy, M. N., Howe, K., Etkin, L. D., Solomon, E., and Freemont, P. S. (1996). PIC 1, a novel ubiquitin-like protein which interacts with the PML component of a multiprotein complex that is disrupted in acute promyelocytic leukemia. Oncogene 13:971-982.  Bodenhausen, G. and Ruben, D. J. (1980). Natural abundance nitrogen-15 NMR by enhanced Heteronuclear spectroscopy. Chem. Phys. Lett. 69:185-189. Bories, J. C , Willerford, D. M., Grevin, D., Davidson, L., Camus, A., Martin, P., Stehelin, D., and Alt, F. W. (1995). Increased T-cell apoptosis and terminal B-cell differentiation induced by inactivation of the Ets-1 proto-oncogene. Nature 377:635-638. Boulton, T. G., Nye, S. H., Robbins, D. J., Ip, N., Y., Radziejewska, E., Morgenbesser, S. D., DePinho, R. A., Panayotatos, N., Cobb, M. H., and Yancopoulos, G. D. (1991). ERKs: a family of protein-serine/threonine kinases that are activated and tyrosine phosphorylated in response to insulin and NGF. Cell 65:663-675. Breitkreutz, A., and Tyers, M. (2002). MAPK signaling specificity: it takes two to tango. Trends Cell Biol. 12:254-257.  Brown, L. A., Rodaway, A. R., Schilling, T. F., Jowett, T., Ingham, P. W., Patient, R. K., and Sharrocks, A. D. (2000). Insights into early vasculogenesis revealed by expression of the ETS-domain transcription factor Fli-1 in wild-type and mutant zebrafish embryos. Mech. Dev. 90:237-252. Brunner, D., Ducker, K., Oellers, N., Hafan, E., Scholz, H., and Klambt, C. (1994). The ETS domain protein pointed-P2 is a target of MAP kinase in the sevenless signal transduction pathway. Nature 370:386-389. Canagarajah, B. J., Khokhlatchev, A., Cobb, M. H., and Goldsmith, E. J. (1997). Activation mechanism of the MAP kinase ERK2 by dual phosphorylation. Cell 179  References  90:859-869. Carrere, S., Verger, A., Flourens, A., Stehelin, D., and Duterque-Coquillard, M. (1998). Erg proteins, transcription factors of the Ets family, form homo, heterodimers and ternary complexes via two distinct domains. Oncogene 16:3261-3268. Chakrabarti, S. R., and Nucifora, G. (1999). The leukemia-associated gene TEL encodes a transcription repressor which associates with SMRT and mSin3A. Biochem. Biophys. Res. Commun. 264:871-877.  Chakrabarti, S. R., Sood, R., Ganguly, S., Bohlander, S., Shen, Z., and Nucifora, G. (1999). Modulation of TEL transcription activity by interaction with the ubiquitinconjugating enzyme UBC9. Proc. Natl. Acad. Sci. USA 96:7467-7472. Chakrabarti, S. R., Sood, R., Nandi, S., and Nucifora, G. (2000). Posttranslational modification of TEL and TEL/AML1 by SUMO-1 and cell- cycle-dependent assembly into nuclear bodies. Proc. Natl. Acad. Sci. USA 97:13281-13285. Chang, C. I., Xu, B. E., Akella, R., Cobb, M. H., and Goldsmith, E. J. (2002). Crystal structures of MAP kinase p38 complexed to the docking sites on its nuclear substrate MEF2A and activator MKK3b. Moi. Cell 9:1241-1249. Chen, Y. H., Yang, J. T., and Martinez, H. M. (1972). Determination of the secondary structures of proteins by circular dichroism and optical rotatory dispersion, Biochemistry 11:4120-4131. Chen, Z., Gibson, T. B., Robinson, F., Silvestro, L., Pearson, G., Xu, B.-e., Wright, A., Vanderbilt, C , and Cobb, M. H. (2001). MAP kinases. Chem. Rev. 101:24492476. Chi, S.-W., Ayed, A., and Arrowsmith, C. H. (1999). Solution structure of a conserved C-terminal domain of P73 with structural homology to the SAM domain. EMBO J. 18:4438-4445. Chinenov, Y., Henzl, M., and Martin, M. E. (2000). The a and (3 subunits of the GAbinding protein form a stable heterodimer in solution. J. Biol. Chem. 275:77497756. Cho, H. S., Liu, C. W., Damberger, F. F., Pelton, J. G., Nelson, H. C. M., and Wemmer, D. E. (1996). Yeast heat shock transcription factor N-terminal activation domains are unstructured as probed by heteronuclear NMR spectroscopy. Protein Sci. 5:262-269. Christendat, D., Yee, A., Dharamsi, A., Kluger, Y., Savchenko, A., Cort, J. R., Booth, V., Mackereth, C. D., Saridakis, V., Ekiel, I., Kozlov, G., Maxwell, K. L., Wu, N., Mcintosh, L. P., Gehring, K., Kennedy, M. A., Davidson, A. R., Pai, E. F., Gerstein, M., Edwards, A. M., and Arrowsmith, C. H. (2000). Structural 180  References  proteomics of an archaeon. Nat. Struct. Biol. 7:903-909  Clore, G. M., Gronenborn, A. M., and Bax, A. (1998). A robust method for determining the magnitude of the fully asymmetric alignment tensor of oriented macromolecules in the absence of structural information. J. Magn. Reson. 133:216-221. 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 13:289-302. Cowley, D. O., and Graves, B. J. (2000). Phosphorylation represses Ets-1 DNA binding by reinforcing autoinhibition. Genes Dev. 14:366-376. Cramer, P., Bushnell, D. A., Fu, J., Gnatt, A. L, Maier-Davis, B., Thompson, N. E., Burgess, R. R., Edwards, A. M., David, P. R., and Kornberg, R. D. (2000). Architecture of RNA polymerase II and implications for the transcription mechanism. Science 288:640-649. Delattre, O., Zucma, J., Plougastel, B., Desmaze, C., Melot, T., Peter, M., Kovar, H., Joubert, I., de Jong, P., Rouleau, G., et al. (1992). Gene fusion with an ETS DNA-binding domain caused by chromosome translocation in human tumours. Nature 359:162-165. Deramuadt, T. B., Remy, P., and Stiegler, P. (2001). Identification of interaction partners for two closely-related members of the ETS protein family, FLI and ERG. Gene 273:169-177. Desterro, J. M., Rodriguez, M. S., and Hay, R. T. (1998). SUMO-1 modification of kBa inhibits N F - K B activation. Moi. Cell 2:233-239. Desterro, J. M., Rodriguez, M. S., Kemp, G. S., and Hay, R. T. (1999). Identification of the enzyme required for activation of the small ubiquitin-like protein SUMO-1. J. Biol. Chem. 274:10618-10624. Dittmer, J., and Nordheim, A. (1998). Ets transcription factors and human disease. Biochim. Biophys. Acta 1377:F1-F11. Donaldson, L. W., Petersen, J. M., Graves, B. J., and Mcintosh, L. P. (1996). Solution structure of the ETS domain from murine Ets-1: a winged helix- turn-helix DNA binding motif. EMBOJ. 15:125-34. Doolittle, R. F. (1995). The multiplicity of domains in proteins. Annu. Rev. Biochem. 64:287-314. Doolittle, R. F., and Bork, P. (1993). Evolutionarily mobile modules in proteins. Sci. Am. 269:50-56. 181  References  Dunker, A. K., Lawson, J. D., Brown, C. J., Williams, R. M., Romero, P., Oh, J. S., Oldfield, C. J., Campen, A. M., Ratliff, C. M., Hipps, K. W., et al. (2001). Intrinsically disordered protein. J. Moi. Graph. Model. 19:26-59. Duterque-Coquillard, M., Niel, C , Plaza, S., and Stehelin, D. (1993). New human erg isoforms generated by alternative splicing are transcriptional activators, Oncogene 8:1865-1873. Dyson, H. J. and Wright, P. E. (2002). Coupling of folding and binding for unstructured proteins. Curr. Opin. Struct. Biol. 12:54-60.  Elion, E. A. (2000). Pheromone response, mating and cell biology. Curr. Opin. Microbiol. 3:573-581.  