UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Structural and functional studies of GlcNAc-modified Tau Cheung, Adrienne Hoyann 2013

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

Item Metadata

Download

Media
24-ubc_2013_spring_cheung_adrienne.pdf [ 1.25MB ]
Metadata
JSON: 24-1.0071928.json
JSON-LD: 24-1.0071928-ld.json
RDF/XML (Pretty): 24-1.0071928-rdf.xml
RDF/JSON: 24-1.0071928-rdf.json
Turtle: 24-1.0071928-turtle.txt
N-Triples: 24-1.0071928-rdf-ntriples.txt
Original Record: 24-1.0071928-source.json
Full Text
24-1.0071928-fulltext.txt
Citation
24-1.0071928.ris

Full Text

Structural and Functional Studies of GlcNAc-modified Tau  by  Adrienne Hoyann Cheung  B.Sc., Queen’s University, 2010  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE  in  THE FACULTY OF GRADUATE STUDIES  (Biochemistry and Molecular Biology)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  April 2013  ©Adrienne Hoyann Cheung, 2013  Abstract O-GlcNAcylation is an abundant post-translational modification found on serine and threonine hydroxyl groups of nucleocytoplasmic proteins. O-GlcNAc transferase (OGT) and O-GlcNAcase (OGA) are the sole enzymes responsible for addition and removal of all GlcNAc moieties, respectively. Although O-GlcNAcylation is involved in many diverse cellular processes and has been linked to many diseases, the molecular mechanisms by which it can modulate the activities of proteins remain largely unknown. As such, the primary goal of my thesis is to gain a better structural and functional understanding of the consequences of O-GlcNAcylation of tau, an intrinsically disordered microtubule-binding protein. Upon hyperphosphorylation, tau aggregates, forming neurofibrillary tangles that are a hallmark of Alzheimer’s disease. In contrast, O-GlcNAc has a reciprocal relationship with phosphorylation and reduces tau aggregation. To address my goals, I used NMR spectroscopy to probe local structural and dynamic effects of O-GlcNAc on a fragment of tau spanning residues 353 to 408 which encompasses a microtubule-binding repeat and Ser400, a key O-GlcNAc acceptor. Although chemical shift perturbations were observed near the site of O-GlcNAc-modification, based on main-chain chemical shifts, 3J-coupling and NOE interactions, there were no significant local structural changes compared to the wild-type tau polypeptide. However, there was a small decrease in the nsec-psec time scale mobility on the main-chain of tau around residue 400. In order to investigate the functional impact of OGlcNAc on tau, I compared the heparin-induced aggregation of the wild-type tau peptide, the O-GlcNAc form, and three serine to aspartate “phosphomimic” mutants. Importantly, the OGlcNAc modification significantly decreased the amount of aggregation, whereas two of the phosphomimic mutants increased aggregation relative to wild-type tau. I postulate that the Oii  GlcNAc modification increases the solubility of tau, thereby stabilizing the monomer in solution, and reducing the stability of aggregates. .  iii  Table of contents Abstract ..................................................................................................................................... ii Table of contents ...................................................................................................................... iv List of figures ........................................................................................................................... vi List of tables ............................................................................................................................ vii Acknowledgements ................................................................................................................... x Chapter 1: Introduction ............................................................................................................. 1 1.1 Post-translational modifications ................................................................................................... 1 1.2 O-GlcNAcylation defines a new class of glycoproteins ............................................................... 1 1.3 GlcNAcylation enzymes: OGT and OGA .................................................................................... 4 1.4 Crosstalk between O-GlcNAcylation and phosphorylation ......................................................... 7 1.5 O-GlcNAc and Alzheimer’s disease related protein tau............................................................... 8 1.7 Thesis overview .......................................................................................................................... 12  Chapter 2: Materials and methods .......................................................................................... 15 2.1 Cloning ....................................................................................................................................... 15 2.1.1 His6-SUMO-tau353-408 ........................................................................................................... 15 2.1.2 Site-directed mutagenesis for aspartic acid mutants ............................................................ 15 2.2 Protein expression and purification ............................................................................................ 17 2.3 NMR spectroscopy ..................................................................................................................... 19 2.4 Aggregation assays ..................................................................................................................... 20  Chapter 3: NMR spectral assignments .................................................................................... 22 3.1 Assignments from main-chain nuclei ......................................................................................... 22 3.2 Cis/trans conformational isomers of tau .................................................................................... 27 3.3 Assignment of O-GlcNAc NMR signals .................................................................................... 28  Chapter 4: Structural, functional, and dynamic studies of tau ................................................ 31 4.1 Chemical shift perturbations....................................................................................................... 31 4.2 Secondary structure propensities derived from chemical shifts ................................................. 31 4.2.1 δ2D ...................................................................................................................................... 31 4.2.2 3JHN-Hα coupling.................................................................................................................... 32 4.2.3 NOESY experiments ........................................................................................................... 33 4.3 15N Relaxation............................................................................................................................. 37 4.4 Cis/trans isomerization............................................................................................................... 39 4.5 Aggregation assays ..................................................................................................................... 40 iv  Chapter 5: Discussion and conclusions................................................................................... 42 5.1 Strategies to overcome experimental challenges ........................................................................ 42 5.1.1 Optimizing protein expression and O-GlcNAcylation yield ............................................... 42 5.1.2 Bigger is better: assigning IDPs with a high field magnet .................................................. 43 5.1.3 Gleaning structural information from chemical shifts in the absence of NOESY-based distance restraints ......................................................................................................................... 44 5.1.4 15N relaxation experiments report on local dynamics .......................................................... 45 5.1.5 Optimizing the aggregation assays ...................................................................................... 45 5.2 Changes in cis/trans population ................................................................................................. 46 5.3 Aggregation of tau ...................................................................................................................... 47 5.4 The global fold of tau ................................................................................................................. 48 5.5 Multiple GlcNAcylation sites ..................................................................................................... 50 5.6 GlcNAcylation in various systems ............................................................................................. 50 5.7 Summary .................................................................................................................................... 52 5.8 Future directions ......................................................................................................................... 52  Works cited ............................................................................................................................. 55 Appendices .............................................................................................................................. 63 Appendix A: Chemical shift tables for tau353-408............................................................................... 63 Appendix B: Chemical shift table for O-GlcNAc ............................................................................ 69 Appendix C: Change in chemical shifts compared to predicted values ........................................... 70  v  List of figures Figure 1.1: Different types of glycoconjugates......................................................................... 2 Figure 1.2: Dynamic O-GlcNAcylation and phosphorylation .................................................. 4 Figure 1.3: OGT structure ......................................................................................................... 6 Figure 1.4: Domain structure of tau. ....................................................................................... 10 Figure 1.5: A model representing global folding of tau.......................................................... 11 Figure 1.6: Post-translational modifications of S396, S400, and S404 .................................. 12 Figure 3.1: Strategies for assigning main-chain resonances. .................................................. 24 Figure 3.2: Nuclei involved in NMR spectral assignments .................................................... 25 Figure 3.3: Assigned 1H-15N HSQC spectra ........................................................................... 26 Figure 3.4: Cis versus trans X-Pro peptide bond assignment ................................................. 28 Figure 3.5: Assignment of O-GlcNAc .................................................................................... 30 Figure 4.1: Change in amide chemical shifts .......................................................................... 34 Figure 4.2: Secondary structure propensity calculated by ∂2D .............................................. 35 Figure 4.3: 3JHN-HA coupling values ........................................................................................ 36 Figure 4.4: 15N T2 and heteronuclear NOE values .................................................................. 38 Figure 4.5: Aggregation assay results for tau353-408 ................................................................ 41 Figure 5.1: Two-layered polyelectrolyte brush model. ............................................................ 49 Figure 5.2: The hexosamine biosynthetic pathway................................................................. 54 Figure C.1: Change in13C’, 13Cα, and 13Cβ chemical shift values ........................................... 70 Figure C.2: Change in1HN, 15N, and 1Hα chemical shift values .............................................. 71  vi  List of tables Table 2.1: Primers used for cloning .........................................................................................16 Table 2.2: Thermocycling protocols for cloning .....................................................................16 Table 2.3: Reagents used for cloning .......................................................................................17 Table 4.1: Residues used to calculate %cis .............................................................................39 Table 4.2: %cis based on peak intensities ................................................................................40 Table A.1: Chemical shifts of 15N/13C wild-type tau353-408 ......................................................63 Table A.2: Chemical shifts of 15N/13C GlcNAc-modified tau353-408 ........................................65 Table A.3: Chemical shifts of 15N/13C S400D tau353-408 ..........................................................67 Table B.1: Chemical shifts of 15N/13C O-GlcNAc ...................................................................69  vii  Abbreviations ε280  molar absorptivity at 280 nm  2D  two-dimensional  3D  three-dimensional  A280  absorption at 280 nm  AD  Alzheimer’s disease  BSA  bovine serum albumin  CDK5  cyclin-dependent kinase 5  EPR  electron paramagnetic resonance  FPLC  fast protein liquid chromatography  FRET  fluorescence resonance energy transfer  GlcNAc  β-N-Acetylglucosamine  GSK3β  glycogen synthase kinase 3  GH  glycoside hydrolase  GT  glycosyl transferase  HAT  histone acetyltransferase  hOGA  human OGA  hOGT  human OGT  HPLC  high-performance liquid chromatography  HSQC  heteronuclear single quantum coherence  IDP  intrinsically disordered peptide  IPTG  isopropyl β-D-1-thiogalactopyranoside  LB  Luria broth  MALDI-TOF matrix-assisted laser desorption/ionization time of flight MS  mass spectrometry viii  MT  microtubules  M9  M9 minimal media  NFT  neurofibrillary tangles  Ni-NTA  nickel-nitrilotriacetic acid  NMR  nuclear magnetic resonance  NOESY  nuclear Overhauser effect spectroscopy  OGA  O-GlcNAcase  OGT  uridine diphosphate-N-acetylglucosamine:polypeptide β-Nacetylglucosaminyltransferase  O.D.600  optical density at 600 nm  PCR  polymerase chain reaction  PHF  paired helical filaments  Pin1  peptidyl-prolyl cis-trans isomerase NIMA-interacting 1  PP2A  protein phosphatase 2A  PTM  post-translational modification  RP-HPLC  reverse phase HPLC  SDS-PAGE  sodium dodecyl sulfate polyacrylamide gel electrophoresis  sWGA  succinylated wheat germ agglutinin  TFA  trifluoroacetic acid  UDP-GlcNAc uridine diphosphate N-acetylglucosamine WT  wild-type  ix  Acknowledgements The completion of this thesis would not have been possible without the efforts of many people, and to those people, I express endless gratitude. In particular, Lawrence, thank you for this opportunity, but more importantly, thank you for your guidance, hard work, and unlimited patience, I could not have hoped for a better supervisor. Dr. Mark Okon, NMR genius, and tennis superstar, I would still be sitting and staring at the magnets without your assistance and expertise. To our collaborators, Dr. David Vocadlo and Dr. Scott Yuzwa, your knowledge, enthusiasm for science, and willingness to help have been invaluable. Members of the McIntosh lab, current and past, thanks for the friendships, the science lessons, the wonderful discussions and troubleshooting sessions. Julien, I am so grateful to have you as a mentor and a friend, thank you for the laughter, the Kleenex for the tears, and the immeasurable support you have provided. To my mom, dad, and brother, although you were 3,000 km away, I would not have made it here without your continuous support and words of encouragement.  