Phosphorylation-Dependent Structural Changes in the N-Terminal Extension of SUR2A NBD1

Phosphorylation-dependent structural changes in the N-terminal extension of SUR2A NBD1

Clarissa Rana Sooklal

A thesis submitted in conformity with the requirements for the degree of the Master of Science Graduate Department of Chemistry University of Toronto

Phosphorylation-dependent structural changes in the N-terminal extension of SUR2A NBD1

Master of Science 2015

Clarissa Rana Sooklal

Graduate Department of Chemistry

University of Toronto

Abstract

ATP-sensitive potassium (KATP) channels are involved in many biological processes and play an important role in sustaining healthy functioning organs. KATP channels are composed of four copies of a pore-forming Kir6.x protein that is surrounded by four copies of regulatory sulfonylurea receptor (SUR) proteins. SUR proteins are members of the ATP-binding cassette

(ABC) superfamily. Nucleotide binding and hydrolysis at the SUR proteins results in channel opening, which is potentiated by SUR protein phosphorylation. The studies conducted during this thesis aimed to understand how phosphorylation of the N-terminal extension (N-tail) of the first nucleotide binding domain (NBD1) of SUR2A alters its structure and influences the N-tail’s interaction with the core NBD1. This biophysical study utilizes nuclear magnetic resonance

(NMR) and other biophysical techniques to understand these structural and dynamic changes in

N-tail with phosphorylation which will allow us to further understand ATP-binding at NBD1 and control of KATP channel conductance.

Acknowledgements

I would first like to thank to my supervisor, Professor Voula Kanelis for all her support, guidance and direction. She has been my biggest inspiration and my role model throughout my graduate studies. Without her help and encouragement I would not have been able to expand my understanding of science and science research. Voula has been my teacher and mentor. And for that, I am forever grateful. I would also like to thank my past colleague Dr. Jorge Pédro Lopez-Alonso who first helped me get started with my project and find my way around the laboratory. Jorge not only made my transition into research an easy one, but also enjoyable. I would like to thank my current lab mates with an extra special thanks to Dr. Elvin de Araujo. His support and endless advice and help has been crucial in getting me through this project. Elvin was the greatest support and was willing to help whenever I ran into issues related to all aspects of my project from purifying the N-tail to using size-exclusion column chromatography and troubleshooting the NMR. And to Sasha Weiditch, who provided her support and comfort making the difficult times bearable. Throughout my graduate studies Sasha became my dear friend and her emotional support helped pull me through the stressful aspects of research. And I am truly grateful to Marijana, our lab technician for helping me obtain my N- tail’s DNA and her countless of favours. I would also like to thank Professor Jumi Shin for allowing me to use her circular dichroism instrument and equipment, and for reading my thesis. I would like to thank Professor Scott Prosser for allowing me to use his NMR magnet. And I would like to thank Dr. Darcy Burns and Dr. Sergiy Nokhrin for the help in conducting my 3D triples resonance NMR experiments and Dr. Peter Mitrakos for his help using a the HPLC at the University of Toronto Mississauga when I was troubleshooting the N-tail phosphorylation reaction experiments. And lastly, I would like to thank Richard Ramsundar and Llana Sooklal for their incredible and endless support!

iii

Table of Contents

1. Introduction 1 1.1 Biological importance of ABC transporters 1 1.2 Structural architecture of ABC transporters 1 1.3 The ABCC subfamily of ABC proteins and the sulfonylurea receptors (SURs) 3

1.4 KATP channels 4 1.5 Intrinsically disordered proteins 5

1.6 Phosphorylation regulation of KATP channels 6 1.7 N-terminal extension (N-tail) of SUR2A 6 1.8 Biophysical techniques to study SUR2A N-tail 7 1.8.1 Nuclear magnetic resonance 7 1.8.2 Circular dichroism 11 1.8.3 Dynamic light scattering 13 1.9 Resonance assignments 15 2. Materials and Methods 19 2.1 Expression and purification of SUR2A N-tail 19 2.2 Phosphorylation of SUR2A N-tail 20 2.3 Protein concentration determination 20 2.4 NMR spectroscopy 21 2.5 Circular dichroism 21 2.6 Size exclusion chromatography 22 2.7 Dynamic light scattering 22 3. Results 23 3.1 Analysis of the primary sequence of N-tail 23 3.2 Expression and purification of SUR2A N-tail S615-E664 construct 23 3.3 Biophysical characterization and structural studies of SUR2A N-tail-S615 26 3.3.1 NMR studies of non-phosphorylated N-tail-S615 26 3.3.2 Phosphorylation of N-tail-S615 27 3.3.3 Structural changes associated with the phosphorylation of N-tail-S615 31

3.3.4 Changes in the interaction between N-tail-S615 and NBD1 with 35 phosphorylation 3.4 Expression and purification of SUR2A N-tail-S608 and N-tail-Q600 37 3.5 Non-phosphorylated and phosphorylated N-tail-S608 NMR spectra and other 41 characterization 3.5.1 Changes in the interaction between N-tail-S608 and NBD1 with 45 phosphorylation 3.6 Resonance Assignment of non-phosphorylated N-tail-S615 48 3.6.1 Selective unlabeling approach to assist in resonance assignment 51 3.6.2 Analysis of assigned N-tail-S615 54 4. Discussion 59 5. References 62 6. Copyright Permissions 68

List of Tables

Table 1. Brief list of the 7 ABC subfamilies in humans and examples 3 Table 2. List of the most important nuclei found in biomolecules and their gyromagnetic ratio 8

List of Figures Figure 1. Schematic of an ABC protein subfamily C (ABCC) 2

Figure 2. Schematic top view of KATP channel 5 Figure 3. Energy separation of nuclear spins in the presence of an external magnetic 8 field Figure 4. Example of a 2-dimensional 15N-HSQC spectrum 10 Figure 5. CD spectra of protein secondary structures 12 Figure 6. Comparison of the hydrodynamic radius under difference salt concentrations 13 monitored by DLS Figure 7. Effect of surface projections on the hydrodynamic radius 13 Figure 8. Determining the hydrodynamic radius of non-spherical particles 14 Figure 9. 3-dimensional NMR experiments used to assign resonances 17 Figure 10. 3-dimensional NMR experiments aided by CCC-TOCSY 3D experiment 18 Figure 11. Agadir predication of helicity based on N-tail-S615 primary sequence 23 Figure 12. SDS gel of the purification of a 1 L cell growth of 6xHis-SUMO N-tail-S615 24 fusion protein Figure 13. Gel sample of the purified N-tail-S615 used for NMR 24 Figure 14. Overlay of NBD1 (615-933) and N-tail-S615 TROSY 15N-HSQCs 26 Figure 15. 15N-HSQC NMR spectrum of 15N N-tail-S615 run at (A) 12 ºC and (B) 30 ºC 27 Figure 16. Monitoring the phosphorylation reaction of N-tail-S615 28 Figure 17. Gel filtration clean-up of (A) phosphorylated N-tail-S615 and 30 (B) phosphorylated N-tail-S615 HSQC Figure 18. SDS gel shift comparing non-phosphorylated and phosphorylated N-tail-S615 30 Figure 19. Comparison of non-phosphorylated and phosphorylated N-tail-S615 15N- 31 HSQCs Figure 20. Comparison of (A) native and (B) denatured non-phosphorylated and 32 phosphorylated N-tail-S615 gel filtrations Figure 21. DLS of non-phosphorylated and phosphorylated N-tail-S615 hydrodynamic 33 radii Figure 22. CD spectra of (A) non-phosphorylated and (C) phosphorylated N-tail-S615 34 with increasing concentrations of trifluoroethanol (TFE) Figure 23. Interaction experiment overlay of non-phosphorylated N-tail-S615 with and 36 without NBD115N-HSQCs

vii

Figure 24. Interaction experiment overlay of phosphorylated N-tail-S615 with and 37 without NBD115N-HSQCs Figure 25 SDS gel of the purification of a 1 L cell growth of 6xHis-SUMO N-tail-S608 39 fusion protein Figure 26 SDS gel of the purification of a 1 L cell growth of 6xHis-SUMO N-tail-Q600 40 fusion protein Figure 27. 15N-HSQC spectrum of 15N N-tail-S608 run at 12 ºC 41 Figure 28. 15N-HSQC spectrum of 15N N-tail-Q600 run at 12 ºC 41 Figure 29. Overlay of 3 non-phosphorylated N-tail constructs’ 15N-HSQC spectra 42 Figure 30. Overlay of non-phosphorylated and phosphorylated of N-tail-S608 15N- 43 HSQCs Figure 31. Overlay of non-phosphorylated and phosphorylated of N-tail-Q600 15N- 43 HSQCs Figure 32. Comparison of native non-phosphorylated and phosphorylated N-tail-S608 gel 44 filtrations Figure 33. CD spectra of (A) non-phosphorylated and (B) phosphorylated N-tail-S608 45 with increasing concentrations of TFE Figure 34. Interaction experiment overlay of non-phosphorylated N-tail-S608 with and 46 without NBD115N-HSQCs Figure 35. Interaction experiment overlay of phosphorylated N-tail-S608 with and 47 without NBD1 15N-HSQCs Figure 36. 3D NMR data used to link clusters and assign N-tail-S615 15N-HSQC peaks 49 Figure 37. Labeled resonance assignment on N-tail-S615 15N-HSQC 50 Figure 38. Overlay of selectively unlabeled lysine N-tail-S615 15N-HSQC 52 Figure 39. Reaction between iodoacetamide and peptide cysteine residue 53 Figure 40. N-tail-S615 modification with iodoacetamide 15N-HSQC 54 Figure 41. CSI calculation based on assigned chemical shift of N-tail-S615 55 Figure 42. Overlay of assigned non-phosphorylated N-tail-S615 and unassigned 55 phosphorylated N-tail 15N-HSQCs Figure 43. Interaction experiments analysis based on chemical shift, volume and intensity 57 changes vs. amino acid of assigned non-phosphorylated N-tail-S615 Figure 44. Schematic of N-tail-S615 phosphorylation regulation of NBD1 58

viii

List of Abbreviations

ABC ATP-binding cassette ATP adenosine triphosphate CD circular dichroism CFTR cystic fibrosis transmembrane conductance regulator DLS dynamic light scattering DSS 4,4-dimethyl-4-silapentane-1-sulfonic acid DTT dithiothreitol ERS early repolarization syndrome FHA forkhead-associated domains HSQC heteronuclear single quantum coherence ICD intracellular domain IDP intrinsically disordered proteins INEPT insensitive nuclei enhanced by polarization transfer IPTG isopropyl-β-D-thio-galactoside

KATP channel ATP-sensitive potassium channel MRP multidrug resistant protein MSD membrane spanning domain MWCO molecular weight cutoff NBD nucleotide binding domain NMR nuclear magnetic resonance TROSY transverse relaxation optimized spectroscopy PgP/PGY1 P-glycoprotein PKA protein kinase A SDS-PAGE sodium dodecyl sulfate-polyacrylamide gel electrophoresis SUMO small ubiquitin-like modifier SUR sulfonylurea receptors

T2 spin-spin relaxation TB tuberculosis TCEP-HCl tris (2-carboxyethyl) phosphine hydrochloride

