«NMR Studies of SH3 Domain Structure and Function by Irina Bezsonova A thesis submitted in conformity with the requirements for the degree of ...»
NMR Studies of SH3 Domain Structure and Function
A thesis submitted in conformity with the requirements
for the degree of Philosophy Doctor
University of Toronto
© Copyright by Irina Bezsonova 2008
NMR studies of SH3 Domain Structure and Function
University of Toronto
SH3 domains are excellent models for probing folding and protein interactions. This thesis
describes NMR studies of several SH3 domains, including the N-terminal SH3 domain of the Drosophila adaptor protein Drk (drkN SH3 domain), the SH3 domain of the proto-oncogene tyrosine-kinase Fyn, and the SH3 domains of the human adaptor protein CIN85, involved in Cblmediated downregulation of epidermal growth factor receptor (EGFR) and other receptor tyrosine kinases (RTKs). The drkN SH3 domain is an ideal system for studying disordered states.
The unique quality of this isolated domain is that it exists in an approximately 50/50 equilibrium between its folded and unfolded states under non-denaturating buffer conditions. Interestingly, the single T22G mutation dramatically stabilizes the domain. Here the NMR structures of the drkN SH3 domain and its T22G mutant are determined and compared in order to illuminate the causes of the marginal stability of the domain. Solvent exposure of the folded and the unfolded drkN SH3 domains are probed and compared with a novel NMR technique using molecular oxygen dissolved in solution as a paramagnetic probe. The changes in partial molar volume along the folding trajectories of the drkN SH3 and Fyn SH3 domains are also studied and analyzed here in terms of changes in protein hydration and packing accompanying folding.
Finally, the interactions between the SH3 domains of CIN85 and ubiquitin are discussed. All three are shown to bind ubiquitin. The structure of the SH3-C domain in complex with ubiquitin is presented and the effect of disruption of ubiquitin binding on ubiquitination of CIN85 and EGFR in vivo is discussed.
SH3 domains are easily amendable to a wide range of NMR approaches and provide a good system for development and testing of novel methods. Through the use of these approaches ii significant insights into details of SH3 domain structure, stability, mechanisms of folding and cellular function have been gained.
iii Acknowledgments First and foremost I would like to thank Julie Forman-Kay for teaching me about science as well as about life. I will be forever grateful to Julie for her patience, understanding and invaluable support through my best and my worst times. I thank Lewis Kay for teaching me everything I know about NMR and for his invaluable guidance. I am grateful to Scott Prosser not only for educating me about paramagnets and NMR but also for being a fun and easy person to work with and a great support when I needed it. I thank Daniela Rotin for providing me an opportunity to study novel SH3 domain/ubiquitin interactions and the biological effects of this interaction.
I would like to thank former and present members of Julie’s and Lewis’ labs, many of whom became my good friends. I thank Karin Crowhurst for our discussions of the drkN SH3 domain and for encouraging me to explore Toronto and Canada when I first came here, Patrick Finerty for teaching me about Linux and being very patient about my computer ignorance and also for teaching me the basics of American slang, Voula Kanelis for her priceless advice on any given topic, Jennifer Baker for being such a good friend and a helpful colleague, and Tanja Mittag for interesting discussions inside and outside the lab. I would like to especially thank Hong Lin for her expertise and great help in protein purification as well as for her warmth and kindness. I am thankful to Andrew Chong for many helpful discussions. I am indebted to Rhea Hudson for reminding me patiently over and over again all I needed to remember about protein cloning and for being always extremely helpful in troubleshooting any cloning difficulties. I thank Joseph Marsh for many interesting discussions of the drkN SH3 domain and potential utilization of oxygen-induced contact shifts as structural restrains for ensemble generation of disordered proteins using the program ENSEMBLE. I thank Ranjith Muhandiram for his crucial help with all aspects of NMR experiments, his expertise and patience.
I am grateful to Dmitry Korzhnev, my husband, my friend and my colleague, who was always an indispensable (even when I wanted to dispense of him) part of my research. I am thankful to my mom, Anna, and my mother-in-law, Katya, for their love and support. And, finally, I would like to thank my kids, Misha and Natalie, for helping me to put everything in perspective.