x  Chapter 1: Introduction 1.1 Post-translational modifications Our genome has approximately 25,000 genes encoding for an even larger number of proteins that must be regulated for proper cellular function1. One of the most important routes to such regulation involves post-translational modifications (PTMs) of amino acid side chains. It is estimated that there are more than 300 types of PTMs, such as methylation, acetylation, ubiquitinylation, SUMOylation, phosphorylation, and glycosylation, to name only a few2. The primary focus of this thesis lies in characterizing the effects of two PTMs, glycosylation and phosphorylation, on the neurofibrillary protein, tau. 1.2 O-GlcNAcylation defines a new class of glycoproteins There are three major classes of glycoconjugates: glycolipids, O-linked glycoproteins, and N-linked glycoproteins (Figure 1.1). In the latter two cases, saccharides are linked posttranslationally to proteins via serine/threonine hydroxyls or asparagine amide groups, respectively. When considering glycoconjugates, it is most common to think of complex branched structures that are used for mediating and modulating cell adhesion and trafficking. However, in the 1980’s, a new class of monosaccharide modification was recognized. While studying N-acetylglucosamine (GlcNAc) found on the surface of intact lymphocytes, Hart and coworkers discovered intracellular proteins modified with a simple O-linked monosaccharide β-N-acetylglucosamine (O-GlcNAc)3. Since then, mono-O-GlcNAc modified proteins have been found to be very abundant and widespread.  1  Glycoproteins Ser/Thr (O-linked)  Proteoglycans Glycolipids  Asn (N-linked)  Ser Ser  Extracellular space  Intracellular space GlcNAc GalNAc  Ser/Thr  Ser/Thr  Fucose  Galactose Glucose Mannose  O-GlcNAc glycoproteins  Xylulose  Figure 1.1: Different types of glycoconjugates in which straight-chain or branched saccharides are covalently linked to lipids or proteins.  O-GlcNAc modifications remained elusive for so many years due to the lack of tools and methods available to study this PTM. The addition of a simple O-GlcNAc to a serine or threonine side chain does not result in a significant change in size or charge, as required for detection by conventional analytical methods. In addition, the O-GlcNAc glycosidic bond is labile, and is preferentially fragmented during ionization, rendering O-GlcNAc-modified residues undetectable by regular protein mass spectrometry (MS) approaches at low concentrations4. Furthermore, there is preferential ionization for the unmodified peptide, suppressing the signal observed for corresponding O-GlcNAc-modified peptides5. However,  2  with the advent of electron-transfer dissociation MS in 2004, O-GlcNAc detection is now possible because the labile glycosidic bond is not cleaved6. Purifying O-GlcNacylated proteins is also challenging. A potential method for purification includes using succinylated wheat germ agglutinin (sWGA) affinity chromatography, since sWGA specifically binds to terminal N-acetylglucosamine. However, this method requires closely clustered GlcNAc residues for sufficient binding affinity7. Recently selective enrichment of O-GlcNAcylated peptides was achieved by using a mutant galactosyl transferase to enzymatically add azidosugars to GlcNAc. Using the chemically reactive azide group, many different tags could then be introduced5. In particular, the addition of a biotin moiety allowed for subsequent purification with streptavidin. Although this increases the detection limits for O-GlcNAc modified peptides, it chemically alters them, making subsequent structural and functional studies difficult to interpret. O-GlcNAc modifications occur on serine and threonine side chain hydroxyl groups. The addition of O-GlcNAc from the activated donor uridine diphosphate Nacetylglucosamine (UDP-GlcNAc) is catalyzed by O-linked N-acetylglucosamine transferase (OGT). The hydrolysis of O-GlcNAc is catalyzed by O-GlcNAcase (OGA). These processes occur on a time scale similar to that of protein phosphorylation (Figure 1.2a)7. Thus far, OGlcNAcylation has been documented in all metazoans, but remarkably there is only one gene that encodes for OGT and one gene that encodes for OGA in any given species7. This stands in marked contrast to the plethora of kinases and phosphatases found in any cell. O-GlcNAcylation has many diverse cellular functions. During metabolism, 2-5% of glucose enters the hexosamine biosynthetic pathway, from which UDP-GlcNAc is synthesized8. As such, the level of O-GlcNAc modifications is postulated to serve as a 3  nutrient/stress sensor that modulates signaling, transcription, and cytoskeletal functions9,10. O-GlcNAc modifications have been linked to cancer, since this PTM is found on many oncogenic proteins and tumor suppressor proteins9. Finally, GlcNAcylation levels have also been linked to neurodegenerative diseases9. Through knockout studies, OGT has been shown to be necessary for development11.  a)  b) H2O  UDP  GlcNAc  ATP  ADP  OGA  kinase  OGT  phosphatase  UDP-GlcNAc  Pi  H2O  Glucose Figure 1.2: Dynamic O-GlcNAcylation and phosphorylation. O-GlcNAcylation and phosphorylation are both reversible reactions that occur on a similar time scale. a) OGT catalyzes the reversible addition of GlcNAc to a protein from the donor UDP-GlcNAc, whereas OGA catalyzes the hydrolysis of the bound GlcNAc. UDP-GlcNAc is synthesized through the hexosamine biosynthetic pathway, which is linked to the amount of glucose available in a system7. b) Hundreds of kinases and phosphatases catalyze the reversible phosphorylation of proteins, often in competition with the same O-GlcNAcylation acceptor serine/threonine.  1.3 GlcNAcylation enzymes: OGT and OGA OGT is an essential glycosyl transferase, in the GT-41 family, composed of a catalytic C-terminal domain and an N-terminal protein-protein interaction domain that is made up of 34-residue helix-loop-helix tetratricopeptide repeats (TPRs)7,12. From the single gene that encodes OGT, there exist three splice variants of human OGT (hOGT), differing by the length of the TPRs. ncOGT is mostly found in the nucleus and cytosol, and has 13.5  4  TPRs; mOGT possesses an N-terminal mitochondrial localization sequence and has nine TPRs; and the shortest form of OGT, sOGT, has three TPRs13. Studies in rats have demonstrated stable expression of ncOGT throughout development, although the levels of expression decrease gradually throughout maturation; in contrast, sOGT levels were almost undetectable during early development and increased significantly after 15 days, suggesting that each isoform of OGT may play a different role throughout development14. Significant progress has been made towards understanding the enzymatic mechanism of OGT. In 2004, the Conti group reported the X-ray crystal structure of the first 11.5 TPRs which form an elongated superhelix resembling importin15. Seven years later, the Walker group reported the binary structures of the catalytic domain and its adjacent 4.5 TPRs in complex with UDP, as well as with a peptide known to be modified by OGT16. Together these studies yielded a model of full-length hOGT (Figure 1.3a). Although these crystal structures aided in answering some functional questions with regards to OGT, the exact molecular mechanisms by which OGT recognizes and glycosylates its substrates remains largely unknown15. Over 3000 proteins have been shown to have O-GlcNAc modifications. Of these, almost half have been reported to be modified at proline-valine-serine sites, whereas the remaining half has no identifiable acceptor motif7. Since TPRs resemble importin, a protein that binds to other proteins and transfers them into the nucleus, they have been hypothesized to play a role in establishing the specificity of OGT by serving as a docking site for target proteins (Figure 1.3b)17.  5  a)  TPR domains  b)  N  Substrate  Catalytic domain  Figure 1.3: a) OGT structure modeled from Conti’s 11.5 TPR crystal structure (left side, blue to red) with Walker’s catalytic domain and 4.5 TPR (right side, blue to yellow)15,17. b) A cartoon model representing the postulated role of the TPR domains in substrate recognition.  OGA also has two domains: an N-terminal GH-84 family glycoside hydrolase domain, and a C-terminal domain that possesses homology to histone acetyltransferases (HATs)12. There are two human splice variants, the full length version localizes to the cytosol (hOGA-L), and a shorter variant, (hOGA-S) that localizes to the nucleus. hOGA-S lacks the 6  HAT-like domain14. In rats, it was found that hOGA-L is expressed at low levels during early development and increases during development. hOGA-S is expressed more during early development than later on14. Interestingly, hOGA-S seems to exhibit much lower enzyme activity in vitro, suggesting that the HAT-like domain is required for OGA to be fully active, although, the role of this domain remains unknown13,18,19. Although OGA regulation is not completely understood, the solved structures of two close bacterial homologues of human OGA provided insights on the catalytic mechanism of this enzyme20-22. In addition, the OGA active site was found to contain a pocket below the acetamido group of the substrate, typical for GH84s, which was exploited to generate inhibitors with high specificity22. These inhibitors now provide an avenue to study the biological relevance of OGA to O-GlcNAcylation. 1.4 Crosstalk between O-GlcNAcylation and phosphorylation Crosstalk between O-GlcNAcylation and phosphorylation is extensive and has often been described as a “yin and yang” relationship23. These two reversible modifications of serine and threonine hydroxyls occur on a similar time scale (Figure 1.2). Furthermore, all the proteins that have been determined to be O-GlcNAcylated have also been found to be phosphorylated7. There are clear cases where the two modifications compete for the same acceptor site. For example, c-Myc, estrogen receptor β, and RNA polymerase II are reciprocally modified on the same serine/threonine residues24-26. Similarly, many proteins, including p53, CAMKIV, and FOXO1, are competitively modified by O-GlcNAc or phosphate at proximal residues27-29.  7  The addition of O-GlcNAc and phosphate is not always antagonistic. Inhibiting a kinase, such as GSK3β, increases the level of O-GlcNAcylation of many cytoskeletal and heat shock proteins, but decreases O-GlcNAcylation of many transcription factors and RNAbinding proteins30. In another study where OGA was inhibited, O-GlcNAcylation increased three-fold, and of the 700 monitored sites, levels of phosphorylation either significantly increased or decreased31. To further illustrate the extent of crosstalk, O-linked GlcNAc cycling enzymes have been found to be phosphorylated. OGT is both tyrosine and serine phosphorylated, and the phosphorylation is thought to activate OGT and assist in substrate recognition32. OGA has also been noted to be serine phosphorylated as well as O-GlcNAcylated8,33. And an ever increasing number of phosphate cycling enzymes are concluded to be O-GlcNAc-modified. A functional complex has been found that contains protein phosphatase I and OGT, allowing the enzyme complex to remove a phosphate and add an O-GlcNAc to the same substrate23. As well, there are examples where OGT and OGA are in complexes with both kinases and phosphatases7. Considering all of this data, it is clear that there is no simple explanation describing the relationship between O-GlcNAcylation and phosphorylation. Additionally, there have been suggestions of crosstalk between O-GlcNAcylation and other PTMs7. 1.5 O-GlcNAc and Alzheimer’s disease related protein tau Second to the pancreas, O-GlcNAcylation is most abundant in the brain7. Interestingly, many studies have indicated a connection between glucose metabolism in the brain and neurodegenerative diseases. In particular, glucose metabolism is impaired in Alzheimer’s disease (AD) neurons, thereby reducing O-GlcNAc levels of proteins, such as tau34. Tau is the major protein comprising neurofibrillary tangles associated with AD, and is 8  extensively and reciprocally phosphorylated and O-GlcNAcylated35. Studies using mouse models have demonstrated that overexpression of OGT in neurons increases OGlcNAcylation and decreases phosphorylation of tau. Conversely, deletion of OGT leads to tau hyperphosphorylation, and ultimately neuronal death36. Thus far, the only known function of tau is to bind to and stabilize microtubules (MTs) and is preferentially localized in neuronal axons37. The largest isoform of tau is comprised of 441 amino acid residues, and is an intrinsically disordered protein in isolation38. Tau possesses an N-terminal domain, referred to as the projection domain, where there is zero, one, or two 29-amino acid long acidic inserts (Figure 1.4a). In the carboxyl half, referred to as the microtubule-binding domain, there are either three or four imperfect 18residue MT-binding repeats. These differences represent the six isoforms of human tau, all derived from the same gene through alternative splicing39. Immediately upstream of the MTbinding repeats is a positively charged proline-rich region. Although its function is not fully characterized, it may contribute to binding negatively-charged MTs as well40. Tau has approximately 80 phosphorylation sites, and when hyperphosphorylated is known to be pathogenic41. Electrostatic repulsion of the negatively-charged phosphate groups with MTs results in an inability to bind, thereby increasing the concentration of free tau42. As the concentration of free tau increases, the probability of aggregation also increases. When tau is misfolded, the likelihood of first forming paired helical filaments (PHFs) and subsequently forming neurofibrillary tangles (NFTs) is increased. Paired helical filaments are formed by the MT-binding repeats stacking to form β-sheets43.  9  Projection domain  Microtubule binding domain  Proline-rich MT binding region repeats  Acidic repeats  1  441  S396 S400 S400  353  408  KIGSLDNITHVPGGGNKKIETHKLTFRENAKAKTDHGAEIVYKSPVVSGDTSPRHL  Figure 1.4: Domain structure of tau. a) Relative to the full-length isoform of human tau, six different isoforms resulting from alternative splicing have either zero, one, or two, acidic repeats, and either three or four MT-binding repeats. b) Tau353-408, the fragment studied herein.  