PROTEIN INTERACTIONS OF MEMBRANE PROTEIN U24 FROM ROSEOLOVIRUS AND IMPLICATIONS FOR ITS FUNCTION by Yurou Sang B.Sc., Tianjin University, 2010 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Chemistry) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) March 2016 © Yurou Sang, 2016 ii Abstract This dissertation describes the investigation of the interactions between the tail-anchored membrane protein U24 from Human Herpesvirus type 6A (HHV-6A) and type 7 (HHV-7) and its potential binding partners. The roles that these interactions play in U24s’ function will be presented. It has been suggested that U24 from HHV-6A (U24-6A) may trigger an autoimmune reaction in multiple sclerosis (MS), through its molecular mimicry of myelin basic protein (MBP). Both versions of U24 have been implicated in endocytic recycling via specific binding partners, namely WW domains. The first part of this thesis is a review of the foundations that this thesis is based on, from the description of Roseoloviruses, associated diseases to molecular characterizations. The two main techniques used in the studies will be described as well. Prior to commencing any structural or interaction studies, a protocol is devised to express and purify recombinant U24 from HHV-7 (U24-7), as well as preliminary studies to prepare samples suitable for structure determination by nuclear magnetic resonance (NMR) spectroscopy. U24-6A was shown to be a mimic of MBP and it was suggested that it could be implicated in MS by competing with MBP for its interactions, such as the interactions with Fyn-SH3 domain. The interactions between U24-6A and Fyn-SH3 domain were therefore probed and found to be weak, calling into question this mimicry hypothesis. Because of the weak binding with Fyn-SH3 domain, alternative functions and binding partners were then explored. WW domains were chosen because their binding ligand, the PPxY (PY) motif, is present in U24 and was identified to be essential for U24’s function. In the next part of this thesis, the investigation of the interactions between U24s and WW domains in Nedd4, which is a key component required for endocytosis, are described. iii U24-7 and phosphorylated U24-6A were found to bind strongly to Nedd4-WW domains, suggesting the negative charge upstream from PY motif in U24 is important for high affinity interactions. Non-canonical Smurf2 WW domains were explored as well. Finally, the results presented in this thesis will be discussed in the context of the function of U24. iv Preface The details of my specific contributions to each chapter are listed below. Chapter 1 and Chapter 6 are written by myself and edited by my supervisor. Except for the figures listed below, adapted from other sources with copyright permission, these chapters are original and unpublished. Figure 1.1A is adapted from http://hhv-6foundation.org/what-is-hhv-6 (accessed on Aug 12th, 2015) and reproduced with the permission of the HHV-6A foundation. Figure 1.1B is adapted from https://en.wikipedia.org/wiki/CD3_(immunology) (accessed September 10th, 2015). It is uploaded by Anriar~commonswiki, and is assumed to be their own work based on copyright claims. It is licensed under Creative Commons Attribution-Share Alike 3.0 Unported (CC BY-SA 3.0). Figure 1.3 is adapted with permission of Nature Publishing Group from “Pathways and mechanisms of endocytic recycling”, Barth D. Grant & Julie G. Donaldson, Nature Reviews Molecular Cell Biology 10, 597-608. The license number issued by Copyright Clearance Center is 3707850678357. Figure 1.15C is adapted with permission of Springer from “Thermodynamics of protein–ligand interactions as a reference for computational analysis: how to assess accuracy, reliability and relevance of experimental data”, Stefan G. Krimmer & Gerhard Klebe, Journal of Computer-Aided Molecular Design. 29(9):867-83. The license number issued by Copyright Clearance Center is 3739221305681. Chapter 2 describes the development of methodology for obtaining recombinant membrane protein U24-7. It is written by me and edited by my supervisor, and my contributions to the experiments described in this chapter are as follows: Designed the project and experiments with my supervisor. v Conducted all of the experiments and data analysis. Prepared all the figures in the chapter. Additional contributions: Marshall Lapawa performed the mass spectrometry in section 2.2.6 using samples prepared by me. The mass spectra are shown in appendix B5. Chapter 3 is part of a co-authored, peer-reviewed journal article. It is adapted with permission from “Probing the Interaction between U24 and the SH3 Domain of Fyn Tyrosine Kinase”, Yurou Sang, Andrew R. Tait, Walter R.P. Scott, A. Louise Creagh, Prashant Kumar, Charles A. Haynes & Suzana K. Straus, Biochemistry. 53:6092–6102. Copyright 2014 American Chemical Society. My contributions are as follows: Designed the project and experiments with my supervisor. Conducted all of the experiments described in the chapter and did most of the data analysis. Prepared the figures in this chapter, except Figure 3.4. Additional contributions: The data fitting of the NMR titration data described in section 3.2.3 was done by my supervisor, Dr. Suzana K. Straus. She also prepared Figure 3.4. A manuscript based on Chapter 4 is in preparation with the following author list: Yurou Sang, Rui Zhang, Walter R.P. Scott, A. Louise Creagh, Charles A. Haynes & Suzana K. Straus. “Investigating the Interactions between U24 and WW Domains of Nedd4 E3 Ubiquitin Ligase”. This chapter is edited by my supervisor, Dr. A. Louise Creagh and Dr. Charles A. Haynes. It will be submitted to peer-reviewed journal as a co-authored article. My contributions are as follows: Designed the project and experiments with Dr. Charles A. Haynes, Dr. A. Louise Creagh and my supervisor. Conducted most of the experiments and all of the data analysis. vi Prepared the figures in this chapter. Additional contributions: The pull-down experiment in section 4.2.1, Figure 4.3B was performed by Rui Zhang using a protocol developed by me. The EXSY pulse sequence used in section 4.2.8 was generated by Dr. Mark Okon. A manuscript based on Chapter 5 is in preparation with the following author list: Yurou Sang, Rui Zhang, Walter R.P. Scott, A. Louise Creagh, Charles A. Haynes & Suzana K. Straus. “Exploration of the Potential Interactions between U24 and the Human Smurf2 WW Domains”. This chapter is edited by my supervisor, Dr. A. Louise Creagh and Dr. Charles A. Haynes. It will be submitted to a peer-reviewed journal as a co-authored article. My contributions are as follows: Designed the project and experiments with Dr. Charles A. Haynes, Dr. A. Louise Creagh and my supervisor. Conducted all experiments and all of the data analysis. Prepared the figures in this chapter. vii Table of Contents Abstract ................................................................................................................. ii Preface ...................................................................................................................iv Table of Contents .................................................................................................vii List of Tables...................................................................................................... xiii List of Figures ......................................................................................................xv List of Abbreviations .......................................................................................... xxi Acknowledgements ..........................................................................................xxvii Dedication ......................................................................................................... xxx Chapter 1 Introduction .......................................................................................... 1 1.1 Roseoloviruses and U24..................................................................................................................... 1 1.1.1 HHV-6 .......................................................................................................................................... 3 1.1.2 HHV-7 .......................................................................................................................................... 3 1.1.3 Membrane protein U24 ........................................................................................................... 4 220.127.116.11 Endosomes ..................................................................................................................... 6 18.104.22.168 Phosphorylation of U24 ............................................................................................ 7 22.214.171.124 U24 contains a proline-rich region.......................................................................... 8 126.96.36.199 Binding domains for the proline-rich region in U24 ......................................... 9 1.2 Possible disease associations of U24 ........................................................................................... 11 1.2.1 U24 and multiple sclerosis.................................................................................................... 13 1.2.2 Mimicking MBP....................................................................................................................... 15 188.8.131.52 Myelin basic protein................................................................................................... 15 184.108.40.206 Posttranslational modification of MBP............................................................... 16 220.127.116.11 Mimicking the interaction of MBP....................................................................... 17 1.2.3 U24 may be implicated in MS by blocking endosomal recycling ............................. 17 18.104.22.168 Minor myelin proteins............................................................................................... 18 viii 22.214.171.124 Nogo-A, a neurite outgrowth inhibitor ............................................................... 18 126.96.36.199 Sodium channels in neurons ................................................................................... 20 188.8.131.52 Evidence of a connection? ...................................................................................... 21 1.3 Potential binding partners of U24 ................................................................................................ 23 1.3.1 Fyn-SH3 domain ..................................................................................................................... 23 184.108.40.206 Architecture of Fyn tyrosine kinase ..................................................................... 24 220.127.116.11 Important role of Fyn in CNS ............................................................................... 25 1.3.2 WW domains in Nedd4 family E3 ubiquitin ligase ...................................................... 26 18.104.22.168 Nedd4 family E3 ubiquitin ligase .......................................................................... 27 22.214.171.124 Functions of Nedd4 and Nedd4L in the CNS ................................................. 29 126.96.36.199 E3 ubiquitin ligase ...................................................................................................... 31 1.4 Methods used to characterize protein interactions .................................................................. 33 1.4.1 Protein NMR spectroscopy ................................................................................................. 33 1.4.2 Quantifying protein interactions using NMR spectroscopy ...................................... 35 1.4.3 Isothermal titration calorimetry.......................................................................................... 40 1.5 Aims of the dissertation .................................................................................................................. 42 Chapter 2 Expression and Purification of U24-7, towards Preliminary Structural and Phosphorylation Studies .............................................................. 44 2.1 Introduction ......................................................................................................................................... 44 2.2 Results .................................................................................................................................................... 46 2.2.1 Cloning and expression optimization ............................................................................... 46 2.2.2 Optimization of U24-7 purification.................................................................................. 49 188.8.131.52 Elimination of the precipitation step................................................................... 50 184.108.40.206 Purification buffer optimization ............................................................................ 52 220.127.116.11 Salt induced acetone precipitation......................................................................... 54 2.2.3 Secondary structure ................................................................................................................ 57 2.2.4 A double Cys to Ser mutation of U24-7 and its CD spectra .................................... 59 2.2.5 Preliminary NMR results ...................................................................................................... 61 2.2.6 Phosphorylation of U24-7 with MAP kinase ................................................................ 64 ix 2.3 Discussion ............................................................................................................................................ 66 2.4 Conclusions .......................................................................................................................................... 69 2.5 Material and methods........................................................................................................................ 70 2.5.1 Gene synthesis ......................................................................................................................... 70 2.5.2 Cloning ....................................................................................................................................... 70 2.5.3 Small-scale protein expression ............................................................................................ 72 2.5.4 Large-scale protein expression ............................................................................................ 72 2.5.5 Protein extraction and purification .................................................................................... 73 2.5.6 SDS-PAGE analysis ................................................................................................................ 75 2.5.7 Acetone precipitation tests and large scale protein isolation ..................................... 76 2.5.8 CD experiment......................................................................................................................... 77 2.5.9 Site-directed mutagenesis...................................................................................................... 78 2.5.10 1H-15N HSQC NMR spectroscopy .................................................................................... 78 Chapter 3 Probing the Interactions between U24-6A or U24-7 and the SH3 Domain of Fyn Tyrosine Kinase ......................................................... 80 3.1 Introduction ......................................................................................................................................... 80 3.2 Results .................................................................................................................................................... 82 3.2.1 Pull-down assays with GST-Fyn-SH3 and U24-7 protein.......................................... 82 3.2.2 ITC Experiments..................................................................................................................... 83 3.2.3 Titration of U24 peptides by NMR .................................................................................. 85 3.2.4 A shorter version of U24-6A peptide may increase the affinity .............................. 89 3.3 Discussion ............................................................................................................................................ 91 3.4 Conclusions .......................................................................................................................................... 95 3.5 Materials and methods...................................................................................................................... 95 3.5.1 GST-SH3 domain pull-down assays with recombinant U24-7 protein.................. 95 3.5.2 Peptide synthesis and purification...................................................................................... 96 3.5.3 ITC experiments ...................................................................................................................... 97 3.5.4 1H-15N HSQC NMR titration experiment....................................................................... 97 x Chapter 4 Investigating the Interactions between U24 and WW Domains of Nedd4 E3 Ubiquitin Ligase ................................................................ 99 4.1 Introduction ......................................................................................................................................... 99 4.2 Results ..................................................................................................................................................103 4.2.1 Pull-down assays with GST-Nedd4-WW domains and U24 proteins ..................103 4.2.2 Affinities between Nedd4-WW domains and U24 peptides ...................................105 4.2.3 The effect of temperature on U24-WW domain interactions ................................108 4.2.4 Phosphorylation on threonine of U24-6A enhances the affinities with WW domains. ...................................................................................................................................110 4.2.5 Electrostatic effect of U24 binding to Nedd4-WW domains .................................113 4.2.6 Binding site residues .............................................................................................................116 4.2.7 Long-range NOEs found in the U24-WW domain complexes..............................119 4.2.8 Slow-exchange of pU24-6A and hNedd4L-WW3* complex..................................121 4.3 Discussion ..........................................................................................................................................125 4.4 Conclusions ........................................................................................................................................134 4.5 Materials and methods....................................................................................................................135 4.5.1 Plasmid construction of GST fusion human WW3* domains...............................135 4.5.2 Expression of GST fusion WW domains .....................................................................136 4.5.3 Purification of GST tagged Nedd4 WW domains .....................................................137 4.5.4 Purification of Nedd4-WW domains .............................................................................137 4.5.5 CD melt experiment .............................................................................................................138 4.5.6 GST tagged WW domains pull-down assays with recombinant U24-6A and U24-7 protein .........................................................................................................................139 4.5.7 ITC experiments ....................................................................................................................139 4.5.8 1H-15N NMR experiments of WW domains ................................................................140 Chapter 5 Exploration of the Potential Interactions between U24 and the Human Smurf2 WW Domains ....................................................................... 143 5.1 Introduction .......................................................................................................................................143 5.2 Results ..................................................................................................................................................146 xi 5.2.1 Pull-down assays with GST-hSmurf2-WW domains and U24 proteins ..............146 5.2.2 Affinities between hSmurf2-WW domains and U24 peptides ................................148 5.2.3 Investigation of the cooperativity in the tandem WW domains of hSmurf2 upon binding to U24 peptides ..........................................................................................149 5.3 Discussion ..........................................................................................................................................152 5.4 Conclusions ........................................................................................................................................156 5.5 Materials and methods....................................................................................................................157 5.5.1 Expression and purification of GST tagged hSmurf2 WW domains ..................157 5.5.2 Purification of hSmurf2 WW domains ..........................................................................158 5.5.3 GST tagged hSmurf2 WW domains pull down experiments with recombinant U24-6A and U24-7 protein ................................................................................................158 5.5.4 ITC experiments ....................................................................................................................159 Chapter 6 Conclusions and Future Perspectives ............................................... 160 6.1 Thesis summary ................................................................................................................................160 6.2 Conclusions ........................................................................................................................................162 6.3 Future perspectives ..........................................................................................................................164 Bibliography ....................................................................................................... 168 Appendices ......................................................................................................... 194 A Examples of SDS-PAGE and HPLC traces..............................................................................194 A1 U24 protein gel images ...........................................................................................................194 A2 Fyn-SH3 and WW domain gel examples ..........................................................................195 A3 HPLC traces for U24-6A and U24-7 peptide purifications ........................................196 B Mass spectrometry results ...............................................................................................................198 B1 MALDI-TOF MS spectra of purified peptides ..............................................................198 B2 LC-MS/MS spectra of purified peptides ..........................................................................201 B3 MALDI-TOF MS spectra for purified recombinant U24-7 protein ........................204 B4 MALDI-TOF MS study for thrombin linkage recombinant U24-7.........................205 B5 MALDI-TOF MS spectra for phosphorylation studies of U24-7 ............................208 xii B6 MALDI-TOF MS spectra for Fyn-SH3 domain and WW domains ........................210 C NMR chemical shift tables and additional NMR spectra ......................................................215 C1 Chemical shift assignment of Fyn-SH3 domain apo and bound to U24-6A ........215 C2 Chemical shift assignment of Nedd4-WW domain apo and bound with U24 peptide (U24-7 is listed as an example)..............................................................................216 C3 U24-7 COSY spectrum...........................................................................................................218 C4 Thermal stability of hNedd4-WW3* and hNedd4L-WW3* domains ....................219 C5 EXSY of pU24-6A/hNedd4L-WW3* complex at 25 °C ...........................................220 C6 R492 regions of EXSY, pU24-6A/hNedd4L-WW3* complex at 15 °C ................221 C7 HSQC overlays of rNedd4-WW3/4 and hNedd4L-WW3* domains: apo form and in complex with U24-6A, U24-7 or pU24-6A peptide. ........................................222 C8 Example of HSQC assignment using 3D HSQC-NOESY and 3D HSQC-TOCSY spectra .........................................................................................................................224 D Additional CD results .......................................................................................................................225 D1 CD spectra of U24-7 protein ...............................................................................................225 D2 An example of CD melt result, rNedd4-WW3/4 domain ..........................................226 E Additional ITC results ......................................................................................................................227 E1 ITC results for Fyn-SH3 domain studies ..........................................................................227 E2 Parameters from ITC experiments for Chapter 4 ..........................................................229 E3 Electrostatic effect on U24 binding to hNedd4-WW3*...............................................231 E4 Parameters from ITC experiments for Chapter 5 ..........................................................232 F Additional figures ...............................................................................................................................233 xiii List of Tables Table 1.1 Proline containing motifs and their binding domains.................................................................... 11 Table 1.2 A summary of the functions of Nedd4 and Nedd4L in the CNS ............................................ 30 Table 2.1 Oligonucleotides used for U24-7 gene synthesize. ......................................................................... 49 Table 2.2 U24-7 concentrations, after recovery and the ratio of the amount recovered relative to the starting concentration of 46.4 µM. ...................................................................................................... 55 Table 2.3 Mass over charge (m/z) of largest molecular weight peak determined in MALDI-TOF MS, upon exposure of U24-7 to MAP kinase. ................................................................................. 65 Table 2.4 Salt stock solutions used in the salt induced acetone precipitation tests ................................. 77 Table 2.5 Primers that was used in site-directed mutagenesis ........................................................................ 78 Table 3.1 Parameters for U24-6A binding to Fyn-SH3 at 25°C .................................................................... 85 Table 3.2 Parameters for U24-6AΔMD binding to Fyn-SH3 at 25°C ............................................................. 91 Table 4.1 Parameters obtained from fitting the ITC data for binding of U24-6A or U24-7 to rNedd4-WW2, rNedd4-WW3/4, hNedd4-WW3* and hNedd4L-WW3* domains at 25 °C. ...................................................................................................................................................................106 Table 4.2 Change in heat capacities upon binding of U24-6A or U24-7 to rNedd4-WW3/4 or hNedd4L-WW3* domains....................................................................................................................110 Table 4.3 Parameters obtained from fitting the ITC data for binding of pU24-6A to rNedd4-WW3/4 and hNedd4L-WW3* domains at 25 °C. ........................................................................113 Table 4.4 Preliminary data fits of pU24-6A peptide binding to hNedd4L-WW3* domain. R2 of the fits are reported to show the goodness of fit.................................................................................125 Table 4.5 Primers used for plasmid construction of human WW3* domains........................................135 Table 5.1 Alignment of canonical and uncanonical WW domains. The conserved tryptophans, or substituted tyrosine or phenylalanines, are highlighted in bold. ...............................................145 Table 5.2 Parameters obtained from fitting the ITC data for binding of U24-6A, pU24-6A or U24-7 binding to hSmurf2-WW3 and hSmurf-WW2+3 domains at 25 °C. ......................................149 Table 5.3 Changes in the heat capacities for binding of U24-6A, pU24-6A or U24-7 to hSmurf2-WW3 or hSmurf2-WW2+3 domains. ...............................................................................................150 xiv Table 5.4 Estimated binding area of U24 peptide binding to hSmurf2-WW domains, obtained using Equation (5.1). ..........................................................................................................................................154 Table B.1 Amino acid sequences and theoretical masses of U24-7 and digested fragments. The theoretical MWs were computed using the ProtParam tool. (303) ..........................................206 Table B.2 Amino acid sequences and theoretical MW of SH3 and WW domains. ...............................210 Table C.1 Chemical shift assignment of Fyn-SH3 domain apo and bound to U24-6A (protein to peptide ratio is 1:12.68)..........................................................................................................................215 Table C.2 Chemical shift assignments of rNedd4-WW3/4 domain, hNedd4L-WW3* domain, both in the apo form and bound to U24-7 peptide (protein to peptide ratio is 1:4 in the case of rNedd4-WW3/4 domain, 1:2 in the case of hNedd4L-WW3* domain). .............................216 Table E.1 Thermodynamic parameters obtained from fitting ITC data of U24-6A or U24-7 binding to rNedd4-WW3/4 and hNedd4L-WW3* domain at 15 °C and 37 °C in 10 mM sodium phosphate, pH 7.4. ..................................................................................................................................229 Table E.2 Thermodynamic parameters obtained from fitting ITC data for U24-6A, pU24-6A or U24-7 binding to rNedd4-WW3/4, hNedd4-WW3* and hNedd4L-WW3* domains at 25 °C in 10 mM sodium phosphate pH 7.4, with additional an 100 mM NaCl or 500 mM NaCl.............................................................................................................................................................230 Table E.3 Thermodynamic parameters obtained from fitting ITC data of U24-6A, pU24-6A or U24-7 binding to hSmurf2-WW3 and hSmurf2-WW2+3 domains at 5 °C and 15 °C in 40 mM HEPES, 10 mM NaCl pH 7.2. ...................................................................................................232 xv List of Figures Figure 1.1 Virus and T-cell receptor: A) Illustration of the envelope of Roseoloviruses. B) Illustration of the T-cell receptor complex (TCR) found on the surface of T-cells. ......... 2 Figure 1.2 Sequence alignment of U24-6A, U24-6B and U24-7 ................................................................... 5 Figure 1.3 Adapted with permission from Pathways and mechanisms of endocytic recycling, Barth D. Grant & Julie G. Donaldson, Nature Reviews Molecular Cell Biology 10, 597-608 ............... 7 Figure 1.4 X-Proline bond isomers and polyproline type I and type II helices. ....................................... 8 Figure 1.5 Healthy and diseased myelin............................................................................................................... 12 Figure 1.6 Alignment of MBP and U24s ............................................................................................................ 14 Figure 1.7 A summary of how U24 may play a part in MS. Figure 1.1A is shown here to represent the Roseoloviruses. ................................................................................................................................ 15 Figure 1.8 Topologies of Nogo-A in the plasma and tubular ER membrane (150) ............................. 19 Figure 1.9 The structure of SH3 domain and proline-rich peptide complex.......................................... 24 Figure 1.10 The structure of a WW domain and ligand containing a PY motif in complex ............... 27 Figure 1.11 Ubiquitination: types, mechanisms, and the structure of ubiquitin ....................................... 32 Figure 1.12 HSQC, coupled HSQC and TROSY-HSQC spectra of small (upper row) and large (bottom row) proteins .......................................................................................................................... 34 Figure 1.13 Protein dynamics probed by NMR spectroscopy. ....................................................................... 36 Figure 1.14 Slow exchange of Apo (A) and Bound protein (B) using EXSY. ........................................... 39 Figure 1.15 ITC data, fitting examples and schematic depiction of ITC.................................................... 41 Figure 2.1 Schematic representation of Mbp-6xHis-U24-7 protein .......................................................... 47 Figure 2.2 Expression tests of U24-7 in pMal-c2x or pMal-p2x vectors, expressed in different strains ......................................................................................................................................................... 48 Figure 2.3 Overall purification process of U24-7 and optimizations done. The steps of U24-7 protein purification are marked in roman numerals from I to VII. The green ovals indicate the sections where various optimizations are described............................................ 50 Figure 2.4 TEV protease cleavage results of Mbp-6xHis-U24-7 in buffers with different detergents. ............................................................................................................................................... 51 xvi Figure 2.5 U24-7 isolation using different buffers for anion-exchange chromatography.................... 54 Figure 2.6 The first six anions and cations in the Hofmeister series.......................................................... 55 Figure 2.7 Supernatants from salt induced acetone precipitation. .............................................................. 57 Figure 2.8 CD spectra of purified recombinant U24-7 in DHPC and detergents. ............................... 58 Figure 2.9 CD spectra of U24-7 dCS protein in citrate-phosphate buffer, 50 mM SDS, 150 mM DPC, pH5.8. ............................................................................................................................................ 60 Figure 2.10 CD spectra of U24-7 dCS in citrate-phosphate buffer, 25 mM SDS, pH 5.8 at different temperatures. ........................................................................................................................................... 61 Figure 2.11 seHSQC spectra of U24-7 dCS protein in citrate-phosphate buffer, pH 5.8 with 25 mM SDS, at different temperatures, recorded on an 850 MHz spectrometer. ............................ 63 Figure 2.12 The A) HSQC, B) seHSQC, C)TROSY-HSQC and D) overlays of above three spectra for U24-7 dCS protein at 45 °C......................................................................................................... 64 Figure 2.13 Transmembrane region of U24-7 shown in helical wheels and wenxiang diagram.......... 68 Figure 3.1 Sequence alignment of MBP, the first 15 residues of U24-6A and U24-7. ........................ 80 Figure 3.2 SDS-PAGE result of GST pull-down experiment with GST-Fyn-SH3 and U24-7. ........ 83 Figure 3.3 ITC data for U24-6A peptide (U24-6A1-15) binding to Fyn-SH3 at 25 °C. .......................... 84 Figure 3.4 Overlays of 15N-1H HSQC spectra of Fyn-SH3 with: a) U24-6A peptide, b) U24-7 peptide added in different ratios. c) The fit of the chemical shifts of U24-6A titration. ... .................................................................................................................................................................... 87 Figure 3.5 Chemical shift perturbations of Fyn-SH3 for a) U24-6A peptide and b) U24-7 peptide NMR titration. ........................................................................................................................................ 88 Figure 3.6 ITC and NMR titration results of U24-6AΔMD (U24-6A3-15) and Fyn-SH3 domain. ........ 90 Figure 3.7 Sequence alignment of Class I and II PxxP motif ligands: U24-6A, U24-6B, MBP, NS5A polyproline region, and three peptides derived from the PI3-kinase polyproline region. ........................................................................................................................................................ 92 Figure 3.8 Structural model of Fyn-SH3 and PI3-kinase peptide complex (PDB:1A0N).................. 94 Figure 4.1 Sequence alignment of the first 15 residues of U24-6A, U24-7 and the PY motifs found in ENaC, Comm protein and the HECT domain of hNedd4L-WW3*. ...........................101 Figure 4.2 Schematic architectures of Nedd4 and Nedd4L protein. .......................................................102 xvii Figure 4.3 SDS-PAGE result of GST pull-down experiment with GST-Nedd4-WW domains and U24 ...........................................................................................................................................................104 Figure 4.4 ΔH plots of U24-6A or U24-7 titrated into the four Nedd4-WW domains.....................107 Figure 4.5 Plots of ΔH vs. temperature for U24-6A or U24-7 titrated into A) rNedd4-WW3/4 and B) hNedd4L-WW3* domains. .........................................................................................................109 Figure 4.6 Sequence alignment of the first 15 residues of U24-6A, pU24-6A and U24-7. ..............111 Figure 4.7 pU24-6A peptide interaction with WW domains at 25 °C A) ITC data with hNedd4L-WW3*. ΔH plot for the three peptide ligands interacting with B) rNedd4-WW3/4 or C) with hNedd4L-WW3* domain. .......................................................................................................112 Figure 4.8 ΔH and ln Ka plots of the three U24 ligands binding to rNedd4-WW3/4 and hNedd4L-WW3* in phosphate buffer, phosphate buffer with additional 100 mM NaCl or 500 mM NaCl at 25 °C. .......................................................................................................................................115 Figure 4.9 Combined chemical shift perturbations mapped on A) rNedd4-WW3/4 and D) hNedd4L-WW3* domain when complexed with U24 peptides. HSQC overlays for B-C) rNedd4-WW3/4 and E-G) hNedd4L-WW3*. ...........................................................................119 Figure 4.10 Schematic figure of NOEs found upon U24 peptide binding to A) rNedd4-WW3/4 and B) hNedd4L-WW3* domain ....................................................................................................120 Figure 4.11 Region of seHSQC, HSQC and EXSY spectra of pU24-6A binding to hNedd4L-WW3* complex at 15 °C ...................................................................................................................123 Figure 4.12 Structural model of rNedd4-WW3/4 and hNedd4L-WW3* domains, coloured using the color-coding scheme described in the text. ..........................................................................130 Figure 4.13 A close-up view of hNedd4L-WW3* domain model bound with pU24-6A peptides. .132 Figure 5.1 Sequence alignment of first 15 residues of U24-6A, U24-7 and the core PY motif of Smad7. .....................................................................................................................................................144 Figure 5.2 SDS-PAGE result of GST pull-down experiment with GST-hSmurf2-WW domains and U24 protein. ...........................................................................................................................................147 Figure 5.3 Plots of ΔH vs. temperature for the three U24 ligands titrated into A) hSmurf2-WW3 and B) hSmurf2-WW2+3. The same data were plotted again for each peptide, with C) U24-6A, D) pU24-6A or E) U24-7 titrated into the two hSmurf2 domains.....................151 Figure A.1 SDS-PAGE of U24-7 purification and U24-6A and U24-7 gel staining with silver or Coomassie stain. M indicates the marker and MWs are marked beside the gel. ..............194 xviii Figure A.2 Example gel images of GST fusion protein purification, including Fyn-SH3 domain and several WW domains. M indicates the marker and the MWs are marked beside the gel. ... .................................................................................................................................................................195 Figure A.3 HPLC gradient table and sample traces for the purification of U24-6A peptide. The purification process includes two sequential HPLC runs: traces of A) 1st run and B) 2nd run are shown. ......................................................................................................................................196 Figure A.4 HPLC gradient table and sample traces for the purification of U24-7 peptide. The purification process includes two sequential HPLC runs: traces of A) 1st run and B) 2nd run are shown. ......................................................................................................................................197 Figure B.1 MALDI-TOF MS spectra of U24-6A peptide. The theoretical MW of U24-6A peptide is 1685.9. The spectra from linear (top) and reflectron (bottom) mode are shown. The zoomed-in regions of the reflectron spectra are shown as an inset (bottom)..................199 Figure B.2 MALDI-TOF MS spectra of U24-7 peptide. The theoretical MW of U24-7 peptide is 1731.9. The spectra from linear (top) and reflectron (bottom) mode are shown. The zoomed-in regions of the reflectron spectra are shown as an inset (bottom)..................200 Figure B.3 LC-MS/MS spectra of U24-6A peptide. The theoretical MW of U24-6A peptide is 1685.9. .....................................................................................................................................................202 Figure B.4 LC-MS/MS spectra of U24-7 peptide. The theoretical MW of U24-7 peptide is 1731.9. .................................................................................................................................................................203 Figure B.5 MALDI-TOF MS spectra of U24-7 protein. The theoretical MW of U24-7 is 9464.0. .... .................................................................................................................................................................204 Figure B.6 MALDI-TOF MS spectra of U24-7 dCS protein. The theoretical MW of U24-7 dCS is 9431.9. .....................................................................................................................................................205 Figure B.7 A) Schematic representation of Mbp-6xHis-U24-7 fusion protein and MALDI-TOF MS spectra of U24-7 protein sample after cleavage using B) Factor Xa and C) thrombin. .................................................................................................................................................................207 Figure B.8 in vitro phosphorylation of U24-7 by MAP kinase detected by MALDI-TOF MS. .......208 Figure B.9 in vitro phosphorylation of U24-6A and MBP by MAP kinase, as detected by MALDI-TOF MS. .................................................................................................................................................209 Figure B.10 MALDI-TOF MS spectra of Fyn-SH3 domain. The theoretical MW of Fyn-SH3 domain is 8016.6. .................................................................................................................................211 xix Figure B.11 MALDI-TOF MS spectra of rNedd4-WW2 domain. The theoretical MW of rNedd4-WW2 domain is 8120.9. Sequence of fragment (6115.2) is listed in Table B.2................211 Figure B.12 MALDI-TOF MS spectra of rNedd4-WW3/4 domain. The theoretical MW of rNedd4-WW3/4 domain is 5659.2.................................................................................................212 Figure B.13 MALDI-TOF MS spectra of hNedd4-WW3* domain. The theoretical MW of hNedd4-WW3* domain is 5195.8. ...................................................................................................................212 Figure B.14 MALDI-TOF MS spectra of hNedd4L-WW3* domain. The theoretical MW of hNedd4L-WW3* domain is 5240.9................................................................................................213 Figure B.15 MALDI-TOF MS spectra of hSmurf2-WW2 domain. The theoretical MW of hSmurf2-WW2 domain is 4962.4. .....................................................................................................................213 Figure B.16 MALDI-TOF MS spectra of hSmurf2-WW3 domain. The theoretical MW of hSmurf2-WW3 domain is 4611.0. The spectra are from reflectron mode...........................................214 Figure B.17 MALDI-TOF MS spectra of hSmurf2-WW2+3 domain. The theoretical MW of hSmurf2-WW2+3 domain is 10104.0............................................................................................214 Figure C.1 The fingerprint region from the COSY spectrum of wild type U24-7 protein. ..............218 Figure C.2 The overlays of HSQC spectra of hNedd4-WW3* and hNedd4L-WW3* domain at different temperatures, from 15 °C to 40 °C. .............................................................................219 Figure C.3 Region of seHSQC, HSQC and EXSY spectra of pU24-6A binding to hNedd4L-WW3* complex at 25 °C. ..................................................................................................................220 Figure C.4 R492 regions of seHSQC, HSQC and EXSY spectra of pU24-6A binding to hNedd4L-WW3* at 15 °C. ....................................................................................................................................221 Figure C.5 HSQC overlays of rNedd4-WW3/4 domain, apo (black), in complex with U24-6A (blue), with U24-7 (red) and with pU24-6A (green). ................................................................222 Figure C.6 HSQC overlays of hNedd4L-WW3* domain, apo (black), in complex with U24-6A (blue), with U24-7 (red) and with pU24-6A (green). ................................................................223 Figure C.7 Example assignment of apo rNedd4-WW3/4 domain using 3D HSQC-NOESY and 3D HSQC-TOCSY spectra. .............................................................................................................224 Figure D.1 CD spectra of U24-7 dCS protein in citrate-phosphate buffer, 200 mM SDS, pH 5.8. ..... ................................................................................................................................................................225 xx Figure D.2 CD melt for rNedd4-WW3/4 domain in 10 mM sodium phosphate buffer, pH 7.4, from 5 to 95 °C.....................................................................................................................................226 Figure E.1 Integrated heat plot from U24-6A and U24-6AΔMD peptides binding to Fyn-SH3 domain.....................................................................................................................................................227 Figure E.2 Typical ITC data obtained for U24-7 binding to Fyn-SH3 domain.....................................228 Figure E.3 ΔH and ln Ka plots of U24-6A and U24-7 ligands binding to hNedd4-WW3* in phosphate buffer, with an additional 100 mM NaCl or 500 mM NaCl at 25 °C.............231 Figure F.1 Alignment of Nedd4-WW, Smurf-WW, KIBRA-WW and other studied WW domains that have strong affinities with PY motif ligands. .....................................................................233 Figure F.2 Alignment of PY motif ligands. .....................................................................................................234 Figure F.3 90 degree view of structural model of rNedd4-WW3/4 and hNedd4L-WW3* domain, shown in Figure 4.12...........................................................................................................................235 xxi List of Abbreviations 6xHis hexahistidine AIDS acquired immune deficiency syndrome Alix apoptosis-linked gene 2 (ALG-2)-interacting protein AMPA Amino-3-hydroxy-5-methyl-isoxazole-4-propionic acid APC/C complex anaphase-promoting complex ARRDC arrestin domain-containing protein BCA bicinchoninic acid BME β-mercaptoethanol BSA bovine serum albumin CAT cationic amino acid transporter Cbl-b Casitas B-lineage Lymphoma-b CCA α-cyano-4-hydroxycinnamic acid CD circular dichroism CD[number] cluster of differentiation [number] CHAPS 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate CID collision-induced dissociation CMC critical micellar concentration CMV cytomegalovirus CNS central nervous system CRL cullin-RING E3 ligase CRISPR clustered regularly interspaced short palindromic repeats CSF cerebrospinal fluid Csk C-terminal Src kinase CaV1.[number] L type, high voltage activated (HVA) calcium channel member [number] Da Dalton DAT dopamine transporter DHB 2,5-dihydroxy benzoic acid xxii DHPC 1,2-dihexanoyl-sn-glycero-3-phosphocholine DSC differential scanning calorimetry Doc sodium deoxycholate DPC n-dodecylphosphocholine dsDNA double-stranded deoxyribonucleic acid DTT dithiothreitol DUB deubiquitinase E. coli Escherichia coli E1 E1 ubiquitin-activating enzyme E2 E2 ubiquitin-conjugating enzyme E3 E3 ubiquitin ligase EAE experimental autoimmune encephalomyelitis EBV Epstein-Barr virus ECM extracellular matrix EDT 1,2-ethanedithiol ER endoplasmic reticulum ESCRT endosomal sorting complex required for transport ETD Electron transfer dissociation EVH1 domain Enabled/VASP Homology 1 domain EXSY exchange spectroscopy Fmoc 9-fluorenylmethyloxycarbonyl GAP GTPase-activating protein GLB Glasgow lysis buffer GLT glutamate transporter golli gene of the oligodendrocyte lineage GST glutathione S-transferase GYF domain Gly-Tyr-Phe domain HA tag hemagglutinin epitope tag HBTU O-(benzotriazol-1-yl)-N,N,N',N'-tetramethyluronium hexafluorophosphate xxiii HECT homologous to the E6-AP carboxyl terminus domain HECW HECT, C2 and WW domain-containing ubiquitin ligase HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid HERC HECT domain and regulator of chromosome condensation 1-like domains HHV-[number] Human herpesvirus type [number] HIV Human immunodeficiency virus HPLC high-performance liquid chromatography Hrs Hepatocyte growth factor-regulated tyrosine kinase substrate HSQC heteronuclear single quantum coherence spectroscopy HSV Herpes simplex virus ILV intralumenal vesicle INEPT insensitive nuclei enhanced by polarization transfer IPTG Isopropylthiogalactoside ITC isothermal titration calorimetry KCNQ[number] potassium voltage-gated channel KQT-like subfamily, member [number] KSHV Kaposi’s sarcoma-associated herpesvirus KV1.[number] potassium voltage-gated channel, shaker-related subfamily, member [number] LB broth Luria Bertani broth LC-MS/MS liquid chromatography–tandem mass spectrometry LMP2A latent membrane protein 2A MAG myelin-associated glycoprotein MALDI-TOF MS matrix-assisted laser desorption ionization time of fly mass spectrometry MAP kinase mitogen-activated protein kinases MBP myelin basic protein Mbp maltose-binding protein † † Both myelin basic protein and maltose-binding protein are referred as “MBP” in the literature. In order to distinguish these two proteins in this dissertation, abbreviations with the same letters but different cases are used. “Mbp” indicates maltose-binding protein. xxiv MDV Marek’s disease virus MES 2-(N-morpholino)ethanesulfonic acid MG myasthenia gravis MOG myelin-oligodendrcyte glycoprotein MOM mitochondrial outer membrane MTLE mesial temporal lobe epilepsy MVB multivesicular body MWCO molecular weight cut-off m/z mass over charge NaV1.x sodium channel, voltage-gated, family 1, type x NDFIPs Nedd4 family-interacting proteins Nedd[number] neural precursor cell expressed developmentally down-regulated protein [number] NEDL Nedd4-like ubiquitin protein ligase NGF nerve growth factor NK cell natural killer cell NKT cell natural killer T cell NMO neuromyelitis optica NMR nuclear magnetic resonance NOESY nuclear Overhauser enhancement spectroscopy Nogo-A neurite outgrowth inhibitor A NPC neural precursor cell OD optical density OPCs oligodendroglial progenitor cells PBS phosphate-buffered saline PCR polymerase chain reaction PDGF platelet-derived growth factor PET photoinduced electron transfer PHD finger plant homeodomain finger Pi prolyl isomerases xxv PI3-kinase phosphatidylinositol 3-kinase PIPES piperazine-N,N’-bis(2-ethanesulfonic acid) PKC protein kinase C PLP proteolipid protein PMA phorbol-12-myristate-13-acetate PNS peripheral nervous system PPI helix polyproline type I helix PPII helix polyproline type II helix PTEN tensin homolog Rab Ras superfamily of monomeric G proteins RING really interesting new gene domain rpm revolutions per minute RRMS relapsing-remitting MS SA sinapinic acid SCF complex Skp, Cullin, F-box containing complex SDS-PAGE sodium dodecyl sulfate polyacrylamide gel electrophoresis seHSQC gradient enhanced & sensitivity enhanced HSQC SGK serum- and glucocorticoid-induced kinase SFK Src family of non-receptor tyrosine kinase SH[number] domain Src homology [number] domain Smad sma and mad (mother against decapentaplegic), two homolog gene Smurf Smad ubiquitin regulatory factor SNARE complex Snap (Soluble N-ethylmaleimide-sensitive factor attachment protein) receptor complex ssDNA single-stranded deoxyribonucleic acid Su(dx) the suppressor of deltex TCA trichloroacetic acid TCEP tris(2-carboxyethyl)phosphine TCR T-cell receptor TES triethylsilane xxvi TEV protease Tobacco Etch Virus protease TFA trifluoroacetic acid TGF-β transforming growth factor-β TMR telomeric repeats TNF-α tumour necrosis factor alpha TOCSY total correlation spectroscopy Trc40 transmembrane domain recognition complex 40 Tris Tris(Hydroxymethyl)aminomethane TROSY transverse relaxation optimized spectroscopy UBD Ubiquitin-binding domain UEV domain ubiquitin E2 variant domain USP ubiquitin-specific protease UV-Vis ultraviolet–visible Vps[number] vacuolar protein sorting [number] WWP WW domain containing E3 ubiquitin protein ligase β2AR β2-adrenergic receptor xxvii Acknowledgements My thesis work would not be finished without the help, advice and support from numerous individuals. I would like to express my sincere gratitude to these people here. First of all, I would like to thank my supervisor, Dr. Suzana K. Straus, for the opportunity to work on this challenging project, and her guidance and patience over the past few years. She has encouraged me to explore different directions in my thesis work, as well as helped me to develop my scientific thinking and spirit of collaboration, when tackling challenging aspects of the project. I truly appreciate her mentorship along the years. I also would like to thank Dr. Jin-Feng Wang for his recommendation that I pursue graduate studies. I want to express my gratitude to Dr. Charles A. Haynes and Dr. A. Louise Creagh for their collaboration on the ITC experiments used to study U24 protein interactions and their edits of Chapter 4 and 5. Their valuable advice and suggestions enabled me to characterize and understand the energetic properties of U24 protein interactions. It is my privilege to have this opportunity to collaborate with them on these projects. I would like to thank the members of my supervisory committee, Dr. Hongbin Li, Dr. Elliott Burnell and Dr. Marco A. Ciufolini for their kind and useful suggestions during the comprehensive exam and committee meeting. I would also like to thank Dr. Li in particular for reading through this thesis. Next, I would like to thank Dr. Lawrence P. McIntosh and Dr. Mark Okon for the use of the high field NMR spectrometers, for their wisdom in editing pulse programs and setting up experiments. A special thank you is for Dr. McIntosh, whose inspiring lectures on protein NMR motivated me to learn more and centre my research around NMR. I also would like to thank Dr. Julie Forman-Kay and Dr. Rhea Hudson for sharing several of the plasmids expressing the different WW domains used in this thesis work. xxviii I am very grateful for the help and companionship from past and current Straus group members: Dr. Andrew R. Tait, Htet Bo, Joseph Lee, Ruqaiba Desmond, Prashant Kumar, Jin Zhang, Angela Dodd, Chantal Mustoe, Nigare Raheem and especially Rui Zhang for her collaboration on U24-6A related projects, help on the pull-down experiments (mentioned in the preface), and encouragement to continue and explore more avenues in my thesis work. In particular, thank you ’s are given to Htet Bo and Dr. Andrew R. Tait, who also worked on the U24 project, for passing on lab skills. I would like to thank the staff of the Biological Services Laboratory, Dr. Elena Polishchuk and Jessie Chen, for the lab skill training they provided, as well as their technical advice. Their efforts in maintaining and acquiring equipment for the facility enabled my productive work during the last few years. Colleagues that worked in Biological Services Laboratory are acknowledged for their friendship and help on my thesis work: Dr. Jie Fang and Dr. Hai-Yan He for advice and help on cloning; Wenbo Liu for suggestions on synthesis and separation of U24-7 peptide; Dr. Niusha Mahmoodi and Dr. Qi Qian for training me on how to use the French press; Dr. Feng Liu for helpful experimental suggestions; and too many others to mention for coordinating equipment sharing and autoclave usage. I would also like to express my sincere gratitude towards the staff of Mass Spectrometry Centre, including Marshall Lapawa, Derek Smith, Marco Yeung and Dr. Yun Ling. Thanks for all of them for running all my samples, over one kilogram of hard copy results, and being patient with my constant bugging, scheduled sample running and special requests. A special thank you to Marshall Lapawa, who figured out the metastable decomposition of U24-7 peptide when analyzing it using MALDI-TOF MS. This enabled my thesis work to proceed as planned. I am grateful for the help from other shops and services in the Department of Chemistry and at UBC, who helped with such things as fixing glassware or instruments, maintaining instruments, making custom parts and so on. I want to thank Dr. Emily Seo and Felix Shuen from the Shared Instrument Facility and Dr. Fred Rosell from Shared Spectroscopy and Kinetics Hub for their help xxix on using CD spectrometers. I would like to thank Dr. Martin E. Tanner for the use of the French press and Dr. John Sherman for the use of the HPLC. My long-time friend, Dr. Xintong Dong, offered a great deal of encouragements during my graduate studies and she also commented on my ideas for future works in Chapter 6, which is much appreciated. I also would like to thank Wanchen Lu for teaching me how to program using command lines, in order to writing scripts for NMRPipe. Last but not least, I would like to thank my entire family for their unconditional love and support, especially my father and deceased mother. They encouraged me to carry on studying and to pursue a Ph.D. degree. I am also very grateful for both material and spiritual support from my father, my cousin-sister, as well as my aunts. Two special thank you ’s are for my uncle Zhi Li, who sent me his painting and calligraphy work to encourage me to carry on writing (one of them is on the next page), and for my godmother Ling Gu who insisted on “watching” me write the first chapter of this thesis. xxx Dedication To my father, & my family At Heron Lodge, by Wang, Zhihuan Mountains cover the white sun, and oceans drain the golden river; But you widen your view three hundred miles by going up one flight of stairs. (Translated by Bynner) 1 Chapter 1 Chapter 1 Introduction The protein interactions between the tail-anchored membrane protein U24 from Human herpesvirus type 6A (HHV-6A) and type 7 (HHV-7), and its potential binding partners are presented in this thesis. U24 from HHV-6A (U24-6A) shares a seven amino acid stretch (PRTPPPS) with myelin basic protein (MBP), a protein which plays an important role in myelination and has therefore been associated with multiple sclerosis (MS). It has been suggested that U24-6A might trigger autoimmunity against MBP in MS through this segment. (1) The proline-rich segment in U24-6A contains both a PxxP and a PPxY (PY) motif. The equivalent region in U24-7 only contains the PY motif. The interactions between these proline-rich motifs and their binding domains are investigated in this thesis and the implications of these interactions are discussed. Chapter 1 reviews four topics that form the basis of the work described in later chapters: U24 and Roseoloviruses, multiple sclerosis as the main associated disease, the structure and function of potential binding partners of U24, and the techniques that were used in this study. Finally, the aims of the dissertation are presented at the very end of this chapter. 