Table of Contents
List of Tables
List of Figures
1 Chapter 1
1.1 Src-homology 3 domains (SH3 domains)
1.2 The drkN SH3 domain
1.2.1 Role of Drk in cellular signal transduction
1.2.2 DrkN SH3 domain as a model for studying protein folding
1.3 NMR spectroscopy in structural characterization of proteins
1.3.1 Chemical shift
1.3.2 Scalar couplings
1.3.3 Nuclear Overhauser Effect (NOE)
1.3.4 Residual dipolar couplings
1.4 Paramagnetic effects in NMR protein structure determination
1.4.1 Hyperfine coupling and its effect on the nuclear chemical shift
1.4.2 The effect of hyperfine coupling on nuclear relaxation
1.4.3 Oxygen as paramagnetic agent in studies of solvent exposure, immersion depth and transient contacts
1.4.4 The effect of dissolved oxygen on chemical shifts
1.4.5 The effect of dissolved oxygen on nuclear relaxation rates
1.5 Unfolded, intermediate and transition states in protein folding
1.5.1 Characterization of disordered (unfolded) protein ensembles
1.5.2 Characterization of transition and intermediate states in protein folding............. 23 2 Chapter 2
2.2 Materials and Methods
2.2.1 Sample preparation
2.2.2 Thermodynamic stability
2.2.3 NMR spectroscopy
2.2.4 Estimation of helical content of the Uexch state
2.2.5 Structure calculations
2.2.6 NMR data and coordinates
2.3 Results and Discussion
2.3.1 Thermodynamic stabilities of the WT and T22G drkN SH3 domains................. 34 2.3.2 NOE and RDC measurements
2.3.3 Structure determination
2.3.4 Comparison of the structures of the WT and T22G mutant of the drkN SH3 domain
2.3.5 Structure refinement with WT NOE restraints
2.3.6 Comparison to the Grb-2 N-terminal SH3 domain
2.3.7 Effect of the T22G mutation on the stability of the drkN SH3 domain................ 48 3 Chapter 3
3.2 Materials and Methods
3.2.1 Free amino acid samples
3.2.2 DrkN SH3 domain samples
3.2.3 Oxygen pressure
3.2.4 Spectral Processing
3.3 Results and Discussion
3.3.1 Contact shift measurements
3.3.3 Analysis of solvent exposure for the Fexch and Uexch states
3.3.4 Analysis of paramagnetic shift data in the context of previous structural information for the Uexch state
3.4 Concluding Remarks
4 Chapter 4
4.2 Materials and Methods
4.2.1 Probing folding of the drkN SH3 domain
4.2.2 Probing folding of the G48M Fyn SH3 domain
4.3 Results and Discussion
4.3.1 The pressure dependence of the folding of the drkN SH3 domain
4.3.2 The pressure dependence of the folding of the G48M Fyn SH3 domain............. 83 4.3.3 Relating volume changes along the folding pathway to hydration and packing.. 85
4.4 Concluding Remarks
5 Chapter 5
5.2 Materials and Methods
5.2.1 Sample preparation:
5.2.2 NMR spectroscopy:
5.2.3 Structural calculations:
5.2.4 Ubiquitination assays:
5.2.5 Deposition of assignment and coordinates:
5.3.1 All three SH3 domains of CIN85 bind to ubiquitin
5.3.2 Data-based structural model of the CIN85 SH3-C domain/Ubiquitin complex. 106
5.3.4 Effect of the CIN85/ubiquitin interaction on ubiquitination of CIN85.............. 113
5.4.1 Structural analysis of the CIN85 SH3-C domain/ubiquitin interaction.............. 115 5.4.2 The CIN85 SH3-C domain/ubiquitin interaction plays a role in EGFdependent ubiquitination
6 Chapter 6
6.2 Future directions
6.2.1 CIN85/ubiquitin interaction
6.2.2 The future of oxygen
Table 2-2 Comparison of the diverging turn geometry of the drkN SH3 domain and its T22G mutant
Table 3-1 Averaged contact shifts, ∆δ O2, and normalized contact shifts, ∆δ O2, for Fexch and Uexch * states of the drkN SH3 domain in the presence of dissolved oxygen at a partial pressure of 60 Atm
Table 3-2 The difference between paramagnetic shifts, ∆∆δ O2, for Fexch and Uexch states of the drkN SH3 domain in the presence of dissolved oxygen at a partial pressure of 60 Atm............. 62 Table 3-3 Paramagnetic shifts, ∆δ O2, from freely dissolved amino acids in the presence of dissolved oxygen at a partial pressure of 30 Atm.