PHFs from AD brains contain a protease-resistant core that coincides with the MTbinding repeats44. This is corroborated by the increased efficiency of the isolated MT-binding repeats aggregating relative to the full-length isoform of tau45. In vivo data has suggested that apoptotic caspases cleave tau at the N-terminal and C-terminal ends, yielding fragments that are more prone to aggregation. C-terminal truncations have been found to occur at Glu391 and Asp421 (Figure 1.5b)  46  . As a result, Mandelkow and coworkers suggest a global  “paperclip fold” for tau based on fluorescence resonance energy transfer (FRET) and electron paramagnetic resonance (EPR) analyses (Figure 1.5a)47. This agrees with the idea that when tau is truncated, the MT-binding repeats become more accessible, thereby increasing the rate of aggregation47. One of their more recent FRET and EPR analyses on phosphomimic  10  mutants to glutamic acid of tau have suggested a further opening and compaction of the paperclip (Figure 1.5c)48.  a)  b)  c) N  N  C  PP P P P  N  C  PP  Figure 1.5: A model representing global folding of a) full length wild-type tau, b) Cterminally truncated tau, and c) with phosphomimic glutamic acid mutations. C-terminally truncated tau results in more accessible MT-binding repeats, whereas the phosphomimics result in a more open and compact conformation of tau47,48.  In addition to extensive phosphorylation, tau is O-GlcNAc modified to a level of approximately 4 mol O-GlcNAc/mol tau35. There has also been a reciprocal relationship observed between O-GlcNAc and phosphorylation49. For example, brains associated with AD posses lower amounts of O-GlcNAc, and intriguingly, O-GlcNAc has not been detected on PHFs or NFTs49,50. As such, O-GlcNAc could be seen as a protective mechanism by limiting the amount of tau phosphorylation, and thus, aggregation51. When OGA is inhibited, tau from healthy rats had reduced overall phosphorylation, as well as 1.5-fold increased Ser400 OGlcNAcylation. This reflects a competition between GlcNAcylation and phosphorylation51,52. Conversely, Ser396, Ser400, and Ser404 can be phosphorylated53. In fact, a sequential phosphorylation pathway has been proposed for this region, where Ser404 phosphorylation by CDK5 primes the cascade, which then facilitates Ser400 and Ser396 phosphorylation by GSK3β53. GlcNAcylation is in direct competition with phosphorylation to modify Ser400, and can disrupt the phosphorylation cascade, thus affecting proximal residues (Figure 1.6). Vocadlo and coworkers also showed that O-GlcNAcylated Ser400 inhibited tau 11  oligomerization relative to wild-type species51. Due to the reciprocal nature of phosphorylation and O-GlcNAcylation on tau, it is an attractive model for studying these PTMs.  a)  b)  P P P  P g P  396 400 404  396 400 404  Figure 1.6: Post-translational modifications of S396, S400, and S404. a) Sequential phosphorylation pathway, where S404 phosphorylation by CDK5 promotes phosphorylation of S400, then S396 by GSK3β. b) O-GlcNAcylation at S400 inhibits this sequential phosphorylation pathway.  1.7 Thesis overview In this thesis, tau will be used as a model to study the structural and functional effects of O-GlcNAcylation. We hypothesize that O-GlcNAcylation of Ser400 results in local structural effects that decrease/prevent the polymerization of tau. Therefore, the goals of this thesis were primarily to compare the local structure and dynamic effects of O-GlcNAcylation on tau, and consequentially, its functional effects on aggregation. To gain a full understanding of the system, the effects of phosphorylation mimics on tau were also considered. In order to study O-GlcNAc-modified tau, an experimental method to posttranslationally modify tau in milligram quantities was required. Our collaborator, Dr. David Vocadlo (SFU) developed a system to O-GlcNAcylate tau in bacterial cells by cotransduction with an inducible plasmid encoding hOGT. Upon induction, both tau and hOGT  12  are produced, and over time, hOGT modifies tau in vivo. Because complete and specific phosphorylation is difficult to achieve in vitro or in vivo, phosphorylation mimics were used to study its effects on tau instead. In mutating the key serine residues (396, 400, and 404) to aspartic acid residues, a single negative charge was introduced, yielding a “phosphomimic” (S396D, S400D, and S404D). The next step was to design a construct that allowed for structural and functional studies of these PTMs on tau. Since local structure and dynamic studies were to be performed using nuclear magnetic resonance (NMR) spectroscopy, studying full-length tau, with 441 residues, was overly challenging due to line broadening and signal overlap. Furthermore, tau is an intrinsically disordered protein, and these tend to have poor chemical shift dispersion, increasing the difficulty of spectral interpretation. Therefore, a smaller model fragment was needed. However, to perform functional aggregation studies, it is necessary to include at least one MT-binding repeat. Finally, there has been an extra C-terminal O-GlcNAcylation site proposed to be on either Ser409, Ser412, or Ser41354. To avoid multiple O-GlcNAcylation sites, the final construct that I chose to study spanned residues 353-408 and is denoted as tau353-408 (Figure 1.4b). Complete NMR assignments were achieved for wild-type, O-GlcNAc-modified, and S400D tau353-408. Based on these chemical shifts, secondary structure propensity calculations revealed that S400D and O-GlcNAc-modified tau353-408 maintains its random coiled conformation. Although its structure was unchanged, 15N relaxation experiments displayed a small decrease in nsec-psec time scale conformational flexibility for amides near position 400. Finally, functional aggregation studies revealed that O-GlcNAc-modified tau353-408 had a  13  reduced amount of aggregation in vitro compared to phosphomimic and wild-type tau353-408 peptides.  14  Chapter 2: Materials and methods 2.1 Cloning 2.1.1 His6-SUMO-tau353-408 Restriction-free cloning was used to create a plasmid encoding His6-SUMO-tau353-408 following published protocols55. SUMO is known to enhance the expression and solubility of proteins in Escherichia coli, this construct was designed with the goal of increasing the yield of tau353-408 56. Primers, listed in Table 2.1, were used to PCR amplify the DNA for human tau353-408 (HGNC ID: 6893)57. PCR protocols and reagents used are listed in Tables 2.2 and 2.3, respectively. The resulting DNA was separated on a 1% agarose gel, purified with a GeneJET Gel Extraction Kit (Fermentas), and inserted into a His6SUMO-containing vector via linear amplification (Tables 2.2 and 2.3). After purification with a GeneJet PCR purification Kit (Fermentas), the resulting plasmid was transformed into E. coli DH5α cells via heat shock. Heat-shocked cells were incubated overnight on Luria broth (LB) agar plates at 37 ºC with selection for kanamycin resistance. The final plasmids were extracted using a GeneJET Plasmid Miniprep Kit (Fermentas) and submitted for DNA sequencing via GENEWIZ, Inc. The confirmed plasmid was transformed into E. coli BL21(λDE3) cells by heat shock. 2.1.2 Site-directed mutagenesis for aspartic acid mutants QuikChange site-directed mutagenesis was performed to generate phosphorylation mimic constructs. Residues 396, 400, and 404 were mutated into aspartic acid residues, independently, using primers listed in Table 2.1. Amplification of desired plasmids was  15  achieved using protocols and reagents listed in Tables 2.2 and 2.3. Purification, plasmid extraction, sequencing, and transformations were performed as previously mentioned. Table 2.1: Primers used for cloning Construct Primer Forward SUMOtau353-408 Reverse S396D S400D S404D  Forward Reverse Forward Reverse Forward Reverse  Sequence 5’- GTTTATCAGGAACAAACGGGGGGTAAGATTGG GTCCCTGGACAATATC-3’ 5’-GTTAGCAGCCGGATCTCAGAGATGCCGTGGAGA CGTGTC-3’ 5'-GCGGAGATCGTGTACAAGGACCCAGTGGTG-3' 5'-CACCACTGGGTCCTTGTACACGATCTCCGC-3' 5'-CCAGTGGTGGATGGGGACACG-3' 5'-CGTGTCCCCATCCACCACTGG-3' 5'-GGGGACACGGATCCACGGCATCTCAGC-3' 5'-GCTGAGATGCCGTGGATCCGTGTCCCC-3'  Table 2.2: Thermocycling protocols used for cloning Steps  PCR amplification  1. Initial 2. Denaturating 3. Annealing 4. Extending 5. Ending Repeat 2-5  Temp. Time 95 ºC 5 min. 95 ºC 30 sec. 65 ºC 1 min. 68 ºC 1 min. 68 ºC 5 min. 35 times  Linear amplification Temp. Time 95 ºC 5 min. 95 ºC 30 sec. 65 ºC 1 min. 72 ºC 13 min. 72 ºC 15 min. 35 times  Site-directed mutagenesis Temp. Time 95 ºC 5 min. 95 ºC 30 sec. 65 ºC 1 min. 72 ºC 14 min. 72 ºC 15 min. 35 times  16  Table 2.3: Reagents used for cloning Reagents  Plasmid/Template PCR product Forward primer Reverse primer dNTPs 10x buffer DMSO ddH2O pfu polymerase Taq polymerase Total  PCR amplification stock volume 50 ng/µL 1 µL 100 pM 2 µL 100 pM 2 µL 10 mM 4 µL 10 µL 80 µL 5 u/µl 1 µL 100 µL  Linear amplification stock volume 50 ng/µL 1 µL 300 ng/µL 2 µL 10 mM 1 µL 5 µL 40 µL 2.5 u/µL 1 µL 50 µL  Site-directed mutagenesis stock volume 50 ng/µL 1 µL 10 µM 1 µL 10 µM 1 µL 10 mM 2 µL 5 µL 2.5 µL 36.5 µL 2.5 u/µL 1 µL 50 µL  2.2 Protein expression and purification To express O-GlcNAc-modified tau353-408, an ampicillin-selectable plasmid containing the OGT gene was co-transformed with the kanamycin-resistant His6-SUMOtau353-408 plasmid into E. coli BL21(λDE3)51. In instances where there were no O-GlcNAc modifications, a catalytically inactive OGT gene was co-transformed as a control instead. Tau constructs were expressed in E. coli BL21(λDE3) cells and grown at 37 ºC until O.D.600 = 0.6. LB media was used to produce unlabeled protein, whereas M9 minimal media supplemented with 1 g/L 15NH4Cl or 1 g/L 15NH4Cl and 3 g/L 13C6-glucose (Sigma-Aldrich) was used to produce  15  N- or  15  N/13C-labeled protein, respectively. Protein expression was  induced with 0.5 mM IPTG for 16 hours at 16 ºC, and harvested by centrifugation at 4000×g in a GSA rotor (Sorvall) for 15 minutes. After one freeze-thaw cycle, the cell pellet was resuspended in Ni-NTA binding buffer (20 mM NaH2PO4 (pH 7.4), 500 mM NaCl, 5 mM imidazole). A minimum of 2 mg/mL of lysozyme was added and the sample was left rocking  17  on ice for 30 minutes. Cells were then disrupted by sonication (Branson Sonifier 250, VWR Scientific) at 60% duty cycle until clarified. The lysate was spun at 26,000×g for 60 minutes in a SS34 rotor (Sorvall). The supernatant was then passed through a 0.8 µm filter before being applied to a 5 mL Ni-NTA column (Qiagen). The column was washed with 20 mM NaH2PO4 (pH 7.4), 500 mM NaCl, 30 mM imidazole, and the protein was resolved using an FPLC (ÄKTA prime plus, General Electric) with 100 mL of buffer increasing the amount of imidazole linearly to 250 mM. Fractions containing the desired His6-SUMO-tau353-408 were identified by 15% SDS-PAGE gels and pooled. The His6-SUMO tag was removed, without leaving any residual amino acids, using the catalytic domain of the Sacchroymces cerevisiae SUMO hydrolase Ulp1. SUMO hydrolase was expressed and purified as published, using a clone provided by Dr. Keith Vosseller (Drexel University)58. Cleavage was performed overnight at room temperature with 5 µg/mL Ulp1 while dialyzing in 20 mM NaH2PO4 (pH 7.0) and 150 mM NaCl. Cleavage was verified on a 15% SDS-PAGE gel. The contents of the dialysis bag were centrifuged at 5000×g for 10 minutes, and the supernatant was applied to a Ni-NTA column to separate uncleaved protein and His6-SUMO from the desired, cleaved tau353-408, found in the flowthrough. Tau353-408 was then concentrated and purified by reverse phase-HPLC (Dionex). Tau353-408 was loaded onto a C18 250×10 mm semi-preparative column (Higgins Analytical, Inc.) and subsequently eluted with a 0%-60% acetonitrile gradient (Fisher Chemical) with 0.1% trifluoroacetic acid (TFA) (Sigma) at 1 mL/min over 80 minutes. Fractions (1 mL) were collected with a Gilson FC205 fraction collector (Mandel Scientific Company Ltd.), and those containing pure tau353-408, were pooled and lyophilized. 18  Since there was only ~10% O-GlcNAcylation in vivo, O-GlcNAc-modified fragments were subjected to a more stringent purification process to separate O-GlcNAc-tau353-408 from unmodified tau353-408. A semi-preparative 250×9.4 mm C8 column (Agilent) was used in RPHPLC. The fragments were moderately separated using a 22%-27% acetonitrile with 0.1% TFA over 100 minutes at 1 mL/min. Again, 1 mL fractions were collected. Using MALDITOF MS, resulting fractions were analyzed for O-GlcNAc-tau353-408. Samples containing mostly O-GlcNAc-tau353-408 were pooled and lyophilized, yielding a ~60% O-GlcNAcmodified tau353-408 sample. 2.3 NMR spectroscopy To characterize tau353-408, NMR spectroscopy was utilized. Optimal sample conditions were determined initially by recording 1H-15N HSQC spectra as a function of pH from 5.58.5 at increments of 0.5 units, followed by optimizing temperature conditions between 5 ºC35 ºC. Final conditions were chosen to be pH 6.0 and 15 ºC in order to minimize signal loss due to amide hydrogen exchange while yielding good quality spectra under near neutral pH conditions. All isotopically-labeled samples contained 10 mM NaH2PO4 (pH 6.0), 1.6% protease stock inhibitor tablet (Roche), and 10% D2O for the signal lock. One protease inhibitor tablet was dissolved in 5 mL of 10 mM NaH2PO4 (pH 6.0), yielding a 1000× stock. Spectra were recorded at 15 ºC using 600 MHz and 850 MHz Bruker Avance III NMR spectrometers equipped with triple resonance cryo-probes. Spectra were processed and analyzed using NMRpipe59,  nmrDraw59,  and  Sparky60.  The  CBCA(CO)NH,  HNCACB,  HNCO,  (H)CC(CO)TOCSY-NH, HNH- (τmix= 200 ms ) and NNH- (τmix= 200 ms) NOESY-HSQC  19  are described as per Sattler et al.61. HNHA and  15  N relaxation (T1, T2, NOE) experiments  were performed as well62,63. 2.4 Aggregation assays Aggregation assays were performed at 37 ºC with the same buffer conditions as NMR experiments. Samples (50 µL) contained 30 µM tau353-408, 7.5 µM low molecular weight heparin (4.0-6.5 kDa) (International Laboratory, USA), and 0.01 mg/mL thioflavin S (Sigma). These samples were placed in a black flat-bottom 386-well plate (Grenier) and sealed with optically clear crystallography tape (Hampton Research). Fluorescence (λexciation = 440 nm, λemission = 480 nm) was measured every 15 minutes for 40 hours using a Varioskan Flash fluorimeter (Thermo Scientific). Aggregation assays are extremely sensitive to protein concentrations, thus they were determined by two different spectrophotometric methods in triplicate. The first method involved measuring the absorbance of tyrosine residues at A280 with a NanoDrop 200c absorbance spectrophotometer (Thermo Scientific). The concentration was calculated according to Beer’s Law, using ε280 = 1490 L/mol•cm as predicted by the ExPASy ProtParam tool64. The second method utilized a bicinchoninic acid assay (Thermo Scientific) which measures protein concentration based on reduction of Cu+2 by peptide bonds65. BSA was used to make a standard curve. The final concentration for each protein sample was determined based on the average of the two assays. Aggregation assays were performed in triplicates with a “no protein” control, as well as a “no heparin” control, since heparin is the aggregation-inducing factor. Three different cultures of 15N-labeled wild-type, S396D, S400D, and S404D were purified individually, and  20  aggregation assays were performed for each batch in triplicates. To ensure consistency, each batch was checked by recording a 1H-15N HSQC spectrum prior to performing aggregation assays. After the assay was optimized, it was carried out with O-GlcNAc-tau353-408 as well. Because O-GlcNAc-tau353-408 was extremely difficult to produce, the sample used in the aggregation assays was preserved from prior NMR experiments.  