1.1 Roseoloviruses and U24 Roseoloviruses belong to the subfamily of Betaherpesvirinae. Roseoloviruses include two main species, Human herpesvirus 6 (HHV-6, the collective name for HHV-6A and HHV-6B) and Human herpesvirus 7 (HHV-7). These viruses are known to cause roseola infantum, or exanthema subitum, in young children. A lifelong latency can be established after the initial infection and severe reactivation of the virus in transplant and immunocompromised patients has been reported, leading to complications, such as encephalitis and multiple sclerosis (MS). (2–4) Both viruses (Figure 1.1A) are ubiquitous and prevalent among all of the population. The specific antibodies against HHV-6 2 and HHV-7 have been detected in more than 90% of adults worldwide. (5, 6) Recently, it has been demonstrated that HHV-6 could be reactivated in vitro. (7) Figure 1.1 Virus and T-cell receptor: A) Illustration of the envelope of Roseoloviruses. B) Illustration of the T-cell receptor complex (TCR) found on the surface of T-cells. Image in A) is adapted from http://hhv-6foundation.org/what-is-hhv-6 (accessed on Aug 12th, 2015) and reproduced with the permission of the HHV-6A foundation. The image in B) is adapted from https://en.wikipedia.org/wiki/CD3_(immunology) (accessed September 10th, 2015). This image, uploaded by Anriar~commonswiki, is licensed under CC BY-SA 3.0. The two chains in red belong to the TCR receptor and the other chains are part of the CD3 co-receptor. An accessory molecule is shown in green. Compared to other herpesviruses, Roseoloviruses have small genomes, with several unique proteins, including putative membrane proteins U20, U21, U23, U24 and U26. (5) Many of these proteins (U21-U24 in particular) have been found not to be required for HHV-6 viral growth in vitro. Indeed, no significant difference in the growth curves of wild type, recombinant (generated using recombinant DNA technology), and revertant (genes of U21-U24 are suppressed) viruses was found. (8) Roseoloviruses have been found to mainly affect T-cells, which are also known as T-lymphocytes. A lymphocyte is a type of white blood cell and plays an important role in immunity. T-cells are different from other types of lymphocytes (e.g. B-cells, natural killer or NK cells) because of the presence of T-cell receptor (TCR) on the cell surface. The TCR is a complex, formed by two TCR chains and a CD3 co-receptor (Figure 1.1B), comprised of γ, δ, and two ε subunits. The CD3 surface expression level is decreased in HHV-6 infected T-cells, suggesting that the virus might encode a protein that down-regulates CD3. (9) A B 3 1.1.1 HHV-6 HHV-6 was firstly isolated from acquired immune deficiency syndrome (AIDS) patients (10) and the isolated strains were later divided into two groups, HHV-6A and HHV-6B. Recently, these two groups have been classified as separate viruses (11). Although the genomes of these two viruses share an overall identity of 90%, they have been distinguished based on restriction endonuclease profiles and cell tropisms. (12, 13) They both utilize CD46 for cell entry, but do so using different mechanisms. (14) Recently, olfactory-ensheathing glial cells were found to support the replication of HHV-6, suggesting that the olfactory pathway is a route of HHV-6 entry to the central nervous system (CNS). (15) HHV-6B is the common cause for roseola infantum in children, whereas there are few symptoms are thought to be associated HHV-6A primary infections. (16) Both HHV-6A and 6B can be integrated into the human chromosome telomeric region. (17) The chromosomally integrated HHV-6, known as ciHHV-6, found in roughly 1% of the population, could be a mechanism to ensure the survival of the viral genome. (18, 19) Such a mechanism could be important to mobilize the viral genome during reactivation. (19, 20) HHV-6 has been implicated in several neurological diseases, such as encephalitis, mesial temporal lobe epilepsy (MTLE) and MS. (21–25) The mRNA of HHV-6A is more frequently found in neuroinflammatory diseases than HHV-6B (e.g. MS and rhombencephalitis). (22, 23, 26) HHV-6A virus was also detected in the cerebrospinal fluid (CSF) of MS patients. (27) 1.1.2 HHV-7 HHV-7 was firstly isolated from the CD4+ T lymphocytes of a healthy individual. (28) It is genetically closely related to HHV-6, but the genome is 10% smaller in length. (29) CD4, a glycoprotein located on immune cells, was found to be crucial for HHV-7 cell entry. CD4 was also found to be down-regulated during the course of infection. (30) Therefore, HHV-7 could potentially interfere with human immunodeficiency virus (HIV) infections, since this latter virus also requires CD4 for cell entry. (30, 31) 4 HHV-7 mainly causes roseola infantum. The primary infection causes febrile illness, with or without rash symptoms. (32) Similar to HHV-6, it is often reactivated in immunocompromised hosts and sometimes accompanied by other viruses, such as cytomegalovirus (CMV). (33, 34) Recently, it has been found that serious neurological diseases could result from a delayed primary HHV-7 infection in adolescents. (35) Although harbouring telomeric repeats, no chromosomally integrated HHV-7 has been reported to date. (19) Moreover, there is little direct evidence that a link between HHV-7 and neurological diseases exists, with the exception of a few studies showing that it could cause encephalitis and be related to MS. (4, 35) 1.1.3 Membrane protein U24 As mentioned above, U24 is a putative membrane protein that is unique to Roseoloviruses. (5) The amino acid sequences of U24 are slightly different in HHV-6A, HHV-6B and HHV-7, as seen in the alignment shown in Figure 1.2. U24 from HHV-6A (U24-6A) and U24 from HHV-6B (U24-6B) are highly conserved, while the sequence remains ca. 28% identical to that from HHV-7 (U24-7). All three U24 proteins contain a transmembrane region at the C-terminus of the protein. This was predicted based on the high abundance of hydrophobic amino acids and subsequently confirmed by studies in which the transmembrane domain recognition complex 40 (Trc40) was deleted. (36) At the N-terminus, all U24s have a proline-rich region. It has been confirmed that all U24s are localized in several endosomal compartments in T-cells. (36) The exact biological role of U24 is currently unknown. U24-6A is one of the HHV-6A early genes, i.e. it is expressed in the infected cells at the early stage of viral infection. (37) In CD4+ T lymphocytes, the mRNA of U24 could be detected immediately after infection by HHV-6A. (37) U24 was also found to be up-regulated in the brain of a bone marrow transplant patient. (38) In addition to its involvement in infection, U24 may be involved in immune evasion. (39) Other proteins which, like U24, are also unique to Roseoloviruses were proven to be crucial for latency maintenance: U20 from HHV-6B and U21 from HHV-7 may down-regulate the cell-surface presenting ligands in order to inhibit the apoptosis activated by tumour necrosis factor alpha 5 (TNF-α), or avoid the recognition by NK cells; (40–42) U83 from HHV-6A is a chemokine homolog and has been found to block endogenous ligand-receptor interactions. (43, 44) Finally, because U24 was found to be up-regulated in the brain tissue of transplant patients, (38) it is highly possible that the function of U24 is also closely related to neurological diseases. conserved polyproline region putative transmembrane ******:* +++ ====================== ++ +++ U24-6A MDPPRTPPPSYSEVLMMDVMCGQVSPHVINDTSFVECIPPPQS-RPAWNLWNNRRKTFSFLVLTGLAIAMILFIVFVLYVFHVNRQRR-- U24-6B MDRPRTPPPSYSEVLMMDVMYGQVSPHASNDTSFVECLPPPQSSRSAWNLWNKRRKTFAFLVLTGLRIAMILFIAFVIYVFNVNRRKK-- U24-7 M-THETPPPSYNDVMLQMFHDHSVFLHQENL-S-----PRTINSTSSSEIKNVRRR-GTFIILACLIISVILCLILILHIFNVRYGGTKP Figure 1.2 Sequence alignment of U24-6A, U24-6B and U24-7 The hydrophobic or putative transmembrane region is found near the C-terminal of the U24s (indicated by double lines), and flanked by positively charged residues (pluses). The conserved polyproline region at the N-terminus is labelled and shows which residues are the same (stars) or similar (colons). The reference number of U24-6A (from HHV-6A, strain U1102) in NCBI is NP_042917.1. The access code of U24-6B (HHV-6B, HST) and U24-7 (HHV-7, RK) in the gene bank are: BAA78245.1 and AAC40737.1. As a unique gene product of Roseolovirus, U24 was found to be able to down-regulate the TCR/CD3 complex (Figure 1.1B), when expressed in T-cells. (36, 45) Similar results were also observed using HHV-6A infection experiments. (9, 46) The TCR/CD3 complex is a membrane bound receptor complex and is essential for T-cells to recognize antigens and receive extracellular signals. A decreased amount of this complex on a cell surface allows the cell to evade detection from the immune system. A low concentration of the TCR/CD3 complex on a cell surface could be due the internalization of this complex. (45) Using an alanine scan of the residues common to all three versions of U24s (Figure 1.2), the proline-rich region at the N-terminus was found to be indispensable for TCR/CD3 down-regulation. Specifically, mutations at Pro8, Pro9 and Tyr11 had a detrimental effect on activity. (36) These residues taken together form PY motif, which will be discussed in detail in sections 18.104.22.168 and 22.214.171.124 below. 6 In order to try to pinpoint how U24 down-regulates the TCR/CD3 complex, a number of experiments were conducted and reported in the literature. One such study showed that U24 and TCR/CD3 showed little to no colocalization in cell membrane compartments. (36) It was also found that U24 causes the exclusion of the receptor from recycling endosomes, a membrane-bounded compartment inside the cell used to transport material in and out of cells (see following section for a description of endosomes). (45) Indeed, instead of being recycled back to the cell surface, TCR/CD3 complexes were found in late endosomes. (9, 45) In order to determine whether U24 also plays a role in early endosomal recycling, its effect on transferrin receptor (TfR) was also investigated. TfR has been used extensively to study early endosomal recycling. (47, 48) It was found that TfR could also be down-modulated by U24 in a CD3 independent manner. (36) This suggests that U24 is not receptor specific and it may affect a third protein. Similar to U24, Nef from HIV-1 was found to reduce the rate of TfR and CD4 recycling back to membrane surface. The TfR was found to accumulate in early endosomes, but how this is exactly achieved is unknown. (49) One possibility is that CD4 down-regulation is a consequence of Nef hijacking an associated protein from membrane protein sorting complex, apoptosis-linked gene 2 (ALG-2)-interacting protein (Alix). (50) Overall, the data have demonstrated that U24 mediates a block of endosomal recycling specific to early endosome, and that receptors recycled from other pathways were not affected. (36, 51) The function of U24 might be similar to that of Nef, but no host protein recruitment was identified for U24 to date. 126.96.36.199 Endosomes Before continuing to discuss U24 in more detail, a brief description of endosomes will be given. In simple terms, an endosome is a membrane-bounded compartment, used for shuttling material from a cell surface into a cell, and in some cases back again, as in the case of endosomal recycling. Endosomes are key transport vehicles and are important in the endocytic pathway, shown in Figure 1.3. There are three basic types: early endosomes, which mature to form late endosomes, and finally recycling endosomes (Figure 1.3). Endosomes are initially formed when part of the cell membrane is pinched to form a vesicle. They carry cargo and fuse with either early or recycling 7 endosomes. The fate of the cargo, which is either to be recycled back to the cell surface or to be degraded in the lysosome, is dependent on the ubiquitin that modifies the cargo. Cargo sorting is done by the endosomal sorting complex required for transport (ESCRT) machinery. Please see further details on the endocytic pathway and all the markers involved (e.g. Rab proteins, Figure 1.3) in these excellent reviews. (52, 53) Figure 1.3 Adapted with permission from Pathways and mechanisms of endocytic recycling, Barth D. Grant & Julie G. Donaldson, Nature Reviews Molecular Cell Biology 10, 597-608 188.8.131.52 Phosphorylation of U24 Like most viral proteins, U24 is extensively posttranslationally modified when expressed in eukaryotes. The size of U24 expressed in T-cells and observed on SDS-PAGE gels (36) was found to be two times larger than the expected molecular weight of 10 kDa (based on amino acid sequence). The exact modifications were, however, not identified. It was previously demonstrated by our group that U24-6A can be phosphorylated in vitro by 8 mitogen-activated protein (MAP) kinases. (54) The phosphorylation site was identified to be Thr6, which is within the optimal recognition motif of MAP kinase, namely Px(S/T)P, with the minimum requirement for MAP kinase recognition being (T/S)P. (55) More importantly, the phosphorylation site identified for U24-6A corresponds to the analogous threonine position which is phosphorylated in MBP. The implications of MBP phosphorylation in terms of MS are described in more detail in section 1.2.2. 184.108.40.206 U24 contains a proline-rich region As mentioned in section 1.1.3, the proline-rich region at the N-terminus appears to be conserved among all three types of U24s. Generally speaking, proline-rich regions are good ligands for protein-protein interactions due to their unique secondary structure and ability to form intermolecular hydrogen bonds (vide infra). In the proline-rich region of U24-6A, PPRTPPPSY, there are two motifs, PxxP and PY motif. The binding domains which can interact with these two motifs will be briefly introduced in section 220.127.116.11. A B trans cis PPI helix PPII helix Figure 1.4 X-Proline bond isomers and polyproline type I and type II helices. A) The trans and cis isomers of X-Pro peptide bonds, where X is any residue preceding the proline. The bonds defining the trans and cis geometry are shown in blue. B) Side views of PPI and PPII helices, comprised of nine proline residues, represented by spheres. The helices were generated using the Build Structure, and Adjust Torsions tools in UCSF Chimera. (56) The helices are oriented such that the top represents the N-terminus. Carbon atoms are shown in grey, nitrogen in blue and oxygen in red. Hydrogen atoms are omitted for clarity. 9 Proline is a unique amino acid. The alkyl side chain is covalently connected to the backbone nitrogen in the peptide bond (Figure 1.4), leaving no amide proton for hydrogen bonding. Because of its particular features, a single proline in the sequence can disrupt secondary structure formation. (57) For example, proline substitutions in amyloid forming peptide resulted in an increase in its solubility and a reduction in fibril formation. (58, 59) In a protein or peptide, the peptide bond between proline and its previous residue has two possible conformations: the cis and trans isomers (Figure 1.4). Conversion between the cis and trans isomers is facilitated by prolyl isomerases (Pi). Proline-rich regions, or polyproline regions, are often found in proteins, in some cases as multiple repeats. (60) Proline-rich regions can adopt two types of helical structure, depending on the conformation of the x-Pro peptide bond. When the x-Pro bond is trans, a left-handed helix, also called type II helix (PPII helix), is formed. It is an elongated helix. As mentioned above, there is no amide proton in proline, thus this helical structure is maintained without hydrogen bonds. Similarly, a cis x-Pro bond results in a right-handed stacked helix or type I helix (PPI helix), which is less common. Proline-rich regions are typically involved in protein-protein interactions because of the distinct structure of the polyproline helix. In a PPII helix, the carbonyl moieties point away from the helix long axis, instead of along the axis as in an α helix. Consequently, there are no intramolecular hydrogen bonds, but rather, the carbonyls easily participate in forming intermolecular hydrogen bonds. (61) In addition, the size and shape of the cyclic proline side chain is often an ideal fit for a hydrophobic pocket situated on a protein surface. (62) Increased potential for intermolecular hydrogen bonding and for specific hydrophobic interactions make proline-rich regions a good ligand for protein-protein interactions. The binding domains will be briefly introduced in the next section and their structure and function will be introduced in section 1.3. 18.104.22.168 Binding domains for the proline-rich region in U24 As described in the section above, two proline-rich motifs are present at the N-terminus of U24-6A, namely a PxxP motif and PY motif. In U24-7, only the PY motif is present. Given the 10 high propensity for such motifs to bind other proteins, it is very likely that the proline-rich motifs found in U24-6A and U24-7 are important for function. Both the PxxP and PY motifs are known to form PPII helices, (61) and have known binding partners: the PxxP motif binds to Src Homology 3 (SH3) domains and the PY motif is the ligand of WW domains. (63) SH3 domain is a non-catalytic domain in kinases and phospholipases, (64) and it is responsible for recognizing substrate. These domains are ubiquitous and small, containing about 50-70 amino acids. Their structure consists mainly of β sheets and some long loops. No distinct binding pocket is located on the domain surface. (65, 66) A PxxP motif containing ligand can bind to SH3 domain in two possible orientations. (67) The residues flanking the core PxxP motif, e.g. an additional arginine up or down-stream, could enhance the affinity with SH3 domain. (67, 68) Details pertaining to the structure and types of proteins which contain a SH3 domain will be introduced in 22.214.171.124. WW domains are named after the two highly conserved tryptophan residues found in the sequence, usually ca. 20-22 residues apart. WW domains are one of the smallest β-stranded folds. They are typically monomers in solution, without any associated cofactors or disulphide bonds. (69) WW domains are responsible for ligand binding and the resulting signalling complexes are implicated in different diseases and cellular activities, such as Liddle’s syndrome, Huntington’s diseases, gene transcription, and ubiquitination. (70–75) There are four groups of WW domains, each with specific preferred proline-rich ligands, i.e. group 1 WW domains prefer the PY motif, while group 4 WW domains prefer the p(S/T)P motif. pS or pT indicates phospho-serine or phospho-threonine, respectively. Typical ligands are listed in Table 1.1. (76) Further details about WW domains will be given in 126.96.36.199. The structural basis of ligand recognition by WW and SH3 domains is the x-Pro (xP) dipeptide, where x is usually a hydrophobic residue. The unique configuration of the xP dipeptide allows these domains to recognize using their xP groove. (61, 77) The small size of the xP groove allows SH3 domains to bind in one of two orientations. In addition to the xP groove, WW domains contain a conserved histidine, which may interact with the tyrosine in PY motif. (78) There are other proline- 11 rich region binding domains, e.g. enabled/VASP Homology 1 (EVH1) domain, Gly-Tyr-Phe (GYF) domains, and ubiquitin E2 variant (UEV) domain, and their preferred binding ligands are summarized in the table below. (79, 80) Table 1.1 Proline containing motifs and their binding domains Φ indicates a hydrophobic residue, and x is any residue. The residues in bold represent the common ligand. Protein domain Proline-rich ligand SH3 domain RxxPxxP, PxxPxR (67, 68) WW domain PPxY, PPLP, PPR, p(S/T)P(76, 81, 82) EVH1 domain FPxΦP, (83, 84) xPPxF (85) GYF domain PPGx(R/K), (61) PPGΦ (80) UEV domain P(S/T)AP (79, 86) 1.2 Possible disease associations of U24 As mentioned in 1.1, Roseoloviruses are very common and believed to be implicated in diseases, such as bone marrow transplantation (BMT), AIDS, malignancies, MS and hypersensitivity syndrome. (27, 87, 88) In the case of MS, there is no conclusive evidence to show that HHV-6 or 7 cause the disease, but there are clues to show that there is a link between them. Specifically, higher level of HHV-6 viral replication has been detected in the lesion areas of MS patients. (89) Being one of the unique proteins coded by Roseoloviruses, U24 might contribute to the diseases associated with HHV-6 and -7. As MS is a focus of this thesis, it will be discussed in more detail in the following sections. For details on how HHV-6 and -7 are linked to other diseases, please consult other references. (87, 88) MS is a neurological inflammatory disease in the central nervous system (CNS). (90) The disease typically arises during young adulthood and is accompanied with several symptoms, such as fatigue, poor balance, speech and cognitive impairment. Most MS patients start with a course of relapsing-remitting MS (RRMS), during which inflammation in the CNS is temporary. With time, 12 however, a chronic neurodegenerative disease develops. There is currently no known cure for MS, only therapeutics which help in the early phases. (90, 91) To date, the cause of MS has not been identified. It is considered by some to be due to environmental and geographic factors, combined with specific high-risk genetic profiles. (90, 92) Some theories suggest that one or more viruses may be possible triggers of MS. (4, 22, 27, 35, 93, 94) Figure 1.5 Healthy and diseased myelin A) Myelin in healthy CNS. Most of the axons are protected by myelin sheaths, which are generated by the oligodendrocytes. B) Diseased myelin from MS patients. One possible cause of the inflammation associated with MS is an autoimmunity attack by a patient’s T-cells. C) Emerging evidences suggest alternative causes for MS, such as the axono- or oligodendropahty. The source of the damage could be either the neuron or oligodendrocytes themselves. In MS patients, lesions are found in the brain tissue as a result of the degradation of the myelin sheath, a scroll-like layer of lipids that wraps around neuronal axons, created by oligodendrocytes. Figure 1.5 is a depiction of healthy (A) and diseased (B, C) myelin. The damaged myelin can be regenerated or remyelinated during the early stage of the disease, but the repair machinery becomes increasingly inefficient and eventually stops, resulting in irreparable lesions in the white matter tissue of the brain. (95, 96) One potential cause of demyelination is linked to an autoimmune attack 13 (Figure 1.5B) by autoreactive T-cells. (22, 90, 94, 97) These T-cells treat the host tissues as alien objects and try to eliminate them by attacking them. The generation of autoreactive T-cells could be triggered by a viral protein. (98) The main autoantigens, i.e. normal constituents of neuronal cells which become the target of an immune responses, are myelin basic protein (MBP), proteolipid protein (PLP), myelin oligodendrocyte glycoprotein (MOG) and myelin-associated glycoprotein (MAG). (99) Another potential cause of MS is dystrophy of the oligodendrocytes and neuron (Figure 1.5C). (91, 100, 101) In this case, the degeneration of these cells would slow down attempts at remyelination, resulting in a net loss of myelin. Specifically, neuron damage may directly contribute to the disability observed in MS patients. (102) Indeed, cortical grey matter lesions can be found even in the early stages of the disease. This neuronal damage could result from both inflammatory and degeneration mechanisms. (102–104) There is also evidence to show that the microglia, the resident macrophages in CNS, could contribute to the inflammatory symptoms found in MS. (105) 1.2.1 U24 and multiple sclerosis As mentioned above, there is some evidence that HHV-6 and -7 may be linked to MS. In terms of U24 more specifically, recent evidence has shown that there might even be a more direct link: a synthetic peptide corresponding to the first 13 amino acids of U24-6A was found to activate more than half of the T-cells in MS patients, T-lymphocytes which recognized and cross-reacted with MBP. (1) Given that this amino acid segment in U24 is located in the proline-rich region (sections 188.8.131.52 and 184.108.40.206) of the N-terminus, a region found to be important for U24 function (section 1.1.3), we hypothesize that the N-terminus of U24 is the “business end” of the protein. A comparison of the sequences of U24-6A or U24-6B and residues 93-107 of MBP (Figure 1.6) shows that the sequences have common features, namely the seven amino acid residue stretch PRTPPPS. (1) U24-7 is also shown in the figure. Besides triggering the initial T-cell response, U24 could play a role in MS progression, in which myelin basic protein has been found to play an important role. Given the sequence similarity, it may 14 be possible that U24 mimics MBP and competes for resources required for important posttranslational modifications, important signalling interactions, etc. This mimicry hypothesis will be described in section 1.2.2. PxxP motif MBP93-107 IVTPRTPPPSQGKGR U24-6A1-15 MDPPRTPPPSYSEVL U24-6B1-15 MDRPRTPPPSYSEVL U24 -71-14 M-THETPPPSYNDVM PY motif Figure 1.6 Alignment of MBP and U24s The numbering of MBP is based on the human 18.5 kDa classic MBP. (106) The identical segment between U24-6A and MBP, PRTPPPS, includes the PxxP motif (marked with the solid line box). In U24-6B, the third residue from the N-terminus is Arg, while in U24-6A it is Pro. U24-7 is aligned with respect to U24-6A and their common PY motifs are marked with a dashed box. Serines or threonines, which were reported to be phosphorylated (54, 106), are shown in bold and red. Thr98 is the most common MAP kinase phosphorylation site in the PxxP motif of MBP. On the other hand, it is possible that U24 plays a different (less direct) role in MS, through its ability to block endosomal recycling (as described in 1.1.3). (36) Recent work has shown that some important myelin sheath proteins, such as MAG and MOG, may traffic in a similar manner to TfR (section 1.1.3). (107) This alternative hypothesis on the role of U24, i.e. that a block of the endosomal recycling by U24 may cause a disruption in remyelination, will be described in section 1.2.3. Alternatively, U24 might affect cell development of neurons and oligodendrocytes via the same mechanism. A summary of the crucial factors which may help to show how U24 may play a part in MS is shown in pictoral form in Figure 1.7 below. 15 Figure 1.7 A summary of how U24 may play a part in MS. Figure 1.1A is shown here to represent the Roseoloviruses. 1.2.2 Mimicking MBP In order to understand how U24 may mimic myelin basic protein, a deeper understanding of MBP is required. MBP is the second most abundant protein in myelin, after PLP, and it is also an initial antigen that triggers experimental autoimmune encephalomyelitis (EAE), a experimental model of MS. (108, 109) In this section, the function, posttranslational modifications and important interaction partners of MBP will be described. 220.127.116.11 Myelin basic protein The gene encoding MBP is part of a larger gene called gene of the oligodendrocyte lineage 16 (golli). The golli consists of ten or eleven exons, seven of which belong to the MBP gene. (110, 111) The size of MBP ranges from 14 kDa to 21.5 kDa due to alternative splicing. The classic isoform is the 18.5 kDa one, which is the most abundant species in myelin. (106, 108) The gene product of golli, called Golli-MBP, can be expressed by oligodendroglial progenitor cells (OPCs) and T-cells. In the OPCs of CNS, Golli-MBP plays a role in myelination and remyelination by enhancing the migration speed of OPCs and stimulating their proliferation and differentiation, steps which are important for CNS development. (112, 113) The Golli-MBP isoform that expresses in T-cells plays a regulation role in the immune system. (114–116) In a compact myelin, MBP is responsible for the adhesion of the cytosolic side of lipid layers, while PLPs may exhibit homophilic interactions in the extracellular space. (108) Together, they help form healthy myelin. When, however an autoimmune response is triggered, MBP acts as an autoantigen, with anti-MBP autoantibodies readily detected in the cerebrospinal fluid (CSF) of MS patients. (117) The number of these antibodies produced was found to increase steadily through the progressive stages of MS. (118) 18.104.22.168 Posttranslational modification of MBP Once expressed in cells, MBP is extensively post-translationally modified. Indeed, there are at least eight different variants for the classic MBP which have been reported. (119) The posttranslational modifications include deamidation, phosphorylation and citrullination, and they generally result in a reduction of the overall charge of MBP. The most abundant and least modified form of MBP in healthy adult myelin contains +19 charges at neutral pH. (120, 121) The positive charges allow it to regulate lipids at the inner leaflet of the cytosolic membrane. (122) Phosphorylation is important for MBP function and is of particular relevance for the work presented in this thesis. In mature myelin, the turnover rate of phosphoryl groups in MBP is rapid, indicating that phosphorylation is occurring recurrently in the cell. (123) Less phosphorylated MBP (at Thr98 (human numbering, underlined) in the PPRTPPPS segment) was found in MS brain than normal tissue. (124) 17 22.214.171.124 Mimicking the interaction of MBP To acquire sufficient modifications, MBP needs to interact with other proteins. One of the important binding partners is Fyn tyrosine kinase, which has a unique role in myelination. The expression level of Fyn was found to be up-regulated during the initial stages of myelination. (125) Fyn mainly promotes gene transcription of MBP through other kinases and transcription factors. This stimulation of Fyn is stage and isoform specific. (125, 126) In in vitro cellular experiments, it is clear that Fyn and MBP are colocalized in oligodendrocytic cell lines. The interaction between Fyn and MBP is mediated by the SH3 domain of Fyn tyrosine kinase (Fyn-SH3) and the PxxP motif in MBP. (127–129) It has been demonstrated that this interaction allows MBP to tether Fyn-SH3 domain to a membrane in vitro. (130) Detailed investigations using NMR revealed that a longer peptide than the core polyproline region is required to form the complex with Fyn-SH3 domain. (131) The detailed architecture and functions of Fyn will be introduced in section 1.3.1. When MBP is phosphorylated at Thr98, the affinity between MBP and Fyn-SH3 is lowered. This is due to the electrostatic repulsion between the negative surface charge on Fyn-SH3 domain, and the negatively charged ligand segment of PxxP in MBP. (129, 130) If U24-6A could mimic MBP by interacting with Fyn-SH3 and competing kinases, U24-6A could potentially be involved in a more complicated regulation system. 1.2.3 U24 may be implicated in MS by blocking endosomal recycling As mentioned in section 1.2, some studies suggest that the initial stage of MS may not be caused by an autoimmune response. (91, 100, 132) Extensive zones of apoptotic oligodendrocytes were found in newly formed lesions with few to no lymphocytes, (133) while lesions in which complete demyelination occurred were packed with T-cells, B-cells and plasma cells. (134) An “inside-out” mechanism was recently proposed for the long-term damage associated with MS, i.e. not the initial response or RRMS. Nikić et al identified a new variant of tissue damage in which axons swell at first, followed by fragmentation and degeneration, but no demyelination. (101) Neurite outgrowth inhibitor could be involved in this tissue damage. What is more, persistent 18 sodium influx through the sodium channels on the neuron could activate the sodium/calcium exchanger, thereby triggering calcium influx, which in turns leads to axonal injury and neuroinflammation. (135) Taken together, this data suggests that other players, such as minor myelin proteins, neurite outgrowth inhibitor or sodium channels, may be important in MS. Since the exact nature of the role of U24 in MS is unknown, it is therefore important to explore whether U24 may function by disrupting the functions of one or more of these proteins (instead of or in addition to competing with MBP). In particular, the hypothesis that U24’s ability to block early endosomal recycling may have an impact on the functions of these proteins will be presented. 126.96.36.199 Minor myelin proteins There are two minor membrane glycoproteins in myelin, MAG and MOG (section 1.2). They are identified as autoantigens in MS and localized at periaxonal loops (the layer next to axon) and abaxonal loops (the layer on the surface of myelin), respectively. (136) Both proteins are endocytosed via a clathrin-dependent pathway, but their fates are different. MAG is targeted to the late endosome and lysosome directly, while MOG is trafficked to the recycling endosome. (107) Located between the neuron and myelin sheath, MAG appears to have two functions, namely to either promote or inhibit neurite outgrowth, depending on the developmental status of the neuron. (137, 138) MOG has been identified as the main autoantigen for MS and EAE, due to its position on the outermost surface of myelin sheaths. It may be involved in the completion and maintenance of the myelin sheath and in cell-cell communication, but exact mechanisms on how MOG accomplishes these functions are unknown. It is speculated to have supporting, regulatory or immune functions, but details are lacking. (139) 188.8.131.52 Nogo-A, a neurite outgrowth inhibitor Neurite outgrowth inhibitor (Nogo-A), or Reticulon 4-A (Rtn4-A), is a member of the Reticulon family and is one of three Nogo isoforms (Nogo-A, Nogo-B and Nogo-C) encoded by the nogo gene. Its role is to maintain the shape of tubular endoplasmic reticulum (ER), and it does so by mainly associating with the ER and plasma membranes. (140, 141) Nogo-A is ubiquitously 19 expressed but it mainly act as inhibitor when it expressed in oligodendrocytes. It functions as a negative regulator of synaptic plasticity when expressed in neurons. (142, 143) Nogo-A has mainly two sections that have inhibitory functions, Nogo-66 and Nogo-A Δ20. (140, 144) Figure 1.8 above shows the details of the topology and localization of these two segments within the protein. Nogo-66 binds to the Nogo-66 receptor (NgR), while Nogo-A Δ20 functions through sphingosine 1-phosphate receptor 2 (S1PR2) and co-receptor tetraspanin-3 (TSPAN3). (145–147) Nogo-A Δ20 is of particular interest due to it is uniqueness among all Nogo homologues. Nogo-A is regulated by Praja2, a member of RING family of E3 ubiquitin ligases. An important feature of Nogo-A is that it contains a PY motif, located within the Nogo-A Δ20 segment, which binds to WWP1 E3 ubiquitin ligase. (148, 149) The difference ubiquitin ligases associated with Nogo-A could be due to the different localizations of this molecule (Figure 1.8). Figure 1.8 Topologies of Nogo-A in the plasma and tubular ER membrane (150) Both Nogo-A Δ20 and Nogo-66 are in the extracellular space when the protein localizes to the plasma membrane. (144, 150) It is proposed that the protein is anchored in the membrane by two transmembrane domains, which could either span the lipid bilayer or be folded into two sections. In the ER membrane, both N- and C- termini are found in the cytoplasm, therefore the Nogo-66 could be in the ER lumen or cytoplasm. The legends are marked with the amino acid sequence numbering in the brackets below the names. 20 In MS brain tissue, the expression of Nogo-A was found to be up-regulated at the edge of lesions in oligodendrocytes, indicating that it may be related to demyelination. (151) The clinical signs of EAE progression are reduced in Nogo-A suppressed mice, and exuberant and expansive myelination is found in mice knockouts after injury. (152–154) On the other hand, Nogo-A can have a positive impact on myelination. The expression level of Nogo-A increased faster than that of MBP and PLP in differentiating oligodendrocytes. Moreover, Nogo-A null mice showed delays in oligodendrocyte differentiation and myelination. (155) Nogo-A Δ20 inhibits the formation of myelin internodes and is necessary for spatial segregation and refinement of myelination in vitro and in vivo. (154) Taken together, the evidence suggests that Nogo-A plays a bifunctional role towards myelination, which may be dependent on the developmental stage of the patient. 184.108.40.206 Sodium channels in neurons Besides the proteins mentioned above, sodium channels could also be implicated in MS. (156) They are essential for generating an action potential on the axon, i.e. a momentary change in electrical potential on the surface of a cell, and might contribute to axonal degeneration. In MS patients, a higher density of sodium channels was found in brain lesions. (157) More specifically, NaV1.6 was more frequently found associated with lesions. (158, 159) NaV1.2 was found in unmyelinated regions and developing CNS nodes, while NaV1.6 replaced NaV1.2 at the nodes of Ranvier as myelination proceeded. (160, 161) This increased density in sodium channels may be an attempt to recover excitability on denuded axons. (162, 163) Besides the axons, microglia, the resident immune cells in CNS, also expresses sodium channels that might contribute to MS and the blockage of these sodium channels could limit the immune responses in EAE. (164–167) It is important to note that many sodium channels, such as epithelial sodium channel (ENaC), undergo endocytosis through a clathrin-dependent pathway, which is regulated by Nedd4 family E3 ubiquitin ligase (to be introduced in section 220.127.116.11). (168) In addition, evidence on the involvement of NaV1.x in endocytosis is also emerging. For example, NaV1.5 has been found to be expressed in the late endosome and to play a role in macrophage endosomal acidification. (169) 21 18.104.22.168 Evidence of a connection? It is clear from sections 22.214.171.124-126.96.36.199 that many molecules have been implicated in MS and that there are many complex pathways that have been proposed to account for the associated degeneration. Although these molecules were presented separately, it is important to note that in biology, they are not necessarily independent. For example, MAG has been found to be regulated by MBP. (170) And MAG and Nogo-66 are two of the three major myelin-associated inhibitors in the adult nervous system. (171) So what is the motivation behind proposing that U24 may play a role in MS via its ability to block early endosomal recycling? The aim of U24 expression in Roseoloviruses may be to down-regulate TCR/CD3 in order to avoid immune detection. This can be achieved by blocking endocytic recycling but may also have additional consequences on other biological functions. As seen above, all the membrane proteins presented are regulated in some way through endocytosis. What if U24 could block endosomal recycling in the context of oligodendrocytes and thereby lead to an imbalance in one or more of these proteins? What if U24 can directly interfere with proteins in neurons? As mentioned above, MAG and Nogo-66 are important bifunctional myelin-associated proteins, so that the change of their amount on the membrane surface could potentially be detrimental to myelin maturation and development. As mentioned in section 1.2, U24 has been shown to regulate endosomal recycling in T-cells (resulting in a down-regulation of TCR and TfR). Sullivan and Coscoy complemented these studies by trying to identify WW domains that would be essential to control this process. (172) They found that the PY motif common to all U24 variants is required for down-regulation and that TCR/CD3ε and TfR are down-regulated through a common mechanism. Since PY motifs are important for group 1 WW domain binding (section 188.8.131.52), they tested the binding capacity of seven human Nedd4 family E3 ubiquitin ligases (Nedd4 family E3s): Nedd4-1, Nedd4-2, AIP4, Smurf1, Smurf2, WWP1, and WWP2 (will be described in section 1.3.2). Immunoprecipitates revealed that U24 binds to Nedd4-2 constructs consisting of WW domains 1 and 2, 1-3, and 1-4, but not domain 1 alone. 22 They also found strong binding to WWP2, which also has 4 WW domains. In this case, all constructs (domain 1 alone, domains 1-2, domains 1-3, and domains 1-4) showed similar binding. A close look at the data in Sullivan’s thesis also suggests binding to WWP1, but this was not discussed nor investigated in detail. On the other hand, as Sullivan argues in his thesis, the interaction between U24 and group 1 WW domains may be an artefact due to the fact that these proteins are allowed to come into contact once the cells are lysed. In other words, it is not clear whether this interaction would be presented in a cell. In addition, experiments using dominant-negative mutants of the Nedd4 family E3 indicate that the function of U24 is not inhibited, leading to the conclusion that Nedd4 family E3s are not necessarily the cognate ligands of U24. Moreover, inhibition of protein kinase C (PKC) can reduce the effect of U24. Therefore, the interaction partner of U24 was concluded to be functioning downstream from phosphorylation events mediated by PKC isoforms. Kidney and brain expressed protein (KIBRA) was proposed to be a very interesting target by Sullivan et al, because it contains two WW domains and is involved in endosomal sorting. There is great value in investigating U24 and Nedd4-WW interactions. From the discussion in Sullivan’s thesis, U24 was found to be able to interact with more than half of the WW domains tried, suggesting it may contain an accessible and optimal sequence for the protein interactions. Indeed, about 12% of the functioning PY motifs reported in the literature can bind to more than 40% of WW domains in the in vitro protein-protein interaction screening. (81) It is therefore of great interest for structural biologists to investigate why U24 can bind so many domains. Since the core PY motif is nearly identical for all cases (PPxY, allowing any residue only at one position), the flanking residues of the binding motif in U24 could be key for its behaviour, although some studies showed that the core PY motif can bind to a WW domain with full affinities. (173, 174) As for the biological relevance, using dominant-negative mutants of Nedd4 in cellular experiments may not rule out Nedd4 being a potential binding partner of U24. U24 might be overexpressed or function as an inhibitor, which might not lead to observable differences even using the dominant -negative mutants of Nedd4. In addition, PKC isoforms are implicated in TCR complex trafficking, as well as endocytosis. It has been demonstrated that Nedd4 could mediate the PKC-dependent regulation of 23 glutamate transporter GLT-1 and cationic amino acid transporter CAT-1. (175, 176) ENaC and voltage gated potassium channel Kv1.5 are also regulated by Nedd4 and PKC. (177, 178) In any case, a more in-depth look at U24 and WW domains should shed light into the optimal sequence of PY motif for most WW domains and provide a molecular mechanism of how U24 functions in endosomal recycling, an area of research where little is understood. In addition to the link between U24 and E3 ligases discussed above, parallel work on peripheral axons has demonstrated that endosomal sorting and targeting to the lysosome may be important in the degeneration of axons found in the peripheral nervous system. (179) Indeed a number of proteins involved in this endosomal pathway (e.g. Rab7) also exist in neurons and may therefore have similar functions in the CNS. 1.3 Potential binding partners of U24 To determine what the role of U24 may be in MS at a molecular level, binding interactions between U24 and specific targets were investigated. For the MBP mimicry hypothesis, the interaction between U24 and Fyn-SH3 was studied. For the endosomal recycling hypothesis, the interaction between U24 and Nedd4 E3 ubiquitin ligase was examined. In the following sections, brief backgrounds on these molecules will be given. 1.3.1 Fyn-SH3 domain The structure of Fyn-SH3 domain has the classic topology of SH3 domains. It binds mainly to PxxP motifs, with the ligand binding in either direction. (180–182) The RT loop and n-Src loop mainly contribute in ligand binding (Figure 1.9). (183) The arginine and threonine in the RT loop are, however, not conserved in all types of SH3 domains. (184) The globular folding is very stable and there is minimal structural change of the SH3 domain upon ligand binding. (66) The interaction between SH3 domain and its ligand will activate downstream biological events, recruiting other kinases or regulating actins in cytoskeleton. (64) Therefore, blocking the interaction has a big impact. 24 The affinities between Fyn-SH3 and its ligands have a broad range: the dissociation constants are normally around 700 nM to 3 mM. (185–187) A PxxPxR ligand from Hepatitis C virus NS5A protein was found to bind to Fyn-SH3 better than its cognate substrate ligand, (185–187) suggesting a possible recruitment or blocking of Fyn’s function for viral survival. Fyn-SH3 has also been found to tolerate the non-proline motif RKxxYxxY within the same binding pocket, but the interaction mechanism may be different. (188) Figure 1.9 The structure of SH3 domain and proline-rich peptide complex The structural model in the above figure is a Fyn-SH3 domain in complex with a proline-rich peptide, PPRPLPVAPGSSKT. The structure was deposited in the protein data bank (PDB) as 1A0N and processed using Pymol. The SH3 domain is shown in cartoon mode, and the RT loop and n-Src loop are highlighted in pink. The prolines in the peptide ligand are shown as sticks. The surface of the complex is coloured according to the APBS functionality in Pymol. (189) Red indicates a negative charge, while blue indicates a positive charge. 184.108.40.206 Architecture of Fyn tyrosine kinase Fyn tyrosine kinase belongs to the Src family non-receptor tyrosine kinases (SFKs). Src tyrosine kinase is the first member of this family and is the product of a proto-oncogene, c-src. The v-src gene, a mutated version originating from c-src, is a transforming gene, which changes a normal cell to a cancerous cell. (190) Similar to c-src, fyn is a proto-oncogene. There are two isoforms of Fyn, depending on whether they are expressed in T-cells or brain cells. The two isoforms have most 25 residues in common, except for about 50 amino acids located in the region between the SH2 and SH1 domains. (191) Fyn has four Src homology domains: Src homology 4 (SH4) domain, a unique domain, SH3, SH2 and SH1 domain from its N-terminus to C-terminus. The SH4 domain is in charge of protein anchoring and binds to the plasma membrane through acylation of N-terminal amino acids. (192) SH3 and SH2 domains are protein interaction domains, which bind to PxxP and phospho-tyrosine respectively, while SH1 domain is the tyrosine kinase domain. The activity of Fyn is regulated by other tyrosine kinases, such as C-terminal Src kinase (Csk). The inactive form of Fyn is one in which the binding sites of SH2 and SH3 domains have been blocked by intramolecular interactions. (193, 194) 220.127.116.11 Important role of Fyn in CNS The connection between Fyn and myelination in CNS has been established by analyzing Fyn-deficient mice. Different levels of hypomyelination were observed in different brain regions, and a severe reduction in the amount of myelination was observed in MAG/Fyn double deficient mice. (195–197) Apart from myelination, cellular experiments have shown that the growth of neurons from Fyn-deficient mice was limited: there are less dendritic spines and there is a lowered density of synaptic button-like structure found on the dendrites. (198) Hence, Fyn is indispensable for CNS development. As a signal transduction molecule, Fyn participates in several cellular processes. It associates with neuron specific integrin α6β1, a transmembrane receptor that promotes cell-extracellular matrix (ECM) interactions, to translate the signal into other cellular events, such as cell survival, membrane formation, or to switch between proliferation and differentiation signals. (199, 200) Downstream from Fyn signalling are a host of other cellular functions, such as the regulation of cytoskeletal dynamics, cell polarity and others (201–203) There is another pathway for Fyn to regulate outgrowth of oligodendrocytes: Fyn could recruit microtubules through its direct interaction with the Tau protein, a cytoskeleton protein stabilized in microtubules. A partial or complete block of 26 downstream signalling from Tau would result in a lowered myelinating capacity. (204, 205) Fyn also promotes oligodendrocyte differentiation, myelination and hippocampal development in the CNS. (206, 207) Outside of neurons, Fyn also plays roles in different cell signalling pathways. It serves in the process of T-cell development and is required for mediating TCR-based cell signalling. (208) In fyn knockout mice, natural killer T cell development was blocked, which indicates that Fyn is unique and plays an important role. (209) What is more, Fyn is also required for intracellular signalling during mitosis, a process that is activated by platelet-derived growth factor (PDGF). Finally, Fyn was detected in the cleavage furrow ingression during meiosis and mitosis. (191, 210, 211) 1.3.2 WW domains in Nedd4 family E3 ubiquitin ligase WW domains are small substrate recognition domains found in a range of proteins. One family in particular is the Neural precursor cell (NPC) expressed developmentally down-regulated protein 4 (Nedd4) E3 ubiquitin ligase, which contains multiple WW domains from different groups, (81) with most of them being group 1 WW domains, i.e. domains which preferentially bind PY motifs. WW domains from Nedd4 typically have three β strands. Some of these domains do not have both conserved tryptophans. (212) From mutational studies, three conserved hydrophobic residues and a histidine, which are localized on the surface of the WW domain, form the binding site. (213) WW domains are essential for the functions of Nedd4 family E3 ubiquitin ligase (Nedd4 family E3), and blocking the protein-ligand interaction could lead to diseases such as Liddle’s syndrome. (71) Normally, the WW domains within one Nedd4 family E3 ubiquitin ligase are not identical. There will be differences at the binding site or beyond for ligand preference selection, via the interaction by itself or by tandem WW domains. The affinities between PY and WW domains seem to depend on the residues in the turn between the first and second β sheets (β1-β2 turn, or β1-β2 loop), and the third β sheets (β3 sheet). For example, a WW3* domain from Drosophila has a very high affinity towards the non-canonical LPSY motif due to the specific residues at the β1-β2 turn. (214) Combining the phosphorylation site up- or down-stream from a PY motif, the ligand could 27 selectively bind to one WW domain, hence guiding the WW domain containing E3 to work sequentially. (215) On the other hand, either phosphorylation of the threonine in the third β strand of WW domain or the tyrosine in the PY motif could abolish or diminish the interaction. (216–218) The PY motif could also bind to more than one domain. Tandem WW domains could cooperate to bind one PY motif, usually with enhanced affinities. (74, 78, 212, 214, 215, 219) Figure 1.10 The structure of a WW domain and ligand containing a PY motif in complex The structure is a human Nedd4-WW3 domain in complex with a peptide ligand derived from epithelia sodium channel proline-rich region. The PDB access code for this complex is 2M3O. The complex is shown in cartoon mode and the PPPxY segment in the ligand is displayed using sticks. The binding pocket on the WW domain is coloured pink, and the surface is shown in transparent, with positive and negative charge coloured by APBS. (189) 18.104.22.168 Nedd4 family E3 ubiquitin ligase Nedd4 is one of ten proteins (Nedd1-10) whose expression levels are reduced during mouse brain development. (220) The Nedd4 gene was recognized as homologous to the E6-AP carboxyl terminus (HECT) domain E3 ubiquitin ligase (for details, see section 22.214.171.124), and it contains an N-terminal C2 domain (spliced out in some isoforms), four WW domains and a C-terminal HECT domain. Currently, there are nine HECT domain E3s, which share similar structures, classified in this E3 family. (221) The C2 domain is a lipid-binding domain, which could anchor the ligase to the plasma membrane or endosome, for protein cargo sorting and labeling. (222, 223) The WW domains, 28 of which there are two to four positioned tandemly in a Nedd4 family E3, are in charge of ligand recognition and selection, while the C-terminal HECT domain carries out the ligation. (71, 224) The members of this family include, Nedd4 (or Nedd4-1), Nedd4-like (Nedd4L or Nedd4-2), ITCH (named after its discovery in itchy mouse), WW domain containing E3 ubiquitin protein ligase 1 (WWP1), WWP2, Smad protein ubiquitin regulatory factor 1 (Smurf1), Smurf2, Nedd4-like ubiquitin protein ligase 1 (NEDL1 or known as HECW1) and NEDL2 (or known as HECW2). (221, 225) This family of E3s is in charge of several key signalling pathways, including cellular proliferation and differentiation. Nedd4 and Nedd4L regulate the sodium ion channels, growth factor receptors, endocytic adaptors and so on. (226) Because they are implicated in the membrane protein trafficking system, it has been shown that viruses could hijack them for budding. (227) The functions of ITCH are related to the immune system, with a mutation or deletion of this protein leading to autoimmune diseases in humans and mice. (228, 229) Smurf E3s are the main regulators of Smad proteins, which transduce extracellular signals from transforming growth factor-β (TGF-β) superfamily ligands. Mutations in Smurf E3s will extend the activation of growth factors and lead to cancer. (230) Similarly, WWP and other Nedd4 family E3s may have roles in cancers. WWP1 regulates Smad in the TGF-β signalling just like the Smurf E3s, while WWP2 and Nedd4 target tumour suppressor phosphatase and tensin homolog (PTEN) for degradation or translocation to the nucleus for its function. (221, 231–233) There are also reports which have shown that Nedd4 could be regulated by PTEN. (234, 235) Nedd4 family E3s are responsible for the regulation of cell development, while being tightly regulated by other regulators. The Nedd4 family E3s are primarily regulated by kinases and other E3s. It is known that kinases and E3s interweave their functions by down-regulating each other. This is achieved through ubiquitination and interaction with a 14-3-3 protein, which recognizes Rxxp(S/T)xP motifs. (221, 236, 237) E3 can form an auto-inhibitory or inactivated conformation, through the interaction between the C2 and HECT domains. Inhibition can be achieved by inter- or intra-molecular interactions, and the cellular concentration of calcium regulates this process. (238–241) Besides the 29 calcium concentration, modification or interaction would activate E3 from its auto-inhibitory state. Phosphorylation by c-Src on the tyrosine of Nedd4 could release it from its auto-inhibitory state. This could also be achieved by activators, like the Nedd4 family-interacting proteins (NDFIPs), binding to the WW domains in Nedd4 family E3. (242–244) Nedd4 family E3 could undergo ubiquitin ligation by itself, or with another member in the family. This is mediated by the interaction between the PY motif within the HECT and WW domain. (245–247) Several family members of Nedd4 are under the regulation of other types of E3, like multi-subunit E3 complex: Skp, Cullin, F-box containing complex (SCF complex) and anaphase-promoting complex (APC/C). (248–250) There are also small ubiquitin-like proteins that block the interactions between E2s and HECT domain. (251) Overall, in order to functionalize as a regulator for other proteins, the Nedd4 family E3s are tightly regulated in the cell. In this thesis, experiments will focus on Nedd4 and Nedd4L, and Smurf2. Nedd4 and Nedd4L proteins (note these are two members of the Nedd4 family E3s) play a role in CNS (section 126.96.36.199), as well as in kidneys and lungs. The Smurf1 and Smurf2 proteins, on the other hand, play a role in the TGF-β signalling pathway. These proteins were chosen primarily because Nedd4 and Nedd4L have canonical WW domains, whereas Smurf2 has only one tryptophan in the conserved WW domains. A comparison of the binding interactions of U24 with these different types of domains will allow us to establish the importance of having canonical domains. This will also be important in establishing the biological role of U24. 