Table 5-1Structural statistics for CIN85 SH3 domain/ubiquitin complex
Figure 1.2 Binding of SH3 domains to proline-rich peptides in class I and class II orientations.
. 3 Figure 1.3 Simultaneous binding of two SH3 domains to a single peptide
Figure 1.4 The Drosophila EGFR signaling pathway.
Figure 1.5 Distribution of secondary Cα and Cβ chemical shifts (δCα, δCβ) in protein secondary structure elements.
Figure 1.6 The Karplus curve describing the variations of 3JHNHα with backbone dihedral angle φ.
Figure 1.7 Magnetic fields generated by nuclear spins I and S.
Figure 1.8 Schematic representation of an NMR spectrum of a nucleus coupled to an unpaired electron through a hyperfine coupling to an S=1/2 species.
Figure 2.1 Comparison of 1HN-15N RDC data recorded for the T22G drkN SH3 domain in the presence and absence of 0.
5 M Na2SO4
Figure 2.2 Superposition of the 10 lowest energy structures of the WT (A) and T22G (B) drkN SH3 domains calculated based on RDC, C'-CSA, dihedral angle and hydrogen bond experimental restraints and WT (C) and T22G (D) structures refined with NOE restraints.
....... 38 Figure 2.3 Ribbon diagram of the lowest energy structures of the (A) WT and (B) T22G drkN SH3 domain.
Figure 2.4 Superposition of the WT (red) and the T22G (blue) structural ensembles represented as tubes with radii proportional to the RMSD values within the ensembles.
Figure 2.5 The difference between the backbone dihedral angles of the WT and T22G structures calculated without (a) and with (b) NOE restraints, shown as a function of residue number.
Figure 3.1 Distribution of average paramagnetic shifts, ∆δ O2, arising from dissolved oxygen for the Fexch (grey bars) and Uexch (black bars) states of the drkN SH3 domain
Figure 3.2 Difference in 13C paramagnetic shifts between the Uexch and Fexch states mapped onto the structure of the folded state, shown in two different orientations
potential structures sampled within the unfolded state ensemble that are consistent with the protection data
Figure 4.1 Ribbon diagrams of the backbone structures of the drkN and Fyn SH3 domains.
..... 76 Figure 4.2 Pressure dependent folding of the drkN SH3 domain.
Figure 4.3 Pressure dependent folding of the G48M Fyn SH3 domain
Figure 5.1 N-HN residual dipolar coupling data for saturated CIN85 SH3–C domain and saturated ubiquitin
Figure 5.2 Paramagnetic relaxation enhancement data for CIN85 SH3-C domain bound to ubiquitin.
Figure 5.3 Homology model of the CIN85 SH3-C domain and comparison with other SH3 domains of CIN85
Figure 5.4 NMR titration data for binding of ubiquitin and CIN85 SH3 domains
Figure 5.5 Binding curves for the ubiquitin/CIN85 SH3 domain interactions.
Figure 5.6 Chemical shift mapping of the CIN85 SH3-C domain/ubiquitin binding interface.
Figure 5.8 Comparison of the CIN85 SH3-C and the Sla1 SH3-3 domains complexes with ubiquitin.
Figure 5.9 Competition between ubiquitin and Cbl peptide for CIN85 SH3-C domain binding.
Figure 5.11 Ubiquitination of the wild type CIN85 and the 3FY (ubiquitin-binding impaired) mutant CIN85
Figure 5.12 Model for the role of ubiquitin binding in the ubiquitination of CIN85................. 119
Figure 6.1 Correlation between oxygen induced 13C contact shifts and solvent accessible surface area.
TOCSY total correlation spectroscopy HSQC heteronuclear single quantum coherence CPMG Carr–Purcell–Meiboom–Gill SAXS small angle X-ray scattering HADDOCK High Ambiguity Driven biomolecular DOCKing TEMPO 4-hydroxy-2,2,6,6-tetramethyl-4-hydroxy-2,2,6,6tetramethylpiperidinooxy DSS 2,2-dimethyl-2-sila-pentane-5-sulfonate Isopropyl β-D-1-thiogalactipyranoside IPTG