21  Chapter 3: NMR spectral assignments 3.1 Assignments from main-chain nuclei A 1H-15N HSQC spectrum is a two-dimensional NMR spectrum that yields a crosspeak at the chemical shifts of each pair of directly bonded 1H and 15N nuclei. Under the conditions used herein, every non-proline amide should yield one signal, and thus the 1H-15N HSQC spectrum provides an “NMR fingerprint” of a polypeptide or protein. Although unstructured polypeptides typically have limited amide 1HN chemical shift dispersion, the 15N chemical shift of each residue is affected by its neighboring residues. As a result, there is often sufficient dispersion in the 15N dimension of a 1H-15N HSQC to yield resolved signals from each amide in even a relatively large intrinsically disordered protein. Furthermore, due to the dynamic behavior of unstructured polypeptides, these signals are usually strong and sharp. Assigning the 1H-15N HSQC spectrum is often the first step in studying protein by NMR spectroscopy. This required the preparation of uniformly  13  C/15N-labeled tau and the  recording of three-dimensional NMR experiments. Of these, the CBCA(CO)NH and HNCACB experiments are most useful for sequential backbone assignments (Figure 3.1)66. The HNCACB correlates the amide 1HN and 15N signals of residue i to its own 13Cα and 13Cβ as well as those of the preceding residue, i-1 (Figure 3.2). The complementary CBCA(CO)NH correlates the amide 1HN and 15N signals of residue i to only the 13Cα and 13Cβ signals of the previous residue, i-1. Together, these experiments provide sequential connections and allow one to distinguish intra- versus inter- residue correlations. Furthermore, the 13Cα and 13Cβ signals may be distinguished as they possess opposite phases in HNCACB experiments. Several additional experiments were also recorded to extend and 22  verify the spectral assignments of tau. These included an (H)CC(CO)-TOCSY-NH which correlates all of the aliphatic 13C signals of residue i-1 with the 1HN and 15N signals of residue i, an HNHA connects the 1Hα, 1HN, and  15  N signals of residue i, and the HNH- and NNH-  NOESY-HSQC experiments which provide intra-residue NOE correlations between 1H’s within 5 Å (Figure 3.2). Finally, the HNCO experiment yields the carbonyl 13C’ signal of i-1 via correlations with the 1HN and  15  N signals of residue i. Collectively, these spectra were  used to obtain the chemical shift assignments for wild-type, GlcNAc-modified, and S400D tau353-408, tabulated in Appendix A, and shown in Figure 3.3. All three tau353-408 species gave 1  H-15N HSQC spectra with very narrow 1HN chemical shift dispersion, indicating that they  are predominantly disordered. In the O-GlcNAc-modified tau353-408 sample, there were chemical shift changes in the residues proximal to Ser400, further discussed in section 4.1. Since only 60% of tau353-408 was modified, based on peak intensities, two sets of peaks were observed for residues adjacent to the site of modification. Of particular interest, the signal from the GlcNAc amide group also appeared in the 1H-15N HSQC spectrum (Figure 3.3b). On the other hand, the phosphomimic S400D tau353-408 mutant was 100% “modified,” and as such, there was only one set of peaks. The signal from Ser400 no longer existed, and a new peak appeared, corresponding to Asp400. Several residues near Asp400 also exhibited minor chemical shift perturbations.  23  I371 CBCA(CO)NH  I371  E372  E372  HNCACB  CBCA(CO)NH  HNCACB  E372Cα  30 K370Cα  K370Cβ  1 1 1 1 I371Cα  I371Cα  I371Cα  13 C (ppm)  40  50  K370Cβ  I371Cβ  60  8.25 8.10 1 H (ppm)  E372Cβ  K370Cβ  8.25 8.10 1 H (ppm)  I371Cβ  I371Cβ  8.55 8.40 1 H (ppm)  8.55 8.40 1 H (ppm)  Figure 3.1: Strategies for assigning main-chain resonances. The spectra are sequentially assigned based on the connectivities of 1HN, 15N, 13Cα, and 13Cβ resonances. The 13Cα and 13 β C resonances of the i, and the i and i-1 residues are detected in the CBCA(CO)NH and HNCACB spectra, respectively. Furthermore, in the HNCACB spectrum, the signs of the 13 α C and 13Cβ peaks are positive (red) and negative (green), respectively.  24  HNCACB  CBCA(CO)NH  (H)CC(CO)NH-TOCSY-NH  HNHA  HNCO  NOESY-HSQC  Figure 3.2: Nuclei involved in NMR spectral assignments. The HNCACB experiments correlate amide 1HN and 15N signals with the i and i-1 13Cα/β, whereas CBCA(CO)NH and HNCO experiments correlate amide signals with i-1 13Cα, 13Cβ and 13C’, respectively. (H)CC(CO)NH-TOCSY-NH experiments correlate amide signals with all i-1 aliphatic carbons. The HNHA experiments correlate the amide signals with the 1Hα of i, and also provides a measure of the 3JHN-Hα scalar coupling. Finally, NOESY-HSQC experiments resolve through-space NOE interactions between adjacent protons (<5 Å) through the 15N shift of an amide.  25  G366  a) 110  b)  G367  G365c  c) G401  G389  110  G365t G401  G401g  T403c  115  T403c  T403cg  T403t  T403t  T386  G355  T403tg  115  T377  S356 T373  S396c  S396c  125  130  N359 S396t S404t  S400g  WT 8.6  D402  8.0  120  D402 V398  GlcNAc  V399g D400  GlcNAc WT  L408  8.4 8.2 w1 H (ppm)  S396t S404t  S404g  w  E391 N381 K383 I360 D402 S400 D358 H407 H407c K385 V398t H362 R406t V363c K369 E380 R406c D387 H374 I371 R379 F378 I392 K370 A390 L376 K375 L357 A382 V398c K395t V399 A384 Y394c V393 Y394t E372  w  H388  120  S404c  T361  N368  w  15 N (ppm)  S404c  8.6  8.4 8.2 w1 H (ppm)  125  V399  S400D WT 8.0  8.6  8.4 8.2 w1 H (ppm)  130 8.0  26  Figure 3.3: Assigned 1H-15N HSQC spectrum of a)wild-type tau353-408 (black). Also shown are the spectra representing b)GlcNAcmodified tau353-408 (green) and c) S400D tau353-408 (red) overlaid on that of the wild-type species, with selected peaks labeled. Distinct signals from residues perturbed by the trans (t) and cis (c) isomers of Val363-Pro364, Ser396-Pro397, and Ser404-Pro405 are identified. Selected chemical shift changes due to modifications have also been indicated by blue lines.  3.2 Cis/trans conformational isomers of tau The 1H-15N HSQC spectra of the three tau353-408 species contained an extra set of weaker signals, attributable to the cis isomer of the three X-Pro (where X is any amino acid) peptide bonds. The cis and trans conformers exchange slowly, on the order of tens of seconds, that results in two distinct peaks in the 1H-15N HSQC spectra67. Usually trans conformers are favored energetically over cis conformers (typically 70%-90% trans population), and hence results in a larger peak relative to the cis conformer67. Importantly, these ratios only apply to X-Pro bonds in unstructured polypeptides or in the dynamic termini and loops of structured proteins. In contrast, usually one conformer is stabilized in the wellordered regions of structured proteins. The signals from residues perturbed by the cis/trans isomerization of Val363-Pro364, Ser396-Pro397, and Ser404-Pro405 in tau were assigned with two methods. First, each gave a separate set of  13  Cα,  13 β 15  C  N, and 1HN correlations in HNCACB and CBCA(CO)NH  spectra. Second, to confirm experimentally which signals arose from which conformer, an (H)CC(CO)-TOCSY-NH spectrum was used to measure the 13Cβ and 13Cγ chemical shifts of the three prolines. If the chemical shift difference between the 13Cβ and 13Cγ is ~10ppm, the proline is in the cis conformation; however, if the chemical shift difference is ~5 ppm, the proline is in the trans conformation68. Indeed, as shown unambiguously in Figure 3.4, the weaker signals arose from sub-populations of tau353-408 in a cis conformation.  27  R406c  R406c  R406t  R406t  (H)CC(CO)TOCSY CBCA(CO)NH (H)CC(CO)TOCSY CBCA(CO)NH  P405cCγ  P405tCγ  30  30  13 C (ppm)  P405tCβ  P405cCβ  P405tCβ  P405cCβ  40  40 8.65  8.50  1 H (ppm)  8.65  8.50  1 H (ppm)  8.40  8.40  8.25  8.25  1 H (ppm)  1 H (ppm)  Figure 3.4: Cis versus trans X-Pro peptide bond assignment. Cis and trans X-Pro conformations are assigned based on the chemical shift differences between the 13Cβ and 13Cγ observed in the (H)CC(CO)TOCSY. A chemical shift difference ~10 ppm indicates the cis conformer, whereas a difference of ~5 ppm indicates the trans conformer. Because 13Cα and 13 β C of the i-1 residue can be seen in both the (H)CC(CO)TOCSY and CBCA(CO)NH, the cis and trans peaks can be identified while assigning signals from sequential main-chain nuclei.  3.3 Assignment of O-GlcNAc NMR signals In addition to characterizing the polypeptide, NMR can also provide insights into the covalently linked O-GlcNAc moiety. Signals from the  13  C/15N-labeled sugar were also  assigned using 1H-15N HSQC, 1H-13C HSQC, HNHA, HNCACB, and HCCH-TOCSY experiments, tabulated in Appendix B, and shown in Figure 3.5. Fortunately, Nacetylglucosamine has several easily distinguished signals in both the 1H-13C HSQC and 1H15  N HSQC as shown in Figure 3.3b and Figure 3.5a, respectively53. The HNHA experiments 28  correlated amide 1H-15N and 1H2 signals, whereas the HNCACB experiments allowed the chemical shifts of C1, C2, and C3, as well as the methyl carbon nuclei to be assigned (refer to Figure 3.5d for the numbering scheme) for O-GlcNAc. An HCCH-TOCSY spectrum correlates the signals from all protons within a 1H-13C spin system to a directly bonded 1H13  C pair. As such, a common pattern of proton chemical shifts appears at 1H-1H planes taken  at each sugar 13C chemical shift. The C6 plane has two of these patterns, since there are two directly-bonded protons, denoted arbitrarily as H6’ and H6’’ (Figure 3.5b). The C4 and C5 carbon chemical shifts were determined by a similar pattern as well, and subsequently mapped on the 1H-13C HSQC. These signals could not be identified unambiguously for C4 or C5 and were instead discriminated based on previously reported NMR data (HMDB ID: 00803)69. Of note, these NMR experiments did not provide a direct link between any nuclei in the O-GlcNAc and tau353-408 residues as required to unambiguously identify Ser400 as the site of covalent attachment. However, the largest chemical shift difference between wild-type and O-GlcNAc-modified tau353-408 occurred for Ser400 and its neighboring nuclei, as seen in the 1H-15N HSQC (Figure 3.3b). Furthermore, the 13Cα and 13Cβ chemical shifts of Ser400 in the HNCACB and CBCA(CO)NH were altered by 2 ppm and 7 ppm, respectively, from the Ser400 chemical shifts in wild-type tau353-408. This is entirely consistent with previous studies that used MS to identify Ser400 as the O-GlcNAc acceptor53. Finally,  15  N-resolved NOESY experiments that correlate the signals of  15  N-labeled  amide protons with those of other protons through space, were performed (Figure 3.2). However, only intra-sugar NOE interactions between the O-GlcNAc amide and the 1H1, 1H2, 1  H3 and CH3 were observed. Importantly, no other NOE correlations were observed between 29  GlcNAc and tau (Figure 3.5c). This reveals that the GlcNAc amide is not localized near the polypeptide chain.  a)  b)  1  H-13C HSQC  HCCH-TOCSY  3.6  3.6  60  w 1 H (ppm)  C2 -H 2  60  H 6’’  H 6’ H 2  3.9  C6 -H 6’’  H4 H5  13 C 63.3ppm  3.8  C6 -H 6’  65  H3  3.6  3.9  3.4  1 H (ppm) w  65  c) NOESY-HSQC  2.0  2.0  w  13 C (ppm)  CH 3 70  70  2.5  2.5  3.0  3.0  d)  75  75  1 H (ppm) w  w  C4 -H 4  H3  3.5  C3 -H 3  3.5 H2  C5 -H 5  4.0  4.0  H1  w  102 w  C1 -H 1 104  104  8.1 8.2  w  102  NH  8.1 8.2  w  4.5  w  4.0  3.5  1 H (ppm)  8.20 8.12 1 H (ppm)  Figure 3.5: Assignment of O-GlcNAc NMR signals. a) 1H-13C HSQC with 13C1 through 13C6 assigned. The acetyl methyl signal at 24.8 and 1.9 ppm is not shown. Although many peaks are seen in this spectrum, the b) HCCH-TOCSY helped discern them from those of the tau polypeptide due to its distinct 1H correlation pattern. c) The HNH-NOESY-HSQC showing only intra-sugar NOE interactions to the amide. d) Numbering scheme for O-GlcNAc. 30  Chapter 4: Structural, functional, and dynamic studies of tau 4.1 Chemical shift perturbations The change in amide chemical shifts for the O-GlcNAc-modified and S400D tau353408  compared to wild-type tau353-408 were all in the vicinity of Ser400 (Figure 4.1). However,  this is difficult to interpret as amide chemical shifts are extremely sensitive to conformational changes and electrostatic effects. On the other hand, the lack of any significant chemical shift perturbations for the rest of tau353-408 strongly indicates that the changes near position 400 do not affect distal residues or the global properties of tau353-408. 4.2 Secondary structure propensities derived from chemical shifts Chemical shift values are very indicative of secondary structure, as such, three different methods were used to interrogate and compare secondary structure in wild-type, OGlcNAc-modified, and S400D tau353-408. 4.2.1 δ2D The main-chain (13Cα, 13Cβ, 13C’, 1Ha, 1HN, and 15N) signals of a residue are strongly dependent upon its secondary structure within a protein70. Therefore, numerous algorithms have been developed to accurately predict secondary structure from chemical shifts alone. Of these,  the  program  δ2D  (http://www-vendruscolo.ch.cam.ac.uk/d2D/index.php)  was  specifically calibrated for intrinsically disordered proteins, and provides the normalized propensities of a residue to adopt an α-helical, β-sheet, polyproline II helical or random coil conformation, with scores totaling to one71. The results from δ2D analysis of the three forms of tau353-408 considered herein are plotted in Figure 4.2. Most significantly, all the residues in each form show random coil 31  scores of larger than 0.7. This is consistent with the 1H-15N HSQC spectra (Figure 3.3), as well as numerous other studies indicating that soluble tau is an intrinsically disordered protein72-74. Upon closer inspection, some small differences are seen between the δ2D results for wild-type, GlcNAc-modified and S400D tau353-408 (Figure 4.2). For example, residues around Ser404 have the highest β-sheet propensity in wild-type tau353-408, whereas residues around Asp400 in S400D tau353-408 have a slightly elevated polyproline II helical propensity. Also, the GlcNAc-modification is predicted to slightly increase the β-sheet propensity around Ser400, and the coil propensity of neighboring residues. However, the significance of these predictions is unclear as the differences in conformational propensities are small. It is important to note that the δ2D database does not contain reference chemical shift values for GlcNAc-modified serine, and treated Ser400 as unmodified. Furthermore, inspection of the assigned NMR chemical shifts for each species relative to those predicted for completely disordered reference polypeptides does not yield an obvious pattern of perturbations indicative of any secondary structure (Appendix C)75. Therefore, GlcNAc modification or mutation of Ser400 does not induce any predominant secondary structure. 