188.8.131.52 Functions of Nedd4 and Nedd4L in the CNS As members of developmental down-regulated proteins, Nedd4 proteins are important for the development of the CNS. Nedd4 or Nedd4L knockout mice are pre-natal lethal, with two different consequences: reduced growth and slow development in nedd4-null mice, while an increased ENaC expression is found in the lung of nedd4l-null mice. (252–254) Subtle gait issues have also been found in nedd4 heterozygous mice which carry only one copy of the functional nedd4 gene. (255) Neuronal survival and outgrowth are regulated by Nedd4 and Nedd4L. Nedd4 promotes axon 30 branching and dendrite outgrowth by interfering with proteins that limit neuron outgrowth, such as Ras-related protein 2A (Rap2A) and PTEN. (256–258) Nedd4L controls the level of nerve growth factor (NGF) receptor and tropomyosin-related kinase receptor (TrkA), which function in neuronal survival and differentiation. (259, 260) Table 1.2 A summary of the functions of Nedd4 and Nedd4L in the CNS Regulator Targets Neuron outgrowth Nedd4 Rap2A, PTEN (256–258) Nedd4L NGF, TrkA (259, 260) Ion channels Nedd4 NaV, CaV1.2 (261, 262) Nedd4L NaV, KCNQ (261, 263) Neurotransmitters Nedd4 AMPA receptors (264) Nedd4L EAATs, DAT (265–267) Not surprisingly, Nedd4 and Nedd4L play a role in regulating the ion channels found in neurons. This includes several voltage-gated sodium channels (NaV), two potassium channels (KCNQ2/3 and KCNQ3/5) and a calcium channel (CaV1.2). (261–263) Through adjusting the number of ion channels present in the neurons, the physiological and electrical properties of neurons are controlled by Nedd4 and Nedd4L. These ubiquitin ligases are also involved in the regulation of neurotransmitters via their transporters and receptors, i.e. amino-3-hydroxy-5-methyl-isoxazole-4-propionic acid (AMPA) receptors, glutamate transporters EAAT, and dopamine transporter (DAT). These processes are also regulated by the kinases, like PKC discussed above, which influence the activities of Nedd4 and Nedd4L. (264–267) Table 1.2 summarizes the function of Nedd4 and Nedd4L in neurons and their identified binding partners. 31 184.108.40.206 E3 ubiquitin ligase Ubiquitination is a post-translational modification of proteins, in which ubiquitin is covalently linked to the targeted protein. The ubiquitin signalling tag serves as a marker in different cellular processes, such as plasma membrane endocytosis, sorting to multivesicular bodies (MVB) or the lysosome, and the regulation of cytoplasmic protein functions. (268) These functions are achieved through three different types of ubiquitination, which were classified based on the number and position of where ubiquitin is attached: monoubiquitination, multimonoubiquitination and polyubiquitination (Figure 1.11). Polyubiquitination is mainly associated with protein degradation. (269) Just like phosphorylation, ubiquitination is a reversible modification. The reverse reaction is called deubiquitination, which is catalyzed by deubiquitinase (DUBs). (270) Ubiquitination is generally carried out sequentially by a set of three enzymes: E1 ubiquitin-activating enzymes (E1s), E2 ubiquitin-conjugating enzymes (E2s), and E3 ubiquitin ligases (E3s). Ubiquitin is expressed as a precursor protein fused to another protein, then cleaved by DUBs or ubiquitin-specific proteases (USPs), and finally activated by E1s so that a covalent linkage to the catalytic Cys is formed, via a thioester bond. Afterwards, the ubiquitin is passed on to the E2s and some E3s by a transthiolation reaction for the final step of ubiquitination (Figure 1.11). (271, 272) Sometimes, the polyubiquitin chain assembly may require an additional ligase E4. (273) G76, the last residue at the C-terminus of ubiquitin is the only residue that can be activated and linked to another protein. (52) The covalent bond formed between G76 and the protein substrate is on a lysine residue from the protein substrate through an isopeptide bond, and sometimes on a serine, threonine, cysteine via an ester bond or the N-terminal residue via a peptide bond (Figure 1.11). (274–277) As for polyubiquitination, the second ubiquitin is attached to one of the seven lysines (K6, K11, K27, K29, K33, K48 and K63) of the previous ubiquitin by E3s. 32 Figure 1.11 Ubiquitination: types, mechanisms, and the structure of ubiquitin A) Types of ubiquitinations. Because there are several choices of lysine on ubiquitin to connect to the previous ubiquitin, the polyubiquitination topologies can be very complicated. It could be a chain built on the same lysine modification, such as K48, or modified on different lysines as the chain builds up. B) Ubiquitination mechanism using E1, E2 and E3 enzymes. The last step involving RING E3 is the same for PHD and U-Box E3s. C) The crystal structure of ubiquitin displayed in cartoon and surface mode. Seven lysines are shown in yellow and G76 is shown in pink. The E3 ubiquitin ligases (E3s) are classified into four types: HECT domain type, really interesting new gene (RING or RING finger) domain type, plant homeodomain finger (PHD finger) type and U-box type. (278, 279) RING domain E3 is the largest family of ubiquitin ligases, and HECT domain E3 is the only type that can catalyze itself, after loading ubiquitin onto its own active-site cysteine from E2. Besides HECT domain E3s, all the other types of E3 act as adaptors to facilitate the transfer of ubiquitins from E2s to substrate proteins. (278) There are mainly three 33 families of HECT domain E3s, Nedd4 family E3s, HERC (HECT domain and regulator of chromosome condensation 1-like domains containing E3) family and other HECT. (225) 1.4 Methods used to characterize protein interactions Protein-protein interactions are important for all biological processes. The identification of protein-protein interactions in vivo (e.g. using fluorescent microscopy, or coimmunoprecipitation experiments) leads to results that are often directly biologically relevant. The interpretation of such experiments can, however, be marred by the presence of other interaction partners or impurities. Using in vitro methods, where the system is simplified to a single protein-ligand pair, allows one to determine interaction strengths and affinities. There are several methods that can be used to study protein-protein interactions in vitro, namely isothermal titration calorimetry (ITC), surface plasmon resonance (SPR), fluorescence spectroscopy, and nuclear magnetic resonance (NMR) spectroscopy. In this dissertation, ITC and NMR spectroscopy are used to characterize the interaction of U24 with its binding partners, and so will be briefly described below. 1.4.1 Protein NMR spectroscopy NMR spectroscopy is commonly used to determine the structure of all sorts of molecules. For proteins, the general approach to determine three-dimensional structure or to investigate protein-ligand interactions relies on the use of 2D 1H-15N heteronuclear single quantum coherence (HSQC) spectroscopy. 15N enriched protein sample is necessary for this type of experiment. Each amide bond in the protein backbone will yield one signal in the spectra, representing the chemical environment of this residue. The N-H side chains of arginine, asparagine, glutamine, tryptophan, and rarely histidine can be recorded in HSQC spectra as well. (280–284) 34 Figure 1.12 HSQC, coupled HSQC and TROSY-HSQC spectra of small (upper row) and large (bottom row) proteins A1) HSQC, A2) coupled HSQC and A3) TROSY-HSQC cross-peaks for a small protein, and B1), B2) and B3) are the corresponding cross-peaks in a large protein. The HSQC chemical shift will be the averaged when decoupled. Instead of yielding a broad decoupled peak seen in HSQC of a large protein as in B1), the TROSY-HSQC experiment select the sharpest peak to yield highly resolved spectra (as in B3). For structure determination, HSQC spectra only provide limited information. Therefore, additional dimensions are added onto the basic HSQC experiment to yield 3D experiments. The third dimension could provide through-space information (nuclear Overhauser enhancement spectroscopy (NOESY)), or proton connectivity information within one residue (total correlation spectroscopy (TOCSY)). The intensities of NOE signals from other protons can serve as distance restraints to yield a protein structure. Such 3D HSQC-NOESY and 3D HSQC-TOCSY experiments are generally amenable to soluble proteins of moderate size. If the protein (or protein complex) is larger, additional pulse sequence elements, such as transverse relaxation optimized spectroscopy 35 (TROSY) need to be used, (285) to give well resolved TROSY-HSQC spectra. Otherwise, the HSQC peaks will be broad due to the fast relaxation found in large proteins. (see Figure 1.12 for an example) For a full review of NMR methods used for protein NMR spectroscopy, please consult an excellent book on the subject. (286) 1.4.2 Quantifying protein interactions using NMR spectroscopy 1H-15N HSQC spectroscopy is often used to study protein-protein interactions. Since chemical shifts are sensitive to the chemical environment around a nucleus, the binding site of a protein can be identified by comparing the spectra of an apo sample (protein without ligand) with one in which the protein is in the bound form. Moreover, binding affinities can be obtained through a NMR titration experiment. Typically, unlabelled ligand (peptide in this thesis) is titrated into a sample composed of the 15N labelled protein domain. HSQC spectra are then recorded before and after each ligand addition. Changes in the chemical shifts or in signal intensities are monitored. How this data is then used to extract an association constant (Ka) then depends on whether the protein is undergoing slow, intermediate, or fast exchange. In NMR, chemical exchange refers to any process in which a nucleus exchanges between two or more environments and is reflected in the NMR parameters (e.g. chemical shift, J-coupling, etc.). In the specific situation of protein-ligand interactions, the two environments experienced by an amide proton in the binding pocket are the environment in the apo and bound forms, respectively. The effect of the exchange event on a given NMR parameter will depend on how fast the event occurs relative to the time scale of the NMR observable. Depending on the conformational exchange rate between the apo and bound form of a certain amide in the protein, the signal containing partial apo and bound form will either appear in the middle, will disappear or will show up at both positions with different intensities. The exchange rate is the sum of forward and backward reaction rates, kex=k’on+koff. The rate k’on is the pseudo first-order reaction rate constant and is related to the association rate constant kon using the equation k’on=kon[ligand]. If the exchange rate is much faster than the difference of resonance frequencies, 36 Δω=ωA – ωB between the apo and bound signals, an averaged resonance will be observed at a fraction-weighted position, namely at ω=pAωA+pBωB, where pA and pB represent the fraction of apo and bound forms, respectively. When the exchange rate is slowed down to near Δω, the peak shape will be broadened and in the extreme case, it will be difficult to distinguish it from the noise. The signals will not be averaged if the exchange rate is much slower than Δω. In this case, both signals will show up at the original frequencies ωA and ωB, but with different intensities, proportional to its fraction. Normally, fast exchange is typical of weak binding interactions, whereas slow exchange is often found in the case of strong binding. Apo+ligand ⇆ Bound (1.1) Figure 1.13 Protein dynamics probed by NMR spectroscopy. Fast, intermediate and slow exchange between apo and bound form protein can be probed by NMR using A) 2D spectra or B) 1D spectra. 100% apo protein is red, ~20% apo or 80% bound is orange, 60% apo or 40% bound is green and 100% bound is blue. 37 Protein dynamics can also be measured using additional methods. (287) In this dissertation, slow exchange is further investigated using exchange spectroscopy (EXSY) applied to the backbone amides. The experiment is also called ZZ exchange and consists of the basic HSQC sequence with an extra delay added before detection. (288–290) This type of experiment can provide kinetic information for a protein-ligand interaction. Without the additional delay, two sets of signals will show up in the spectra (slow exchange), representing the apo and bound form, respectively. When the extra delay is in the sequence, cross peaks from exchanging 1H nuclei will arise, resulting in four peaks (Figure 1.14A). The intensities of the cross-peaks will grow as a function of the length of the delay, while the original signals, auto-peaks, will decrease. A schematic representation of the pulse sequence building blocks and observed amides are shown in Figure 1.14B. INEPT is short for insensitive nuclei enhanced by polarization transfer and is a common method in NMR to transfer magnetization from sensitive nuclei, such as 1H, to insensitive nuclei, such as 13C and 15N. The evolution time is a period in pulse sequences of multi-dimensional NMR experiments, in which the magnetization, i.e. 15N in Figure 1.14B, is allowed to evolve as a function of time t1. The resulting EXSY data set consists of four peaks, which include two auto-peaks, the peaks from Apo (AA) and Bound (BB) themselves, and two exchange cross-peaks, Apo to Bound (AB) and Bound to Apo (BA). An example of the peaks is shown in the Figure 1.14. The exchange rate kex can be obtained by fitting the intensities of the peaks in the data set obtained as a function of T, using the equations below. (289, 291–293) A set of typical curves is shown in Figure 1.14C. (289, 290) ddt[MzA(T)MzB (T)] = [-R1A0 -pBkex pAkexpBkex -R1B0 -pAkex] [MzA(T)MzB(T)] (1.2) The solution of the equation above is: [MzA (T)MzB(T)] = [aAA(T) aBA (T)aAB(T) aBB(T)] [MzA(0)MzB(0)] (1.3) 38 in which aAA(T)=MAA(T)MAA(0)=12[(1+R1A0 -R1B0 +kex (pB-pA)λ+-λ-)e-λ+T+ (1-R1A0 -R1B0 +kex(pB-pA)λ+-λ-) e-λ-T] aBB(T)=MBB(T)MBB(0)=12[(1-R1A0 -R1B0 +kex(pB-pA)λ+-λ-) e-λ+T+(1+R1A0 -R1B0 +kex (pB-pA)λ+-λ-) e-λ-T] aAB(T)=MAB(T)MAA(0)=kexpBλ+-λ-[e-λ-T-e-λ+T] aBA(T)=MBA(T)MBB(0)=kexpAλ+-λ-[e-λ-T-e-λ+T] (1.4) and λ±=12[R1A0 +R1B0 +kex ±√(R1A0 -R1B0 +kex (pB -pA))2+4pApBkex2 ] (1.5) In the equations above, pA represent the fraction of apo form. R1A0 is the longitudinal relaxation rate constant of the apo form. MzA(T) is the magnetization of apo form in Z direction†. Similarly, the symbols with subscript B represent the fraction, longitudinal relaxation rate constant and Z direction magnetization of the bound form. † MzA(T) here noted is ΔMzA(T) in the original version, (291). ΔMzA(T)=MzA(T)- MA0 . MA0 represents the equilibrium longitudinal magnetization of apo form, and MA0 =pAM0 . M0 is the total equilibrium longitudinal magnetization. It sometimes is ignored in the data fitting, as it is considered to be much smaller than the developed magnetization. 39 Figure 1.14 Slow exchange of Apo (A) and Bound protein (B) using EXSY. a) Examples of HSQC spectra, and EXSY with delays. Red is the signal of Apo (A), and blue is Bound (B) protein. Blue/red and red/blue peaks represent the exchanged signals. b) Schematic representation of HSQC pulse components and observed N -H amides. c) The intensities of four types of peaks change as a function of the mixing time T. The intensities for AB and BA peaks should be equal according to the equations, but are drawn differently in the figure for clarity. 40 1.4.3 Isothermal titration calorimetry Heat is a general phenomenon of chemical and physical processes. Unlike NMR, where only localized information is measured, ITC captures the overall effect of protein-ligand interactions in a very precise manner. The ligand is titrated into the protein solution, which is housed inside an adiabatic jacket (Figure 1.15C). The ligand is added using a syringe, which is fused to a paddle-shaped needle. The solution inside the sample cell is stirred. An identical cell, filled with deionized water, is used as reference. Both cells are maintained at the same temperature and a constant reference power is applied to the reference cell. A feedback power, which is connected to the heat sensor between the two cells, is applied to the sample cell. The calorimeter measures the temperature difference between two cells by recording the applied power differences between the reference cell and sample cell during the experiment. This recorded data is also referred as DP signal. Upon ligand addition, heat generated by an exothermic reaction will cause an increase of sample cell temperature, namely ΔT≠0. In order to maintain two cells isothermal, less feedback power is needed to apply to the sample cell. This results in a negative change in the DP signal. This signal is plotted for each small addition of ligand (typically 20-26 in total) until the ligand is fully bound to the protein. The change of DP is no longer observed upon ligand addition. The integrated areas of the resulting peaks represent the heat released from the sample cell during an injection. These data are normalized against sample concentrations and plotted against ligand/protein molar ratios. 41 Figure 1.15 ITC data, fitting examples and schematic depiction of ITC. A) An example of ITC raw data (top) and normalized integrated heat plot (bottom). An example is marked in red. The heat of dilution, which is shown in grey triangles, was subtracted from the integrated heat before data fitting. Blue circles represent the first injection, which will be deleted before data fitting. B) The data is fitted using Origin 7, using a “one set of sites” model (a fitting method described in the Origin 7; it fits n, Ka, and ΔH for a single set of identical binding sites using standard Marquardt methods), where the resulting n may not be exactly 1.0. Ka is extracted from the slope at the inflection point of the fitted curve (red). C) Schematic depiction of ITC. A ligand solution (magenta) is injected into the protein solution (green) in sample cell. This part of the figure is adapted from (294) with permission of Springer. The raw data in A) record the DP signal, differential power between feedback power and reference power. 42 c =nKaMtot (1.6) Typically, the first injection will not be used for data fitting due to systematic errors, such as syringe bending or syringe leaking during the long equilibrium time which precedes data acquisiting (shown in blue circles in Figure 1.15A). The normalized integrated heat is fitted using non-linear regression to obtain the equilibrium association constant (Ka), the number of binding sites (n) and the enthalpy of the titration (ΔH) (Figure 1.15B). The quality of the data depends on how well the protein-ligand interaction is saturated and the curvature of the recorded data points, c value. This c value is defined by the Equation (1.6). (294) Mtot indicates the concentration of protein sample in the cell. According to the manuals of the microcalorimeters (VP-ITC and iTC200) used in this thesis, the optimal c value is 10-50 for VP-ITC, and 5-100 for iTC200. The c value is controlled to be 3-40 for the experiments performed on VP-ITC, and 1-10 on iTC200 in this dissertation, unless there are difficulties in making the protein sample, as mentioned in the methods. 1.5 Aims of the dissertation There are many emerging theories and models for cancers and neurodegenerative diseases such as MS where viruses are said to play a potential role. Viruses have been known to recruit host proteins to achieve their replication and survival. Roseoloviruses are very common among the population and play a direct role in a number of diseases such as some cancers, chronic fatigue syndrome, encephalitis and complications from surgery. These viruses are also suggested to have a link to MS. In this dissertation, I focus on the tail-anchored membrane protein U24 from Roseoloviruses HHV-6A and HHV-7, in order to identify its potential role in disease. As detailed above, it has been suggested that U24 may have two functions, namely that of mimicking MBP and the other of blocking endocytic recycling (Figure 1.7). In order to test which of these functions may be most important for MS, I have expressed and purified recombinant U24-7 protein. I have also studied the interaction of a peptide consisting of the first 15 residues of U24 with Fyn-SH3 and WW domains. 43 The specific studies described in this thesis are: 1. Development and optimization of the purification of recombinant U24-7, to yield large quantities of pure U24, required for structural studies (described in Chapter 2). 2. Preliminary structural studies of U24-7 using CD and NMR (described in Chapter 2). 3. Determination of whether U24-6A mimics MBP by competitively binds with Fyn-SH3 domain, as verified using NMR and ITC (Chapter 3). 4. Investigation of the interactions between U24s and the WW domains from Nedd4, using NMR and ITC (Chapter 4). 5. Exploration of the interaction of U24 with non-canonical WW domains from hSmurf2 using ITC (Chapter 5). In the studies, the work focused on both U24-7 and U24-6A since both proteins have been found to down-regulate T-cell receptors. However compared to HHV-6A, HHV-7 is less virulent and associated with fewer diseases. From the structure and function studies of the homologue from two highly related viruses, it is possible understand the differences between these two proteins, and perhaps even the two viruses. 44 Chapter 2 Chapter 2 Expression and Purification of U24-7, towards Preliminary Structural and Phosphorylation Studies 2.1 Introduction Approximately 20 to 30% of all genes in most genomes are membrane proteins, (295) and they are important for cell functions. Membrane proteins work as transporters or ion channels, receptors, enzymes and membrane supporting molecules. (296) Structural studies are essential to understand and rationalize the mechanisms associated with protein function. Membrane protein structure elucidation is considered to be challenging. These proteins can dynamically interact with lipids or detergents, so it is often very difficult to obtain crystals for X-ray crystallography. NMR spectroscopy, which does not require crystallization, is a promising method to investigate the structure of membrane proteins, especially for small ones. Previously in our group, the structure of U24-6A was studied using circular dichroism (CD) spectroscopy, and its secondary structure was found to be mainly α-helical. Further experiments using NMR spectroscopy were not successful due to the aggregation of the sample. (297, 298) U24 has been characterized as a unique type of tail-anchored membrane protein. (36) Apart from the C-terminus, most of the protein is out of the membrane and faces the cytosol. This unique topology requires different sets of molecular chaperones for membrane insertion. U24 can be inserted into the ER membrane or the mitochondrial outer membrane (MOM), before being delivered to its final location by vesicular transport. (299) The membrane insertion is carried out by protein complexes, such as ATPase Asna1/Trc40. (300) In general, the correct localization of tail- 45 anchored membrane proteins is essential for their function. (36) These functions are both membrane and cytosol related, such as protein translocation, regulation of other membrane proteins and vesicular traffic. For example, it is the bundle of tail-anchored membrane proteins in the Snap (soluble N-ethylmaleimide-sensitive factor attachment protein) receptor complex (SNARE complex) that mediates the transient fusion, known as kiss-and-run fusion, in neuron synapsis. (301, 302) In the case of U24, it is of great interest to study its structure, as such knowledge will help us to understand its function (which is reviewed in Chapter 1) and how it interacts with other proteins in various membrane compartments. It has been recently demonstrated that U24-7 functions similarly to U24-6A, and investigations of the U24-7 structure could be useful to interpret the function of U24 in general. Before any function or structural study, a method needs to be worked out to obtain recombinant U24-7 that is suitable for further studies. The procedures of obtaining milligram quantities of U24-6A has been developed by Tait et al, (297, 298) and the production of U24-7 should be obtainable with similar fusion constructs and solubilisation strategies. However, there are two cysteines in the transmembrane domain of U24-7, which are not in the transmembrane domain of U24-6A. This pair of cysteines might be oxidized during expression or purification. There are only seven amino acids in between the two of cysteines, which may be too short a length for forming an intramolecular disulphide bond. The impact of U24-7 forming a dimer, by intermolecular disulphide bonds formed in the transmembrane region of two U24-7 molecules, is unknown. This may give rise to challenges for the protocol development. (297) In addition, once U24-7 is produced and isolated, aggregation could still be an issue, just as i t was for U24-6A. The hydrophobic segment of U24-7 and U24-6A are exactly the same length, and amino acid types in both transmembrane domains are limited to hydrophobic residues. A continuous leucine and isoleucine segment in the transmembrane region of U24-7 could make NMR spectral assignment difficult (Figure 1.2). What is more, the sample conditions for structural studies cannot be directly “borrowed” from the case of U24-6A. Extensive optimization is required to ensure that 46 the solubilisation of U24-7 transmembrane domain does not denature the cytosolic domain. U24-6A has been proven to be a mimic of MBP in the sense that it can be phosphorylated and recognized in vitro by MAP kinase. (54) Unlike U24-6A, U24-7 does not contain the identical amino acid stretch to MBP, PPRTPPPS, thus U24-7 is not a molecular mimic of MBP. But, it does contain a threonine in the same position (HETPPPS), therefore it may still be recognized by MAP kinase or considered a phosphorylation mimic of MBP. Overall, this chapter focuses on the purification of recombinant U24-7 that could be used for NMR studies. Preliminary structural studies were done to validate the protein purification and sample preparation protocols are suitable. Finally, phosphorylation was performed to check if U24-7 could be phosphorylated in vitro. 2.2 Results 2.2.1 Cloning and expression optimization To obtain sufficient amounts of U24-7 for structural studies, a recombinant plasmid was designed consisting of a maltose-binding protein (Mbp)-hexahistidine fusion tag (6xHis tag) followed by a Tobacco Etch Virus protease (TEV protease) cleavage site (ENLYFQ/S). Factor Xa cleavage site on the vector was kept for troubleshooting (Figure 2.1). Thrombin was not used due to unspecific cleavage in the middle of U24-7 (data shown and discussed in appendix B4). TEV protease is a highly restrictive cysteine protease that works at 4 °C. It was chosen because of its high specificity and low working temperature. 47 Figure 2.1 Schematic representation of Mbp-6xHis-U24-7 protein The gene sequence of U24-7 was obtained through overlapping PCR using synthetic oligonucleotides (Integrated DNA Technologies). The oligonucleotides and primers used for the gene synthesis are listed in Table 2.1 at the end of this section. The synthesized full length DNA fragment, containing a 6xHis tag and a TEV protease cleavage site at the N-terminus of protein sequence, was cloned into vectors pMal-c2x and pMal-p2x using BamHI/HindIII sites, down-stream of the Mbp encoding malE gene. The cloned vectors were sent to NAPS unit (Michael Smith Laboratories, UBC) for DNA sequencing using pMal-seq primer (Table 2.1) to confirm correct insertion of the desired fragment. The difference between these two vectors is that pMal-c2x will direct the gene product to the cytoplasm, whereas pMal-p2x will direct it to the periplasm during expression. The expression levels of U24-7 were tested on both pMal-c2x and pMal-p2x constructs in three different E. coli strains, which showed promising yields for full-length U24-6A expression. (298) These three strains are SHuffle express (New England Biolabs, SHE), Origami 2 (Novagen, Or2) and C41 (DE3) (Lucigen). Two expression temperatures, 18 °C and 37 °C (30 °C in the case of SHE as per manufacturer ’s instructions) were tried. The cells were harvested and lysed before analysis using SDS-PAGE. The gel was stained using Coomassie G-250. The results (Figure 2.2) show that the pMal-p2x-U24-7 construct did not overexpress the full-length Mbp-6xHis-U24-7 in all three strains, independent of temperature. On the other hand, the expression yield of the pMal-c2x construct at any condition was high enough for obtaining purified U24-7. Expression in M9 media was also tried and no obvious difference in yield was observed. 48 Figure 2.2 Expression tests of U24-7 in pMal-c2x or pMal-p2x vectors, expressed in different strains Three E.coli strains, SHuffle express (SHE), Origami 2 (Or2) and C41 (DE3), were tested. Arrows point at designated MW of Mbp-6xHis-U24-7. “M” indicates the maker lanes. lane 1: No induction, pMal-c2x, SHE. lane 2: 30 °C, pMal-c2x-U24-7 in SHE. lane 8: 37 °C, pMal-p2x-U24-7 in Or2. lane 3: 18 °C, pMal-c2x-U24-7 in SHE. lane 9: 18 °C, pMal-p2x-U24-7 in Or2. lane 4: 30 °C, pMal-p2x-U24-7 in SHE. lane 10: 37 °C, pMal-c2x-U24-7 in C41. lane 5: 18 °C, pMal-p2x-U24-7 in SHE. lane 11: 18 °C, pMal-c2x-U24-7 in C41. lane 6: 37 °C, pMal-c2x-U24-7 in Or2. lane 12: 37 °C, pMal-p2x-U24-7 in C41. lane 7: 18 °C, pMal-c2x-U24-7 in Or2. lane 13: 18 °C, pMal-p2x-U24-7 in C41. 49 Table 2.1 Oligonucleotides used for U24-7 gene synthesize. Name Oligonucleotides sequence from 5’ to 3’ Oligo 1 CCAGGATCCCATCACCATCACCATCACTCTTCCGGTCTGGTTCCGCGTGGTTCTATGACTCACGAGACGCCACCACCTTCTTATAATGAT Oligo 2 ATTGATCGTACGTGGAGACAGATTTTCCTGATGCAGAAACACGGAATGATCGTGGAACATTTGCAGCATTACATCATTATAAGAAGGTGG Oligo 3 CTGTCTCCACGTACGATCAATTCCACTTCTTCCTCTGAAATCAAGAATGTGCGTCGCCGTGGCACTTTTATTATTCTGGCATGTCTGATC Oligo 4 CCTAAGCTTTCACGGTTTCGTACCGCCATAGCGAACATTGAAGATGTGCAGAATCAGAATCAGACACAGGATAACGGAAATGATCAGACATGCCAGAAT TEV read oligo 1 GGTATCCAGGATCCCATCACCATCACCATCACTCTTCTGGTCGTGAAAACCTGTACTTCCAGTCTATGACTCACGAGACGCCACCACC TEV read oligo 4 CATCACCTAAGCTTTCACGGTTTCGTACCGCCATAGC Primer 1 CCAGGATCCCATCACCAT Primer 2 ATTGATCGTACGTGGAGACAG Primer 3 CTGTCTCCACGTACGATCAAT Primer 4 CCTAAGCTTTCACGGTTTC Long P1 GGTATCCAGGATCCCATCAC Long P4 CATCACCTAAGCTTTCACGG pMal-seq Primer GGTCGTCAGACTGTCGATGAAGCC pMal-seqrv Primer CGCCAGGGTTTTCCCAGTCACGAC 2.2.2 Optimization of U24-7 purification Purification of U24-7 was carried out similarly to that of U24-6A, but the yield was very low. The final step (marked as step VII in Figure 2.3) was to precipitate U24-7 from solution using cold acetone, but nearly little to no protein was isolated at the end. Three ways were considered to circumvent this problem: 1) avoid this step, 2) change the solution in which U24 is dissolved before the precipitation step, or 3) optimize the acetone precipitation. In each subsection below the outcome of these possibilities is presented and discussed. An overview of the purification process and optimization is in Figure 2.3. 50 Figure 2.3 Overall purification process of U24-7 and optimizations done. The steps of U24-7 protein purification are marked in roman numerals from I to VII. The green ovals indicate the sections where various optimizations are described. 220.127.116.11 Elimination of the precipitation step An experiment was designed firstly trying to eliminate the acetone precipitation step. Without this step, the yield may be enhanced and there would be no need to work with precipitated protein, a denatured form of protein. The reason for an acetone precipitation step is to remove the detergent used throughout the purification process, Triton X-100. It is a common detergent for membrane protein extraction and purification and has been used for U24-6A extraction and purification. (297) However, Triton X-100 can not be removed by dialysis, due to its extremely small critical micellar concentration (CMC), 0.24 mM or 0.0155% w/v. Thus a precipitation step is required. Changing detergent during the purification could remove this step in the protocol, and this strategy will be described in this section. 51 Ideally, the solubilizing detergent for membrane protein purification can be easily removed by dialysis against the buffer used in the structural studies. The yield could be increased because there is no need for additional steps, such as dissolving the protein precipitate, removing all the salts by dialysis, then renaturing the protein. Learning from the experiences of purifying U24-6A, three detergents were selected to try during purification: sodium dodecyl sulphate (SDS), sodium deoxycholate (NaDoc) and 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS). Detergent could be changed during the 6xHis tag affinity chromatography step, in which the tagged fusion protein was immobilized on the column (Figure 2.3, step II). Then, the tagged fusion protein could be eluted with the elution buffer containing one of the detergents mentioned above. To ascertain which detergent is best suited for the purification process, the purification of U24-7 was carried out. The next step after the affinity chromatography is dialysis and TEV protease cleavage (Figure 2.3, step IV). The efficiency of the cleavage is important in selecting which detergent is compatible for the purification of the U24-7 fusion protein with a TEV cleavage site. Figure 2.4 TEV protease cleavage results of Mbp-6xHis-U24-7 in buffers with different detergents. Mbp-6xHis-U24-7, Mbp-6xHis tag and U24-7 are indicated with arrows on the gel. The digested fractions of Mbp-6xHis-U24-7 in buffers with different detergents were loaded on the gel. lane 1: 0.5 % Triton X-100 lane 2: 12 mM SDS lane 3: 6 mM NaDoc lane 4: 10 mM CHAPS 52 The experiment on the histrap column was carried out as described in section 2.5.5, and eluted fusion proteins Mbp-6xHis-U24-7 were divided into three parts. They were then dialyzed against three different buffers, 50 mM Tris pH 7.8 with (A) 12 mM SDS, (B) 6 mM NaDoc or (C) 10 mM CHAPS, respectively, before digestion by TEV protease. An experiment in which Triton X-100 detergent was used was performed in parallel as a control experiment. The digested fractions were analyzed using SDS-PAGE in Tris-Tricine buffers and stained with silver stain. The digestion result showed that there is no full-length fusion protein left in lane 1, where Triton X-100 is used, while all the other three lanes still have residual Mbp-6xHis-U24-7, as evidenced by the band at around 55 kDa. What is more, in the case of SDS, the cleaved U24-7 appears to have different charged isomers. Since the TEV protease is fused with a 6xHis tag, increasing the amount of protease would simply make it more difficult to remove during later purification. Therefore, keeping Triton X-100 throughout the purification was found to be the best, and further optimizations will make use of this detergent. 18.104.22.168 Purification buffer optimization The second method tested in order to extract more U24-7 was to change the solvent used to dissolve U24-7, before the acetone precipitation step. In the section above, the buffer, 50 mM Tris pH 7.8, was found to work well for the last step purification, which is the anion-exchange chromatography. This step was used to trap the impurities on the column, while U24-7 was collected in the flow-through fraction (Figure 2.3, step VI). Anion-exchange chromatography binds negatively charged protein, when the isoelectric point (pI) of the protein is smaller than the pH of its buffer. The pI of the cleaved recombinant U24-7 is 8.8, as calculated by ProtParam, (303) and U24-7 would have a positive charge in any buffer with a pH below 8.8. Since under the current conditions, U24-7 does not bind to the column but simply flows through, the buffer used for anion-exchange chromatography needed to be optimized. Phosphate buffered saline (PBS) buffer, pH 7.4, is a promising choice, as U24-6A could be easily precipitated from PBS buffer. Another buffer, 2-(N-morpholino)ethanesulfonic acid (MES) buffer, pH 6.5, was chosen due to its low pH. If the 53 purification using these new buffer systems could be achieved, then the acetone precipitation might be a viable next step. U24-7 was purified up to the step involving the second histrap column (as described in section 2.5.5). It was then directly separated into three aliquots without dialysis. Next, two of the aliquots were dialyzed against PBS or 50 mM MES buffer, respectively. 1 mM DTT was supplemented in both dialysis buffers. The purification using anion-exchange chromatography was carried out in the same fashion and the flow-through fractions (containing U24-7 protein), elution fractions and washing fractions were collected. In the case where the flow-through samples were too dilute, a concentrated flow-through faction was prepared as well. These samples were analyzed using SDS-PAGE, with Tris-Tricine buffer. The resulting gels were stained by silver stain and are shown below. It is important to note in examining Figure 2.5 that the thickness of the protein band is not directly proportional to the amount of the protein when using silver staining. The band thickness depends instead on the exposure time to the stain. The flow-through fractions shown in Figure 2.5, lanes A1, B1 and C1 show the presence of the U24-7 monomer. There are also U24-7 dimers in lane B1 and C1, indicating the U24-7, which was isolated using PBS and MES buffer, was a mixture of monomer and dimer. Less dimer was found in the flow-through fraction using Tris buffer. Very little U24-7 was detected in the washing fractions. As for the elution step, A3, B2 and C4 in the gel above, mainly contained the cleaved Mbp-6xHis tag. There were other impurities eluted in the case of PBS and MES buffer, including the dimer of U24-7. This indicates that both of these buffers were not efficient in separating U24-7 for anion-exchange chromatography. The yield was also found to be low. In conclusion, the tests suggest that even though it complicated the acetone precipitation step, Tris buffer was the best to isolate U24-7. 54 Figure 2.5 U24-7 isolation using different buffers for anion-exchange chromatography SDS-PAGE of different fractions using A) Tris buffer, B) PBS buffer, and C) MES buffer for anion-exchange chromatography. The arrows point to the band of U24-7 monomer. “x” indicates an empty lane. lane A1: flow-through fraction, Tris lane B3: concentrated flow-through fraction, PBS lane A2: washing fraction, Tris lane C1: flow-through fraction, MES lane A3: elution fraction, Tris lane C2: concentrated flow-through fraction, MES lane B1: flow-through fraction, PBS lane C3: washing fraction, Tris lane B2: elution fraction, PBS lane C4: elution fraction, MES 22.214.171.124 Salt induced acetone precipitation As mentioned above, U24-7 could not be precipitated out by cold acetone. Chloroform-methanol extraction was tried as well, but not much U24-7 was isolated from the Tris buffer. Considering that U24-6A could easily fall out of PBS buffer by adding cold acetone, similar approaches should work for its homologue U24-7. It is highly probable that a component of the PBS buffer (i.e. salt) plays a role in inducing the precipitation of U24-6A. Therefore experiments were carried out to test the effect of salt in order to improve the precipitation of U24-7. The Hofmeister series is a classification of ions based on the ability of the ion to disturb protein solubility. In the Hofmeister series, the first six cations and anions that decrease protein solubilities are listed in Figure 2.6. Coincidentally, PBS contains NaCl, KCl, Na2HPO4 and KH2PO4, nearly all of which are in the Hofmeister series. Because the anions normally have bigger effects on 55 protein solubilities than the cations, (304) NaCl, and Na2HPO4 were chosen for the test. (NH4)2SO4 was also tested, because it should be the worst salt to maintain the protein in solution, in theory. The aim of this experiment was to test whether U24-7 can be precipitated from the Tris buffer by adding salt before performing the acetone precipitation, and whether the precipitated U24-7 could be easily dissolved in 10 mM SDS solution, useful conditions for NMR studies. Anions: F - ≈ SO42 - > HPO42 - > acetate > Cl - > NO3 - Cations: NH4+ > K + > Na + > Li + > Mg2 + > Ca2 + Figure 2.6 The first six anions and cations in the Hofmeister series. Table 2.2 U24-7 concentrations, after recovery and the ratio of the amount recovered relative to the starting concentration of 46.4 µM. The concentrations were determined using the BCA assay. Salt added U24-7 concentrations (After recovery) Ratio (After/Before) (NH4)2SO4 25.3 μM 54.4% a KH2PO4 - b No precipitation NaCl 47.5 μM 102% Na2HPO4 39.9 μM 86.0% a The colour of the BCA assay mixture turned blue. b No precipitate was observed after the addition of cold acetone. The concentration of purified U24-7, which was in Tris buffer, containing 0.5 % of Triton X-100, is 46.4 μM. This solution was divided into four parts and supplemented with different salt solutions, before the addition of cold acetone (see details in section 2.5.7). The concentrations of U24-7, before the precipitation and after redissolution in 10 mM SDS, are shown in Table 2.2. KH2PO4 did not yield precipitated protein from the solution and was therefore used as a control. The percentages of the sample recovered from the precipitation indicate the efficiency of the extraction. SDS-PAGE was used to detect the residual U24-7 in the supernatant in the acetone solution. These samples were dried in a fume hood before loading on the gel. The gel was run in 56 Tris-Tricine buffer and stained by silver stain. The results are shown in Figure 2.7, where a lane showing no band for U24-7 protein represents the best condition for protein precipitate, as in this case, it is absent from the supernatant. All three salts tried, NaCl, Na2HPO4 and (NH4)2SO4, could induce precipitation of U24-7 from acetone. However, identifying which is the best salt to use is difficult based on the numbers shown in Table 2.2 because in the case of (NH4)2SO4 side reactions most likely occurred, resulting in a blue solution. In addition, the use of certain salts may make subsequent steps in the purification or in the structure determination of the protein complicated. For instance, NaCl was shown to work with cold acetone for U24-7 extraction, but the precipitates were difficult to dissolve in 10 mM SDS. The sample was sonicated for hours before finally yielding a clear solution. Therefore, NaCl was not included in the following study using SDS-PAGE. For a salt to greatly induce the precipitation, there should be little to no protein detected on the gel. In the figure above, the supernatants resulting from (NH4)2SO4- and KH2PO4-induced acetone precipitation has bands for U24-7 and its dimer, while nearly no protein could be detected in the lane where Na2HPO4 was used. Combining the qualitative ratio of the amount recovered, Na2HPO4 appears as the best choice to induce precipitation. The precipitate resulting from Na2HPO4-induced acetone precipitation is fluffy and could be easily dissolved. As a result, the yield of U24-7 could be largely increased and this method was used in the following sample preparation for structural studies. 57 Figure 2.7 Supernatants from salt induced acetone precipitation. U24-7 and U24-7 dimer are indicated by arrows. The supernatants are from different salt induced acetone precipitations. lane 1: (NH4)2SO4 lane 2: KH2PO4 lane 3: Na2HPO4 lane 4: diluted dialysed U24-7 in Tris buffer. 2.2.3 Secondary structure The secondary structure of U24-7 was first characterized by circular dichroism (CD) spectroscopy. Different lipids or detergent micelles were tried to reconstitute U24-7. Using CD could therefore also provide useful sample conditions for the structure elucidation using NMR spectroscopy. The concentrations required for CD experiments are not high, typically 25-100 μM. CD spectra were recorded at room temperature for U24-7 in short chain lipids, 1,2-dihexanoyl-sn-glycero-3-phosphocholine (DHPC), Triton X-100, SDS and NaDoc. The sample concentrations were determined using BCA assay, and the raw ellipticities (in milidegree) were converted to mean residue ellipticities. The results are showed in Figure 2.8. 58 Figure 2.8 CD spectra of purified recombinant U24-7 in DHPC and detergents. A) CD spectra of U24-7 in 20 mM DHPC. B) CD spectra of U24-7 in 10 mM NaCl and 0.5% Triton X-100 (green curve), 10 mM NaCl and 10 mM SDS (red curve), or 10 mM NaCl and 0.25% NaDoc (blue curve). The CD spectra of U24-7 in DHPC micelles have two minima at 205 nm and 225 nm, which are features characteristic of helical structure. U24-7 clearly adopts a well defined helical structure in DHPC, even in the absence of salt. As for the tests with detergents, 10 mM NaCl was added to increase solubility. U24-7 in Triton X-100 clearly yields a sample where the protein is not structured. 59 The CD spectra obtained in the presence of Triton also display a lot of noise from absorptions of the detergent itself, around 275 nm and 283 nm (Product information sheet, Sigma-Aldrich). The spectra from U24-7 in SDS and NaDoc micelles are similar to those obtained in DHPC, indicating both detergents are suitable for structural studies. The differences in mean residue ellipticities between the spectra of U24-7 in DHPC and detergent may be due to the fact that detergent may interfere with the concentration determination using BCA assay. The error associated with determining the concentration is carried to the mean residue ellipticity. 2.2.4 A double Cys to Ser mutation of U24-7 and its CD spectra Efforts were made to make a well-behaved NMR sample for U24-7, but the native protein could never be fully dissolved. 5-10 mM DTT was always supplemented in the sample. None of the detergents, such as SDS, NaDoc, n-dodecylphosphocholine (DPC) or lauryldimethylamine oxide (LDAO), was helpful to make a sample around 0.2 mM, the minimum concentration for NMR studies. This concentration is much higher than the ones used for CD spectroscopy, or acetone precipitation recovery. In samples were the protein was partially dissolved (e.g. U24-7 sample in 100 mM DPC) the supernatant did yield signals in a correlation spectroscopy (COSY) experiment, but the resulting spectra were not resolved, i.e. most of the amide protons are clustered (Figure C.1). After filtering the sample through a 0.22 μm membrane, much fewer amide protons could be detected in the COSY spectra. This is an indication of loss of protein and could be due to aggregation. Although the supernatant contained U24-7, the protein was not dissolved as individual molecules. The aggregation could be due to the interaction between transmembrane cysteines, since a maximum of 10 mM DTT was used, or alternatively, it could be due to the high hydrophobicity of the transmembrane segment. In order to try to circumvent the problems described above, a mutated version, where two cysteines are mutated to serines (dCS), was created. The U24-7 dCS was purified and isolated using the same method outlined above for wild type U24-7. A concentrated U24-7 dCS was made in citrate-phosphate buffer containing 50 mM SDS and 150 mM DPC, pH 5.8. The CD spectrum for 60 this sample was recorded at 30 °C. This sample was then filtered using a 0.8 μm filter, followed by a 0.22 μm filter. CD spectra were recorded after each filtration. This experiment aims to test whether these is significant aggregation in the sample or whether the sample is so hydrophobic that much of it sticks to the filter. The concentration of this dCS version of U24-7 was hard to determine using BCA assay, because there are no cysteine or tryptophan residues in the sequence. The concentration is also difficult to determined by comparison with protein standards like bovine serum albumin (BSA). The concentration was estimated by comparing to typical cell pellet yields (ca. 0.3 mM). The CD spectra of the filtering experiment are shown in Figure 2.9. Figure 2.9 CD spectra of U24-7 dCS protein in citrate-phosphate buffer, 50 mM SDS, 150 mM DPC, pH5.8. The CD spectra of the sample before filtrations (marked as recovered U24-7) and after filtrations using 0.8 μm and 0.22 μm membranes are shown in different colors. The CD spectra of recovered U24-7 dCS, before and after filtration were found to be nearly the same, even though they do not contain the characteristic two distinct minimal absorptions. It was hard to determine the secondary structure of U24-7 dCS using these data. Considering there are only two residues different from the wild type, it is unlikely that such a small sequence change would cause such a large secondary structure rearrangement, leading to e.g. the exclusive formation of beta sheets. At the same time, the CD spectra above showed no distinct features of β sheets of random -150 -100 -50 0 50 200 210 220 230 240 250 Raw ellipticity (mdeg) Wavelength (nm) 150mM DPC/50mM SDS in CP buffer recovered U24-7 0.8 µm filter 0.8 µm and 0.22 µm filter 61 coil, as the presence of distinctive negative or positive peaks around 215 nm are absent. In order to determine whether the use of other detergents would result in a more helical structure, samples were prepared instead with SDS alone. When 200 mM SDS was used, similar spectra to those observed in Figure 2.9 were obtained (Figure D.1). In contrast, when a higher proportion of SDS was used, the helical character of U24-7 dCS was observed, as seen in Figure 2.10. The CD spectra of U24-7 dCS in citrate-phosphate buffer, 25 mM SDS also demonstrated a temperature stability in the range from 15 °C to 55 °C (Figure 2.10). The data in the wavelength range between 190 nm to 205 nm was omitted because they are very noisy, as is typical for many detergent containing samples. Figure 2.10 CD spectra of U24-7 dCS in citrate-phosphate buffer, 25 mM SDS, pH 5.8 at different temperatures. 2.2.5 Preliminary NMR results As mentioned in Chapter 1, 1H-15N HSQC spectroscopy is commonly used for protein studies. 15N uniformly labelled U24-7 dCS mutant was prepared as described in sections 2.5.5 and 2.5.7, and then dissolved in citrate-phosphate buffer, pH 5.8, 25 mM SDS. The sample concentration was 62 estimated to be about 0.2 mM. Three types of HSQC type experiment, namely HSQC, gradient enhanced & sensitivity enhanced HSQC (seHSQC), (280–283) and TROSY-HSQC, (305–310) were performed at 15, 25, 35 and 45 °C, on an 850 MHz NMR spectrometer equipped with a cryo probe (see section 2.5.10 for details). The overlays of the seHSQC spectra obtained at different temperatures are shown in Figure 2.11. Most chemical shifts of the amide protons are between 7.5 to 8.7 ppm, which is a small range (Figure 2.11). Most regions of the spectra are well resolved, which is completely different from that in the aggregated sample of U24-6A. (298) The intensities of the resonances are low at lower temperatures except for a handful of resonances. The spectra of higher temperatures show more resonances, and the signals appear to be well resolved as well. Increasing amide resonances has been observed in the case of U24-6A. The number of the resonances observed on the spectra is increased from 30 at 15 °C to 70 at 45 °C. Figure 2.12 shows the HSQC spectra that were recorded using the different pulse programs. In general, all three methods, HSQC, seHSQC and TROSY-HSQC, result in well resolved spectra for U24-7 dCS protein. Compared to the conventional HSQC, seHSQC selectively enhances the intensity of some resonances, and this is clearly evident when comparing the peaks in Figure 2.12B versus those in panel A. As seen in Figure 2.12C, the TROSY-HSQC experiment seems to be able to resolve more peaks than HSQC and seHSQC, making it the most promising method to study U24-7 dCS in the future. The side chain region in TROSY-HSQC spectra is much weaker than in the other two spectra. This is because the TROSY-HSQC pulse sequence is longer than seHSQC and HSQC, and the side chain amide magnetization relaxes faster than that of amides. In the overlays of the three spectra, it is easy to see TROSY-HSQC was the best of all three methods. Note that for reasons explained in Chapter 1, the chemical shifts from the TROSY-HSQC are slightly different from those of a regular HSQC. (Figure 2.12D) 63 Figure 2.11 seHSQC spectra of U24-7 dCS protein in citrate-phosphate buffer, pH 5.8 with 25 mM SDS, at different temperatures, recorded on an 850 MHz spectrometer. The spectra are recorded using the exact same pulse program and are shown using the same starting contour and number of contour levels. 64 Figure 2.12 The A) HSQC, B) seHSQC, C)TROSY-HSQC and D) overlays of above three spectra for U24-7 dCS protein at 45 °C. The spectra in A), B) and C) are shown at the same contour level, and are coloured red to blue depending on the contour level. The dashed blue box in D) marks the amide side chain region and the solid blue box marks the side chain resonances for the arginines. These peaks are folded in. 2.2.6 Phosphorylation of U24-7 with MAP kinase As mentioned in Chapter 1, a key hypothesis that was tested as part of this thesis work is whether U24 can mimic MBP. In particular, the amount of phosphorylated MBP is decreased in MS cases. Although U24-7 does not have a PxxP motif in its sequence, it does contain the minimal recognition sequence of MAP kinase, (T/S)P. (55) Based on cellular experiments, U24-7 is highly 65 likely to be phosphorylated in vivo. (36) To test if the construct described here could be phosphorylated in vitro, experiments were carried out on U24-7 using a similar protocol to the one previously used to show that U24-6A can be phosphorylated. Wild type U24-7 was prepared as described in section 2.5.5 and the phosphorylation experiments were carried out as described before. (54, 298) Two concentrations of ATP, 0.1 mM and 1 mM, were used for the phosphorylation of U24-7. In parallel, U24-6A and MBP (bovine, ≥90%, M1891, Sigma-Aldrich) with 0.1 mM ATP were prepared as a control. The phosphorylation mixture was sampled before the addition of kinase, 3 hours after the addition, and approximately 12 hours after the addition (labelled overnight in Table 2.3). These samples were precipitated using trichloroacetic acid (TCA) and cold acetone. The precipitated protein samples were dried on a heat block. Mass spectra of the samples were obtained using matrix-assisted laser desorption ionization time of fly mass spectrometry (MALDI-TOF MS). Resulting spectra are attached in appendix B5. Table 2.3 Mass over charge (m/z) of largest molecular weight peak determined in MALDI-TOF MS, upon exposure of U24-7 to MAP kinase. Mass over charge (m/z) Theoretical 0 hour 3 hours Overnight U24-7 (0.1 mM ATP) 9464.1 9463.0 9608.5 9468.2 9465.9 9480.4 U24-7 (1 mM ATP) 9469.5 9465.5 U24-6A (0.1 mM ATP) 10235.0 10264.6 10253.7 10334.0 10334.0 10411.4 MBP (0.1 mM ATP) 18468.3 18542.7 18538.7 18637.1 18698.2 Table 2.3 is a summary of the MALDI-TOF MS data. It is clear that additional peaks, observed at approximately 80 Da or 160 Da more than the m/z of the unmodified protein at time zero, could be detected in the cases of U24-6A and MBP. No similar pattern was observed for the phosphorylation of U24-7, even with 10 times of ATP concentration being used. These results suggest that the U24-7 construct used here could not be phosphorylated by MAP kinase. 66 2.3 Discussion This chapter describes how U24-7 membrane protein was expressed and purified, thereby enabling preliminary structural studies and in vitro phosphorylation of U24-7 to be tested. Membrane proteins are notoriously difficult to express and purify, therefore optimizations of protocols are not only important to make sufficient material for structural biology studies, but for providing interesting insights on the properties of U24-7 and, more generally, membrane proteins. The main aim in optimizing the purification of U24-7 is to obtain non-aggregated samples for NMR study, with as high a yield as possible. In the case of U24-6A, structure elucidation using NMR was not feasible because U24-6A resulted in poorly dispersed NMR spectra. Such poor resolution can be due to a number of reasons: i) lack of structure because of poor reconstitution of the membrane protein into an environment which mimics its true membrane environment; or ii) the formation of higher order aggregates, e.g. extended helical protein aggregates, which also display a characteristically narrow range in chemical shift. In order to address the reconstitution issue, a number of different sample conditions were explored. When choosing a detergent for purification, one must consider its impact on yield, compatibility with other reagents (e.g. proteases), and whether it can be used directly for structural studies or easily swapped out for other lipids (e.g. does it have a CMC in a useful range?). From section 2.2.2, the results suggest that Triton X-100 is the best for the affinity chromatography step, while all the other three detergents are not compatible with TEV protease cleavage. It is possible that Triton X-100 helps to unfold protein a little bit, thereby exposing the linkage to be cleaved by TEV. As for the other detergents tested, they were able to solubilize U24. However, the folding of Mbp-6xHis-U24-7 in a non-denaturing detergent, like CHAPS and NaDoc, might result in a globular fold, where the linkage between 6xHis tag and U24-7 is unexposed. In the case of SDS, it is possible that SDS denatures the TEV protease itself, rendering it ineffective. Besides the affinity chromatography step, which must be followed by TEV protease cleavage, the last chromatography step in our protocol is an important step in which the detergent can be replaced by something more 67 suitable for structural studies. Experiments were designed and carried out for this last step, using cation-exchange chromatography instead of anion-exchange chromatography. In theory, U24-7 would be retained by the resin when loaded in a buffer with a pH smaller than the pI of the protein. Yet, the cleaved U24-7 did not bind efficiently to the column, regardless of the pH of loading buffer used, making this option inpractical (data not shown). Therefore, Triton X-100 was kept in the protocol, but further optimization was required to obtain good yields of U24-7. The buffer system for anion-exchange chromatography was optimized. As presented above, the best conditions were found to be the use of Tris. During the tests performed, high proportions of U24-7 dimers were found in the fraction purified using either PBS or MES buffer. Also, the elution fractions from both buffer systems contained U24-7 dimers. This surprising result suggests that U24-7 is present as a dimer in the PBS and MES buffer before it is loaded onto the column. The only step that was different from the Tris buffer protocol is overnight dialysis for buffer exchange. Although 1 mM DTT was supplemented in both buffers, dimer still forms. It is highly possible that time is the issue here. Indeed, U24-7 dimer formation was observed in a sample composed of isolated U24-7 after it had been stored in the fridge for one day (data not shown). What is more, U24-7 dimer or larger oligomers may be more stable than the monomer. Both cysteine residues are in the transmembrane region and oligomerization could reduce the hydrophobic area exposed to the water, prior to membrane insertion. In a helical wheel representation, which is the illustration of the top view of a helical structure, the two cysteines are on the same face of the wheel, making oligomerization plausible (see Figure 2.13 for details). The additional bands that were detected in the elution fraction, in lanes B2 and C4 in Figure 2.5, could be these oligomers. They are different by nearly 10 kDa and were never found when purified using Tris buffer. In order to avoid dimerization or oligomerization of U24-7, which may lead to aggregated samples, the number and length of purification steps after TEV cleavage were kept to a minimum. 