TFE trifluoroethanol TRIS 2-amino-2-hydroxymethyl-propane-1,3-diol Ulp1 (ubiquitin-like protein)-specific protease 1 UV-Vis spectroscopy ultraviolet visible spectroscopy

1. Introduction

1.1 Biological importance of ABC transporters The ATP-Binding Cassette (ABC) transporter superfamily is one of the largest and diverse families of proteins (1, 2). ABC proteins are integral, multi-spanning membrane proteins that are found in all kingdoms of life (2, 3, 4). There are over 6000 known ABC proteins. In prokaryotes, ABC proteins function as both importers and exporters (4). In eukaryotic cells, ABC proteins export solutes from the cytoplasm, regulate the activity of other proteins, or function as channels (4). ABC proteins control diverse processes, such as in the maintenance of the blood-brain barrier, antigen presentation to T-cell and iron trafficking (2, 3, 4). Given that ABC proteins are involved in a number of diverse and crucial physiological processes, it is no surprise that they can contribute to a number of human health-related diseases and disorders (2, 4). For example, cystic fibrosis is caused by a mutation of the gene encoding ABCC7 or the cystic fibrosis transmembrane conductance regulator (CFTR), which results in reduced activity of the CFTR ion channel rendering the membrane impermeable to chloride ions (2). Mutation of the ABCA1 gene causes Tangier disease, a rare disorder resulting in abnormal deposition of cholesterol and atherosclerosis (2). Dubin-Johnson syndrome, adrenoleukodystrophy and immunodeficiency are a few more amongst many other ABC-related human diseases (2). In addition to their normal role in the cell, over-activation of ABC proteins also affects human health. For example, the ABC proteins P-glycoprotein (PgP) and the multidrug resistance proteins (MRP1 and MRP2) are widely distributed in mammalian tissues to export toxins out of cells (2). PgP plays an important role in chemotherapeutic resistance as well as in the efflux of immunosuppressant drugs contributing to organ rejection (2). Prokaryotic ABC proteins confer multi-drug resistance to bacterial cells. Thus, ABC proteins are of vast medical importance. An understanding of the molecular mechanisms underlying ABC protein function and activity is an important field of current research (5, 6, 7).

1.2 Structural architecture of ABC transporters Although ABC proteins transport a wide variety of substrates, ABC transporters exhibit an overall similar molecular architecture (3). Proteins of the ABC superfamily, at minimum,

1 consists of two membrane spanning domains (MSDs) and two cytoplasmic nucleotide binding domains (NBDs) shown in Figure 1 (8, 9, 10). The MSDs are extremely hydrophobic consisting of 6 membrane spanning domains (2). In general, the MSDs determine substrate specificity of the transporter (2). Thus the MSDs are involved with interactions of the specific substrates, whereas the hydrophilic NBDs of the ABC proteins are the sites of nucleotide binding/hydrolysis. Since the MSDs of different ABC’s interact with different substrates there is a considerable degree of variability amongst the MSDs of different ABC proteins unlike the NBDs (1). In contrast, the NBDs are significantly conserved with a high primary sequence identity among different ABC proteins (1, 2).

Figure 1: Schematic of an ABC protein subfamily C illustrating the extra MSD (MSD0) connected to the first MSD (MSD1) via an L0 linker (10).

Crystal structures of several full length ABC proteins demonstrate that the transmembrane helices of the MSDs extends from the membrane into the cytoplasm of the cell (11, 12, 13). These long helical extensions are connected by short irregular helices, known as coupling helices (13). Together the long helical extensions and the coupling helices form the intracellular domains (ICDs). The coupling helices bind the NBDs and serve as the mechanical link between the MSDs and the NBDs (1, 12). Consequently, conformational changes associated with ATP-binding at the NBDs are translated through the coupling helices and ICDs to the MSDs enabling substrate transport (1).

1.3 The ABCC subfamily of ABC proteins and the sulfonylurea receptors (SURs)

ABC transporters are classified into sub-families based on the sequence of the NBDs. There are 7 subfamilies (ABCA to ABCG) for human ABC proteins (Table 1) (4). Table 1: The 7 ABC subfamilies in humans and examples (14) Subfamily Examples ABCA ABCA1 – protein regulator of cholesterol efflux (14)

ABCB ABCB1 – P-glycoprotein (14)

ABCC ABCC8/9 – sulfonylurea receptors, ABCC7 – CFTR (14) multidrug resistant proteins (MRPs) (15) ABCD ABCD1 – adrenoleukodystrophy (ALDP) protein (16)

ABCE ABCE1 – RNAse L inhibitor (14)

ABCF

ABCG ABCG5/8 – sterol transport (17)

My work focuses on studies of a regulatory region in the sulfonylurea receptor 2A (SUR2A ABCC9), which is a member of the C subfamily of ABC proteins (ABCC). Other members of the ABCC subfamily include the multidrug resistant proteins (e.g. MRP1/ABCC1 and MRP2/ABCC2), the cystic fibrosis conductance regulator (CFTR/ABCC7) and the sulfonylurea receptor 1 (SUR1/ABCC8). Members of the ABCC subfamily are further distinguished from other ABC proteins by the presence of a region N-terminal to the minimum architecture (18). The N-terminal extension varies in length, and thus members of the ABCC subfamily can be further classified as ‘long” or “short” (18). ABCC members with the “long” arrangement contain an extra MSD (MSD0) that is connected to MSD1 via a cytoplasmic L0 linker (Figure 1). In ABCC members with the “short” arrangement, the N-terminal extension is cytoplasmic. The SUR proteins and many of the MRPs possess the “long” arrangement, while the rest of the C subfamily members possess the “short” configuration (18). SUR proteins are atypical members of the ABCC subfamily as they do not possess any + intrinsic transport activity, but instead form the regulatory subunits of ATP sensitive K (KATP) channels (4). There are 2 genes encoding SUR proteins, SUR1 (ABCC8) and SUR2 (ABCC9).

In addition, there are multiple splice isoforms of SUR2, (SUR2A-2D) splice variants (2, 19). SUR1 is abundant in the brain and pancreas while SUR2A is predominantly expressed in the heart and SUR2B within vascular smooth muscle cells (2).

1.4 KATP channels

ATP-sensitive potassium (KATP) channels are multimeric protein complexes that are found in metabolically active tissues and are highly selective to potassium ions (10). By sensing changes in the ATP and ADP concentrations, KATP channels couple metabolism of the cell to the membrane potential (10). As a result, these channels regulate many biological processes such as insulin secretion in the pancreas, the release of neurotransmitters in the brain, and triggering of action potentials in the heart (10, 20). In response to cardiac stress in the heart (e.g. ischemia, hypertension, physical exertion) KATP channels open causing an efflux of potassium ions (10, 20). Consequently, this causes an efflux of calcium that results in depolarization of the cell and shortening of action potentials which is believed to protect the heart against arrhythmias and is known to reduce potential cardiac damage (10, 20).

KATP channels are comprised of four copies of a pore-forming inwardly rectifying potassium channel (Kir6.1 or Kir6.2) that are surrounded by four copies of a SUR protein

(Figure 2) (21, 22, 23). Different combination of Kir6.x and SUR proteins form different KATP channels in different tissues (Table 1) and have different pharmacological and nucleotide sensitivities. Proper regulation of gating of the KATP channel pore by the SUR proteins is critical, as mutations in the SUR proteins are associated with diseases, such as hyperinsulinism, diabetes, epilepsy, and cardiac disorders (10, 20).

Gating of KATP channels is a complex process that involves lipids, nucleotide, intracellular pH, interacting proteins (e.g. syntaxin 1A), and post-translational modifications (24,

25, 26, 27). Nucleotide-dependent regulation of KATP channel gating involves both the Kir6.x and SUR subunits. In the absence of Mg2+, ATP binding at the Kir6.x proteins cause inhibition of the KATP channel, whereas MgATP binding and hydrolysis at the SUR NBDs causes channel activation (2, 21, 23). Conformational changes resulting from ATP binding and hydrolysis in the SUR NBD1/NBD2 heterodimer are transmitted to the MSDs, via the coupling helices, and also likely to the L0 linker. Structural changes in the SUR MSDs and NBDs are relayed to the Kir6.x

subunits through MSD0 and the L0 linker, respectively. The affinity of the Kir6.x subunits for ATP decreases and the channel opens.

SUR protein Kir6.x subunit

Minimum ABC protein

MSD0

Figure 2: Schematic top view of a KATP channel. In blue represents the 4 copies of Kir6.x and in red and yellow represents the 4 SUR proteins. Figure modified from 28*.

1.5 Intrinsically disordered proteins The traditional idea that the three dimensional structure is key for protein function has been challenged (29). Although proteins containing disorder and the occurrence of unfolded proteins were accepted, it was only until recently did the idea exist that these disordered residues contribute to protein function (30). Intrinsically disordered proteins (IDPs) have no well defined structure. Instead, they can adopt a wide ensemble of conformations in solution in their functioning states (30). Notably, IDPs exhibit biological activity (31). In fact there are far more IDPs and proteins with intrinsically disordered regions than there are folded proteins (31). Eukaryotes have more IDPs or proteins with significant regions of disorder than do both prokaryotes and Archaea, implying that they must have some functional role related to the complexity of the organism (31, 32). For example, IDPs and intrinsically disordered regions of proteins are common and necessary for cellular signaling pathways (32). In fact most disease- associated proteins contain significant amounts of disorder (31, 32). Due to the flexible nature of IDPs, the wide range of conformations often renders them promiscuous, binding to multiple targets (32). Upon interacting with its target molecule, the best- fitting conformation in equilibrium of the IDP is selected (33). Thus interactions of IDPs with their target molecules is described by an induced fit mechanism (32).

NBD1 of SUR2A has been predicted to contain an intrinsically disordered region at its N- terminus (herein referred to as N-tail), which is believed to have a transient interaction with the core NBD1 portion of the protein that is controlled by phosphorylation of the N-tail (10). Detailed nuclear magnetic resonance (NMR) studies of NBD1 containing the N-tail are not possible due to limited solubility of the protein. Thus, specific information regarding changes in N-tail conformations with phosphorylation were lacking and was a focus of this work.