4.2.2 3JHN-Hα coupling The three-bond 3JHN-Hα coupling constant is dependent upon the φ dihedral angle of a residue and thus is also a measure of secondary structure. For example, residues in helices have 3JHN-Hα values around 3-4 Hz, those in β-sheets have values around 9-10 Hz, and those in random coils have couplings around 7 Hz62. Using a HNHA experiment the 3JHN-Hα coupling constants of the three tau353-408 species were measured62. As shown in Figure 4.3, the measured couplings were uniformly indicative of a conformational disorder in wild-type, 32  GlcNAc-modified and S400D tau. An increase in helical or sheet propensity should appear as a cluster of adjacent residues with 3JHN-Hα values less than or greater than 7 Hz, respectively, and this was not observed. 4.2.3 NOESY experiments NOESY experiments correlate through-space dipolar interactions between protons within ~5 Å. In addition to providing tertiary structural information, regular patterns of NOE interactions between main-chain protons are also very diagnostic of secondary structure. These include NOE interactions between the amide 1HN of residue i with its neighbor’s amide protons (i+1 and i-1), as well as i to the 1Hα of itself (i) and the preceding residue (i-1). The latter are often denoted as dαN(i,i) and dαN(i-1,i), respectively70. Furthermore, a cluster of adjacent residues with dαN(i,i) versus dαN(i-1,i) NOE intensity ratios of ~6 are indicative of an α-helical structure, whereas ratios of ~0.25 arise from β-sheet structures. In the case of a random coil polpeptide, the dαN(i,i)/dαN(i-1,i) ratio is in the range of 0.3576. HNH- and NNH-NOESY-HSQC spectra were recorded for the tau353-408 species. These spectra further confirmed that tau353-408 is globally unfolded, since many residues had dαN(i,i)/dαN(i-1,i) ratios ~0.35 (not shown). However, focusing on the residues proximal to residue 400, a trend could not be found in tau353-408 to discern whether there were any local structural changes due to modifications. This is attributed mainly to poor dispersion in the 1H dimension. As a result of overlapping 1Hα and 1HN signals, the dαN(i,i)/dαN(i-1,i) intensity ratios could not be calculated accurately. Furthermore, there are two proline residues (which lack amide protons required for this experiment) and a glycine residue in this region. Since, the two glycine 1Hα signals were not resolved, correct dαN(i,i)/dαN(i-1,i) intensity ratios could not be determined. 33  1 0.9  Change in chemical shift (ppm)  0.8 0.7 0.6 0.5 0.4 0.3 0.2  0  K353 I354 G355 S356 L357 D358 N359 I360 T361 H362 V363 P364 G365 G366 G367 N368 K369 K370 I371 E372 T373 H374 K375 L376 T377 F378 R379 E380 N381 A382 K383 A384 K385 T386 D387 H388 G389 A390 E391 I392 V393 Y394 K395 S396 P397 V398t V399 S/D400 G401 D402 T403 S404 P405 R406 H407 L408  0.1  GlcNAc  S400D  Figure 4.1: Change in amide chemical shifts (Δδ = [(Δδ1H)2 + (0.2×Δδ15N)2]1/2) of the trans conformers of GlcNAc-modified and S400D tau353-408 compared to wild-type tau353-408. Significant perturbations occurred only near the 400 residue. Additional minor chemical shift perturbations clustered around histidine residues, and most likely reflect minor changes in sample pH value which was close to the pKa value of ~6 for a histidine side chain. 34  Random coil propensity 0.000 I354 G355 S356 L357 D358 N359 I360 T361 H362 V363 P364 G365 G366 G367 G368 G369 K370 I371 E372 T373 H374 K375 L376 T377 F378 R379 E380 N381 A382 K383 A384 K385 T386 D387 H388 G389 A390 E391 I392 V393 Y394 K395 S396 P397 V398 V399 S400 G401 D402 T403 S404 P405 R406 H407  Polyproline II helical propensity  β-sheet propensity  α-helical propensity 0.04  0.02  0.00  0.20  0.00  0.30  0.20  0.10  0.00  1.000  0.500  WT GlcNAc S400D  35  Figure 4.2: Secondary structure propensity calculated by ∂2D for each residue in wild-type (black), GlcNAc-modified (green), and S400D (red) tau353-408. Note: the y-axes have different scales.  K353 I354 G355 S356 L357 D358 N359 I360 T361 H362 V363 P364 G365 G366 G367 N368 K369 K370 I371 E372 T373 H374 K375 L376 T377 F378 R379 E380 N381 A382 K383 A384 K385 T386 D387 H388 G389 A390 E391 I392 V393 Y394 K395 S396 P397 V398 V399 S/D400 G401 D402 T403 S404 P405 R406 H407 L408 GlcNAc  HN-HA (Hz) 3J  9 8 7 6 5 4 3 2 1 0  WT  GlcNAc  S400D  Figure 4.3: 3JHN-HA coupling values for non-glycine residues in tau353-408. The three-bond 3JHN-HA coupling values around 3-4 Hz are typical for α-helices, whereas values around 9-10 Hz are typical for β-sheets. Finally, values around 7 Hz, which is observed above, is typical for random coils. Ambiguous values due to spectra overlap have not been shown. Error bars are based on spectral signal-tonoise ratios.  36  4.3 15N Relaxation Based on chemical shifts and  3  JHN-Hα coupling constants, wild-type tau is  predominantly disordered and neither GlcNAc-modification nor S400D mutation induces any persistent secondary structure. To investigate this further, amide  15  N relaxation (T2, 1H-15N-  NOE) experiments were performed to gain insights on backbone dynamics. Note that the heteronuclear NOE values decrease from 0.8 to -0.3 with increasing mobility of the amide 15  N-1HN bond vector on the nsec-psec time scale. Decreasing NOE values accompanied with  increasing T2 values indicate an increase in flexibility, and the reverse trend indicates an increase in rigidity70. The amide  15  N relaxation data for the three tau species are presented in Figure 4.3.  Most notably, the central regions of each have roughly uniform heteronuclear NOE values ~0.35 and T2 values ~300 msec, whereas the termini have reduced NOE values and increased T2 values. This pattern is diagnostic of a conformationally disordered polypeptide. However, upon closer inspection, the NOE values of Ser400 and Gly 401 increase slightly relative to those of the wild-type species upon GlcNAc-modification, whereas the T2 values decrease slightly relative to those of the wild-type species upon GlcNAc-modification, whereas the T2 values decrease slightly. Similar effects are noticed for the S400D sample. This suggests that the presence of the GlcNAc or Asp400 slightly dampens the motions of these two amides; however, the effect is small. These patterns were also observed within the two populations found in the ~60% modified GlcNAc sample, indicating that these results are not due to differences in experimental conditions. The motions of the GlcNAc amide were also probed by these  15  N relaxation measurements. As shown on the very right hand side of Figure 4.4,  the sugar amide has an unusually low NOE value of -0.1 and a high T2 value of ~700 msec 37  800 700  T2 (msec)  600 500 400 300 200 100 0 0.6  NOE  0.4 0.2 0  -0.4  K353 I354 G355 S356 L357 D358 N359 I360 T361 H362 V363 P364 G365 G366 G367 N368 K369 K370 I371 E372 T373 H374 K375 L376 T377 F378 R379 E380 N381 A382 K383 A384 K385 T386 D387 H388 G389 A390 E391 I392 V393 Y394 K395 S396 P397 V398 V399 S400 G401 D402 T403 S404 P405 R406 H407 L408 GlcNAc  -0.2  WT  GlcNAc  S400D  38  Figure 4.4: 15N T2, and heteronuclear NOE values for wild-type (black), GlcNAc-modified (green), and S400D (red) tau353-408. These were measured on an 850 MHz Bruker AVANCE III NMR spectrometer. T2 values and their errors were fit by SPARKY60. NOE values were calculated by the ratio of peak intensities in experimental and control spectra, and errors were calculated by using signalto-noise ratios. Missing data correspond to proline residues or residues with signal overlap.  Therefore, it is highly mobile on the nsec-psec timescale, and is not conformationally restrained by the tau polypeptide. 4.4 Cis/trans isomerization The possibility that the GlcNAc-modification or S400D mutation altered the cis/trans ratio of the adjacent to Ser396-Pro397 or Ser404-Pro405 was also examined. These ratios were determined from the relative 1H-15N HSQC peak intensities of amides giving resolved signals attributable to each X-Pro bond (Table 4.1). Note that this assumes similar NMR relaxation properties for the two conformers such that relative peak intensities reflect relative populations. The introduction of a negatively charged aspartic acid, in the place of serine (S400D), results in no significant change for either Ser396-Pro397 or Ser404-Pro405 (Table 4.2). In contrast, it appears that the GlcNAc modification on tau353-408 has decreased the cis populations slightly for Pro405 and Pro397. However, it is difficult to say with certainty that these values are significant. Table 4.1: Residues that were used to calculate relative cis populations, where y denotes a peak that was used. P364 P397 P405 a b c G365 Y394 K395 S396 V398 T403 S404 R406 H407 y y y y y y y Wild-type y y y y y y GlcNAc y y y y y y y y S400D a The Gly365 cis peak was not observed for GlcNAc-modified tau353-408, this may be the result of a minor chemical shift change in this region. Residue  b  Lys395 was not used for any of the constructs due to an unidentified cis peak.  c  Val398c was only observed for wild-type and S400D tau353-408, however, there were overlapping intensities observed in the wild-type spectrum.  39  Table 4.2: Relative cis populations calculated based on peak intensities of n number of residues about the X-Pro bonds. Errors were calculated based on signal-to-noise values for each of the peaks. Construct Wild-type GlcNAc S400D  P364 8.0 7.0  n 1 0 1  %cis P397 8.7±0.13 7.8±0.43 8.8±0.02  n 2 2 3  P405 10.6±0.13 7.6±0.44 10.1±0.01  n 4 4 4  4.5 Aggregation assays Valuable insight into tau’s ability to aggregate in vivo can be obtained using polyanions, such as heparin to induce aggregation in vitro. Thioflavin S can bind to the resulting aggregates, increasing its relative fluorescence emission at 480 nm by 35-fold77. These assays are viewed somewhat qualitatively in terms of the rate and extent of indirectlydetected aggregation. In this experiment, wild-type, S400D, and O-GlcNAc-modified tau353-408 were tested alongside two additional phosphomimic constructs. The latter being S396D and S404D, thus representing all three serine residues that become phosphorylated in wild-type tau via a phosphorylation cascade (Figure 1.6). As shown in Figure 4.5, O-GlcNAc-modified tau353-408 yielded the lowest amount of fluorescence, indicating the lowest of aggregation. S396D and S400D tau353-408 mutants had increased aggregation relative to wild-type tau353-408, whereas S404D had decreased aggregation compared to wild-type tau353-408. Overall, it is striking that these single mutations and GlcNAc modifications have such a pronounced effect on the aggregation of this tau fragment, especially given that they are distal to the microtubulebinding repeat.  40  3  2.5  Relative fluorescence units  2 WT 1.5  GlcNAc S396D S400D  1  S404D control 0.5  0 0 -0.5  5  10  15  20  25  30  35  40  45  Time (hours)  Figure 4.5: Aggregation assay results for tau353-408. S400D (red snowflakes) and S396D (purple crosses) have higher amounts of aggregation relative to wild-type (black triangles), whereas S404D (blue circles) has lower aggregation. Most significantly, GlcNAc-modified tau353-408 showed very little aggregation under these conditions. Average data for triplicates are shown, with error bars representing standard deviations.  41  Chapter 5: Discussion and conclusions The goals of my thesis were to determine the effects of O-GlcNAc-modified Ser400 on the local structure and dynamics of tau353-408 as well as its effects on heparin-induced aggregation. My initial hypothesis was that there would be a change in local structure which could correlate with reduced aggregation. Indeed, the O-GlcNAc modification did decrease the level of tau353-408 aggregation, however, no significant structural changes were observed. 5.1 Strategies to overcome experimental challenges In order to achieve the goals of my thesis, I had to overcome a number of experimental challenges that will be discussed in this section. 5.1.1 Optimizing protein expression and O-GlcNAcylation yield The original pET28a vectors encoding full-length or C-terminal half constructs of tau, provided by our collaborator Dr. David Vocadlo (SFU) did not yield milligram levels of expressed protein required for typical NMR studies. Therefore, I developed a SUMO-fusion system in order to increase the expression of tau fragments in E. coli. Such fusions have proven useful for increasing protein expression and also have the advantage of leaving no extra amino acid acids after cleavage by a SUMO-specific hydrolase58. Furthermore, to simplify the resulting NMR spectra, a shorter construct spanning the fourth MT-binding repeat and including Ser400 was chosen. In the end, a plasmid encoding His6SUMO-tau353-408 was created and expressed high level of tau353-408 required for structural and functional characterization. The next step was to create a GlcNAc-modified peptide. To date, GlcNAc-modified peptides have mostly been produced synthetically with Fmoc chemistry78. However, a 56-residue  13  C/15N-labeled peptide would have been very  42  costly and difficult to synthesize. Alternatively, in vitro methods have been used to OGlcNAcylate peptides, Smet-Nocca et al. achieved 8.2% O-GlcNAcylation after incubating their tau peptide (residues 392-411) for two days at 37 ºC with recombinant ncOGT and a 10fold excess of UDP-GlcNAc53. Although a viable approach with a biosynthetically-labeled peptide, it would still be very costly to produce 13C/15N-UDP-GlcNAc as required to label the monosaccharide. The Vocadlo group developed an in vivo system that resulted in ~10% OGlcNAcylation tau353-408. This was preferred to the in vitro method because of the slightly improved levels of O-GlcNAcylation without the need to produce and purify OGT, and more significantly, it allowed for uniform labeling of both the peptide and GlcNAc, as required for NMR studies. However, in order to study the local dynamic and structural effects of OGlcNAcylation, spectral assignments were first required. 