68 Figure 2.13 Transmembrane region of U24-7 shown in helical wheels and wenxiang diagram A) The diagram was generated by online tool Helical Wheel (http://kael.net/helical.htm, accessed on September 25th, 2015) The N-terminus is pointed inside and cysteine residues are coloured yellow, while histidine is coloured blue. B) The wenxiang diagram is generated using online tool Wenxiang2 (http://www.jci-bioinfo.cn/wenxiang2, accessed on September 25th, 2015) The N-terminus of the helix starts from the middle of the diagram. Polar or positive charged residues are coloured blue, non-polar residues are coloured red and the cysteines are coloured yellow. The step of acetone precipitation was optimized by the use of salt. It has been demonstrated that Na2HPO4 could efficiently disrupt the interaction between U24-7 and Triton X-100, hence inducing more U24-7 precipitate from Tris buffer. In addition, the resulting lyophilized U24-7 is fluffy, unlike the observed U24-6A before, suggesting that it would be easier to reconstitute. This was indeed the case for wild type U24-7, but they tend to aggregate when dissolved in buffer regardless of the amount of DTT supplemented in the solution. When samples were prepared at high concentrations, oligomer formation was again observed using SDS-PAGE (data not shown). This aggregation problem of U24-7 caused by cysteines, which is different from that in the case of U24-6A, can be resolved by using the U24-7 dCS construct. Although CD spectroscopy has been used routinely to check sample structure, the sample A B 69 conditions could not be directly applied in the NMR experiments. The reason is that the sample concentration required in CD spectroscopy is very low, and high protein concentration will increase the absorbance, which leads to noisy spectra, like those seen in Figure 2.9. Noise could also be found around 190 nm to 200 nm for a 50 μM sample. NMR spectra of low concentration samples could in theory be recorded, but it would take a very long time to obtain decent signal to noise ratios. The CD spectra of the wild type and U24-7 dCS showed that with proper reconstitution, the protein adopts mainly helical structure. In addition, the spectra obtained contained a similar proportion of helical content, indicating that the mutation does not perturb the structure significantly. Since U24-7 dCS remains monomeric, this construct should probably be used for a full structural study of U24-7. Preliminary NMR data, shown in Figure 2.12, shows promise that many of the resonances could be assigned, by using a combination of HSQC and higher dimensional experiments, e.g. 3D HSQC-NOESY or other triple resonance experiments. Due to time constraints, this work was not completed, but could be the subject of future work (see details in section 6.3). The temperature dependence of the structure was also investigated. It was found that the protein is stable in the temperature range of 15-55 °C. More resonances were observed in the NMR spectra at higher temperatures, which is typical of membrane proteins. Moreover, more resolved peaks were observed for U24-7 dCS than for U24-6A. (311) Finally, the in vitro phosphorylation experiment performed in U24-7 showed that this construct could not be as readily phosphorylated as U24-6A or MBP. One possibility is that the MAP kinase used in the experiment is highly selective for its recognition sequence, not present in U24-7, thereby explaining why U24-7 cannot be phosphorylated in this experiment. 2.4 Conclusions In this chapter, the expression and purification of U24-7 is described. This is the first ever report of the production of this protein, in yields high enough to make structural studies possible. CD and preliminary NMR data show that the protein adopts a primarily helical structure in a 70 membrane environment. The sample and temperature conditions tested here form a good starting point for future structure elucidation of this protein. The U24-7 protein is not denatured upon heating and the native structure could be maintained in detergent solutions, suggesting that it is stable in vivo. Unlike U24-6A, the post-translational modification of U24-7 is still unclear. Considering the function of U24-6A and U24-7 are highly similar in vivo, it is possible that U24-7 can function the same even without the phosphorylation. 2.5 Material and methods 2.5.1 Gene synthesis The gene sequence of U24-7 was obtained from NCBI (Gene ID: 3289481, protein sequence reference GenBank: AAC40737.1) and optimized according to the relative codon usage in E.coli. (312) Polymer chain reaction (PCR) experiments were carried out using High-Fidelity DNA polymerase Pfu (Fermentas) or Phusion (New England Biolabs) and a T100 Thermal Cycler (Bio-rad) to synthesize or amplify DNA fragments. Four oligonucleotides (ssDNA, in Table 2.1), with 10 to 20 overlapped base pairs (bps), were used to synthesize the gene sequence of U24-7. Oligo 1 and 2 were used to synthesize the first half of the target gene sequence, dsDNA A fragment, and Oligo 3 and 4 were used to get the dsDNA B fragment. These dsDNA fragments (A and B) were analysed and purified using agarose gel electrophoresis and QIAquick Gel Extraction Kit (Qiagen). These fragments were amplified and purified a second time. The overlapping and purification steps were repeated again using these two dsDNA fragments, A and B, to obtain the full-length dsDNA fragment containing restriction sites of BamHI and HindIII, a 6xHis tag, a protease cleavage site (thrombin protease (LVPR/GS) as in Figure B.7A or TEV protease (ENLYFQ/S) as in Figure 2.1) the gene sequence of U24-7 and a stop codon. 2.5.2 Cloning The constructed full-length DNA fragment was digested using two restricted endonuclease BamHI and HindIII at 37 °C for 1 hour. The vectors used in this study, pMal-c2x and pMal-p2x 71 (New England Biolabs), were also digested using these two endonucleases at 37 °C for 2 hours. The digested full-length dsDNA fragments (or insert) and vectors were purified using agarose gel electrophoresis and Qiaquick Gel Extraction Kit. This step could be omitted if both insert and vector fragments were purified before digestion or if the vectors had been treated with dephosphorylation enzymes, like Alkaline Phosphatase (New England Biolabs). The concentration of the extracted insert fragments and vectors were determined by absorption at 260 nm using a UV-Vis spectrophotometer (Cary). Ligation experiments were carried out using 1 μL (400 units) of T4 DNA ligase (New England Biolabs), 50 to 100 ng digested vector fragments and the insert fragments calculated using Equation (2.1). The ligation mixtures were incubated at room temperature for 1 hour, and then 5 μL was used to transform 60 μL competent XL1-Blue (Stratagene) E. coli. (313) The heat shocked and recovered E. coli were plated on Luria Bertani (Lennox L Broth Base, Life technologies) agar plates containing 50 μg/mL carbenicillin (LB-CBC agar plates) and grown at 37 °C overnight. mass(insert)=3×DNA length in bp (insert)DNA length in bp (vector)×mass(vector) (2.1) After the overnight incubation, 3 single colonies from the plate were used to inoculate three 5 mL of Luria Bertani broth containing 50 μg/mL carbenicillin (LB-CBC media) respectively, and these cultures were allowed to grow for another 12 to 16 hours at 37 °C, 225 rpm. Colony PCRs experiments were conducted to check if there were DNA fragments inserted in the vectors. Taq DNA polymerase or OneTaq 2X master mix (New England Biolabs) were used to amplify the sequence between two primers, pMal-seq and pMal-seqrv. These two primers are located up- and down-stream from BamHI and HindIII cleavage sites. 1 μL of the overnight culture was added in the PCR mixture as a template. An addition PCR experiment was setup as a negative control using 1 μL pMal-c2x vector as the template. The PCR cycle began with a 3 to 5 minutes initial heating step at 95 °C in order to lyse the E. coli. The resulting PCR products were analyzed using agarose gel electrophoresis. The fragments amplified from the plasmid containing the correct inserts should be around 300 bps, larger than that from the negative control. The plasmids correctly inserted with 72 U24-7 gene sequence were extracted from the corresponding 5 mL culture using QIAprep Spin Miniprep Kit (Qiagen). The plasmids were submitted to NAPS unit (UBC, Michael Smith Laboratory) for sequencing to confirm the insertion of U24-7 gene. 2.5.3 Small-scale protein expression Two constructed U24-7 plasmids, pMal-c2x-U24-7 and pMal-p2x-U24-7, were transformed into three E. coli strains (C41(DE3), Origami 2 and SHuffle express) and the transformed E. coli were plated on LB-CBC agar plates, incubated at 37 ºC overnight (all the grow ups for SHuffle express were done at 30 ºC and 250 rpm if shaking as per manufacturer’s instructions). Single colonies on each plate were used to inoculate 5 mL LB-CBC media and the cultures were grown at 37 ºC, 225 rpm, for 16 hours. 1 mL from each of the previous 5 mL cultures was used to inoculate 100 mL LB media with 100 µg/mL ampicillin (LB-amp media), and these cultures were shaking for about 2 to 3 hours. The optical density at 600 nm (OD600) was checked every 30 minutes using a UV-Vis spectrophotometer. When the OD600 reached 0.5, the cultures were divided into two parts in order to test for growth at 37 ºC and at 18 ºC. To the culture at 37 ºC, a final concentration of 300 µM isopropyl β-D-1-thiogalactopyranoside (IPTG) was directly added to start a three hours induction, whereas the one grown at 18 ºC required additional ice incubation for 10 minutes prior to the addition of IPTG and 20 hours of induction. 500 μL of all the induction cultures and the cultures before induction were collected and centrifuged. The cell pellets were stored in a freezer for further analysis by electrophoresis. 2.5.4 Large-scale protein expression The large-scale expression was done under the optimized conditions for U24-7, pMal-c2x-U24-7 in Origami at 37 ºC. This expression is preceded by a small-scale expression, where single colonies were picked and grown in a 5 mL LB-CBC media, and this culture was grown for 3 to 5 hours at 37 ºC, 225 rpm. 1 mL of this culture was used to inoculate a 100 mL fresh LB-amp media. This 100 mL culture was shaking at 37 ºC for 15 hours, 225 rpm. Every 25 mL of this culture was used to inoculate 1 L of fresh LB-amp media and in total four 1 L fresh LB-amp media 73 were inoculated. They were grown at 37 ºC, 225 rpm until the OD600 reached 0.5 to 0.8. A final concentration of 300 µM IPTG were used for a 4 hour induction period, and the cells were harvested by centrifugation at 5000 × g, 4 ºC for 15 minutes. The harvested cells were resuspended in 120 mL phosphate buffered saline (PBS buffer), transferred to four 50 mL centrifuge tubes (30 mL each) and centrifuged at 5000 × g again. These tubes, each containing cell pellet from 1 L culture, were stored in -80 ºC freezer until further use. 500 μL of all the induction cultures and the cultures before induction were collected for further analysis by electrophoresis in order to confirm the protein expression. M9 media (6.78 g/L Na2HPO4, 3 g/L KH2PO4, 1 g/L NH4Cl, 0.5 g/L NaCl, 1% dextrose, 1 mM MgSO4, 0.1 mM CaCl2, 0.1 µg/L thiamine) and C41(DE3) strains were used to express 15N labelled U24-7 used in NMR experiments. The same constructed plasmid, pMal-c2x-U24-7 was used. The starter culture was grown in 5 mL LB-CBC media as described above, and 1 mL of this culture was used to inoculate 80 mL M9 media with 100 µg/mL ampicillin (M9-amp media, 1 g/L 15NH4Cl was supplemented for 15N protein labelling) for overnight growth. Baffled flasks were used to increase the aeration during shaking. 35 mL of the overnight culture was used to inoculate 1 L fresh M9-amp media and in total two 1 L cultures were prepared and grown until the OD600 reached 0.5, or up to three hours. The cells were harvested and stored as described above. 2.5.5 Protein extraction and purification One tube of frozen cell pellet (from 1 L culture) was thawed on ice and resuspended using 25 mL solubilization buffer (20 mM KH2PO4, 0.5 M NaCl, 1% Triton X-100, 10 mM Imidazole, pH 7.4), which is supplemented with trace lysozyme (Bio Basic Inc), DNase I (Roche) and EDTA-free protease inhibitor cocktail for histagged proteins (Sigma-Aldrich). Cells were lysed by passing them three times through a French Press (Aminco) or by sonication (BRANSON, 5 second pulses) on ice for 3 to 5 minutes. The lysate was then transferred to a beaker and stirred for 2 hours with 4 M urea, at 4 ºC. Then, the lysate was centrifuged at 25,000 × g at 4 ºC for 1 hour, and the supernatant was filtered using a 0.45 µm syringe-drive filter unit (Millipore) and transferred to 74 another tube. All the steps described in this paragraph were carried out in a cold room at 4 ºC. An Econo-Pac column (Bio-rad) that was packed with 7 to 10 mL of Ni-NTA agarose (Qiagen), further referred to as histrap, was used for the purification of Mbp-6xHis-U24-7. A peristaltic pump (Gilson) was used during the purification to provide a flow speed of 5 mL/min for all steps. The histrap was firstly equilibrated with binding buffer (20 mM KH2PO4, 0.5 M NaCl, 1% Triton X-100, 10 mM Imidazole, pH 7.4). The supernatant described from the paragraph was loaded onto the column using the peristaltic pump and the flow-through fraction was loaded back onto the column another two times. After the loading steps, 150 mL of binding buffer (washing fraction 1 to 3), and 25 mL of washing buffer (washing fraction 4, 20 mM KH2PO4, 0.5 M NaCl, 1% Triton X-100, 25 mM Imidazole, pH 7.4) was applied to the column. Finally, Mbp-6xHis-U24-7 was eluted using 35 mL of elution buffer (20 mM KH2PO4, 0.5 M NaCl, 1% Triton X-100, 500 mM Imidazole, pH 7.4) and was dialyzed against 2 L Tris buffer (50 mM Tris, pH 7.8) for over 5 hours using 2,000 MWCO dialysis membrane (Spectrum Laboratories). The histrap column was washed using 100 mL deionized water before stored in 20% ethanol at 4 ºC. The dialyzed sample containing Mbp-6xHis-U24-7 was transferred to a 50 mL centrifuge tube and 0.2 mg purified recombinant histagged TEV protease was added. (314) A final concentration of 1 mM DTT was supplemented for the digestion. The solution was rotated end-over-end for 16 hours at 4 ºC. To carry out the detergent exchange experiments, additional steps were needed after the washing steps using binding and washing buffers described in the above paragraph. 100 mL of NaDoc washing buffer (20 mM KH2PO4, 0.5 M NaCl, 6 mM NaDoc, 25 mM Imidazole, pH 7.4) was used to wash the column before the elution step. The Mbp-6xHis-U24-7 protein was eluted from the column using 35 mL NaDoc elution buffer (20 mM KH2PO4, 0.5 M NaCl, 6 mM NaDoc, 500 mM Imidazole, pH 7.4). The eluted fraction was divided into three parts and dialysed against three different buffers: 50 mM Tris, pH 7.8 with (A) 12 mM SDS, (B) 6 mM NaDoc or (C) 10 mM CHAPS, 1 L each. 0.1 mg of TEV protease and 1 mM DTT were added into the dialyzed sample for 75 digestion. The steps described in this paragraph were carried out at 4 ºC. The histrap column was used again to purify the cleaved U24-7 protein. The column was firstly equilibrated with Tris buffer, and the digestion mixture was loaded onto the histrap column. The flow-through fraction, which contains mainly cleaved U24-7, was collected. 100 mL Tris buffer was then used to wash the column, and the first 5 mL of washing buffer was collected and combined with the U24-7 containing fraction. The Mbp-6xHis, undigested Mbp-6xHis-U24-7 and TEV protease, were eluted using elution buffer. The flow-through fraction was used in the next part of purification. An Econo-Pac column containing 5 mL Q Sepharose Fast Flow media (GE Healthcare) was used in the anion-exchange chromatography. This step was done in order to isolate U24-7 from Mbp and other contaminants. The column was equilibrated with 100 mL Tris buffer, and then the previous flow-through fraction, approximately 45 mL, was applied to the column. The purified U24-7 was again collected in the flow-through and combined with the first 5 mL of washing buffer. Mbp and other contaminants were eluted using Qs elution buffer (50 mM Tris, 1 M NaCl, 1% Triton, pH 7.8) , and the column was washed by 100 mL water before stored in 20% ethanol. 10 µl of sample during each phase were taken and were analyzed using SDS-PAGE to monitor the purification process. An example of the SDS-PAGE results is shown in Figure A.1. 2.5.6 SDS-PAGE analysis Fraction samples that were collected during the various purification steps (10 µl each) were mixed with 10 µl 2X tricine SDS sample buffer (Novex, Life technologies) supplemented with 5% β-mercaptoethanol (2X BME gel loading buffer), and were heated in a 95 ºC digital dry bath (Labnet) for 15 minutes. The samples were cooled and spun before loading onto polyacrylamide gels that were made according to standard protocols. The gels were run in Tris-Glycine buffer to resolve 10 to 200 kDa proteins, or Tris-Tricine buffers to resolve 2 to 100 kDa proteins. 10 µl Mark12 unstained standards (Life technologies) or 8 µl Unstained protein marker (New England Biorlabs, 2-212 kDa) were used as a reference in each gel. 76 Cell pellet samples were first thawed, then were resuspended in 50 to 100 µl 50% glycerol and heated in a 95 ºC digital dry bath for 10 minutes. The mixtures were cooled to room temperature before mixing with 2X BME gel loading buffer. Another 20 minutes incubation at 95 ºC was done to lyse E. coli and denature all cellular proteins. The samples were cooled and spun before gel loading and were run in Tris-Glycine buffer. The resulting gels were stained in colloidal Coomassie G-250 solution (recipe according to (315)) for 2 hours and then destained in water. The gels containing cleaved U24-7 were stained using silver stain. The band of U24-7 protein stained using Coomassie blue is fuzzy, therefore was not used (Figure A.1). Gel images were recorded on a GelDoc XR system (Bio-rad). 2.5.7 Acetone precipitation tests and large scale protein isolation The purified U24-7 from the anion-exchange column flow-through fraction was dialysed against water using 2,000 MWCO dialysis membrane for the precipitation tests. The dialyzed U24-7 was concentrated using Amicon ultra-15 centrifugal filter unit with a 3 kDa MWCO membrane (Millipore) from approximately 45 mL to 10 mL. This concentrated 10 mL U24-7 was again dialyzed against water, and was finally concentrated to 4 mL using a centrifugal filter unit. In the salt-inducted precipitation tests, 50 μL of different concentrated salt solutions were added in 200 μL of the concentrated U24-7 fraction, in a micro centrifuge tube (details in Table 2.4). 750 μL cold acetone was added into each tube and the sample were stored in -20 °C freezer for 2 hours. White precipitate would form after the incubation. The supernatant was removed by centrifugation at 13,000 rpm for 5 minutes, and this process was repeated three times. The final precipitate was dried in the fume hood and dissolved in 250 μL 10 mM SDS solution. The concentrations of the samples before and after the precipitation were determined by bicinchoninic acid (BCA) protein assay (Pierce). As for the samples used in SDS-PAGE analysis, the supernatants could be obtained after the centrifuge step. About 1 mL supernatant from each tube was removed and was kept in a fume hood for a day. The residual sample was then diluted with 10 mM SDS solution till 50 μL in total and then analyzed by gel electrophoresis. 77 Table 2.4 Salt stock solutions used in the salt induced acetone precipitation tests Salt Stock concentration (w/v) (NH4)2SO4 25% KH2PO4 25% NaCl 30% Na2HPO4 25% (suspension, not a solution) In the large-scale protein isolation, the concentrated U24-7, about 7 to 10 mL, was transferred to a 50 mL centrifuge tube. 1 g of Na2HPO4 was added to the U24-7 protein solution and this mixture was mixed using a vortex generator for 5 minutes. 30 mL of cold acetone was added into the sample and this mixture was stored in -20 °C freezer overnight. The mixture was then centrifuged at 3000 × g for 10 minutes at 4 °C, and the supernatant was removed. The process was repeated another two times. The final precipitate were dried in the fume hood and dissolved in 15 mL of deionized water via vortex mixing and dialyzed three times against 2 L deionized water for over 48 hours. The dialyzed solution, containing white cloudy precipitates, were lyophilized and part of the sample was submitted to Mass Spectrometry Centre (Chemistry, UBC) to confirm the protein MW using MALDI-TOF MS (Bruker Autoflex). The resulting mass spectra are in the appendix B3. 2.5.8 CD experiment The lyophilized purified recombinant U24-7 was dissolved in water or phosphate buffer first, then concentrated detergent containing water or phosphate buffer was added in to make the final sample for CD experiments. Blank samples that do not contain protein sample were prepared for baseline corrections. The spectra were recorded using a J-815 CD spectrometer (Jasco) flushed with nitrogen gas. A quartz cuvette with path length of 0.1 cm was loaded with 200 μL protein samples at room temperature. Samples were scanned at a rate of 100 nm/min with a step size of 0.1 nm, from 190 nm to 250 nm. Spectra were averaged over four scans and the blank sample was scanned with the same program for background subtraction. 78 2.5.9 Site-directed mutagenesis The DNA fragment containing U24-7 gene was cloned into the BamHI and HindIII sites of pUC19 vector using the method described. This plasmid will be used as a template for the mutation experiments. Two pairs of primers for C57S and C65S (Table 2.5), were designed according to the method described in the Quikchange Site Directed Mutagenesis Kit instructions (Stratagene). The PCR was setup with 4 ng template using Phusion High-Fidelity DNA polymerase. A negative control with mismatched primers was setup at the same time with 20 ng templates. DpnI and CutSmart buffer (New England Biolabs) were added into the PCR products and the mixtures were incubated at 37 °C for 2 hours. 5 μL of the digested mixture was then transformed to competent XL1-Blue E. coli, and the E. coli were plated on LB-CBC agar plates. Three colonies from the experiment plates were grown in 5 mL LB-CBC media overnight. The plasmids were extracted using the QIAprep Spin Miniprep Kit and sent to NAPS unit for sequencing to confirm the site-directed mutagenesis. The process was repeated again with the second pair of primers to obtain a double cysteine to serine mutation (dCS mutant). The DNA fragment containing dCS mutants were digested from pUC19 using BamHI and HindIII and cloned into pMal-c2x and pMal-p2x using the methods described in section 2.5.2. Table 2.5 Primers that was used in site-directed mutagenesis Name Primers from 5’ to 3’ C57S_For CTTTTATTATTCTGGCAAGTCTGATCATTTC C57S_Rev GAAATGATCAGACTTGCCAGAATAATAAAAG C65S_For CATTTCCGTTATCCTGAGTCTGATTCTGATTC C65S_Rev GAATCAGAATCAGACTCAGGATAACGGAAATG 2.5.10 1H-15N HSQC NMR spectroscopy A sample of 15N labelled U24-7 dCS was prepared as described above (2.5.5), and the lyophilized protein (from 500 mL of cell pellet) was gradually dissolved in 0.5-1 mL citrate-phosphate buffer with 25 mM SDS, pH 5.8. The citrate-phosphate buffer was made of 30 mM 79 disodium phosphate, 10 mM citric acid, and then the pH was adjusted to 5.8 by adding citric acid or sodium hydroxide. This citrate-phosphate buffer halved the salt amount of the original recipe. The sample was sonicated at room temperature for about 10 minutes before being filtered through a 0.22 μm filter. The resulting protein sample was transferred into a 5 mm NMR tube for experiments carried out on a Bruker Avance III 850 MHz NMR spectrometer. The spectra of three different types of HSQC were recorded after sample equilibration at the target temperature for 10 minutes. The pulse program used for the HSQC experiment was hsqcetgpspf3, while for seHSQC it was hsqcetfpf3gpsi, (280–283) with additional modifications of the pulse gradients by Dr. Mark Okon. TROSY experiments were done using trosyetf3gpsi, with pulse gradients edited by Dr. Mark Okon. (305–310) Typical delays between acquisitions, d1, used in these experiments were 1 s. The 90 ° hard pulse p1 was calibrated each time at different temperatures and they were normally around 9 μs. The data sets were processed using NMRpipe, (316) and visualized using Sparky. (317) 80 Chapter 3 Chapter 3 Probing the Interactions between U24-6A or U24-7 and the SH3 Domain of Fyn Tyrosine Kinase† 3.1 Introduction In this chapter, the potential interactions between U24-6A and Fyn tyrosine kinase SH3 domain (Fyn-SH3 domain) are investigated. As reviewed in the first chapter, U24-6A shares a seven amino acid segment with myelin basic protein at its N-terminus (Figure 3.1), and a peptide containing this segment was shown to activate T-cells that cross-reacted with MBP (1). It has also been demonstrated that, like MBP, U24-6A protein can be phosphorylated at the threonine position within this segment. (54) This evidence suggests that U24-6A may be a molecular mimic of MBP. MBP93-107 IVTPRTPPPSQGKGR U24-6A1-15 MDPPRTPPPSYSEVL U24 -71-15 M-THETPPPSYNDVML Figure 3.1 Sequence alignment of MBP, the first 15 residues of U24-6A and U24-7. The numbering of MBP is based on the human 18.5 kDa classic MBP. (106) The common sequence between U24-6A and MBP is aligned and in orange. A PxxP motif (marked with a solid line box) and a PY motif (marked with a dashed line box) are highlighted. U24-7 is aligned to U24-6A’s PY motif, but it does not contain the PxxP motif. Confirmed phosphorylation site in the PxxP motif of MBP and U24-6A are shown in bold and red. † This chapter is adapted with permission from Sang Y, Tait AR, Scott WRP, Creagh AL, Kumar P, Haynes CA, Straus SK. 2014. Probing the Interaction between U24 and the SH3 Domain of Fyn Tyrosine Kinase. Biochemistry. 53:6092–6102. Copyright 2014 American Chemical Society. (423) 81 The shared sequence between U24-6A and MBP may also be important because it represents a PxxP recognition motif site that binds to SH3 domain of Fyn tyrosine kinase, (127) an interaction that is important for oligodendrocyte cell morphology and protein trafficking. (129) As the cell morphology is important for the maturation of oligodendrocytes and the formation of myelin sheaths, it is possible that U24 may act to disrupt the development of oligodendrocytes by blocking MBP’s interaction with Fyn tyrosine kinase, again by way of mimicry. Fyn tyrosine kinase has been shown to have a pivotal function in translating complex communication signals into different cellular responses to direct central nervous system (CNS) myelination. (201) Fyn is considered thus far to mediate downstream signalling of three major pathways: i) the Rho-family GTPases which in turn regulate actin cytoskeleton dynamics essential to cell survival and morphological differentiation, ii) control the transport, stability and translation of mRNA of myelin proteins, especially MBP, and iii) recruitment of microtubule-associated protein Tau to stabilize the cytoskeleton. (201) The interaction with Tau is mediated by Fyn-SH3 domain, and the downstream signaling of this interaction could promote myelination. (204, 205) For more detailed information on the function and structure of Fyn, please see section 1.3.1 in Chapter 1. It is conceivable that a foreign protein, such as one from a virus or bacterium, can bind to the Fyn-SH3 domain and compete with essential interactions, such as the Fyn-Tau or Fyn-MBP association, thereby bringing about dysfunction in myelination. Hepatitis C-encoded NS5A protein has relatively high affinity for Fyn-SH3 (Kd = 629 nM) and could potentially compete with native host proteins for binding to Fyn-SH3. (185) As discussed above, U24-6A protein has the PRTPPPS sequence in common with MBP (Figure 3.1). Therefore, it is possible that U24 may bind to Fyn-SH3 to prevent it from interacting with its cytoskeleton-organizing partners such as Tau or MBP. The goal of this study was to probe whether a specific interaction between U24-6A and Fyn-SH3 exists and to be able to further hypothesize how the interaction may relate to a dysfunction in myelination. In other words, a specific interaction between U24-6A and Fyn-SH3 via the PxxP motif would support U24-6A in the role of molecular mimicry of MBP, (54, 297) namely that 82 U24-6A might compete in all important binding interactions of MBP. To enable detection of interactions between U24-6A and Fyn-SH3, GST pull-downs, NMR spectroscopy and ITC were utilized. Two 15-mer synthetic peptides, representing the N-terminus of U24-6A or U24-7, were used to represent the full-length membrane protein. U24-6A and NΔ9 U24-6A, a full-length construct of the protein with the first 9 residues removed, have been tested in pull -down experiments to further test whether the interaction is specific to the N-terminus of U24-6A. (298) 3.2 Results 3.2.1 Pull-down assays with GST-Fyn-SH3 and U24-7 protein Unlike U24-6A, U24-7 does not contain a PxxP motif, which is required for SH3 domain interaction. It has been demonstrated that the interaction between Fyn-SH3 and U24-6A is very weak, (298) so no interaction was expected in theory. However, peptides devoid of the PxxP motif have recently been shown to nevertheless bind SH3 domains. (318, 319) As for Fyn, a non proline motif, RKxxYxxY, could bind to its SH3 domain. (188) Therefore, it is necessary to investigate the possibility that an interaction still exists, despite the lack of PxxP motif. GST-Fyn-SH3, GST and U24-7 were purified prior the pull-down experiment. Figure 3.2 shows a silver stained SDS-PAGE gel demonstrating that full-length U24-7 could interact with GST-Fyn-SH3 bound to glutathione beads (lane 4) , even without the required PxxP motif. It is a much weaker band stained by silver stain compared to the pull-down result with U24-6A. (298) The result suggests that U24-7 can bind to Fyn-SH3 domain in an uncanonical fashion. Although there is no explicit PxxP motif in U24-7, the PY motif in the N-terminal sequence (Figure 3.1) and the hydrophobic nature of the N-terminal sequence could be the reason for uncanonical interactions. 83 Figure 3.2 SDS-PAGE result of GST pull-down experiment with GST-Fyn-SH3 and U24-7. U24-7 and GST-Fyn-SH3 used in the pull-down experiment were loaded as a reference in lanes 1 and 2, respectively. The gel was stained with silver stain. The marker and corresponding molecular weights were indicated in lane M. The input gel shows that GST-Fyn-SH3 and GST were loaded equally, due to the similar thickness of the bands. The presence of two bands where GST-Fyn-SH3 was loaded indicated that the sample consisted mainly of GST-Fyn-SH3, with a minor contribution from GST tag. Because GST tag and U24-7 do not interact (lane 3), the presence of this minor impurity does not affect the outcome of the pull-down experiment. U24 -7 appears to bind only very weakly to Fyn-SH3, as seen from the very faint band in lane 4. 3.2.2 ITC Experiments In order to ascertain the strength of the binding interaction between U24 and Fyn-SH3, ITC experiments were conducted. Since the pull-down results suggest that the strongest interaction is between U24-6A and Fyn-SH3, this pair was investigated in detail. In order to detect weak interactions, the proteins and ligand concentration must be high enough to generate enough heat. To simplify the model of titration, U24 membrane protein was replaced by the corresponding 15-mer peptide, noted as U24-6A and U24-7 peptides. 14.4 kDa 6 kDa 36.5 kDa 31 kDa GST-Fyn-SH3 GST tag U24-7 Input 1 2 3 4 M GST tag − − + − GST-Fyn-SH3 − + − + U24-7 + − + + 84 Figure 3.3 ITC data for U24-6A peptide (U24-6A1-15) binding to Fyn-SH3 at 25 °C. The other runs performed looked nearly identical. (Upper) Raw titration data for nineteen 2 µL injections of 27.6 mM U24-6A peptide into the ITC cell containing 1.48 mM Fyn-SH3 in 40 mM sodium phosphate pH 6.0. (Lower) Integrated heat data (points) and best fit (red line) to a “one set of sites” model. The mean thermodynamic parameters averaged from three independent experiments are reported in Table 3.1, with a characteristic ITC experiment (with heat of dilution run subtracted) shown in Figure 3.3. Control experiments of buffer into Fyn-SH3 solution showed no evidence of protein aggregation or dissociation phenomena. Less binding was observed by ITC for U24-7 peptide titrations into Fyn-SH3 domain and the result was estimated to be around Kd ≈ 8-12 mM. (see the raw data in Figure E.2). 85 Table 3.1 Parameters for U24-6A binding to Fyn-SH3 at 25°C Kd (mM) ΔG (kJ/mol) ΔH (kJ/mol) ΔS (J/mol K) ITCa 5.1 ± 0.3c -13.1 ± 0.2 -12.5 ± 0.6 2 ± 2 NMRb 5 ± 1 - a The data was obtained from an average of three runs. The errors represent the standard deviation. b The binding constant was determined by averaging the Kd value determined for 4 of the perturbed residues: R96, T97, W119s and I33. c The stoichiometry, n, was fixed at 1.0 while fitting the ITC data. 3.2.3 Titration of U24 peptides by NMR As the ITC demonstrates that the interaction between U24-6A peptide and Fyn-SH3 is weak, 1H-15N HSQC NMR spectroscopy was performed, using 15N-labelled Fyn-SH3 and titrating in increasing amounts of U24-6A and U24-7 peptides representing the 15 residues at the N-terminus of both versions of U24 protein, respectively. The assignment of Fyn-SH3 is based on that reported by Mal et al., (320) and verified using a 3D HSQC-NOESY experiment. Several amide chemical shifts of Fyn-SH3 changed in response to the addition of U24-6A peptide (Figure 3.4a), and continued to change until a sufficiently high peptide to protein ratio was reached, indicative of binding saturation. The most notable shift changes were for residues R96 and T97, which are known to be part of the RT loop, (321) a flexible region of SH3 domains that is critical for peptide selectivity and binding. These residues were also perturbed in the case of titration with U24-7 peptide, but to a significantly lesser extent, further suggesting that U24-7 interacts more weakly and non-specifically with Fyn-SH3 than the U24-6A. Figure 3.5, containing the chemical shift perturbation plot and insets of Arg96 and Thr97, demonstrate the detailed chemical shift perturbations by U24-6A and U24-7 on Fyn-SH3. It is clear that the perturbations in the case of U24-6A is specific and limited to amino acids in a binding site, compared to that of U24-7. Little line broadening was observed during the titration, indicating that exchange between the bound and unbound states is fast on the chemical shift time scale. Under these conditions, the shift 86 in peak position relative to the spectrum of Fyn-SH3 alone is directly proportional to the fraction of the bound state, assuming two-state binding. Fitting the binding data for U24-6A peptide using the method by Zarrine-Afsar et al., (322) the dissociation constant was found to be 5 ± 1 mM (Table 3.1), using a number of perturbed amide resonances. This is in agreement with the values obtained from ITC. The fraction of peptide bound, a parameter obtained during the Kd determination, (322) is plotted for a few representatives residues in Figure 3.4c for U24-6A peptide. Only the changes in proton chemical shifts are presented here, and the “H” at the end of each residue in the legend refers to the 1H chemical shift. Similar calculations were performed based on the 15N chemical shift data, but are not shown. No dissociation constants were determined for U24-7. 87 Figure 3.4 Overlays of 15N-1H HSQC spectra of Fyn-SH3 with: a) U24-6A peptide, b) U24-7 peptide added in different ratios. c) The fit of the chemical shifts of U24-6A titration. a) Fyn-SH3 to U24-6A peptide ratio= 1:0 (black); 1:1.08 (purple); 1:2.55 (red); 1:5.10 (green); 1:9.12 (blue) and 1:12.68 (cyan); b) Fyn-SH3 to U24-7 peptide ratio= 1:0 (black); 1:0.98 (purple); 1:3.10 (red); 1:5.00 (green); and 1:8.30 (blue). Some of the key residues affected are indicated. Side-chains are indicated by “s”; c) The fit of the chemical shift data from a) yielded the fraction of U24-6A bound as a function of peptide concentration for select residues: R96 (black, solid), T97 (black, dashed), W119s (grey, solid), and I133 (grey, dotted), as shown in c). 88 Figure 3.5 Chemical shift perturbations of Fyn-SH3 for a) U24-6A peptide and b) U24-7 peptide NMR titration. Insets are shown for residues R96 and T97 in both cases as the normalized chemical shift difference for all residues. The normalization factor in a) was 0.786 for the nitrogen shifts (black bars) and 0.199 for the proton shifts (grey bars). The normalization factor in b) was 0.295 for the nitrogen shifts (black bars) and 0.065 for the proton shifts (grey bars). The absolute values of the normalized shifts were plotted. 89 3.2.4 A shorter version of U24-6A peptide may increase the affinity Compared to the optimal binding ligand of SH3 domain, PxxPxR or RxxPxxP, U24-6A does not contain a positive flanking charged residue on one or both sides. On the contrary, U24-6A has an aspartic acid preceding the core PxxP motif. In this section, a truncated version of U24-6A peptide, with its first two residues deleted (U24-6AΔMD=PPRTPPPSYSEVL), was synthesized and investigated. In theory, it is expected to bind better than the wild type U24-6A peptide. ITC and NMR experiments were carried out to confirm this. The ITC and NMR data were fitted using the same approach as described in last two sections and the resulting values are reported in Table 3.2. The ITC experiment was only performed once to see the rough trend. An integrated heat plot containing the titration curves of both U24-6AΔMD and U24-6A titration is shown in appendix Figure E.1 to demonstrate the difference in titration curves and affinities. There is no line broadening observed during NMR titration and the chemical shifts of I133, T97, R96, W119, S135 and W119S were used to fit Kd. Unlike the interaction with U24-6A peptide, most proton chemical shifts are not significantly perturbed. Therefore, the 15N chemical shifts were included in the fit of Kd. The fraction of peptide bound is plotted against the concentration for U24-6AΔMD peptide. (322) Data points and fit lines are shown in panel C of Figure 3.6. 90 Figure 3.6 ITC and NMR titration results of U24-6AΔMD (U24-6A3-15) and Fyn-SH3 domain. A) Overlays of 15N-1H HSQC spectra of Fyn-SH3 with U24-6AΔMD peptide added in Fyn-SH3 to U24-6A peptide ratio= 1:0 (black); 1:0.9 (orange); 1:2.0 (green); 1:3.1 (blue); 1:4.8 (cyan); 1:7.3 (purple) and 1:11 (red); B) ITC data for U24-6AΔMD peptide binding to Fyn-SH3 at 25 °C, containing (Upper) raw titration data and (Lower) integrated heat data (points) and best fit (red) to a “one set of sites” model. C) Fit of the proton chemical shifts data for S135 (black, solid) and T97 (black, dotted); 15N chemical data for W119S (grey, solid) and T97 (grey, dashed) from U24-6AΔMD/Fyn-SH3 NMR titration. 91 Table 3.2 Parameters for U24-6AΔMD binding to Fyn-SH3 at 25°C Kd (mM) ΔG (kJ/mol) ΔH (kJ/mol) ΔS (J/mol K) ITCa 1.9 ± 0.1c -16 -7 ± 0.1 27 NMRb 1.1 ± 0.1 - a The data was obtained from one run. The errors on Kd and ΔH represent the uncertainty of the fitting. b The binding constant was determined by averaging the Kd value determined for 3 of the perturbed residues: A95H, T97H, S135H, T97N, S114N and W119SN. H and N indicate either 1H or 15N chemical shifts are used in data fitting. c The stoichiometry, n, was fixed at 1.0 while fitting the ITC data. 3.3 Discussion Fyn tyrosine kinase is localized mainly in the oligodendrocyte plasma membrane and plays an important role in a range of signalling pathways via integrins and Ras activation during CNS development. (191, 323) Fyn is particularly important in oligodendrocyte differentiation and myelination. (201) It has been demonstrated that the SH3 domain of Fyn interacts with myelin basic protein (MBP) via the PxxP binding motif in MBP. (129) As a result of this interaction, an increase in the length of membrane processes and branching complexity in N19-OLG cultures was observed. (OLG is referred to the cell line of oligodendrocytes) It was therefore concluded that interactions of classic MBP isoforms with SH3-domain-containing proteins may play a physiological role in oligodendrocytes and myelin formation, compaction, and overall stability. (129) Furthermore, disruption of normal interactions between MBP and its intended SH3-domain partners in oligodendrocytes (i.e.: Fyn) may lead to myelin dysfunction, leading to diseases like MS. The study presented herein aimed to determine whether two similar but not identical forms of a viral protein, U24-6A and U24-7, can potentially mimic MBP binding to the SH3 domain of Fyn. U24-6A shares an identical segment and a PxxP motif with MBP, whereas U24-7 does not. However, both U24 proteins share a common PY motif. The data presented provides evidence that it is mainly the PxxP region in U24-6A which interacts with the SH3 domain of Fyn tyrosine kinase. GST- 92 tagged Fyn-SH3 can successfully bind to U24-6A via its PxxP region, and titrating a peptide of U24-6A containing this same region into Fyn-SH3 confirms a weak but specific interaction between the two partners (Kd = 5 mM, Table 3.1). When bound to a SH3 domain, U24-6A fulfills the structural requirements for binding by forming a PPII helix, (298) as also found for MBP. (127) On the other hand, U24-7 shows an even weaker interaction with Fyn-SH3. Evidence for this comes from the very weak band seen in the GST pull-down experiment (Figure 3.2B) and in the estimated Kd from NMR and ITC of ca. 8-12 mM. Overall, the data suggest that interaction with Fyn-SH3 occurs through a weak interaction with the PxxP motif of U24-6A. This segment may be of importance and may explain why the binding interaction between phosphatidylinositol 3-kinase (PI3-kinase) peptide and Fyn-SH3 has a range of Kd between 16 μM to 3 mM. (187, 324) These peptides all have a flanking arginine around the PxxP motifs, and extra prolines around the core PxxP motif (Figure 3.7). Class I ligand RxxPxxP Class II ligand PxxPxR Affinities (Kd) U24-6A MDPPRTPPPSYSEVL 5 mM U24-6AΔMD PPRTPPPSYSEVL 2 mM U24-6B MDRPRTPPPSYSEVL - MBP IVTPRTPPPSQGK - NS5A PPRSPPIPPPRKKR 629 nM (185) PI3-kinase Peptide 1 KKISPPTPKPRPPR 3 mM (187) Peptide 2 PPRPLPVAPGSSKT 50 μM Peptide 3 PPRPTPVAPGSSKT 300 μM Figure 3.7 Sequence alignment of Class I and II PxxP motif ligands: U24-6A, U24-6B, MBP, NS5A polyproline region, and three peptides derived from the PI3-kinase polyproline region. The sequences were aligned through their PxxP motifs, highlighted in orange. The positively charged residues are marked in blue and negative ones in red. Proline residues are shown in bold. 93 The binding constants shown in Figure 3.7 suggest that flanking charges play an important role in binding. Indeed, the U24-6AΔMD resulted in stronger binding between the 13-residue peptide and Fyn-SH3. The fitted ITC and NMR results suggest an enhancement of roughly a factor of three (Figure 3.6 and Figure E.1). Without the negatively charged residue upstream from the PxxP motif in U24-6A, binding is stronger. A number of studies on the importance of residues surrounding the PxxP motif for binding to Src homology 3 domains have been reported in the literature. (185, 187, 325) There are two classes of PxxP ligands that are optimal for SH3 domains, RxxPxxP and PxxPxR (Figure 3.7). Flanking arginines are found to interact with the D99 in the RT loop of Src SH3. (67, 68) Similarly, acidic amino acids were found in the same positions in the SH3 domain of Hck and Sem-5 protein. (326, 327) In the case of Fyn, D100 is at the conserved position and D99, D100, E116 and D118 are potentially using their charged side-chains to stabilize the additional positive charge (Figure 3.8). Arginine fits into the negatively charge cleft upstream from the xP grooves. The optimal ligand is therefore one in which the charged residues are at the exact position up or down stream from PxxP motif, as seen in the NS5A polyproline region (Figure 3.8). This peptide binds to Fyn-SH3 with high affinity (Kd = 629 nM). (185, 328) The precision of the up or down-stream required arginines rule out the possibility of U24-6B being a good ligand of SH3 domain. However, there is an exception in the case of peptide 1 from PI3-kinase. It has positively charged residues up and downstream from the PxxP motif, but it has a very low affinity with Fyn-SH3 domain. The reason could be those extra positive charges further up and down stream from the PxxP motif. Although a Kd value was never reported for full-length MBP protein and Fyn-SH3, a crude estimate based on the data reported in (129) can be made (Kd ≈ 0.1-1 mM). More recent data on the MBP/Fyn-SH3 interaction reported the affinity to be 4-8 μM, in which a long peptide xα2 (containing S38–S107 of MBP, murine numbering, -3 compared to human numbering) is used instead of MBP protein. (131) No data was obtained for just the polyproline segment of MBP as shown in Figure 3.7. Since MBP shares the amino acid stretch starting from a proline in U24-6AΔMD, negative charges should not affect the interactions. However, the presence of an arginine and threonine inside of the PxxP motif may have an impact. There is no known example of high affinity 94 binding PxxP motif containing an arginine at one of the x positions. The PI3-kinase peptide 1 contains a threonine at the same position as U24-6A and MBP, and has poor binding affinities. A Thr to Leu mutation at a position upstream of the PxxP motif, changing PI3-kinase peptide 3 to peptide 2, appears to increase the affinity six fold. This suggests that there are more factors, other than just have the flanking arginines present, which are required to achieve high affinity binding of Fyn-SH3 domain. Figure 3.8 Structural model of Fyn-SH3 and PI3-kinase peptide complex (PDB:1A0N). The structural model in the above figure is a Fyn-SH3 domain in complex with PI3-kinase peptide 2, PPRPLPVAPGSSKT. The access code in PDB is 1A0N and model 4 is shown in cartoon and surface display mode using Pymol. The Fyn-SH3 domain is shown in grey ribbons and the surface of the Fyn-SH3 domain is coloured using the APBS tools in Pymol in ligand free mode. (189) Red indicates a negative charge, while blue indicates a positive charge. D99, D100, E116 and D118 in this domain are shown in red sticks. The ligand is shown in yellow, and its upstream arginine is shown in blue sticks. The PxxP motif is shown in yellow sticks. 95 3.4 Conclusions Previous work has shown that U24-6A can be phosphorylated like MBP and hence it was suggested that U24-6A may be a mimic for MBP and be potentially involved in MS. The extent of the phosphorylation of U24-6A, however, is much weaker than for MBP. (54) Here we find that a binding interaction between U24-6A and Fyn-SH3 exists, but it is again most likely much weaker than the interaction between MBP and Fyn-SH3. The Kd between MBP and Fyn-SH3 was estimated to be 0.1-1 mM. (129) A segment upstream from the PxxP motif in MBP is involved in the interaction, which will have a low micromolar affinity with Fyn-SH3 domain. (131) This suggests that mimicry is possible, but that perhaps explicit competition between U24-6A and MBP in cells is less likely, unless in vivo conditions tip the balance in favour of U24-6A. Some studies have suggested that weak protein-protein interactions may nonetheless be a determinant of many biological processes. (325) It is also possible that U24 brings about myelin dysfunction by a completely different mechanism which could occur in conjunction to the mimicry of MBP or be entirely separate. 3.5 Materials and methods 3.5.1 GST-SH3 domain pull-down assays with recombinant U24-7 protein The GST protein and GST-Fyn-SH3 used in this study was prepared as described before, (298) and pull-down experiments was carried out similarly. U24-7 was purified and dialysed against 2 L pull down buffer using the protocol in section 2.5.5, and its concentration was determined by BCA protein assay. 1 mM DTT was supplemented in the U24-7 sample and 750 μL of the U24-7 stock was used for the interaction with GST-Fyn-SH3. After washing with Glasgow lysis buffer (GLB), the resin was heated at 95 ºC for 5 minutes with 20 μL 2X BME gel loading buffer. 10 μL was removed for electrophoresis using Tris-Tricine buffer system. 10 μL of U24-7, GST protein and GST-Fyn-SH3 were mixed with 2X BME gel loading buffer and loaded on the gel for reference. 96 3.5.2 Peptide synthesis and purification Two 15-mer peptides representing the N-terminus of U24-6A and U24-7 (see sequence in Figure 3.1) were synthesized using N-9-fluorenylmethyloxycarbonyl (N-Fmoc) protected α-amino acids on an automated peptide synthesizer (CS Bio). Preloaded Fmoc-Leu-Wang resin (Advanced ChemTech) and O-(benzotriazol-1-yl)-N,N,N',N'-tetramethyluronium hexafluorophosphate (HBTU, Advanced ChemTech) were used in the coupling protocols. Double coupling was required for the last nine residues of U24-6A (MDPPRTPPP) and last eight residues of U24-7 (MTHETPPP). The crude peptide was deprotected and cleaved from the resin in 90% trifluoroacetic acid (TFA) with trace water, 1,2-ethanedithiol (EDT) and triethylsilane (TES) as scavengers. The experiments were carried out based on a protocol described previously. (329) The excess solvents were removed under reduced pressure, and the peptide was precipitated through diluting into cold diethyl ether. The crude peptide powder was dried in vacuo, redissolved in de-ionized water and lyophilized to remove all residual solvents. 20 μL of the crude peptide solution was lyophilized separately and submitted to Mass Spectrometry Centre to confirm the mass of intact peptide using MALDI-TOF MS. Both peptides were purified using a C18 reversed-phase high-performance liquid chromatography (HPLC) column (Phenomenex, Jupiter, 10 µm 300 Å, 250 x 21.2 mm) on a Waters 600 system, monitored by UV absorption at 229 nm and 279 nm using a photodiode array detector. The peptides were eluted with a gradient from 0 to 30% of acetonitrile for over 40 minutes at a flow rate of 10 mL/min. The purified peptides were lyophilized and purified using HPLC for a second time to increase the purity. The peptides, which were purified twice, were submitted to Mass Spectrometry Centre for molecular weight confirmation using MALDI-TOF MS and sequencing by liquid chromatography–tandem mass spectrometry (LC-MS/MS, Bruker HCTultra ion trap spectrometer). Sample HPLC traces, MS spectra and LC-MS/MS spectra are shown in appendices A3, B1 and B2, respectively. The shorter version of U24-6A peptide (U24-6AΔMD peptide), without the first two N-terminal amino acids, was synthesized and purified using the same method for U24-6A peptide described above. No additional mass spectrometry or HPLC traces are shown in the appendix for U24-6AΔMD peptide. 97 3.5.3 ITC experiments Fyn-SH3 was expressed and purified as previously described. (298) The cleaved SH3 domain was dialysed to a 40 mM sodium phosphate buffer (pH 6.0) overnight and then concentrated using an Amicon ultra-4 filter device with a MWCO of 3 kDa (Millipore). The final protein concentration was determined by the BCA protein assay to be around 1.10 to 1.48 mM and the pH was 6.09. Two 20x peptide solutions were made. After dissolving purified and lyophilized U24-6A peptides and U24-7 peptides in the dialyzed protein buffer, the pH was checked with a micro pH meter (Thermo Scientific) and was adjusted to match that of the protein solution. The protein sample was filtered using a 0.22 μm filter (Millipore). The peptide solution was placed in an eppendorf tube (1.5 mL) and centrifuged on a tabletop microcentrifuge (Fisher Scientific) at 13,000 rpm for 20 minutes at 4 °C. The supernatant was transferred to a new tube. Both protein and peptide were degassed at 25 °C for 4 to 8 minutes before loading into the sample cell and syringe, respectively. Isothermal titration calorimetry (ITC) experiments were performed using an iTC200 MicroCalorimeter (GE Healthcare) at 25 °C. The titration protocol was comprised of a preliminary injection of 0.4 μL of the peptide solution, followed by 19 consecutive 2 μL injections into the sample cell (200 μL) containing the Fyn-SH3 protein. The time between each injection was 150 seconds. Control titrations of peptide sample into protein-free buffer, as well as buffer into Fyn-SH3 sample (to verify lack of protein oligomerization effects), were also conducted. The heats of dilution for the peptide were subtracted from the original heats prior to data fitting to a bimolecular interaction model to obtain Ka (association constant, in M-1) and ∆H (enthalpy of binding, in kcal/mol) at a binding stoichiometry n (number of binding sites per Fyn-SH3) of 1.0. The ITC experiment was repeated three times with mean values and standard deviations reported. 3.5.4 1H-15N HSQC NMR titration experiment Solutions of 15N labelled Fyn-SH3 protein were prepared by successive rounds of concentration and dilution with NMR buffer (40 mM sodium phosphate, 10% D2O, 0.5 mM benzamidine, 0.1% sodium azide, pH 6.0) using an Amicon Ultra-15 centrifugal filter device with a 98 MWCO of 3 kDa. Protein solutions were centrifuged briefly and diluted to a final concentration about 0.4 to 0.6 mM with NMR buffer. The final protein concentration was determined by the UV absorption at 280 nm in 6 M Guanidine HCl, and calculated with an extinction coefficient of ε280 = 16,960 cm-1M-1, which was computed using the ProtParam tool. (303) The concentration was also verified using the BCA protein assay. The concentrated sample was placed in a 5 mm NMR tube and the 1H-15N HSQC spectra were recorded at 25 °C using a Bruker Avance III 850 MHz NMR spectrometer (Milton, Ontario, Canada), equipped with a TCI probe. In the case of U24-7 peptide titration, the spectra were recorded on a Bruker Avance 500 MHz NMR spectrometer equipped with a TXI probe at 30 °C. Purified U24 peptide stocks were prepared in the NMR buffer with a concentration of 30 to 40 mM, and then were added into the Fyn-SH3 sample step by step. HSQC spectra were recorded after each subsequent addition. A total of 14 to 15 different protein:peptide ratios were examined between 1:0 to 1:12.6 for U24-6A and 1:13.5 for U24-7. The pH of the solution was measured to be 6.3 at the start of the titration and found to be identical after the last addition of peptide. The amide chemical shifts of Fyn-SH3 were assigned using previously published assignments by Mal et al, and verified using a 3D HSQC-NOESY and 3D HSQC-TOCSY dataset obtained on a Bruker Avance III 850 MHz spectrometer. The assignments are listed in appendix C1. The dissociation constant (Kd) was calculated according to the method described by Zarrine-Afsar et al. (322) The chemical shift at a concentration C of peptide is defined as: Ω(C)=Ω0+fB(Ωf - Ω0)+mC (3.1) with Ω0=Ω(0) and Ωf =Ω(∞), and m is a baseline correction factor. In this case, the fraction of Fyn-SH3 bound to the peptide is related to Kd : fB=CC+Kd (3.2) The parameters m and Kd are allowed to vary to optimize the fit between the experimental and calculated chemical shift values. 99 Chapter 4 Chapter 4 Investigating the Interactions between U24 and WW Domains of Nedd4 E3 Ubiquitin Ligase† 4.1 Introduction The interactions between U24 (U24-6A or U24-7) and WW domains of Nedd4 E3 ubiquitin ligase are studied in this chapter. As described in Chapter 1, the PY motif in U24 is crucial for its function of blocking endosomal recycling. (36, 45) In T-cells, the surface level of TCR/CD3 and TfR could be down-regulated by U24. However, it was shown that this effect is not achieved through a direct interaction. No colocalization between U24 and TCR/CD3 was detected. (36) Consequently, a protein which plays a role in endosomal recycling, such as Nedd4, could mediate the down-regulation of TCR/CD3 and TfR and thus be a specific binding partner of U24. In other words, a direct interaction between WW domains and U24 or other more indirect pathways would result in the blocking of endosomal recycling. In this chapter, we will examine whether a direct interaction exists between Nedd4 WW domains and U24 from HHV-6A and -7. E3 ubiquitin ligase Nedd4 and Nedd4-like (Nedd4L) proteins normally contain two to four WW domains depending on the isoforms. As an E3 ubiquitin ligase, not only does Nedd4 induce the endocytosis of membrane proteins, but it is also involved in the subsequent steps. (330) For instance, hepatocyte growth factor-regulated tyrosine kinase substrate (Hrs), a component of the ESCRT complex, was shown to be responsible for endosomal sorting of ENaC, after being † A version of this chapter will be published as Sang Y, Zhang R, Scott WRP, Creagh AL, Haynes CA, Straus SK. 2016. Investigating the Interactions between U24 and Nedd4-WW Domains. 