Everett, R. D., Lomonte, P., Sternsdorf, T., van Driel, R., and Orr, A. (1999). Cell cycle regulation of PML modification and ND10 composition. J. Cell Sci. 112:4581-8. Fantz, D. A., Jacobs, D., Glossip, D., and Kornfeld, K. (2001). Docking sites on substrate proteins direct extracellular signal-regulated kinase to phosphorylate specific residues. J. Biol. Chem. 276:27256-27265. Farrow, N. A., Muhandiram, R., Singer, A. U., Pascal, S. M., Kay, C. M., Gish, G., Shoelson, S. E., Pawson, T., Forman-Kay, J. D., and Kay, L. E. (1994). Backbone dynamics of a free and phosphopeptide-complexed Src homology 2 domain studied by 15N NMR relaxation. Biochemistry 33:5984-6003. Fenrick, R., Amann, J. M., Lutterbach, B., Wang, L, Westendorf, J. J., Downing, J., and Hiebert, S. W. (1999). Moi. Cell. Biol. 19:6566-6574. Fenrick, R., Wang, L, Nip, J., Amann, J. M., Rooney, R. J., Walker-Daniels, J., Crawford, H. C , Hulboy, D. L, Kinch, M. S., Matrisian, L. M., and Hiebert, S. W. (2000). Moi. Cell. Biol. 20:5828-5839. Flory, E., Hoffmeyer, A., Smola, U., Rapp, U. R., and Bruder, J. T. (1996). Raf-1 kinase targets GA-binding protein in transcriptional regulation of the human immunodeficiency virus type 1 promoter. J. Virol. 70:2260-2268. Fromm, L, and Burden, S. J. (1998). Genes Dev. 12:3074-3083. Fromm, L, and Burden, S. J. (2001). Neuregulin-1-stimulated phosphorylation of GABP in skeletal muscle cells. Biochemistry 40:5306-5312. Goetz, T. L, Gu, T. L, Speck, N. A., and Graves, B. J. (2000). Auto-inhibition of Ets-1 is counteracted by DNA binding cooperativity with core-binding factor alpha2. Moi. Cell. Biol. 20:81-90.  182  References  Golub, T. R., Barker, G. F., Lovett, M., and Gilliland, D. G. (1994). Fusion of PDGF receptor p to a novel ets-like gene, tel, in chronic myelomonocytic leukemia with t(5;12) chromosomal translocation. Cell 77:307-316. Golub, T. R., Goga, A., Barker, G. F., Afar, D. E., McLaughlin, J., Bohlander, S. K., Rowley, J. D., Witte, O. N., and Gilliland, D. G. (1996). Oligomerization of the ABL tyrosine kinase by the Ets protein TEL in human leukemia. Moi. Cell. Biol. 16:4107-4116. Goodrich, J. A., Cutler, G., and Tijan, R. (1996). Contacts in context: promoter specificity and macromolecular interactions in transcription. Cell 84:825-830. Gorlach, M., Wittekind, M., Beckman, R. A., Mueller, L., and Dreyfuss, G. (1992) Interaction of the RNA-binding domain of the hnRNP C proteins with RNA. EMBO J. 11:3289-3295. Gostissa, M., Hengstermann, A., Fogal, V., Sandy, P., Schwarz, S. E., Scheffner, M. and Del Sal, G. (1999). Activation of p53 by conjugation to the ubiquitin-like protein SUMO-1. EMBO J. 18:6462-6471. Graves, B. J., and Petersen, J. M. (1998). Specificity within the ets family of transcription factors. In Advances in cancer research, G. V. Woude, and G. Klein, eds. (San Diego, CA, Academic Press), pp. 1-55. Gronwald, W., Huber, F., Grunewald, P., Sporner, M., Wohlgemuth, S., Herrmann, C , and Kalbitzer, H. R. (2001). Solution structure of the Ras binding domain of the protein kinase Byr2 from Schizosaccharomyces pombe. Structure (Camb) 9:1029-41. Grzesiek, S. and Bax, A. (1993a). The importance of not saturating H20 in protein NMR - application to sensitivity enhancement and NOE measurements. J. Am. Chem. Soc. 115:12593-12594. Grzesiek, S., Anglister, J., and Bax, A. (1993b). Correlation of backbone amide and aliphatic side-chain resonances in C-13/N-15-enriched proteins by isotropic mixing of C-13 magnetization. J. Magn. Reson. Ser. B 101:114-119. Guidez, F., Petrie, K., Ford, A. M., Lu, H., Bennett, C. A., MacGregor, A., Hanemann, J., Ito, Y., Ghysdael, J., Greaves, M., Wiedemann, L. M., and Zelant, A. (2000). Recruitment of the nuclear receptor corepressor N-CoR by the TEL moiety of the childhood leukemia-associated TEL-AML1 oncoprotein. Stood 96:2557-2561. Gustin, M. C , Albertyn, J., Alexander, M., and Davenport, K. (1998). MAP kinase  pathways in the yeast Saccharomyces cerevisiae. Microbiol. Moi. Biol. Rev.  62:1264-1300.  Hahn, S. L, Wasylyk, B., Criqui-Filipe, P., and Criqui, P. (1997). Modulation of ETS-1 183  References  transcriptional activity by huUBC9, a ubiquitin- conjugating enzyme. Oncogene 15:1489-95. Hanks, S., Quinn, A., and Hunter, T. (1988). The protein kinase family: conserved features and deduced phylogeny of the catalytic domains. Science 241:42-52. Harbury, P. B., Kim, P. S., and Alber, T. (1994). Crystal structure of an isoleucinezipper trimer. Nature 371:80-83. Harbury, P. B., Zhang, T., Kim, P. S., and Alber, T. (1993). A switch between two-, three-, and four-stranded coil coils in GCN4 leucine zipper mutants. Science 262:1401-1407. Hart, A., Melet, F., Grossfeld, P., Chien, K., Jones, C , Tunnacliffe, A., Favier, R., and Bernstein, A. (2000). Fli-1 is required for murine vascular and megakaryocytic development and is hemizygously deleted in patients with thrombocytopenia. Immunity 13:167-177.  Hewett, P. W., Nishi, K., Daft, E. L, and Murray, J. C. (2001). Selective expression of erg isoforms in human endothelial cells. Int. J. Biochem. Cell. Biol. 33:347-355. Hill, J. M., Vaidyanathan, H., Ramos, J. W., Ginsberg, M. H., and Werner, M. H. (2002). Recognition of ERK MAP kinase by PEA-15 reveals a common docking site within the death domain and death effector domain. EMBO J. 21:6494-6504. Ho, J. M., Beattie, B. K., Squire, J. A., Frank, D. A., and Barber, D. L. (1999). Fusion of the ets transcription factor TEL to Jak2 results in constitutive Jak-Stat signaling. Blood 93:4354-4364.  Hochstrasser, M. (2001). Sp-ring for sumo: new functions bloom for a ubiquitin-like protein. Cell 107:5-8.  Hoffmeyer, A., Avots, A., Flory, E., Weber, C. K., Serfling, E., and Rapp, U. R. (1998). The GABP-responsive element of the interleukin-2 enhancer is regulated by JNK/SAPK-activating pathways in T lymphocytes. J. Biol. Chem. 273:1011210119. Holland, P. M., and Cooper, J. A. (1999). Protein modification: docking sites for kinases. Curr. Biol. 9:R329-R331. •  Hung, A. T. and Sheng, M. (2002). PDZ domains: structural modules for protein complex assembly. J. Biol. Chem. 277:5699-5702. Ikura, M., Kay, L. E., and Bax, A. (1990). A novel approach for sequential assignment of 1H, 13C, and 15N spectra of proteins: Heteronuclear triple-resonance threedimensional NMR spectroscopy. Application to calmodulin. Biochemistry 29:4659-4667. 184  References  Iwamoto, M., Higushi, Y., Koyama, E., Enomoto-lwamoto, M., Kurisu, K., Yeh, H., Abrams, W. R., Rosenbloom, J., and Pacifici, M. (2000). J. Cell Biol. 150:27-39. Jacobs, D., Glossip, D., Xing, H., Muslin, A. J., and Kornfeld, K. (1999). Multiple docking sites on substrate proteins form a modular system that mediates recognition by ERK MAP kinase. Genes Dev. 13:163-175. Johnson, E. S., and Gupta, A. A. (2001). An E3-like factor that promotes SUMO conjugation to the yeast septins. Cell 106:735-44. Jousset, C., Carron, C., Boureux, A., Quang, C. T., Oury, C., Dusanter-Fourt, I., Charon, M., Levin, J., Bernard, O., and Ghysdael, J. (1997). A domain of TEL conserved in a subset of ETS proteins defines a specific oligomerization interface essential to the mitogenic properties of the TEL-PDGFRb oncoprotein. EMBO J. 16:69-82. Kabsch, W. (1976). A solution for the best rotation to relate two sets of vectors. Acta Crystallogr. A32:922-923.  Kadonaga, J. T. (1998). Eukaryotic transcription: an interlaced network of transcription factors and chromatin-modifying machines. Cell 92:307-313. Kagey, M. H., Melhuish, T. A., and Wotton, D. (2003). The Polycomb protein Pc2 is a SUMO E3. Cell 113:127-137. Kanelis, V., Donaldson, L. W., Muhandiram, D. R., Rotin, D., Forman-Kay, J. D., and Kay, L. E. (2000). Sequential assignment of proline-rich regions in proteins: application to modular binding domain complexes. J. Biomol. NMR 16:253-259. Kar, S., Sakaguchi, K., Shimohigashi, Y., Samaddar, S., Banerjee, R., Basu, G., Swaminathan, V., Kundu, T. K., and Roy, S. (2002). Effect of phosphorylation on the structure and fold of transactivation domain of p53. J. Biol. Chem. 277:1557915585. Kasten, M., and Giordano, A. (2001). Cdk10, a Cdc2-related kinase, associates with the Ets2 transcription factor and modulates its transactivation activity. Oncogene 2:1832-1838. Kay, L. E. (1993). Pulsed-field gradient-enhanced three-dimensional NMR experiment for correlating 13Ca/p, 13C, and 1Ha chemical shifts in uniformly carbon-13labeled proteins dissolved in water. J. Am. Chem. Soc. 115:2055-2057. Kay, L. E., Keifer, P., and Saarinen, T. (1992). Pure absorption gradient enhanced Heteronuclear single quantum correlation spectroscopy with improved sensitivity. J. Am. Chem. Soc. 114:10663-10665.  