5.1.2 Bigger is better: assigning IDPs with a high field magnet Tau353-408 is an intrinsically disordered peptide (IDP), and thus has poor spectral dispersion in the 1H dimension. Use of a recently installed high field magnet (850 MHz) tremendously helped in improving spectral resolution as well as sensitivity for tau. Standard main-chain directed 3D experiments became feasible for backbone assignment, which was further helped by the fact that most  13  Cα and  13 β  C chemical shifts matched the expected  random coil values of each amino acid79. This allowed for the full assignment of the signals from backbone nuclei in wild-type, S400D, and O-GlcNAcylated tau353-408. Furthermore, cis conformers signals were also assigned for many residues in the vicinity of the three X-Pro bonds. Since GlcNAc was also 13C/15N-labeled, it was possible to assign the signals from all of the NMR-active nuclei as well.  43  In the 1H-15N HSQC spectra, residues adjacent to position 400 underwent chemical shift perturbations. There were no drastic changes in chemical shift, with the exception of Asp400 in the S400D tau353-408 1H-15N HSQC. There were minor chemical shift changes observed near histidine residues (Figure 4.1), as they titrate around pH 6.0, the pH at which NMR experiments were conducted. Since chemical shift perturbations were observed, further structural investigations were made. 5.1.3 Gleaning structural information from chemical shifts in the absence of NOESYbased distance restraints Residues adjacent to position 400 showed amide 1HN and 15N chemical shift changes upon mutation or O-GlcNAcylation of this site, indicating that any conformational perturbations were local rather than global. Accordingly, the secondary structures for all three species of tau353-408 were interrogated by three approaches. The first involved the algorithm δ2D, which predicts structure from main-chain chemical shifts. The second involved measuring the three-bond 3JHN-HA, scalar coupling values, which depend upon backbone φ angles. The third involved NOESY experiments, indicative of through-space proton-proton interactions, however, accurate values could not be calculated for residues in the 400 region. Based on each of these methods, I concluded that there was no obvious change in secondary structure induced upon O-GlcNAcylation or the mutation of Ser400. This is in agreement with a study performed on full-length tau, where residual β-sheet, αhelical, and polyproline II helical structure was found throughout tau, but not for residues spanning 345 to 43072. Although structural changes were not observed, probing the dynamic properties between all three species of tau353-408 resulted in differences.  44  5.1.4 15N relaxation experiments report on local dynamics Amide 15N relaxation experiments were recorded for all three species of tau353-408. For residues near GlcNAc-Ser400 and Asp400, there was a very small increase in NOE values and a decrease in  15  N T2 values, indicating a slight dampening in the nsec-psec timescale  motions of the amide  15  N-1HN bond vectors. Even more notable, the GlcNAc amide had  exceptionally high T2 and low NOE values, together with the lack of any detected saccharidepolypeptide NOESY correlations, it can be concluded that, although covalently linked to Ser400, GlcNAc does not interact with tau353-408 and is very mobile on the nsec-psec timescale. This is consistent with earlier findings, where glycosylation results in a shortrange decrease in backbone flexibility for highly flexible peptides80-82. With structural and dynamic properties assessed, functional aggregation assays were then performed. 5.1.5 Optimizing the aggregation assays The aggregation assays were performed at a physiological temperature of 37 ºC, however, this quickly led to uneven sample evaporation due to heating edge effects. To circumvent this problem, optically clear crystallography tape was used to seal the wells, such that each well maintained the same volume throughout the experiment. In the heparininduced aggregation assays, O-GlcNAc-modified tau353-408 reproducibly aggregated much less than wild-type tau353-408. In contrast, S396D and S400D exhibited increased aggregation compared to wild-type tau353-408, whereas S404D aggregated somewhat less. Thus despite the lack of any structural perturbations, these modifications had a very pronounced effect on the aggregation of tau353-408.  45  5.2 Changes in cis/trans population In the 1H-15N HSQC spectra of tau353-408, more than the expected number of peaks were observed. The subset of weak extra signals has been assigned to residues adjacent to the cis isomers of the three X-Pro bonds present in tau353-408. Although full-length tau has 48 prolines and has been completely assigned by NMR, there has been no mention of the cis isoform73,83. This is surprising and may simply reflect their relatively low populations and difficulty to observe in the very crowded spectra of the full-length species. However, in a study on the same O-GlcNAc site on tau, in a construct spanning residues 392-411 (and thus overlapping with tau353-408) there was no mention of a cis population either. Regardless, there are many important implications of cis/trans isomerization. Several studies have reported a change in X-Pro cis/trans isomer ratio due to PTMs. In the conserved loop region of nicotinic acetylcholine receptor, Rickert and Imperiali noted that cis percentage decreased from 45% to 30% due to N-linked glycosylaion84. However, in disordered regions, such as the octapeptide studied in the mucin domain of MAdCAM-1, the cis population was found to be a more typical 8-10%. It was noted that for this octapeptide, O-GlcNAc alone caused no change in cis/trans isomerization, whereas the addition of a steric branched fucose resulted in a cis percentage decrease by two-fold, from 8% to 4%85. Finally, the introduction of a negative charge, such as a phosphate group, or a negatively charged amino acid, has been suggested to favor the cis isomer86,87. For tau353-408, a small decrease in the cis to trans ratio of Ser396-Pro397 was observed as a result of the O-GlcNAc modification, and no effects were observed for the S400D mutation, relative to wild-type. The small increase in relative trans population due to O-GlcNAcylation may have an interesting implication. A decrease in PP2A phosphatase activity has been shown to increase 46  hyperphosphorylation of tau in mice88. The levels of phosphorylation for residues Ser202, Thr205, Thr231, and Ser235 were affected by phosphatase activity of PP2A89. Interestingly, proline-directed phosphatases, such as PP2A, can only dephosphorylate pSer/Thr-Pro residues in the trans conformer90. PP2A activity is thus stimulated by Pin1, a cis/trans prolyl isomerase91. Pin1 has been observed to accelerate the isomerization of amyloid precursor protein by over 1000-fold92. Pin1 knockouts in mice results in an accumulation of phosphate at residues 202, 205, 231, and 23593. The level of soluble Pin1 in AD brains is reduced by five-fold relative to age-matched normal brains94. Recently, the development of cis- and trans-specific antibodies have shown that it is cis, but not trans, phosphorylated tau (p-tau) that appears early on in the brains of humans exhibiting mild cognitive impairment since the cis isomer cannot promote microtubule assembly. Using these antibodies, Lu and coworkers discovered that Pin1 catalyzes the cis/trans interconversion of p-tau to prevent AD tau pathology95. O-GlcNAcylation at Ser400 may be playing the same role by increasing the amount of tau in a phosphatase susceptible trans conformation, thereby reducing p-tau aggregation. 5.3 Aggregation of tau In solution, tau is intrinsically disordered, and when hyperphosphorylated, will form paired helical filaments that lead to neurofibrillary tangles. However, O-GlcNacylation has been seen to reciprocally affect PHF formation35. Glutamic and aspartic acid residues are commonly used to mimic the effects of phosphorylation since they possess a negative charge. It was shown that a single phosphomimic was sufficient to cause a decrease in MT-binding, to induce conformational changes in the first and second MT-binding repeats, and most importantly, to increase tau aggregation96-98. In concordance with this observation, the S396D  47  and S400D mutations increased the aggregation of tau353-408 as measured by thioflavin S fluorescence. In contrast, the S404D construct modestly reduced aggregation. Most dramatically, the O-GlcNAc-modified tau353-408 showed very little aggregation. This indicates that O-GlcNAcylation can inhibit oligomerization of tau either by destabilizing PHF formation, or by stabilizing the monomer in solution. The effect of GlcNAcylation may also arise from its interplay with tau phosphorylation. Assuming that aspartic acid is a functional mimic of phosphoserine, I surmise that the phosphorylation of Ser396 and Ser400 are more crucial in inducing paired helical formation than the phosphorylation of Ser404. Given that the O-GlcNAcylation at Ser400 is reported to block the phosphorylation cascade, it could further reduce aggregation by indirectly reducing phosphorylation53. To better understand the reciprocal relationship in tau353-408, it would be valuable to test the different combinations of phosphomimics. Would two or three phosphomimics increase the amount of aggregation? Could one O-GlcNAc counteract the effects of the phosphomimics? Finally, this study would be more accurate if a phosphate group (bearing almost two negative charges at neutral pH) was added, instead of mutating it to an aspartic acid residue (which only has one negative charge). 5.4 The global fold of tau Although the site of modification being studied is distal to the MT-binding repeats, it has the ability to change the amount of aggregation. What is surprising is that just one new negative charge or one monosaccharide can cause such an effect. Considering the global “paperclip” fold of tau (Figure 1.5), where both C-terminal truncations and phosphomimics  48  result in more accessible MT-binding repeats, it is clear that the C-terminus is involved in paired helical formation47,48. Furthermore, both Glu391 and Asp 421 truncations have been observed to enhance the rates of tau filament formation in vitro and have been found in NFTs99,100. Conceivably, C-terminally truncated tau353-408 results in an accessible MT-binding repeat, and the phosphomimics further amplifies this effect. Although the effects of modifications on the global folding of tau have not been studied, presumably it results in a less accessible MT-binding repeat. Now considering the global structure of aggregated tau, recently, Müller et al. proposed a tau paired helical filament model based on immunogold labeling transmission electron micrography experiments. They demonstrated that N- and C-terminal tails are not central to filament formation, but form a “two-layered polyelectrolyte brush”101. This suggests that, around the aggregated MT-binding domains, there is a rigid layer of fibrils with positive charges surrounded by a sparse and soft brush layer with negative charges (Figure 5.1). I hypothesize that the introduction of aspartic acid residues, beyond creating a more accessible MT-binding repeat, also forms strong interactions with positive charges found in the “brush”, resulting in a sturdier framework around the already-formed filaments. Furthermore, the introduction of a bulky GlcNAc moiety may lead to a less densely packed core.  49  C  N aggregation  paired helical filaments Neutral Negative Positive  Figure 5.1: When tau fragments aggregate in to paired helical filaments, they form a twolayered polyelectrolyte brush.  5.5 Multiple GlcNAcylation sites Although I was careful in avoiding multiple O-GlcNAc sites at the C-terminal end of the tau353-408 construct, there was an O-GlcNAc modification site present at the N-terminal end of the construct. That is, Ser356 has been observed to be O-GlcNAcylated. However, this residue is part of the fourth MT-binding repeat, and considered to be required for aggregation studies. From NMR experiments, O-GlcNAcylation was never observed at Ser356, and only observed at Ser400. As such, I speculate that the TPR domain of OGT recognizes peptides sequences N-terminal to the site of modification. 5.6 GlcNAcylation in various systems O-GlcNAc has been shown to induce turn-like structures in small peptide models of the C-terminal domain of RNA polymerase II and the N-terminus of murine estrogen receptor β24. This could result from hydrogen bonding between the N-acetyl moiety and the polypeptide backbone. The evidence for turn formation was based primarily on NOE 50  connectivities, and in the case of RNA polymerase II, further corroborated with 1H/2H exchange experiments78. Both peptides exhibited NOE patterns diagnostic for turns upon GlcNAcylation. In particular, Li and coworkers concluded that in a small peptide from murine estrogen receptor β, O-GlcNAc promotes turn formation based on a newly observed NOESY peak between the modified residue and its i+1 neighbor. Conversely, they suggest that phosphorylation at this location creates a more extended conformation largely based on the absence of a NOESY signal otherwise seen in the unmodified peptide. In the case of Simanek et al. using 1H/2H exchange experiments, they found that amide protons belonging to the turn-like structure exchanged more than 20× slower than the remaining amide protons. In contrast to these examples, no clear evidence was found for a GlcNAc-induced turn in tau353-408. In HNH-NOESY-HSQC spectra of tau353-408 (Figure 3.5c), the only NOE correlations observed to the GlcNAc amide were intra-sugar connectivities. Even though OGlcNAcylation at Ser400 resulted in minor local chemical shift perturbations, there were no indications of any local structural changes (Figure 4.1). Carbohydrates can modulate proteins in many ways. N-linked glycans usually have a common core structure, but differ terminally, with complex, and sometimes, branched, structures102. Proteins undergo glycosylation in the lumen of the endoplasmic reticulum, as a cotranslational event. As such, the addition of large polar sugars is considered to play a role in their correct folding, as well as increasing their solubility. Furthermore, due to their polar nature, sugars are usually found on the exterior of a protein102. For example, O-linked glycans are often located in hinge or linker regions between folded globular domains103. Glycans can shield the protein surface to prevent aggregation, and increase thermal stability, 104,105  . However, tau is intrinsically disordered and extremely soluble, yet it still possesses  51  several glycosylation sites72. Thus, the link between increased O-GlcNAcylation and decreased tau aggregation must lie in the monosaccharide increasing the solubility of tau 353408  .  5.7 Summary The goals of this thesis were to compare the local structure and dynamic effects of OGlcNAcylation and phosphomimic on tau, and their functional effects on aggregation. As far as we know, this is the longest peptide to be studied with an O-GlcNAc modification. Using NMR experiments, I determined that neither O-GlcNAcylation nor aspartate mutations (considered to be a phosphomimic) induced a change in secondary structure. Although there were minimal changes in the local dynamics around position 400, no specific conformational perturbations were detected. Since there were no structural changes, the modifications effects on tau’s aggregation must arise for other reasons. In performing functional aggregation assays, I confirmed that indeed aggregation propensity decreased for O-GlcNAc-modified tau353-408 and increased for phosphomimics. 