1.6 Phosphorylation regulation of KATP channels Phosphorylation is one of the most common post-translational modifications that regulate protein function within cells (4). Phosphorylation sites have been identified in almost all eukaryotic ABC transporters (4). Therefore it is believed that all transporters in the ABC superfamily are regulated by phosphorylation to some extent (4). Phosphorylation of specific serine and threonine residues in SUR2A NBD1 by protein kinase A (PKA) has a stimulatory effect on KATP channels in cardiomyocytes (24, 25). Two phosphorylation sites have previously been identified in rat SUR2A NBD1 at T632 and S636 that are located in the N-terminus of NBD1 (10, 33, 34). The N-terminal is believed to be a disordered region (33, 34). Previous work in our laboratory has shown that phosphorylation mimics the removal of the N-terminal tail (10). This and other data suggest that phosphorylation disrupts the N-tail’s interaction with NBD1 enhancing ATP binding and consequently activates KATP channels (10). Phosphorylation and the N-tail must therefore play an important role in regulating ATP binding at the NBDs and ultimately on the activation of KATP channel conductance. 1.7 N-terminal extension (N-tail) of SUR2A Since we have previously shown that phosphorylation of the N-tail disrupts its interaction with NBD1, we wanted to focus our future studies on the isolated N-tail to more thoroughly understand the regulation mechanism (10). The work in this thesis employed constructs of N-tail with boundaries S615-E664, S608-E664 and Q600-E664. With the inclusion of the N-terminal tail, the NBD1 construct used in our laboratory was predicted to have an N-terminus boundary between Q600 and T618 (33). Studies of the various NBD1 constructs that includes the N-tail found that S615-L933 was soluble and could be studied at high (>500 µM) concentrations. The NBD1 protein from S608-L933 was also soluble, but aggregated at high concentrations and Q600-L933 was completely insoluble (33). Since domain boundaries affect solubility, protein

6 constructs aim to include all structural portions and exclude flexible regions which contribute to protein aggregation (35). This explains why the inclusion of 8 to 15 extra residues from the flexible N-terminal tail contributes to compromised solubility of NBD1 S608-L933 and Q600- L933 respectively. Thus the studies of NBD1 conducted in our lab utilized the S615-L933 NBD1 boundaries (33). For comparison, the N-tail proteins with the N-terminal boundary of S615 (i.e. S615-E664) was used. Since the exact N-terminal of NBD1 begins between somewhere between Q600 and T618 and the S615-E664 (N-tail-S615) construct under study may not provide a complete representation of the N-tail, the other two boundaries for N-tail were also considered (S608-E664 and Q600-E664), N-tail-S608 and N-tail-Q600, respectively.

1.8 Biophysical techniques to study SUR2A N-tail

1.8.1 Nuclear magnetic resonance Nuclear magnetic resonance (NMR) spectroscopy is a powerful technique that can provide an atomic resolution description about protein structure, dynamics and reaction kinetics (36, 37). By exploiting the magnetic properties of atomic nuclei, NMR spectroscopy provides information about the chemical environment of atoms and, in turn provides atomic-level structural data about biomolecules (37). Nuclear magnetism and the notion of NMR spectroscopy stems from the theory of quantum mechanics, in particular the spin angular momentum (the nuclear spin quantum number, I) (37). The NMR signal solely depends on the existence of a non-zero nuclear spin (37). In general, nuclei with odd mass numbers have fractional half spin quantum numbers, i.e. x/2, where x is a non-zero integer number. Nuclei with even mass numbers and odd atomic numbers have a positive, non-zero integer quantum number. And nuclei with even mass numbers and even atomic numbers have a zero spin quantum number rendering them NMR inactive and thus invisible in NMR experiments (37). NMR spectroscopy of biomolecules, such as proteins and nucleic acids involves 1H, 13C, 15N, and 31P nuclei, all of which have an I = ½ (37). The 19F nucleus, which also has an I = ½, can be employed with specific labeling of the protein (38). Nuclei with non-zero spin angular momentum possess nuclear magnetic moments, µ that are proportional to I µ = γI,

7 where γ is the gyromagnetic ratio, a characteristic constant for each nuclei and determines the sensitivity of the given nucleus within the NMR spectrometer (Table 2) (37). Table 2: List of the most important nuclei found in biomolecules and their gyromagnetic ratios Nucleus I γ (Ts)-1

1H ½ 2.675 × 108

13C ½ 6.728 × 107

15N ½ -2.713 × 107

19F ½ 2.518 × 108

31P ½ 1.084 × 108

In the absence of an external magnetic field (Bo), nuclear spins are oriented in all possible directions (39). However when Bo is applied, all spins become aligned parallel or anti-parallel to

Bo (39). By convention, Bo is always along the z-axis. The predominant nuclei used in NMR studies of biomolecules have I=½. Thus, these nuclei have only two possible nuclear spin states, m= +½ and m= -½ (37, 39). The energy of each spin state depends on the presence and its interaction with Bo (37, 39).

E=-µBo

Like other spectroscopic techniques, a transition occurs from the ground state to the excited state via the absorption of a photon (39). However, unlike UV-Vis where an electron is promoted in the transition, short pulses of radiofrequency (RF) electromagnetic radiation are used to promote a nuclear spin from its ground state to the excited state in NMR (37, 39). An external magnetic field such as that of Bo is required to generate the differences in energy between the ground and excited states as illustrated in Figure 3 (37, 39).

Figure 3: Energy separation of nuclear spins in the presence of an external magnetic field. Upon applying a magnetic field to a spin ½ particle, the energy separation between the ground and excited state increases with increasing magnetic field strength (39).

Longitudinal magnetization describe the projection of the spins along the z-axis, while transverse magnetization describes the projection of spins along the x and y-axes, which is random and incoherent at thermal equilibrium, causing them to cancel out (37, 39). In addition to causing an orientation of the spins along the +z and –z axes, the presence of

Bo also induces a torque, which causes the spins to precess about the Bo field with a frequency known as the Larmor frequency (ωo) (37).

Ωo = -γBo The Larmor frequency is the resonance frequency of each nucleus, which ultimately is the signal generated by NMR (39). Since this frequency depends on Bo it becomes important to recognize that although Bo is static, not all nuclei detect the magnitude of this magnetic field due to chemical shielding by electron density (37, 39). Electrons are also known to precess due to the

Bo field generating an additional magnetic field that can enhance or oppose the static Bo (39). The net magnetic field felt by each nucleus is therefore slightly different depending on their chemical environment (39). As the electron density increases, the local magnetic field, Blocal, experienced by the nuclei is decreased lowering the resonance frequency (39).

Blocal = (1-σ)Bo Since the resonance frequency of a nucleus is altered by its chemical environment, it is also often called a chemical shift (39). Thus, due to different amounts of local electron density, nuclei experience different local magnetic fields resulting in different chemical shifts that reflect their environment (39). The NMR detects the precession of the bulk magnetization, M, by a coil in the xy plane.

Since at equilibrium, M is parallel to Blocal in the z-axis there is no signal produced (37). A short 90º radiofrequency (RF) pulse flips M to the xy plane. Once the RF is turned off, M will precess around Bo with its Larmor frequency, which ultimately generates a detectable signal in the xy plane as the transverse magnetization is now coherent and nonzero (37, 39). The precession of this coherent magnetization stimulates a current in the receiver coil creating a signal, known as the free induction decay (FID), which encodes the resonance frequency or chemical shift of each nucleus (39). An important pulse sequence for protein NMR is the heteronuclear single quantum coherence (HSQC) experiment that describes the correlation between protons (1H nuclei) and their attached heteronuclear spins, such as nitrogen (15N nuclei) (39). When studying large

9 biomolecules like proteins, heteronuclear NMR is advantageous since it reduces peak overlap and resonances are more dispersed and thus have distinct chemical shifts (39). In general an HSQC involves the transfer of magnetization from the proton to the heteronucleus (e.g. 15N) using the insensitive nuclei enhanced by polarization transfer (INEPT) step (39). Subsequently, chemical shift evolution occurs for each 15N nucleus and a second INEPT sequence transfers the magnetization back to proton for chemical shift detection. Consequently all amides (15N1H groups) generate a peak in the 2D HSQC spectrum. Since all residues, with the exception of proline, contain an amide 1H, each peak correlates to an amino acid of the protein, with a few additional peaks from 15N1H groups present in side chains, as shown in Figure 4. In general, the 15 1 15 1 + N-terminus N H proton is not observed, as it is an N H3 group that is in rapid exchange with the bulk solvent. Because residues in folded proteins are positioned in unique environments, their peaks are well dispersed and easily discrete from one another. However, in IDPs, the high degree of dynamics and flexibility causes the residues to experience an overall similar, averaged environment, causing significant resonance overlap, especially in the 1H dimension. Nonetheless, resonance assignment is possible for unfolded proteins. Once each signal in the HSQC has been mapped to a specific 15N1H pair in the macromolecule, the HSCQ provides residue-level information about protein structure.

Figure 4: Example of a 2-dimensional (2D) 15N HSQC spectrum with the 1-dimensional (1D) spectra for proton and nitrogen illustrated to the left and above the 2D spectrum, respectively. Each peak represents the signal from every 15N1H pair. Thus residues with additional side chain 15N1H groups will produce more than one peak per amino acid (39).

1.8.2 Circular dichroism Circular dichroism (CD) is a useful technique to determine the global protein secondary structure, conformation and stability in a wide range of environmental conditions (40, 41). In general, macromolecules are asymmetric or chiral, such that they are non-superimposable with their mirror image (41). The characteristic of chirality renders biological molecules, such as proteins, nucleic acids, and carbohydrates, optically active allowing them to rotate light passing through the sample either left or right (41). Overall there are two major methods that exploit chirality and optical activity to provide structural data of biomolecules (41). The first method is optical rotatory dispersion in which chiral molecules rotate linearly polarized monochromatic light by an angle of rotation, α (41). The other method is circular dichroism, in which chiral molecules differentially absorb the left (AL) and right (AR) circularly polarized light components of plane polarized light to give

ΔA = AL – AR = εLlc – εRlc = Δεlc where ε is the molar extinction coefficient, l is the path length, and c is concentration of the analyte (41). This differential absorption creates elliptically polarized light, where the elliptic angle is represented by Θ and are related by

Θ = 2.303(AL-AR)/4l (40). In order to compare the CD spectra of different proteins, mean residue ellipticity [Θ] is used and is defined as [Θ] = Θ ×100×M/cln, where M is molecular weight and n is the number of residues in a protein (41). CD is predominantly used to predict the secondary structure of proteins since the peptide bond is asymmetric or chiral, thus rendering proteins optically active (41). To use the amide bond (i.e. peptide bond) as a chromophore, CD must be measured below 250 nm in the far UV

* * range (41). Amides have two electronic transitions in the far UV-range, n  π and πo  π exhibiting a CD peak at 215-230 nm and 185-200 nm, respectively (40). CD spectroscopy is therefore usually collected in the far UV-range (180-250 nm) where negative signals are due to the absorption of left circularly polarized light and positive signals are caused by right circularly light absorbance (41). A positive peak at approximately 195 nm and a double negative band at 208 and 222 nm is characteristic of α helical secondary structure, whereas a significant negative peak around 200 nm and no signal at 215 nm is indicative of an unfolded or disordered protein

11 illustrated in Figure 5 (41). Thus CD can provide a quick estimation of whether a protein is unfolded and overall can also be used for determining the fraction of residues involved in a particular secondary conformation (41).

Figure 5: CD spectra of protein secondary structures – α-helical (●), β-structured (○) and random coil (□). For a typical α-helical protein, the spectrum consists of a large positive peak at 195 nm and 2 negative peaks at 208 and 222 nm. Like the α-helix, β-structured proteins produce a CD spectrum with a positive peak at 195 nm, but only a single negative peak at 215 nm. The random coil spectrum, unlike the other spectra, have a negative peak at 195 nm and no signal at 215 nm. Each transition is due to a characteristic electronic transition (42).