100 ubiquitinated by Nedd4. This requires an interaction between ENaC subunits and Hrs, mediated by Nedd4. (331) Similarly, the suppressor of deltex (Su(dx)), a Nedd4 family E3 ubiquitin ligase found in Drosophila melanogaster, was reported to target the Notch receptor to the late endosome by sorting it to a Hrs-enriched subdomain. (332) Nedd4 functions widely in the cellular system and its role in early endosomal protein/cargo sorting makes it a potential protein that U24 might interfere with. Therefore, Nedd4 is an appealing target to study U24’s functions. Many studies have shown that a virus is capable of exploiting cellular machinery. Nedd4 could be recruited by a virus for budding, a process needed to release the replicated virus particles after an infection. This recruitment is normally achieved via an interaction between a viral protein and a WW domain of Nedd4 (Nedd4-WW domains), rarely via the C2 domain. (333–337) Depending on the types of viruses, Nedd4 could be recruited directly by a viral protein or through a cellular adaptor protein, such as Alix and arrestin domain-containing protein 1 (ARRDC1). (333, 335) Apart from virus budding, ITCH and Nedd4L have been shown to be recruited by Epstein-Barr virus (EBV) to promote ubiquitination of Lyn and Syk, two Src family non-receptor tyrosine kinases (SFKs). This is achieved through the PY motif in latent membrane protein 2A (LMP2A) from EBV. (338) It is conceivable that Nedd4 might be recruited by U24, leading to endosomal sorting or ubiquitination events that block the endosomal recycling of TCR/CD3. TCR/CD3 down-regulation could also indirectly involve Nedd4. Some of Nedd4 or Nedd4L’s targets, such as serum- and glucocorticoid-induced kinase (SGK) and Casitas B-lineage Lymphoma-b (Cbl-b), function as regulators in the cells. (237, 339, 340) SGK facilitates the endosomal recycling of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor and KCNQ1/KCNE1 potassium channels. (341, 342) At the same time, SGK inactivates Nedd4L via phosphorylation, which implies that these two proteins reciprocally regulate one another. (237, 343) On the other hand, Cbl-b is a RING family E3 ubiquitin ligase that mediates protein endocytosis and is implicated in cancer and autoimmune diseases. (344, 345) Mice with a cbl-b gene knock-out are highly susceptible to EAE, (346) and there are also clues to suggest a link between Cbl-b variants and MS. 101 (347) This suggests that there could be an indirect connection between Nedd4 and MS, via endocytosis. The potential interaction between U24 and Nedd4 could perturb the balance between Nedd4 and related proteins, which could result in changes in the endocytosis of Nogo-A, MAG or sodium channels, leading to MS (see details in section 1.2.3). Overall, this evidence makes Nedd4 a very attractive binding partner to study. Position Numbering PPx Y -6 -5 -4 -3 -2 -10 1 2 34 567 89 10 11 12 U24-6A MDPP RTP PPS Y SE V- L U24-7 M-TH ETP PPS Y ND VML rβENaC LPIP GTP PPN Y DS L hαENaC TAP PPA Y AT LG Comm TG LPS Y DE ALH HECT-PY RL D LPP Y ET FED LR E K Figure 4.1 Sequence alignment of the first 15 residues of U24-6A, U24-7 and the PY motifs found in ENaC, Comm protein and the HECT domain of hNedd4L-WW3*. Two ENaC sequences, from rat and human respectively, are aligned. ENaC contains three subunits and α β or γ indicates the subunit number. (78, 348) The PY motif in HECT-PY is the atypical sequence LPPY, but it nevertheless binds to hNedd4L-WW3* domain. (247) The PY motifs are shown in bold. The positively and negatively charged residues are marked in blue and red, respectively. The alignment of additional ligands is in Figure F.2. Most of the WW domains in Nedd4 belong to group 1 (section 126.96.36.199), which prefers to bind a PY motif. (81) U24 was found to bind most of the WW domains screened in work done by Sullivan and Coscoy (172), indicating that it may contain an optimal sequence for group 1 WW domains. The sequences that contain PY motifs which bind to group 1 WW domains are aligned in Figure 4.1. There is little difference between the core residues of the motif, but the flanking residues are unique in each ligand. The ligand preference of Drosophila melanogaster Nedd4 WW3* domain (dNedd4-WW3*) has been investigated using alanine scanning and NMR spectroscopy. The threonine at the -1 position was found to be crucial for high affinity binding, (214) even though an earlier study suggested that only the core PY motif residues, position 0 to 4, as the smallest high 102 affinity sequence. (173, 174) In the case of U24, U24-6A and U24-7 both contain threonine at the -1 position, but oppositely charged residues at the -2 position. This may cause some differences in the binding affinities of U24-6A and U24-7 to Nedd4-WW domains. Figure 4.2 Schematic architectures of Nedd4 and Nedd4L protein. The WW domains used in this chapter are highlighted in pink in the figure above. The sizes and distances between WW domains are scaled drawing of WW domains and the linkers, while C2 and HECT domain are drawn in half and a third of their size for clarity. The dashed line and C2 domain in hNedd4L indicates some hNedd4L isoforms do not have a C2 domain. The access codes for these three proteins are, AAB48949.1 (rNedd4), AAI52453.1 (hNedd4) and NP_001138442.1 or Q96PU5-5 (hNedd4L). The sequence alignment of these WW domains and additional WW domains are listed in Figure F.1. There are several isoforms of Nedd4, with slight variations in the WW domain sequences. Based on the primary screening done by Sullivan and Coscoy, (172) Nedd4-WW1 was found to be the only domain that does not bind U24. In order to compare with reported binding affinities and structural signatures, four WW domains were chosen for further study here: Rattus norvegicus Nedd4 WW2 and WW3/4 (rNedd4-WW2 and rNedd4-WW3/4), and two WW3* domains from Homo sapiens Nedd4 (hNedd4-WW3* and hNedd4L-WW3*). These WW3 domains are starred because this type of WW domain is missing in rat and mouse Nedd4. Moreover, the third WW domain in rNedd4 is a homologue of hNedd4-WW4 (Figure 4.2), so it is written as rNedd4-WW3/4. (214, 349) Figure 4.2 shows the positioning of the WW domains studied in this thesis within the Nedd4 proteins and their alignment with respect to other Nedd4 proteins. The explicit amino acid sequence alignment of the WW domains is given in appendix Figure F.1. 103 As stated above, the aim of this chapter was to test whether U24 can bind to Nedd4-WW domains with high affinity. The affinities and binding sites of the interactions between U24 and Nedd4-WW domains were determined using ITC and NMR. Since there are differences in the N-terminal segments of U24-6A and U24-7, it will be interesting to find out whether those positions are crucial for the interactions. This could help to explain the differences in the two versions of U24 and whether U24 may bind better than other PY ligands. 4.2 Results 4.2.1 Pull-down assays with GST-Nedd4-WW domains and U24 proteins It has been demonstrated that U24-6A protein could coimmuprecipitate with several types of WW domain containing proteins. (172) To determine whether Nedd4-WW domains can interact with both U24-6A and U24-7 proteins, pull-down assays using glutathione S-transferase (GST)-tagged Nedd4-WW domains were performed. The WW domains mentioned above, rNedd4-WW2, rNedd4-WW3/4, hNedd4-WW3* and hNedd4L-WW3*, were fused to the C-terminus of GST protein individually. Once immobilized on the GST 4B resin, these GST-tagged proteins can be used as “bait” proteins to isolate potential binding partners, such as U24-6A and U24-7, from a solution. The pull-down assays were conducted using the protocol described in section 4.5.6. 104 Figure 4.3 SDS-PAGE result of GST pull-down experiment with GST-Nedd4-WW domains and U24 Pull-down experiment using A) GST-rNedd4-WW2 and GST-rNedd4-WW3/4, B) GST-hNedd4-WW3* and GST-hNedd4L-WW3* and U24-6A and U24-7. U24-6A and U24-7 are loaded in the first lane as references, and the GST tag alone plus U24 are in the second lane. Lanes 3 and 5 are loaded with GST-Nedd4-WW domains alone. GST-Nedd4-WW domains plus U24 are loaded in lanes 4 and 6. The marker lanes are indicated by lane M. Based on the molecular weight ladder of the marker, the positions of GST and GST-Nedd4-WW domains are marked at the right side of the gel. In each case, the bottom gel shows that the amount of GST-tagged protein used is uniform for each combination tested. 105 The pull-down results were analysed using SDS-PAGE, and the gel images are shown in Figure 4.3. Besides the section of the gel corresponding to lower molecular weight (MW) proteins (i.e. containing U24), the section containing GST or GST-Nedd4-WW domains are presented to demonstrate that a uniform quantity of protein was used for each condition tested (i.e. in each lane). The protein bands detected at the U24-6A or U24-7 MWs in lanes 4 and 6 indicated that all four WW domains tested can interact with both U24-6A and U24-7 protein. This is similar to the findings of Sullivan and Coscoy. (172) The gel was stained using silver stain, so the intensity of the bands cannot be used to directly assess the strength of the binding interaction. 4.2.2 Affinities between Nedd4-WW domains and U24 peptides In order to compare the affinities between U24-6A and U24-7, dissociation constants for interactions between the U24/Nedd4-WW domain pairs tested above were determined using ITC. 15-mer peptides representing the N-terminus of U24-6A and U24-7 (see section 3.5.2 for the protocol) were used as the ligand or injectant. Titration experiments using these peptides were carried out on all four Nedd4-WW domains in 10 mM sodium phosphate, pH 7.4, at 25 °C. Control experiments in which the phosphate buffer was titrated into the protein solutions were also conducted to ensure the proteins are stable during the titration with peptide. Figure 4.4 shows the ΔH plots of U24-6A or U24-7 titrated into the four Nedd4-WW domains. Each point is an integrated heat, which is normalized and corrected for the effect of U24 peptide dilution. The curves are fitted using the “one set of sites”† model and all parameters are allowed to vary, including the number of binding sites on the WW domain, n. The mean values of the fitted thermodynamic parameters from three ITC experiments for each titration are reported in Table 4.1. The curves in Figure 4.4C and D have a symmetrical S-shape and saturation is achieved much faster than in the curves shown in Figure 4.4A and B. This indicates that the U24s have higher † “one set of sites” model is a fitting method described in the Origin 7, ITC manual; it fits n, Ka, and ΔH for a single set of identical binding sites using standard Marquardt methods. 106 affinities with hNedd4-WW3* and hNedd4L-WW3* than the other two domains. A comparison of all four panels in Figure 4.4 clearly shows an enthalpy difference between the titrations using U24-6A and U24-7 in all cases. Extrapolating to a molar ratio of 0, one can see that the red curves (U24-7) show more negative enthalpies than the blue ones (U24-6A). This similar trend can also be seen in the ΔH values found in Table 4.1. The difference in ΔH values obtained for U24-6A versus U24-7 is about 20 kJ/mol. From the fitted data in the table, the standard deviations of rNedd4-WW2 experiments are significantly bigger than the others. This is because it was necessary to run the ITC experiments at low concentrations and therefore at a c value less than 1 (see details in section 1.4.3). The rNedd4-WW2 domain was found to be prone to degradation and aggregation during sample preparation, so preparation of a concentrated protein solution was avoided prior to the ITC experiment. Table 4.1 Parameters obtained from fitting the ITC data for binding of U24-6A or U24-7 to rNedd4-WW2, rNedd4-WW3/4, hNedd4-WW3* and hNedd4L-WW3* domains at 25 °C. The parameters are obtained from fitting the data from three separate runs and averaging them. The errors represent ± one standard deviation. Protein concentrations used in these titrations are about 30 to 80 μM. U24-6A U24-7 Kd (μM) ΔH (kJ/mol) ΔS (J/mol K) Kd (μM) ΔH (kJ/mol) ΔS (J/mol K) rNedd4-WW2 155 ± 12 -54 ± 5 -110 ± 19 87 ± 5 -76 ± 3 -176 ± 11 rNedd4-WW3/4 43 ± 2 -46 ± 2 -70 ± 7 21.0 ± 0.5 -63.5 ± 0.5 -123 ± 2 hNedd4-WW3* 9.7 ± 0.2 -53.0 ± 0.1 -81.7 ± 0.5 1.98 ± 0.04 -70.1 ± 0.2 -126 ± 1 hNedd4L-WW3* 6.3 ± 0.3 -55.3 ± 0.3 -86 ± 1 1.22 ± 0.01 -68.8 ± 0.1 -117.8 ± 0.2 107 Figure 4.4 ΔH plots of U24-6A or U24-7 titrated into the four Nedd4-WW domains. Normalized heats and fits are shown for U24-6A peptide in blue and U24-7 peptide in red. The peptides are titrated into a solutions of A) rNedd4-WW2 domain, B) rNedd4-WW3/4 domain, C) hNedd4-WW3* domain and D) hNedd4L-WW3* domain. Protein concentrations used in these titrations are about 30 to 80 μM. All titrations are done at 25 °C. The horizontal axis is the molar ratio between U24 and WW domain. The vertical axis is the normalized heat in kcal/mol. 108 4.2.3 The effect of temperature on U24-WW domain interactions The affinities of the interactions between U24-7 and WW domains are much stronger than those of U24-6A and the WW domains tested. Since there is no difference between the two PY motifs in U24-6A and U24-7 ligands, it is plausible that the residues up- or downstream from the PY motif affect the interaction. One possibility is that most WW domains prefer the flanking residues of the PY motif in U24-7, in particular the negatively charged glutamic acid at position -2 (Figure 4.1). This residue could either contribute to or even directly participate in the interaction. In the case of SH3 domain, it has been found that the flanking arginine in the PxxP motif could increase the affinity ten times. (187) The importance of flanking residues could be similar in the case of the WW domains. If there is an additional contact between the WW domain and those flanking residues in U24-7, the binding site area between U24 and the WW domain may be different for U24-6A than in U24-7. In general, the binding of PY motif ligands to WW domains is mainly mediated through their hydrophobic residues, (78) leading to a burial of the hydrophobic interface. The change of heat capacity upon binding at constant pressure of the interaction, ΔCP, can be used to characterize the burial of a hydrophobic surface. (350) ΔCP can be measured using techniques such as differential scanning calorimetry (DSC) or can be derived from the heat of interaction obtained from ITC experiments run at different temperatures (Equation (4.1)). Therefore in order to determine whether hydrophobic burial plays a role in the interaction of U24 with WW domains, additional ITC experiments of U24 titrated into solutions of the WW domains were carried out at 15 °C and 37 °C, in 10 mM sodium phosphate, pH 7.4. In addition, CD melt experiments were conducted to ensure the stability of all four WW domains between 5 to 95 °C. A typical result is shown in appendix Figure D.2. rNedd4-WW2 and hNedd4-WW3* domains were found not to be stable above 25 °C, so ITC experiments as a function of temperature were not conducted. Since the CD melt did not show a significant transition for hNedd4-WW3* and hNedd4L-WW3*, NMR experiments were carried out to verify further the structural instability of hNedd4-WW3* (Figure C.2). The values of the thermodynamic parameters obtained for U24-6A or U24-7 interacting with rNedd4-WW3/4 and 109 hNedd4L-WW3* are reported in Table E.1. They show classic enthalpy-entropy compensation in biomolecular interactions. (351) The mean ΔH values obtained from three individual experiments was then plotted as a function of temperature (Figure 4.5). A linear fit of the data in Figure 4.5 yields ΔCP. The fitted ΔCP values and the associated errors obtained from linear regression are reported in Table 4.2. ∆CP = (∂H∂T)P (4.1) Figure 4.5 Plots of ΔH vs. temperature for U24-6A or U24-7 titrated into A) rNedd4-WW3/4 and B) hNedd4L-WW3* domains. A) The data points in blue circles represent the average ΔH from the titrations using U24-6A titrate into a solution of rNedd4-WW3/4 at 15 °C, 25 °C and 37 °C. The error bars represent the standard deviation of ΔH from three independent experiments. The pink squares represent the analogous values for U24-7. The blue line is the linear fit of the series of ΔH at different temperatures for U24-6A, while the red line is the linear fit for U24-7. B) The ΔH data and linear fit for U24-6A (blue circles and blue line) or U24-7 (pink squares and red line) titrated into hNedd4L-WW3* domain at 15 °C, 25 °C and 37 °C. The error bars and lines have the analogous meaning as in A. 110 Table 4.2 Change in heat capacities upon binding of U24-6A or U24-7 to rNedd4-WW3/4 or hNedd4L-WW3* domains. The error in the slope is obtained from a linear regression analysis. Protein concentrations used in these experiments are about 40 to 150 μM. ΔCP (kJ/mol K) U24-6A U24-7 rNedd4-WW3/4 -1.11 ± 0.09 -0.9 ± 0.1 hNedd4L-WW3* -0.9 ± 0.2 -0.93 ± 0.07 In Figure 4.5, all four lines are nearly parallel, indicating that there is little difference in the change in heat capacity amongst the various U24 and WW domain combinations. The ΔCP values reported in Table 4.2 however exhibit interesting features. The heat capacity change of U24-6A appears to be slightly larger than that of U24-7 when interacting with rNedd4-WW3/4, but smaller than that for U24-7 interacting with hNedd4L-WW3*. It has to be noted, however, that the errors are large (as is typical) and thus the statements above should be interpreted with caution. S ince only three different temperatures were sampled, this results in a large uncertainty in data fitting. Overall, the trends observed in Table 4.2 suggest that there is no distinct difference between the change of heat capacities between U24-6A and U24-7. 4.2.4 Phosphorylation on threonine of U24-6A enhances the affinities with WW domains. It has been demonstrated that U24-6A protein is extensively posttranslationally modified in vivo and that Thr6 of U24-6A can be phosphorylated in vitro. (36, 54) It is possible that these modifications play a role in the interactions with WW domains. Moreover, there is an arginine upstream from the PY motif in U24-6A, while a glutamic acid is found at the same position (-2 position in Figure 4.1) in U24-7. The opposite charge found at the -2 position might be a reason for the different affinities observed for a given WW domain. If Thr6 in U24-6A is phosphorylated, then there will be a negative charge near the -2 position. The phosphorylated 15-mer U24-6A peptide, noted as pU24-6A (Figure 4.6) was used to test whether the charge at the -2 position is crucial for 111 U24-WW domain interaction. If the charge is crucial for the interaction, pU24-6A should have a better affinity with Nedd4-WW domains than U24-6A. At the same time, pU24-6A peptide is a more biologically relevant model ligand than U24-6A. It could be helpful to explain how U24-6A behaves in vivo. U24-6A M DPPR T PPPSYS E VL pU24-6A M DPPRpTPPPSYS E VL U24 -7 M -THE T PPPSYN D VML Figure 4.6 Sequence alignment of the first 15 residues of U24-6A, pU24-6A and U24-7. The PY motifs of three ligands are highlighted with a yellow background. The positively charged residues are shown in blue, while the negative ones are shown in red. The phosphorylated threonine residue in pU24-6A is shown in green. ITC experiments were carried out at 25 °C in 10 mM sodium phosphate, pH 7.4, as described in section 4.5.7 using pU24-6A peptide (95% purity, custom synthesis from Genscript, New Jersey) and rNedd4-WW3/4 or hNedd4L-WW3* domains. The ITC data, including an example of the raw data obtained, are shown in Figure 4.7. The mean values of the fitted thermodynamic parameters and binding site n are reported in Table 4.3. The raw ITC data for the pU24-6A/hNedd4L-WW3* interaction (Figure 4.7A) has a typical s-shape curve. The curves for pU24-6A, shown in green (Figure 4.7B and C), demonstrate that saturation is reached faster than for U24-6A or U24-7. These ΔH plots also show a very interesting normalized heat pattern, as the heat of pU24-6A is between that of U24-6A and U24-7. From Table 4.3, the affinities between pU24-6A and Nedd4-WW domains are better than those between U24-6A or U24-7 and the given Nedd4-WW domains. In particular, the Kd of pU24-6A binding to hNedd4L-WW3* is 758 ± 27 nM, which is the best affinity characterized to date between a U24 peptide ligand and WW domains. 112 Figure 4.7 pU24-6A peptide interaction with WW domains at 25 °C A) ITC data with hNedd4L-WW3*. ΔH plot for the three peptide ligands interacting with B) rNedd4-WW3/4 or C) with hNedd4L-WW3* domain. A) An example of the ITC data of pU24-6A peptide binding to hNedd4L-WW3* at 25 °C, containing raw titration data (upper) and the corresponding ΔH plot (bottom). B) Normalized heats and the fitting curves shown for U24-6A peptide (blue), pU24-6A (green) and U24-7 peptide (red) titrated into B) rNedd4-WW3/4 domain and C) hNedd4L-WW3* domain. The horizontal axis is the molar ratio between U24 peptide and the WW domain. Protein concentrations used in these experiments are about 40 to 60 μM. 113 Table 4.3 Parameters obtained from fitting the ITC data for binding of pU24-6A to rNedd4-WW3/4 and hNedd4L-WW3* domains at 25 °C. The parameters are obtained from fitting the data from three separate runs and averaging them. The errors represent ± one standard deviation. Protein concentrations used in these experiments are about 40 to 60 μM. pU24-6A n (binding sites) Kd (μM) ΔH (kJ/mol) ΔS (J/mol K) rNedd4-WW3/4 0.999 ± 0.009 13.5 ± 0.4 -54 ± 1 -87 ± 4 hNedd4L-WW3* 1.179 ± 0.005 0.76 ± 0.03 -58.7 ± 0.4 -79.8 ± 0.9 4.2.5 Electrostatic effect of U24 binding to Nedd4-WW domains The last section demonstrated that phosphorylation at Thr6 enhances the affinity of U24-6A with Nedd4-WW domains. The phospho-threonine adds a negative charge around the -2 to -1 position of the PY motif in U24-6A (Figure 4.1), making it more similar to the PY motif of U24-7. Both pU24-6A and U24-7 bind WW domains strongly. It is plausible that the negative charge upstream from PY motif promotes binding in the WW domains via electrostatic interactions. There are several conserved lysines or arginines in Nedd4-WW domains (see alignment in Figure F.1) that could provide the right counter charge for the pU24-6A and U24-7 ligands. Electrostatic interactions, also known as salt bridges, are often found to facilitate protein-protein interactions. A typical distance between the relevant functional groups is about 4 Å. (352, 353) The strength of electrostatic interactions depends on salt concentrations and the pH of the solution, as external charges can shield the functional groups and thereby reduce the interaction strength. In order to find out whether electrostatic interactions play a role in the binding of U24 peptides to Nedd4-WW domains, a set of ITC experiments were performed as a function of increasing salt concentration at 25 °C. The experiments described in sections 4.2.2 to 4.2.4 used the same buffer, namely 10 mM sodium phosphate, pH 7.4. This buffer has been also used in other WW domain studies reported in the literature. (78) In the experiments described in this section, ITC 114 experiments were performed using U24-6A, pU24-6A and U24-7, binding to rNedd4-WW3/4 and hNedd4L-WW3* in 10 mM sodium phosphate, 100 mM NaCl, pH 7.4, and 10 mM sodium phosphate, 500 mM NaCl, pH 7.4. The fitted data is reported in Table E.2. Affinities and enthalpies of each U24-WW domain pair as a function of salt concentration are plotted in Figure 4.8. The equilibrium constant Ka is plotted as ln Ka in Figure 4.8 (C and D), as is typically done in literature. (354, 355) From the ΔH plot in Figure 4.8A, little difference was observed among the ΔH values for a given ligand binding to rNedd4-WW3/4, as a function of increasing salt concentration. Considering the error bars shown, no clear trend can be obtained from a comparison of the ΔH values, suggesting that there might be no strong electrostatic interactions at play in this protein-ligand interaction. On the other hand, in panel B of Figure 4.8, there are statistically significant differences in the case of pU24-6A and U24-7. Interestingly, the trends observed are different for the two ligands. The ΔH for U24-7 interacting with hNedd4L-WW3* becomes more negative with increasing salt concentration, whereas that of pU24-6A becomes less negative at high salt concentration (500 mM NaCl). This could be result from specific salt bridges between pU24-6A and hNedd4L-WW3* being disturbed by shielding effects from the surrounding ions. The ln Ka of all three U24 ligands binding to rNedd4-WW3/4 increase with increasing salt concentration (Figure 4.8C), indicating that the affinities are getting higher. The ln Ka plot for hNedd4L-WW3* shows an opposite trend. (Figure 4.8D) The affinities between pU24-6A or U24-7 and hNedd4L-WW3* are found to decrease as the salt concentration increased, while that of U24-6A seems to be unaffected. Moreover, the equilibrium constant of pU24-6A is reduced more significantly than that of U24-7 upon additional salt. The Ka values of pU24-6A and U24-7 in 500 mM NaCl are almost the same. Additional experiments with different salt concentrations were also performed on hNedd4-WW3*, resulting in similar trends to those seen for hNedd4L-WW3* (Figure E.3 and Table E.2). 115 Figure 4.8 ΔH and ln Ka plots of the three U24 ligands binding to rNedd4-WW3/4 and hNedd4L-WW3* in phosphate buffer, phosphate buffer with additional 100 mM NaCl or 500 mM NaCl at 25 °C. ΔH plot of U24-6A (blue bars), pU24-6A (green bars) and U24-7 (red bars) titrated into A) rNedd4-WW3/4 or B) hNedd4L-WW3* solutions in three different buffers. Protein concentrations used in these titration experiments are about 40 to 80 μM. ln Ka plot of the three ligands titrated into C) rNedd4-WW3/4 or D) hNedd4L-WW3* in three different buffers. Solid bars indicate that the experiments were done in 10 mM sodium phosphate, pH 7.4. Bars with coloured dots were for experiments done in phosphate plus an extra 100 mM NaCl, while bars filled with grids were for experiments where an extra 500 mM NaCl was added to the phosphate buffer. Error bars indicate ± one standard deviation, obtained from three repeats (n=3). n.s., nonsignificant; *P≤0.037, **P≤0.0088, ***P≤0.001, ****P≤0.0001 by Student’s t test. 116 4.2.6 Binding site residues In addition to determining the affinities between U24 and the Nedd4-WW domains using ITC, 1H-15N HSQC NMR spectroscopy was performed to characterize the binding site in rNedd4-WW3/4 and hNedd4L-WW3* domains upon ligand binding. 15N-labelled rNedd4-WW3/4 and hNedd4L-WW3* were used in the experiment and the spectra of apo WW domain were recorded at 25 °C in 10 mM phosphate buffer, pH 7.4. HSQC spectra of six protein-peptide complexes, involving three U24 ligands and two WW domains, were also recorded under the same conditions. The assignment of apo rNedd4-WW3/4 and hNedd4L-WW3* domains and the complexes were achieved using 3D HSQC-NOESY and 3D HSQC-TOCSY experiments (see details in section 1.4.1). The changes in 1H and 15N chemical shifts were combined to yield Δδcomb, obtained using the equation in section 4.5.8. The combined changes in chemical shift are divided into five groups, using the method described by Schumann et al. (356) The nitrogen atoms of the perturbed amides in the two WW domains studied are shown as spheres in Figure 4.9. The spheres are coloured based on the magnitude of Δδcomb. Generally, the Δδcomb determined for the three ligands U24-6A, pU24-6A and U24-7 were very similar (Figure 4.9A and D). Indeed, the peak positions seen in the overlays of the HSQC spectra in Figure 4.9B-C) and E-G) are generally quite close, with a few exceptions. The full titration overlays are not shown because most U24 and Nedd4-WW domain interactions were found to be in the intermediate exchange regime (see details in section 1.4.2). As shown in Figure 4.9A, the residues perturbed the most in rNedd4-WW3/4 were on its third β sheet (β3 sheet), and the loop between β1 and β2 (β1-β2 loop), which is important for binding PY motif ligands. (78) Similarly, strong perturbation was be found on the hNedd4L-WW3* at β3, but only moderate perturbation was found in the β1-β2 loop. This difference between the two WW domains could be due to differences in the amino acid sequence. Moreover, hNedd4L-WW3* contains a proline in the β1-β2 loop, which might limit the flexibility of the loop, resulting in fewer perturbations of the backbone. In addition, the residues in the β2-β3 loop and at the end of β3 in rNedd4-WW3/4 are more strongly perturbed than the corresponding residues in hNedd4L-WW3*. On the other hand, strong perturbation on β2 117 of hNedd4L-WW3* is observed, but not in rNedd4-WW3/4. As mentioned above, the HSQC overlays of rNedd4-WW3/4 and hNedd4L-WW3* are shown in Figure 4.9B-C) and E-G), respectively. The residue numbering of rNedd4-WW3/4 and hNedd4L-WW3* corresponds to rNedd4 (GenBank: AAB48949.1) and hNedd4L isoform 5 (Uniprot: Q96PU5-5). In the case of rNedd4-WW3/4, most amide resonances showed similar chemical shift perturbations when complexed with U24-6A and pU24-6A, indicating that the backbone is perturbed similarly by these peptides. There are however some residues, such as R468, V475, F476 and the W487 side-chain, which are perturbed differently by the three ligands. Presumably many of these residues, which are in close promixity (Figure 4.9A), may be involved in a direct interaction with the phospho-threonine in pU24-6A. These differences can be easily observed in the spectra (section C7) but are very difficult to quantify since the Δδcomb is calculated using the root mean square of chemical shift perturbations in both 15N and 1H dimensions. For the interaction with hNedd4L-WW3*, a comparison of the spectra obtained upon titration of U24-6A and pU24-6A showed similar chemical shift perturbations. Indeed, most of the blue and green peaks in Figure 4.9 panels E-G) are overlapped. Here there are two notable exceptions, for residues R492 and W505, where distinct chemical shifts were observed upon binding to each of the three U24 ligands. These residues are found in an analogous region to those which were perturbed in rNedd4-WW3/4, suggesting that this region may again be important for specific interactions between residues N-terminus to the PY motif and the WW domain. 118 119 Figure 4.9 Combined chemical shift perturbations mapped on A) rNedd4-WW3/4 and D) hNedd4L-WW3* domain when complexed with U24 peptides. HSQC overlays for B-C) rNedd4-WW3/4 and E-G) hNedd4L-WW3*. A) Coloured model of rNedd4-WW3/4 and a PY motif peptide (PDB: 1I5H). The WW domain is shown in green and the peptide is shown as a grey ribbon. The PY motif is highlighted in yellow. The nitrogens are shown as spheres and coloured according to the Δδcomb scale described in section 4.5.8. B-C) Regions of the HSQC spectra of apo rNedd4-WW3/4 (black), and rNedd4-WW3/4 bound with U24-6A (blue, protein:peptide=1:4), with pU24-6A (green, 1:3) or with U24-7 (red, 1:4). D) Coloured model of hNedd4L-WW3* and a PY motif peptide (PDB: 2MPT). The WW domain is shown in purple. The Δδcomb data was mapped onto this domain as described in A). E-G) Regions of HSQC spectra of apo hNedd4L-WW3* (black), and hNedd4L-WW3* bound with U24-6A (blue, 1:2), with pU24-6A (green, 1:2) or with U24-7 (red, 1:2). 4.2.7 Long-range NOEs found in the U24-WW domain complexes In order to further characterize differences in the binding interaction between the various U24 peptides and the WW domains, 3D HSQC-NOESY experiments were conducted. The NOESY mixing time for these experiments was set to 200 ms, so as to yield long-range NOE signals, which may include intermolecular NOEs. Two sets of three U24/WW domain complexes were inspected using 3D HSQC-NOESY experiments: rNedd4-WW3/4 saturated with U24-6A, pU24-6A or U24-7, as well as hNedd4L-WW3* saturated with the same three U24 ligands. Since the focus in this section was to find differences between the three complexes, the NMR data was analyzed by comparing strip plots. When comparing the NOE signals from a given residue among the three complexes side by side, different numbers of NOEs or different intensities could be observed. About 60 NOEs that could reflect differences based on number of contacts or intensities in three ligands from each WW domain were analysed. If the binding of one U24 ligand causes two residues to come closer, it is possible that the motion of the protein at that position is reduced or, in other words, that the structure is more rigid. To visualize these results, for each NOE, the strongest signal or shortest distance among the three complexes was represented by a coloured solid line (Figure 4.10). For new or unique NOEs observed for a given complex, the contact was indicated by a coloured dotted line. If a new NOE is observed but the assignment of the residue it interacts with is ambiguous, such as from a nearby proline or from the peptide ligand, a coloured star is drawn next to the residue. 120 Figure 4.10 Schematic figure of NOEs found upon U24 peptide binding to A) rNedd4-WW3/4 and B) hNedd4L-WW3* domain Stronger/new NOE signals detected when U24-6A (blue solid/dotted lines), pU24-6A (green solid/dotted lines) and U24-7 (red solid/dotted lines) are complexed with A) rNedd4L-WW3/4 domain or B) hNedd4L-WW3* domain. 121 Overall, Figure 4.10 shows that there are a number of new or more intense NOEs observed, in particular when U24-7 and pU24-6A are added. For rNedd4L-WW3/4 domain, most of these stem from for the sample consisting of U24-7 and the WW domain, whereas for hNedd4L-WW3*, most arise for the pU24-6A sample. It is also interesting to note where these additional or stronger contacts arise. In Figure 4.10A, there are several lines around the loop region, mainly caused by U24-7. D472 seems to have a central role in the β1-β2 loop and the additional NOEs suggest that it may become more rigid upon ligand binding. pU24-6A seems to pull the end of β2 closer to β1, while there is little unique impact of U24-6A. As for the ambiguous NOEs, most of them are around the binding site, which is the second conserved tryptophan. In panel B, there are several lines outside of the loop region, indicating that the sheets are possibly more restricted than in rNedd4L-WW3/4. Moreover, N490 in hNedd4L-WW3* domain, located in an analogous position to D472 in rNedd4L-WW3/4, has many additional contacts for the sample to which pU24-6A was added. The unidentified NOEs in this WW domain seem to cover a larger area in the case of pU24-6A, centered around W505. Finally, there are more unique NOEs (dotted lines, Figure 4.10B) in the hNedd4L-WW3* domain than in rNedd4-WW3/4 domain, especially for pU24-6A ligand. Overall, this analysis suggests that binding of the ligands has an impact on the structure of the WW domains studied here. A full structure determination may be of interest in the future to quantify more specifically what these changes are. 4.2.8 Slow-exchange of pU24-6A and hNedd4L-WW3* complex In solution, protein conformational changes caused by folding, aggregation or ligand binding happen in real time, and can be directly observed using NMR spectroscopy. In section 1.4.2, the NMR methods used to detect such changes were briefly introduced. Through the observation of how the titration peaks shift, disappear or how their intensities change, one can find the relative exchange rate of protein, leading to an estimation of the dissociation constant. In the studies of the binding of U24 to WW domains described in section 4.2.6, most of the amide resonances underwent intermediate exchange, since they disappeared during the titration and reappeared at another frequencies. However, for pU24-6A binding to the hNedd4L-WW3* domain (Kd = 0.76 ± 122 0.03 μM from ITC, Table 4.3), there were two sets of peaks detected during titration (Figure 1.14). This indicates that the exchange rate is slow on the chemical shift time scale. In this case, the dynamics of slow protein exchange can be investigated further using EXSY. Before conducting the EXSY experiment, a seHSQC spectrum of a sample of hNedd4L-WW3*:pU24-6A=1:0.6 was recorded to confirm that both Apo (A) and Bound (B) peaks are narrow. This sample was prepared using the protocol described in sections 4.5.4 and 4.5.8. Besides the peaks from the A and B form, some weak cross-peaks were observed in the spectrum (Figure 4.11 left middle). The seHSQC spectra shown in Chapter 2 (Figure 2.11) consisted of a single resonance per amide group. The appearance of additional peaks in the seHSQC experiment was therefore a bit surprising, since this experiment is designed to enhance sensitivity (by using two consecutive reverse INEPT pulse elements (281)), but not specifically for studying chemical exchange. A possible explanation for the observed cross-peaks is that chemical exchange happens right after transferring the nitrogen magnetization back to proton, during the second INEPT (i.e. within the ca. 4 ms delay). Such phenomena have been suggested in the literature. (291) To confirm whether slow exchange was occurring in the hNedd4L-WW3*:pU24-6A sample, a series of HSQC and EXSY spectra were recorded using a Bruker Avance III 850 MHz spectrometer at 15 °C and 25 °C. Selected spectra from the experiments at 15 °C are shown using the same starting contour and number of contour levels in Figure 4.11. The results for the experiments run at 25 °C are shown in appendix C5. 123 Figure 4.11 Region of seHSQC, HSQC and EXSY spectra of pU24-6A binding to hNedd4L-WW3* complex at 15 °C From left top to bottom, apo (black) and bound (green) seHSQC overlays of hNedd4L-WW3* or bound with pU24-6A at (1:2); seHSQC of the hNedd4L-WW3*:pU24-6A=1:0.6 sample; HSQC of this same sample, with a mixing time T=0 ms (Figure 1.14). From right top to bottom, EXSY spectra with mixing time T=8 ms, 24 ms or 60 ms delay. All the spectra, except the overlays of apo and bound spectra (top left), are plotted using the same starting contour and number of contour levels. 124 From Figure 4.11, cross-peaks are clearly observed in the seHSQC spectrum for residues E484 and E506, but are absent in the HSQC spectra. The intensities of both the apo and bound peaks (auto-peaks, AA and BB) are stronger in the seHSQC spectra as well. The cross-peaks of these two residues are not observed in the EXSY spectra for mixing times (T) smaller than 8 ms. Similarly to what is observed in the seHSQC spectra, cross-peaks for both E484 and E506 are observed in the EXSY spectrum with T=8 ms, while for this mixing time, residues H498 and I496 displayed only one cross-peak. As the mixing time was increased, the intensities of the cross-peaks increased as well, but the overall intensities of the AA and BB peaks decreased, presumably due to the relaxation of the spins. In the last EXSY spectrum with T=60 ms, the intensities of cross-peaks are strong and some of them were almost as intense as the auto-peaks that the magnetization exchanged from, e.g. E506-AB and E506-AA. At the same time, the dominant resonance of bound form, E506-BB and E506-BA, is still E506-BB. This indicates that the rates of relaxation of these two species are most likely different. Similar findings were observed in the experiment at 25 °C (Figure C.3). In the spectra obtained at 25 °C, the exchange peaks in the EXSY spectra with 4 ms delay were similar to those seen in the seHSQC spectra, but with lower intensities. Although the spectra shown in Figure 4.11 focussed on select resonances, the EXSY cross-peaks were observed for most residues. However, due to strong overlap, most groups of four resonances were difficult to identify and analyze further. Therefore, intensities for only a handful of residues were fitted to Equations (1.2) to (1.5). Only one set of spectra, with various T values, was collected, so the integrated peak volumes are associated with random errors. The data fitting shown in this section is a preliminary result. The exchange kex, relaxation rates of apo and bound form (R1A0 and R1B0 ) were allowed to vary, while pA was fixed at 0.4. The non-linear regression fitting was done using generalized reduced gradient algorithm and the results for two sets of peaks from residues E484, R492 and E506 are reported in Table 4.4. The two glutamic acids are next to the two conserved tryptophans, while R492 was shown perturbed differently by three slightly different U24 peptide ligands. Both R492 and E506 are near the N-terminus of the PY motif, namely the phospho-threonine, while E484 is on β1, where it might not directly contribute to the ligand binding 125 (Figure 4.10). The non-linear regression fit generated assuming R1A0 ≠ R1B0 has a very small relaxation rate for apo form R492 and E506, which yielded a long relaxation time T1A of 4 to 8 s. This result does not make much sense since the AA peak of apo form peaks are very weak when the delay is 60 ms (Figure 4.11 and Figure C.4). These preliminary fitting values are not accurate as the coefficient of determination is low (R2 ≈0.8). The other method, assuming R1A0 = R1B0 , yields more resonable relaxation rates for both forms but the R2 is even lower. Nevertheless, the fitted kex values are similar in both cases, around 40 s-1, and are well within the slow exchange range of around 0.2 to 100 s-1. (357) Overall, the EXSY data shows interesting effects occur for the pU24-6A/hNedd4L-WW3* complex. More experiments and careful analysis of the data may provide complementary insights to the NOE data described above and allow one to pinpoint how binding affects the WW domain structure and dynamics. Table 4.4 Preliminary data fits of pU24-6A peptide binding to hNedd4L-WW3* domain. R2 of the fits are reported to show the goodness of fit. Fits assuming R1A0 ≠ R1B0 Fits assuming R1A0 = R1B0 E484 (s-1) R492 (s-1) E506 (s-1) E484 (s-1) R492 (s-1) E506 (s-1) kex 45 42 35 43 41 30 R1A0 2.4 0.12 0.24 6.3 1.8 3.0 R1B0 9.6 4.8 4.1 6.3 1.8 3.0 R2 (unitless) 0.81 0.81 0.83 0.80 0.79 0.82 4.3 Discussion The results presented in sections 4.2.1 and 4.2.2 show that U24-6A and U24-7 protein, and peptides consisting of the first 15 residues in the sequence can bind to all four WW domains tested , namely rNedd4-WW2, rNedd4-WW3, hNedd4-WW3* and hNedd4L-WW3*. Interestingly, a 126 phosphorylated version of U24-6A, pU24-6A, is found to have the best affinities with the WW domains (section 4.2.4). Overall, the dissociation constants between the U24 peptides and the WW domains are in the range of low micromolar, indicating strong binding. Both hNedd4 and hNedd4L-WW3* have much better affinities with U24 peptides than the other two WW domains. These two domains are homologues to the high-affinity WW3* domains characterized in the literature, (214) and hNedd4-WW3* has been proven to be the most important for interactions with PY motifs, since a construct containing 3 WW domains from hNedd4 (i.e. WW2+WW3*+WW4 of hNedd4) was found to bind no differently than hNedd4-WW3* alone. (348) Interestingly, the affinities of U24 peptides binding to the WW3* domain of hNedd4 or hNedd4L at 25 °C, especially pU24-6A, are better than most of the ligands reported to date in the literature on single WW3* domains. (247, 358, 359) This strong interaction suggests that U24 could compete or inhibit the intrinsic interactions of WW3* domains in vivo. In order to function in a cell, Nedd4 protein associates with many other proteins and its WW domain has been found to interact with a substrate, an adaptor or an inhibitor. In the latter case, the inhibitor is normally part of the Nedd4 protein itself. (71, 242, 358, 360) The exact function of U24 is currently unknown, but the work described here has demonstrated that it can strongly interact with Nedd4-WW domains, suggesting it may play one or more of the roles listed above. Since there is no lysine in the cytosolic segment of U24 protein, it is unlikely that the HECT domain of Nedd4 could ubiquitinate U24, therefore making U24 an unlikely substrate of Nedd4. U24 is very small and its cytosolic segment is even smaller (about 50 residues). Most of the adaptors of Nedd4 will have another signalling domain or motif besides the PY motif. (225, 361, 362) Still, some of the adaptors can activate and regulate by only using the PY motif, such as NDFIP, an adaptor of Nedd4 family E3. (242) Considering that both U24 and NDFIP are endosomal membrane proteins containing PY motifs and have a potential role in protein trafficking, (363) it is possible that U24’s function is similar to that of NDFIP. Indeed, the known function of U24, down-regulating TCR/CD3, could be due to the activation of Nedd4 family E3 in sorting endosomes, which result in the altered fate for these receptors. Finally, U24 might be an inhibitor of Nedd4, i.e. interfering with the functions 127 of NDFIP by blocking one or many of its downstream signalling pathways. (364, 365) NDFIP1 is crucial for neuron survival, as it is responsible for regulating the level of iron in neuronal cells. (366, 367) Iron accumulation has been detected in the lesion area of MS brain tissues and found to contribute to the pathogenesis and progression of MS. (368, 369) Interfering with NDFIP, U24 may promote iron accumulation in lesion areas. Besides NDFIP, other substrates or adaptors of Nedd4 protein may not be able to bind to Nedd4 if U24 strongly associates with Nedd4-WW domains. In other words, the strong binding observed and described here may mean that U24 can reduce the opportunities for other ligands to bind Nedd4. The ramifications of this can be very complicated. Even though the exact function of U24 is still uncertain, it is highly possible U24 function is closely related to that of Nedd4. The thermodynamic parameters obtained in this thesis for the U24/WW domain interactions show that these interactions are enthalpy-driven since a large release of heat was observed upon binding. It suggests the interaction is dominated by bond formation, which could be formed intra- or intermolecularly. Specifically, the interactions with U24-7 were always found to have a more negative ΔH than those with U24-6A and pU24-6A, suggesting that more bonds are formed. This observation may be linked to the differences in the flanking residues of U24-6A versus U24-7, such as the charged or nonpolar residues upstream from the PY motif (Figure 4.6). The ΔH of pU24-6A binding to WW domains is between that of U24-6A and U24-7, although pU24-6A has the best affinities with the WW domains. This suggests that the modified phosphoryl group does not simply act as a replacement for the acidic side chain of the glutamic acid in U24-7. Since the phosphoryl group is larger than the Glu side chain, interactions between the N-terminus on the pU24-6A peptide and the WW domain are more likely to occur, translating into a higher affinity binding ligand. From the ITC studies as a function of temperature, no large differences between the changes in heat capacity of U24-6 and U24-7 binding to WW domains were found. It has been demonstrated that ΔCP is mainly affected by the changes in the solvent-accessible non-polar and polar areas, ΔAnp and ΔAp (change of area, non-polar and polar) upon binding. (370, 371) Equation (4.2) describes the 128 relationship between ΔCP and these two terms, where ΔAp and ΔAnp are in Å2. (372) If the entire interaction surface is considered to be non-polar, an estimated ΔAnp (noted as ΔAnp,e) could be calculated from this equation. That is, assuming ΔAp=0, the non-polar area which becomes buried upon binding to rNedd4-WW3/4 can be estimated to be around 680-820 Å2 and 670-700 Å2 for hNedd4L-WW3*. These |ΔAnp,e| values are pretty close to the reported total binding areas (ΔAtotal, defined in Equation (4.3) of PY ligand/WW domain complexes. (78, 214) |ΔAnp,e| represents a lower limit for the binding area for U24/Nedd-WW domain interactions because the contribution from the polar surface is ignored. Therefore, the actual ΔAtotal value of these interactions should be bigger than |ΔAnp,e|. Indeed, a larger interaction surface area is reported for Comm/dNedd4-WW3* complex, 895 ± 40 Å2. (214) Although dNedd4-WW3* is a homologue of hNedd4L-WW3*, the sequence of Comm and U24 are quite different (Figure F.2). Comparing the different U24s and WW domains, the calculated areas are found to be similar and fall within the ranges given above. This lack of distinction may be due to the partial cancellation of ΔAnp and ΔAp, to the errors in data fitting or to the discrepancies observed in the literature. (370, 373) Overall, the studies of U24 binding to Nedd4-WW domain at different temperatures show similar negative ΔCP for U24-6A and U24-7 peptide, suggesting that there is not a big difference in binding surface. ∆CP° =(0.32±0.04)∆Anp-(0.14±0.04)∆Ap Cal/mol K (4.2) ΔAtotal =|ΔAnp|+|ΔAp| (4.3) In addition, the effect of electrostatics on the interaction was investigated. The two WW domains, rNedd4-WW3/4 and hNedd4L-WW3*, have been found to behave differently. In the case of rNedd4-WW3/4, the affinities with all three U24 peptides are all slightly enhanced with increasing salt concentration. Increased ionic strength does not decrease U24 binding to rNedd4-WW3/4 domains. This suggests that there is little to no electrostatic interactions required to form the complex. The enhanced affinities could be due to the enhanced diffusion-controlled association rate, kon. (352) As for hNedd4L-WW3* domain, different trends were observed among the three ligands. The affinities of U24-6A did not change much upon increasing salt concentration, while the interactions with pU24-6A and U24-7 were weakened. The dependence of Ka on salt concentration, 129 symbolized by -SKa, can be used to estimate the total number of cations, anions or water molecules released as a result of the interaction, as defined in Equation (4.4). (354, 374, 375) In this equation, [MX] represents the concentration of salt, where M is the cation and X is the anion. The -SKa calculated for pU24-6A and U24-7 were about -0.2 and -0.1, indicating that there are essentially no ions or water released upon binding. The experimental data showed that the |ΔH| of pU24-6A is significantly reduced at high salt concentrations. This suggests that fewer bonds are formed upon binding. Indeed, an increase in entropy of the system is observed as well (Table E.2). At the same time, the smaller |ΔH| value observed might also be an error caused by the uncertainty in the pU24-6A peptide concentration. Overall, there is no strong electrostatic interaction found between U24 and Nedd4-WW domains. A very weak salt bridge might form in the case of pU24-6A binding to hNedd4L-WW3*. -SKa=-d logKa/d log[MX] (4.4) NMR spectroscopy was used to identify the differences among the three ligands at a molecular level. The data presented in section 4.2.6 suggests that all U24 ligands largely bind to the WW domains similarly, except for some residues. This is surprising given the different binding affinities obtained from ITC. In order to emphasize better the differences in binding of the three U24 ligands, the NMR chemical shifts were therefore re-analyzed. Instead of calculating a combined chemical shift for nitrogen and hydrogen, the chemical shifts of the six U24/Nedd4-WW complexes were inspected in both dimensions, respectively. For a given ligand, the peak position was deemed different if the chemical shift changed by 0.14 ppm or more in the 15N dimension or 0.02 ppm or more in the 1H dimension. Most residues in the WW domains binding to pU24-6A and U24-6A were perturbed similarly. U24-7, on the other hand, yielded a wider range of chemical shift perturbations. Figure 4.12 shows structural models of the rNedd4-WW3/4 and hNedd4-WW3* domains color-coded according to which residues were deemed different according to the criteria given above. In this way, the specific perturbation from one ligand could be highlighted. From Figure 4.12, U24-7 ligand mainly perturbs the region where the second half of the PY motif peptide (middle residue to C-terminus) binds, while perturbations in the β1-β2 loop are mostly 130 introduced by the pU24-6A ligand. The residues around the conserved tryptophan are perturbed differently among three ligands. It is surprising to see that, based on this analysis, most of the unique chemical perturbations are the result of the interaction of U24-7 and not pU24-6A with the WW domain. Interestingly, this analysis suggests that the two WW domains show slightly different patterns of U24 peptide binding. In particular, the phospho-ligand affects different residues on the WW domains. A similar figure showing a 90° rotated view of these two structural models are shown in Figure F.3. Some residues at the back of the PY motif binding site are perturbed specifically by pU24-6A and U24-7. This suggests a change around the residues peripheral to binding site upon ligand binding. Figure 4.12 Structural model of rNedd4-WW3/4 and hNedd4L-WW3* domains, coloured using the color-coding scheme described in the text. Left: Coloured model of rNedd4-WW3/4 and PY motif peptide. (PDB: 1I5H) The WW domain is shown in green and peptide is shown as a grey ribbon. The nitrogen atoms are shown as spheres and coloured based on the ligand that is perturbed the most. Blue means U24-6A, red is U24-7, green is pU24-6A, yellow is three of them are very different and white is no ligand is special. The N- and C-terminus of the ligand is marked. Right: Coloured model of hNedd4L-WW3* and PY motif peptide. (PDB: 2MPT) The WW domain is shown in purple. Other colour scheme is as in A). 131 The NOE data shown in section 4.2.7 suggests most of the loop regions are more rigid upon ligand binding, and different ligands prefer different kinds of WW domains. pU24-6A clearly has a bigger impact on hNedd4L-WW3* than rNedd4-WW3/4. The number of NOE constraints increased around the β1-β2 loop region in hNedd4L-WW3*, implying pU24-6A may limit the motion of this region. pU24-6A also caused three β sheets to come closer around the β2-β3 loop end, the binding site for the tyrosine in the PY motif. It is possible that this β sandwich is getting more compact with one ligand bound in it. U24-7 seems to affect rNedd4-WW3/4 domain more, but this could be due to the fact that pU24-6A does not prefer the β1-β2 loop in this domain. More unique NOEs in all loop regions are detected when rNedd4-WW3/4 in complex with U24-7, than pU24-6A. EXSY was used to further investigate the pU24-6A/hNedd4L-WW3* complex. The cross-peaks intensities build up equally for most residues, whereas for E484 and E506 the build-up occurs earlier (Figure 4.11), suggesting these residues are special. Indeed, these glutamic acids are right next to the conserved two tryptophans, which makes them of structural and functional importance. In order to explain the special nature of E484 and E506, a more in-depth investigation on this slow chemical exchange system is required. The preliminary fitting based on one set of EXSY spectra is not very reliable and the origin of the huge data variation observed could be the fitting of the AB peak. Two main reasons of how the experiments were set up might have caused this. The first one is the concentration used for the experiments (0.25 mM). Apo hNedd4L-WW3* was found to be extremely unstable after condensation and could precipitate out of solution at room temperature at concentrations higher than 0.3 mM. The low concentrations required mean that the intensities integrated from the HSQC spectra may not be reliable. The second reason is that the sample was prepared as hNedd4L-WW3*:pU24-6A=1:0.6, in which case pA<pB. This introduces a difference between the magnetizations of apo and bound forms at T=0 ms. The normalized intensities of one set of peaks (aAA(T), aAB(T), aBA(T) and aBB(T)) are fitted to Equation (1.4) and they are normalized by dividing MAA(0) or MBB(0). (289, 291, 293) This means that the calculated normalized intensities contain the errors from two terms. Specifically for the AB peak, both MAB(T) and MBB(0) can 132 contribute to the error. In addition, it would appear that the relaxation rate of the apo and bound forms are very different, R1A0 ≠R1B0 , since the fits obtained with the condition, R1A0 =R1B0 , are worse than those without. This may be caused by the structural instability of the apo hNedd4L-WW3* domain, or the slight differences between the populations of the A and B forms. The spectra obtained at 25 °C show similar results, except that there are more noises in the spectra (Figure C.3). What is more, the fraction between AA peaks and BB peaks are different from that in the spectra obtained at 15 °C. Because of lacking the intensities of AA peaks, the data at 25 °C is very hard to be fitted. In order to get a well-fitted data set, a more concentrated sample (ca. 1 mM) should be prepared, (289) and additional experiments need to be done on the apo and bound form protein to measure their longitudinal relaxation rates. Figure 4.13 A close-up view of hNedd4L-WW3* domain model bound with pU24-6A peptides. The structural model was generated from model 9 of 2MPT and the peptide was mutated to pU24-6A using the mutagenesis tool in Pymol. The phosphoryl group on the ligand is generated using plugin in Pymol, PyTMs. (377) The positions of the residues are optimized using the sculpting function. The hNedd4L-WW3* domain is shown in blue and the peptide backbone is shown in grey. The PY motif is shown in yellow. N490, R492, W483, W505 are shown as sticks and phospho-threonine is shown as sticks as well. The sticks are coloured by elements and the carbons of phospho-theronine are shown in pink, of N490 and R492 are shown in blue, of W483 and W505 are shown in green. 133 The observed high affinities of hNedd4L-WW3* domain with pU24-6A and U24-7 are considered to be due to the negative charge upstream from the PY motif, even though there is no strong electrostatic effect observed for the interactions. Earlier conclusions regarding the flanking residues around the PY motif were derived from studies performed on the YAP65 WW domain, (173, 174) and this observed enhancement with the negative charge from the PY motifs was limited to the WW domains investigated, namely rNedd4-WW3/4 and hNedd4L-WW3* domains. From the electrostatic effects studies, it suggests that the flanking residues may participate in the interaction with hNedd4L-WW3* domains. As a neurite outgrowth inhibitor, Nogo-A has a PY motif ligand, in the segment IKHEPENPPPYEEAM. This ligand has two glutamic acids upstream from PY motif, at the -1 and -3 position. The Kd determined for Nogo-A peptide and hNedd4L-WW3* is 3.9 ± 0.2 μM at 25 °C in 10 mM sodium phosphate, pH 7.4. This interaction is not as strong as hNedd4L-WW3* with pU24-6A, or with U24-7 peptide, suggesting these two glutamic acids may not contribute much to the interaction. The question which remains then is why this happens? One possibility is that the interaction in the pU24-6A/hNedd4L-WW3* complex is not a purely electrostatic interaction. A simple addition of one more charged residue, as in the Nogo-A peptide, does not improve binding. Other reasons may include a geometrical preference of salt bridge formation among charged residues. (353) The phosphoryl group might have similar preferences with glutamic acid and aspartic acid. This could explain the small dissociation constants observed for both pU24-6A and U24-7. Not only does the phosphoryl group elongate the side chain of the threonine, it also contains two negative charges in a pH 7.4 buffer, providing a large polar interface to interact with the WW domain. As expected, more chemical shift perturbations by pU24-6A can be observed around β1-β2 loop region, where the N-terminal region of the ligand is bound. N490 and R492 are two potential residues that could form complex hydrogen bonds/salt bridges with pU24-6A. The side chains of these residues, guanidine or side chain amide group, may contribute to the interaction with the phosphoryl group in pU24-6A (Figure 4.13). They are in the same region and near the second conserved W505 in the WW domain, the binding site of PY motif. They could provide either a positive charge, or a hydrogen to bond with a negative or highly electronegative 134 oxygen of the phosphoryl group. Evidence to support this could be found in the resonances of the N490 side chain amide. The distance between two peaks that generated by the side chain amide was measured to be 2.6 ppm in the pU24-6A/hNedd4L-WW3* complex, while it was only 1.7 or 2.0 ppm when bound to U24-7 or U24-6A, respectively. This suggests that the phosphoryl group can affect the side chain amide group differently. Indeed, both N490 and R492 have been identified as a binding site for phospho-serine in a triphosphorylated Smad3 peptide. (215) Smad will be introduced in Chapter 5. The dissociation constant between triphosphorylated Smad3 and hNedd4L-WW3* is in the range of high nanomolar, (215) similar to that of pU24-6A. This strongly suggests that these two residues could be key to the phospho-ligand preference of hNedd4L-WW3*. Even though it is a very weak electrostatic interaction, an enhancement in affinities could be observed with pU24-6A, when compared to U24-6A. This is possibly due to an increased value of kon, which could be confirmed using SPR, as was done for a PY and phospho-PY ligand shown to have a similar koff when binding to a WW domain, but different kon. (376) In the case of U24-7, there is a glutamic acid upstream from the PY motif, which might have a similar role as the phosphoryl group. This negative charge is closer the backbone of the peptide, but further away from both N490 and R492 in order to form complex H-bond/salt bridges. It might only have a long-range electrostatic interaction with R492, which explains the SKa calculated above. Overall, the phosphorylation or negative charged residue upstream from PY motif can enhance the affinities with hNedd4L-WW3* domain. 