5.8 Future directions The effects of O-GlcNAcylation and serine to aspartic acid mutations on tau were studied in this thesis. Although tau was an interesting model to investigate due to its link with Alzheimer’s disease, there was only ~10% modification and the process of purifying the resulting O-GlcNAc modified peptide to yield milligram quantities was difficult. Studying a peptide for which OGT has a higher activity for modifying in vivo could result in more amenable samples. Furthermore, studying other peptides could provide more insights into the common and distinct structural and dynamic changes due to O-GlcNAcylation.  52  Selective O-GlcNAc labeling could circumvent the issue of low modification, and would help in further investigating the dynamics of the linked O-GlcNAc moiety. In the case of the 56-residue tau353-408, the NMR signals from the sugar amide were easily identifiable. However, in order to monitor the dynamics of O-GlcNAc in larger proteins, selective labeling might be required. In principle, this can be achieved by using in vivo modification by providing exogenous labeled GlcNAc. However, to achieve high levels of GlcNAc labeling without also labeling the protein an E. coli strain that is auxotrophic for UDPGlcNAc would be required. UDP-GlcNAc is the end product of the hexosamine biosynthetic pathway, and thus it is important to understand this pathway in order to create the desired auxotrophic strain (Figure 5.2)106. In E. coli GlcNAc and GlcN are taken up into the cell by specific permesases of the phosphotransferase system. This system converts GlcNAc into GlcNAc-6-P. With a removal of the acetate group, it becomes GlcN-6-P, which is isomerized to GlcN-1-P. Both of these become re-acetylated and are then converted to UDP-GlcNAc. However, GlcN-6-P may also be synthesized from fructose-6-P (made during glycolysis) and glutamine by GlmS. In order to maximize the use of labeled GlcNAc, a GlmS knock-out strain must be created so that unlabeled GlcNAc (from glycolysis) will not be incorporated into the protein. Rather, all labeled GlcNAc or GlcN will be derived from exogenous sources. Although E. coli strains with the GlmS mutation have been isolated and characterized, they are not suitable with today’s expression systems107. Thus, recombineering technology can be used to generate a GlmS knock-out in E. coli BL21(λDE3). Finally, 15N-labeled GlcNAc may be synthesized chemo-enzymatically. Fungi grown on  15  N-ammonia will result in ~60% labeled glucosamine in their cell wall. Subsequent 53  depolymerization and acetylation of their cell wall will yield  15  N-labeled GlcNAc. This in  turn can be provided to the GlmS auxotrophic strain in order to generate GlcNAcylated proteins with only the monosaccharide labeled.  GlcN  ManXYZ  GlcNAc  NagK GlcNAc-6-P NagA NagB Fructose-6-P  GlcN-6-P GlmS  GlmM GlcN-1-P  Glycolysis  GlmU GlcNAc-1-P GlmU UDP-GlcNAc  Peptidoglycan  Figure 5.2: The hexosamine biosynthetic pathway. In order to create an auxotrophic strain for UDP-GlcNAc, it will be necessary to create a GlmS knockout.  54  Works cited 1.  Finishing the euchromatic sequence of the human genome. Nature 431, 931-45 (2004).  2.  Jensen, O.N. Modification-specific proteomics: characterization of post-translational modifications by mass spectrometry. Curr Opin Chem Biol 8, 33-41 (2004).  3.  Torres, C.R. & Hart, G.W. Topography and polypeptide distribution of terminal Nacetylglucosamine residues on the surfaces of intact lymphocytes. Evidence for Olinked GlcNAc. J Biol Chem 259, 3308-17 (1984).  4.  Greis, K.D. & Hart, G.W. Analytical methods for the study of O-GlcNAc glycoproteins and glycopeptides. Methods Mol Biol 76, 19-33 (1998).  5.  Wang, Z. et al. Enrichment and site mapping of O-linked N-acetylglucosamine by a combination of chemical/enzymatic tagging, photochemical cleavage, and electron transfer dissociation mass spectrometry. Mol Cell Proteomics 9, 153-60 (2010).  6.  Syka, J.E., Coon, J.J., Schroeder, M.J., Shabanowitz, J. & Hunt, D.F. Peptide and protein sequence analysis by electron transfer dissociation mass spectrometry. Proc Natl Acad Sci U S A 101, 9528-33 (2004).  7.  Hart, G.W., Slawson, C., Ramirez-Correa, G. & Lagerlof, O. Cross talk between OGlcNAcylation and phosphorylation: roles in signaling, transcription, and chronic disease. Annu Rev Biochem 80, 825-58 (2011).  8.  Butkinaree, C., Park, K. & Hart, G.W. O-linked beta-N-acetylglucosamine (OGlcNAc): Extensive crosstalk with phosphorylation to regulate signaling and transcription in response to nutrients and stress. Biochim Biophys Acta 1800, 96-106 (2010).  9.  Slawson, C., Copeland, R.J. & Hart, G.W. O-GlcNAc signaling: a metabolic link between diabetes and cancer? Trends Biochem Sci 35, 547-55 (2010).  10.  Dias, W.B. & Hart, G.W. O-GlcNAc modification in diabetes and Alzheimer's disease. Mol Biosyst 3, 766-72 (2007).  11.  Shafi, R. et al. The O-GlcNAc transferase gene resides on the X chromosome and is essential for embryonic stem cell viability and mouse ontogeny. Proc Natl Acad Sci U S A 97, 5735-9 (2000).  12.  Park, B.H., Karpinets, T.V., Syed, M.H., Leuze, M.R. & Uberbacher, E.C. CAZymes Analysis Toolkit (CAT): web service for searching and analyzing carbohydrate-active enzymes in a newly sequenced organism using CAZy database. Glycobiology 20, 1574-84 (2010).  55  13.  Vocadlo, D.J. O-GlcNAc processing enzymes: catalytic mechanisms, substrate specificity, and enzyme regulation. Curr Opin Chem Biol 16, 488-97 (2012).  14.  Liu, Y. et al. Developmental regulation of protein O-GlcNAcylation, O-GlcNAc transferase, and O-GlcNAcase in mammalian brain. PLoS One 7, e43724 (2012).  15.  Jinek, M. et al. The superhelical TPR-repeat domain of O-linked GlcNAc transferase exhibits structural similarities to importin alpha. Nat Struct Mol Biol 11, 1001-7 (2004).  16.  Lazarus, M.B., Nam, Y., Jiang, J., Sliz, P. & Walker, S. Structure of human OGlcNAc transferase and its complex with a peptide substrate. Nature 469, 564-7 (2011).  17.  Gorlich, D., Prehn, S., Laskey, R.A. & Hartmann, E. Isolation of a protein that is essential for the first step of nuclear protein import. Cell 79, 767-78 (1994).  18.  Butkinaree, C. et al. Characterization of beta-N-acetylglucosaminidase cleavage by caspase-3 during apoptosis. J Biol Chem 283, 23557-66 (2008).  19.  Toleman, C., Paterson, A.J., Whisenhunt, T.R. & Kudlow, J.E. Characterization of the histone acetyltransferase (HAT) domain of a bifunctional protein with activable O-GlcNAcase and HAT activities. J Biol Chem 279, 53665-73 (2004).  20.  Rao, F.V. et al. Structural insights into the mechanism and inhibition of eukaryotic OGlcNAc hydrolysis. EMBO J 25, 1569-78 (2006).  21.  Dennis, R.J. et al. Structure and mechanism of a bacterial beta-glucosaminidase having O-GlcNAcase activity. Nat Struct Mol Biol 13, 365-71 (2006).  22.  Macauley, M.S., Whitworth, G.E., Debowski, A.W., Chin, D. & Vocadlo, D.J. OGlcNAcase uses substrate-assisted catalysis: kinetic analysis and development of highly selective mechanism-inspired inhibitors. J Biol Chem 280, 25313-22 (2005).  23.  Wells, L., Kreppel, L.K., Comer, F.I., Wadzinski, B.E. & Hart, G.W. O-GlcNAc transferase is in a functional complex with protein phosphatase 1 catalytic subunits. J Biol Chem 279, 38466-70 (2004).  24.  Chen, Y.X. et al. Alternative O-GlcNAcylation/O-phosphorylation of Ser16 induce different conformational disturbances to the N terminus of murine estrogen receptor beta. Chem Biol 13, 937-44 (2006).  25.  Chou, T.Y., Hart, G.W. & Dang, C.V. c-Myc is glycosylated at threonine 58, a known phosphorylation site and a mutational hot spot in lymphomas. J Biol Chem 270, 18961-5 (1995).  56  26.  Kelly, W.G., Dahmus, M.E. & Hart, G.W. RNA polymerase II is a glycoprotein. Modification of the COOH-terminal domain by O-GlcNAc. J Biol Chem 268, 1041624 (1993).  27.  Yang, W.H. et al. Modification of p53 with O-linked N-acetylglucosamine regulates p53 activity and stability. Nat Cell Biol 8, 1074-83 (2006).  28.  Dias, W.B., Cheung, W.D., Wang, Z. & Hart, G.W. Regulation of calcium/calmodulin-dependent kinase IV by O-GlcNAc modification. J Biol Chem 284, 21327-37 (2009).  29.  Housley, M.P. et al. O-GlcNAc regulates FoxO activation in response to glucose. J Biol Chem 283, 16283-92 (2008).  30.  Wang, Z., Pandey, A. & Hart, G.W. Dynamic interplay between O-linked Nacetylglucosaminylation and glycogen synthase kinase-3-dependent phosphorylation. Mol Cell Proteomics 6, 1365-79 (2007).  31.  Wang, Z., Gucek, M. & Hart, G.W. Cross-talk between GlcNAcylation and phosphorylation: site-specific phosphorylation dynamics in response to globally elevated O-GlcNAc. Proc Natl Acad Sci U S A 105, 13793-8 (2008).  32.  Whelan, S.A., Lane, M.D. & Hart, G.W. Regulation of the O-linked beta-Nacetylglucosamine transferase by insulin signaling. J Biol Chem 283, 21411-7 (2008).  33.  Beausoleil, S.A. et al. Large-scale characterization of HeLa cell nuclear phosphoproteins. Proc Natl Acad Sci U S A 101, 12130-5 (2004).  34.  Liu, F. et al. Reduced O-GlcNAcylation links lower brain glucose metabolism and tau pathology in Alzheimer's disease. Brain 132, 1820-32 (2009).  35.  Arnold, C.S. et al. The microtubule-associated protein tau is extensively modified with O-linked N-acetylglucosamine. J Biol Chem 271, 28741-4 (1996).  36.  O'Donnell, N., Zachara, N.E., Hart, G.W. & Marth, J.D. Ogt-dependent Xchromosome-linked protein glycosylation is a requisite modification in somatic cell function and embryo viability. Mol Cell Biol 24, 1680-90 (2004).  37.  Binder, L.I., Frankfurter, A. & Rebhun, L.I. The distribution of tau in the mammalian central nervous system. The Journal of Cell Biology 101, 1371-1378 (1985).  38.  Cleveland, D.W., Hwo, S.Y. & Kirschner, M.W. Purification of tau, a microtubuleassociated protein that induces assembly of microtubules from purified tubulin. J Mol Biol 116, 207-25 (1977).  39.  Andreadis, A., Broderick, J.A. & Kosik, K.S. Relative exon affinities and suboptimal splice site signals lead to non-equivalence of two cassette exons. Nucleic Acids Research 23, 3585-3593 (1995). 57  40.  Goode, B.L. et al. Functional interactions between the proline-rich and repeat regions of tau enhance microtubule binding and assembly. Mol Biol Cell 8, 353-65 (1997).  41.  Ittner, A. et al. Brief update on different roles of tau in neurodegeneration. IUBMB Life 63, 495-502 (2011).  42.  Chen, J., Kanai, Y., Cowan, N.J. & Hirokawa, N. Projection domains of MAP2 and tau determine spacings between microtubules in dendrites and axons. Nature 360, 674-7 (1992).  43.  Daebel, V. et al. β-Sheet Core of Tau Paired Helical Filaments Revealed by SolidState NMR. Journal of the American Chemical Society 134, 13982-13989 (2012).  44.  Wischik, C.M. et al. Structural characterization of the core of the paired helical filament of Alzheimer disease. Proc Natl Acad Sci U S A 85, 4884-8 (1988).  45.  Friedhoff, P., Schneider, A., Mandelkow, E.M. & Mandelkow, E. Rapid assembly of Alzheimer-like paired helical filaments from microtubule-associated protein tau monitored by fluorescence in solution. Biochemistry 37, 10223-30 (1998).  46.  Rissman, R.A. et al. Caspase-cleavage of tau is an early event in Alzheimer disease tangle pathology. J Clin Invest 114, 121-30 (2004).  47.  Jeganathan, S., von Bergen, M., Brutlach, H., Steinhoff, H.J. & Mandelkow, E. Global hairpin folding of tau in solution. Biochemistry 45, 2283-93 (2006).  48.  Jeganathan, S., Chinnathambi, S., Mandelkow, E.M. & Mandelkow, E. Conformations of microtubule-associated protein Tau mapped by fluorescence resonance energy transfer. Methods Mol Biol 849, 85-99 (2012).  49.  Liu, F., Iqbal, K., Grundke-Iqbal, I., Hart, G.W. & Gong, C.X. O-GlcNAcylation regulates phosphorylation of tau: a mechanism involved in Alzheimer's disease. Proc Natl Acad Sci U S A 101, 10804-9 (2004).  50.  Ledesma, M.D., Bonay, P., Colaço, C. & Avila, J. Analysis of microtubule-associated protein tau glycation in paired helical filaments. Journal of Biological Chemistry 269, 21614-21619 (1994).  51.  Yuzwa, S.A. et al. Increasing O-GlcNAc slows neurodegeneration and stabilizes tau against aggregation. Nat Chem Biol 8, 393-9 (2012).  52.  Yuzwa, S.A. et al. A potent mechanism-inspired O-GlcNAcase inhibitor that blocks phosphorylation of tau in vivo. Nat Chem Biol 4, 483-90 (2008).  53.  Smet-Nocca, C. et al. Identification of O-GlcNAc sites within peptides of the Tau protein and their impact on phosphorylation. Mol Biosyst 7, 1420-9 (2011).  58  54.  Yuzwa, S.A. et al. Mapping O-GlcNAc modification sites on tau and generation of a site-specific O-GlcNAc tau antibody. Amino Acids 40, 857-68 (2011).  55.  van den Ent, F. & Lowe, J. RF cloning: a restriction-free method for inserting target genes into plasmids. J Biochem Biophys Methods 67, 67-74 (2006).  56.  Marblestone, J.G. et al. Comparison of SUMO fusion technology with traditional gene fusion systems: enhanced expression and solubility with SUMO. Protein Sci 15, 182-9 (2006).  57.  Seal, R.L., Gordon, S.M., Lush, M.J., Wright, M.W. & Bruford, E.A. genenames.org: the HGNC resources in 2011. Nucleic Acids Res 39, D514-9 (2011).  58.  Weeks, S.D., Drinker, M. & Loll, P.J. Ligation independent cloning vectors for expression of SUMO fusions. Protein Expr Purif 53, 40-50 (2007).  59.  Delaglio, F. et al. NMRPipe: a multidimensional spectral processing system based on UNIX pipes. J Biomol NMR 6, 277-93 (1995).  60.  Goddard, T.D., Kneller, D.G. (ed.) SPARKY 3, (University of California, San Francisco).  61.  Sattler, M., Schleucher, J. & Griesinger, C. Heteronuclear multidimensional NMR experiments for the structure determination of proteins in solution employing pulsed field gradients. Progress in Nuclear Magnetic Resonance Spectroscopy 34, 93-158 (1999).  62.  Kuboniwa, H., Grzesiek, S., Delaglio, F. & Bax, A. Measurement of HN-H alpha J couplings in calcium-free calmodulin using new 2D and 3D water-flip-back methods. J Biomol NMR 4, 871-8 (1994).  63.  Kay, L.E., Torchia, D.A. & Bax, A. Backbone dynamics of proteins as studied by nitrogen-15 inverse detected heteronuclear NMR spectroscopy: application to staphylococcal nuclease. Biochemistry 28, 8972-8979 (1989).  64.  Gasteiger E., H.C., Gattiker A., Duvaud S., Wilkins M.R., Appel R.D., Bairoch A. (ed.) Protein Identification and Analysis Tools on the ExPASy Server, 571-607 (Humana Press, 2005).  65.  Smith, P.K. et al. Measurement of protein using bicinchoninic acid. Analytical Biochemistry 150, 76-85 (1985).  66.  Grzesiek, S. & Bax, A. An efficient experiment for sequential backbone assignment of medium-sized isotopically enriched proteins. Journal of Magnetic Resonance (1969) 99, 201-207 (1992).  67.  Schmid, F.X., Mayr, L.M., Mucke, M. & Schonbrunner, E.R. Prolyl isomerases: role in protein folding. Adv Protein Chem 44, 25-66 (1993). 59  68.  Schubert, M., Labudde, D., Oschkinat, H. & Schmieder, P. A software tool for the prediction of Xaa-Pro peptide bond conformations in proteins based on 13C chemical shift statistics. J Biomol NMR 24, 149-54 (2002).  69.  Wishart, D.S. et al. HMDB 3.0--The Human Metabolome Database in 2013. Nucleic Acids Res 41, D801-7 (2013).  70.  Wuthrich, K. (ed.) NMR of Proteins and Nucleic Acids, (Wiley, New York, 1986).  71.  Camilloni, C., De Simone, A., Vranken, W.F. & Vendruscolo, M. Determination of secondary structure populations in disordered states of proteins using nuclear magnetic resonance chemical shifts. Biochemistry 51, 2224-31 (2012).  