Certain additives can be incorporated into sample buffers to increase the secondary structural propensity of a protein (41). For example, trifluoroethanol (TFE), hexafluoroisopropanol, ethylene glycol and glycerol are a few of such additives used in circular dichroism studies (41). TFE is the most commonly used and has been shown to increase the helicity of peptides that have a tendency to adopt α-helical structure (41). TFE excludes water forcing the peptide backbone to hydrogen bond with itself (43). This induces α-helicity only in proteins and peptides with α-helical propensities.

1.8.3 Dynamic light scattering Dynamic light scattering (DLS) is a spectroscopic technique that determines the dimensions of biomolecules by calculating their hydrodynamic radius. By measuring Brownian

12 motion, DLS can relate the velocity of a particle to particle’s overall size (44). Brownian motion describes the movement of particles as a result of colliding with solvent molecules (44). In general, larger molecules exhibit a slower Brownian motion whereas smaller particles move faster. Smaller particles move further and faster upon colliding with solvent molecules compared to larger particles (44). From the Brownian velocity (translational diffusion coefficient, D), the hydrodynamic radius (dr) can be calculated using the Stokes-Einstein equation:

dr = kT/6πηD where k is Boltzmann’s constant, T is the absolute temperature and η is the viscosity of the solution (44). The hydrodynamic radius is expressed as the radius of a sphere that has the same translational diffusion coefficient, D, as that of the particle (44). Factors that affect D and ultimately the diffusion velocity of a particle are salt concentration, surface structure and core size as described in Figures 6 to 8 (44, 45).

Figure 6: Comparison of the hydrodynamic diameter under different salt concentrations. (A) Under low salt concentrations, the particle has a larger hydrodynamic diameter compared to the same particle under (B) high salt concentrations (modified from 44, 45). Green represents the particle of interest and red represents the predicted size of the particle in different salt concentrations.

Figure 7: Effect of surface projections on the hydrodynamic diameter. The degree of projections from the particle surface or the compactness of the particle can also influence the apparent size. For example, (A) more extended structures will exhibit a larger hydrodynamic diameter in comparison to (B) a particle that is more globular and compact (modified from 44, 45).

Figure 8: Determination of the hydrodynamic diameter of non-spherical particles. Non-spherical particles that are extended exhibit a diameter of a sphere that has the same translational diffusion speed as the non- spherical particle of interest (modified from 44, 45).

To measure Brownian motion, DLS applies a beam of light through the sample and measures the rate at which the intensity of the transmitted, scattered light fluctuates as the particles move (44). When the particles are illuminated by the DLS laser, the detector screen displays a “speckle pattern” (44). Scattered light additions create dark spots in areas of destructive interference and form light spots where there is constructive interference (44). Since the particles are in constant motion, the “speckles” are also moving (44). The rate at which these intensity fluctuations occur ultimately depends on the particle size, as smaller particles move more quickly causing an increase in intensity fluctuations compared to larger ones (44). Since it is difficult to measure these intensity fluctuations directly, the DLS instrument uses a correlator (44). A correlator compares the changes in intensity with time. In general, it has been found that for smaller particles the signal decays quicker over time compared to larger particles. Using the correlator, DLS computes a correlation function that can be related to the scattered light intensity (44). And with the use of algorithms, the hydrodynamic radius (i.e. size) can be calculated from the correlation function in one of two ways (44). The first uses a single exponential fit of the correlation function to provide the mean size, or z = average diameter (44). The second method uses multiple exponential fits to determine the distribution of particle sizes (44). The distribution is a plot of relative light intensity versus particle size, also known as the intensity size distribution (44). Intensity can also be converted into volume by using Mie theory to generate a volume distribution (44).

1.9 Resonance assignments

One of the key and most powerful uses of NMR spectroscopy is its ability to provide residue-level information about proteins. To extract such information, it is necessary to generate chemical shift assignments that relate each 15N1H peak of an HSQC to a particular amino acid in the protein sequence (36). When resonance assignments are available, structural and dynamic information provided by NMR spectroscopy can be associated to specific atoms and amino acids in the protein (36). Chemical shift assignments are primarily obtained through 3D triple resonance experiments (36). As the name suggests, these experiments correlate three types of nuclei: 1H, 13C, and 15N. Thus, proteins need to be uniformly labeled with 15N and 13C for these experiments (see Materials and Methods). The 2D 15N1H HSQC requires only 15N-labelling. Many experiments have been designed to relate the three nuclei comprising the protein backbone (36). These experiments are often analyzed in pairs – experiment A records the intra- (i) and inter- residue (i-1) resonances while experiment B only records resonances of the i-1 residue (36). The additional dimension of the 3D triple resonance assignment experiments compared to a 2D 15N1H HSQC allows for a better resolution of overlapped peaks into the 13C (third) dimension (36). Depending on the size of the protein, typically 4 to 6 experiments are used to assign the backbone. For small proteins (< 25 kDa), the two most important for backbone 15 1 assignments are the 3D HNCαCβ and the CβCα(CO)NNH (36). The HNCαCβ correlates the N H 15 13 13 and N resonances of residue i with the Cα and Cβ chemical shifts of i and i-1 residues 13 13 (Figure 9) (36). In addition, the Cα and Cβ signals are of opposite signs in the HNCαCβ and are coloured differently (e.g. red for negative peaks and black for positive peaks) allowing the Cα 15 1 15 and Cβ carbons to be distinguished. The CβCα(CO)NNH experiment links the N H and N 13 13 resonances of a residue with the Cα and Cβ of i-1. Comparison of the HNCαCβ and 1 15 CβCα(CO)NNH experiments at the same H and N plane, including the sign of the signals in the 13 13 HNCαCβ, allows one to distinguish the Cα and Cβ peaks of the i and i-1 residue (36). 13 13 Further, the Cβ resonances in the HNCαCβ experiment, the Cβ chemical shifts are 13 critical in backbone assignment of proteins, as they have less ambiguity than Cα chemical 13 shifts. Further, certain Cβ chemical shifts can identify specific residues (36). For example, only 13 13 the Cβ of alanine amino acids resonate at ~20 ppm, which is upfield of all other Cβ shifts and

13 only the C chemical shifts of serine and threonine resonate at > 60 ppm. As glycine does not have a Cβ it also gives rise to a unique pattern in the HNCαCβ and CβCα(CO)NNH experiments 13 13 (36). For ambiguous Cα / Cβ chemical shifts, a third useful experiment is the CCC-TOCSY which correlates the 15N1H and 15N resonances of residue i with the 13C of the i-1 residue (36). This experiment helps distinguish amino acids by recording the chemical shifts of the aliphatic 13C atoms (α, β, γ, δ, etc.), which together are distinct from residue to residue (Figure 10) (36). 13 13 Figure 9 links various HNCαCβ and CβCα(CO)NNH strips ( Cα / Cβ chemical shift combinations) in order to assign an 15N1H correlation in the HSQC to an amino acid in the 13 13 primary sequence. For example, panel 3 (HNCαCβ) illustrates the Cβ resonances in red and Cα resonances in black. The two downfield peaks at 62 and 68 ppm are resonances from threonine in the i-1 position. Threonine must be in the i-1 position because it correlates with panel 2 13 13 (CβCα(CO)NNH) that only shows the Cα and Cβ from the i-1 position and H and N chemical 13 13 shifts of amino acid i. Since threonine has characteristic Cα and Cβ resonances that are unique 13 from all other amino acids, these resonances are easily identifiable as being from threonine Cα 13 and Cβ. With these two peaks confirmed as the i-1 peaks, the remaining 2 peaks above must be 13 from the i residue. In this case it is alanine determined by the chemical shift of the Cβ resonance. Using these two experiments we can also move in the i+1 direction. From panel 3, we can locate panel 4 (CβCα(CO)NNH) that now puts our alanine carbon chemical shifts in the i-1 position. Thus the corresponding HNCαCβ (panel 5) illustrates alanine in the i-1 position and a 13 new amino acid in the i position yet to be assigned. In this case there is no observable Cβ and 13 with a distinguishable Cα chemical shift, the i+1 amino acid is assigned as a glycine. Glycine would then be moved to the i-1 position using the corresponding CβCα(CO)NNH illustrated in panel 6 and the process is repeated to assign the i+2 position. An analogous approach can be used to move through the sequence in the reverse direction.

i-1 i i+1

O H H O H H O H H O H α α α – C – N – C – C – N – C – C – N – C – C – N – α α α HC –OH C H H β β 3

CH3 Thr Ala Gly

1 2 3 4 5 6

HNCαCβ CβCα(CO)NNH HNCαCβ CβCα(CO)NNH HNCαCβ CβCα(CO)NNH

Cβi Cαi

Cβi-2 Cαi+1

13C Cαi Cβi

Cαi-2

Cβi-1 Cβi-1

Cαi-1 Cαi-1

Thr Ala Gly

Figure 9: Three dimensional NMR experimental results used for chemical shift assignments illustrating the protein sequence Arg-Thr-Ala-Gly. HNCαCβ and CβCα(CO)NNH strips used to assign back bone peaks.

13 13 For indistinguishable Cα and Cβ chemical shifts, a CCC-TOCSY experiment can be

used in addition to the HNCαCβ and CβCα(CO)NNH. Using panel 1’s HNCαCβ and its 13 13 corresponding CβCα(CO)NNH (panel 0) in Figure 10, the Cα and Cβ of the i-1 of our assigned 13 13 threonine (i) can be confirmed, based on its unique Cα and Cβ chemical shifts. However 13 13 unlike our previously assigned residues, the Cα and Cβ of the i-1 could correspond to a number of different amino acids. By including the CCC-TOCSY and thus resonances of the

aliphatic side chain 13C of the i-1 position (Figure 10 panel 7), we can confirm that the i-1 residue must be an arginine. 0 1 7 C βCα(CO)NNH HNCαCβ CCC-TOCSY

Cβi-1 Cβi-1 Cγi-1

13C Cδi-1

Cαi-1 Cαi-1

1H Thr Arg

Figure 10: Three dimensional CCC-TOCSY NMR experimental results used for chemical shift assignments of the protein sequence Arg-Thr. The CCC-TOCSY helps distinguish ambiguous amino 13 13 13 acids (i.e. with similar Cα and Cβ chemical shifts) based on the resonances of aliphatic C.

The peaks within each panel are grouped as a cluster that corresponds to a peak in the 2D HSQC, because they all have the same 1H and 15N chemical shifts. Thus by identifying the i residue of each cluster, the HSQC peak is consequently assigned. Due to phosphorylation or an interaction, a selected group of HSQC peaks shift due to a change in their chemical environment and can now be related to the protein’s primary sequence.