185  References  Khokhlatchev, A. V., Canagarajah, B., Wilsbacher, J., Robinson, M., Atkinson, M., Goldsmith, E., and Cobb, M. H. (1998). Phosphorylation of the MAP kinase ERK2 promotes its homodimerization and nuclear translocation. Cell 93:605-15. Kim, C. A., Gingery, M., Pilpa, R. M., and Bowie, J. U. (2002a). The SAM domain of polyhomeotic forms a helical polymer. Nat. Struct. Biol. 9:453-457. Kim, C. A., Phillips, M. A., Kim, W., Gingery, M., Tran, H. H., Robinson, M. A., Faham, S., and Bowie, J. U. (2001a). Polymerization of the SAM domain of TEL in leukemogenesis and transcriptional repression. EMBO J. 20:4173-4182. Kim, K. I., Baek, S. H., and Chung, C. H. (2002b). Versatile protein tag, SUMO: its enzymology and biological function. J. Cell Physiol. 191:257-68. Kim, S., Cullis, D. N., Feig, L. A., and Baleja, J. D. (2001b). Solution structure of the repsl EH domain and characterization of its binding to NPF target sequences. Biochemistry 40:6776-6785.  Kirsh, O., Seeler, J. S., Pichler, A., Gast, A., Muller, S., Miska, E., Mathieu, M., HarelBellan, A., Kouzarides, T., Melchior, F., and Dejean, A. (2002). The SUMO E3 ligase RanBP2 promotes modification of the HDAC4 deacetylase. EMBO J. 21:2682-2691. Klambt, C. (1993). The Drosophila gene pointed encodes two ETS-like proteins which are involved in the development of the midline glial cells. Development 117:163176. Knezevich, S. R., McFadden, D. E., Tao, W., Lim, J. F., and Sorensen, P. H. (1998). A novel ETV6-NTRK3 gene fusion in congenital fibrosarcoma. Nat. Genet. 18:184187. Kodandapani, R., Pio, F., Ni, C. Z., Piccialli, G., Klemsz, M., McKercher, S., Maki, R. A., and Ely, K. R. (1996). A new pattern for helix-turn-helix recognitions revealed by the PU.1 ETS-domain DNA complex. Nature 380:456-460. Kornberg, R. D. and Lorch, Y. (1999). Twenty-five years of the nucleosome, fundamental particle of the eukaryotic chromosome. Cell 98:285-294. Kotaja, N., Karvonen, U., Janne, O. A., and Palvimo, J. J. (2002). PIAS proteins modulate transcription factors by functioning as SUMO-1 ligases. Moi. Cell. Biol. 22:5222-5234. Kovacs, H., O'Donoghue, S. I., Hoppe, H.-J., Comfort, D., Reid, K. B. M., Campbell, I. D., and Nilges, M. (2002). Solution structure of the coiled-coil trimerization domain from lung surfactant protein D. J. Biomol. NMR 24:89-102. Kraulis, P. J. (1991). MOLSCRIPT: a program to produce both detailed and schematic 186  References  plots of protein structures. J. Appl. C7yste//ogr.24:945-949. Kullander, K., Mather, N. K., Diella, F., Dottori, M., Boyd, A. W., and Klein, R. (2001). Kinase-dependent and kinase-independent functions of EphA4 receptors in major axon tract formation in vivo. Neuron 29:73-84. Kuboniwa, H., Grzesiek, S., Delaglio, F., and Bax, A. (1994). Measurement of H -H J couplings in calium-free calmodulin using 2D and 3D water-flip-back methods. J. N  a  Biomol. NMR 4:871-878.  Kwiatkowski, B. A., Bastian, L. S., Bauer Jr., T. R., Tsai, S., Zielinska-Kwiatkowska, A. G., and Hickstein, D. D. (1998). The ets family member Tel binds to Fli-1 oncoprotein and inhibits its transcriptional activity. J. Biol. Chem. 273:1752517530. Kyba, M., and Brock, H. W. (1998). The SAM domain of polyhomeotic, RAE28, and Scm mediates specific interactions through conserved residues. Dev. Genet. 22:74-84. Lacronique, V., Boureux, A., Valle, V. D., Poirel, H., Quang, C. T., Mauchauffe, M., Berthou, C., Lessard, M., Berger, R., Ghysdael, J., and Bernard, O. A. (1997). A TEL-JAK2 fusion protein with constitutive kinase activity in human leukemia, Science 278:1309-1312.  Lai, Z. C., and Rubin, G. M. (1992). Negative control of photoreceptor development in Drosophila by the product of the yan gene, an ETS domain protein. Cell 70:609620. Landschulz, W. H., Johnson, P. F., and McKnight, S. L. (1988). The leucine zipper: a hypothetical structure common to a new class of DNA binding proteins. Science 240:1759-1764. Lelievre, E., Lionneton, F., Soncin, F., and Vandenbunder, B. (2001). The Ets family contains transcriptional activators and repressors involved in angiogenesis. Int. J. Biochem. Cell. Biol. 33:391-407. Lee, H., Mok, K. H., Muhindiram, 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 activation domain of human p53. J. Biol. Chem. 275:2942629432. LeMarco, K., Thompson, C. C , Byers, B. P., Walton, E. M., and McKnight, S. L. (1991). Identification of ets- and notch-related subunits in GA binding protein. Science 253:789-792.  Li, H. and Chen, J. D. (2000). PML and the oncogenic nuclear domains in regulating transcriptional repression. Curr. Opin. Cell Biol. 12:641-644. 187  References  Li, R., Pei, H., Watson, D. K., and Papas, T. S. (2000). EAP1/Daxx interacts with ETS1 and represses transcriptional activation of ETS1 target genes. Oncogene 19:745753. Liang, H., Mao, X., Olejniczak, E. T., Nettesheim, D. G., Yu, L., Meadows, R. P., Thompson, C. B., and Fesik, S. W. (1994). Solution structure of the ets domain of Fli-1 when bound to DNA. Nat. Struct. Biol. 1:871-875.  Lin, D., Tatham, M. H., Yu, B., Kim, S., Hay, R. T., and Chen, Y. (2002). Identification of a substrate recognition site on Ubc9. J. Biol. Chem. 277:21740-21748. Linge, J. P. and Nilges, M. (1999). Influence of non-bonded parameters on the quality of NMR structures: a new force-field for NMR calculation. J. Biomol. NMR 13:5159. Linge, J. P., O'Donoghue, S. I., and Nilges, M. (2001). Assigning ambiguous NOEs with ARIA. Methods Enzymol. 339:71-90. Lipari, G. and Szabo. A. (1982). Model-free approach to the interpretation of nuclear magnetic resonance relaxation in macromolecules. 1. Theory and range of validity. J. Am. Chem. Soc. 104:4546-4559. Liu, Q., Jin, C , Liao, X., Shen, Z., Chen, D. J., and Chen, Y. (1999a). The binding interface between an E2 (UBC9) and a ubiquitin homologue (UBL1). J. Biol. Chem. 274:16979-16987. Liu, Q., Yuan, Y. C , Shen, B., Chen, D. J., and Chen, Y. (1999b). Conformational flexibility of a ubiquitin conjugation enzyme (E2). Biochemistry 38:1415-1425. Logan, T. M., Olejniczak, E. T., Xu, E. T., and Fesik, S. W. (1992). Side chain and backbone assignments in isotopically labeled proteins from two Heteronuclear triple resonance experiments. FEBS Lett. 314:413-418. Lovejoy, B., Choe, S., Cascio, D., McRorie, D. K., DeGrado, W. F., and Eisenberg, D. (1993). Crystal structure of a synthetic triple-stranded alpha-helical bundle. Science 259:1288-1293.  