4.4 Conclusions Earlier studies on the function of U24 found that it can down-regulate TCR/CD3 in T-cells and that the PY motif is crucial for this function. This could have broad implications for the biology as U24 would be targeting a process, and not a specific protein. (36) The purpose of expressing U24 in vivo could be latency maintenance and immune response modulation, in order to ensure virus survival. 135 In this chapter, a group of potential binding partners of U24, Nedd4-WW domains, were investigated. Both versions of U24 can bind to human Nedd4 and Nedd4L WW3* domain strongly in vitro and pU24-6A ligand can bind to these WW domains at an even higher affinity. This suggests that U24 could affect Nedd4 via its WW domain in vivo. Whether Nedd4 is the protein affected during TCR/CD3 down-modulation cannot be confirmed using the studies above. Further in vivo experiments would be required to test this. 4.5 Materials and methods 4.5.1 Plasmid construction of GST fusion human WW3* domains Two cDNA clones containing human E3 ubiquitin-protein ligase Nedd4 (KIAA0093) and Nedd4-like protein (KIAA0439) were ordered from Kazusa DNA research institute (Chiba, Japan). The third WW domain of both proteins was amplified by PCR using the primers listed in Table 4.5 and Phusion DNA polymerase. The PCR products were analysed and purified using agarose gel electrophoresis and QIAquick Gel Extraction Kit and then were amplified again with a pair of shorter primers for higher yield. The amplification and purification were done twice more to increase the purity of the products. Table 4.5 Primers used for plasmid construction of human WW3* domains Name Primers from 5’ to 3’ KIAA0093_BamHI_F AATAGGATCCATTGAGCAAGGATTCCTTC KIAA0093_XhoI_R CAATAATTTACTCGAGCTACAGATGGGCTGG KIAA0439_BamHI_F AATAGGATCCGTCACACAGAGCTTCTTG KIAA0439_XhoI_R CAATAATTTACTCGAGCTACATATGTACTGG KIAA0093_Short_F GTTAAATAGGATCCATTGAG KIAA0439_Short_F GTTAAATAGGATCCGTCACAC Universal_Short_R CAATAATTTACTCGAGCTAC pGEX 5’ GGGCTGGCAAGCCACGTTTGGTG pGEX 3’ CCGGGAGCTGCATGTGTCAGAGG 136 The PCR products were inserted into BamHI and XhoI sites, C-terminal to the Glutathione S-transferase tag (GST tag) protein, into the pGEX-4T2 and pGEX-6P1 (GE Healthcare) vectors, using the method described in section 2.5.2. pGEX 5’ and 3’ primers were used for colony PCR. The cloned plasmids were submitted to NAPS unit to confirm the insertion of WW3* domain genes. 4.5.2 Expression of GST fusion WW domains The plasmids containing GST fusion proteins of the second and third WW domains (WW2 and WW3/4) of rNedd4 were kind gifts from Dr. Julie D. Forman-Kay (Hospital for Sick Children, Toronto). The plasmids containing GST fusion proteins of two WW3* domains were constructed as described above. The pGEX-4T2 versions of both constructs were used. All plasmids were transformed into E. coli BL21(DE3) for expression. The 5 mL starting LB-CBC culture was grown at 37 °C for 5 to 6 hours and then was used to inoculate 800 mL fresh LB-amp. The cultures were then grown till OD600 reached 0.5, and were cooled down in the cold room for 5 minutes before a induction period of 16 hours at 25 °C. A final concentration of 400 μM IPTG was found to be optimal for protein expression. Expression tests were carried out separately using the protocol described in section 2.5.3. 300 μL samples from before and after induction were pelleted for electrophoresis analysis to confirm protein expression. The cells were harvested by centrifugation as described in section 2.5.4 and the cell paste was suspended in 30 mL PBS buffer. Then, the cells were transferred to three 15 mL centrifuge tubes and centrifuged again. Each tube contained cell paste from approximate 300 mL culture, and was stored in -80 °C freezer until further use. 15N labelled protein was expressed as described in section 2.5.4 and induced with 400 μM IPTG for 16 hours at 25 °C. 800 mL culture was grown. The cells were harvested by centrifugation as described in section 2.5.4, suspended using 20 mL PBS, divided into two 15 mL centrifuge tubes and centrifuged again. Each tube contained the cell pellet from approximate 400 mL culture, and was frozen and stored in -80 °C freezer until further use. 137 4.5.3 Purification of GST tagged Nedd4 WW domains The cell paste from approximate 600 mL culture was thawed on ice and resuspended in PBS/Triton buffer (PBS buffer supplemented with 1% Triton X-100). Lysozyme, protease inhibitor and DNAse were added in the mixture. The mixture was incubated on ice for half hour, then was lysed in an ice bath by sonication (5 second pulses were used except 2 second pulses for GST-hNedd4L-WW3* domain) for 3 minutes. After being centrifuged at 9,000 rpm for 1 hour, the supernatant was filtered using a 0.45 μm filter, and then applied to 2 mL of PBS buffer washed Glutathione Sepharose 4B resin (GE Healthcare, GST 4B resin), loaded into a 15 mL centrifuge tube. An end-over-end rotator (SARSTEDT) was used to mix the resin. The resin slurry with cell lysate was incubated for 2 hours at room temperature, and then washed with PBS/Triton buffer three times. The resin was collected by centrifugation at 600 × g for 5 minutes. A Wash4 buffer (0.5 mM Glutathione reduced in PBS buffer) was used to wash off some impurities. The sample was washed further using three washes of PBS/Triton buffer. 2 mL of GST elution buffer (50 mM Tris-HCl, 10 mM reduced glutathione, 1% Triton X-100, pH 8.0) was added into the mixture and the sample was mixed for a further half an hour using the end-over-end rotator. 3 mL of GST elution buffer was used to wash the resin two more times. The eluted fusion protein was dialyzed again 2 L of pull down buffer (20 mM KH2PO4, 75 mM NaCl, 0.5 % Triton X-100, pH 7.4) twice overnight, in order to remove all of the reduced glutathione bound by GST tag. 10 μL of each fraction was collected and analyzed with Tris-Glycine SDS-PAGE. After the overnight dialysis, the GST fusion WW domain was filtered using a 0.45 μm filter. The concentration of the sample was determined by BCA protein assay (Pierce). 4.5.4 Purification of Nedd4-WW domains The proteins were extracted and immobilized on GST 4B resin, as described above. After the washing with Wash4 buffer, the resin was washed three times with PBS buffer. 3 mL PBS buffer and 150 units of thrombin (GE Healthcare) were added into the pelleted resin and the slurry was incubated at room temperature for 16 hours. 138 The supernatant containing cleaved WW domains were collected on the second day by centrifugation. 4 mL of PBS buffer was used to wash the resin 2 more times. 200 μL of p-Aminobenzamidine-agarose (Sigma-Aldrich) was washed with PBS buffer and then added into the combined mixture of WW domain containing factions. This resin mixture was incubated at 4 °C for 2 hours, and the resin was pelleted by centrifugation at 3,900 rpm for 10 minutes. 100 μL of PBS washed GST 4B resin was used to trap residual GST protein in the WW domain fractions. 10 μL of each fraction were collected and analyzed with Tris-Glycine and Tris-Tricine SDS-PAGE. The isolated WW domains were dialyzed to water and lyophilized before submitting them to the Mass Spectrometry Centre to confirm the mass of each WW domain using MALDI-TOF MS. Please see appendix B6 for spectra. The WW domain contained fraction was dialyzed against 2 L phosphate buffer (10 mM sodium phosphate, pH 7.4) using 2,000 MWCO dialysis membrane overnight for ITC or NMR experiments, or phosphate buffer with various salt concentrations (phosphate buffer with 100 mM or 500 mM NaCl). 4.5.5 CD melt experiment The stabilities of the Nedd4-WW domains at different temperatures were determined by CD melt experiments. The protein samples were prepared as described in section 4.5.4 and diluted to 50 μM using phosphate buffer. The spectra were recorded using a J-815 CD spectrometer (Jasco) flushed with nitrogen gas. A quartz cuvette with path length of 0.1 cm was loaded with 300 to 400 μL protein samples and its top was sealed with parafilm. A Peltier type CD/FL cell holder connected with thermo regulator was used. Samples were scanned at a rate of 100 nm/min with a step size of 0.1 nm, from 190 nm to 250 nm. All spectra were recorded after a temperature equilibration period of 10 seconds, every five degrees from 5 °C to 95 °C. Spectra were averaged over three scans and the sample-free buffer was scanned with the same program for background subtraction. An example result from a CD melt experiment is shown in Figure D.2. 139 4.5.6 GST tagged WW domains pull-down assays with recombinant U24-6A and U24-7 protein The GST-Nedd4 WW domains were prepared as described in sections 3.5.1 and 4.5.3, and approximate 1 to 2 nmole of protein was immobilized on 20 μL GST 4B resin individually. U24-6A and U24-7 were prepared as described in section 2.5.5. (298) The resin was incubated with 600 μL U24-6A and U24-7 protein stock solution in pull down buffer at 4 °C. U24-7 stock solution was supplemented with 1 mM DTT. After 1 hour of incubation, the resin was washed three times in order to remove the free U24 protein from the system. Before being mixed with 2X BME gel loading buffer, the resin was washed using GLB for a last time. The samples were heated at 95 °C for 5 minutes and then were centrifuged. 3 to 10 μL of the supernatant was removed, and loaded onto a gel that was run using a Tris-Tricine buffer system. The gels were stained with silver stain according to standard protocols. 4.5.7 ITC experiments After overnight dialysis, the cleaved Nedd4 WW domains in phosphate buffer, with different salt concentrations, were ready for use in ITC experiments. The dialysis buffer was kept as a sample-free buffer to measure the heats of dilution. If the proteins were not deemed to be sufficiently pure based on the SDS-PAGE results, another 100 μL of washed GST 4B resin would be used to immobilize the GST tagged impurities. The sample was then concentrated to 40 to 150 μM, depending on the estimated Kd value, using Microsep centrifugal devices with a MWCO of 1 kDa. The concentration of the sample was determined by the absorbance at 280 nm in 6 M Guanidine HCl on a UV-Vis spectrophotometer, and calculated using the theoretical extinction coefficient obtained from the ProtParam tool. A 15x to 30x concentrated peptide stock solution was made by dissolving the purified peptides, U24-6A, pU24-6A and U24-7, in the same buffer as the protein and its pH was adjusted to match the buffer within 0.05. The peptide amounts were determined gravimetrically and calibrated according to the absorbance at 280 nm in 6M Guanidine HCl. Both protein and peptide stock solutions were filtered and degassed before loading into the sample cell 140 and injection syringe, respectively. ITC experiments were performed on a VP-ITC MicroCalorimeter (MicroCal) at 15 °C, 25 °C and 37 °C. The titration protocol was comprised of a preliminary injection of 2 μL of the peptide solution, followed by 20 or 25 consecutive 10 μL injections into the sample cell (1.4 mL) containing the Nedd4-WW domain studied. The time between each injection was 300 seconds. Control titrations of peptide solution into protein-free buffers were also conducted at different temperatures. The heats of dilution for the peptide were subtracted from the original heats prior to data fitting to a one-to-one bimolecular interaction model to obtain Ka and ∆H. The ITC experiments were repeated three times with mean values and standard deviations reported. The changes of heat capacities upon binding were calculated using the linear fit of the enthalpy at three different temperatures. The experiments of buffer titrated into a Nedd4-WW domain were carried out at different conditions to monitor the state of the protein during titration, and a small aliquot of sample taken before and after titration was analyzed on SDS-PAGE to confirm there was no degradation. During sample preparation, rNedd4-WW2 and hNedd4-WW3* were found not to be stable. In our rNedd4-WW2, there is a linker downstream from the WW2 domain which was found to be degraded. The WW2 domain itself was however folded (Figure B.11). It was not concentrated using ultra centrifugal tubes to prevent further degradation. In the case of hNedd4-WW3*, very poor heat stabilities were found for the apo form and the purified protein (Figure C.2) cannot be stored at 4 °C for more than 36 hours. 4.5.8 1H-15N NMR experiments of WW domains NMR experiments were conducted on 15N labelled rNedd4-WW3/4, hNedd4-WW3* and hNedd4L-WW3* domains. All three domains were handled in the same manner, so they are denoted as WW3 domain in this paragraph for simplification. Starting from purified protein, as described in section 4.5.4, 15N labelled WW3 domain protein was prepared by overnight dialysis against 10 mM sodium phosphate, pH 7.4. The protein sample was concentrated using a Microsep centrifugal device with a MWCO of 1 kDa. Protein solutions were centrifuged several times until a final 141 concentration of about 0.4 to 0.6 mM in 10 mM sodium phosphate, pH 7.4 was reached. The sample was then supplemented with 10% D2O, 0.5 mM benzamidine, 0.1% sodium azide. The concentration determination and buffer matching of the WW3 domains and three peptide stock solutions was conducted as in section 4.5.7. NMR experiments on rNedd4-WW3/4 and hNedd4L-WW3* were performed at 25 °C using a Bruker Avance III 850 MHz NMR spectrometer (Milton, Ontario, Canada), equipped with a TCI probe. 1H-15N HSQC spectra of apo protein, in a 5 mm NMR tube, were recorded. Then, the U24 peptide stock solution, normally 3-8 mM, was added into the NMR sample in small aliquots and additions were stopped when there were no additional chemical shift changes observed. The 1H-15N HSQC of six U24/WW domain complexes, three U24 ligands with rNedd4-WW3/4 and hNedd4L-WW3*, were recorded using the same program as used above for the apo protein. The data sets were processed using NMRpipe, (316) and visualized using Sparky. (317) The resulting spectra of apo and bound protein were assigned using additional 3D HSQC-NOESY and 3D HSQC-TOCSY spectra. A table listing the assignment of all residues can be found in appendix. The chemical shift changes of the amide protons, ΔδHN in the proton dimension and ΔδN in the nitrogen dimension, were calculated according to Equation (4.5). The Δδcomb was calculated (using Equations (4.6) and (4.7)) (356, 378) for each residue to represent the chemical shift perturbations upon U24 ligand binding. 3D HSQC-NOESY experiments were performed on six U24/WW domain complexes, with the sample loaded in a Shigemi tube. The pulse program used was noesyhsqcf3gpsi3d, (280–282) which was edited by Dr. Mark Okon to improve phase cycling, pulse gradients and the selective water flip-back. The NOESY mixing time d8 for all spectra was set to 200 ms and d1=1 s for hNedd4L-WW3* or 1.5 s for rNedd4-WW3/4 domain related experiments. 3D HSQC-TOCSY experiments were set up for assignment and resonance identification using the pulse program dipsihsqcf3gpsi3d. (280–282) Dr. Mark Okon optimized the pulse gradients and water suppression. The 3D data was processed and visualized using NMRpipe and Sparky. (316, 317) For the NOESY data, the strip plots for each amide resonance were examined and the peaks with different intensities were analyzed. The volumes of the NOE signals were integrated and converted 142 to a distance value using Equation (4.8). The rij indicates the distance between proton i and proton j, with a reference distance, rref=2.8 Å, used here for the side chain amide proton ε1 and ζ2 proton of the first conserved tryptophan, W465 (rNedd4-WW3/4) and W483 (hNedd4L-WW3*). About 60 NOE signals were analyzed for each WW domain. Δδ=δbound - δapo (4.5) Δδcomb=[ΔδHN2 +(ΔδN Rscale⁄ )2]12⁄ (4.6) Rscale=〈σδ〉HN〈σδ〉N=√120∑ ∆δHN220k=1√119∑ ∆δN219k=1 (4.7) rij=rref (aref/aij)1/6 (4.8) To determine the thermal stability of the WW3* domains, 1H-15N HSQC spectra of apo hNedd4-WW3* and hNedd4L-WW3* were recorded at various temperature using a Bruker Avance III 600 MHz NMR spectrometer (Milton, Ontario, Canada), equipped with a TCI probe. The spectra were recorded at 15 °C to 40 °C after 10 minutes of temperature equilibration, shimming, matching, tuning and pulse calibration. The data were processed and visualized as above. The EXSY or ZZ exchange experiments were carried out using an unsaturated sample, hNedd4L-WW3*:pU24-6A=1:0.6 that was prepared using the methods mentioned in section 4.5.4 and as above. The pulse program used was hsqcetgpspf3 with a d21 inserted between the evolution of 15N magnetization and reverse INEPT by Dr. Mark Okon. (288) An additional water pre-saturation pulse was added to avoid an echo problem during data acquisition (-Dpresat in zgoption). Spectra were recorded at 5 °C, 15 °C and 25 °C. The exchanged peak at 5 °C takes very long time to build-up therefore no data was shown. The data were processed and visualized as above. 143 Chapter 5 Chapter 5 Exploration of the Potential Interactions between U24 and the Human Smurf2 WW Domains† 5.1 Introduction As mentioned in Chapter 1, there are several members in the Nedd4 family E3 ubiquitin ligase. Smurf, Smad ubiquitination regulatory factor, is studied in this chapter. There are two highly related but independent Smurf proteins, Smurf1 and Smurf2. They share 80% sequence identity and have a similar architecture to Nedd4 protein, namely a C2 domain, two or three WW domains and a HECT domain. Via their WW domains, Smurf proteins target Smad proteins or Smad complexes for proteosomal degradation. (379, 380) Smad proteins are the central players of the classic TGF-β signalling pathway and can be translocated into the nucleus, where they act as transcription factors. There are eight Smad proteins, divided into three classes depending on their specific role in the pathway: receptor-regulated Smad (R-Smad), common-mediator Smad (Co-Smad) and inhibitory Smad (I-Smad). (381) I-Smad, including Smad6 and Smad7, negatively regulates the signals of TGF-β. This regulation is achieved via Smurf E3 ubiquitin ligase. Smad7 acts as the adaptor for Smurf to recruit activated TGF-β receptor, which will result in the degradation of both Smad7 and TGF-β receptor. (382) This process is mediated by the interaction between the Smurf-WW domain and the PY motif in Smad7. (382, 383) U24, which also contains a PY motif, may interfere with the † A version of this chapters will be published as Sang Y, Zhang R, Scott WRP, Creagh AL, Haynes CA, Straus SK. 2016. Exploration of the Potential Interactions between U24 and the Human Smurf2 WW Domains. 144 process by binding to Smurf-WW domains. The PY motifs in U24 and Smad7 are aligned in Figure 5.1. Position Numbering PPx Y -6 -5 -4 -3 -2 -10 1 2 34 567 89 10 11 12 U24-6A MD PP RTP PPS Y SE V-L U24-7 M- TH ETP PPS Y ND VML Smad7 C EL ESP PPP Y SR YPM DFL Figure 5.1 Sequence alignment of first 15 residues of U24-6A, U24-7 and the core PY motif of Smad7. The positively and negatively charged residues are marked in blue and red, respectively. The TGF-β signalling pathway is vital for embryo development and maintaining cytostatic and apoptotic actions in mature tissue. (384) For more details regarding this pathway, please consult the following excellent review articles. (381, 384–386) At a cellular level, TGF-β is involved in many activities, i.e. cell division, differentiation and apoptosis. As a result, it promotes tissue growth and cell differentiation in embryos, but it can also have a negative effect on cell growth in mature tissues. (381, 384) It is therefore not surprising that cancer cells involve mutant proteins from this pathway, e.g. inactive TGF-β type II receptors have been identified in colon cancer cells. (387) TGF-β is important in maintaining cytostatic and immune self-tolerance. It has been previously observed that the level of TGF-β isoforms and TGF-β receptors is increased in the lesion area of MS patients. (388) There is evidence that TGF-β signalling induces the development of T-helper-17 (TH17) lineage, a group of T-cells that is required to initiate the autoimmune of EAE. (389–391) However, TGF-β also shows a protective function in MS models as well. It can delay the onset of EAE, enhance myelin formation via cellular signalling, and reduce the demyelination in a viral-induced MS model. (392–395) As TGF-β is involved in a variety of cellular process, interruption of its signalling pathway might cause aberrant behaviour of the immune system, leading to autoimmune diseases. U24 could interfere with the TGF-β signalling pathway through its negative regulator, the Smurf proteins. 145 Table 5.1 Alignment of canonical and uncanonical WW domains. The conserved tryptophans, or substituted tyrosine or phenylalanines, are highlighted in bold. WW domain Sequence hNedd4L-WW3* FLPPGWEMRIAPNGRPFFIDHNTKTTTWEDPRLKF hNedd4L-WW4 PLPPGWEERIHLDGRTFYIDHNSKITQWEDPRLQN hSmurf1-WW1 ELPEGYEQRTTVQGQVYFLHTQTGVSTWHDPRIPS hSmurf1-WW2 PLPPGWEVRSTVSGRIYFVDHNNRTTQFTDPRLHH hSmurf2-WW1 DLPDGWEERRTASGRIQYLNHITRTTQWERPTRPA hSmurf2-WW2 DLPEGYEQRTTQQGQVYFLHTQTGVSTWHDPRVPD hSmurf2-WW3 PLPPGWEIRNTATGRVYFVDHNNRTTQFTDPRLSA Although U24 shows strong affinities with Nedd4-WW domains (Chapter 4), it is not a given that binding to Smurf-WW domains will be equally strong. Smurf-WW domains are uncanonical WW domains. Some Smurf-WW domains, such as Smurf2-WW1, do not recognize PY motifs, even though they have both conserved tryptophans. Other Smurf-WW domains, unlike the Nedd4-WW domains, only contain one tryptophan in each domain, with the other conserved site being replaced by either a tyrosine or phenylalanine (see Table 5.1 for alignment). The first conserved tryptophan is critical for protein stability. If it is changed to any other residues, the WW domain will not maintain a stable folded structure, and hence will not bind to a PY ligand. (212, 396) Indeed, Smurf1-WW1 and Smurf2-WW2 cannot recognize PY motifs on their own (Table 5.1). (397) Replacing the second conserved tryptophan with phenylalanine or tyrosine typically results in lower affinities with the PY motif ligands. (212, 396, 398) However, these WW domains do not necessarily bind more weakly than their Nedd4-WW domain counterparts, as cooperativity plays an important role in the binding interaction. Smurf-WW domains that work in pairs have extremely high affinity with one PY motif ligand, normally in Smad proteins. Tandem WW domains in Nedd4 protein can bind PY motif ligands as well, but normally bind one ligand each or two ends of an extended ligand. (215, 359) Even though Smurf1-WW1 and Smurf2-WW2 do not bind to PY motif by themselves, they could enhance the affinities when linked with another WW domain. For instance, the coupled WW2+3 domain (i.e. including WW2-linker-WW3) from hSmurf2 binds to the Smad PY motif approximately 146 three to ten times stronger than hSmurf2-WW3 domain alone. (212, 359) Cooperativity enhances the selectivity, thereby making the signalling pathway of TGF-β more tightly regulated. In this chapter, the ability of U24 to interact with human Smurf2-WW domains was evaluated. The affinities of the interactions between U24 and hSmurf2-WW domains were determined using ITC. The associated heat capacities for these interactions were investigated as well. Rui Zhang, a former graduate student in the lab, previously characterized these interactions by NMR. (399) The results obtained here are compared to those obtained for the Nedd4-WW domains, discussed in Chapter 4. Since the exact function of U24 is unknown and since its tissue localization is unknown, it will be interesting to compare binding affinities between these two types of WW domains, to try to better understand the biological role of U24. 5.2 Results 5.2.1 Pull-down assays with GST-hSmurf2-WW domains and U24 proteins In order to determine if U24 proteins can bind to hSmurf2-WW domains, GST pull-down assays using GST-hSmurf2-WW domains were conducted. Purified recombinant U24-6A and U24-7 were used, as in Chapter 4. The hSmurf2-WW domains investigated were the WW2 domain on its own, the WW3 domain on its own, and a tandem WW2-WW3 domain (WW2+3 domain). Since previous work had shown that hSmurf2-WW2 could not bind to PY ligands on its own, it was expected that there might not be interaction with U24 protein for this WW domain. (212) However, the cooperativity observed between the hSmurf2-WW2 and WW3 domains suggested that the best binding might be expected to occur between U24 and the construct with the tandem domains, hSmurf2-WW2+3. The pull-down experiments were performed as described in sections 4.5.6 and 5.5.3. 147 Figure 5.2 SDS-PAGE result of GST pull-down experiment with GST-hSmurf2-WW domains and U24 protein. Pull-down experiment using GST-hSmurf2-WW2, GST-hSmurf2-WW3, GST-hSmurf2-WW2+3 and the two versions of U24. U24-6A and U24-7 are loaded in the first lane as references, and the pull-down of GST alone and U24 are loaded in the second lane. Lanes 3, 5 and 7 are loaded with GST-hSmurf2-WW domains alone. Lanes 4, 6, and 8 are from the pull-down of U24 and GST-hSmurf2-WW2, GST-hSmurf2-WW3, GST-hSmurf2-WW2+3, respectively. The marker lanes are indicated as lane M. The corresponding molecular weights of the marker and the positions of the GST and GST-Nedd4-WW domains are marked on the right side of the gel. The input of GST tagged proteins shown below the results indicated that the proteins were loaded equally. The pull-down results, analyzed using SDS-PAGE, are shown in Figure 5.2. The inputs of the GST tagged proteins are shown below the pull-down results to demonstrate that a uniform quantity of protein was used for each condition tested (i.e. in each lane). GST-hSmurf2-WW2+3 pull-down experiments were carried out with supplemented DTT to increase the binding fraction of this fusion protein to the GST 4B Sepharose resin. In lanes 6 and 8, there are bands found at the molecular weights of U24-6A and U24-7, indicating that both hSmurf2-WW3 and hSmurf2-WW2+3 interact with U24. Clearly, and perhaps not surprisingly, there was no band observed for U24-6A nor U24-7 in lane 4, meaning that hSmurf2-WW2 does not bind to either U24 protein. As in section 4.2.1, 148 there were some minor impurities detected in the input section. This may be due to the potential degradation of the GST fusion protein and to the inherent sensitivity of silver staining. As mentioned before, the intensity of the bands cannot be used to directly assess the strength of the binding interaction. 5.2.2 Affinities between hSmurf2-WW domains and U24 peptides To determine the exact binding affinities between U24 and hSmurf2-WW3 and hSmurf2-WW2+3, ITC experiments were conducted. In particular, these experiments were carried out to see whether hSmurf2-WW2 can enhance the affinities between U24 and hSmurf2-WW domains, namely whether hSmurf2-WW2+3 binds better than hSmurf2-WW3, since the GST pull-down results were not conclusive in this regard. As in Chapter 4, three 15-mer peptides representing U24-6A, pU24-6A and U24-7 were titrated into protein solutions containing hSmurf2-WW3 or hSmurf2-WW2+3 using an iTC200 MicroCalorimeter. An experiment with hSmurf2-WW2 was also carried out, but no endo- or exothermic events were detected except for the heat of dilution, as anticipated from the pull-down for this construct. These experiments were carried out in 40 mM HEPES, 10 mM NaCl, pH 7.2, at 25 °C (section 5.5.4). Phosphate-based buffer was not used (as in Chapter 4) because apo hSmurf2-WW2+3 domain is not stable in this buffer. (399) Although hSmurf2-WW3 is stable in phosphate buffer, this sample condition was also discarded because the heats generated in U24-6A/hSmurf2-WW3 titration in this buffer were too small to yield reliable fits (data not shown). Control experiments in which buffer was titrated into the protein solution were also conducted to ensure that hSmurf2-WW3 and hSmurf2-WW2+3 were stable during the titration. The titration data were fitted using “one set of sites” model and all parameters were allowed to vary during data fitting. The mean values of the fitted thermodynamic parameters from three ITC experiments from each pair of titrations are reported in Table 5.2. As seen in Table 5.2, there is no significant enhancement in the affinities with hSmurf2-WW2+3 as compared to hSmurf2-WW3, regardless of which ligand was used. These interactions are in the range of 50 to 300 μM, weaker than those with Nedd4-WW domains. Remarkably, pU24- 149 6A still binds the best among the three ligands, followed by U24-7 and U24-6A. The affinities with pU24-6A are nearly five times better than its non-phosphorylated counterpart, U24-6A. The ΔH values obtained for the interactions with the hSmurf2-WW domains are also much smaller than those for the Nedd4-WW domains. The weaker binding observed may be due to the fact that the missing tryptophan in hSmurf2-WW domains leads to a reduction in the hydrophobic area, required for the interaction. Table 5.2 Parameters obtained from fitting the ITC data for binding of U24-6A, pU24-6A or U24-7 binding to hSmurf2-WW3 and hSmurf-WW2+3 domains at 25 °C. The parameters are obtained from fitting the data from three separate runs and averaging them. The errors represent ± one standard deviations. The protein concentrations used in the titration experiments are about 150 to 300 μM. U24-6A pU24-6A U24-7 Kd (μM) ΔH (kJ/mol) ΔS (J/mol K) Kd (μM) ΔH (kJ/mol) ΔS (J/mol K) Kd (μM) ΔH (kJ/mol) ΔS (J/mol K) hSmurf2-WW3 329±15 -24±3 -14±9 67.8±0.5 -42.0±0.4 -61±1 175±3 -42.8±0.1 -71.7±0.6 hSmurf2-WW2+3 302±5 -30.6±0.9 -35±3 50.3±0.8 -45.3±0.5 -70±2 133±4 -55±1 -110±4 5.2.3 Investigation of the cooperativity in the tandem WW domains of hSmurf2 upon binding to U24 peptides As there is no significant affinity enhancement (e.g. by a factor of 3-10, (212)) observed for binding to hSmurf2-WW2+3 over the hSmurf-WW3 domain alone, the ITC data presented above could not be used to conclusively prove whether the tandem WW domains bind to U24 peptide cooperatively or not. For a certain ligand, i.e. U24 peptide used here, the difference between binding to an individual WW domain and to the tandem WW domains that have two WW domains is that this ligand can be docked into one or two binding pockets at the same time. On the other hand, for an individual WW domain or tandem WW domains that having two WW domains, the binding surface covered upon ligand binding will be one binding pocket, or two binding pockets at the same time. The change in polar or non-polar surfaces of hSmurf2-WW3 and hSmurf2-WW2+3 domains 150 upon U24 ligand binding can be used as an indication whether the WW2+3 can cooperatively bind to U24 ligand, and this value can be obtained from the changes in heat capacities. (372) Consequently, the changes in heat capacities were determined for all combination of U24 peptide/WW domain pairs. As previously described (section 4.2.3), heat capacity data can be derived by conducting ITC experiments at different temperatures. Hence, experiments on the three U24 l igands titrated into hSmurf2-WW3 and hSmurf2-WW2+3 tandem domains were conducted in 40 mM HEPES, 10 mM NaCl, pH 7.2, at 5 °C and 15 °C. The values of the thermodynamic parameters of all titrations are reported in Table E.3. The mean ΔH values averaged from three runs were plotted as a function of temperature, and a least-squares fit was done for each titration. The ΔH values and linear fits are compared for each WW domain (Figure 5.3A-B), or for each U24 peptide (Figure 5.3C-E). The fitted ΔCP values and associated errors, calculated from the linear regression analysis, are given in Table 5.3. Table 5.3 Changes in the heat capacities for binding of U24-6A, pU24-6A or U24-7 to hSmurf2-WW3 or hSmurf2-WW2+3 domains. The error in the slope is obtained from a linear regression analysis. R2 of the fittings are shown in the brackets below the numbers. The protein concentrations used in these experiments are about 100 to 300 μM. ΔCP (kJ/mol K) U24-6A pU24-6A U24-7 hSmurf2-WW3 0.2 ± 0.1 (0.69) -0.67 ± 0.02 (0.97) -1.01 ± 0.02 (0.99) hSmurf2-WW2+3 -0.6 56 ± 0.04 (0.94) -1.35 ± 0.03 (0.97) -2.14 ± 0.04 (0.99) 151 Figure 5.3 Plots of ΔH vs. temperature for the three U24 ligands titrated into A) hSmurf2-WW3 and B) hSmurf2-WW2+3. The same data were plotted again for each peptide, with C) U24-6A, D) pU24-6A or E) U24-7 titrated into the two hSmurf2 domains. ΔH values for the six titrations are plotted as a function of temperature, namely at 5 °C, 15 °C and 25 °C, in A-E). The protein concentrations used in these experiments are about 100-300 μM. The least-squares fit lines of the mean values are shown as well. The results from the six titrations are grouped in terms of the different WW domains in A) hSmurf2-WW3 and B) hSmurf2-WW2+3. They are shown again as groups of different titration ligands in C) U24-6A, D) pU24-6A and E) U24-7. Legends are described at the bottom of the figure. The blue circles represent the ΔH for U24-6A/hSmurf2-WW3, while the black circles are the ΔH values for U24-6A/hSmurf2-WW2+3. The green diamonds are for pU24-6A/hSmurf2-WW3 and the dark green diamonds for pU24-6A/hSmurf2-WW2+3. The pink and purple squares are the ΔH values for U24-7 titrated into hSmurf2-WW3 and hSmurf2-WW2+3, respectively. The error bars represent the standard deviation of ΔH from three runs. 152 As shown in Figure 5.3A, the ΔCP of pU24-6A and U24-7 titrated into hSmurf2-WW3 were both negative and of similar magnitude, while that of U24-6A/hSmurf2-WW3 was associated with quite large errors (Table 5.3). This number may not be representative of ΔCP of binding for this interaction, as the error bars of ΔH at 15 °C and 25 °C are very large. Also, due to sample preparation difficulties, low concentrations were used for these titrations (c values less than 3, see details in section 1.4.3) making the data less reliable. In Figure 5.3B, all the fits were found have negative slopes, including U24-6A. The error bars for the ΔH values were all small, resulting in smaller errors in ΔCP (Table 5.3). In this case, the c values for the titrations of U24-6A/hSmurf2-WW2+3 were also around 3. There linear fits of these data are also associated with errors so these data will only be used as a trend for further analysis The |ΔCP| for the three ligands binding to hSmurf2-WW2+3 ranked from highest to lowest are U24-7 > pU24-6A > U24-6A. This trend could also be found in Table 5.3. In Figure 5.3C-E, these data is recast in terms of the ligand for comparison. In panel C, ΔCP of U24-6A binding to hSmurf2-WW3 and hSmurf2-WW2+3 had different signs, although most of the affinities are similar for both WW domains (Table E.3). In the case of pU24-6A and U24-7, Figure 5.3D-E, steeper slopes were found for hSmurf2-WW2+3 than for hSmurf2-WW3. The actual values show an interesting trend. Indeed, the ΔCP of U24 peptide binding to hSmurf2-WW3 (except for U24-6A) is smaller than the ΔCP when the same peptides bind to hSmurf2-WW2+3. This suggests the interactions with hSmurf2-WW2+3 are associated with larger surface area. Please see discussion part for further details. 5.3 Discussion Although U24 does not bind to hSmurf2-WW domains at a high affinity, it has been previously demonstrated using NMR spectroscopy that U24-6A binds to the typical binding pocket of hSmurf2-WW3 domain, the same binding pocket where the Smad7 peptide binds. (397) The chemical shifts of similar residues in hSmurf2-WW3 domain are perturbed, i.e. T323, V316, H318, and R321. (399) In keeping with these NMR results, the hSmurf2-WW3 and hSmurf2-WW2+3 domains were also found to bind to U24 using ITC. The hSmurf2-WW2 domain on its own did not 153 bind to any of the U24 ligands tested. Generally, the affinities between the three U24 peptides and hSmurf2-WW3 or WW2+3 domains are lower than that of Smad7 peptides and the same Smurf domains. (212, 397) This could be because U24 is missing a tyrosine downstream from the PY motif, which is present in the PY tail of Smad7. The replacement of tryptophan with phenylalanine at the second conserved tryptophan position in WW3 lowers the affinity between the PY motif peptide and WW3 domains, (397) but enables it to be more selective towards its binding partner. From the data shown here, no affinity enhancement was observed when U24 interacted with hSmurf2-WW2+3. One possible reason for this is that U24 does not contain key residues for affinity enhancement, like the tyrosine mentioned above, or the arginine down-stream from the PY motif (Figure 5.1). Overall the findings presented here and those obtained by Rui Zhang both show that U24 does not strongly bind to hSmurf2-WW domains. (399) This suggests that U24 may not strongly interfere with the interactions between Smad7 and hSmurf2-WW domains, and hence not play a significant role in the TGF-β signalling pathway. ∆Aest=∆Anp-0.44∆Ap=∆CP°0.32 (5.1) Although the data presented here showed that the affinities between hSmurf2-WW3 and U24 could not be enhanced by the additional WW2 domain, this does not indicate that the tandem WW2+3 is only binding U24 via the WW3 segment alone. Indeed, the temperature dependence of the U24/hSmurf2-WW domain interactions presented in section 5.2.3 suggest otherwise. Using the measured ΔH at three different temperatures and Equation (4.1), the ΔCP for binding of three U24 peptides binding to WW3 or WW2+3 domains was calculated (Table 5.3). The ΔCP value for the U24-7/hSmurf2-WW3 interaction was found to be similar to that of U24/Nedd4-WW domains, suggesting that a similar non-polar/polar surface is buried/exposed upon binding. In order to get a rough estimate of the relative binding surface, the estimated changed in surface upon U24 ligand binding, ΔAest was calculated according to Equation (5.1), which represents the total buried non-polar surface and nearly half of the exposure of the polar surface. This equation is rearranged from 154 Equation (4.2). Although an accurate non-polar and polar surface area involved in the interaction cannot be calculated, ΔAest could provide a clue for the relative binding surface. The calculated values are listed in Table 5.4. Table 5.4 Estimated binding area of U24 peptide binding to hSmurf2-WW domains, obtained using Equation (5.1). The error listed is propagated from the error associated with the slope fitting for ΔCP. Contrary to Chapter 4, the value of ΔAest is calculated here as a number of discrepancies have been observed using Equation (4.2). (370, 373) These estimates should only be used as relative numbers, i.e. to identify trends. ΔAest (×1000 Å2) U24-6A pU24-6A U24-7 hSmurf2-WW3 0.2 ± 0.1 -0.50 ± 0.08 -0.76 ± 0.06 hSmurf2-WW2+3 -0.4 ± 0.1 -1.0 ± 0.2 -1.6 ± 0.2 A comparison of the numbers found in Table 5.4 shows that for pU24-6A and U24-7, the binding area for the interaction with hSmurf2-WW2+3 is nearly double that for hSmurf2-WW3 alone. The errors that were propagated from the linear fits are quite large, but it is clear that the slope is doubled. This could suggest that roughly twice the number of WW3 binding sites are buried when binding to the U24 peptide in this case. Since the fitted binding site number is n=1 (Table E.3), it is unlikely that two WW3 domains are involved during the interactions with U24 peptide. Another alternative explanation for this doubling of the area could be that WW2+3 binds U24 cooperatively, without any enhancement in affinities. This is completely different from the well-studied model of Smad7 peptide binding to hSmurf2-WW3, or hSmurf2-WW2+3 domain. There is a distinct affinity enhancement for the tandem WW domains and there are clues to show that WW2 and the linker region are involved in the interaction. (212) Using NMR spectroscopy, more perturbations of the residues in the WW2 and linker part and different perturbations of the residues in WW3 part were observed. (212) When using U24-6A as a ligand, slightly different chemical shift perturbations of the residues in the hSmurf2-WW3 domain alone was observed as in the hSmurf2-WW2+3 domain. 155 (399) In this case, it is hard to conclude whether there is cooperativity between the two WW domains in hSmurf2-WW2+3 upon U24-6A binding. The ΔCP of hSmurf2-WW3 interact with U24-6A peptide cannot be accurately determined, therefore it cannot be used to compare with the ΔCP of hSmurf2-WW2+3. In fact, it is an indication for forming stable complexes in the case of hSmurf2-WW2+3 binding to U24-6A because its ΔCP can be determined using ITC even the affinities are in the same range with hSmurf2-WW3. There could be some degrees of increased binding area when U24-6A binds to tandem WW domains while no other evidence has been found to support this. Nevertheless, the additional phosphoryl group in pU24-6A probably enables an increase in binding affinity which leads to a doubling in the binding surface compared to hSmurf2-WW3 domain. This suggests that this additional phosphoryl group on U24-6A, which is essentially an additional negative charged “side chain” upstream from PY motif, is important for the high affinities with hSmurf2-WW domains. That the negative charge upstream from PY motif, may be the key to the high affinities and cooperativity of hSmurf2-WW2+3, is further supported by the similar affinity enhancement and doubling of the binding surface observed in U24-7 ligand, which is not phosphorylated but contains a glutamic acid residue upstream from the PY motif. This glutamic acid may hold the key to bridge the two WW domains in hSmurf2-WW2+3 together. Not surprisingly, the glutamic acid in Smad7, the same position as in U24-7 ligand was found to interact with the tryptophan (W288) in hSmurf2-WW2 using NMR spectroscopy. (212) The average distance measured between the oxygen in the glutamic acid backbone and the NH in the tryptophan indole ring is 1.84 Å, (PDB: 2KXQ) suggesting that this could be a hydrogen bond, O…H-N. Similar bonds might be found in the U24-7/hSmurf2-WW2+3 interaction. As for pU24-6A, it is hard to pinpoint the position of the phosphoryl group without a structural model of the complex, but it is highly possible that an oxygen atom in the phosphoryl group can form a hydrogen bond with the side chain in W288, i.e. that the interaction is also stabilized by a hydrogen bond. Alternatively, temperature might be required for cooperativity. In Figure 5.3B, the heat measure at 15 °C for U24-6A, pU24-6A and U24-7 binding to hSmurf2-WW2+3 domain are highly similar, while the heat measured for the titrations at 5 °C and 25 °C are very different, leading to different slopes. This 156 might be a coincidence, but another explanation is that the overall behaviour of the change in heat capacity (ΔCP) is not constant. ΔCP could be change with temperature and only three data points cannot reveal the true relationship between them. This might be the result from the instability of the WW2+3 domain and the atypical buffer used in the ITC experiemnts, namely 40 mM HEPES and 10 mM NaCl (as compared to the typical 10 mM sodium phosphate as in the case of Nedd4-WW studies). Very few publications can be found on the specific contribution of temperature to tandem protein domain cooperativity, (400) but this is one possibility that cannot be ruled out from the data presented here. Comparing the ΔAest obtained for the pU24-6A interactions versus those for the U24-7 interactions, it is not hard to notice that the |ΔAest(pU24-6A)| values are smaller than the |ΔAest(U24-7)|. Since pU24-6A binds hSmurf2-WW domain better than U24-7, these values are very interesting. One possibility is that pU24-6A is not fully docked in the binding site of hSmurf2-WW3 or hSmurf2-WW2+3. One end of the peptide is bound to hSmurf2-WW domain, but the other end may not be. If the phosphoryl group in the phospho-threonine forms an interaction like the -2 position glutamic acid in Smad7 peptide, the backbone might be “pushed away” from the WW domains. The distance between pU24-6A peptide and hSmurf2-WW2+3 domain would be increased and water molecules may be located in between to fill the gap. This could thus account for a reduction in the non-polar interaction surface. Another potential reason is that the polar area exposed is increased in the pU24-6A/hSmurf2-WW domain complex. The phospho-threonine might again be the reason here. If no binding site, or groove in the WW domain can harbour all of the phosphoryl group, part of the surface of the phosphoryl group would be solvent exposed, leading to a reduction in the ΔAest value. 5.4 Conclusions As in Chapter 4, a potential binding partner for U24 was investigated. Unlike the Nedd4-WW domains, hSmurf2-WW domains bind to U24 with low affinities. As Smurf is a regulator of the TGF-β pathway, this suggests that U24 may not have the ability to directly interfere in this signalling 157 pathway via Smurf-WW domains. However, U24 may have advantages in vivo. U24 is a stable membrane protein and it has the potential to accumulate in the cell membrane as a function of time. Indeed, the down-regulation of TCR/CD3 is time-dependent. (36) An accumulation of U24 in the membrane of a given compartment may result in a sufficiently high concentration of U24 to counteract its low binding affinity. The negative charge upstream from the PY motif, identified in the Nedd4-WW domain studies, is crucial here for the high affinities with hSmurf2-WW domains. The glutamic acid and aspartic acid are conserved in all Smad protein at position -2 (see alignment in Figure F.2), which could explain why Smurf-WW domains interact preferentially with U24-7 and pU24-6A. Compared to U24-6A, the affinity enhancement and potential cooperativity observed in pU24-6A suggests that the phosphoryl group could be the factor to activate U24-6A in vivo, since phosphorylation is typical in signalling pathways. This phosphorylation is located in the shared amino acid stretch with MBP protein, so the mimicry of U24-6A may give arise to its true function in vivo. 5.5 Materials and methods 5.5.1 Expression and purification of GST tagged hSmurf2 WW domains The plasmids containing GST fusion proteins of WW2, WW3 and WW2+3 domains of hSmurf2 were kind gifts from Dr. Julie D. Forman-Kay (Hospital for Sick Children, Toronto). The expression and harvesting of the proteins were done as described in section 4.5.4. GST tagged hSmurf2 WW2 and WW3 were extracted and purified using the method described in section 4.5.3. GST tagged hSmurf2 WW2+3 was supplemented with 4 mM dithiothreitol (DTT). The purified WW2 and WW3 domains were dialyzed against 2 L of pull down buffer and WW2+3 was dialyzed against 2 L of 1 mM DTT supplemented pull down buffer. 158 5.5.2 Purification of hSmurf2 WW domains The hSmurf2 WW domains were extracted and immobilized on GST 4B resin as described previously. After the washing with Wash4 buffer, the resin was washed three times by HRV 3C buffer (50 mM Tris-HCl, 150 mM NaCl, 1mM DTT, pH 8.0). 4 mL HRV 3C buffer and 240 units of HRV 3C protease (Pierce) were added in the resin slurry for WW2, WW3 or WW2+3. An additional 1 mM DTT was supplemented in the slurry of WW2+3. The resin slurries were incubated at 4 °C for 16 hours. The WW domain containing fraction was collected as described in section 4.5.4, and 150 μL of HRV 3C buffer washed GST 4B resin was used to trap residual GST protein. Samples were taken from each fraction and analyzed with Tris-Glycine and Tris-Tricine SDS-PAGE. The resulting purified WW domains were submitted to the Mass Spectrometry Centre (Chemistry, UBC) to confirm the mass using MALDI-TOF MS. The purified WW2 and WW3 fractions were dialyzed against 2 L phosphate buffer or HEPES buffer (40 mM HEPES, 10 mM NaCl, pH 7.2) for five hours to overnight, and the purified WW2+3 domain was dialyzed against 2 L HEPES buffer with 0.5 mM TCEP supplemented. Then the dialyzed fractions were concentrated to approximate 4 mL using Microsep centrifugal devices with MWCO of 1 kDa. The concentrated samples were dialyzed against 2 L of freshly made phosphate buffer, HEPES buffer or 1 L HEPES buffer with 1 mM TCEP supplemented for WW2+3 overnight for ITC or NMR experiments. 5.5.3 GST tagged hSmurf2 WW domains pull down experiments with recombinant U24-6A and U24-7 protein The experiment was carried out as described in section 4.5.6 and 1 mM DTT was kept in the pull down buffer for WW2+3. Additional DTT was added to make the WW2+3 stock solution to a final DTT concentration of 4 mM. 1 mM of DTT was kept in all the buffers that were used to wash WW2+3 containing resin. Finally, 5 to 10 μL of the samples were used for gel electrophoresis. 159 5.5.4 ITC experiments The cleaved and dialyzed hSmurf2 WW domains were prepared as described above and were concentrated to 150 to 300 μM using Microsep centrifugal devices with a MWCO of 1 kDa. The peptide samples were prepared as described in Chapter 4 and the concentration of protein and peptide were determined as previously. hSmurf2 WW2 and WW3 domains were loaded directly after filtration and degassing, while hSmurf2 WW2+3 was supplemented with fresh 1 mM TCEP in the solution, and degassed again. The 1 mM TCEP was also supplemented in the peptide solution and protein free buffer for WW2+3. ITC experiments were done on an iTC200 MicroCalorimeter at 5 °C, 15 °C and 25 °C. Data showed that these proteins might not be stable at higher temperatures. (399) The titration protocol was the same as in section 3.5.3. The Ka and ∆H were obtained using the same method and the changes of heat capacities were calculated the same way. The control experiments for protein stabilities were carried out as well. 160 Chapter 6 Chapter 6 Conclusions and Future Perspectives 6.1 Thesis summary This thesis mainly described my work on studying the membrane protein U24-7, from obtaining the recombinant protein, preliminary structural studies, to finding its potential binding partners. Methods to obtain purified recombinant U24-7 were developed. Methods previously developed for the expression and purification of U24-6A could not be directly applied to U24-7 to obtain the protein in sufficient yields. In Chapter 2, the detailed purification procedures were presented and the optimizations discussed. Overall, the protocol yielded U24-7 in sufficient quantities (from as little as 500 mL of M9 media) and well dispersed in solution, so that structural studies using NMR are now feasible. Preliminary structural data on U24-7 shows it to be helical and very stable upon heating. In this thesis, two hypotheses were tested to determine whether U24 plays a role in MS. The first was the mimicry hypothesis, namely that U24 may mimic myelin basic protein. The second was that U24 interferes with other proteins implicated in MS, e.g. Nogo-A, MAG, etc., via its endosomal recycling role and its interaction with WW domains. A number of studies have suggested that U24-6A might be implicated in MS via molecular mimicry of MBP. (1, 54) As introduced in Chapter 1, U24 possibly functions in vivo via a binding partner. Since U24-6A contains a PxxP motif, the interaction with Fyn-SH3 domain, which has been previously identified bind to MBP, (129) was studied in Chapter 3. The data obtained from ITC and NMR demonstrated that this interaction is very weak. The dissociation constant is in the range of 5-10 mM. Although weak, this interaction is specific, with residues in the RT loop of Fyn-SH3 being most perturbed. There is no specific interaction between U24-7 (where the PxxP motif is absent) 161 and Fyn-SH3 domain. Besides molecular mimicry or the PxxP motif mentioned above, U24 might be implicated in MS via its PY motif. This motif is responsible for down-regulation of TCR/CD3 and TfR in T-cells, and Nedd4-WW domains are potential binding partners of U24. ITC data between three U24 peptides and Nedd4-WW domains indicated that U24 can bind to these WW domains strongly. In particular, hNedd4L-WW3* binds to U24 peptides better than the other WW domains tested. Among all three peptide ligands, pU24-6A was found to be the best ligand when binding to WW domains, followed by U24-7 peptide. The strongest affinity detected is 0.76 ± 0.03 μM, which is between pU24-6A peptide and hNedd4L-WW3*. To compare, the binding affinity between a peptide consisting of residues IKHEPENPPPYEEAM in Nogo-A and hNedd4L-WW3* yielded a Kd = 3.9 ± 0.2 μM at 25 °C. To better understand how phosphorylation has such a favourable impact on binding, further ITC experiments were conducted. Heat capacities were determined and it was found that there were no big differences in ΔCP between U24-6A and U24-7. It was also found that salt can slightly interfere with the pU24-6A/hNedd4L-WW3* interaction. Overall, the ITC data could not explain why pU24-6A is the better ligand. The NMR data showed similar combined chemical shift perturbations in WW domains for the three ligands. More NOE signals were detected in the loop regions of WW3 domains when they were complexed with pU24-6A or U24-7 than in complexes with U24-6A, suggesting that the peptide is more tightly held by the protein, consistent with the measured binding affinities. In order to better understand the necessary requirements for good binding to WW domains, a different type of WW domain, namely Smurfs were studied. These are non-canonical domains, i.e. one W is replaced by a Y or F. ITC experiments were carried out between U24 peptides and hSmurf2-WW domains. Unlike Nedd4-WW domains, hSmurf2-WW domains did not bind to U24 strongly, including pU24-6A. The dissociation constants are in the middle to high micromolar range. In addition, the affinities were not enhanced when U24 bound to the coupled WW domains, hSmurf2-WW2+3. However, both domains may be involved in binding to U24, based on the heat capacity data. 162 Overall, the studies of U24 presented in this thesis describe, for the first time, the structure of U24-7 protein and in vitro studies of U24 binding to Fyn-SH3 and several WW domains. From the affinities of these interactions, U24 binds to Nedd4-WW domains better than all the other binding partners tested. The negatively charged residues or phosphorylated residue upstream of the PY motif is the key to strong binding or the cooperativity observed in the case of hSmurf2-WW tandem domains. 6.2 Conclusions U24-6A is proposed to be a potential trigger of the immune response in MS and a mimic of MBP since U24-6A has a seven amino acid stretch which is identical to MBP. (1) To test this mimicry theory, U24-6A was found to be phosphorylated at the same site in the shared region. (54) These findings suggested that U24-6A could be treated by the immune system in the same way as MBP. In this thesis, the interaction between U24-6A and Fyn-SH3 domain, a protein domain MBP binds to, was studied and the association was found to be weak. This interaction might not be stronger in vivo as Fyn-SH3 domain prefers ligands with upstream positive charges, which U24-6A cannot acquire via post-translational modifications. Thus, a weak association is expected for U24-6A and Fyn-SH3 domain in vivo, weakening the hypothesis that U24-6A functions via Fyn-SH3 domain binding, leading to immature myelin development or less modified MBP. The data presented here suggests that U24-6A does not behave like a mimic of MBP with respect to binding to Fyn-SH3 domain. However, some studies have suggested that weak binding may still be biologically relevant. (401) It is also possible that mimicry occurs in a different manner to that explored here (e.g. MBP undergoes many other post-translational modifications, shifting the balance to diseased myelin). (106) On the other hand, how U24 is implicated in different diseases can be studied from the angle of viral infection. During viral infection, U24 is believed to modulate host immune response by down-regulation of TCR/CD3. (36) U24 can interfere with other cellular processes that have a similar endocytic recycling pathway to TCR/CD3. This points to the second hypothesis that U24 may function by interfering with the regulation of membrane proteins in the neuron, which is done 163 through WW domains. Unlike Fyn-SH3 domain, Nedd4-WW domains can bind to both versions of U24 with high affinities. The Kd ranges from 760 nM to 160 μM. The estimated affinities in vivo may be even higher than in vitro if one considers that there are two to four WW domains in one Nedd4 protein. Even though no cooperativity has been reported for Nedd4-WW domains, it is quite possible that an enhancement in binding could be achieved through using multiple WW domains to bind multiple ligands or multiple ends of a given ligand. (149, 215, 358) The strong interactions between U24 and Nedd4-WW domains support the hypothesis that U24 function may be intricately entwined with the function of Nedd4-WW domains. As discussed in Chapter 4, the functions of U24 could interfere with those of Nedd4, e.g. activating Nedd4 in the wrong membrane compartment or blocking its down-stream interactions. Some studies in the literature have shown how hijacking of Nedd4 protein can occur by viral or cellular proteins. (333, 335, 338) Aside from this evidence, there is little to indicate how U24 affects its cellular target, other than the known fact that it causes down-regulation of TCR/CD3 via its PY motif. Nevertheless, the strong interaction found here between U24 and Nedd4-WW domains suggests that further in vivo studies on how interference with the signaling pathway regulated by Nedd4 or other E3s may impact disease. This could provide new insights into steps involved during neuronal development or myelin formation, i.e. the importance in the regulation of structural and regulatory membrane proteins (MOG, MAG and Nogo-A), ion channels or exchangers or even iron levels on neuron or microglia cells (section 1.2.2). Ultimately, such information may provide clues as to how a neurodegenerative disease like MS could be the result from the poor regulation of one or more of these components. An interesting outcome from the work described in this thesis was the observation that phosphorylated U24-6A was found to bind better than U24-6A. This suggests that this may be the relevant functional form of U24-6A. The mimicry theory may be connected to the function of U24-6A here. A strong affinity between phosphorylated U24-6A protein and WW domains is expected in vivo, perhaps in neurons or oligodendrocytes, and this could be the reason why T-cells target it as an antigen. It is a complex containing an alien protein, phosphorylated U24-6A. The effort to eliminate this interaction via neutralizing the active part of this protein, PRpTPPPSY, may 164 trigger the autoimmune reaction against MBP protein. The sequence of this region in MBP, PRpTPPPS, is highly conserved among mammals. (402) Although there is no direct evidence from the literature to support a link between autoimmune response and phosphorylation of MBP, it is possible that U24-6A’s function makes it a target of T-cells. Although the phosphorylation of U24-7 was not proven using MAP kinase in this thesis, it is possible that phosphorylation occurs in vivo. Overall, this dissertation described the study of the function of U24 protein via its structural stability and its interaction with different binding partners. Many of the interactions were found to be weak and may therefore not be significant in vivo. Nevertheless, U24 binds to Nedd4-WW domains tightly, which may be how it functions. As for the two hypotheses detailed in Chapter 1, the evidence obtained to support that U24 might function through the mimicry of MBP and thereby interfere with MBP function was not strong. The results described in this dissertation rather suggest that the blocking endosmal recycling may be more important. Not only were strong and tight interactions observed, but the importance of phosphorylation was also revealed. This opens up the possibility that U24 would implicated in MS through Nedd4. As mentioned in Chapters 1 and 4, such an interaction may have an impact on protein levels of such MS players as Nogo-A or MAG or alternatively on iron levels. Further studies will help to pinpoint U24’s true function and its connection to this debilitating disease. 