2. Materials and methods

2.1 Expression and purification of SUR2A N-tail Three constructs comprising the N-tail of rat SUR2A NBD1 were expressed as fusion proteins with an N-terminal-6xHis-SUMO cleavable tag in E.coli BL21 (DE3) CodonPlus-RIL cells (Stratagene). These constructs differed in the N-terminal boundary used. The construct extending residues S615-E664 is referred to as N-tail-S615. The construct extending from S608- E664 is referred to as N-tail-S608, while the longest construct extends from Q600-E664 and is referred to as N-tail-Q600. The cell cultures were grown at 37 ºC, with constant agitation in 13 15 100% M9 minimal media containing C-glucose and/or NH4Cl, for isotopic enrichment necessary for NMR spectroscopic studies. When the OD600 of the cultures reached 0.8 the incubation temperature was immediately reduced to 18 °C. The cells were incubated at 18 °C for 30 minutes at which point 0.75 mM isopropyl β-D-thiogalactoside (IPTG) was added to induce gene expression. After 20 hours of agitation at 18 °C, the cells were harvested by centrifugation and the pellets were stored at -20 °C. Purification of the 6xHis-SUMO SUR2A N-tail fusion proteins was conducted at 4 °C. Cell pellets from the 1 L culture were resuspended in 15 mL of lysis buffer (20 mM tris pH 7.6, 100 mM arginine, 150 mM NaCl, 2 mM β-mercaptoethanol, 5% (v/v) glycerol, 0.2% (v/v) Triton X-100, 2 mg/mL deoxycholic acid, 1 mg/mL lysozyme, 5 mM 6-aminocaproic acid, 5 mM benzamidine, and 1 mM PMSF). The cells were lysed by sonication and centrifuged for 40 minutes to remove cellular debris. The sonication and centrifugation steps were repeated and the cell lysates were combined, filtered through a 0.45 μM filter, and loaded onto a 5 mL High Performance Ni2+-NTA affinity column (GE Healthcare) that was pre-equilibrated with equilibration buffer (20 mM tris pH 7.6, 150 mM NaCl, 5 mM imidazole, 5% (v/v) glycerol, and 2 mM β-mercaptoethanol). The column was washed with 30 mL of equilibration buffer and the protein was eluted in 5 mL fractions with elution buffer (20 mM tris pH 7.6, 150 mM NaCl, 400 mM imidazole, 5% (v/v) glycerol, and 2 mM β-mercaptoethanol). Eluents containing the 6xHis-SUMO SUR2A N-tail fusion protein were combined and diluted 3-fold to reduce the final imidazole concentration to less than 200 mM. The 6xHis-SUMO tag was removed from N-tail with the addition of 6xHis-Ulp1 protease. The mixture containing, isolated N-tail, the 6xHis- SUMO tag, the 6xHis-Ulp1, and contaminants that co-eluted from the Ni2+ column, was

19 concentrated using a 3 kDa molecular weight cutoff (MWCO) centrifugal filter (Millipore) to 3 mL. The concentrated mixture was subsequently loaded onto a 24 mL size exclusion column (Superdex 75, GE Healthcare) that was pre-equilibrated with gel filtration buffer (20 mM tris, pH 7.6, 150 mM NaCl, 5% (v/v) glycerol, and 2 mM β-mercaptoethanol). Fractions containing the N-tail were collected, combined, and purified to homogeneity by a reverse Ni2+-NTA affinity column in 20 mM tris pH 7.6, 150 mM NaCl, 25 mM imidazole, 5% (v/v) glycerol, and 2 mM β- mercaptoethanol. The reverse Ni2+-NTA affinity step is necessary to remove remaining 6xHis- SUMO contaminants that were not separated in the gel filtration. Confirmation of the isolation and purity of the N-tail was determined by SDS-PAGE, mass spectrometry, and amino acid analysis. Mass spectrometry and amino acid analysis were conducted by the SPARC facility at the Hospital for Sick Children. For NMR studies, the purified N-tail was buffer exchanged into NMR buffer (20 mM Na+ phosphate, pH 7.2, 150 mM NaCl, 2% (v/v) glycerol, and 2 mM TCEP-HCl) using 3 kDa MWCO concentrator.

2.2 Phosphorylation of SUR2A N-tail Phosphorylation reactions were carried out at 30 °C on 200 μM purified samples of the different N-tail proteins. Since the NMR buffer of the non-phosphorylated N-tail does not contain MgATP, 15 mM ATP and 10 mM MgCl2 were added to the sample, along with 1000 units of the catalytic subunit of protein kinase A (PKA, Promega). The phosphorylation reaction was monitored by recording a series of 2 hour 2D 15N1H HSQC spectra over the course of 16 hours. Phosphorylation was confirmed by mass spectrometry on tryptic fragments of non- phosphorylated and phosphorylated N-tail (SPARC, Sick Kids Hospital). The phosphorylated N- tail was isolated from ATP and PKA, and exchanged into the NMR buffer using size exclusion chromatography.

2.3 Protein concentration determination

Protein concentration was determined by A280 readings of samples in 6 M guandinium HCl using an extinction coefficient of 8380 M-1cm-1. The extinction coefficient was calculated using ε280 nm = (# Trp)(5500) + (# Tyr)(1490) + (# Cys)(125) (45)

2.4 2D NMR spectroscopy NMR spectra of the non-phosphorylated and phosphorylated N-tail were recorded at 12 °C on a 600 MHz Varian VNMR S spectrometer equipped with a room temperature H(F)CN triple resonance probe or at 10 °C on an Agilent DD2 700 with a 5 mm Xsens cold probe. Chemical shifts were referenced to 4,4-dimethyl-4-silapentane-1-sulfonic acid (DSS) for all spectra (46, 47). Spectra were recorded on 200 µM samples of non-phosphorylated and phosphorylated N-tail (N-tail-S615, N-tail-S608, and N-tail-Q600) proteins in NMR buffer. Samples were buffer exchanged into the same stock of NMR buffer to ensure that both samples of the N-tail used were recorded in identical solution conditions. NMR spectral data were processed using NMRPipe/NMRDraw and were analyzed with NMRView (48, 49). Resonance assignments of non-phosphorylated N-tail-S615 were obtained using the standard HNCαCβ, CβCαCONNH, CCC-TOCSY, and HNCO triple resonance experiments. These experiments were recorded on a 700 MHz Agilent DD2 700 with a 5 mm Xsens cold probe at 10

°C (50). Three dimensional HNCαCβ, CβCα(CO)NNH and CCC-TOCSY spectra of the phosphorylated N-tail-S615 were also recorded at 10 °C on a 700 MHz NMR spectrometer. Chemical shifts were referenced to DSS for all spectra (47). Spectra were recorded on 400 µM samples of N-tail-S516 in NMR buffer. NMR spectral data were processed using NMRPipe/NMRDraw and were analyzed with NMRView (48, 49). 15N1H HSQC spectra used to compare structural changes in N-tail proteins with phosphorylation were recorded at 12 °C on a 600 MHz Varian VNMR S spectrometer. 15N1H HSQC spectra were also recorded to assess interaction of non-phosphorylated and phosphorylated N-tail-S615 or N-tail-S608 with NBD1. The interaction experiment samples contained 40 μM of 15N-labelled non-phosphorylated or phosphorylated N-tail-S615 or N-tail- S608 with 400 μM unlabeled NBD1. Control spectra containing 40 µM of 15N-labelled non- phosphorylated and phosphorylated N-tail-S615 or N-tail-S608 were also recorded.

2.5 Circular dichroism Circular dichroism (CD) spectra of the non-phosphorylated and phosphorylated N-tail- S615 and N-tail-S608 with increasing concentrations of trifluoroacetic acid (TFE) were recorded from 190 to 260 nm. The CD experiments were run at 12 °C on an Aviv 250 CD spectrometer

(Aviv Biomedical Inc., Lakeview, NJ), with a bandwidth of 1 nm and using a 1 mm path length quartz cell. Each spectrum was averaged from a total of five scans and were blank corrected. Samples contained 50 μM non-phosphorylated or phosphorylated N-tail-S615 or N-tail-S608 in 20 mM Na+ phosphate pH 7.2, 150 mM NaCl, 2% (v/v) glycerol, and 2 mM TCEP, and 0% (v/v) to 80% TFE (v/v).

2.6 Analytical size exclusion chromatography Purified non-phosphorylated and phosphorylated N-tail-S615 or N-tail-S608 samples were applied and run on a 24 mL size exclusion column in the NMR buffer at 0.5 ml/min. Sample volumes were 100 μL and sample concentrations were 200 μM. Standard samples (Gel Filtration Calibration Kit LMW, GE Healthcare) were also run on the size exclusion column under identical conditions. The standard samples used were blue dextran and the proteins conalbumin, carbonic anhydrase, ovalbumin, ribonuclease A and aprotinin. Samples of non- phosphorylated N-tail-S615 or N-tail-S608 (100 μL, 200 μM) and phosphorylated N-tail-S615 or N-tail-S608 (100 μL, 200 μM) were also run on the size exclusion column in NMR buffer containing 6 M guanidinium HCl and a flow rate of 0.5 mL/min.

2.7 Dynamic light scattering Dynamic light scattering (DLS) experiments were performed in triplicates on a Zetasizer Nano Series Nano-2S (Malvern Instruments, UK) at 12 °C. Samples contained 50 µM of non- phosphorylated and phosphorylated N-tail-S615 in NMR buffer.

3. Results

3.1 Analysis of the primary sequence of N-tail-S615 Previous NMR in our laboratory on NBD1 proteins that either contain (S615-L933) or lack (D665-L933) N-tail, indicated that N-tail-S615 is disordered (10). PONDR predictions of SUR2A also suggests that N-tail is unfolded (10, 33, 51-53). The program AGADIR also indicates that N-tail is primarily disordered although it predicts some helical propensity at the C- terminal of N-tail-S615 (Figure 11).

T G S M G S W R T G 619 0.0 0.0 0.0 0.1 0.1 0.1 0.4 0.4 0.4 0.4

E G T L P F E S C K 629 0.1 0.0 0.0 0.0 0.1 0.2 0.3 0.4 0.4 0.4

K H T G V Q S K P I 639 0.3 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.1 0.2

N R K Q P G R Y H L 649 0.2 0.2 0.1 0.0 0.1 0.1 0.8 0.8 0.9 1.2

D N Y E Q A R R L R 659 1.4 3.2 6.6 8.8 9.6 9.6 9.4 8.1 5.0 0.0

P A E T E 664 0.0 0.0 0.0 0.0 0.0

Figure 11: SUR2A N-tail-S615 primary sequence and AGADIR prediction of secondary structure. T632 and S363 shown in red are the N-tail residues that are the sites of phosphorylation. In green are the few residues remaining from the tag. Residues predicted to have helical propensity are shown in blue.