Lupas, A., van Dyke, M., and Stock, J. (1991). Predicting coiled coils from protein sequences. Science 252:1162-1164. MacKay, J. P., Shaw, G. L., and King, G. F. (1996). Backbone dynamics of the c-Jun leucine zipper: 15N NMR relaxation studies. Biochemistry 35: Mackereth, C. D., Arrowsmith, C. H., Edwards, A. M., and Mcintosh, L. P. (2000). Zincbundle structure of the essential RNA polymerase subunit RPB10 from  Methanobacterium thermoautotrophicum. Proc. Natl. Acad. Sci. USA 97:6316188  References  6321. Mackereth, C. D., Scharpf, M., Gentile, L. N., and Mcintosh, L. P. (2002). Chemical shift and secondary structure conservation of the PNT/SAM domains from the Ets family of transcription factors. J. Biomol. NMR 24:71-72. Mahajan, R., Delphin, C., Guan, T., Gerace, L, and Melchior, F. (1997). A small ubiquitin-related polypeptide involved in targeting RanGAPI to nuclear pore complex protein RanBP2. Cell 88:97-107. Marion, D., Kay, L. E., Sparks, S. W., Torcia, D. A., and Bax, A. (1989a). J. Am. Chem. Soc. 3-Dimensional Heteronuclear NMR of N-15-labeled proteins. 111:15151517. Marion, D., Driscoll, P., Kay, L, Wingfield, P., Bax, A., Gronenborn, A., and Clore, G. (1989b). Overcoming the overlap problem in the assignment of H-1-NMR spectra of larger proteins by use of 3-dimensional Heteronuclear H-1-N-15 HartmanHahn multiple quantum coherence and nuclear overhauser multiple quantum coherence spectroscopy - application to interleukin-1-beta. Biochemistry 28:6150-6156. Maroulakou, I. G., and Bowe, D. B. (2000). Expression and function of Ets transcription factors in mammalian development: a regulatory network. Oncogene 19:6432-6442. Matunis, M. J., Coutavas, E., Blobel, G. (1996). A novel ubiquitin-like modification modulates the partitioning of the Ran-GTPase-activating protein RanGAPI between the cytosol and the nuclear pore complex. J. Cell Biol. 135:1457-1470. Maul, G. G., Negorev, D., Bell, P., and Ishov, A. M. (2000). Properties and assembly mechanisms of ND10, PML bodies, or PODs. J. Struct. Biol. 129:278-287. Mavrothalassitis, G., and Ghysdael, J. (2000). Proteins of the ETS family with transcriptional repressor activity. Oncogene 19:6524-6532. Mcintosh, L. P., Brun, E., and Kay, L. E. (1997). Stereospecific assignment of the NH2 resonances from the primary amides of asparagine and glutamine side chains in isotopically labeled proteins. J. Biomol. NMR 9:306-312. McLaughlin, F., Ludbrook, V. J., Cox, J., von Carlowitz, I., Brown, S., and Randi, A. M. (2001). Combined genomic and antisense analysis reveals that the transcription factor Erg is implicated in endothelial cell differentiation. Blood 98:3332-3339. McLaughlin, F., Ludbrook, V. J., Kola, I., Campbell, C. J., and Randi, A. M. (1999). Characterisation of the tumor necrosis factor (TNF)-a response elements in the human ICAM-2 promoter. J. Cell Sci. 112:4695-4703. 189  References  McLean, T. W., Ringold, S., Neuberg, D., Stegmaier, K., Tantravahi, R., Ritz, J., Koeffler, H. P., Takeuchi, S., Janssen, J. W., Seriu, T., et al. (1996). TEL-AML-1 dimerizes and is associated with a favorable outcome in childhood acute lymphoblastic leukemia. Blood 88:4252-4258. Melet, F., Motro, B., Rossi, D. J., Zhang, L., and Bernstein, A. (1996). Generation of a novel Fli-1 protein by gene targeting leads to a defect in thymus development and a delay in Friend virus-induced erythroleukemia. Moi. Cell. Biol. 16:27082718. Merritt, E. A. and Bacon, D. J. (1997). Raster3D: photorealistic molecular graphics. Methods Enzymol. 277:505-524.  Mo, Y., Vaessen, B., Johnston, K., and Marmorstein, R. (1998). Structures of SAP-1 bound to DNA targets from the E74 and c-fos promoters: insights into DNA sequence discrimination by Ets proteins. Moi. Cell 2, 201-212. Mo, Y., Vaessen, B., Johnston, K., and Marmorstein, R. (2000). Structure of the elk-1DNA complex reveals how DNA-distal residues affect ETS domain recognition of DNA. Nat. Struct. Biol. 7:292-297.  Montelione, G. T., Lyons, B. A., Emerson, S. D., and Tashiro, M. (1992). An efficient triple resonance experiment using C-13 isotropic mixing for determining sequence-specific resonance assignments of isotopically-enriched proteins. J. Am. Chem. Soc. 114:10974-10975.  Muhandiram, D. and Kay, L. (1994). Gradient-enriched triple-resonance 3-dimensional NMR experiments with improved sensitivity. J. Magn. Reson. Ser. B 103:203216. Muller, S., Matunis, M. J., and Dejean, A. (1998). Conjugation with the ubiquitin-related modifier SUMO-1 regulates the partitioning of PML within the nucleus. EMBO J. 17:61-70. Muller, S., Berger, M., Lehembre, F., Seeler, J. S., Haupt, Y., and Dejean, A. (2000). C-Jun and p53 activity is modulated by SUMO-1 modification. J. Biol. Chem. 275:13321-13329. Muller, S., Hoege, C , Pyrowolakis, G., and Jentsch, S. (2001). SUMO, ubiquitin's mysterious cousin. Nat. Rev. Moi. Cell. Biol. 2:202-10. Murakami, K., Mavrothalassitis, G., Bhat, N. K., Fisher, R. J., and Papas, T. S. (1993). Human ERG-2 protein is a phosphorylated DNA-binding protein -- a distinct member of the ets family. Oncogene 8:1559-1566. Muthusamy, N., Barton, K., and Leiden, J. M. (1995). Defensive activation and survival of T cells lacking the Ets-1 transcription factor. Nature 377:639-642. 190  References  Nakagawa, K., and Yokosawa, H. (2002). PIAS3 induces SUMO-1 modification and transcriptional repression of IRF-1. FEBS Lett. 530:204-8. Nassar, N., Horn, G., Herrmann, C., Scherer, A., McCormick, F., and Wittinghofer, A. (1995). The 2.2 A crystal structure of the Ras-binding domain of the serine/threonine kinase c-Raf1 in complex with RaplA and a GTP analogue. Nature 375:554-560.  Neri, D., Szyperski, T., Otting, G., Senn, H., and Wuthrich, K. (1989). Stereospecific nuclear magnetic resonance assignments of the methyl groups of valine and leucine in the DNA-binding domain ofthe 434 repressor by biosynthetically directed fractional 13C labeling. Biochemistry 28:7510-7516. Netzer, W. J., and Hartl, F. U. (1997). Recombination of protein domains facilitated by co-translational folding in eukaryotes. Nature 388:343-349. Nicholls, A., Sharp, K., and Honig, B. (1991). Protein folding and association: insights from the interfacial and thermodynamic properties of hydrocarbons. Proteins 11:281-296 Nilges, M. (1995). Calculation of protein structures with ambiguous distance restraints. Automated assignment of ambiguous NOE crosspeaks and disulphide connectivities. J. Moi. Biol. 245:645-660. Nilges, M. and O'Donoghue, S. (1998). Ambiguous NOEs and automated NOE assignment. Prog. NMR Spect. 32:107-139. Nishida, T., and Yasuda, H. (2002). PIAS1 and PIASxalpha function as SUMO-E3 ligases toward androgen receptor and repress androgen receptor-dependent transcription. J. Biol. Chem. 277:41311-41317.  Oikawa, T., and Yamada, T. (2003). Molecular biology ofthe Ets family of transcription factors. Gene 303:11-34. Okazaki, N., Okazaki, K., Tanaka, K., and Okayama, H. (1991). The ste4+ gene, essential for sexual differentiation of Schizosaccharomyces pombe, encodes a protein with a leucine zipper motif. Nucl. Acids Res. 19:7043-7047. O'Neill, E. M., Rebay, I., Tijan, R., and Rubin, G. M. (1994). The activities of two Etsrelated transcription factors required for Drosophila eye development are modulated by the Ras/MAPK pathway. Cell 78:137-147. Ottiger, M., Delaglio, F., and Bax, A. (1998). Measurement of J and dipolar couplings from simplified two-dimensional NMR spectra. J. Magn. Reson. 131:373-378. Papadopoulos, P., Ridge, S. A., Boucher, C. A., Stocking, C , and Wiedemann, L. M. 191  References  (1995). The novel activation of ABL by fusion to an ets-related gene, TEL. Cancer Res. 55:34-38.  Parker, D., Jhala, U. S., Radhakrishnan, I., Yaffe, M. B., Reyes, C , Shulman, A. I., Cantley, L. C , Wright P. E., and Montminy, M. (1998). Analysis of an activator:coactivator complex reveals an essential role for secondary structure in transcriptional activation. Moi. Cell 2:353-359. Pascal, S. M., Muhandiram, D. R., Yamazaki, T., Kay, J. D. F., and Kay, L. (1994). Simultaneous acquisition of N-15-edited and C-13 edited NOE spectra of proteins dissolved in H20. J. Magn. Reson. 