72.  Mukrasch, M.D. et al. Structural polymorphism of 441-residue tau at single residue resolution. PLoS Biol 7, e34 (2009).  73.  Narayanan, R.L. et al. Automatic assignment of the intrinsically disordered protein Tau with 441-residues. J Am Chem Soc 132, 11906-7 (2010).  74.  Peterson, D.W., Zhou, H., Dahlquist, F.W. & Lew, J. A soluble oligomer of tau associated with fiber formation analyzed by NMR. Biochemistry 47, 7393-404 (2008).  75.  Tamiola, K., Acar, B. & Mulder, F.A.A. Sequence-Specific Random Coil Chemical Shifts of Intrinsically Disordered Proteins. Journal of the American Chemical Society 132, 18000-18003 (2010).  76.  Maltsev, A.S., Ying, J. & Bax, A. Impact of N-terminal acetylation of alphasynuclein on its random coil and lipid binding properties. Biochemistry 51, 5004-13 (2012).  77.  Friedhoff, P., Schneider, A., Mandelkow, E.-M. & Mandelkow, E. Rapid Assembly of Alzheimer-like Paired Helical Filaments from Microtubule-Associated Protein Tau Monitored by Fluorescence in Solution†. Biochemistry 37, 10223-10230 (1998).  78.  Simanek, E.E. et al. Glycosylation of Threonine of the Repeating Unit of RNA Polymerase II with β-Linked N-Acetylglucosame Leads to a Turnlike Structure. Journal of the American Chemical Society 120, 11567-11575 (1998).  79.  Ulrich, E.L. et al. BioMagResBank. Nucleic Acids Research 36, D402-D408 (2008).  80.  Wormald, M.R. & Dwek, R.A. Glycoproteins: glycan presentation and protein-fold stability. Structure 7, R155-60 (1999).  81.  Wormald, M.R. et al. Conformational studies of oligosaccharides and glycopeptides: complementarity of NMR, X-ray crystallography, and molecular modelling. Chem Rev 102, 371-86 (2002).  60  82.  Imperiali, B. & Rickert, K.W. Conformational implications of asparagine-linked glycosylation. Proc Natl Acad Sci U S A 92, 97-101 (1995).  83.  Harbison, N.W., Bhattacharya, S. & Eliezer, D. Assigning Backbone NMR Resonances for Full Length Tau Isoforms: Efficient Compromise between Manual Assignments and Reduced Dimensionality. PLoS One 7, e34679 (2012).  84.  Rickert, K.W. & Imperiali, B. Analysis of the conserved glycosylation site in the nicotinic acetylcholine receptor: potential roles in complex assembly. Chem Biol 2, 751-9 (1995).  85.  Wu, W.-g. et al. Structural Study on O-Glycopeptides:  Glycosylation-Induced Conformational Changes of O-GlcNAc, O-LacNAc, O-Sialyl-LacNAc, and O-SialylLewis-X Peptides of the Mucin Domain of MAdCAM-1. Journal of the American Chemical Society 121, 2409-2417 (1999).  86.  Grathwohl, C. & Wüthrich, K. Nmr studies of the rates of proline cis–trans isomerization in oligopeptides. Biopolymers 20, 2623-2633 (1981).  87.  Evans, C.A. & Rabenstein, D.L. Nuclear magnetic resonance studies of the acid-base chemistry of amino acids and peptides. II. Dependence of the acidity of the Cterminal carboxyl group on the conformation of the C-terminal peptide bond. Journal of the American Chemical Society 96, 7312-7317 (1974).  88.  Kins, S. et al. Reduced protein phosphatase 2A activity induces hyperphosphorylation and altered compartmentalization of tau in transgenic mice. J Biol Chem 276, 38193200 (2001).  89.  Landrieu, I. et al. Molecular Implication of PP2A and Pin1 in the Alzheimer's Disease Specific Hyperphosphorylation of Tau. PLoS One 6, e21521 (2011).  90.  Lu, K.P. & Zhou, X.Z. The prolyl isomerase PIN1: a pivotal new twist in phosphorylation signalling and disease. Nat Rev Mol Cell Biol 8, 904-16 (2007).  91.  Zhou, X.Z. et al. Pin1-dependent prolyl isomerization regulates dephosphorylation of Cdc25C and tau proteins. Mol Cell 6, 873-83 (2000).  92.  Pastorino, L. et al. The prolyl isomerase Pin1 regulates amyloid precursor protein processing and amyloid-beta production. Nature 440, 528-34 (2006).  93.  Liou, Y.C. et al. Role of the prolyl isomerase Pin1 in protecting against agedependent neurodegeneration. Nature 424, 556-61 (2003).  94.  Lu, P.J., Wulf, G., Zhou, X.Z., Davies, P. & Lu, K.P. The prolyl isomerase Pin1 restores the function of Alzheimer-associated phosphorylated tau protein. Nature 399, 784-8 (1999).  61  95.  Nakamura, K. et al. Proline isomer-specific antibodies reveal the early pathogenic tau conformation in Alzheimer's disease. Cell 149, 232-44 (2012).  96.  Kiris, E. et al. Combinatorial Tau Pseudophosphorylation: MARKEDLY DIFFERENT REGULATORY EFFECTS ON MICROTUBULE ASSEMBLY AND DYNAMIC INSTABILITY THAN THE SUM OF THE INDIVIDUAL PARTS. Journal of Biological Chemistry 286, 14257-14270 (2011).  97.  Fischer, D. et al. Conformational changes specific for pseudophosphorylation at serine 262 selectively impair binding of tau to microtubules. Biochemistry 48, 1004755 (2009).  98.  Chang, E., Kim, S., Schafer, K.N. & Kuret, J. Pseudophosphorylation of tau protein directly modulates its aggregation kinetics. Biochim Biophys Acta 1814, 388-95 (2011).  99.  Novak, M., Kabat, J. & Wischik, C.M. Molecular characterization of the minimal protease resistant tau unit of the Alzheimer's disease paired helical filament. EMBO J 12, 365-70 (1993).  100.  Abraha, A. et al. C-terminal inhibition of tau assembly in vitro and in Alzheimer's disease. J Cell Sci 113 Pt 21, 3737-45 (2000).  101.  Wegmann, S., Medalsy, I.D., Mandelkow, E. & Muller, D.J. The fuzzy coat of pathological human Tau fibrils is a two-layered polyelectrolyte brush. Proc Natl Acad Sci U S A 110, E313-21 (2013).  102.  Taylor, M.E., and Drickamer, K. (ed.) Introduction to Glycobiology, (Oxford University Press, Oxford, 2003).  103.  Petrescu, A.J., Milac, A.L., Petrescu, S.M., Dwek, R.A. & Wormald, M.R. Statistical analysis of the protein environment of N-glycosylation sites: implications for occupancy, structure, and folding. Glycobiology 14, 103-14 (2004).  104.  Rudd, P.M. & Dwek, R.A. Glycosylation: Heterogeneity and the 3D Structure of Proteins. Critical Reviews in Biochemistry and Molecular Biology 32, 1-100 (1997).  105.  Wang, C., Eufemi, M., Turano, C. & Giartosio, A. Influence of the carbohydrate moiety on the stability of glycoproteins. Biochemistry 35, 7299-307 (1996).  106.  Park, J.T. & Uehara, T. How bacteria consume their own exoskeletons (turnover and recycling of cell wall peptidoglycan). Microbiol Mol Biol Rev 72, 211-27, table of contents (2008).  107.  Sarvas, M. Mutant of Escherichia coli K-12 defective in D-glucosamine biosynthesis. J Bacteriol 105, 467-71 (1971).  62  Appendices Appendix A: Chemical shift tables for tau353-408 Table A.1: Chemical shifts for15N/13C wild-type tau353-408 13 13 α 13 β 1 N C' C C H Residue (ppm) (ppm) (ppm) (ppm) K353 176.4 61.48 38.37 I354 173.9 44.98 8.624 G355 174.7 58.21 63.76 8.151 S356 177.1 54.91 41.91 8.419 L357 175.8 54.34 40.95 8.164 D358 175 53.18 38.52 8.274 N359 176.3 61.14 37.95 7.99 I360 174.2 61.92 69.57 8.177 T361 173.8 55.64 29.65 8.454 H362 174.4 59.18 33.97 7.994 V363 177.5 63.42 31.9 P364 174.8 45.21 8.587 G365 173.8 45.17 8.324 G366 174.2 45.08 8.324 G367 175.1 53.06 38.66 8.309 N368 176.3 56.23 32.83 8.255 K369 176.2 56.19 32.75 8.313 K370 176 60.82 38.41 8.2 I371 176.3 56.03 30.34 8.513 E372 174.1 61.94 69.55 8.222 T373 174.2 55.41 29.67 8.39 H374 175.6 57.39 32.73 8.173 K375 177.3 55.08 42.14 8.319 L376 174 61.54 69.86 8.056 T377 175.4 57.9 39.5 8.265 F378 175.8 55.86 30.76 8.142 R379 176.2 56.6 30.12 8.329 E380 175 53.04 38.61 8.458 N381 177.7 53.07 18.98 8.21 A382 176.3 56.12 32.75 8.149 K383 177.7 52.37 19.12 8.158 A384  15  N (ppm) 114.2 115.6 124.1 120.5 118.7 120.6 118.2 121.4 121.8 110.4 108.7 108.7 118.5 121.9 123.4 123.5 126.3 116.1 122.3 123.8 123.9 115 122.8 123.2 122.2 120 124.5 120.2 125  63  Table A.1: Chemical shifts for 15N/13C wild-type tau353-408 13  Residue K385 T386 D387 H388 G389 A390 E391 I392 V393 Y394c Y394t K395t S396c S396t P397c P397t V398c V398t V399 S400 G401 D402t T403c T403t S404c S404t P405c P405t R406c R406t H407 L408  C' (ppm) 176.6 174 176.1 175.2 173.7 177.5 176.2 175.7 175.5 175.2 175.1 175.4 172.8 172.7 176.2 176.5 175.8 176.2 175.9 174.8 173.7 176.7 173.6 174.6 172.7 172.4 176 176.7 175.9 175.7 173.5 175.5  Cα (ppm) 56.29 61.55 54.08 55.75 45.24 52.23 56.34 60.96 61.81 57.6 57.98 55.3 56.42 62.85 63.03 61.7 62.21 61.87 58.23 45.12 54.23 61.48 61.52 55.66 56.86 62.69 63.14 56.22 55.82 55.08 56.66 13  Cβ (ppm) 32.85 69.74 41.05 29.01 19.18 29.97 38.28 32.67 38.81 38.81 33.36 64.09 63.02 32.12 31.99 32.79 32.46 32.64 63.72 41.01 69.96 69.4 63.87 62.93 34.22 31.91 30.61 30.69 29.43 42.83 13  1  HN (ppm) 8.292 8.093 8.275 8.429 8.451 8.123 8.345 8.139 8.1 8.308 8.377 8.05 8.032 8.282 8.242 8.157 8.281 8.414 8.428 8.208 8.046 8.174 8.136 8.257 8.576 8.342 8.507 8.143  15  N (ppm) 120.9 114.6 122.7 119.5 109.9 123.7 120.3 123.3 125.5 125.5 126 125 117 119.4 124.6 121.1 125.2 120.3 111.4 120.5 114.1 114.4 117.8 120 122 121.4 120.9 129.7  64  Table A.2: Chemical shifts for 15N/13C GlcNAc-modified tau353-408 13  Residue GlcNAc K353 I354 G355 S356 L357 D358 N359 I360 T361 H362 V363 P364 G365 G366 G367 N368 K369 K370 I371 E372 T373 H374 K375 L376 T377 F378 R379 E380 N381 A382 K383 A384 K385 T386 D387 H388  C' (ppm) 177.1 176.4 173.9 174.7 177.1 175.8 175 176.3 174.3 173.3 177.5 174.7 174.8 173.8 175.1 176.3 176.2 176 176.3 174 174.1 176.2 177.2 174 175.4 175.8 176.2 175 177.7 176.3 177.7 176.8 174 176 175.1  Cα (ppm) 61.47 44.99 58.2 55.06 54.35 53.32 61.12 61.88 55.41 63.39 45.21 45.19 45.07 53.13 56.17 56.19 60.77 56.01 61.92 55.19 56.21 55.04 61.48 57.83 55.87 56.59 53.28 52.75 56.14 52.37 56.33 61.47 54.05 55.59 13  Cβ (ppm) 38.37 63.79 41.96 40.96 38.53 38.26 69.59 29.41 33.97 31.88 38.72 32.85 32.72 38.42 30.37 69.54 29.39 32.83 42.17 69.86 39.51 30.76 30.12 38.64 18.96 32.79 19.11 32.89 69.75 41.06 28.85 13  1  HN (ppm) 8.143 8.618 8.146 8.413 8.164 8.273 7.981 8.167 8.488 7.975 8.571 8.317 8.316 8.305 8.251 8.308 8.193 8.505 8.217 8.413 8.187 8.335 8.06 8.262 8.142 8.324 8.452 8.208 8.145 8.151 8.287 8.082 8.27 8.448  15  N (ppm) 122.2 114.2 115.6 124.1 120.5 118.7 120.6 118.1 121.3 121.5 110.4 108.7 108.7 118.5 121.9 123.3 123.5 126.3 116.2 122 123.6 124 115 122.8 123.1 122.2 120 124.5 120.2 125 120.9 114.5 122.7 119.3  65  Table A.2: Chemical shifts for 15N/13C GlcNAc-modified tau353-408 13  Residue G389 A390 E391 I392 V393 Y394c Y394t K395c K395t S396c S396t P397c P397t V398c V398t V399 V399g S400 S400g G401 G401g D402t D402tg T403c T403t T403tg S404c S404t P405c P405t R406c R406t H407c H407t L408  C' (ppm) 173.7 177.5 176.2 175.7 175.5 175.1 174.9 175.4 176 176.5 176.2 175.9 175.8 174.8 174.2 173.7 176.7 176.7 173.6 174.6 176 176.7 175.9 175.7 173.4 -  Cα (ppm) 45.26 52.22 56.35 60.9 61.79 57.89 55.73 55.26 56.36 62.29 62.97 62.29 61.85 61.72 58.22 56.25 45.13 45.16 56.05 54.24 61.55 61.43 61.39 56.88 62.69 63.09 56.24 55.77 55.44 54.96 56.63 13  Cβ (ppm) 19.18 29.98 38.29 32.65 38.89 38.84 33.39 33.38 64.13 63 32.46 31.94 32.53 32.68 32.93 63.76 70.39 46.7 41.12 70.02 69.41 69.36 63.97 62.88 34.24 31.89 30.65 30.71 29.32 29.32 42.82 13  1  HN (ppm) 8.453 8.13 8.34 8.137 8.098 8.304 8.372 8.049 8.029 8.278 8.245 8.15 8.275 8.406 8.352 8.42 8.359 8.204 8.184 8.171 8.198 8.135 8.267 8.57 8.336 8.516 8.376 8.164  15  N (ppm) 109.9 123.7 120.4 123.2 125.5 125.5 126 125 117 119.4 124.6 121.2 125.2 120.2 119.5 111.3 111.7 120.5 120.5 114.4 114.5 117.8 120 122 121.4 120.8 120.4 129.8  66  Table A.3: Chemical shifts for 15N/13C S400D tau353-408 13  Residue K353 I354 G355 S356 L357 D358 N359 I360 T361 H362 V363c P364c P364t G365c G365t G366 G367 N368 K369 K370 I371 E372 T373 H374 K375 L376 T377 F378 R379 E380 N381 A382 K383 A384 K385 T386 D387  C' (ppm) 176.4 173.9 174.7 177.1 175.8 175 176.3 174.3 173.4 176.9 177.5 174.7 174.7 173.8 175.1 176.2 176.2 176 176.3 174 174.2 176.2 177.2 174 175.4 175.8 176.2 175 177.7 176.3 177.7 176.7 174 176  Cα (ppm) 61.46 44.99 58.21 55.06 54.35 53.3 61.12 61.9 55.58 59.05 58.85 63.38 45.22 45.2 45.08 53.07 56.18 56.19 60.78 56.03 61.92 55.32 57.11 55.04 61.54 57.82 55.89 56.6 53.16 52.75 56.15 52.36 56.31 61.44 54.05 13  Cβ (ppm) 38.37 63.8 41.94 40.97 38.54 38.47 69.58 29.63 34 32.66 31.89 38.72 32.86 32.79 38.41 30.36 69.54 29.63 32.83 42.17 69.93 39.52 30.75 30.14 38.64 18.97 32.81 19.12 32.9 69.76 41.07 13  1  HN (ppm) 8.616 8.146 8.411 8.16 8.267 7.985 8.17 8.453 7.958 8.579 8.583 8.318 8.318 8.304 8.247 8.305 8.192 8.506 8.215 8.385 8.168 8.312 8.051 8.252 8.14 8.325 8.448 8.204 8.142 8.149 8.283 8.084 8.271  15  N (ppm) 114.2 115.6 124.1 120.5 118.7 120.6 118.2 121.4 121.4 110.3 110.4 108.7 108.7 118.5 121.9 123.4 123.5 126.3 116.1 122.3 123.7 123.8 115 122.8 123.1 122.2 120 124.5 120.2 125 120.9 114.5 122.7  67  Table A.3: Chemical shifts for 15N/13C S400D tau353-408 13  Residue H388 G389 A390 E391 I392 V393 Y394c Y394t K395c K395t S396c S396t P397c P397t V398c V398t V399 D400 G401 D402 T403c T403t S404c S404t P405c P405t R406c R406t H407c H407t L408  C' (ppm) 175.2 173.7 177.5 176.2 175.7 175.4 175.1 174.8 175.3 176.2 176.5 176.1 175.6 176.5 173.9 176.5 173.6 174.6 176 176.7 175.9 175.8 173.5 -  Cα (ppm) 55.71 45.27 52.22 56.38 60.9 61.78 57.51 57.91 55.69 55.28 56.36 62.93 63.05 61.49 62.28 61.8 54.51 45.27 54.21 61.59 61.5 55.62 56.88 62.64 63.11 56.2 55.71 55.05 54.99 56.61 13  Cβ (ppm) 28.98 19.2 29.98 38.3 32.72 38.89 38.86 33.42 33.41 64.15 63.02 32.11 31.92 32.97 32.48 32.76 41.07 41.09 69.95 69.36 63.97 62.88 34.24 31.87 30.64 30.72 29.35 29.38 42.87 13  1  HN (ppm) 8.424 8.444 8.118 8.341 8.13 8.093 8.3 8.366 8.045 8.015 8.292 8.197 8.134 8.253 8.38 8.301 8.209 7.989 8.128 8.128 8.225 8.565 8.339 8.525 8.359 8.138  15  N (ppm) 119.4 109.9 123.7 120.3 123.1 125.4 125.5 125.9 125 117 119.5 124.1 121.1 124.9 124.8 109.6 120.6 114 114.2 117.9 119.9 122 121.4 120.8 120.3 129.7  68  Appendix B: Chemical shift table for O-GlcNAc Table B.1: Chemical shifts for 15N/13C O-GlcNAc 13  Atom C1-H1 C2-H2 C3-H3 C4-H4 C5-H5 C6-H6' C6-H6'' CH3  1 C H (ppm) (ppm) 103.269 4.418 57.93 3.587 76.424 3.421 72.432 3.329 78.48 3.344 63.33 3.642 63.31 3.82 24.831 1.907  69  ∂C'  Appendix C: Change in chemical shifts compared to predicted values 1.5 1 0.5 0 -0.5 -1 -1.5 1.0 ∂Cα  0.0 -1.0 -2.0 -3.0  10 5 0 -5 -10 -15 -20 -25  I354 G355 S356 L357 D358 N359 I360 T361 H362 V363 P364 G365 G366 G367 N368 K369 K370 I371 E372 T373 H374 K375 L376 T377 F378 R379 E380 N381 A382 K383 A384 K385 T386 D387 H388 G389 A390 E391 I392 V393 Y394t K395t S396t P397 V398t V399 S400 G401 D402t T403t S404t P405t R406t H407  ∂Cβ  -4.0  WT  GlcNAc  S400D  70  Figure C.1: Change in13C’, 13Cα, and 13Cβ chemical shift values for wild-type, O-GlcNAc-modified, and S400D tau353-408 compared to predicted chemical shift values 75.  0.20  ∂HN  0.10 0.00 -0.10 -0.20 -0.30  6.00  ∂N  4.00 2.00 0.00 -2.00 0.1  ∂Hα  0.05 0 -0.05  -0.15  I354 G355 S356 L357 D358 N359 I360 T361 H362 V363 P364 G365 G366 G367 N368 K369 K370 I371 E372 T373 H374 K375 L376 T377 F378 R379 E380 N381 A382 K383 A384 K385 T386 D387 H388 G389 A390 E391 I392 V393 Y394t K395t S396t P397 V398t V399 S400 G401 D402t T403t S404t P405t R406t H407  -0.1  WT  GlcNAc  S400D  71  Figure C.2: Change in1HN, 15N, and 1Hα chemical shift values for wild-type, O-GlcNAc-modified, and S400D tau353-408 compared to predicted chemical shift values 75.  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0071928/manifest

Comment

Related Items