6.3 Future perspectives The preliminary structural study of U24-7 shows that it is promising to study its structure using NMR. Unlike the NMR spectra collected for U24-6A, the peaks are well resolved so it should be possible to determine the solution structure of U24-7 via NOESY or triple-resonance spectra in detergent micelles. More structural information could be extracted from the measurement of residual internuclear dipolar couplings (RDCs). (403) RDCs can provide averaged orientational information of an internuclear vector, which is very useful to refine protein structures. (404) For membrane proteins, lipid bicelles could be used to solubilize them, as well as align the membrane protein for RDCs measurement. (405, 406) Although there are challenges in distinguishing 165 multiplets and measuring the RDCs in very crowded spectra, it is a method that allows one to study membrane proteins in a water and lipid rich environment. The interactions in this thesis are studied mostly using ITC and NMR, methods which do not directly provide time-dependent parameters, such as reaction rate constants kon and koff. A better understanding of the kinetics involved in these interactions could be probed using SPR. It is a very sensitive method and capable of direct measurement of kon and koff using very limited amount of sample. In order to better understand the reason for the strong interactions between pU24-6A and hNedd4L-WW3*, it could be informative to examine the kinetic parameters related to the interactions between three U24 ligands and WW domains. It has been previously observed that the phosphorylated PY ligand from β and γ subunits of rat ENaC binds to WW domain at an enhanced kon, but similar koff. (376) This could be the same for pU24-6A. EXSY allows one to study protein binding dynamics as well but there are some limitations using NMR. The amide magnetization relaxes very rapidly. What is more, U24-6A does not bind to Nedd4-WW domain strongly, so the apo and bound form are not in the slow exchange regime, relative to the chemical shift time scale. One possible approach to use NMR to investigate the kinetics of the interaction is to use only the nitrogen magnetization Nz, (289, 290) instead of correlated amides. This could solve the relaxation problem observed and yield reliable kinetic data. SPR could also be used to study the kinetics of U24 binding to hSmurf2-WW domains. The cooperativity of tandem WW domains does not enhance the affinities with U24. The reason could be that kon and koff are increased or decreased to the same extent. The two domains in hSmurf2-WW2+3 may hinder the initial binding of PY ligands onto WW3 domain. Alternatively, the two domains may work together to strengthen the interaction by forming a stable complex with PY ligands. Performing SPR experiments would allow us to understand the details of the interactions. NMR spectroscopy could also be used to investigated the potential weak salt bridges which may be present in the pU24-6A/hNedd4L-WW3* complex. In addition to using 15N and 1H chemical shifts in the WW domain, the chemical shift and coupling constant of 31P in the pU24-6A 166 peptide is sensitive to the chemical environment of the phospho-threonine. 31P is 100% natural abundant, has a long longitudinal relaxation times (T1) and is more sensitive than 13C or 15N. If there are hydrogen bonds formed in between the guanidine or side chain amide group and phosphoryl group, intermolecular coupling of three-bond or two-bond, 31P-15N or 31P-1H, in a potential hydrogen bond such as N-H…O-P, could be detected. (407) In addition, an NOE might be observed between 31P and the surrounding protons, which could be used to determine relative atomic distances in the local structure. 17O enriched phospho-threonine peptide could be used as well. Although it is very hard to label, 17O NMR may provide detailed information of intermolecular hydrogen bonds. (408) All of the interactions presented in this thesis used 15-mer peptides, which contain the core interaction motifs. A longer ligand could be used as it mimics the protein better. It would be interesting to see whether affinities could be enhanced, especially for tandem WW domains. The affinity enhancement observed in the literature for hSmurf2-WW2+3 may be due to the lengths of the length. (212) On the other hand, other types of WW domains could be tested as well, especially those that are not from E3 ubiquitin ligases. As a conserved domain, WW domains exist in several kinds of proteins, i.e. peptidyl-prolyl cis/trans isomerase 1 (PIN1), WW domain-containing oxidoreductase (WWOX), transcription elongation regulator 1 (TCERG1) and kidney and brain expressed protein (KIBRA). (172, 409) Their WW domain might be able to bind to U24 at very high affinity, too. As affinities between Nedd4-WW and Smurf2-WW domain differ a lot, these WW domains may or may not bind tightly to U24, depending on their ligand preferences. Since there is no clear biological evidence which protein might interact with U24, a range of WW domains could be potential binding partners of U24, making them worth studying. Finally, to complete the picture, it would be important to understand how the interactions found to be important here translate at the cellular level. In particular with regards to MS, it would be important to demonstrate whether U24 affects endosomal recycling in oligodendrocytes or whether increased levels of U24 can be correlated to demyelination. Another approach could be to 167 use animal models. For instance, Caenorhabditis elegans is a kind of transparent nematode that is commonly used as a model organism to study neuronal development. C. elegans could be used to study the function of U24 to see if it can interfere with neuronal development, i.e. whether it interferes specifically with the downstream function of Nedd4 via binding to its WW domain. Y92H12A.2 in C. elegans is an ortholog of human Nedd4L protein. Genetic manipulation of C. elegans is easy and the phenotypes can be identified via microscopy. (410, 411) What is more, the effect of U24 endocytosis may be observed in this model organism. In vivo investigations of endocytic recycling in C. elegans have been reported (412, 413), so it is possible that such a system could be used to determine the localization and function of U24. In this way, not only the binding partner of U24 could be identified, but the phenotype resulting from expressing U24 could be observed as well. Such studies could provide a clue for U24’s implication in MS, with the caveat that the neurons in C. elegans are not myelinated, so this system is not the best model for MS. A more relevant model to establish whether a direct link between U24 and MS exists could be a mouse model. Along with the development of genome engineering, multiple mutations can easily be deleted or introduced to embryos or mature tissues of mice using clustered regularly interspaced short palindromic repeats (CRISPR) associated RNA guided endonucleases Cas-9. (414–416) U24 could be integrated into the genome of mice using this method. The function of U24 in such a mouse model could thus be characterized, providing knowledge on Roseoloviruses infection and MS. Understanding the basic role of U24 from Roseolovirus is important to explain the role of this protein in the virus and its associated diseases. HHV-6, and to some extent HHV-7, is highly prevalent in humans. Infection typically occurs during early childhood, a period where neuronal development is at its highest. As more and more evidence comes to light of the involvement of viruses in neurodegenerative diseases, it could become increasingly important carry out both the in vitro and in vivo studies suggested above in order to pinpoint the role of U24. 168 Bibliography 1. Tejada-Simon M V., Zang YCQ, Hong J, Rivera VM, Zhang JZ. 2003. Cross-reactivity with myelin basic protein and human herpesvirus-6 in multiple sclerosis. Ann. Neurol. 53(2):189–97 2. Scheurer ME, Pritchett JC, Amirian ES, Zemke NR, Lusso P, Ljungman P. 2012. HHV-6 encephalitis in umbilical cord blood transplantation: a systematic review and meta-analysis. Bone Marrow Transplant. 48(4):574–80 3. Rieux C, Gautheret-Dejean A, Challine-Lehmann D, Kirch C, Agut H, Vernant JP. 1998. Human herpesvirus-6 meningoencephalitis in a recipient of an unrelated allogeneic bone marrow transplantation. Transplantation. 65(10):1408–11 4. Nora-krukle Z, Chapenko S, Logina I, Millers A. 2011. Human Herpesvirus 6 and 7 Reactivation and Disease Activity in Multiple Sclerosis. Medicina (B. Aires). 47(10):527–31 5. Caselli E, Di Luca D. 2007. Molecular biology and clinical associations of Roseoloviruses human herpesvirus 6 and human herpesvirus 7. New Microbiol. 30(3):173–87 6. Stoeckle MY. 2000. The spectrum of human herpesvirus 6 infection: from roseola infantum to adult disease. Annu. Rev. Med. 51:423–30 7. Arbuckle JH, Medveczky MM, Luka J, Hadley SH, Luegmayr A, et al. 2010. The latent human herpesvirus-6A genome specifically integrates in telomeres of human chromosomes in vivo and in vitro. Proc. Natl. Acad. Sci. U. S. A. 107(12):5563–68 8. Jasirwan C, Tang H, Kawabata A, Mori Y. 2015. The human herpesvirus 6 U21-U24 gene cluster is dispensable for virus growth. Microbiol. Immunol. 59(1):48–53 9. Lusso P, Markham PD, Tschachler IE, Di F, Veronese M, et al. 1988. In vitro Cellular Tropism of Human B-Lymphotropic Virus (Human Herpesvirus-6). J. Exp. Med. 167(May):1659–70 10. Salahuddin SZ, Ablashi D V, Markham PD, Josephs SF, Kaplan M, et al. 1986. Isolation of a New Virus , HBLV , in Patients with Disorders Lymphoproliferative. Science. 234(4776):596–601 11. Ablashi D, Agut H, Alvarez-Lafuente R, Clark DA, Dewhurst S, et al. 2014. Classification of HHV-6A and HHV-6B as distinct viruses. Arch. Virol. 159(5):863–70 12. Dominguez G, Dambaugh TR, Stamey FR, Dewhurst S, Inoue N, Pellett PE. 1999. Human herpesvirus 6B genome sequence: coding content and comparison with human herpesvirus 6A. J. Virol. 73(10):8040–52 13. Ablashi D V, Balachandran N, Josephs SF, Hung CL, Krueger GR, et al. 1991. Genomic polymorphism, growth properties, and immunologic variations in human herpesvirus-6 isolates. Virology. 184(2):545–52 14. De Bolle L, Van Loon J, De Clercq E, Naesens L. 2005. Quantitative analysis of human herpesvirus 6 cell tropism. J. Med. Virol. 75(1):76–85 15. Harberts E, Yao K, Wohler JE, Maric D, Ohayon J, et al. 2011. Human herpesvirus-6 entry into the central nervous system through the olfactory pathway. Proc. Natl. Acad. Sci. U. S. A. 108(33):13734–39 169 16. Bolle L De, Naesens L, Clercq E De. 2005. Update on Human Herpesvirus 6 Biology , Clinical Features , and Therapy Update on Human Herpesvirus 6 Biology , Clinical Features , and Therapy. Clin. Microbiol. Rev. 18(1):217–45 17. Nacheva EP, Ward KN, Brazma D, Virgili A, Howard J, et al. 2008. Human Herpesvirus 6 Integrates Within Telomeric Regions as Evidenced by Five Different Chromosomal Sites. J. Med. Virol. 80:1952–58 18. Tanaka-Taya K, Sashihara J, Kurahashi H, Amo K, Miyagawa H, et al. 2004. Human herpesvirus 6 (HHV-6) is transmitted from parent to child in an integrated form and characterization of cases with chromosomally integrated HHV-6 DNA. J. Med. Virol. 73(3):465–73 19. Osterrieder N, Wallaschek N, Kaufer BB. 2014. Herpesvirus Genome Integration into Telomeric Repeats of Host Cell Chromosomes. Annu. Rev. Virol. 1(1):215–35 20. Kaufer BB, Jarosinski KW, Osterrieder N. 2011. Herpesvirus telomeric repeats facilitate genomic integration into host telomeres and mobilization of viral DNA during reactivation. J. Exp. Med. 208(3):605–15 21. Fotheringham J, Jacobson S. 2005. Human herpesvirus 6 and multiple sclerosis: potential mechanisms for virus-induced disease. Herpes. 12(1):4–9 22. Leibovitch EC, Jacobson S. 2014. Evidence linking HHV-6 with multiple sclerosis: an update. Curr. Opin. Virol. 9:127–33 23. Challoner PB, Smith KT, Parker JD, MacLeod DL, Coulter SN, et al. 1995. Plaque-associated expression of human herpesvirus 6 in multiple sclerosis. Proc. Natl. Acad. Sci. U. S. A. 92(16):7440–44 24. Isaacson E, Glaser CA, Forghani B, Amad Z, Wallace M, et al. 2005. Evidence of human herpesvirus 6 infection in 4 immunocompetent patients with encephalitis. Clin. Infect. Dis. 40(6):890–93 25. Fotheringham J, Donati D, Akhyani N, Fogdell-Hahn A, Vortmeyer A, et al. 2007. Association of human herpesvirus-6B with mesial temporal lobe epilepsy. PLoS Med. 4(5):0848–57 26. Crawford JR, Kadom N, Santi MR, Mariani B, Lavenstein BL. 2007. Human herpesvirus 6 rhombencephalitis in immunocompetent children. J. Child Neurol. 22(11):1260–68 27. Alenda R, Álvarez-Lafuente R, Costa-Frossard L, Arroyo R, Mirete S, et al. 2014. Identification of the major HHV-6 antigen recognized by cerebrospinal fluid IgG in multiple sclerosis. Eur. J. Neurol. 21(8):1096–1101 28. Frenkel N, Schirmer EC, Wyatt LS, Katsafanas G, Roffman E, et al. 1990. Isolation of a new herpesvirus from human CD4+ T cells. Proc. Natl. Acad. Sci. U. S. A. 87(2):748–52 29. Nicholas J. 1996. Determination and analysis of the complete nucleotide sequence of human herpesvirus. J. Virol. 70(9):5975–89 30. Lusso P, Secchiero P, Crowley RW, Garzino-demo A, Berneman ZN, Gallo RC. 1994. CD4 is a critical component of the receptor for human herpesvirus 7: Proc. Natl. Acad. Sci. U. S. A. 91(April):3872–76 31. Dalgleish AG, Beverley PC, Clapham PR, Crawford DH, Greaves MF, Weiss RA. 1984. The 170 CD4 (T4) antigen is an essential component of the receptor for the AIDS retrovirus. Nature. 312(5996):763–67 32. Yoshikawa T, Ihira M, Akimoto S, Miyake F, Suga S, et al. 2004. Detection of Human Herpesvirus 7 DNA by Loop-Mediated Isothermal Amplification. J. Clin. Virol. 42(3):1348–52 33. Hall CB, Caserta MT, Schnabel KC, McDermott MP, Lofthus GK, et al. 2006. Characteristics and acquisition of human herpesvirus (HHV) 7 infections in relation to infection with HHV-6. J. Infect. Dis. 193(8):1063–69 34. Lautenschlager I, Lappalainen M, Linnavuori K, Suni J, Höckerstedt K. 2002. CMV infection is usually associated with concurrent HHV-6 and HHV-7 antigenemia in liver transplant patients. J. Clin. Virol. 25(Suppl 2):S57–61 35. Schwartz KL, Richardson SE, Ward KN, Donaldson C, MacGregor D, et al. 2014. Delayed Primary HHV-7 Infection and Neurologic Disease. Pediatrics. 133(6):e1541–47 36. Sullivan BM, Coscoy L. 2010. The U24 protein from human herpesvirus 6 and 7 affects endocytic recycling. J. Virol. 84(3):1265–75 37. Yao K, Mandel M, Akyani N, Maynard K, Sengamalay N, et al. 2006. Differential HHV-6A Gene Expression in T Cells and Primary Human Astrocytes Based on Multi-Virus Array Analysis. Glia. 53:789–98 38. Yao K, Akyani N, Donati D, Sengamalay N, Fotheringham J, et al. 2006. Detection of HHV-6B in post-mortem central nervous system tissue of a post-bone marrow transplant recipient: a multi-virus array analysis. J. Clin. Virol. 37(Suppl 1):57–62 39. Vossen MTM, Westerhout EM, Söderberg-Nauclér C, Wiertz EJHJ. 2002. Viral immune evasion: A masterpiece of evolution. Immunogenetics. 54(8):527–42 40. Glosson NL, Gonyo P, May NA, Schneider CL, Ristow LC, et al. 2010. Insight into the mechanism of human herpesvirus 7 U21-mediated diversion of class I MHC molecules to lysosomes. J. Biol. Chem. 285(47):37016–29 41. Glosson NL, Hudson AW. 2007. Human herpesvirus-6A and -6B encode viral immunoevasins that downregulate class I MHC molecules. Virology. 365(1):125–35 42. Kofod-Olsen E, Ross-Hansen K, Schleimann MH, Jensen DK, Moller JML, et al. 2012. U20 Is Responsible for Human Herpesvirus 6B Inhibition of Tumor Necrosis Factor Receptor-Dependent Signaling and Apoptosis. J. Virol. 86(21):11483–92 43. Dewin DR, Catusse J, Gompels UA. 2006. Identification and characterization of U83A viral chemokine, a broad and potent beta-chemokine agonist for human CCRs with unique selectivity and inhibition by spliced isoform. J. Immunol. 176(1):544–56 44. Catusse J, Parry CM, Dewin DR, Gompels UA. 2007. Inhibition of HIV-1 infection by viral chemokine U83A via high-affinity CCR5 interactions that block human chemokine-induced leukocyte chemotaxis and receptor internalization. Blood. 109(9):3633–39 45. Sullivan BM, Coscoy L. 2008. Downregulation of the T-cell receptor complex and impairment of T-cell activation by human herpesvirus 6 u24 protein. J. Virol. 82(2):602–8 46. Lusso P, Malnati M, De Maria A, Balotta C, DeRocco SE, et al. 1991. Productive infection of CD4+ and CD8+ mature human T cell populations and clones by human herpesvirus 6: 171 Transcriptional down-regulation of CD3. J. Immunol. 147(2):685–91 47. Dunn KW, McGraw TE, Maxfield FR. 1989. Iterative fractionation of recycling receptors from lysosomally destined ligands in an early sorting endosome. J. Cell Biol. 109(6 II):3303–14 48. Presley JF, Mayor S, McGraw TE, Dunn KW, Maxfield FR. 1997. Bafilomycin A1 treatment retards transferrin receptor recycling more than bulk membrane recycling. J. Biol. Chem. 272(21):13929–36 49. Madrid R, Janvier K, Hitchin D, Day J, Coleman S, et al. 2005. Nef induced Alteration of the Early or Recycling Endosomal Compartment Correlates with Enhancement of HIV-1 Infectivity. J. Biol. Chem. 280(6):5032–44 50. Amorim NA, Da Silva EML, De Castro RO, Da Silva-Januário ME, Mendonça LM, et al. 2014. Interaction of HIV-1 Nef with the Host Protein Alix Promotes Lysosomal Targeting of CD4 Receptor. J. Biol. Chem. 289(40):27744–56 51. Barral DC, Cavallari M, McCormick PJ, Garg S, Magee AI, et al. 2008. CD1a and MHC class I follow a similar endocytic recycling pathway. Traffic. 9(9):1446–57 52. MacGurn JA, Hsu P-C, Emr SD. 2012. Ubiquitin and Membrane Protein Turnover: From Cradle to Grave. Annu. Rev. Biochem. 81(1):231–59 53. Grant BD, Donaldson JG. 2009. Pathways and mechanisms of endocytic recycling. Nat. Rev. Mol. Cell Biol. 10(9):597–608 54. Tait AR, Straus SK. 2008. Phosphorylation of U24 from Human Herpes Virus type 6 (HHV-6) and its potential role in mimicking myelin basic protein (MBP) in multiple sclerosis. FEBS Lett. 582(18):2685–88 55. Garai A, Zeke A, Gogl G, Toro I, Fordos F, et al. 2012. Specificity of Linear Motifs That Bind to a Common Mitogen-Activated Protein Kinase Docking Groove. Sci. Signal. 5(245):ra74–ra74 56. Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, et al. 2004. UCSF Chimera-A visualization system for exploratory research and analysis. J. Comput. Chem. 25(13):1605–12 57. Pace CN, Scholtz JM. 1998. A helix propensity scale based on experimental studies of peptides and proteins. Biophys. J. 75(1):422–27 58. Moriarty DF, Raleigh DP. 1999. Effects of sequential proline substitutions on amyloid formation by human amylin20-29. Biochemistry. 38(6):1811–18 59. Wood SJ, Wetzel R, Martin JD, Hurle MR. 1995. Prolines and amyloidogenicity in fragments of the Alzheimer’s peptide beta/A4. Biochemistry. 34(3):724–30 60. Williamson MP. 1994. The structure and function of proline-rich regions in proteins. Biochem. J. 297(Pt 2):249–60 61. Zarrinpar A, Bhattacharyya RP, Lim WA. 2003. The structure and function of proline recognition domains. Sci. Signal., pp. 1–10 62. Kay BK, Williamson MP, Sudol M. 2000. The importance of being proline: the interaction of proline-rich motifs in signaling proteins with their cognate domains. FASEB J. 14(2):231–41 63. Chen HI, Sudol M. 1995. The WW domain of Yes-associated protein binds a proline-rich ligand that differs from the consensus established for Src homology 3-binding modules. Proc. 172 Natl. Acad. Sci. U. S. A. 92(17):7819–23 64. Mayer BJ. 2001. SH3 domains: complexity in moderation. J. Cell Sci. 114(Pt 7):1253–63 65. Musacchio A, Noble M, Pauptit R, Wierenga R, Saraste M. 1992. Crystal structure of a Src-homology 3 (SH3) domain. Nature. 359(6398):851–55 66. Yu H, Chen JK, Feng S, Dalgarno DC, Brauer AW, Schreiber SL. 1994. Structural basis for the binding of proline-rich peptides to SH3 domains. Cell. 76(5):933–45 67. Feng S, Chen JK, Yu H, Simon JA, Schreiber SL. 1994. Two binding orientations for peptides to the Src SH3 domain: development of a general model for SH3-ligand interactions. Science. 266(5188):1241–47 68. Feng S, Kasahara C, Rickles RJ, Schreiber SL. 1995. Specific interactions outside the proline-rich core of two classes of Src homology 3 ligands. Proc. Natl. Acad. Sci. U. S. A. 92(26):12408–15 69. Ibragimova GT, Wade RC. 1998. Importance of explicit salt ions for protein stability in molecular dynamics simulation. Biophys. J. 74(6):2906–11 70. Huang X, Poy F, Zhang R, Joachimiak A, Sudol M, Eck MJ. 2000. Structure of a WW domain containing fragment of dystrophin in complex with beta-dystroglycan. Nat. Struct. Biol. 7(8):634–38 71. Staub O, Dho S, Henry PC, Correa J, Ishikawa T, et al. 1996. WW domains of Nedd4 bind to the proline-rich PY motifs in the epithelial Na+ channel deleted in Liddle’s syndrome. EMBO J. 15(10):2371–80 72. Ermekova KS, Zambrano N, Linn H, Minopoli G, Gertler F, et al. 1997. The WW domain of neural protein FE65 interacts with proline-rich motifs in Mena, the mammalian homolog of Drosophila enabled. J. Biol. Chem. 272(52):32869–77 73. Faber PW, Barnes GT, Srinidhi J, Chen J, Gusella JF, MacDonald ME. 1998. Huntingtin interacts with a family of WW domain proteins. Hum. Mol. Genet. 7(9):1463–74 74. Mosser EA, Kasanov JD, Forsberg EC, Kay BK, Ney PA, Bresnick EH. 1998. Physical and functional interactions between the transactivation domain of the hematopoietic transcription factor NF-E2 and WW domains. Biochemistry. 37(39):13686–95 75. Wiesner S, Stier G, Sattler M, Macias MJ. 2002. Solution structure and ligand recognition of the WW domain pair of the yeast splicing factor Prp40. J. Mol. Biol. 324(4):807–22 76. Ilsley JL, Sudol M, Winder SJ. 2002. The WW domain: Linking cell signalling to the membrane cytoskeleton. Cell. Signal. 14(3):183–89 77. Nguyen JT, Turck CW, Cohen FE, Zuckermann RN, Lim WA. 1998. Exploiting the basis of proline recognition by SH3 and WW domains: design of N-substituted inhibitors. Science. 282(5396):2088–92 78. Kanelis V, Rotin D, Forman-Kay JD. 2001. Solution structure of a Nedd4 WW domain–ENaC peptide complex. Nat. Struct. Biol. 8(5):407–12 79. Pornillos O, Alam SL, Davis DR, Sundquist WI. 2002. Structure of the Tsg101 UEV domain in complex with the PTAP motif of the HIV-1 p6 protein. Nat. Struct. Biol. 9(11):812–17 80. Kofler MM, Freund C. 2006. The GYF domain. FEBS J. 273(2):245–56 173 81. Hu H, Columbus J, Zhang Y, Wu D, Lian L, et al. 2004. A map of WW domain family interactions. Proteomics. 4(3):643–55 82. Sudol M, Hunter T. 2000. NeW Wrinkles for an old domain. Cell. 103(7):1001–4 83. Prehoda KE, Lee DJ, Lim W a, Francisco S. 1999. Structure of the Enabled/VASP Homology 1 Domain–Peptide Complex: A Key Component in the Spatial Control of Actin Assembly. Cell. 97:471–80 84. Fedorov AA, Fedorov E, Gertler F, Almo SC. 1999. Structure of EVH1, a novel proline-rich ligand-binding module involved in cytoskeletal dynamics and neural function. Nat. Struct. Biol. 6(7):661–65 85. Ball LJ, Jarchau T, Oschkinat H, Walter U. 2002. EVH1 domains: Structure, function and interactions. FEBS Lett. 513(1):45–52 86. Sundquist WI, Schubert HL, Kelly BN, Hill GC, Holton JM, Hill CP. 2004. Ubiquitin recognition by the human TSG101 protein. Mol. Cell. 13(6):783–89 87. Kosuge H. 2000. HHV-6, 7 and their related diseases. J. Dermatol. Sci. 22(3):205–12 88. Mori Y, Yamanishi K. 2007. HHV - 6A, 6B, and 7: pathogenesis, host response, and clinical disease. In Human Herpesviruses: Biology, Therapy, and Immunoprophylaxis., ed A A, C-F G, M E. 4: Cambridge: Cambridge University Press 89. Opsahl ML, Kennedy PGE. 2005. Early and late HHV-6 gene transcripts in multiple sclerosis lesions and normal appearing white matter. Brain. 128(3):516–27 90. Compston A, Coles A. 2008. Multiple sclerosis. Lancet. 372(9648):1502–17 91. Trapp BD, Nave K-A. 2008. Multiple sclerosis: an immune or neurodegenerative disorder? Annu. Rev. Neurosci. 31:247–69 92. Rosati G. 2001. The prevalence of multiple sclerosis in the world: an update. Neurol. Sci. 22:117–39 93. Borkosky SS, Whitley C, Kopp-Schneider A, Hausen H, de Villiers EM. 2012. Epstein-barr virus stimulates torque teno virus replication: A possible relationship to multiple sclerosis. PLoS One. 7(2):32160 94. Wucherpfennig KW, Strominger JL. 1995. Molecular mimicry in T cell-mediated autoimmunity: viral peptides activate human T cell clones specific for myelin basic protein. Cell. 80(5):695–705 95. Clanet M. 2008. Jean-Martin Charcot. Int. MS J. 15:59–61 96. Chari DM. 2007. The Neurobiology of Multiple Sclerosis. Int. Rev. Neurobiol. 79(07):589–620 97. Reynaud JM, Horvat B. 2013. Human Herpesvirus 6 and Neuroinflammation. ISRN Virol. 2013: 98. Zhao ZS, Granucci F, Yeh L, Schaffer PA, Cantor H. 1998. Molecular mimicry by herpes simplex virus-type 1: autoimmune disease after viral infection. Science. 279(5355):1344–47 99. Sospedra M, Martin R. 2005. Immunology of Multiple Sclerosis. Annu. Rev. Immunol. 23(1):683–747 100. Nakahara J, Maeda M, Aiso S, Suzuki N. 2012. Current concepts in multiple sclerosis: Autoimmunity versus oligodendrogliopathy. Clin. Rev. Allergy Immunol. 42(1):26–34 101. Nikić I, Merkler D, Sorbara C, Brinkoetter M, Kreutzfeldt M, et al. 2011. A reversible form of 174 axon damage in experimental autoimmune encephalomyelitis and multiple sclerosis. Nat. Med. 17(4):495–99 102. Vercellino M, Masera S, Lorenzatti M, Condello C, Merola A, et al. 2009. Demyelination, inflammation, and neurodegeneration in multiple sclerosis deep gray matter. J. Neuropathol. Exp. Neurol. 68(5):489–502 103. Peterson JW, Bö L, Mörk S, Chang A, Trapp BD. 2001. Transected neurites, apoptotic neurons, and reduced inflammation in cortical multiple sclerosis lesions. Ann. Neurol. 50(3):389–400 104. Calabrese M, Magliozzi R, Ciccarelli O, Geurts JJG, Reynolds R, Martin R. 2015. Exploring the origins of grey matter damage in multiple sclerosis. Nat. Rev. Neurosci. 16(3):147–58 105. Howell OW, Rundle JL, Garg A, Komada M, Brophy PJ, Reynolds R. 2010. Activated Microglia Mediate Axoglial Disruption That Contributes to Axonal Injury in Multiple Sclerosis. J. Neuropathol. Exp. Neurol. 69(10):1017–33 106. Harauz G, Boggs JM. 2013. Myelin management by the 18.5-kDa and 21.5-kDa classic myelin basic protein isoforms. J. Neurochem. 125(3):334–61 107. Winterstein C, Trotter J, Krämer-Albers E-M. 2008. Distinct endocytic recycling of myelin proteins promotes oligodendroglial membrane remodeling. J. Cell Sci. 121(Pt 6):834–42 108. Quarles RH, Macklin WB, Morell P. 2006. Myelin Formation, Structure and Biochemistry. In Basic Neurochemistry: Molecular, Cellular and Medical Aspects, pp. 51–72. Elsevier 109. Fritz RB, Chou CH, McFarlin DE. 1983. Relapsing murine experimental allergic encephalomyelitis induced by myelin basic protein. J. Immunol. 130(3):1024–26 110. Pribyl TM, Campagnoni CW, Kampf K, Kashima T, Handley VW, et al. 1993. The human myelin basic protein gene is included within a 179-kilobase transcription unit: expression in the immune and central nervous systems. Proc. Natl. Acad. Sci. U. S. A. 90(22):10695–99 111. Givogri M, Bongarzone E, Campagnoni A. 2000. New insights on the biology of myelin basic protein gene: the neural-immune connection. J. Neurosci. Res. 59(2):153–59 112. Paez PM, Cheli VT, Ghiani CA, Spreuer V, Handley VW, Campagnoni AT. 2012. Golli myelin basic proteins stimulate oligodendrocyte progenitor cell proliferation and differentiation in remyelinating adult mouse brain. Glia. 60(7):1078–93 113. Paez PM, Fulton DJ, Spreuer V, Handley V, Campagnoni CW, et al. 2009. Golli myelin basic proteins regulate oligodendroglial progenitor cell migration through voltage-gated Ca2+ influx. J. Neurosci. 29(20):6663–76 114. Marty MC, Alliot F, Rutin J, Fritz R, Trisler D, Pessac B. 2002. The myelin basic protein gene is expressed in differentiated blood cell lineages and in hemopoietic progenitors. Proc. Natl. Acad. Sci. U. S. A. 99(13):8856–61 115. Feng JM, Givogri IM, Bongarzone ER, Campagnoni C, Jacobs E, et al. 2000. Thymocytes express the golli products of the myelin basic protein gene and levels of expression are stage dependent. J. Immunol. 165(10):5443–50 116. Feng JM, Fernandes AO, Campagnoni CW, Hu YH, Campagnoni AT. 2004. The golli-myelin basic protein negatively regulates signal transduction in T lymphocytes. J. Neuroimmunol. 152(1-2):57–66 175 117. Panitch HS, Hooper CJ, Johnson KP. 1980. CSF Antibody to Myelin Basic Protein: Measurement in Patients With Multiple Sclerosis and Subacute Sclerosing Panencephalitis. Arch. Neurol. 37:206–9 118. Reindl M, Linington C, Brehm U, Egg R, Dilitz E, et al. 1999. Antibodies against the myelin oligodendrocyte glycoprotein and the myelin basic protein in multiple sclerosis and other neurological diseases: a comparative study. Brain. 122:2047–56 119. Chou FCH, Jen Chou CH, Shapira R, Kibler RF. 1976. Basis of microheterogeneity of myelin basic protein. J. Biol. Chem. 251(9):2671–79 120. Moscarello MA, Wood DD, Ackerley C, Boulias C. 1994. Myelin in multiple sclerosis is developmentally immature. J. Clin. Invest. 94(1):146–54 121. Boggs JM. 2006. Myelin basic protein: A multifunctional protein. Cell. Mol. Life Sci. 63(17):1945–61 122. Lee DW, Banquy X, Kristiansen K, Kaufman Y, Boggs JM, Israelachvili JN. 2014. Lipid domains control myelin basic protein adsorption and membrane interactions between model myelin lipid bilayers. Proc. Natl. Acad. Sci. U. S. A. 111(8):E768–75 123. Ulmer JB, Braun PE. 1986. In vivo phosphorylation of myelin basic proteins: single and double isotope incorporation in developmentally related myelin fractions. Dev. Biol. 117(2):502–10 124. Kim JK, Mastronardi FG, Wood DD, Lubman DM, Zand R, Moscarello MA. 2003. Multiple Sclerosis, An important role for post-translational modifications of myelin basic protein in pathogenesis. Mol. Cell. Proteomics. 2(7):453–62 125. Lu Z, Ku L, Chen Y, Feng Y. 2005. Developmental abnormalities of myelin basic protein expression in fyn knock-out brain reveal a role of fyn in posttranscriptional regulation. J. Biol. Chem. 280(1):389–95 126. Umemori H, Kadowaki Y, Hirosawa K, Yoshida Y, Hironaka K, et al. 1999. Stimulation of myelin basic protein gene transcription by Fyn tyrosine kinase for myelination. J. Neurosci. 19(4):1393–97 127. Polverini E, Rangaraj G, Libich DS, Boggs JM, Harauz G. 2008. Binding of the proline-rich segment of myelin basic protein to SH3 domains: spectroscopic, microarray, and modeling studies of ligand conformation and effects of posttranslational modifications. Biochemistry. 47(1):267–82 128. Smith GST, Homchaudhuri L, Boggs JM, Harauz G. 2012. Classic 18.5- and 21.5-kDa myelin basic protein isoforms associate with cytoskeletal and SH3-domain proteins in the immortalized N19-oligodendroglial cell line stimulated by phorbol ester and IGF-1. Neurochem. Res. 37(6):1277–95 129. Smith GST, De Avila M, Paez PM, Spreuer V, Wills MKB, et al. 2012. Proline substitutions and threonine pseudophosphorylation of the SH3 ligand of 18.5-kDa myelin basic protein decrease its affinity for the Fyn-SH3 domain and alter process development and protein localization in oligodendrocytes. J. Neurosci. Res. 90(1):28–47 130. Homchaudhuri L, Polverini E, Gao W, Harauz G, Boggs JM. 2009. Influence of membrane surface charge and post-translational modifications to myelin basic protein on its ability to tether the Fyn-SH3 domain to a membrane in vitro. Biochemistry. 48(11):2385–93 176 131. De Avila M, Vassall KA, Smith GST, Bamm VV, Harauz G. 2014. The proline-rich region of 18.5 kDa myelin basic protein binds to the SH3-domain of Fyn tyrosine kinase with the aid of an upstream segment to form a dynamic complex in vitro. Biosci. Rep. 34(6):775–88 132. Lucchinetti C, Bruck W, Parisi J, Scheithauer B, Rodriguez M, Lassmann H. 2000. Heterogeneity of Multiple Sclerosis Lesions: Implications for the Pathogenesis of Demyelination. Ann. Neurol. 47:707–17 133. Barnett MH, Prineas JW. 2004. Relapsing and Remitting Multiple Sclerosis: Pathology of the Newly Forming Lesion. Ann. Neurol. 55:458–68 134. Henderson APD, Barnett MH, Parratt JDE, Prineas JW. 2009. Multiple sclerosis Distribution of inflammatory Cells in Newly Forming Lesions. Ann. Neurol. 66:739–53 135. Stys PK, Sontheimer H, Ransom BR, Waxman SG. 1993. Noninactivating, tetrodotoxin-sensitive Na+ conductance in rat optic nerve axons. Proc. Natl. Acad. Sci. U. S. A. 90(15):6976–80 136. Berger T, Reindl M. 2007. Multiple sclerosis: Disease biomarkers as indicated by pathophysiology. J. Neurol. Sci. 259(1-2):21–26 137. Mukhopadhyay G, Doherty P, Walsh FS, Crocker PR, Filbin MT. 1994. A novel role for myelin-associated glycoprotein as an inhibitor of axonal regeneration. Neuron. 13(3):757–67 138. Cai D, Qiu J, Cao Z, McAtee M, Bregman BS, Filbin MT. 2001. Neuronal cyclic AMP controls the developmental loss in ability of axons to regenerate. J. Neurosci. 21(13):4731–39 139. Johns TG, Bernard CCA. 1999. The structure and function of myelin oligodendrocyte glycoprotein. J. Neurochem. 72(1):1–9 140. Grandpre T, Nakamura F, Vartanian T, Strittmatter SM. 2000. Identification of the Nogo inhibitor of axon regeneration as a Reticulon protein. Nature. 403(27):439–44 141. Voeltz GK, Prinz WA, Shibata Y, Rist JM, Rapoport TA. 2006. A class of membrane proteins shaping the tubular endoplasmic reticulum. Cell. 124(3):573–86 142. Zemmar A, Weinmann O, Kellner Y, Yu X, Vicente R, et al. 2014. Neutralization of nogo-a enhances synaptic plasticity in the rodent motor cortex and improves motor learning in vivo. J. Neurosci. 34(26):8685–98 143. Huber AB, Weinmann O, Brösamle C, Oertle T, Schwab ME. 2002. Patterns of Nogo mRNA and protein expression in the developing and adult rat and after CNS lesions. J. Neurosci. 22(9):3553–67 144. Oertle T, van der Haar ME, Bandtlow CE, Robeva A, Burfeind P, et al. 2003. NogoA inhibits neurite outgrowth and cell spreading with three discrete regions. J. Neurosci. 23(13):5393–5406 145. Kempf A, Tews B, Arzt ME, Weinmann O, Obermair FJ, et al. 2014. The Sphingolipid Receptor S1PR2 Is a Receptor for Nogo-A Repressing Synaptic Plasticity. PLoS Biol. 12(1): 146. Thiede-Stan NK. 2015. Composition and dynamics of the Nogo-A-Δ20 receptor complex. ETH-Zürich 147. Thiede-stan NK, Tews B, Albrecht D, Ristic Z, Ewers H. 2015. Tetraspanin-3 is an organizer of the multi-subunit Nogo-A signaling complex. J. Cell Sci. 128(19):3583–96 148. Sepe M, Lignitto L, Porpora M, Delle Donne R, Rinaldi L, et al. 2014. Proteolytic control of neurite outgrowth inhibitor NOGO-A by the cAMP/PKA pathway. Proc. Natl. Acad. Sci. U. S. 177 A. 111(44):15729–34 149. Qin H, Pu HX, Li M, Ahmed S, Song J. 2008. Identification and structural mechanism for a novel interaction between a ubiquitin ligase WWP1 and Nogo-A, a Key inhibitor for central nervous system regeneration. Biochemistry. 47(51):13647–58 150. Schmandke A, Schmandke A, Schwab ME. 2014. Nogo-A: Multiple Roles in CNS Development, Maintenance, and Disease. Neurosci. 20(4):372–86 151. Satoh J-I, Onoue H, Arima K, Yamamura T. 2005. Nogo-A and nogo receptor expression in demyelinating lesions of multiple sclerosis. J. Neuropathol. Exp. Neurol. 64(2):129–38 152. Karnezis T, Mandemakers W, McQualter JL, Zheng B, Ho PP, et al. 2004. The neurite outgrowth inhibitor Nogo A is involved in autoimmune-mediated demyelination. Nat. Neurosci. 7(7):736–44 153. Yang Y, Liu Y, Wei P, Peng H, Winger R, et al. 2010. Silencing Nogo-A promotes functional recovery in demyelinating disease. Ann. Neurol. 67(4):498–507 154. Chong SYC, Rosenberg SS, Fancy SPJ, Zhao C, Shen Y-AA, et al. 2012. From the Cover: Neurite outgrowth inhibitor Nogo-A establishes spatial segregation and extent of oligodendrocyte myelination. Proc. Natl. Acad. Sci. U. S. A. 109(4):1299–1304 155. Pernet V, Joly S, Christ F, Dimou L, Schwab ME. 2008. Nogo-A and myelin-associated glycoprotein differently regulate oligodendrocyte maturation and myelin formation. J. Neurosci. 28(29):7435–44 156. Waxman SG. 2006. Axonal conduction and injury in multiple sclerosis: the role of sodium channels. Nat. Rev. Neurosci. 7(12):932–41 157. Moll C, Mourre C, Lazdunski M, Ulrich J. 1991. Increase of sodium channels in demyelinated lesions of multiple sclerosis. Brain Res. 556(2):311–16 158. Bouafia A, Golmard JL, Thuries V, Sazdovitch V, Hauw JJ, et al. 2014. Axonal expression of sodium channels and neuropathology of the plaques in multiple sclerosis. Neuropathol. Appl. Neurobiol. 40(5):579–90 159. Craner MJ, Newcombe J, Black JA, Hartle C, Cuzner ML, Waxman SG. 2004. Molecular changes in neurons in multiple sclerosis: Altered axonal expression of Nav1.2 and Nav1.6 sodium channels and Na+/Ca2+ exchanger. Proc. Natl. Acad. Sci. U. S. A. 101(21):8168–73 160. Kaplan M, Cho M, Ullian E, Isom L. 2001. Differential Control of Clustering of the Sodium Channels Nav1.2 and Nav1.6 at Developing CNS Nodes of Ranvier. Neuron. 30(1):105–19 161. Boiko T, Rasband MN, Levinson SR, Caldwell JH, Mandel G, et al. 2001. Compact myelin dictates the differential targeting of two sodium channel isoforms in the same axon. Neuron. 30(1):91–104 162. Bostock BYH, Sears TA. 1978. The internodal axon membrane: elctrical excitability and continuous conduction in segmental demyelination. J. Physiol. 280:273–301 163. Foster RE, Whalen CC, Waxman SG. 1980. Reorganization of the Axon Membrane in Demyelinated Peipheral Nerve Fibers: Morphological Evidence. Science. 210:661–63 164. Black JA, Liu S, Waxman SG. 2009. Sodium channel activity modulates multiple functions in microglia. Glia. 57(10):1072–81 178 165. Morsali D, Bechtold D, Lee W, Chauhdry S, Palchaudhuri U, et al. 2013. Safinamide and flecainide protect axons and reduce microglial activation in models of multiple sclerosis. Brain. 136(4):1067–82 166. Craner MJ, Damarjian TG, Liu S, Hains BC, Lo AC, et al. 2005. Sodium Channels Contribute to Microglia or Macrophage Activation and Function in EAE and MS. Glia. 49:220–29 167. Black JA, Waxman SG. 2012. Sodium channels and microglial function. Exp. Neurol. 234:302–15 168. Rotin D, Staub O, Haguenauer-Tsapis R. 2000. Ubiquitination and endocytosis of plasma membrane proteins: Role of Nedd4/Rsp5p family of ubiquitin-protein ligases. J. Membr. Biol. 176(1):1–17 169. Carrithers MD, Dib-Hajj S, Carrithers LM, Tokmoulina G, Pypaert M, et al. 2007. Expression of the voltage-gated sodium channel Nav1.5 in the macrophage late endosome regulates endosomal acidification. J. Immunol. 178(12):7822–32 170. Nakahara J, Nakahara J, Tan-takeuchi K, Tan-takeuchi K, Seiwa C, et al. 2001. Myelin basic protein is necessary for the regulation of myelin- associated glycoprotein expression in mouse oligodendroglia. Neurosci. Lett. 298:163–66 171. Liao H, Duka T, Teng FYH, Sun L, Bu WY, et al. 2004. Nogo-66 and myelin-associated glycoprotein (MAG) inhibit the adhesion and migration of Nogo-66 receptor expressing human glioma cells. J. Neurochem. 90(5):1156–62 172. Sullivan BM. 2009. Immune Modulation by Human Roseoloviruses. University of California, Berkeley. 155 pp. 173. Ball LJ, Kühne R, Schneider-Mergener J, Oschkinat H. 2005. Recognition of proline-rich motifs by protein-protein-interaction domains. Angew. Chemie Int. Ed. 44(19):2852–69 174. Pires JR, Taha-nejad F, Toepert F, Ast T, Hoffmu U, et al. 2001. Solution Structures of the YAP65 WW Domain and the Variant L30 K in Complex with the Peptides GTPPPPYTVG, N-(n-octyl)-GPPPY and PLPPY and the Application of Peptide Libraries Reveal a Minimal Binding Epitope. J. Mol. Biol. 314:1147–56 175. Vina-Vilaseca A, Bender-Sigel J, Sorkina T, Closs EI, Sorkin A. 2011. Protein kinase C-dependent ubiquitination and clathrin-mediated endocytosis of the cationic amino acid transporter CAT-1. J. Biol. Chem. 286(10):8697–8706 176. García-Tardón N, González-González IM, Martínez-Villarreal J, Fernández-Sánchez E, Giménez C, Zafra F. 2012. Protein kinase C (PKC)-promoted endocytosis of glutamate transporter GLT-1 requires ubiquitin ligase Nedd4-2-dependent ubiquitination but not phosphorylation. J. Biol. Chem. 287(23):19177–87 177. Andersen MN, Skibsbye L, Tang C, Petersen F, MacAulay N, et al. 2015. PKC and AMPK regulation of Kv1.5 potassium channels. Channels. 9(3):121–28 178. Booth RE, Stockand JD. 2003. Targeted degradation of ENaC in response to PKC activation of the ERK1/2 cascade. Am. J. Physiol. Renal Physiol. 284(5):F938–47 179. Bogdanik LP, Sleigh JN, Tian C, Samuels ME, Bedard K, et al. 2013. Loss of the E3 ubiquitin ligase LRSAM1 sensitizes peripheral axons to degeneration in a mouse model of Charcot-Marie-Tooth disease. Dis. Model. Mech. 6(3):780–92 179 180. Noble ME, Musacchio A, Saraste M, Courtneidge SA, Wierenga RK. 1993. Crystal structure of the SH3 domain in human Fyn; comparison of the three-dimensional structures of SH3 domains in tyrosine kinases and spectrin. EMBO J. 12(7):2617–24 181. Rickles RJ, Botfield MC, Weng Z, Taylor JA, Green OM, et al. 1994. Identification of Src, Fyn, Lyn, PI3K and Abl SH3 domain ligands using phage display libraries. EMBO J. 13(23):5598–5604 182. Carducci M, Perfetto L, Briganti L, Paoluzi S, Costa S, et al. 2012. The protein interaction network mediated by human SH3 domains. Biotechnol. Adv. 30(1):4–15 183. Musacchio A, Saraste M, Wilmanns M. 1994. High-resolution crystal structures of tyrosine kinase SH3 domains complexed with proline-rich peptides. Nat. Struct. Biol. 1(8):546–51 184. Larson SM, Davidson AR. 2000. The identification of conserved interactions within the SH3 domain by alignment of sequences and structures. Protein Sci. 9(11):2170–80 185. Shelton H, Harris M. 2008. Hepatitis C virus NS5A protein binds the SH3 domain of the Fyn tyrosine kinase with high affinity: mutagenic analysis of residues within the SH3 domain that contribute to the interaction. Virol. J. 5:24 186. Solheim SA, Petsalaki E, Stokka AJ, Russell RB, Taskén K, Berge T. 2008. Interactions between the Fyn SH3-domain and adaptor protein Cbp/PAG derived ligands, effects on kinase activity and affinity. FEBS J. 275(19):4863–74 187. Morton CJ, Pugh DJ, Brown EL, Kahmann JD, Renzoni DA, Campbell ID. 1996. Solution structure and peptide binding of the SH3 domain from human Fyn. Structure. 4(6):705–14 188. Kang H, Freund C, Duke-Cohan JS, Musacchio A, Wagner G, Rudd CE. 2000. SH3 domain recognition of a proline-independent tyrosine-based RKxxYxxY motif in immune cell adaptor SKAP55. EMBO J. 19(12):2889–99 189. Baker NA, Sept D, Joseph S, Holst MJ, McCammon JA. 2001. Electrostatics of nanosystems: application to microtubules and the ribosome. Proc. Natl. Acad. Sci. U. S. A. 98(18):10037–41 190. Martin GS. 2001. The hunting of the Src. Nat. Rev. Mol. Cell Biol. 2(6):467–75 191. Resh MD. 1998. Fyn, a Src family tyrosine kinase. Int. J. Biochem. Cell Biol. 30(11):1159–62 192. Rawat A, Nagaraj R. 2010. Determinants of membrane association in the SH4 domain of Fyn: Roles of N-terminus myristoylation and side-chain thioacylation. Biochim. Biophys. Acta - Biomembr. 1798(10):1854–63 193. Nada S, Yagi T, Takeda H, Tokunaga T, Nakagawa H, et al. 1993. Constitutive activation of Src family kinases in mouse embryos that lack Csk. Cell. 73(6):1125–35 194. Bergman M, Mustelin T, Oetken C, Partanen J, Flint NA, et al. 1992. The human p50csk tyrosine kinase phosphorylates p56lck at Tyr-505 and down regulates its catalytic activity. EMBO J. 11(8):2919–24 195. Umemori H, Sato S, Yagi T, Aizawa S, Yamamoto T. 1994. Initial events of myelination involve Fyn tyrosine kinase signalling. Nature. 367:572–76 196. Sperber BR, Boyle-Walsh EA, Engleka MJ, Gadue P, Peterson AC, et al. 2001. A unique role for Fyn in CNS myelination. J. Neurosci. 21(6):2039–47 197. Biffiger K, Bartsch S, Montag D, Aguzzi A, Schachner M, Bartsch U. 2000. Severe 180 hypomyelination of the murine CNS in the absence of myelin-associated glycoprotein and fyn tyrosine kinase. J. Neurosci. 20(19):7430–37 198. Morita A, Yamashita N, Sasaki Y, Uchida Y, Nakajima O, et al. 2006. Regulation of dendritic branching and spine maturation by semaphorin3A-Fyn signaling. J. Neurosci. 26(11):2971–80 199. Colognato H, Ramachandrappa S, Olsen IM, Ffrench-Constant C. 2004. Integrins direct Src family kinases to regulate distinct phases of oligodendrocyte development. J. Cell Biol. 167(2):365–75 200. Colognato H, Baron W, Avellana-Adalid V, Relvas JB, Baron-Van Evercooren A, et al. 2002. CNS integrins switch growth factor signalling to promote target-dependent survival. Nat. Cell Biol. 4(11):833–41 201. Krämer-Albers E-M, White R. 2011. From axon-glial signalling to myelination: the integrating role of oligodendroglial Fyn kinase. Cell. Mol. Life Sci. 68(12):2003–12 202. Liang X, Draghi NA, Resh MD. 2004. Signaling from integrins to Fyn to Rho family GTPases regulates morphologic differentiation of oligodendrocytes. J. Neurosci. 24(32):7140–49 203. Bryan BA, Li D, Wu X, Liu M. 2005. The Rho family of small GTPases: Crucial regulators of skeletal myogenesis. Cell. Mol. Life Sci. 62(14):1547–55 204. Klein C, Kramer E-M, Cardine A-M, Schraven B, Brandt R, Trotter J. 2002. Process outgrowth of oligodendrocytes is promoted by interaction of fyn kinase with the cytoskeletal protein tau. J. Neurosci. 22(3):698–707 205. Belkadi A, LoPresti P. 2008. Truncated Tau with the Fyn-binding domain and without the microtubule-binding domain hinders the myelinating capacity of an oligodendrocyte cell line. J. Neurochem. 107(2):351–60 206. Grant SG, O’Dell TJ, Karl KA, Stein PL, Soriano P, Kandel ER. 1992. Impaired long-term potentiation, spatial learning, and hippocampal development in fyn mutant mice. Science. 258(5090):1903–10 207. Osterhout DJ, Wolven A, Wolf RM, Resh MD, Chao MV. 1999. Morphological Differentiation of Oligodendrocytes Requires Activation of Fyn Tyrosine Kinase. J. Cell Biol. 145:1209–18 208. Palacios EH, Weiss A. 2004. Function of the Src-family kinases, Lck and Fyn, in T-cell development and activation. Oncogene. 23(48):7990–8000 209. Gadue P, Morton N, Stein PL. 1999. The Src family tyrosine kinase Fyn regulates natural killer T cell development. J. Exp. Med. 190(8):1189–96 210. Roche S, Fumagalli S, Courtneidge SA, Roche S, Fumagalli S, Courtneidgett SA. 1995. Requirement for Src FamilyProtein Tyrosine Kinases in G2 for Fibroblast Cell Division. Science. 269(5230):1567–69 211. Levi M, Maro B, Shalgi R. 2010. Fyn kinase is involved in cleavage furrow ingression during meiosis and mitosis. Reproduction. 140(6):827–34 212. Chong PA, Lin H, Wrana JL, Forman-Kay JD. 2010. Coupling of tandem Smad ubiquitination regulatory factor (Smurf) WW domains modulates target specificity. Proc. Natl. Acad. Sci. U. S. A. 107(43):18404–9 213. Macias MJ, Hyvönen M, Baraldi E, Schultz J, Sudol M, et al. 1996. Structure of the WW 181 domain of a kinase-associated protein complexed with a proline-rich peptide. 214. Kanelis V, Bruce MC, Skrynnikov NR, Rotin D, Forman-Kay JD. 2006. Structural determinants for high-affinity binding in a Nedd4 WW3* domain-comm PY motif complex. Structure. 14(3):543–53 215. Aragón E, Goerner N, Zaromytidou A-I, Xi Q, Escobedo A, et al. 2011. A Smad action turnover switch operated by WW domain readers of a phosphoserine code. Genes Dev. 25(12):1275–88 216. Morales B, Ramirez-Espain X, Shaw AZ, Martin-Malpartida P, Yraola F, et al. 2007. NMR structural studies of the ItchWW3 domain reveal that phosphorylation at T30 inhibits the interaction with PPxY-containing ligands. Structure. 15(4):473–83 217. Sotgia F, Lee H, Bedford MT, Petrucci T, Sudol M, Lisanti MP. 2001. Tyrosine phosphorylation of β-dystroglycan at its WW domain binding motif, PPxY, recruits SH2 domain containing proteins. Biochemistry. 40(48):14585–92 218. Otte L, Wiedemann U, Schlegel B, Pires JR, Beyermann M, et al. 2003. WW domain sequence activity relationships identified using ligand recognition propensities of 42 WW domains. Protein Sci. 12(3):491–500 219. Lu PJ, Zhou XZ, Shen M, Lu KP. 1999. Function of WW domains as phosphoserine- or phosphothreonine-binding modules. Science. 283(5406):1325–28 220. Kumar S, Tomooka Y, Noda M. 1992. Identification of a Set of Genes with Developmentally Down-regulated Expression in the Mouse Brain. Biochem. Biophys. Res. Commun. 185(3):1155–61 221. An H, Krist DT, Statsyuk AV. 2014. Crosstalk between kinases and Nedd4 family ubiquitin ligases. Mol. Biosyst. 10(7):1643–57 222. Dunn R, Klos DA, Adler AS, Hicke L. 2004. The C2 domain of the Rsp5 ubiquitin ligase binds membrane phosphoinositides and directs ubiquitination of endosomal cargo. J. Cell Biol. 165(1):135–44 223. Angers A, Ramjaun AR, McPherson PS. 2004. The HECT Domain Ligase Itch Ubiquitinates Endophilin and Localizes to the trans-Golgi Network and Endosomal System. J. Biol. Chem. 279(12):11471–79 224. Huibregtse JM, Scheffner M, Beaudenon S, Howley PM. 1995. A family of proteins structurally and functionally related to the E6-AP ubiquitin-protein ligase. Proc. Natl. Acad. Sci. U. S. A. 92(7):2563–67 225. Rotin D, Kumar S. 2009. Physiological functions of the HECT family of ubiquitin ligases. Nat. Rev. Mol. Cell Biol. 10(6):398–409 226. Boase NA, Kumar S. 2015. NEDD4: The founding member of a family of ubiquitin-protein ligases. Gene. 557(2):113–22 227. Davey NE, Travé G, Gibson TJ. 2011. How viruses hijack cell regulation. Trends Biochem. Sci. 36(3):159–69 228. Matesic LE, Copeland NG, Jenkins N a. 2008. Itchy mice: The identification of a new pathway for the development of autoimmunity. Curr. Top. Microbiol. Immunol. 321:185–200 229. Lohr NJ, Molleston JP, Strauss KA, Torres-Martinez W, Sherman EA, et al. 2010. Human 182 ITCH E3 Ubiquitin Ligase Deficiency Causes Syndromic Multisystem Autoimmune Disease. Am. J. Hum. Genet. 86(3):447–53 230. Izzi L, Attisano L. 2004. Regulation of the TGFbeta signalling pathway by ubiquitin-mediated degradation. Oncogene. 23(11):2071–78 231. Trotman LC, Wang X, Alimonti A, Chen Z, Teruya-Feldstein J, et al. 2007. Ubiquitination Regulates PTEN Nuclear Import and Tumor Suppression. Cell. 128(1):141–56 232. Maddika S, Kavela S, Rani N, Palicharla VR, Pokorny JL, et al. 2011. WWP2 is an E3 ubiquitin ligase for PTEN. Nat. Cell Biol. 13(6):728–33 233. Wang X, Trotman LC, Koppie T, Alimonti A, Chen Z, et al. 2007. NEDD4-1 Is a Proto-Oncogenic Ubiquitin Ligase for PTEN. Cell. 128(1):129–39 234. Hsia H-E, Kumar R, Luca R, Takeda M, Courchet J, et al. 2014. Ubiquitin E3 ligase Nedd4-1 acts as a downstream target of PI3K/PTEN-mTORC1 signaling to promote neurite growth. Proc. Natl. Acad. Sci. U. S. A. 111(36):13205–10 235. Ahn Y, Hwang CY, Lee S-R, Kwon K-S, Lee C. 2008. The tumour suppressor PTEN mediates a negative regulation of the E3 ubiquitin-protein ligase Nedd4. Biochem. J. 412(2):331–38 236. Ichimura T, Yamamura H, Sasamoto K, Tominaga Y, Taoka M, et al. 2005. 14-3-3 Proteins modulate the expression of epithelial Na+ channels by phosphorylation-dependent interaction with Nedd4-2 ubiquitin ligase. J. Biol. Chem. 280(13):13187–94 237. Zhou R, Snyder PM. 2005. Nedd4-2 Phosphorylation Induces Serum and Glucocorticoid-regulated Kinase (SGK) Ubiquitination and Degradation. J. Biol. Chem. 280(6):4518–23 238. Wiesner S, Ogunjimi AA, Wang HR, Rotin D, Sicheri F, et al. 2007. Autoinhibition of the HECT-Type Ubiquitin Ligase Smurf2 through Its C2 Domain. Cell. 130(4):651–62 239. Wan L, Zou W, Gao D, Inuzuka H, Fukushima H, et al. 2011. Cdh1 regulates osteoblast function through an APC/C-independent modulation of smurf1. Mol. Cell. 44(5):721–33 240. Wang J, Peng Q, Lin Q, Childress C, Carey D, Yang W. 2010. Calcium activates Nedd4 E3 ubiquitin ligases by releasing the C2 domain-mediated auto-inhibition. J. Biol. Chem. 285(16):12279–88 241. Cui G, Wei P, Zhao Y, Guan Z, Yang L, et al. 2014. Brucella infection inhibits macrophages apoptosis via Nedd4-dependent degradation of calpain2. . 174:195–205 242. Mund T, Pelham HRB. 2009. Control of the activity of WW-HECT domain E3 ubiquitin ligases by NDFIP proteins. EMBO Rep. 10(5):501–7 243. Persaud A, Alberts P, Mari S, Tong J, Murchie R, et al. 2014. Tyrosine phosphorylation of NEDD4 activates its ubiquitin ligase activity. Sci. Signal. 7(346):1–11 244. Gallagher E, Gao M, Liu Y, Karin M. 2010. Activation of the E3 ubiquitin ligase Itch through a phosphorylation-induced conformational change. . 107(16):7616 245. Scialpi F, Malatesta M, Peschiaroli A, Rossi M, Melino G, Bernassola F. 2008. Itch self-polyubiquitylation occurs through lysine-63 linkages. Biochem. Pharmacol. 76(11):1515–21 246. Fukunaga E, Inoue Y, Komiya S, Horiguchi K, Goto K, et al. 2008. Smurf2 induces ubiquitin-dependent degradation of Smurf1 to prevent migration of breast cancer cells. J. Biol. Chem. 283(51):35660–67 183 247. Escobedo A, Gomes T, Aragón E, Martín-Malpartida P, Ruiz L, Macias MJ. 2014. Structural Basis of the Activation and Degradation Mechanisms of the E3 Ubiquitin Ligase Nedd4L. Structure. 22(10):1446–57 248. Kannan M, Lee S-J, Schwedhelm-Domeyer N, Stegmuller J. 2012. The E3 ligase Cdh1-anaphase promoting complex operates upstream of the E3 ligase Smurf1 in the control of axon growth. Development. 139(19):3600–3612 249. Cui Y, He S, Xing C, Lu K, Wang J, et al. 2011. SCFFBXL15 regulates BMP signalling by directing the degradation of HECT-type ubiquitin ligase Smurf1. EMBO J. 30(13):2675–89 250. Liu J, Wan L, Liu P, Inuzuka H, Liu J, et al. 2014. SCFβ-TRCP-mediated degradation of NEDD4 inhibits tumorigenesis through modulating the PTEN/Akt signaling pathway. Oncotarget. 5(4):1026–37 251. Malakhova OA, Zhang DE. 2008. ISG15 inhibits Nedd4 ubiquitin E3 activity and enhances the innate antiviral response. J. Biol. Chem. 283(14):8783–87 252. Boase NA, Rychkov GY, Townley SL, Dinudom A, Candi E, et al. 2011. Respiratory distress and perinatal lethality in Nedd4-2-deficient mice. Nat. Commun. 2:287 253. Cao XR, Lill NL, Boase N, Shi PP, Croucher DR, et al. 2008. Nedd4 controls animal growth by regulating IGF-1 signaling. Sci. Signal. 1(38):1–10 254. Donovan P, Poronnik P. 2013. Nedd4 and Nedd4-2: Ubiquitin ligases at work in the neuron. Int. J. Biochem. Cell Biol. 45(3):706–10 255. Camera D, Boase NA, Kumar S, Pow D V, Poronnik P. 2013. Subtle gait abnormalities in Nedd4 heterozygous mice. Behav. Brain Res. 260:1–10 256. Kawabe H, Neeb A, Dimova K, Young SM, Takeda M, et al. 2010. Regulation of Rap2A by the Ubiquitin Ligase Nedd4-1 Controls Neurite Development. Neuron. 65(3):358–72 257. Drinjakovic J, Jung H, Campbell DS, Strochlic L, Dwivedy A, Holt CE. 2010. E3 Ligase Nedd4 Promotes Axon Branching by Downregulating PTEN. Neuron. 65(3):341–57 258. Liu K, Lu Y, Lee JK, Samara R, Willenberg R, et al. 2010. PTEN deletion enhances the regenerative ability of adult corticospinal neurons. Nat. Neurosci. 13(9):1075–81 259. Georgieva MV., de Pablo Y, Sanchis D, Comella JX, Llovera M. 2011. Ubiquitination of TrkA by Nedd4-2 regulates receptor lysosomal targeting and mediates receptor signaling. J. Neurochem. 117(3):479–93 260. Arévalo JC, Waite J, Rajagopal R, Beyna M, Chen ZY, et al. 2006. Cell Survival through Trk Neurotrophin Receptors Is Differentially Regulated by Ubiquitination. Neuron. 50(4):549–59 261. Fotia AB, Ekberg J, Adams DJ, Cook DI, Poronnik P, Kumar S. 2004. Regulation of neuronal voltage-gated sodium channels by the ubiquitin-protein ligases Nedd4 and Nedd4-2. J. Biol. Chem. 279(28):28930–35 262. Rougier JS, Albesa M, Abriel H, Viard P. 2011. Neuronal precursor cell-expressed developmentally down-regulated 4-1 (NEDD4-1) controls the sorting of newly synthesized Cav1.2 calcium channels. J. Biol. Chem. 286(11):8829–38 263. Schuetz F, Kumar S, Poronnik P, Adams DJ. 2008. Regulation of the voltage-gated K channels KCNQ2/3 and KCNQ3/5 by serum- and glucocorticoid-regulated kinase-1. Am. J. Physiol. - 184 Cell Physiol., pp. 73–80 264. Lin A, Hou Q, Jarzylo L, Amato S, Gilbert J, et al. 2011. Nedd4-mediated AMPA receptor ubiquitination regulates receptor turnover and trafficking. J. Neurochem. 119(1):27–39 265. Sopjani M, Alesutan I, Dërmaku-Sopjani M, Fraser S, Kemp BE, et al. 2010. Down-regulation of Na+-coupled glutamate transporter EAAT3 and EAAT4 by AMP-activated protein kinase. J. Neurochem. 113:1426–35 266. Boehmer C, Henke G, Schniepp R, Palmada M, Rothstein JD, et al. 2003. Regulation of the glutamate transporter EAAT1 by the ubiquitin ligase Nedd4-2 and the serum and glucocorticoid-inducible kinase isoforms SGK1/3 and protein kinase B. J. Neurochem. 86(5):1181–88 267. Vina-Vilaseca A, Sorkin A. 2010. Lysine 63-linked polyubiquitination of the dopamine transporter requires WW3 and WW4 domains of Nedd4-2 and UBE2D ubiquitin-conjugating enzymes. J. Biol. Chem. 285(10):7645–56 268. Mukhopadhyay D, Riezman H. 2007. Proteasome-independent functions of ubiquitin in endocytosis and signaling. Science. 315(5809):201–5 269. Pickart CM, Fushman D. 2004. Polyubiquitin chains: Polymeric protein signals. Curr. Opin. Chem. Biol. 8(6):610–16 270. Komander D, Clague MJ, Urbé S. 2009. Breaking the chains: structure and function of the deubiquitinases. Nat. Rev. Mol. Cell Biol. 10(8):550–63 271. Wiborg O, Pedersen MS, Wind A, Berglund LE, Marcker KA, Vuust J. 1985. The human ubiquitin multigene family: some genes contain multiple directly repeated ubiquitin coding sequences. EMBO J. 4(3):755–59 272. Kerscher O, Felberbaum R, Hochstrasser M. 2006. Modification of proteins by ubiquitin and ubiquitin-like proteins. Annu. Rev. Cell Dev. Biol. 22:159–80 273. Koegl M, Hoppe T, Schlenker S, Ulrich HD, Mayer TU, Jentsch S. 1999. A novel ubiquitination factor, E4, is involved in multiubiquitin chain assembly. Cell. 96(5):635–44 274. Cadwell K, Coscoy L. 2005. Ubiquitination on nonlysine residues by a viral E3 ubiquitin ligase. Science. 309(5731):127–30 275. Wang X, Herr RA., Chua WJ, Lybarger L, Wiertz EJHJ, Hansen TH. 2007. Ubiquitination of serine, threonine, or lysine residues on the cytoplasmic tail can induce ERAD of MHC-I by viral E3 ligase mK3. J. Cell Biol. 177(4):613–24 276. Ciechanover A, Ben-Saadon R. 2004. N-terminal ubiquitination: More protein substrates join in. Trends Cell Biol. 14(3):103–6 277. Wang X, Herr RA, Hansen TH. 2012. Ubiquitination of substrates by esterification. Traffic. 13(1):19–24 278. Nakayama KI, Nakayama K. 2006. Ubiquitin ligases: cell-cycle control and cancer. Nat. Rev. Cancer. 6(5):369–81 279. Saritas-Yildirim B, Silva EM. 2014. The role of targeted protein degradation in early neural development. Genesis. 299(March):287–99 280. Kay LE, Keifer P, Saarinen T. 1992. Pure Absorption Gradient Enhanced Heteronuclear Single 185 Quantum Correlation Spectroscopy with Improved Sensitivity. J. Am. Chem. Soc., pp. 10663–65 281. Palmer AGI, Cavanagh J, Wright PE, Rance M. 1991. Sensitivity Improvement in Proton-Detected Two-Dimensional Heteronuclear Correlation NMR Spectroscopy. J. Magn. Reson. 