103:197-201. Pawson, T., Gish, G. D., and Nash, P. (2001). SH2 domains, interaction modules and cellular wiring. Trends Cell. Biol. 11:504-511. Pelton, J. G., Torcia, D. A., Meadow, N. D., and Roseman, S. (1993). Tautomeric states of the active-site histidines of phosphorylated and unphosphorylated IIIGIc, a signal-transducing protein from Escherichia coli, using two-dimensional Heteronuclear NMR techniques. Protein Sci. 2:543-558. Petersen, J. M., Skalicky, J. J., Donaldson, L. W., Mcintosh, L. P., Alber, T., and Graves, B. J. (1995). Modulation of transcription factor Ets-1 DNA binding: DNAinduced unfolding of an alpha helix. Science 269:1866-9. Peterson, A. J., Kyba, M., Bornemann, D., Morgan, K., Brock, H. W., and Simon, J. (1997). A domain shared by the Polycomb group proteins Scm and ph mediates heterotypic and homotypic interactions. Moi. Cell. Biol. 17:6683-6692. Pichler, A., Gast, A., Seeler, J. S., Dejean, A., and Melchior, F. (2002). The nucleoporin RanBP2 has SUM01 E3 ligase activity. Cell 108:109-20. Pichler, A., and Melchior, F. (2002). Ubiquitin-related modifier SUM01 and nucleocytoplasmic transport. Traffic 3:381-7. Poirel, H., Lopez, R. G., Lacronique, V., Delia Valle, V., Mauchauffe, M., Berger, R., Ghysdael, J., and Bernard, O. A. (2000). Characterization of a novel ETS gene, TELB, encoding a protein structurally and functionally related to TEL, Oncogene 19:4802-4806. Ponting, C. P. (1995). SAM: a novel motif in yeast sterile and Drosophila polyhomeotic proteins. Protein Sci. 4:1928-1930. Potter, M. D., Buijs, A., Kreider, B., van Rompaey, L., and Grosveld, G. C. (2000). Identification and characterization of a new human ETS-family transcription factor, TEL2, that is expressed in hematopoietic tissues and can associate with TEL1/ETV6. Blood 95:3341-3348. 192  References  Prasad, D. D., Rao, V., N., Lee, L, and Reddy, E. S. P. (1994). Differentially spliced erg-3 product functions as a transcriptional activator. Oncogene 9:669-673. Price, M. D., and Lai, Z. (1999). The yan gene is highly conserved in Drosophila and its expression suggests a complex role throughout development. Dev. Genes Evol. 209:207-217. Rad, M. R., Xu, G., and Hollenberg, C. P. (1992). STE50, a novel gene required for activation of conjugation at an early step in mating in Saccharomyces cerevisiae. Moi. Gen. Genet. 236:145-54.  Ramachander, R., Kim, C. A., Phillips, M. L., Mackereth, C. D., Thanos, C. D., Mcintosh, L. P., and Bowie, J. U. (2002). Oligomerization-dependent association of the SAM domains from Schizosaccharomyces pombe Byr2 and Ste4, J. Biol. Chem. 277:39585-39593. Rao, V., N., Papas, T. S., and Reddy, E. S. P. (1987). erg, a human ets-related gene on chromosome 21: alternative splicing, polyadenylation, and translation. Science 237:635-639.  Raouf, A., and Seth, A. (2000). Ets transcription factors and targets in osteogenesis, Oncogene 19:6455-6463. Reddy, E. S. P., and Rao, V., N. (1991). erg, an ets-related gene, codes for sequencespecific transcriptional activators. Oncogene 6:2285-2289. Reddy, E. S. P., Rao, V., N., and Papas, T. S. (1987). The erg gene: a human gene related to the ets oncogene. Proc. Natl. Acad. Sci. USA 84:6131-6135. Remy, P., and Baltzinger, M. (2000). The Ets-transcription factor family in embryonic development: lessons from the amphibian and bird. Oncogene 19:6417-6431. Rigaut, G., Shevchenko, A., Rutz, B., Wilm, M., Mann, M., and Seraphin, B. (1999). A generic protein purification method for protein complex characterization and proteome exploration. Nat. Biotechnol. 17:1030-1032. Rodriguez, M. S., Desterro, J. M., Lain, S., Midgley, C. A., Lane, D. P., and Hay, R. T. (1999). SUMO-1 modification activates the transcriptional response of p53. EMBO J. 18:6455-6461. Rodriguez, M. S., Dargemont, C., and Hay, R. T. (2001). SUMO-1 conjugation in vivo requires both a consensus modification motif and nuclear targeting. J. Biol. Chem. 276:12654-9. Rogge, R., Green, P. J., Urano, J., Horn-Saban, S., Mlodzik, M., Shilo, B. Z., Hartenstein, V., and Banerjee, U. (1995). The role of yan in mediating the choice between cell division and differentiation. Development *\21:3947-3958. 193  References  Rubnitz, J. E., Pui, C.-H., and Downing, J. R. (1999). The role of TEL fusion genes in pediatric leukemias. Leukemia 13:6-13. Rustandi, R. P., Baldisseri, D. M., and Weber, D. J. (2000). Structure of the negative regulatory domain of p53 bound to S100B(SB). Nat. Struct. Biol. 7:570-574. Sanij, E., Hatzistavrou, T., Hertzog, P., Kola, I., and Wolvetang, E. J. (2001). Ets-2 is induced by oxidative stress and sensitizes cells to H 0 -induced apoptosis: implications for Down's syndrome. Biochem. Biophys. Res. Commun. 287:10031008. 2  2  Santoro, J. and King, G. C. (1992). A constant-time 2D overbodenhausen experiment for inverse correlation of isotopically enriched species. J. Magn. Reson. 97:202207. Sapetschnig, A., Rischitor, G., Braun, H., Doll, A., Schergaut, M., Melchior, F., and Suske, G. (2002). Transcription factor Sp3 is silenced through SUMO modification by PIAS1. EMBO J. 21:5206-15. Sattler, M., Schleucher, J., and Griesinger, C. (1999). Heteronuclear multidimensional NMR experiments for the structure determination of proteins in solution employing pulsed field gradients. Prog. Nucl. Magn. Reson. Spec. 34:93-158. Schaeffer, L., Duclert, N., Huchet-Dymanus, M., and Changeux, J. P. (1998). Implication of a multisubunit Ets-related factor in synaptic expression of the nicotinic acetylcholine receptor. EMBO J. 17:3078-3090. Scheffzek, K., Grunewald, P., Wohlgemuth, S., Kabsch, W., Tu, H., Wigler, M., Wittinghofer, A., and Herrmann, C. (2001). The Ras-Byr2RBD complex: structural basis for Ras effector recognition in yeast. Structure (Camb) 9:1043-50. Schlessinger, J. (2000). Cell signaling by receptor tyrosine kinases. Cell 103:211-225. Schmidt, D., and Muller, S. (2002). Members of the PIAS family act as SUMO ligases for c-Jun and p53 and repress p53 activity. Proc. Natl. Acad. Sci. USA 99:28722877. Schneikert, J., Lutz, Y., and Wasylyk, B. (1992). Two conserved activation domains in c-Ets-1 and c-Ets-2 located in nonconserved sequences of the ets gene family. Oncogene 7:249-256.  Schubert, M., Labudde, D., Oschkinat, H., and Schmeider, P. (2002). A software tool for the prediction of xaa-pro peptide bond conformations in proteins based on 13C chemical shift statistics. J. Biomol. NMR 24:149-154. Schultz, J., Milpetz, F., Bork, P., and Ponting, C. P. (1998). SMART, a simple modular 194  References  architecture tool: identification of signalling domains. Proc. Natl. Acad. Sci. USA 95:5857-5864. Schultz, J., Ponting, C. P., Hofmann, K., and Bork, P. (1997). SAM as a protein interaction domain involved in developmental regulation. Protein Sci. 6:249-253. 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. 16:127-37. Sementchenko, V. I., and Watson, D. K. (2000). Ets target genes: past, present and future. Oncogene 19:6533-6548. Senn, H., Werner, B., Messerle, B. A., Weber, C , Traber, R., and Wuthrich, K. (1989). Stereospecific assignment of the methyl 1H NMR lines of valine and leucine in polypeptides by nonrandom 13C labeling. FEBS Lett. 249:113-118. Sharrocks, A. D. (2001). The ETS-domain transcription factor family. Nat. Rev. Moi. Cell. Biol. 2:827-837.  