3.2 Expression and purification of SUR2A N-tail-S615 The 6xHis-SUMO N-tail-S615 fusion protein DNA was expressed in E. coli BL21 (DE3) CodonPlus-RIL (Stratagene) cells. A pre- and post-induction sample of the culture was taken and loaded onto an SDS gel. Based on molecular weight the SDS gel confirmed that the N-tail-S615 fusion protein had successfully expressed in the E. coli cells. Purification of 6xHis-SUMO N- tail-S615 using Ni2+ affinity chromatography was confirmed using an SDS gel based on molecular weight (~20 kDa = ~13 kDa SUMO + 6.3 kDa N-tail-S615) (Figure 12A). A band corresponding to the molecular weight of the N-tail-S615 fusion protein was present in the

supernatant samples and not in the samples from the insoluble fraction, indicating that N-tail- S615 was expressed completely solubly (Figure 12A). Removal of the 6xHis-SUMO tag from N- tail-S615 was accomplished with Ulp1, a protease that recognizes the 3-dimensional structure at the C-terminus of SUMO (53). The purification of the Ulp1 digested-N-tail-S615 was continued with size exclusion chromatography where the N-tail-S615 (6.2 kDa) eluted at 14.34 mL, partially co-eluting with the larger 6xHis-SUMO tag (~13 kDa), a protein double the size of N- tail-S615 (Figures 12B, C). An elution volume of 14.34 mL is larger than the expected elution volume of 16.5 mL for aprotinin, a globular protein of 6.5 kDa used as a standard. The early elution volume for N-tail-S615 is evidence that N-tail-S615 is disordered, resulting in an increased hydrodynamic radius, and thus causes N-tail-S615 to elute at an earlier volume like a larger protein. Also note that the 6xHis-SUMO tag and N-tail elute with a 1:1 ratio, SUMO does

not contain Trp residues, thus reducing the A280 observed. N-tail-S615 contains one Trp. Due to the partial co-elution of N-tail-S615 with the 6xHis-SUMO tag (Figure 12C), fractions containing the N-tail-S615 were applied to a second (reverse) Ni2+ affinity chromatography column. Since the SUMO tag possess the 6xHis sequence, SUMO remains bound to the column, while the N-tail-S615 is found in the flow through and wash (Figure 12D). kDa STD P1 S1 P2 S2 L FT W E1 E2 E3 E4 E5 250

98 50 A 22 6xHis-SUMO N-tail-S615

600 N-tail-S615 500 400

B 280 300 A 200 6xHis-SUMO

100 0 10 11 12 13 14 15 16 17 18 19 20 Elution Volume (mL) 24

kD ST kD ST FT W W Ni a B a D 250 D 250 50 50

25 6xHis-SUMO 25

N-tail-S615 20 20 15 C 15 D 6xHis-SUMO 10 6xHis-SUMO 10

N-tail-S615 5 N-tail-S615

5 2

Figure 12: Purification of a 1 L cell growth of 6xHis-SUMO N-tail-S615 fusion protein. (A) Ni2+ affinity chromatography of 6xHis-SUMO N-tail-S615. Lanes (P1 and P2) and (S1 and S2) show the insoluble and soluble fractions, respectively, from two sonications, and L represents the combined lysates loaded onto the Ni2+ affinity chromatography column. The flow through (FT) and wash (W) lanes illustrate the proteins that did not bind the Ni2+ column and the E1-E5 lanes correspond to the five, 5 mL elution fractions. (B) A280 trace from the gel filtration purification step of N-tail-S615. The large absorbance peak is due to the elution of N-tail-S615 from the column with the 6xHis-SUMO tag eluting slightly earlier. (C) SDS PAGE analysis of size exclusion chromatography purification of N-tail-S615 from Ulp1 and 6xHis-SUMO tag. Fractions with the elution volume corresponding to 13.5 to 15.5 mL contain N-tail- S615. The elution volumes of molecular weight standard proteins (GE Healthcare) are indicated at the top of the trace in panel B. (D) Reverse Ni2+ affinity chromatography of N-tail-S615 following the gel filtration purification. The lanes labeled FT and W show the flow through and wash, respectively, of unbound N-tail-S615 to the Ni2+ column and the lane labeled NiB depict the 6xHis-SUMO tag retained on the Ni2+ beads. The beads were washed for a third time using the same wash buffer to remove additional N-tail-S615 that is non-specifically bound to the Ni2+ column (not shown).

The sample of N-tail-S615 used in all biophysical studies was purified to homogeneity as shown in Figure 13. To determine the purity of N-tail-S615, 30 µL of an N-tail-S615 sample at 500 µM was loaded onto an SDS gel in order to visualize small amounts of contaminants that failed to be removed during the purification process (Figure 13). The lack of contaminants observed indicates that our N-tail-S615 samples are highly pure. kDa STD NMR 250 50

22 16

6 N-tail-S615

Figure 13: SDS-PAGE analysis of purified N-tail-S615. This sample, and other samples of equal purity, is used for all biophysical studies and illustrates the success of the purification of N-tail-S615 to homogeneity.

3.3 Biophysical and structural characterization of SUR2A N-tail-S615

3.3.1 NMR studies of non-phosphorylated N-tail-S615 The purified sample of N-tail-S615 was concentrated to 200 µM and was characterized by NMR spectroscopy. The 2D 15N1H HSQC spectrum, recorded at 30 ºC of non-phosphorylated N-tail-S615 is shown in Figure 14. N-tail-S615 in its non-phosphorylated state exhibits an NMR spectrum characteristic of disordered proteins (Figure 15A). Due to the rapid conversion of disordered proteins between conformations, nuclei experience a similar averaged chemical environment, resulting in significant overlap within a narrow range of the spectrum between 7.8 to 8.5 ppm in the 1H dimension (37). This confirmed the prediction that N-tail-S615 remains disordered even when isolated from NBD1 (Figure 14) (10). Figure 15 illustrates that the disordered residues in the spectrum of NBD1 overlap with the isolated N-tail-S615 peaks (10).

Figure 14: Overlay of NBD1 (615-933) in black and isolated N-tail-S615 in orange. The disordered resonances of NBD1 overlap with peaks corresponding to the N-tail-S615 suggesting that the N-terminal tail of NBD1 is disordered and when isolated, the N-terminal tail remains unfolded (10). Figure taken from (10). Both spectra recorded are TROSY HSQCs at 30 ºC.

Due to significant amide exchange with the bulk solvent at high temperatures (e.g. 25 ºC and 30 ºC) and basic pH, NMR spectra of N-tail-S615 was collected at low temperature (12 ºC) and at pH 7.2 to minimize these effects and increase peak intensity (55). Peak intensity is much greater when spectra are recorded at 12 ºC rather than at 30 ºC (compare Figures 15A and B).

110.0 Non-phosphorylated N-tail-S615 at 12 °C

115.0

A 120.0

(ppm) N

125.0

130.0 10.00 8.00 7.50 7.00 9.50 9.00 8.50 1H (ppm)

110.0 Non-phosphorylated N-tail-S615 at 30 °C

115.0

B 120.0

(ppm) N 15

125.0

130.0 10.00 9.50 9.00 8.50 8.00 7.50 7.00 6.50 1 H (ppm) Figure 15: NMR spectra at different temperatures. (A) 15N1H HSQC recorded at 12 ºC at a pH of 7.2. (B) 15N1H HSQC spectrum recorded at 30 ºC at a pH of 7.6. Both samples had the same concentration and were recorded for the same amount of time. Reduced temperature and pH significantly increases the signal intensity in 15N1H HSQC spectra of N-tail-S615.

3.3.2 Phosphorylation of N-tail-S615 Previous studies have shown that the two phosphorylation sites in SUR2A NBD1 are actually in the N-tail (10, 25). Previous NMR studies also indicated that phosphorylation of N- tail disrupts its interactions with the NBD1 core (10). In addition to understanding the effect of phosphorylation on NBD1, we were interested in the structural changes to N-tail itself that occur with phosphorylation. Non-phosphorylated N-tail-S615 was incubated with PKA, the natural kinase of SUR2A, and MgATP. The reaction was monitored by NMR (Figure 16) based on the

27 shift of the indole 15N1H resonance of W616. As the phosphorylation reaction proceeds, this peak (10.0760 ppm, 129.4996 ppm) disappears and reappears downfield at 10.1211 ppm, 129.5878 ppm. Thus phosphorylation was monitored by the decrease in the non-phosphorylated peak (10.0760 ppm, 129.4996 ppm) and appearance of the new phosphorylation peak (10.1211 ppm, 129.5878 ppm) (Figure 16). Once phosphorylation was complete, the protein was exchanged into its original NMR buffer for comparison with non-phosphorylated N-tail-S615 (Figure 17, 18). In the initial buffering conditions, dithiothreitol (DTT) was used as the reducing reagent. And consequently, Bradford assays were used to measure concentration. However, the phosphorylated N-tail-S615 poorly binds the Bradford reagent leading to inaccurate concentration estimates. TCEP was used in place of DTT which absorbs at 280 nm. Therefore, without DTT present A280 concentration readings were used in place of a Bradford Assay as described in Materials and Methods 2.3. All non-phosphorylated N-tail-S615 experiments were repeated using TCEP in place of DTT and yielded identical results. All ATP is removed from phosphorylated N-tail-S615 samples since it also absorbs at 280 nm (Figure 17).

2 h after phosphorylation 110.0 of N-tail-S615

16 h after phosphorylation of N-tail-S615 115.0

A 120.0

(ppm) N

125.0

130.0 10.00 9.50 9.00 8.50 8.00 7.50 7.00 1 H (ppm)

W616* 129.0

129.4 B

N (ppm) N 129.8 15

130.2 10.16 10.12 10.08 10.02 1 H (ppm)

P NP P NP P NP P NP 2 hrs 4 hrs 6 hrs 8 hrs 129.40

N (ppm) N 15

129.80 10.120 10.080 10.120 10.080 10.120 10.080 10.120 10.080 1H (ppm) 1H (ppm) 1H (ppm) 1H (ppm) Ci

P NP P NP P NP P NP

10 hrs 12 hrs 14 hrs 16 hrs 129.40

N (ppm) N 15

129.80 10.120 10.080 10.120 10.080 10.120 10.080 10.120 10.080 1 1 1 1 H (ppm) H (ppm) H (ppm) H (ppm)

100

60 Non-phosphorylated N-tail-S615 Phosphorylated N-tail-S615 Cii 40

Intensity (%) Intensity 20

0 2 4 6 8 10 12 14 16 Time (hours)

Figure 16: Phosphorylation of N-tail-S615. Following the addition of PKA and MgATP, a series of 2 hour 15N1H HSQC NMR experiments are recorded at 30 ºC to monitor the phosphorylation reaction. (A) Phosphorylation reaction of N-tail-S615 monitored by NMR. The black spectrum represents 2 hours following the addition of PKA and MgATP and, in red represents the final spectrum of the same sample recorded 16 hours later. >99% phosphorylation was noted after 6 hours. (B) Phosphorylation was monitored by the decrease in the non-phosphorylated peak and appearance of the new phosphorylation peak shown in Ci and Cii. The peak monitored corresponds to the 15N1H indole from the only Trp present in the N-tail (W616). And its identity was determined based on its downfield chemical shift in 1H and 15N and by mutagenesis of NBD1 (10). In B and C, black represents the non-phosphorylated state and red corresponds to phosphorylation.

Buffer exchange of the phosphorylated N-tail-S615 into the original NMR buffer (i.e. the removal of MgATP and PKA from the phosphorylated N-tail) was accomplished by gel filtration (Figure 17A). Our NMR studies of phosphorylated N-tail-S615 confirmed that N-tail-S615 remains disordered with phosphorylation due to the limited resonance dispersion (Figure 17B).

600 ATP

500

400

280 300 A A 200 Phosphorylated N-tail-S615 100

0 11 12 13 14 15 16 17 18 19 20 Elution Volume (mL)

Phosphorylated N-tail-S615 110.0

115.0

B 120.0

(ppm) N 15

125.0

130.0

10.00 9.50 9.00 8.50 8.00 7.50 7.00 1H (ppm)

Figure 17: (A) Gel filtration trace used to restore the phosphorylated N-tail-S615 into the original phosphate buffer following the addition of MgATP and enzyme PKA. (B) Phosphorylated N-tail-S615.