93:151–70 282. Schleucher J, Schwendinger M, Sattler M, Schmidt P, Schedletzky O, et al. 1994. A general enhancement scheme in heteronuclear multidimensional NMR employingpulse field gradients. J. Biomol. NMR. 4:301–6 283. Grzesiek S, Bax A. 1993. The Importance of Not Saturating H2O in Protein NMR. Application to Sensitivity Enhancement and NOE Measurements. J. Am. Chem. Soc. 115:12593–94 284. Piotto M, Saudek V, Sklenář V. 1992. Gradient-tailored excitation for single-quantum NMR spectroscopy of aqueous solutions. J. Biomol. NMR. 2(6):661–65 285. Pervushin K, Riek R, Wider G, Wüthrich K. 1997. Attenuated T2 relaxation by mutual cancellation of dipole-dipole coupling and chemical shift anisotropy indicates an avenue to NMR structures of very large biological macromolecules in solution. Proc. Natl. Acad. Sci. U. S. A. 94(23):12366–71 286. Cavanagh J, Fairbrother WJ, Palmer AGI, Rance M, Skelton NJ. 2007. Protein NMR Spectroscopy: Principles and Practice. San Diego, CA: Academic press. 885 pp. second ed. 287. Mittermaier AK, Kay LE. 2009. Observing biological dynamics at atomic resolution using NMR. Trends Biochem. Sci. 34(12):601–11 288. Wider G, Neri D, Wuthrich K. 1991. Studies of slow conformational equilibria in macromolecules by exchange of heteronuclear longitudinal 2-spin-order in a 2D difference correlation experiment. J. Biomol. NMR. 1:93–98 289. Farrow NA, Zhang O, Forman-Kay JD, Kay LE. 1994. A heteronuclear correlation experiment for simultaneous determination of 15N longitudinal decay and chemical exchange rates of systems in slow equilibrium. J. Biomol. NMR. 4(5):727–34 290. Robson SA, Peterson R, Bouchard L-S, Villareal VA, Clubb RT. 2010. A heteronuclear zero quantum coherence Nz-exchange experiment that resolves resonance overlap and its application to measure the rates of heme binding to the IsdC protein Material includes: S1 S2. J. Am. Chem. Soc. 132(8):9522–23 291. Palmer AGI, Kroenke CD, Loria JP. 2001. Nuclear magnetic resonance methods for quantifying microsecond-to-millisecond motions in biological macromolecules. Methods Enzymol. 339:204–38 292. Sahu D, Clore GM, Iwahara J. 2007. TROSY-Based z-Exchange NMR: Appl Determination Activation Energy for Intermol Protein Translocation between Specific Sites on Different DNA. J. Am. Chem. Soc. 129(5):13232–37 293. Kloiber K, Spitzer R, Grutsch S, Kreutz C, Tollinger M. 2011. Longitudinal exchange: An alternative strategy towards quantification of dynamics parameters in ZZ exchange spectroscopy. J. Biomol. NMR. 51:123–29 294. Krimmer SG, Klebe G. 2015. Thermodynamics of protein-ligand interactions as a reference for computational analysis: how to assess accuracy, reliability and relevance of experimental data. J. 186 Comput. Aided. Mol. Des. 29(9):867–83 295. Krogh A, Larsson B, von Heijne G, Sonnhammer EL. 2001. Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J. Mol. Biol. 305(3):567–80 296. Almén MS, Nordström KJ V, Fredriksson R, Schiöth HB. 2009. Mapping the human membrane proteome: a majority of the human membrane proteins can be classified according to function and evolutionary origin. BMC Biol. 7:50 297. Tait AR, Straus SK. 2011. Overexpression and purification of U24 from Human Herpesvirus Type-6 in E. coli: unconventional use of oxidizing environments with a maltose binding protein-hexahistine dual tag to enhance membrane protein yield. Microb. Cell Fact. 10(1):51 298. Tait AR. 2011. Isolation and characterization of recombinant U24, a membrane protein from Human Herpesvirus Type 6. The University of British Columnbia. 189 pp. 299. Hegde RS, Keenan RJ. 2011. Tail-anchored membrane protein insertion into the endoplasmic reticulum. Nat. Rev. Mol. Cell Biol. 12(12):787–98 300. Favaloro V, Vilardi F, Schlecht R, Mayer MP, Dobberstein B. 2010. Asna1/TRC40-mediated membrane insertion of tail-anchored proteins. J. Cell Sci. 123(Pt 9):1522–30 301. Borgese N, Colombo S, Pedrazzini E. 2003. The tale of tail-anchored proteins: Coming from the cytosol and looking for a membrane. J. Cell Biol. 161(6):1013–19 302. Rizo J, Xu J. 2015. The Synaptic Vesicle Release Machinery. Annu. Rev. Biophys. 44(1):339–67 303. Gasteiger E, Hoogland C, Gattiker A, Duvaud S, Wilkins MR, et al. 2005. Protein Identification and Analysis Tools on the ExPASy Server. In The Proteomics Protocols Handbook, pp. 571–607 304. Yang Z. 2009. Hofmeister effects: an explanation for the impact of ionic liquids on biocatalysis. J. Biotechnol. 144(1):12–22 305. Czisch M, Boelens R. 1998. Sensitivity enhancement in the TROSY experiment. J. Magn. Reson. 134(1):158–60 306. Pervushin K V, Wider G, Wüthrich K. 1998. Single Transition-to-single Transition Polarization Transfer (ST2-PT) in [15N,1H]-TROSY. J. Biomol. NMR. 12(1):345–48 307. Meissner A, Schulte-Herbrüggen T, Briand J, Sørensen OW. 1998. Double spin-state-selective coherence transfer. Application for two-dimensional selection of multiplet components with long transverse relaxation times. Mol. Phys. 95(6):1137–42 308. Weigelt J. 1998. Single Scan, Sensitivity- and Gradient-Enhanced TROSY for Multidimensional NMR Experiments. J. Am. Chem. Soc. 120(41):10778–79 309. Rance M, Loria JP, Palmer AGI. 1999. Sensitivity improvement of transverse relaxation-optimized spectroscopy. J. Magn. Reson. 136(1):92–101 310. Zhu G, Kong XM, Sze KH. 1999. Gradient and sensitivity enhancement of 2D TROSY with water flip-back, 3D NOESY-TROSY and TOCSY-TROSY experiments. J. Biomol. NMR. 13(1):77–81 311. Ellena JF, Liang B, Wiktor M, Stein A, Cafiso DS, et al. 2009. Dynamic structure of lipid-bound synaptobrevin suggests a nucleation-propagation mechanism for trans-SNARE complex formation. Proc. Natl. Acad. Sci. U. S. A. 106(48):20306–11 187 312. Sharp PM, Cowe E, Higgins DG, Shields DC, Wolfe KH, Wright F. 1988. Codon usage patterns in Escherichia coli, Bacillus subtiiis, Saccharomyces certvisiae, Schizjosaccharomycespombt, Drosophila mclanogaster and Homo sapiens; a review of the considerable within-species diversity. Nucleic Acids Res. 16(17):8207–11 313. Drew D, Lerch M, Kunji E, Slotboom D-J, de Gier J-W. 2006. Optimization of membrane protein overexpression and purification using GFP fusions. Nat. Methods. 3(4):303–13 314. Chan HW. 2012. Inversigating the Enzymatic Mechanisms of the Inverting and Retaining Glycosyltransferases by NMR spectroscopy. The University of British Columbia. 270 pp. 315. Kang D, Gho YS, Suh M, Kang C. 2002. Highly Sensitive and Fast Protein Detection with Coomassie Brilliant Blue in Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis. Bull. Korean Chem. Soc. 23(11):1511–12 316. Delaglio F, Grzesiek S, Vuister GW, Zhu G, Pfeifer J, Bax A. 1995. NMRPipe: A multidimensional spectral processing system based on UNIX pipes. J. Biomol. NMR. 6(3):277–93 317. Goddard TD, Kneller DG. SPARKY 3 318. Jia CYH, Nie J, Wu C, Li C, Li SS-C. 2005. Novel Src homology 3 domain-binding motifs identified from proteomic screen of a Pro-rich region. Mol. Cell. Proteomics. 4(8):1155–66 319. Kaneko T, Kumasaka T, Ganbe T, Sato T, Miyazawa K, et al. 2003. Structural Insight into Modest Binding of a Non-PXXP Ligand to the Signal Transducing Adaptor Molecule-2 Src Homology 3 Domain. J. Biol. Chem. 278(48):48162–68 320. Mal TK, Matthews SJ, Kovacs H, Campbell ID, Boyd J. 1998. Some NMR experiments and a structure determination employing a [15N,2H] enriched protein. J. Biomol. NMR. 12(2):259–76 321. Arold S, O’Brien R, Franken P, Strub MP, Hoh F, et al. 1998. RT loop flexibility enhances the specificity of Src family SH3 domains for HIV-1 Nef. Biochemistry. 37(42):14683–91 322. Zarrine-Afsar A, Mittermaier A, Kay LE, Davidson AR. 2006. Protein stabilization by specific binding of guanidinium to a functional arginine-binding surface on an SH3 domain. Protein Sci. 15(1):162–70 323. Manie SN, Astier A, Haghayeghi N, Canty T, Druker BJ, et al. 1997. Regulation of Integrin-mediated p130 Cas Tyrosine Phosphorylation in Human B Cells. J. Biol. Chem. 272(25):15636–41 324. Renzoni DA, Pugh DJR, Siligardi G, Das P, Morton CJ, et al. 1996. Structural and thermodynamic characterization of the interaction of the SH3 domain from Fyn with the proline-rich binding site on the p85 subunit of PI3-kinase. Biochemistry. 35(49):15646–53 325. Sparks AB, Rider JE, Hoffman NG, Fowlkes DM, Lawrence A, Kay BK. 1996. Distinct ligand preferences of Src homology 3 domains from Src ,Yes, Abl, Cortactin, p53bp2, PLCγ , Crk, and Grb2. Proc. Natl. Acad. Sci. U. S. A. 93(4):1540–44 326. Lee CH, Leung B, Lemmon MA, Zheng J, Cowburn D, et al. 1995. A single amino acid in the SH3 domain of Hck determines its high affinity and specificity in binding to HIV-1 Nef protein. EMBO J. 14(20):5006–15 327. Lim WA, Richards FM. 1994. Critical residues in an SH3 domain from Sem-5 suggest a 188 mechanism for proline-rich peptide recognition. Nat. Struct. Biol. 1(4):221–25 328. Macdonald A, Mazaleyrat S, McCormick C, Street A, Burgoyne NJ, et al. 2005. Further studies on hepatitis C virus NS5A-SH3 domain interactions: Identification of residues critical for binding and implications for viral RNA replication and modulation of cell signalling. J. Gen. Virol. 86(4):1035–44 329. Cheng JTJ. 2010. Investigatin the structure-function relationship of cationic antimicrobial peptides and lipopeptides. The University of British Columbia. 231 pp. 330. Kabra R, Knight KK, Zhou R, Snyder PM. 2008. Nedd4-2 Induces Endocytosis and Degradation of Proteolytically Cleaved Epithelial Na+ Channels. J. Biol. Chem. 283(10):6033–39 331. Zhou R, Kabra R, Olson DR, Piper RC, Snyder PM. 2010. Hrs Controls Sorting of the Epithelial Na+ Channel between Endosomal Degradation and Recycling Pathways. J. Biol. Chem. 285(40):30523–30 332. Wilkin MB, Carbery A-M, Fostier M, Aslam H, Mazaleyrat SL, et al. 2004. Regulation of Notch Endosomal Sorting and Signaling by Drosophila Nedd4 Family Proteins. Curr. Biol. 14:2237–44 333. Sette P, Jadwin JA, Dussupt V, Bello NF, Bouamr F. 2010. The ESCRT-associated protein Alix recruits the ubiquitin ligase Nedd4-1 to facilitate HIV-1 release through the LYPXnL L domain motif. J. Virol. 84(16):8181–92 334. Blot V, Perugi F, Gay B, Prévost M-C, Briant L, et al. 2004. Nedd4.1-mediated ubiquitination and subsequent recruitment of Tsg101 ensure HTLV-1 Gag trafficking towards the multivesicular body pathway prior to virus budding. J. Cell Sci. 117(Pt 11):2357–67 335. Rauch S, Martin-Serrano J. 2011. Multiple interactions between the ESCRT machinery and arrestin-related proteins: implications for PPXY-dependent budding. J. Virol. 85(7):3546–56 336. Votteler J, Sundquist WI. 2013. Virus budding and the ESCRT pathway. Cell Host Microbe. 14(3):232–41 337. Freed EO. 2002. Viral Late Domains. J. Virol. 76(10):4679–87 338. Winberg G, Matskova L, Chen F, Plant P, Rotin D, et al. 2000. Latent Membrane Protein 2A of Epstein-Barr Virus Binds WW Domain E3 Protein-Ubiquitin Ligases That Ubiquitinate B-Cell Tyrosine Kinases. Mol. Cell. Biol. 20(22):8526–35 339. Zhimin L, Tony H. 2009. Degradation of Activated Protein Kinases by Ubiquitination. Annu. Rev. Biochem. 78(1):435–75 340. Magnifico A, Ettenberg S, Yang C, Mariano J, Tiwari S, et al. 2003. WW Domain HECT E3s Target Cbl RING Finger E3s for Proteasomal Degradation. J. Biol. Chem. 278(44):43169–77 341. Liu W, Yuen EY, Yan Z. 2010. The Stress Hormone Corticosterone Increases Synaptic-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic Acid (AMPA) Receptors via Serum- and Glucocorticoid-inducible Kinase (SGK) Regulation of the GDI-Rab4 Complex. J. Biol. Chem. 285(9):6101–8 342. Seebohm G, Strutz-Seebohm N, Birkin R, Dell G, Bucci C, et al. 2007. Regulation of endocytic recycling of KCNQ1/KCNE1 potassium channels. Circ. Res. 100(5):686–92 343. Debonneville C, Flores SY, Kamynina E, Plant PJ, Tauxe C, et al. 2001. Phosphorylation of Ned4-2 by Sgk1 regulates epithelial Na+ channel cell surface expression. EMBO J. 20(24):7052– 189 59 344. Gay DL, Ramón H, Oliver PM. 2008. Cbl- and Nedd4-family ubiquitin ligases: balancing tolerance and immunity. Immunol. Res. 42(1-3):51–64 345. Zhang J, Liu Q, Langdon WY. 2014. Cbl-b: Roles in T Cell Tolerance, Proallergic T Cell Development, and Cancer Immunity. IInflammation Cell Signal. 1(e146):7–11 346. Chiang YJ, Kole HK, Brown K, Naramura M, Fukuhara S, et al. 2000. Cbl-b regulates the CD28 dependence of T-cell activation. Nature. 403(6766):216–20 347. Sanna S, Pitzalis M, Zoledziewska M, Zara I, Sidore C, et al. 2010. Variants within the immunoregulatory CBLB gene are associated with multiple sclerosis. Nat. Genet. 42(6):495–97 348. Lott JS, Coddington-Lawson SJ, Teesdale-spittle PH, Donald FJMC. 2002. A single WW domain is the predominant mediator of the interaction between the human ubiquitin-protein ligase Nedd4 and the human epithelial sodium channel. Biochem. J. 488:481–88 349. Henry PC, Kanelis V, O’Brien MC, Kim B, Gautschi I, et al. 2003. Affinity and Specificity of Interactions between Nedd4 Isoforms and the Epithelial Na+ Channel. J. Biol. Chem. 278(22):20019–28 350. Stites W. 1997. Protein-protein interactions: interface structure, binding thermodynamics, and mutational analysis. Chem. Rev. 97:1233–50 351. Cooper A. 1999. Thermodynamic analysis of biomolecular interactions. Curr. Opin. Chem. Biol. 3(5):557–63 352. Sheinerman F. 2000. Electrostatic aspects of protein-protein interactions. Curr. Opin. Struct. Biol. 10(2):153–59 353. Donald JE, Kulp DW, DeGrado WF. 2011. Salt bridges: Geometrically specific, designable interactions. Proteins Struct. Funct. Bioinforma. 79(3):898–915 354. Ha JH, Capp MW, Hohenwalter MD, Baskerville M, Record MT. 1992. Thermodynamic stoichiometries of participation of water, cations and anions in specific and non-specific binding of lac repressor to DNA. Possible thermodynamic origins of the “glutamate effect” on protein-DNA interactions. J. Mol. Biol. 228(1):252–64 355. Boudker O, Ryan RM, Yernool D, Shimamoto K, Gouaux E. 2007. Coupling substrate and ion binding to extracellular gate of a sodium-dependent aspartate transporter. Nature. 445(7126):387–93 356. Schumann FH, Riepl H, Maurer T, Gronwald W, Neidig K-P, Kalbitzer HR. 2007. Combined chemical shift changes and amino acid specific chemical shift mapping of protein-protein interactions. J. Biomol. NMR. 39(4):275–89 357. Kleckner IR, Foster MP. 2011. An introduction to NMR-based approaches for measuring protein dynamics. Biochim. Biophys. Acta. 1814(8):942–68 358. O’Hayre M, Gutkind JS, Hurley JH. 2014. Structural and biochemical basis for ubiquitin ligase recruitment by arrestin-related domain-containing protein-3 (ARRDC3). J. Biol. Chem. 289(8):4743–52 359. Aragón E, Goerner N, Xi Q, Gomes T, Gao S, et al. 2012. Structural basis for the versatile interactions of Smad7 with regulator WW domains in TGF-β Pathways. Structure. 20(10):1726– 190 36 360. Bruce MC, Kanelis V, Fouladkou F, Debonneville A, Staub O, Rotin D. 2008. Regulation of Nedd4-2 self-ubiquitination and stability by a PY motif located within its HECT-domain. Biochem. J. 415(1):155–63 361. Becuwe M, Herrador A, Haguenauer-Tsapis R, Vincent O, Léon S. 2012. Ubiquitin-mediated regulation of endocytosis by proteins of the arrestin family. Biochem. Res. Int. 2012: 362. Nabhan JF, Pan H, Lu Q. 2010. Arrestin domain-containing protein 3 recruits the NEDD4 E3 ligase to mediate ubiquitination of the beta2-adrenergic receptor. EMBO Rep. 11(8):605–11 363. Shearwin-Whyatt LM, Brown DL, Wylie FG, Stow JL, Kumar S. 2004. N4WBP5A (Ndfip2), a Nedd4-interacting protein, localizes to multivesicular bodies and the Golgi, and has a potential role in protein trafficking. J. Cell Sci. 117(Pt 16):3679–89 364. Dalton HE, Denton D, Foot NJ, Ho K, Mills K, et al. 2011. Drosophila Ndfip is a novel regulator of Notch signaling. Cell Death Differe1. Dalt. HE, Dent. D, Foot NJ, Ho K, Mills K, al. 2011. Drosoph. Ndfip is a Nov. Regul. Notch signaling. Cell Death Differ. 18(7)1150–60ntiation. 18(7):1150–60 365. Mund T, Pelham HRB. 2010. Regulation of PTEN/Akt and MAP kinase signaling pathways by the ubiquitin ligase activators Ndfip1 and Ndfip2. Proc. Natl. Acad. Sci. U. S. A. 107(25):11429–34 366. Howitt J, Putz U, Lackovic J, Doan A, Dorstyn L, et al. 2009. Divalent metal transporter 1 (DMT1) regulation by Ndfip1 prevents metal toxicity in human neurons. Proc. Natl. Acad. Sci. U. S. A. 106(36):15489–94 367. Sang Q, Kim MH, Kumar S, Bye N, Morganti-Kossman MC, et al. 2006. Nedd4-WW domain-binding protein 5 (Ndfip1) is associated with neuronal survival after acute cortical brain injury. J. Neurosci. 26(27):7234–44 368. Williams R, Buchheit CL, Berman NEJ, Levine SM. 2012. Pathogenic implications of iron accumulation in multiple sclerosis. J. Neurochem. 120(1):7–25 369. Hametner S, Wimmer I, Haider L, Pfeifenbring S, Brück W, Lassmann H. 2013. Iron and neurodegeneration in the multiple sclerosis brain. Ann. Neurol. 74(6):848–61 370. Cooper A. 2005. Heat capacity effects in protein folding and ligand binding: a re-evaluation of the role of water in biomolecular thermodynamics. Biophys. Chem. 115(2-3):89–97 371. Spolar RS, Livingstone JR, Record MT. 1992. Use of liquid hydrocarbon and amide transfer data to estimate contributions to thermodynamic functions of protein folding from the removal of nonpolar and polar surface from water. Biochemistry. 31(16):3947–55 372. Spolar RS, Rocord MTJ. 2003. Coupling of Local Folding to Site-Specific Binding of Proteins to DNA. Science. 263(5148):777–84 373. Talhout R, Villa A, Mark AE, Engberts JBFN. 2003. Understanding binding affinity: A combined isothermal titration calorimetry/molecular dynamics study of the binding of a series of hydrophobically modified benzamidinium chloride inhibitors to trypsin. J. Am. Chem. Soc. 125(35):10570–79 374. Bergqvist S, O’Brien R, Ladbury JE. 2001. Site-specific cation binding mediates TATA binding 191 protein - DNA interaction from a hyperthermophilic archaeon. Biochemistry. 40(8):2419–25 375. O’Brien R, DeDecker B, Fleming KG, Sigler PB, Ladbury JE. 1998. The Effects of Salt on the TATA Binding Protein-DNA Interactino from a Hyperthermophilic Archaeon. J. Mol. Biol. 279:117–25 376. Shi H, Asher C, Chigaev A, Yung Y, Reuveny E, et al. 2002. Interactions of beta and gamma ENaC with Nedd4 can be facilitated by an ERK-mediated phosphorylation. J. Biol. Chem. 277(16):13539–47 377. Warnecke A, Sandalova T, Achour A, Harris RA. 2014. PyTMs: a useful PyMOL plugin for modeling common post-translational modifications. BMC Bioinformatics. 15(1):370 378. Mulder FA, Schipper D, Bott R, Boelens R. 1999. Altered flexibility in the substrate-binding site of related native and engineered high-alkaline Bacillus subtilisins. J. Mol. Biol. 292(1):111–23 379. Zhang Y, Chang C, Gehling DJ, Hemmati-Brivanlou A, Derynck R. 2001. Regulation of Smad degradation and activity by Smurf2, an E3 ubiquitin ligase. Proc. Natl. Acad. Sci. U. S. A. 98(3):974–79 380. Zhu H, Kavsak P, Abdollah S, Wrana JL, Thomsen GH. 1999. A SMAD ubiquitin ligase targets the BMP pathway and affects embryonic pattern formation. Nature. 400(6745):687–93 381. Shi Y, Massague J. 2003. Mechanisms of TGF-beta Signaling from Cell Membrane to the Nucleus. Cell. 113(Figure 2):685–700 382. Kavsak P, Rasmussen RK, Causing CG, Bonni S, Zhu H, et al. 2000. Smad7 Binds to Smurf2 to Form an E3 Ubiquitin Ligase that Targets the TGFβ Receptor for Degradation. Mol. Cell. 6(6):1365–75 383. Ebisawa T, Fukuchi M, Murakami F, Chiba T, Tanaka K, et al. 2001. Smurf1 Interacts with Transforming Growth Factor-beta Type I Receptor through Smad7 and Induces Receptor Degradation. J. Biol. Chem. 276(16):12477–80 384. Massagué J, Gomis RR. 2006. The logic of TGFβ signaling. FEBS Lett. 580(12):2811–20 385. Feng X-H, Derynck R. 2005. Specificity and Versatility in Tgf-Β Signaling Through Smads. Annu. Rev. Cell Dev. Biol. 21(1):659–93 386. Dijke P Ten, Hill CS. 2004. New insights into TGF-β-Smad signalling. Trends Biochem. Sci. 29(5):265–73 387. Markowitz S, Wang J, Myeroff L, Parsons R, Sun L, et al. 1995. Inactivation of the type II TGF-β receptor in colon cancer cells with microsatellite instability. Science. 268(5215):1336–38 388. De Groot CJ, Montagne L, Barten AD, Sminia P, Van Der Valk P. 1999. Expression of transforming growth factor (TGF)-beta1, -beta2, and -beta3 isoforms and TGF-beta type I and type II receptors in multiple sclerosis lesions and human adult astrocyte cultures. 389. Mangan PR, Harrington LE, O’Quinn DB, Helms WS, Bullard DC, et al. 2006. Transforming growth factor-β induces development of the TH17 lineage. Nature. 441(7090):231–34 390. Veldhoen M, Hocking RJ, Flavell RA, Stockinger B. 2006. Signals mediated by transforming growth factor-beta initiate autoimmune encephalomyelitis, but chronic inflammation is needed to sustain disease. Nat. Immunol. 7(11):1151–56 391. Cua DJ, Sherlock J, Chen Y, Murphy CA, Joyce B, et al. 2003. Interleukin-23 rather than 192 interleukin-12 is the critical cytokine for autoimmune inflammation of the brain. Nature. 421(6924):744–48 392. Kuruvilla AP, Shah R, Hochwald GM, Liggitt HD, Palladino MA, Thorbecke GJ. 1991. Protective effect of transforming growth factor beta 1 on experimental autoimmune diseases in mice. Proc. Natl. Acad. Sci. U. S. A. 88(April):2918–21 393. Mirshafiey A, Mohsenzadegan M. 2009. TGF-β as a promising option in the treatment of multiple sclerosis. Neuropharmacology. 56(6-7):929–36 394. Drescher KM, Murray PD, Lin X, Carlino JA, Rodriguez M. 2000. TGF-β2 Reduces Demyelination, Virus Antigen Expression, and Macrophage Recruitment in a Viral Model of Multiple Sclerosis. J. Immunol. 164(6):3207–13 395. Diemel LT, Jackson SJ, Cuzner ML. 2003. Role for TGF-beta1, FGF-2 and PDGF-AA in a myelination of CNS aggregate cultures enriched with macrophages. J. Neurosci. Res. 74(6):858–67 396. Jennings MD, Blankley RT, Baron M, Golovanov AP, Avis JM. 2007. Specificity and autoregulation of notch binding by tandem wwdomains in suppressor of deltex. J. Biol. Chem. 282(39):29032–42 397. Chong PA, Lin H, Wrana JL, Forman-Kay JD. 2006. An expanded WW domain recognition motif revealed by the interaction between Smad7 and the E3 ubiquitin ligase Smurf2. J. Biol. Chem. 281(25):17069–75 398. McDonald CB, Buffa L, Bar-Mag T, Salah Z, Bhat V, et al. 2012. Biophysical Basis of the Binding of WWOX Tumor Suppressor to WBP1 and WBP2 Adaptors. J. Mol. Biol. 422(1):58–74 399. Zhang R. 2015. Probing the interaction between U24 from HHV-6A and Smurf2 WW domains. The University of British Columbia 400. Freiburger LA, Baettig OM, Sprules T, Berghuis AM, Auclair K, Mittermaier AK. 2011. Competing allosteric mechanisms modulate substrate binding in a dimeric enzyme. Nat. Struct. Mol. Biol. 18(3):288–94 401. Vaynberg J, Fukuda T, Chen K, Vinogradova O, Velyvis A, et al. 2005. Structure of an ultraweak protein-protein complex and its crucial role in regulation of cell morphology and motility. Mol. Cell. 17(4):513–23 402. Harauz G, Libich DS. 2009. The classic basic protein of myelin--conserved structural motifs and the dynamic molecular barcode involved in membrane adhesion and protein-protein interactions. Curr. Protein Pept. Sci. 10(3):196–215 403. Brunner E. 2001. Residual Dipolar Couplings in Protein NMR. Concepts Magn. Reson. 13(4):238–59 404. Bax A, Grishaev A. 2005. Weak alignment NMR: A hawk-eyed view of biomolecular structure. Curr. Opin. Struct. Biol. 15:563–70 405. Tjandra N, Bax A. 1997. Crystalline Medium. Scien. 278(November):1111–14 406. Dürr UHN, Gildenberg M, Ramamoorthy A. 2012. The Magic of Bicelles Lights Up Membrane Protein Structure. Chem. Rev. 112:6054−6074 193 407. Mishima M, Hatanaka M. 2000. Couplings Across Hydrogen Bonds Formed between a Protein and a Nucleotide. J. Am. Chem. Soc., pp. 5883–84 408. Wu G, Yamada K, Dong S, Bldg LM, Street SG, et al. 2000. Intermolecular Hydrogen-Bonding Effects on the Amide Oxygen Electric-Field-Gradient and Chemical Shielding Tensors of Benzamide. J. Am. Chem. Soc. 122(21):4215–16 409. Ingham RJ, Colwill K, Howard C, Dettwiler S, Lim CSH, et al. 2005. WW domains provide a platform for the assembly of multiprotein networks. Mol. Cell. Biol. 25(16):7092–7106 410. Dong X, Liu OW, Howell AS, Shen K. 2013. An Extracellular Adhesion Molecule Complex Patterns Dendritic Branching and Morphogenesis. Cell. 155(2):296–307 411. Mello CC, Kramer JM, Stinchcomb D, Ambros V. 1991. Efficient gene transfer in C.elegans: extrachromosomal maintenance and integration of transforming sequences. EMBO J. 10(12):3959–70 412. Shi A, Grant BD. 2015. In vivo analysis of recycling endosomes in Caenorhabditis elegans. Biophys. Methods Cell Biol. 130:1–18 413. Sato K, Norris A, Sato M, Grant BD. 2014. C. elegans as a model for membrane traffic. In WormBook, pp. 1–47 414. Wang H, Yang H, Shivalila CS, Dawlaty MM, Cheng AW, et al. 2013. One-step generation of mice carrying mutations in multiple genes by CRISPR/cas-mediated genome engineering. Cell. 153(4):910–18 415. Hsu PD, Lander ES, Zhang F. 2014. Development and Applications of CRISPR-Cas9 for Genome Engineering. Cell. 157(6):1262–78 416. Hisano Y, Sakuma T, Nakade S, Ohga R, Ota S, et al. 2015. Precise in-frame integration of exogenous DNA mediated by CRISPR/Cas9 system in zebrafish. Sci. Rep. 5:8841 417. Strohalm M, Hassman M, Kosata B, Kodícek M. 2008. mMass data miner: an open source alternative for mass spectrometric data analysis. Rapid Commun. Mass Spectrom. 22:905–8 418. Strohalm M, Kavan D, Nova P, Volny M, Vladimı ́r Havlı ́cek. 2010. mMass 3: A Cross-Platform Software Environment for Precise Analysis of Mass Spectrometric Data. Anal. Chem. 82(11):4648–51 419. Kessner D, Chambers M, Burke R, Agus D, Mallick P. 2008. ProteoWizard: open source software for rapid proteomics tools development. Bioinformatics. 24(21):2534–36 420. Chambers MC, Maclean B, Burke R, Amodei D, Ruderman DL, et al. 2012. A cross-platform toolkit for mass spectrometry and proteomics. Nat. Biotechnol. 30(10):918–20 421. Chalkley RJ, Hansen KC, Baldwin MA. 2005. Bioinformatic methods to exploit mass spectrometric data for proteomic applications. Methods Enzymol. 402(1997):289–312 422. Schuchardt BJ, Mikles DC, Hoang LM, Bhat V, McDonald CB, et al. 2014. Ligand binding to WW tandem domains of YAP2 transcriptional regulator is under negative cooperativity. FEBS J. 281(24):5532–51 423. Sang Y, Tait AR, Scott WRP, Creagh AL, Kumar P, et al. 2014. Probing the Interaction between U24 and the SH3 Domain of Fyn Tyrosine Kinase. Biochemistry. 53:6092–6102 194 Appendices A Examples of SDS-PAGE and HPLC traces A1 U24 protein gel images Figure A.1 SDS-PAGE of U24-7 purification and U24-6A and U24-7 gel staining with silver or Coomassie stain. M indicates the marker and MWs are marked beside the gel. A) Histrap purification of Mbp-6xHis-U24-7. lane 1: insoluble cell pellet; lane 2: lysate; lane 3: flow-through; lane 4/5/6/7: washing fraction 1/2/3/4; lane 8: eluted Mbp-6xHis-U24-7. B) 2nd histrap and anion-exchange chromatography (anion-ex). lane 1: Mbp-6xHis-U24-7; lane 2: TEV digested fraction; lane 3: 2nd histrap flow-through; lane 4: 2nd histrap washing fraction; lane 5: 2nd histrap elution; lane 6: anion-ex flow-through; lane 7: anion-ex washing fraction; lane 8: anion-ex elution. C)/D) are purified U24-6A and U24-7 analyzed with DTT and BME, stained using silver/Coomassie. lane 1: 100 ng U24-6A; lane 2: 200 ng U24-6A; lane 3: 100 ng U24-7; lane 4: 200 ng U24-7. Dotted box indicates U24-6A/-7 monomers, and solid box indicates their dimers. 195 A2 Fyn-SH3 and WW domain gel examples Figure A.2 Example gel images of GST fusion protein purification, including Fyn-SH3 domain and several WW domains. M indicates the marker and the MWs are marked beside the gel. A) GST resin binding of GST-Fyn-SH3. lane 1: cell lysate; lane 2: flow-through; lane 3/4/5: washing fraction 1/2/3; lane 6/7: washing fraction 6/8. B) Post-digestion purification. lane 1: post-digestion fraction; lane 2: 3rd wash of post-digestion; lane 3: combined three washes of post-digestion fractions with protease (thrombin) removed; lane 4: elution of p-Aminobenzamidine-agarose (thrombin); lane 5: elution fraction. C)-F) are the section of the gels for different WW domains to show the purity and MW of the protein used in experiments. Lane 1-3 in C)/E)/F) are the same as lane 1-3 in B). They were from the purification of C) rNedd4-WW3 domain, E) hSmurf2-WW2+3 domain and F) hNedd4L-WW3* domain. D) purification result of rNedd4-WW2 domain. The digestion step was carried out after GST-rNedd4-WW2 domain eluted from the resin. The digested solution was bound again to clean GST resin. lane 1: after 2 times of rebinding; lane 2: after 3 times of rebinding; lane 3: elution of the resin used to bind GST protein. 196 A3 HPLC traces for U24-6A and U24-7 peptide purifications Figure A.3 HPLC gradient table and sample traces for the purification of U24-6A peptide. The purification process includes two sequential HPLC runs: traces of A) 1st run and B) 2nd run are shown. The HPLC gradient list is in the box above. The numbers represent the composition of buffer A or B. Example traces obtained at two wavelengths, 229 nm and 278 nm (in blue inset, only detect the tyrosine in the peptide), are shown for both runs. U24-6A collected from the 1st run, from injecting crude synthetic peptide, is the 42-44 min fraction in A). This fraction is lyophilized and injected again, for the 2nd run in B). The fraction at 49-50 min was collected as purified U24-6A peptide. 197 Figure A.4 HPLC gradient table and sample traces for the purification of U24-7 peptide. The purification process includes two sequential HPLC runs: traces of A) 1st run and B) 2nd run are shown. The HPLC gradient list is in the box above. The numbers represent the composition of buffer A or B. Example traces obtained at two wavelengths, 229 nm and 278 nm (in blue inset, only detect the tyrosine in the peptide), are shown for both runs. U24-7 collected from the 1st run, from injecting crude synthetic peptide, is the 33-35 min fraction in A). This fraction is lyophilized and injected again, for the 2nd run in B). The fraction at 29-32 min was collected as purified U24-7 peptide 198 B Mass spectrometry results B1 MALDI-TOF MS spectra of purified peptides Peptide samples were lyophilized before submitting to the Mass Spectrometry Centre. The molecular weights of each batch of purified peptides used in the titration experiments in Chapter 3, Chapter 4 and Chapter 5, were confirmed using MALDI-TOF MS. An additional peak (1244.8 in Figure B.2 inset) always appeared in the spectra of purified U24-7 peptide run in reflectron mode. However, this mass does not correspond to any possible truncation products of the peptide, and it would be disappeared when the same sample was run in linear mode. This phenomenon was observed with U24-7 peptide samples from different synthetic batches and different purification fractions (see traces above). It was proposed by Mashall Lapawa (Mass Spectrometry Centre, UBC) that this peak may be the result of a metastable decomposition of the peptide. Another reason could be that the resolution of the peak at 1244 is lower than the full-length peak at 1731, (see Figure B.2 below) which cannot be attributed to the noise in the spectra, as the intensity of both peaks are pretty high. This indicates that the reflector cannot focus the ions that may be caused by the decomposition. No similar peak was detected in the case of U24-6A peptide. The resolution of the peak at 1490 is poor due to low intensity. The spectra of the same sample run in linear and reflectron mode are shown in Figure B.1. The MALDI-TOF samples were prepared by the staff in the Mass Spectrometry Centre using the submitted peptide and α-cyano-4-hydroxycinnamic acid (CCA) or 2,5-dihydroxy benzoic acid (DHB). The spectra were converted from Bruker file format to mzML format using ProteoWizard and visualized using mMass. (417–420) 199 Figure B.1 MALDI-TOF MS spectra of U24-6A peptide. The theoretical MW of U24-6A peptide is 1685.9. The spectra from linear (top) and reflectron (bottom) mode are shown. The zoomed-in regions of the reflectron spectra are shown as an inset (bottom). 200 Figure B.2 MALDI-TOF MS spectra of U24-7 peptide. The theoretical MW of U24-7 peptide is 1731.9. The spectra from linear (top) and reflectron (bottom) mode are shown. The zoomed-in regions of the reflectron spectra are shown as an inset (bottom). 201 B2 LC-MS/MS spectra of purified peptides The purified peptides were submitted to the Mass Spectrometry Centre for sequencing on a Bruker HCTultra ion trap spectrometer. The fragmentation method used was collision-induced dissociation (CID). Different types of ions induced by MS/MS can be computed for peptides or proteins, here as U24-6A and U24-7, using the MS-product tool from ProteinProspector. (421) The tool can be accessed at http://prospector.ucsf.edu/prospector/mshome.htm. A series of peaks representing b and y ions (fragments extending from the N-termini are b ions, those from the C-termini are y ions), the m/z of which were computed by the MS-product tool above, could be found in the spectra. This could confirm the sequential connections of 13 residues in U24-7, while only 10 residues could be connected in the U24-6A peptide. This is because there are very few fragments generated in the case of U24-6A, even after repeated steps of ion isolation and fragmentation. Another fragmentation method electron transfer dissociation (ETD) could be tried if more sequential connectivities in U24-6A are required. The spectra shown below were visualized using MestreNova (Mestrelab Research). 202 Figure B.3 LC-MS/MS spectra of U24-6A peptide. The theoretical MW of U24-6A peptide is 1685.9. The ions at 1685.90 m/z, then 1439.80 m/z, then 1421.80 m/z were isolated sequentially and fragmented. The resulting spectra are shown from top to bottom. 203 Figure B.4 LC-MS/MS spectra of U24-7 peptide. The theoretical MW of U24-7 peptide is 1731.9. The ions at 1732.80 m/z and 866.90 m/z (+2) were isolated and fragmented. The 1132.60 m/z ion from 1732.80 m/z were also isolated. The resulting spectra are shown from top to bottom. 204 B3 MALDI-TOF MS spectra for purified recombinant U24-7 protein Figure B.5 MALDI-TOF MS spectra of U24-7 protein. The theoretical MW of U24-7 is 9464.0. The sample was prepared in a sinapinic acid (SA) sandwich run in linear mode. The peak at 4734.7 is the doubly charged ion. The peak at 10857.4, or doubly charged ion 5431.1, could be impurities. The spectra were converted from Bruker file format to mzML format using ProteoWizard and visualized using mMass. (417–420) 205 Figure B.6 MALDI-TOF MS spectra of U24-7 dCS protein. The theoretical MW of U24-7 dCS is 9431.9. The sample was prepared in a sinapinic acid (SA) sandwich run in linear and reflectron mode. The inset spectrum is from reflectron mode. The peak at 4707.4 is the doubly charged ion. The spectra were converted from Bruker file format to mzML format using ProteoWizard and visualized using mMass. (417–420) B4 MALDI-TOF MS study for thrombin linkage recombinant U24-7 Thrombin is an easy to use, efficient protease that is often used in purification of recombinant proteins to remove an expression tag. Early U24-7 constructs consisted of a plasmid with a thrombin cleavage site. Thrombin digestion of the resulting U24-7 protein was unspecifically cleaved. The design of Mbp-6xHis-U24-7 is shown in Figure B.7A. The cleaved fragment was detected by MALDI-TOF. Experiments were carried then out to make use of Factor Xa and thrombin to cleave the fusion protein Mbp-6xHis-U24-7, after elution from affinity chromatography and buffer exchange. The segment from Factor Xa digestion was still linked with 6xHis, denoted as 6xHis-U24-7, (see Figure B.7) while only U24-7 was obtained from the thrombin digestion. After 206 digestion, affinity chromatography was performed to remove Mbp. Different fractions were collected: elution fraction when Factor Xa was used, and flow-through fraction when thrombin was used. These fractions were desalted and dried after TCA and acetone precipitation. The resulting samples were sent to the Proteomics Core Facility (Michael Smith Laboratories, UBC). The results are shown in Figure B.7. The sequence and predicted MW of both full-length U24-7 and identified cleaved fragments are listed in Table B.1. The result shown in Figure B.7 indicates thrombin uncanonical cleavage at LSPR/TI, instead of at LVPR/GS. This breaks the U24-7 protein into two fragments, which were detected in the MALDI-TOF MS spectra. This protocol can therefore not be used for the study of U24-7 function, but could be useful to study the structure of the cytosolic region of U24-7 (no transmembrane region). It would be interesting to see if this segment can adopt the same structure as in the full-length protein. Table B.1 Amino acid sequences and theoretical masses of U24-7 and digested fragments. The theoretical MWs were computed using the ProtParam tool. (303) Name Sequence MW U24-7 GSMTHETPPPSYNDVMLQMFHDHSVFLHQENLSPRTINSTSSSEIKNVRRRGTFIILACLIISVILCLILILHIFNVRYGGTKP 9521.1 6xHis-U24-7 ISEFGSHHHHHHSSGLVPRGSMTHETPPPSYNDVMLQMFHDHSVFLHQENLSPRTINSTSSSEIKNVRRRGTFIILACLIISVILCLILILHIFNVRYGGTKP 11661.4 Fragment 1 GSMTHETPPPSYNDVMLQMFHDHSVFLHQENLSPR 4080.5 Fragment 2 TINSTSSSEIKNVRRRGTFIILACLIISVILCLILILHIFNVRYGGTKP 5458.6 207 Figure B.7 A) Schematic representation of Mbp-6xHis-U24-7 fusion protein and MALDI-TOF MS spectra of U24-7 protein sample after cleavage using B) Factor Xa and C) thrombin. B): fusion protein cleaved by Factor Xa, resulting in 6xHis-U24-7. The theoretical MW is 11661.4. C): fusion protein cleaved by thrombin, resulting in U24-7. The theoretical MW is 9521.1 but two smaller fragments, 4081.50 and 5460.11 were detected. The sum of these fragments is close to the mass of the full-length protein, suggesting the protein was cleaved in the middle. 208 B5 MALDI-TOF MS spectra for phosphorylation studies of U24-7 Figure B.8 in vitro phosphorylation of U24-7 by MAP kinase detected by MALDI-TOF MS. A) Phosphorylation experiment with 1 mM ATP B) 10 mM ATP. The insets are zoomed-in regions of the region of interest. The spectra of the sample before phosphorylation are coloured in black, of the sample 3 hours after the reaction are in blue and of the sample from the overnight reaction are in red. These spectra were recorded by Marshall Lapawa. 209 Figure B.9 in vitro phosphorylation of U24-6A and MBP by MAP kinase, as detected by MALDI-TOF MS. A) Phosphorylation of U24-6A with 1 mM ATP, B) Phosphorylation of MBP with 1 mM ATP. The insets are zoomed-in regions of the region of interest. The spectra of the sample before phosphorylation are coloured in black, of the sample 3 hours after the reaction are in blue and of the sample from the overnight reaction are in red. These spectra were recorded by Marshall Lapawa. 210 B6 MALDI-TOF MS spectra for Fyn-SH3 domain and WW domains In this section, MALDI-TOF MS spectra of the protein domains that were studied in Chapter 3, Chapter 4 and Chapter 5 are shown. The purified protein samples were dialyzed against water for over 48 hours and lyophilized before submitting to the Mass Spectrometry Centre. All of the spectra were collected by the staff in the Mass Spectrometry Centre, using a Bruker Autoflex MALDI-TOF spectrometer. Most of the spectra shown below were analyzed in linear mode, except hSmurf2-WW3. Linear mode spectra have lower resolution than those from reflectron mode, especially at high molecular weight range, so the mass of labeled peak is not very accurate, typically off by 1 to 2 Da. This inaccuracy can also be found in sections B3 and B4. The MALDI-TOF samples were prepared by the staff in the Mass Spectrometry Centre using the submitted protein powder and sinapinic acid (SA). The sequence of these protein domains, after protease cleavage, and their theoretical masses are listed in Table B.2 below. The theoretical molecular weights are calculated using the Protparam tool. (303) The spectra shown below were converted from Bruker file format to mzML format using ProteoWizard and visualized using mMass. (417–420) Table B.2 Amino acid sequences and theoretical MW of SH3 and WW domains. Name Sequence MW Fyn-SH3 GSPGISGGGGGILDTGVTLFVALYDYEARTEDDLSFHKGEKFQILNSSEGDWWEARSLTTGETGYIPSNYVAPVD 8016.6 rNedd4-WW2 GSEEQPTLPVLLPTSSGLPPGWEEKQDDRGRSYYVDHNSKTTTWSKPTMQDDPRSKIPAHLRGKTPVDSKNSS 8120.9 degraded segment of rNedd4-WW2 GSEEQPTLPVLLPTSSGLPPGWEEKQDDRGRSYYVDHNSKTTTWSKPTMQDDPR 6116.6 rNedd4-WW3/4 GSPVDSNDLGPLPPGWEERTHTDGRVFFINHNIKKTQWEDPRMQNVAITG 5659.2 hNedd4-WW3* GSIEQGFLPKGWEVRHAPNGRPFFIDHNTKTTTWEDPRLKIPAHL 5195.8 hNedd4L-WW3* GSVTQSFLPPGWEMRIAPNGRPFFIDHNTKTTTWEDPRLKFPVHM 5240.9 hSmurf2-WW2 GPLGSPPDLPEGYEQRTTQQGQVYFLHTQTGVSTWHDPRVPRDL 4962.4 hSmurf2-WW3 GPLGSGPLPPGWEIRNTATGRVYFVDHNNRTTQFTDPRLSAN 4611.0 hSmurf2-WW2+3 GPLGGSPPDLPEGYEQRTTQQGQVYFLHTQTGVSTWHDPRVPRDLSNINCEELGPLPPGWEIRNTATGRVYFVDHNNRTTQFTDPRLSAN 10104.0 211 Figure B.10 MALDI-TOF MS spectra of Fyn-SH3 domain. The theoretical MW of Fyn-SH3 domain is 8016.6. Figure B.11 MALDI-TOF MS spectra of rNedd4-WW2 domain. The theoretical MW of rNedd4-WW2 domain is 8120.9. Sequence of fragment (6115.2) is listed in Table B.2. 212 Figure B.12 MALDI-TOF MS spectra of rNedd4-WW3/4 domain. The theoretical MW of rNedd4-WW3/4 domain is 5659.2. Figure B.13 MALDI-TOF MS spectra of hNedd4-WW3* domain. The theoretical MW of hNedd4-WW3* domain is 5195.8. 213 Figure B.14 MALDI-TOF MS spectra of hNedd4L-WW3* domain. The theoretical MW of hNedd4L-WW3* domain is 5240.9. Figure B.15 MALDI-TOF MS spectra of hSmurf2-WW2 domain. The theoretical MW of hSmurf2-WW2 domain is 4962.4. 214 Figure B.16 MALDI-TOF MS spectra of hSmurf2-WW3 domain. The theoretical MW of hSmurf2-WW3 domain is 4611.0. The spectra are from reflectron mode. Figure B.17 MALDI-TOF MS spectra of hSmurf2-WW2+3 domain. The theoretical MW of hSmurf2-WW2+3 domain is 10104.0. 215 C NMR chemical shift tables and additional NMR spectra C1 Chemical shift assignment of Fyn-SH3 domain apo and bound to U24-6A Table C.1 Chemical shift assignment of Fyn-SH3 domain apo and bound to U24-6A (protein to peptide ratio is 1:12.68) Amino acid Apo Bound (U24-6A) Amino acid Apo Bound (U24-6A) 15N 1H 15N 1H 15N 1H 15N 1H G83 110.586 8.505 110.551 8.503 N113 114.978 7.527 114.906 7.513 V84 119.83 7.965 119.84 7.964 S114 121.593 8.841 121.888 8.894 T85 119.858 8.43 119.852 8.429 S115 117.154 8.059 117.29 8.107 L86 125.689 8.276 125.704 8.277 E116 121.728 8.708 121.783 8.746 F87 120.95 8.799 120.919 8.795 G117 107.849 8.281 107.737 8.268 V88 119.513 9.805 119.517 9.8 D118 119.068 8.51 119.189 8.522 A89 126.633 9.074 126.655 9.076 W119 120.793 7.636 120.686 7.603 L90 126.087 9.492 126.102 9.481 W119s 128.639 9.92 128.667 9.886 Y91 112.17 7.142 112.236 7.133 W120 124.223 9.177 124.203 9.168 D92 117.547 8.281 117.534 8.275 W120s 128.319 9.598 128.378 9.604 Y93 120.566 8.339 120.392 8.305 E121 123.83 8.693 123.883 8.699 E94 128.148 7.249 128.175 7.26 A122 131.818 9.421 131.892 9.428 A95 126.116 8.178 126.237 8.203 R123 118.766 8.961 118.785 8.964 R96 121.393 9.605 121.168 9.557 R123s 137.895 7.004 137.919 7.005 R96s 139.666 7.316 139.683 7.261 S124 119.947 8.763 119.932 8.75 T97 112.388 8.421 112.373 8.356 L125 130.857 8.996 130.87 9.014 E98 118.678 8.769 118.913 8.758 T126 116.368 8.443 116.346 8.444 D99 116.391 8.082 116.315 8.092 T127 108.362 8.067 108.364 8.073 D100 120.047 8.001 120.04 8.003 G128 111.172 7.963 111.16 7.964 L101 119.746 8.215 119.697 8.204 E129 120.816 8.029 120.82 8.032 S102 114.487 8.016 114.628 8.018 T130 113.218 8.372 113.236 8.366 F103 116.355 8.78 116.451 8.794 G131 111.232 8.801 111.311 8.815 H104 116.441 8.682 116.493 8.662 Y132 118.999 8.713 118.98 8.751 K105 121.814 9.213 121.827 9.21 I133 112.577 9.238 112.675 9.255 G106 114.602 8.948 114.628 8.951 S135 121.72 7.644 121.732 7.639 E107 124.582 8.18 124.636 8.184 N136 114.067 8.107 114.099 8.112 216 Amino acid Apo Bound (U24-6A) Amino acid Apo Bound (U24-6A) 15N 1H 15N 1H 15N 1H 15N 1H K108 120.993 8.098 121.055 8.095 Y137 119.377 7.823 119.465 7.813 F109 115.91 9.215 115.873 9.206 V138 109.049 7.135 109.03 7.124 Q110 119.624 8.64 119.629 8.627 A139 122.257 8.804 122.237 8.807 I111 126.648 9.459 126.603 9.468 V141 121.44 7.951 121.43 7.953 L112 128.722 8.946 128.692 8.949 D142 121.891 8.246 121.888 8.249 C2 Chemical shift assignment of Nedd4-WW domain apo and bound with U24 peptide (U24-7 is listed as an example) Table C.2 Chemical shift assignments of rNedd4-WW3/4 domain, hNedd4L-WW3* domain, both in the apo form and bound to U24-7 peptide (protein to peptide ratio is 1:4 in the case of rNedd4-WW3/4 domain, 1:2 in the case of hNedd4L-WW3* domain). rNedd4-WW3/4 domain hNedd4L-WW3* domain Amino acid Apo Bound (U24-7) Amino acid Apo Bound (U24-7) 15N 1H 15N 1H 15N 1H 15N 1H L461 122.618 8.731 122.474 8.752 L479 122.144 8.407 122.009 8.484 G464 112.413 8.815 112.625 9.003 G482 112.695 8.979 112.978 9.071 W465 118.285 7.677 117.515 7.739 W483 118.105 7.788 117.307 7.831 W465S 129.463 10.608 130.159 10.609 W483S 130.088 10.289 129.932 10.322 E466 121.207 9.409 120.573 9.374 E484 122.203 9.494 120.094 9.408 E467 124.711 8.718 125.325 8.813 M485 125.665 8.961 124.572 8.944 R468 125.984 8.521 125.101 8.426 R486 126.573 8.442 125.356 8.084 T469 113.596 8.089 115.148 8.237 I487 119.149 8.196 119.257 8.255 H470 127.44 9.752 127.456 9.855 A488 131.506 8.917 131.21 9.113 T471 119.904 8.42 116.95 6.948 N490 111.68 7.544 111.788 7.634 D472 - - 118.62 7.388 N490S1 107.632 7.384 106.429 7.379 G473 108.436 8.173 108.266 8.112 N490S2 107.605 6.035 106.43 5.617 R474 121.408 8.2 121.962 8.205 G491 108.746 8.634 108.888 8.636 V475 122.53 8.38 124.24 8.479 R492 122.623 7.769 123.772 7.814 F476 121.275 8.634 120.876 8.915 F494 118.168 8.798 118.358 9.249 F477 115.949 8.725 115.432 8.965 F495 117.594 8.816 116.864 9.172 217 rNedd4-WW3/4 domain hNedd4L-WW3* domain Amino acid Apo Bound (U24-7) Amino acid Apo Bound (U24-7) 15N 1H 15N 1H 15N 1H 15N 1H I478 120.734 9.292 121.68 9.539 I496 122.55 9.248 123.564 9.415 N479 123.903 8.361 123.416 8.509 D497 124.685 8.166 124.496 8.335 N479S2 109.516 6.962 112.291 7.235 H498 123.793 8.728 123.142 8.947 H480 124.241 8.609 122.967 8.727 N499 115.674 8.333 115.268 8.365 N481 117.655 8.368 115.787 8.106 N499S1 117.909 8.065 118.249 7.975 N481S1 113.925 7.197 115.193 7.446 N499S2 117.89 7.662 118.273 7.709 N481S2 113.923 7.531 115.193 7.446 T500 104.301 6.651 103.83 6.466 I482 108.81 6.764 107.793 6.434 K501 119.595 8.05 119.433 7.629 K483 119.677 7.552 118.31 7.144 T502 110.902 7.383 107.589 7.223 K484 119.069 7.431 117.436 7.258 T503 115.348 8.286 109.015 7.989 T485 118.622 8.367 112.064 8.011 T504 118.124 9.28 113.662 9.217 Q486 119.597 9.497 113.089 9.21 W505 125.005 8.675 126.019 9.022 Q486S1 113.618 6.823 115.947 7.096 W505S 128.748 10.066 129.343 10.453 Q486S2 113.608 7.67 115.948 8.287 E506 122.146 8.741 120.617 8.868 W487 120.359 8.727 121.111 9.076 D507 126.144 8.197 126.55 8.118 W487S 128.577 9.975 129.239 10.409 R509 118.6 8.286 118.654 8.277 E488 118.111 8.11 116.843 8.554 R509S 113.851 8.726 115.123 8.764 D489 124.905 8.392 124.901 8.354 L510 117.4 7.215 117.103 7.149 R491 118.468 8.69 119.044 8.652 1 M492 116.542 7.424 116.317 7.335 11 218 C3 U24-7 COSY spectrum Figure C.1 The fingerprint region from the COSY spectrum of wild type U24-7 protein. The protein amides peaks are crowded and unresolved. 219 C4 Thermal stability of hNedd4-WW3* and hNedd4L-WW3* domains Figure C.2 The overlays of HSQC spectra of hNedd4-WW3* and hNedd4L-WW3* domain at different temperatures, from 15 °C to 40 °C. Top: The overlays of HSQC spectra of hNedd4-WW3*. There are fewer peaks when the temperature is over 30 °C and the residual peaks are moving towards the center of the spectra. Bottom: The HSQC spectra overlays of hNedd4L-WW3*. 220 C5 EXSY of pU24-6A/hNedd4L-WW3* complex at 25 °C Figure C.3 Region of seHSQC, HSQC and EXSY spectra of pU24-6A binding to hNedd4L-WW3* complex at 25 °C. From top to bottom, apo (black) and bound (green) seHSQC overlays of hNedd4L-WW3* or bound with pU24-6A at (1:2); seHSQC of hNedd4L-WW3*:pU24-6A=1:0.6; HSQC of this same sample; EXSY spectra with 4 ms delays; EXSY spectra with 20 ms delays. All the spectra, except the overlays of apo and bound spectra, are plotted using the same starting contour and number of contour levels. *The noise in the HSQC spectra (no exchange) is higher due to the partial degradation of the sample. 221 C6 R492 regions of EXSY, pU24-6A/hNedd4L-WW3* complex at 15 °C Figure C.4 R492 regions of seHSQC, HSQC and EXSY spectra of pU24-6A binding to hNedd4L-WW3* at 15 °C. From top to bottom, apo (black) and bound (green) seHSQC overlays of hNedd4L-WW3*/pU24-6A (1:2) complex; seHSQC; HSQC; EXSY with 8 ms delays; EXSY with 24 ms; EXSY with 60 ms of an exchanged sample, hNedd4L-WW3*:pU24-6A=1:0.6. All the spectra, except the overlays of apo and bound spectra, are plotted using the same starting contour and number of contour levels. 222 C7 HSQC overlays of rNedd4-WW3/4 and hNedd4L-WW3* domains: apo form and in complex with U24-6A, U24-7 or pU24-6A peptide. Figure C.5 HSQC overlays of rNedd4-WW3/4 domain, apo (black), in complex with U24-6A (blue), with U24-7 (red) and with pU24-6A (green). The assignments of apo rNedd4-WW3/4 domain (black peaks and labels) and in complex with U24-7 (red peaks and labels) are shown. Grey dotted lines circle the peaks of the same residues in three bound complexes if they are away from each other. 223 Figure C.6 HSQC overlays of hNedd4L-WW3* domain, apo (black), in complex with U24-6A (blue), with U24-7 (red) and with pU24-6A (green). The assignments of apo hNedd4L-WW3* domain (black peaks and labels) and in complex with U24-7 (red peaks and labels) are shown. Grey dotted lines circle the peaks of the same residues in three bound complexes if they are away from each other. 224 C8 Example of HSQC assignment using 3D HSQC-NOESY and 3D HSQC-TOCSY spectra Figure C.7 Example assignment of apo rNedd4-WW3/4 domain using 3D HSQC-NOESY and 3D HSQC-TOCSY spectra. The strips of both 3D HSQC-NOESY (red to blue gradient colour) and 3D HSQC-TOCSY spectra (black) of assigned connected residues (T481 to D489) are shown. The connection between the protons in the previous residue (i-1, in TOCSY) and the current residues (i, in NOESY) are shown in dotted lines. 225 D Additional CD results D1 CD spectra of U24-7 protein Figure D.1 CD spectra of U24-7 dCS protein in citrate-phosphate buffer, 200 mM SDS, pH 5.8. The CD spectra of the sample before (recovered U24-7) and after filtration using 0.8 μm and 0.22 μm membranes are shown in different colors. These spectra are highly similar to Figure 2.9 U24-7 dCS protein dissolved in citrate-phosphate buffer, 50 mM SDS and 150 mM DPC, pH 5.8. -200 -150 -100 -50 0 50 100 200 210 220 230 240 250 Raw ellipticity (mdeg) Wavelength (nm) 150mM DPC/50mM SDS in CP buffer 0.8 µm and 0.22 µm filters 0.8 µm filter recovered U24-7 226 D2 An example of CD melt result, rNedd4-WW3/4 domain Figure D.2 CD melt for rNedd4-WW3/4 domain in 10 mM sodium phosphate buffer, pH 7.4, from 5 to 95 °C. The structural features start to disappear around 45 °C, as marked by the dotted black line. 227 E Additional ITC results E1 ITC results for Fyn-SH3 domain studies Figure E.1 Integrated heat plot from U24-6A and U24-6AΔMD peptides binding to Fyn-SH3 domain. The red curve is for U24-6AΔMD and the black curve is for U24-6A. In the red curve, saturation is achieved faster than in the black one, indicating a better binding affinity 228 Figure E.2 Typical ITC data obtained for U24-7 binding to Fyn-SH3 domain. The Kd was estimated using the “one set of binding site” model to fit the points in the bottom plot. The plot was fitted with binding stoichiometry n fixed at 1. 229 E2 Parameters from ITC experiments for Chapter 4 Table E.1 Thermodynamic parameters obtained from fitting ITC data of U24-6A or U24-7 binding to rNedd4-WW3/4 and hNedd4L-WW3* domain at 15 °C and 37 °C in 10 mM sodium phosphate, pH 7.4. The numbers in the brackets are the error, i.e. ± one standard deviation from three runs. n indicates the number of binding sites. Temperature (°C) U24-6A U24-7 n (sites) Kd (μM) ΔH (kJ/mol) ΔS (J/mol K) n (sites) Kd (μM) ΔH (kJ/mol) ΔS (J/mol K) rNedd4-WW3/4 15 1.217 (0.007) 16.4 (0.4) -36.6 (0.6) -35 (2) 1.03 (0.02) 5.07 (0.08) -52 (2) -80 (9) 37 1.126 (0.006) 93 (3) -61 (2) -119 (5) 1.03 (0.01) 43 (2) -72.5 (0.7) -150 (3) hNedd4L-WW3* 15 1.120 (0.006) 2.55 (0.09) -49.3 (0.2) -64.2 (0.9) 1.09 (0.05) 0.49 (0.02) -58.2 (0.4) -81 (2) 37 1.13 (0.01) 17.3 (0.1) -68.94 (0.02) -131 (0) 1.129 (0.006) 5.0 (0.2) -78.8 (0.6) -153 (2) 230 Table E.2 Thermodynamic parameters obtained from fitting ITC data for U24-6A, pU24-6A or U24-7 binding to rNedd4-WW3/4, hNedd4-WW3* and hNedd4L-WW3* domains at 25 °C in 10 mM sodium phosphate pH 7.4, with additional an 100 mM NaCl or 500 mM NaCl. The numbers in the brackets are the error, i.e. ± one standard deviation from three runs. n indicates the number of binding sites. extra NaCl concentration in 10 mM Na phosphate U24-6A U24-7 n (sites) Kd (μM) ΔH (kJ/mol) ΔS (J/mol K) n (sites) Kd (μM) ΔH (kJ/mol) ΔS (J/mol K) rNedd4-WW3/4 100 mM NaCl 1.169 (0.001) 35.7 (0.7) -45.7 (0.5) -68 (2) 1.04 (0.02) 14.6 (0.1) -66 (3) -129 (11) 500 mM NaCl 1.191 (0.004) 30.6 (0.9) -48.2 (0.7) -75 (3) 1.117 (0.005) 13.2 (0.4) -65.5 (0.6) -126 (2) hNedd4L-WW3* 100 mM NaCl 1.102 (0.005) 6.9 (0.2) -58.3 (0.5) -97 (2) 1.173 (0.002) 1.84 (0.06) -69.4 (0.5) -123 (2) 500 mM NaCl 1.142 (0.004) 6.1 (0.2) -57.4 (0.4) -93 (2) 1.107 (0.002) 2.09 (0.04) -73.1 (0.1) -136.3 (0.6) hNedd4-WW3* 100 mM NaCl 1.07 (0.01) 13.8 (0.4) -53.3 (0.3) -85 (1) 1.043 (0.006) 2.96 (0.03) -68.3 (0.3) -123.2 (0.9) 500 mM NaCl 1.093 (0.007) 11.4 (0.2) -54.34 (0.09) -87.7 (0.5) 1.172 (0.006) 3.8 (0.2) -67.9 (0.7) -123 (3) extra NaCl concentration in 10 mM Na phosphate pU24-6A n (sites) Kd (μM) ΔH (kJ/mol) ΔS (J/mol K) rNedd4-WW3/4 100 mM NaCl 1.046 (0.004) 12 (1) -56 (2) -93 (6) 500 mM NaCl 1.03 (0.06) 9.5 (0.1) -55.4 (0.7) -90 (3) hNedd4L-WW3* 100 mM NaCl 1.125 (0.004) 1.33 (0.03) -61.1 (0.3) -92 (1) 500 mM NaCl 1.174 (0.006) 1.97 (0.03) -53.8 (0.2) -71.3 (0.9) 231 E3 Electrostatic effect on U24 binding to hNedd4-WW3* Figure E.3 ΔH and ln Ka plots of U24-6A and U24-7 ligands binding to hNedd4-WW3* in phosphate buffer, with an additional 100 mM NaCl or 500 mM NaCl at 25 °C. A) ΔH plot of U24-6A (blue bars) and U24-7 (red bars) titrated into hNedd4-WW3* in three different buffers. B) ln Ka plot of the three ligands titrated into hNedd4-WW3* in three different buffers. The patterns of the bars indicate the buffer used: a solid bar indicates the experiments were done in 10 mM sodium phosphate, pH 7.4; bars with coloured dots were done in an extra 100 mM NaCl; while bars with a coloured grid were done in an extra 500 mM NaCl. Error bars indicate the standard deviation from three repeats. 232 E4 Parameters from ITC experiments for Chapter 5 Table E.3 Thermodynamic parameters obtained from fitting ITC data of U24-6A, pU24-6A or U24-7 binding to hSmurf2-WW3 and hSmurf2-WW2+3 domains at 5 °C and 15 °C in 40 mM HEPES, 10 mM NaCl pH 7.2. The numbers in the brackets are the error, i.e. ± one standard deviation from three runs. n indicates the number of binding sites. Temperature (°C) U24-6A pU24-6A n (sites) Kd (μM) ΔH (kJ/mol) ΔS (J/mol K) n (sites) Kd (μM) ΔH (kJ/mol) ΔS (J/mol K) hSmurf2-WW3 5 1.02 (0.02) 132 (2) -28.6 (0.3) -29 (1) 1.06 (0.01) 11.2 (0.2) -28.5 (0.4) -8 (2) 15 1.0 (0.1) 193 (11) -24 (3) -11 (9) 1.088 (0.001) 25.7 (0.4) -33.3 (0.4) -28 (1) hSmurf2-WW2+3 5 1.06 (0.02) 128 (4) -19.5 (0.2) 4.4 (0.6) 1.016 (0.004) 8.9 (0.5) -18.3 (0.5) 31 (1) 15 1.09 (0.02) 225 (6) -27.4 (0.6) -25 (2) 1.001 (0.009) 20 (1) -28.0 (0.7) -7 (3) Temperature (°C) U24-7 n (sites) Kd (μM) ΔH (kJ/mol) ΔS (J/mol K) hSmurf2-WW3 5 1.00 (0.01) 24 (1) -22.5 (0.5) 7 (2) 15 0.98 (0.04) 79 (2) -34 (1) -39 (4) hSmurf2-WW2+3 5 0.87 (0.01) 23.9 (0.8) -12.18 (0.07) 44.6 (0.5) 15 1.12 (0.02) 56 (2) -30.0 (0.3) -23 (1) 233 F Additional figures WW domain Sequence hNedd4-WW1 PLPPGWEERQDILGRTYYVNHESRRTQWKRPTPQD hNedd4-WW2 GLPPGWEEKQDERGRSYYVDHNSRTTTWTKPTVQA hNedd4-WW3* FLPKGWEVRHAPNGRPFFIDHNTKTTTWEDPRLKI hNedd4-WW4 PLPPGWEERTHTDGRIFYINHNIKRTQWEDPRLEN hNedd4L-WW1 PLPPGWEEKVDNLGRTYYVNHNNRTTQWHRPSLMD hNedd4L-WW2 GLPSGWEERKDAKGRTYYVNHNNRTTTWTRPIMQL hNedd4L-WW3* FLPPGWEMRIAPNGRPFFIDHNTKTTTWEDPRLKF hNedd4L-WW4 PLPPGWEERIHLDGRTFYIDHNSKITQWEDPRLQN rNedd4-WW1 PLPPGWEERQDVLGRTYYVNHESRTTQWKRPSPED rNedd4-WW2 GLPPGWEEKQDDRGRSYYVDHNSKTTTWSKPTMQD rNedd4-WW3/4 PLPPGWEERTHTDGRVFFINHNIKKTQWEDPRMQN dNedd4.1-WW3* PLPPRWSMQVAPNGRTFFIDHASRRTTWIDPRNGR hSmurf1-WW1 ELPEGYEQRTTVQGQVYFLHTQTGVSTWHDPRIPS hSmurf1-WW2 PLPPGWEVRSTVSGRIYFVDHNNRTTQFTDPRLHH hSmurf2-WW1 DLPDGWEERRTASGRIQYLNHITRTTQWERPTRPA hSmurf2-WW2 DLPEGYEQRTTQQGQVYFLHTQTGVSTWHDPRVPD hSmurf2-WW3 PLPPGWEIRNTATGRVYFVDHNNRTTQFTDPRLSA Su(dx)-WW1 PLPAGWEIRLDQYGRRYYVDHNTRSTYWEKPT Su(dx)-WW2 PLPPGWEIRKDGRGRVYYVDHNTRKTTWQRPN Su(dx)-WW3 PLPDGWEKKIQSDNRVYFVNHKNRTTQWEDPRTQG Su(dx)-WW4 PLPPGWEIRYTAAGERFFVDHNTRRTTFEDPRPGA YAP2-WW1 PLPAGWEMAKTSSGQRYFLNHIDQTTTWQDPRKAM YAP2-WW2 PLPDGWEQAMTQDGEIYYINHKNKTTSWLDPRLDP KIBRA-WW1 PLPEGWEEARDFDGKVYYIDHTNRTTSWIDPRDRY KIBRA-WW2 ELPLGWEEAYDPQVGDYFIDHNTKTTQIEDPRVQW Figure F.1 Alignment of Nedd4-WW, Smurf-WW, KIBRA-WW and other studied WW domains that have strong affinities with PY motif ligands. The conserved tryptophan, or tyrosine, phenylalanine or isoleucine at the tryptophan conserved positions are highlighted in bold and red. The residues in the β1-β2 loop of WW domain, determine its ligand preference, are highlighted in yellow background. 234 PY motif -9 -8 -7 -6 -5 -4 -3 -2 -10 1 2 34 567 89 10 11 12 PPx Y U24-6A M DPP R TP PPS Y SE V-L U24-7 M -TH E TP PPS Y ND VML pU24-6A M DPPR pTP PPS Y SE V-L rαENaC MTPP LAL TAP PPA Y AT LG rβENaC L PIP GTP PPN Y DS L rγENaC GS TVP GTP PPR Y NT L R hαENaC TAP PPA Y AT LG hβENaC G RP PPN Y DS LR Comm TG LPS Y DE AAH HECT-PY R LD LPP Y ET FED LRE K Nogo-A I KHE P EN PPP Y EE AM Smad6 CGP E SP PPP Y SR LSP RDE Smad7 CEL E SP P PP Y SR YPM DFL Smad1 MPA D TP PPA Y LP P ED PMT pSmad1* TS pSD P G pSPF QMP A D TP PPA Y LP P ED PMT Q Smad2 YIP E TP PPG Y IS EDG ETS pSmad2 PE pTP PPG Y IS EDG Smad3 NIP E TP PPG Y LS EDG ETS pSmad3* IPE pTP PPG Y IS EDG ETS DQQ LNQS M DTG pSP A EL pSP TT Smad5 LPA D TP PPA Y MP P DD QMG ARRDC3 PY1 ER P EA PPS Y AE VVT E ARRDC3 PY2 EFR FLP PPL Y SE IDP N N1 A KQ PPS Y ED CIK LATS1 Y QGP PPP Y PK HL Figure F.2 Alignment of PY motif ligands. The PY motif is highlighted in bold, arginines and lysines are highlighted in blue, while glutamic acid and aspartic acid residues are highlighted in red. Phosphorylated serines or threonines are shown in green. Except for rENaC (including αβγ subunits) ligand, all ligands listed above bind to at least one WW domain strongly (Kd≤15 μM). The ARRDC3 ligand binds to hNedd4L-WW domains, N1 binds to Su(dx)-WW domains, (396) and LATS1 binds to YAP2 WW domains. (422) The stars indicate ligand containing more than on binding motifs. longer ligand, The phospho-serine motif pSmad1 and pSmad3 in are roughly aligned to PY motif. 235 Figure F.3 90 degree view of structural model of rNedd4-WW3/4 and hNedd4L-WW3* domain, shown in Figure 4.12. Left: Coloured model of rNedd4-WW3/4 and PY motif peptide. (PDB: 1I5H) The WW domain is shown in green and peptide is shown as a grey ribbon. The nitrogen atoms are shown as spheres and coloured based on the ligand that is perturbed the most. Blue means U24-6A, red is U24-7, green is pU24-6A, yellow is three of them are very different and white is no ligand is special. The N- and C-terminus of the ligand is marked. Right: Coloured model of hNedd4L-WW3* and PY motif peptide. (PDB: 2MPT) The WW domain is shown in purple. Other colour scheme is as in A).
UBC Theses and Dissertations
Protein interactions of membrane protein U24 from Roseolovirus and implications for its function Sang, Yurou 2016
Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.
- 24-ubc_2016_may_sang_yurou.pdf [ 18.6MB ]
- JSON: 24-1.0228339.json
- JSON-LD: 24-1.0228339-ld.json
- RDF/XML (Pretty): 24-1.0228339-rdf.xml
- RDF/JSON: 24-1.0228339-rdf.json
- Turtle: 24-1.0228339-turtle.txt
- N-Triples: 24-1.0228339-rdf-ntriples.txt
- Original Record: 24-1.0228339-source.json
- Full Text