Sharrocks, A. D., Yang, S. H., and Galanis, A. (2000). Docking domains and substrate specificity determination for the MAP kinases. Trends Biochem. Sci. 25:448-453. Shekhtman, A., Ghose, R., Wang, D., Cole, P. A., and Cowburn, D. (2001). Novel mechanism of regulation of the non-receptor protein tyrosine kinase Csk: inights from NMR mapping studies and site-directed mutagenesis. J. Moi. Biol. 314:129138. Slupsky, C. M., Gentile, L. N., Donaldson, L. W., Mackereth, C. D., Seidel, J. J., Graves, B. J., and Mcintosh, L. P. (1998a). Structure of the Ets-1 pointed domain and mitogen-activated protein kinase phosphorylation site. Proc. Natl. Acad. Sci. USA 95:12129-12134. Slupsky, C. M., Gentile, L. N., and Mcintosh, L. P. (1998b). Assigning the NMR spectra of aromatic amino acids in proteins: analysis of two Ets pointed domains. Biochem. Cell Biol. 76:379-390. Smalla, M., Schmieder, P., Kelly, M., Ter Laak, A., Krause, G., Ball, L, Wahl, M., Bork, P., and Oschkinat, H. (1999). Solution structure of the receptor tyrosine kinase EphB2 SAM domain and identification of two distinct homotypic interaction sites. Protein Sci. 8:1954-1961.  Spyropoulos, D. D., Pharr, P. N., Lavenburg, K. R., Jackers, P., Papas, T. S., Ogawa, M., and Watson, D. K. (2000). Hemorrhage, impaired hematopoiesis, and lethality in mouse embryos carrying a targeted disruption of the Flit transcription factor, Moi. Cell. Biol. 20:5643-5652. Stapleton, D., Balan, I., Pawson, T., and Sicheri, F. (1999). The crystal structure of an 195  References  Eph receptor SAM domain reveals a mechanism for modular dimerization. Nat.  Struct. Biol. 6:44-49.  Starck, J., Doubeikovski, A., Sarrazin, S., Gonnet, C., Rao, G., Skoultchi, A., Godet, J., Dusanter-Fourt, I., and Morle, F. (1999). Spi-1/PU.1 is a positive regulator ofthe Fli-1 gene involved in inhibition of erythroid differentiation in friend erythroleukemic cell lines. Moi. Cell. Biol. 19:121-135. Sternsdorf, T., Jensen, K., and Will, H. (1997). Evidence for covalent modification of the nuclear dot-associated proteins PML and Sp100 by PIC1/SUMO-1. J. Cell Biol. 139:1621-1634.  Struhl, K. (1996). Chromatin structure and RNA polymerase II connection: implications for transcription. Cell 84:179-182. Sumarsono, S. H., Wilson, T. J., Tymms, M. J., Venter, D. J., Corrick, C. M., Kola, R., Lahoud, M. H., Papas, T. S., Seth, A., and Kola, I. (1996). Down's syndrome-like skeletal abnormalities in Ets2 transgenic mice. Nature 379:534-537. Tanoue, T., Adachi, M., Moriguchi, T., and Nishida, E. (2000). A conserved docking motif in MAP kinases common to substrates, activators and regulators. Nat. Cell Biol. 2:110-116.  Thanos, C. D., Faham, S., Goodwill, K. E., Cascio, D., Phillips, M. A., and Bowie, J. U. (1999a). Monomeric structure ofthe human EphB2 sterile a motif domain. J. Biol. Chem. 274:37301-37306.  Thanos, C. D., Goodwill, K. E., and Bowie, J. U. (1999b). Oligomeric structure ofthe human Eph receptor SAM domain. Science 283:833-836. Thompson, C. C , Brown, T. A., and McKnight, S. L. (1991). Convergence of Ets- and notch-related structural motifs in a heteromeric DNA binding complex. Science 253:762-768. Tran, H. H., Kim, C. A., Faham, S., Siddall, M.-C, and Bowie, J. U. (2002). Native interface of the SAM domain polymer of TEL. BMC Struct. Biol. 2:5. Triezenberg, S. J., LeMarco, K. L., and McKnight, S. L. (1988). Functional dissection of VP16, the trans-activator of herpes simplex virus immediate early gene expression. Genes Dev. 2:730-. Truong, A. H. L., and Ben-David, Y. (2000). The role of Fli-1 in normal cell function and malignant transformation. Oncogene 19:6482-6489. Tu, H., Barr, M., Dong, D. L., and Wigler, M. (1997). Multiple regulatory domains on the Byr2 protein kinase. Moi. Cell. Biol. 17:5876-5887. 196  References  Vlaeminck-Guillem, V., Carrere, S., Dewitte, F., Stehelin, D., Desbiens, N., and Duterque-Coquillaud, M. (2000). Mech. Dev. 91:331-335. Vuister, G. W. and Bax, A. (1992). Resolution enhancement and spectral editing of uniformly C-13-enriched proteins by homonuclear broad-band C-13 decoupling. J. Magn. Reson. 98:428-435. Vuister, G. W., Wang, A. C , and Bax, A. (1993). Measurement of 3-bond nitrogen carbon-J couplings in proteins uniformly enriched in N-15 and C-13. J. Am. Chem. Soc. 115:5334-5335.  Wai, D. H., Knezevich, S. R., Lucas, T., Jansen, B., Kay, R. J., and Sorensen, P. H. (2000). The ETV6-NTRK3 gene fusion encodes a chimeric protein tyrosine kinase that transforms NIH3T3 cells. Oncogene 19:906-915. Walunas, T. L, Wang, B., Wang, C. R., and Leiden, J. M. (2000). Cutting edge: the Ets1 transcription factor is required for the development of NK T eels in mice. J. Immunol. 164:2857-2860. Wang, H., Mcintosh, L. P., and Graves, B. J. (2002). Inhibitory module of Ets-1 allosterically regulates DNA binding through a dipole-facilitated phosphate contact. J. Biol. Chem. 277:2225-2233.  Wang, L., and Hiebert, S. W. (2001). TEL contacts multiple co-repressors and specifically assciates with histone deacetylase-3. Oncogene 20:3716-3725. Wang, Y., Xu, H. P., Riggs, M., Rodgers, L, and Wigler, M. (1991). byr2, a Schizosaccharomyces pombe gene encoding a protein kinase capable of partial suppression of the ras1 mutant phenotype. Moi. Cell. Biol. 11:3554-3563. Wasylyk, C , Schlumberger, S. E., Criqui-Filipe, P., and Wasylyk, B. (2002). Sp100 interacts with ETS-1 and stimulates its transcriptional activity. Moi. Cell. Biol. 22:2687-2702. Werner, M. H., Clore, G. M., Fisher, C. L., Fisher, R. J., Trinh, L., Shibach, J., and Gronenborn, A. M. (1997). Correction of the NMR structure of the Ets1/DNA complex. J. Biol. NMR 10:317-328. Wilsbacher, J. L., and Cobb, M. H. (2001). Bacterial expression of activated mitogenactivated protein kinases. Methods Enzymol. 332:387-400. Wishart, D. S., and Sykes, B. D. (1994). The 13C Chemical Shift Index: a simple method for the identification of protein secondary structure using 13C chemical shift data. J. Biomol. NMR 4:171-180. Wittekind, M., Reizer, J., Deutscher, J., Saier, M. H., and Klevit, R. E. (1989). Common structural changes accompany the functional inactivation of HPr by seryl 197  References  phosphorylation or by serine to aspartate substitution. Biochemistry 28:99089912. Wittekind, M. and Mueller, L. (1993). HNCACB, a high-sensitivity 3D NMR experiment to correlate amide-proton and nitrogen resonances with the alpha-carbon and beta-carbon resonances in proteins. J. Magn. Reson. Ser. B 101:201-205. Wolf, E., Kim, P. S., and Berger, B. (1997). MultiCoil: a program for predicting two- and three-stranded coiled coils. Protein Sci. 6:1179-1189. Wolvetang, E. J., Wilson, T. J., Sanij, E., Busciglio, J., Hatzistavrou, T., Seth, A., Hertzog, P. J., and Kola, I. (2003). ETS2 overexpression in transgenic models and in Down syndrome predisposes to apoptosis via the p53 pathway. Hum. Moi. Genet. 12:247-255.  Wright, P. E., and Dyson, H. J. (1999). Intrinsically unstructured proteins: re-assessing the protein structure-function paradigm. J. Moi. Biol. 293:321-331. Yamazaki, T., Forman-Kay, J. D., and Kay, L. (1993). 2-Dimensional NMR experiments for correlating C-13-beta and H-1-delta/epsilon chemical-shifts of aromatic residues in C-13-labeled proteins via scalar couplings. J. Am. Chem. Soc. 115:11054-11055. 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. Moi. Cell. Biol. 16:538-547. Yordy, J. S., and Muise-Helmericks, R. C. (2000). Signal transduction and the Ets family of transcription factors. Oncogene 19:6503-6513. Zhong, S., Muller, S., Ronchetti, S., Freemont, P. S., Dejean, A., and Pandolfi, P. P. (2000). Role of SUMO-1-modified PML in nuclear body formation. Blood 95:2748-2752.  