Although NMR can be used to confirm phosphorylation, the addition of two phosphate groups illustrates a slight gel shift (Figure 18) due to the small difference in molecular weight between non-phosphorylated and phosphorylated N-tail-S615. kDa STD NP NP P P

98 50

22 16

Figure 18: SDS gel of non-phosphorylated (NP) and phosphorylated (P) N-tail-S615. Although there is a small mass difference of two phosphate groups (189.94 g/mol) between the two states, phosphorylated samples display a gel shift.

3.3.3 Structural changes associated with the phosphorylation of N-tail-S615 As shown in Figure 19, one of the most significant chemical shift changes due to phosphorylation occurs to the peak corresponding to the indole 15N1H resonance of W616. Because known phosphorylation sites at T632 and S636 are 24 and 30 residues away, respectively, in the disordered N-tail-S615, chemical shift changes of the W616 resonance implies that the N-tail-S615 possesses some residual structure, possibly bringing W616 close in space to the phosphorylation sites. To test this hypothesis we applied samples of the non- phosphorylated and phosphorylated N-tail-S615 to analytical native (Figure 20A) and denaturing (Figure 20B) gel filtration chromatography.

Non-phosphorylated N-tail-S615 110.0 Phosphorylated N-tail-S615

115.0

120.0

(ppm) N 15

125.0

130.0 10.00 9.50 9.00 8.50 8.00 7.50 7.00 1H (ppm)

Figure 19: Overlay of non-phosphorylated (black) and phosphorylated (red) N-tail-S615 15N1H HSQC spectra. This phosphorylated sample was restored into the original NMR buffer (i.e. PKA and MgATP was removed) to allow for an accurate comparison with the non-phosphorylated sample.

The phosphorylated N-tail-S615 elutes earlier (12.71 mL) than its non-phosphorylated counterpart (13.02 mL) (Figure 20A). Although both forms of the N-tail-S615 are extended, the non-phosphorylated form appears to be less extended. As a control, and to determine whether the phosphorylated N-tail-S615 was as extended as a random coil, we applied both samples to a denaturing column. Under denaturing conditions N-tail-S615 completely unfolds and thus is fully extended. As illustrated in Figure 20B both forms of N-tail-S615 under denaturing conditions elute at 10.56 mL, which is significantly earlier than either the non-phosphorylated

31 and phosphorylated forms under native conditions. These data are consistent with a model in which the N-tail exhibits some residual structure that is reduced, but not eliminated with phosphorylation.

Carbonic Anhydrase (29 kDa) Ribonuclease A (13.7 kDa) Aprotinin (6.5 kDa) 70 Non-phosphorylated N-tail-S615 60 Phosphorylated N-tail-S615 50

A 40 280

A 30

0 11 11.5 12 12.5 13 13.5 14 14.5 Elution Volume (mL)

50 Non-phosphorylated N-tail-S615 Phosphorylated N-tail-S615 40

280 A B 20

0 9 9.5 10 10.5 11 11.5 12 12.5 13 13.5 14 Elution Volume (mL)

Figure 20: (A) Elution profile of non-phosphorylated and phosphorylated N-tail-S615 from a native gel filtration column. (B) Elution profile from a denaturing column, equilibrated in 6 M guanidinium HCl in N-tail buffer.

We also performed dynamic light scattering (DLS) on the non-phosphorylated and phosphorylated forms of N-tail-S615. The hydrodynamic radius of non-phosphorylated N-tail- S615 is 0.856 nm, compared with 2.151 nm for the phosphorylated N-tail-S615 (Figure 21). Thus, both the gel filtration and DLS data indicate that N-tail-S615 is compact and that this compact structure is disrupted upon phosphorylation.

40 Non-phosphorylated N-tail-S615 35 Phosphorylated N-tail-S615

volume % volume 15

0 0.1 1 10 size (r.nm)

Figure 21: Dynamic light scattering of N-tail-S615. The hydrodynamic diameter of the N-tail-S615 increases with phosphorylation.

To study the apparent residual structure of the N-tail we performed circular dichroism experiments on N-tail-S615. According to our CD spectrum in Figure 22, the non- phosphorylated N-tail-S615 exhibits some α-helical structure since a positive and negative band appear at 195 nm and 222 nm respectively, characteristic of α-helices (41). We measured various CD spectra with increasing concentrations of trifluoroethanol (TFE) to the non-phosphorylated N-tail-S615 to test its helical propensity. Increasing TFE in the solutions results in increased helical secondary structure of N-tail-S615 (Figure 22A). In comparison, the phosphorylated N- tail-S615 has higher mean residue ellipticity values without TFE and displays no changes in the CD spectra with increasing TFE (Figure 22B, C). Thus, the CD and TFE experiments indicate that the non-phosphorylated N-tail-S615 possesses residual helical structure that is reduced with phosphorylation. TFE concentrations from 10% to 50% (v/v) consistently enhanced helicity in N-tail-S615. At concentrations >50% (v/v) this pattern was not observed and also precipitated N- tail-S615. Thus the remaining studies were performed using up to 50% (v/v) TFE.

3000

2000 1000 A 0 195 205 215 225 235 245 255 -1000 wavelength (nm) -2000

A cm²dmol¯¹) -3000 0% TFE -4000 20% TFE -5000 30% TFE (deg Ellipticity Residue Mean 40% TFE -6000

6000

4000 2000

) 1

- 0 190 200 210 220 230 240 250 260 -2000 wavelength (nm)

B dmol 2

-4000 cm 40% TFE -6000 50% TFE 60% TFE -8000 (deg Ellipticity Residue Mean 70% TFE 80% TFE -10000 6000

4000

2000

) 1 - 0 C dmol 2 190 200 210 220 230 240 250 260

cm wavelength (nm) -2000 0%TFE 20%TFE

-4000 30%TFE Mean Residue Ellipticity (deg Ellipticity Residue Mean 40%TFE 50%TFE -6000

Figure 22: Circular dichroism of (A) non-phosphorylated N-tail-S615 with increasing concentrations of TFE (0-50%). (B) non-phosphorylated N-tail-S615 with even higher concentrations of TFE (40 -80%). Although helical propensity is increasing, samples precipitated with >50% TFE. (C) Phosphorylated N- tail-S615 with increasing concentrations of TFE (0-50%).

3.3.4 Changes in the interaction between the N-tail-S615 and NBD1 with phosphorylation

Previous work in the lab indicated that when linked to NBD1, the N-tail makes transient interactions to the NBD1 core and that these residual interactions are disrupted with phosphorylation of N-tail (10). However, these studies employed constructs of N-tail linked to NBD1. The spectral overlap limited our ability to probe the interaction from the perspective of N-tail. In addition, resonance assignment could only be obtained with an NBD1 construct lacking N-tail. Thus residual-level detail of N-tail is lacking. In the NMR interaction experiments the non-phosphorylated N-tail-S615 is 15N-labeled, while NBD1 is unlabeled, and consequently invisible in the NMR experiment. Thus any change in spectra of N-tail-S615 upon addition of NBD1 is indicative of an interaction. Also, because we have assigned the resonances in the N- tail-S615 spectra to specific residues, chemical shift changes in N-tail-S615 from the addition of NBD1 will give information on which N-tail residues bind NBD1. And from our work with NBD1, phosphorylation should reduce the interaction between N-tail and NBD1, and consequently exhibit a spectrum with minimal to no changes (10). Since the N-tail and NBD1 are linked in the same protein a weak interaction between them is expected, which consequently makes it difficult to monitor these changes. In order to see these changes, we saturated N-tail- S615 binding by used a 10:1 ratio of NBD1:N-tail-S615. A 400 µM sample of NBD1 was used as higher concentrations results in NBD1 dimerization, and 40 µM of N-tail-S615 provides sufficient NMR signal with longer (8 hour) acquisition times. As shown in Figure 23A and B, there are some minor chemical shifts and changes in intensity in N-tail-S615 resonances in the presence of NBD1. Since some peaks remain the same while others shift and decrease in intensity, it suggests that the N-tail residues corresponding to these peaks must be in contact with NBD1. One possible way to promote binding would be to use TFE in the interaction experiment, as the regions of N-tail with helical propensity may be involved in binding of NBD1, as has been seen in other proteins (56, 57). When a similar experiment was conducted using the phosphorylated N-tail-S615, fewer changes were observed (Figure 24A, B). An almost perfect overlay of the two spectra (APO-phosphorylated N-tail-S615 with NBD1-phosphorylated N-tail- S615) supported our hypothesis that phosphorylation reduces the N-tail’s interaction with NBD1. In N-tail-S615 and NBD1 interaction experiments, the small changes observed are expected because the two proteins are naturally linked within the KATP channel and as a result only require

35 a weak association. Furthermore the weak interaction between N-tail-S615 and NBD1 could also be explained by the entropic cost of binding.

APO non-phosphorylated 110.0 N-tail-S615 NBD1 non-phosphorylated N-tail-S615

115.0

A 120.0 (ppm) N

125.0

130.0 10.00 9.50 9.00 8.50 8.00 7.50 7.00 1H (ppm)

122.5 120.0 B

N (ppm) N N (ppm) N

121.0 15 123.5

8.200 8.000 8.550 8.450 1 1 H (ppm) H (ppm)

Figure 23: Interaction experiment comparing non-phosphorylated N-tail-S615 with/without NBD1. (A) A 10:1 ratio of 14N-ΔN-NBD1 and non-phosphorylated 15N-tail-S615, and corresponding control NMR spectra were recorded and compared. In the background, the black spectrum is APO-non-phosphorylated N-tail-S615 (i.e. without NBD1), and the blue spectrum in the foreground is a sample containing a 10:1 ratio of 14N- ΔN-NBD1 and 15N-tail-S615. (B) Some peaks shift and decrease in intensity suggesting that non-phosphorylated N-tail-S615 must be in contact with NBD1 via the residues corresponding to these peaks.

APO phosphorylated 110.0 N-tail-S615 NBD1 phosphorylated N-tail-S615

115.0

A 120.0

(ppm) N 15

125.0

130.0 10.00 9.50 9.00 8.50 8.00 7.50 7.00

1H (ppm)

122.5 120.0 B

121.0 N (ppm) N

(ppm) N 15

123.5 8.550 8.450 1 8.200 8.000 H (ppm) 1 H (ppm)

Figure 24: Interaction experiment comparing phosphorylated N-tail-S615 with/without NBD1. (A) A 10:1 ratio of 14N- ΔN-NBD1 and phosphorylated 15N-tail-S615, and corresponding control NMR spectra were recorded and compared. In the background, the red spectrum is APO-phosphorylated N-tail-S615 (i.e. without NBD1), and the blue spectrum in the foreground is a sample containing a 10:1 ratio of 14N- ΔN-NBD1 and phosphorylated 15N-tail-S615. (B) In the presence of NBD1 there are fewer changes to the phosphorylated N-tail-S615 spectrum.