198  Appendix 1  oo  cn  CD  u C  CO  oo  .O C5 00  00 T—  c O CO  CD  ~ .-  T  TO  4 0  CD CD  T3  a:  o  c  o  co  03 O  i l : 05  CD X CD  CN C  oo  00  co  CD  CD  CD  CD  CD  CD  CD  CD ><  5 c  o  CD -C  o  o o  O O  CN  O O CN  O O CN  O O CM  00  co  CD  CD  CO  o o  c  CD  ©  CO UJ  O m  o o  ~  CM  CD CN  CM  co oo  oo oo  in oo  CD  -t CO  CD  CM l>-  CM r*-  ;E  §  CO  co  o co  o o  o o  o o  o o  o o  o o  oo  in  m  in  in  CN o  oo  co  oo  O  o CN  o CM  o CM  o in  o  m  CN o  O  o  o  o  o  o  o  h-  r- r*-  o co  O 1"-  o o o  CN  O  o  o  o f-  >CO O  o z  in  o "3in  o ^jm  O m  < x  CD X  z X  o m  >  5^-  >-  o  CO  E  — I in  o  oo  03  >  U-  CO  O O N-  I  CD  s  oo  o  >CO  CO  o  CO  o *  CM  5  oo co  CL  rco  CD in  CN O  in CM  to  o  5  CD in CM  CM  E  CO Q.  CD  CN  CN o  CO  Q.  CD *~  m  c  r--  CD CD  CD  m  o  CD CD  i  05  O  CD CD  i  m  .s  CD T-  CM  m  CD  i-  00  (0  c  CO  CD ID  o CO  to  CO  in  o co  >,  i_  CN  o  to  OJ  •G  T _  CD  IT)  CN  co  CM  oo  X) oo O) CD  CD  r--  E  T _  co O cn  CD  CD  CD  .o. co O)  -co  co  aQ  CO iii  o z  CD  o 'i • CD  o CO  o CD  o  »  E  199  E  CO  CN  T _  o o rCD O  c  o  c  o O f-  O  CD  X  o  E  o CO  o co  CN  CN  in  o CD  o  o  T—  ro-  CD  o c  O  a  CO X I  z  m  ro-  o CD  "co CN X  o a  Appendix 1  CM 05  CM CO CO  CO  co  CO T  CD O  c  X CO CQ  X CD CQ  T3 C  T3 C CD  co  Q>  '<n CD  O)  0  CD  CD  co  O0  CM  S  CD  ~  CO  a> a> , _ ,  CO CD  ~  CO -+»«  CD  N CO  N CO  co E  co E  CO  CO  >-  CO  CM  CO  O0  oo  CM O0  CM O0  CM  T  ~  CM  _j 75  co * © t-. CD c .92 C co co  CD W  CD  a5 c CO  O0  in  CO  O0  If) CD CO  co  O  oo  co  co  to  o  o  o o co co  a  co oo  TJ-  CO  CM O  co CM  to  oo CO CD  m  CO  co  in CM CO  o  o  oo CD  oo co  o o o in  o o o in  O0  CO  .E c 0) CD  i CO  o o < o CQ o CQ  X  O O < O  CD  O  CQ X  co oo  co in  o  CM  CM  o  o  CD CN 00 OO CO  CO CM oo O) CD  co CN 00 O) CO  (0  c  CO  co  CM  5  g  a5 c co  CM  co  co  ^ CD CO  CM  CM  CO  «M § o oo r : O) OO _ co" <» co COO  CM  CO CO  > CD CO X * CO CQ  0)  c  g  T-  .  0  5  r  03^ r a> 15 CD co ^  -i o  co  CO  CM  CD  CM  in  o o o  "co CD  ™ ' »  -•o  o o o  CO <D . CD CD CD  c "  CD  co co P. — c  O  C — I o ffl x  co 0)  3  i5  •o c  TO  CO  < = 2 a s  JD  co  ,_  0  co  oo  co co  CM  TO  c  c co •o c  CD  0  CO  oo  co  00  CM CO  co  in  o 00  oo  £  oo  ?8>  co CO CO  co oo co oo  CM  CD CQ  in CM  co  CO  Q X Q  LU X UJ  O  o  O  O CD O  o o CQ CQ  X  Q  CQ  O  CQ  X  o o CO  CO  in  CM CO  — I  o o o I  g>  ID  > CO  o O H  I  O O  o  o  o o o  in  5  >-  eo  CO  CO  CO  co oo  CM  00  CM CO  CM CO  oo  o o o  o o CD  o  o  CM  00 co CO  CO  o oo fCO  CD  CO  o 00  o oo rco^-  co  o o  -a-  -tf  o  CM  o  CM  o  CM  cq CM oo oo CD  CO d o o co  co d o o CD  co CM oo O) CD  CD CM co CO co  o o  CO X I  o co  o co  200  CM  O  <  < o X  < o o  o < < X o CQ o CQ X  CM  oo  in co  CO  in  CM O)  o  in  in  oo oo  oo  oo  co  co oo  co  co  o o -3CM  d CM oo  CM O  CD CM oo CO CD  5  >  o o  CO  II  o o X X  >CO  o o  — I  o o X I  >CO  o o H  o o X I  CD  in co  CO  m  CM  5  CO  g  CO  o CQ  o < o CO  in  g CQ  O < O  Appendix 1  O  00  CN CJ> CJ) CN CN  CD O  c  2!  <  2 5 g>x•D CZ CD CD  m  CD  CO  CD ^.  CD CD co CD  „  CD CD CD ©  cz  *-  *-  i_  1  ^:  0) CD 0) CD 0? CO 00 CO •5 CD '5 CD  (5  -  CD  CD  N  CZ CJ)  00 co S oo .0.  CD  CN CJ) CJ) CN - °2 C» _= „ CD DJ x T  ? § 2 a ~ .= co -a CQ cz "co "c5 "co co co £ co^-g-o CD ai CD CJ)  CD •4^1  CD  CD  "co o CO co  ieni  O  CJ)  -o -a co co co co CD CD CJ> cj> CJ> CD CD CD CD  1  CD  —7  a.  CO  a)  CZ CZ CD CD CZ C CO >^ CoD CD CD =3 CO j 2 cz co 5 , cz CO > CD T3  O DO  CD  E  CN C  2  co  CD  CN  o  00  00  CN CO  CN  co  CO  CN CO  CN CO  CO  O O CN CO  O  CO  00  O  00  co  •a  o o CD LO CN CJ)  CN  o CN co  cx>  o CN co  1^  CN CD  X  cQ) a  Q. <  co  co  o o in o o o  d  CD CJ)  CO LO  co  00  1  1  00 CN  co LO CN  co as  co LO CN  CJ) ai CJ)  CJ) ai CJ)  o  o  CD T—  5 CD CJ)  0 0 CN  CN  CO  CJ) CO  0  •* CN 0  CN O  CN  CN O  CO CN  co CN  CD c\i  CD CN  CO CN  CO CN  CD CN  co CN  CJ) CD  CXi CD  CJ> CO  CJ) co  CJ) CD  CJ) CD  CJ) CO  •tf  CN O  00  ai  O) 0> ID  «fr  CN  00  ai a)  o> LO  00  00 CD CO  o o I  CO  o o  I  CD o CD D) cz co O) cz o  >CO LU  CD o CD CD c  co i_  CD CZ o  O  IS  00  o  00  E  00  00  CD LO  •<fr  o o  o o  o o o  o o  o o o o  CO CN  CN  00  00  00 as CD  CO  E o CN  >-  >  o  CO LU  CO UJ  z  ~  g  O  o O O O aco O aCO CO X X  a  CD  o  CD CD  co  CN  CJ> CO  O .s O CO X c o CD CD  CN  LO  oi  CN  •«* CO CJ>  3 C  C  CD  CN  s  CN  a>  c o o  co  CO co  CD  c  CJ)  00  CO ai LO  T -  CN CD  CD CO  CJ)  E E O O co CD  ti  CO  201  CO X I  z  LO  X  CJ  ro Eo co  Appendix 2 Appendix 2. Chemical shift table for Erg Residue  N  C  C  A  60.07  S  other  CP 64.63  H  120.63  C , 120.02 (7.04)  M108  115.37  175.94  55.52(4.55)  32.95 (1.98, 2.12)  C , 31.89 (2.5, 2.58); C , 17.09 (2.07)  E109  121.9 (8.38)  176.4  56.6 (4.39)  30.41 (1.93, 1.93)  C , 36.39 (2.23, 2.23)  E110  122.37 (8.36)  176.35  56.7 (4.21)  30.46(1.93,1.93)  C , 36.38 (2.21,2.21)  Kill  121.7 (8.23)  175.18  56.16(4.25)  33.1 (1.69, 1.69)  C , 24.7 (1.31, 1.31); C , 29.04 (1.61, 1.61);  6 2  Y  E  Y  Y  Y  5  C , 42.16(2.94, 2.94) E  55.91 (4.61)  30.78 (3.02, 3.09)  C , 120.06 (7.03)  123.31 (8.23)  53.01 (4.75)  32.72(1.89, 1.99)  C\ 32.13 (2.48, 2.55); C , 16.96 (2.06)  P114  138.29  61.56 (4.65)  30.82  P115  136.2  P116  134.68  176.61  63.2 (4.38)  32.07  C , 27.39(*, *); C , 50.42(*, *)  N117  117.65 (8.39)  176.05  53.3 (4.62)  38.79 (2.81,2.77)  N  M118  120.79 (8.3)  176.2  55.75 (4.51)  33.2 (2.01, 1.91)  C , 32.02 (2.35, 2.51); C , 17 (2.02)  69.92 (4.24)  C , 21.59 (1.18)  H112  120.58 (8.28)  M113  174.85  5 2  E  61.22(4.67) Y  5  6 2  , 112.821 (7.584, 6.851)  y  E  T119  114.9 (8.2)  174.69  61.91 (4.43)  T120  115.73 (8.15)  173.48  61.59 (4.38)  69.9 (4.24)  C , 21.49 (1.18)  N121  126.33 (8.06)  175.25  54.23 (4.46)  39.4 (*, 2.73)  N  E122  120.46 (8.39)  176.3  57.1 (4.2)  30.19(2.03,1.92)  C , 36.33 (2.24, 2.24)  R123  121.68 (8.25)  175.77  56.1 (4.26)  30.5 (1.72, 1.78)  C , 27.21 (1.33, 1.33); C , 4 3 . 3 7 (2.94, 2.94)  R124  122.09 (8.19)  175.61  56.01 (4.31)  31.09(1.74, 1.74)  C , 27.12 (1.57, 1.57); C , 43.38 (3.14, 3.14)  V125  121.72 (7.85)  174.97  61.57(4.03)  32.72 (1.77)  1126  126.78 (8.32)  175.18  60.83 (4.08)  37.53 (1.78)  Y 2  7 2  8 2  , 112.52 (7.45, 6.75)  y  Y  5  Y  8  C  Y L  , 20.93 (0.7); C , 20.91 (0.78)  C  Y L  , 26.96 (1.04, 1.34); C  y 2  Y 2  , 17.45 (0.75); C , 8 1  12.38 (0.71) V127  119.89 (7.61)  P128  133.99  176.45  A129  124.81 (8.47)  176.7  D130  115.07 (7.22)  P131  134.79  T132 L133  58.26 (4.18)  32.71 (1.37)  C " , 21.88 (0.48); C  63.13 (4.35)  32.24(1.87, 2.2)  C , 27.33 (1.95, *); C , 50.16 (3.61, 3.75)  1  , 18.93 (0.57)  Y 2  Y  S  54.29 (4.07)  18.96(1.38)  49.53 (4.76)  41.6 (2.39, 2.8)  177.25  61.98 (1.36)  30.48 (0.43, 0 10^  C , 27.32 (0.97, 1.19); C , 50.28 (3.76, 3.41)  107.85 (7.85)  176.51  64.81 (3.52)  68.88 (3.98)  C , 22.32 (1.14)  120.01 (7.71)  178.34  54 (4.41)  41.86 (1.76, 1.76)  C , 26.68 (1.44); C  Y  S  y2  y  6 1  , 25.67 (0.95); C , 22.18 6 2  (0.76) W134  120.02 (7.5)  178.81  55.49 (5.55)  30.65 (3.33, 3.56)  C , 122.06 (6.34); C 6 1  E 3  , 119.64 (7.15); C  125.35 (5.75); C ^ , 1 1 3 . 9 7 (7.02); C , (6.06); N 127.61 (11.42)  176.13  57 (5.06)  T136  111.52 (9.15)  177.94  65.33 (4.47)  68.35 (4.33) 41.49 (2.33,2.37) 30.69 (2.16,3.38)  D137  122.67 (7.73)  178.13  H138  121.23 (7.48)  177.48  59.3 (4.08)  ,  , 126.52(10.24)  65.95 (4.08, 4.56)  S135  57.43 (4.52)  E L  2  121.79  Q  2  C , 22.1 (1.25) y2  C  6 2  , 115.16(6.9); C  254.56(*); N  E 2  , 138.96 (7.84); N  E L  5 1  ,  , 166.32(*)  V139  118.31 (8.07)  177.25  67.51 (3.8)  32.32 (2.57)  C  R140  118.99 (7.58)  177.8  60.36 (3.86)  29.05 (1.97, 2.12)  C , 27.09 (1.51, 1.63); C , 43.17 (3.16, 3.23);  Y L  , 21.38 (1.27); C , 24.13 (1.41) y 2  Y  8  N " , 83.78(*, *); N 1  7 1 2  , 83.78(*, *);  Q141  116.96 (8.09)  179.31  59.46 (3.92)  29.3 (1.94,2.03)  C , 34.91 (2.51, 2.25); N  W142  122.71 (8.54)  177.04  62 (4.25)  27.98 (2.53, 3.35)  C , 126.68 (7.12); C  y  5 1  E 3  E 2  , 110.12(6.99, 6.71)  , 120.4 (7.8); C  1 2  ,  123.39 (7.11); C^ , 114.13 (7.78); C ^ , 121.38 2  (6.96); N L143  119.83 (8.63)  178.7  58.15 (3.24)  42.81 (0.98, 1.87)  E L  3  , 133.49(11.9)  C , 26.39 (1.75); C y  5 1  , 26.75 (0.24); C  8 2  , 24.11  (0.16) E144  116.3 (8.59)  179.75  59.98 (3.69)  29.89 (1.91,2.09)  C , 37.71 (2.56,2.12)  W145  120.55 (7.89)  177.19  61.52 (4.02)  28.31 (3.21,3.64)  C  Y  8 1  , 126.91 (7.18); C , 120.43 (7.26); C 8 3  1 2  ,  124.41 (7.04); C ^ , 114.12 (7.25); C , 121.32 2  (6.59); N  202  E L  , 129.33 (9.89)  p  Appendix 2  (Appendix 2. Continued) other  Residue  N  C  C  A146  123.82 (9.07)  179.7  54.93 (2.87)  17.37 (0.77)  V147  115.53 (8.5)  178.77  66.6 (3.36)  31.6 (2.11)  C  K148  118.69 (6.78)  178.33  58.21 (3.97)  32.51 (1.7, 1.7)  C , 24.91 (1.29, 1.39); C , 28.96 (1.59, 1.59);  a  CP  Y L  , 20.76 (0.95); C  , 22.09 (0.98)  Y 2  Y  5  C , 42.11 (2.92,2.92) E  E149  121.73 (8.29)  178.18  58.35 (3.47)  28.6 (0.15, 0.37)  C , 33.68 (6:87, 1.34)  Y150  112.11 (7.83)  176.03  57.06 (4.49)  37.96 (2.42,3.19)  C  Y  8 1  , 132.38 (6:86); C  117.95 (6.51); C G151  110.47 (7.19)  L152  118.82 (7.86)  174.72  8 2  , 132.38 (6.86); C  E L  ,  , 117.95 (6.51)  E 2  47.65 (3.81, 51.97(4.68)  41.34 (1.32, 1.74)  C , 26.54 (1.55); C Y  8 1  , 25.42 (0.61); C  8 2  , 22.56  (0.83) P153  136.25  176.64  63.47 (4.48)  32.25 (2.16, 1.93)  D154  116.3 (8.31)  175.77  54.5 (4.42)  39.77 (2.69, 2.69)  C , 26.8 (1.99, 1.99); C , 50.51 (3.88,4) R  8  V155  119.77 (8.02)  174.99  62.7 (3.97)  32.01 (1.96)  C  N156  126.74 (8.93)  176.26  51.44 (4.87)  37.1 (3.21,2.81)  N ,  1157  124.33 (8.43)  178.33  64.54 (3.69)  37.96(1.93)  C  Y L  , 20.96 (0.61); C  Y L  Y 2  , 22.74 (0.77)  109.51 (7.68, 6.94)  8 2  , 28.8 (1.12, 1.62); C  Y 2  , 18.12 (0.95); C  8 1  ,  13.34 (0.8) L158  120.45 (7.94)  180.16  57.51 (4.13)  41.14(1.86, 1.6)  C , 27.31 (1.72); C Y  8 1  , 24.73 (0.95); C  8 2  , 23.39  (0.9) L159  116.77 (7.65)  177.68  56.46(4.15)  40.96(1.43, 1.84)  C , 25.1 (1.5); C ' , 26.39 (1.15); C Y  8  8 2  , 21.64  (0.83) F160  115.88 (7.92)  176.67  57.22 (4.45)  39.21 (2.91,3.35)  C  8 1  , 132.71 (7.15); C  130.67 (6.97); C  8 2  , 132.71 (7.15); C ' , E  , 130.67 (6.97); C5, 128.87  E 2  (6.89) Q161  120.31 (7.39)  174.99  58.87 (4.28)  29.43 (2.07, 2.31)  C , 33.57 (2.32, 2.45); N  N162  114.64 (8.72)  174.14  52.95 (4.93)  38.79 (2.63,3.01)  N , 112.96 (7.57, 6.88)  1163  123.38 (7.8)  173.29  62.16(4)  38.52 (2.21)  C  Y  E 2  , 110.64 (7.48, 6.84)  8 2  Y L  , 27.51 (1.49, 0.99); C  Y 2  , 18.2 (1.09); C  8 1  ,  13.31 (0.68) D164  123.89 (7.6)  176.44  51.33 (4.58)  G165  105.84 (8.69)  175.56  48.51 (3.6,  K166  121.35(7.81)  179.18  41.53 (2.73,3.38)  4.54) 59.73 (3.82)  31.85 (1.85, 1.93)  C , 24.8 (1.44, 1.56); C , 29.27 (1.7, 1.7); C , Y  8  E  42.01 (3.02, 3.02) E167  119.07 (7.83)  179.46  58.6 (4.04)  29.67 (2.07, 2.14)  C , 35.8 (2.31,2.31)  L168  122.23 (9.02)  179.79  58.15 (4.14)  43.68 (1.72,1.99)  C , 26.44 (1.71); C  Y  Y  8 1  , 24.32 (1.29); C  5 2  , 27.87  (0.98) C169  111.17(7.96)  174.46  63.24 (4.01)  27.34 (2.67, 2.77)  K170  117.72 (7.38)  177.07  56.03 (4.38)  33.58 (1.86, 1.86)  C , 24.65 (1.53, 1.5); C , 29.45 (1.63, 1.63); Y  6  C , 42.02 (2.92, 2.92) E  M171  120.67 (7.34)  176.82  57.62 (4.41)  34.81 (2.03, 2.29)  C , 32.28 (2.42, 3); C , 15.19 (1.33)  T172  114.16(9.39)  176.28  60.03 (4.65)  72.26 (4.75)  C , 21.78 (1.38)  K173  121.54 (9.19)  178.04  61.09 (3.98)  31.91 (1.86,1.86)  C , 25.04 (1.32, 1.41); C , 29.19 (1.67, 1.6);  Y  E  Y 2  Y  8  C , 42.11(*, 2.93) E  57.51 (4.32)  40.76 (2.49, 2.55)  179.36  57.42 (4.37)  41.15 (2.66, 2.97)  178.76  63.01 (3.98)  40.8(3.01,3.21)  D174  117.22 (8.07)  178.41  D175  118.8(7.55)  F176  118.96 (8.42)  C  8 1  , 131.16 (7.32); C  131.95 (7.44); C  E 2  8 2  , 131.16 (7.32); C  E L  ,  , 131.95 (7.44); C , 129.66 ;  (7.32) 116.73 (8.45)  177.05  57.45 (4.96)  28.43 (2.16, 2.21)  C , 35.4 (2.55, 2.39); N  R178  117.01 (7.27)  177.04  58.18(4.16)  30.38 (1.85, 1.85)  C , 28.29 (1.86, 1.63); C , 43.37 (3.24, 3.24)  L179  116.91 (7.58)  176.45  55.27 (4.26)  43.67(1.58, 1.07)  C , 26.84 (1.4); C ' , 25.11 (0.08); C  Q177  Y  Y  Y  (0.43) T180  116.03 (8.18)  59.27 (4.98)  72.28 (4.2)  203  C , 18.73 (1.25) Y 2  E 2  , 110.78 (7.42, 6.89)  8  8  5 2  , 22.53  Appendix 2  (Appendix 2. Continued) Residue  N  C  C  P181  133.76  176.6  63.11 (4.47)  30.98  S182  122.29 (8.01)  174.81  61.6  62.7 (3.91,3.96)  Y183  118.85 (7.89)  177.02  59.6 (4.37)  38.16(2.87,3.14)  a  C  other  p  C  5 1  , 133.71 (7.12); C  118.64 (6.86); C , N184  117.1 (7.16)  176.86  54.42 (4.4)  38.42 (2.33,3.01)  A185  122.7 (8.24)  177.73  55.3 (3.63)  19.1 (1.43)  D186  115.44 (8.14)  179.02  57.63 (4.22)  40.52 (2.61,2.57)  1187  120.5 (7.06)  178.22  63.92 (3.56)  37.56(1.66)  6 2  , 133.71 (7.12); C  E L  ,  118.64 (6.86)  2  N , 110.79 (7.46, 6.56) 6 2  C \ 28.67 (0.95, 1.34); C R  T 2  , 17.17 (0.33); C  5 1  ,  12.26 (0.6) L188  119.86 (8.03)  178.64  58.32 (4.02)  41.68 (1.44, 1.59)  C , 26.66 (1.63); C T  5 1  , 25.32 (0.68); C , 22.17  8 1  , 24.94 (0.74); C  5 2  (0.39) L189  117.61 (8.22)  179.48  58.21 (3.81)  41.77 (1.36, 1.51)  C , 27.26 (1.11); C R  8 2  , 24.87  (0.78) S190  113.94 (7.76)  176.96  61.81 (4.21)  H191  122.88 (8.08)  177.53  61.5(4.53)  62.62 (3.96, 3.96) 31.28 (3.64,3.41)  C  L192  120.07 (8.51)  178.82  58.08 (3.85)  41.96(1.64, 1.72)  C , 26.84 (1.39); C  8 2  , 119.97 (7.03); C  R  8 1  e l  , 137.69 (7.71)  , 23.56 (0.34); C  8 2  , 25.3  (0.53) H193  116.47 (8.28)  177.5  59.54(4.14)  29.73 (3.18,3.18)  C  8 2  , 119.31 (6.91); C  210.26(*); N Y194  120.26 (7.8)  178.08  60.72 (4.27)  37.78 (3.21,3.32)  E 2  119.37(8.06)  178.94  56.97 (3.87)  42.22(1.8, 1.27)  , 138.28 (7.94); N  5 1  ,  S L  ,  , 179.59(*)  C , 133.34 (7.08); C 8 1  118.41 (6.82); C L195  E L  E 2  6 2  , 133.34 (7.08); C  , 118.41 (6.82)  C , 26.22 (1.71); C R  8 1  , 26.17 (0.86); C  8 2  , 22.68  (0.88) R196  116.01 (7.73)  176.81  57.68 (3.94)  31.03 (1.69, 1.9)  C , 27.39 (1.78, 1.32); C , 43.73 (2.95, 3.02)  E197  118.44 (7.65)  176.31  T198  117.62 (7.71)  P199  139.07  L200  124.36(8.18)  176.51  R  56.55 (4.18)  30.16(1.83, 1.98)  C , 36.36(2.12,2.12)  60.23 (4.4)  69.91 (3.92)  C , 21.2(1.12)  63.07 (4.38)  31.95  C , 27.36(*, *); C , 50.98(*, *)  52.89 (4.52)  41.61 (1.55, 1.55)  C , 26.93 (1.67); C  T 2  R  R  (0.86) P201  8  R  140.82  204  5  8 1  , 25.39 (0.9); C  8 2  , 23.49  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items