3.4 Expression and Purification of SUR2A N-tail-S608 and N-tail- Q600

Although these interaction experiments between NBD1 and N-tail-S615 agree with our hypothesis, the shifts illustrated in Figures 23 and 24 are not significant enough to confirm our claim. This led us to re-evaluate the N-tail construct used in our experiments. Previous work in our lab indicates that the linker between the end of transmembrane segment 11 and the linker between the last transmembrane helix in MSD1 and NBD1 is comprised of residues Q600-E664 (1, 10). Homology modelling, sequence alignments and predictions of transmembrane segment 11 based on other ABC protein structures suggests that V599 is the end of helix 11. And residues Q600-L607 then extend helix 11 into the cytoplasm with the core folded NBD1 beginning at

37 residue D665 (10, 31, 57-67). Consequently, Q600, S608 or S615 were used as the N-terminal and E664 as the C-terminal boundaries of the N-tail construct. Because we wish to compare the isolated N-tail studies to studies of N-tail linked to NBD1, we must take into consideration the solubility of NBD1-containing constructs with different N-terminal boundaries. NBD1-containing construct starting at Q600 is insoluble, while an NBD1-containing construct starting at S608 is soluble but aggregated at high concentrations (8). Because NBD1-containing constructs starting at S615 were soluble and could be studied at high (>500 µM) concentrations, studies of NBD1 conducted in our lab utilized the S615-L933 boundary (8, 21, 58). For comparison, the N-tail was therefore first studied using boundaries S615-E664 of NBD1. However, because the MSD1-NBD1 linker begins at Q600, the S615-E664 construct under study may not provide a complete picture of the N-tail and may also affect the interaction studies with NBD1. This led to our study on other N-tail constructs with boundaries S608-E664 (N-tail-S608) and Q600-E664 (N-tail-Q600). As described in the Materials and Methods sections, N-tail-S608 and N-tail-Q600 were expressed and purified with the identical method used for the N-tail-S615 construct (Figure 26A- D and 27A-D). Since these constructs contain 7 and 15 more residues than N-tail-S615, the N- tail-S608 and N-tail-Q600 elute earlier (13.57 mL and 13.44 mL, respectively) from the size exclusion column compared to N-tail-S615 (14.38 mL), causing it to further co-elute with SUMO (Figure 25B-C, 26B-C). After applying the mixture of the N-tail-S608 or N-tail-Q600 and 6xHis-SUMO to the second Ni2+ affinity chromatography column, pure N-tail samples were obtained and used for biophysical studies (Figure 25D, 26D). In a 1 L purification of E. coli culture, N-tail-S615 yields approximately 3.5 g of protein, whereas N-tail-S608 yields 1.8 g and N-tail-Q600 yields 0.6 g of protein, respectively.

kDa STD P1 S1 P2 S2 L FT W E1 E2 E3 E4 E5 250 98 64

50 A 36

16 6 400 N-tail-S608 300

200 B 280

A 100 6xHis-SUMO

0 10 11 12 13 14 15 16 17 18 19 20 Elution Volume (mL)

kDa STD kDa STD 1NMR 2NMR 50 50 C 36 D 36

16 22 16

6 6 Figure 25: Purification of a 1 L cell growth of N-tail-S608 6x-His-SUMO fusion protein. (A) Ni2+ affinity chromatography. Lanes (P1 and P2) and (S1 and S2) show the insoluble and soluble fractions respectively from two sonications, and L represents the combined lysates loaded onto the Ni2+ affinity chromatography column. The flow through (FT) and wash (W) lanes illustrate the proteins that did not 2+ bind the Ni column and the E1-E5 lanes correspond to the five, 5 mL elution fractions. (B) A280 trace of from the gel filtration purification step of N-tail-S608. The large absorbance peak is due to the elution of N-tail-S608 from the column with the 6xHis-SUMO tag eluting slightly earlier. Although the tag and N- tail elute with a 1:1 ratio, SUMO does not contain Trp residues thus reducing the A280 observed. The N- tail-S608 contains 1 Trp amino acid. (C) SDS PAGE analysis of size exclusion chromatography purification of N-tail-S608 from the 6xHis-SUMO tag. Fractions with elution volume corresponding to 13.5 to 15.5 mL contain N-tail-S608. (D) SDS-PAGE analysis of purified N-tail-S608. This sample, and other samples of equal purity, are used for all biophysical studies and illustrates the success of the purification of N-tail-S608 to homogeneity.

kDa STD P- P+ P1 S1 P2 S2 L FT W E1 E2 E3 E4 98 50 36 A 16

300 250 200 6xHis-SUMO

280 150 and B A N-tail-Q600 100 50 0 8 9 10 11 12 13 14 15 16 17 Elution Volume (mL)

kDa STD 1 2 3 kDa STD 98 98 50 50 36 36

C 16 D 16 6 6

Figure 26: Purification of a 1 L cell growth of N-tail-Q600 6xHis-SUMO fusion protein. (A) Ni2+ affinity chromatography. Lanes (P1 and P2) and (S1 and S2) show the insoluble and soluble fractions respectively from two sonications, and L represents the combined lysates loaded onto the Ni2+ affinity chromatography column. The flow through (FT) and wash (W) lanes illustrate the proteins that did not bind the Ni2+ column and the E1-E5 lanes correspond to the five, 5 mL elution fractions. (B) A280 trace of from the gel filtration purification step of N-tail-Q600. The outlined absorbance peak is due to the elution of N-tail- Q600 from the column co-eluting with the 6xHis-SUMO tag. Although the tag and N-tail-Q600 elute with a 1:1 ratio, SUMO does not contain Trp residues thus reducing the A280 observed. The N-tail-Q600 contains 1 Trp amino acid. (C) SDS PAGE analysis of size exclusion chromatography purification of N- tail-Q600 from the 6xHis-SUMO tag. Fractions with elution volume corresponding to 13.5 to 15.5 mL contain N-tail-Q600. (D) SDS-PAGE analysis of purified N-tail-Q600. This sample, and other samples of equal purity, is used for all biophysical studies and illustrates the success of the purification of N-tail- Q600 to homogeneity.

3.5 Non-phosphorylated and phosphorylated N-tail-S608 and N- tail-Q600 NMR spectra and other characterization

To determine whether the additional residues S608-D614 and Q600-D614 alter the structure of the N-tail, NMR studies were conducted. 2D 15N1H HSQC spectra of N-tail-S608 and N-tail-Q600 have limited dispersion in the 1H dimension, demonstrating that N-tail comprising residues S608-E664 and Q600-E664 are also disordered (Figure 27 and 28). Thus, residues S608-D614 and Q600-D614 do not induce any structure to the N-tail (Figure 29). Recall, PONDR predicts that residues L607-E664 of SUR2A as being primarily disordered (10, 33, 51-53, 58, 60-68). Non-phosphorylated 110.0 N-tail-S608

115.0

N (ppm) N 120.0 15

125.0

130.0 10.00 9.50 9.00 8.50 8.00 7.50 7.00 1H (ppm) 15 Figure 27: HSQC NMR spectrum of a 200 µM sample of non -phosphorylated N N-tail-S608. Non-phosphorylated 110.0 N-tail-Q600

115.0

120.0

N (ppm) N 15

125.0

130.0

10.00 9.50 9.00 8.50 8.00 7.50 7.00 1 H (ppm) Figure 28: HSQC NMR spectrum of a 50 µM sample of non -phosphorylated 15N N-tail-Q600.

110.0 N-tail-S615 N-tail-S608 N-tail-Q600

115.0

N (ppm) N

15 120.0

125.0

130.0 10.00 9.50 9.00 8.50 8.00 7.50 7.00 1H (ppm) Figure 29: Overlay of the 3 N-tail constructs. In green is N-tail-S615, in pink is N-tail-S608 and in black is the N-tail-Q600 construct. The overlap of resonances indicates that the three constructs generate similar disordered N-tail proteins.

Phosphorylated N-tail-S608 and N-tail-Q600 also produce spectra characteristic of disordered proteins. Like N-tail-S615, a phosphorylation peak (circled in red in Figure 30A) appears in spectra of N-tail-S608. However, unlike N-tail-S615, the 15N1H of the indole side chain of W616 in spectra of N-tail-S608 upon phosphorylation does not shift (Figure 30B). Since the indole corresponds to residue W616, it is likely that phosphorylation has a greater impact when W616 is closer to the N-terminus (i.e. in the N-tail-615 construct). The phosphorylation- dependent changes of N-tail are even less apparent in the N-tail-Q600 construct (Figure 31). As shown in Figure 31, the indole 15N1H of W616 is not present in the spectra of N-tail-Q600 and the phosphorylation peak is weak in intensity. In addition, the sample of phosphorylated N-tail- Q600 is dilute resulting in peak intensity that is only slightly greater than the noise. Our data confirms that the phosphorylation sites transiently interact with the N-terminus of N-tail. With additional N-terminal residues in N-tail-S608 and N-tail-Q600, the charged N-terminus is further from the phosphorylation sites which may also explain the varied chemical shifts we see with W616. However, our other data is consistent amongst the three constructs and ultimately confirms that the N-tail is not completely extended as it exhibits some compact structure. With poor spectra of phosphorylated N-tail-Q600, N-tail-S608 was instead used to confirm the trends illustrated by N-tail-S615.

Non-phosphorylated N-tail-S608 110.0 Phosphorylated N-tail-S608

115.0

A 120.0

(ppm) N 15

125.0

130.0 10.00 9.50 9.00 8.50 8.00 7.50 7.00 1H (ppm)

129.0 W616*

129.4

N (ppm) N 15 129.8

130.2 10.16 10.12 10.08 10.04 1H (ppm) Figure 30: Overlay of 200 µM non-phosphorylated (black) and 200 µM phosphorylated (red) N-tail-S608 15N1H HSQC.

110.0 Non-phosphorylated N-tail-Q600 Phosphorylated N-tail-Q600

115.0

120.0

(ppm) N

125.0

130.0 10.00 9.50 9.00 8.50 8.00 7.50 7.00 1H (ppm) Figure 31: Overlay of 50 µM non-phosphorylated (black) and 20 µM phosphorylated (red) N-tail-Q600 15N1H HSQC.

We used analytical gel filtration chromatography to determine whether the disordered N- tail-S608 protein possesses compact structure that is also altered by phosphorylation. As observed with N-tail-S615, the non-phosphorylated N-tail-S608 is more compact than its phosphorylated counterpart (Figure 32), again suggesting that non-phosphorylated N-tail may have some residual structure that is reduced with phosphorylation. However, the phosphorylated form elutes 0.010 mL earlier than the non-phosphorylated N-tail-S608, less apparent than phosphorylated N-tail-S615 that eluted 0.39 mL earlier than non-phosphorylated N-tail-S615.

25 Non-phosphorylated N-tail-S608 Phosphorylated N-tail-S608 20

15 280 A 10

0 11 11.5 12 12.5 13 13.5 14 Elution Volume (mL)

Figure 32: Elution profile of non-phosphorylated and phosphorylated N-tail-S608 from a gel filtration column in native N-tail buffer.

To further test the apparent residual structure of our N-tail-S608 construct, we recorded CD spectra of the protein in the presence of increasing concentrations of TFE. Like N-tail-S615, the non-phosphorylated form of N-tail-S608 possesses helical propensity For example, the mean residue ellipticity at 222 nm with 50% TFE is -3153.69 deg cm2 dmol-1 in non-phosphorylated N-tail-S615 and similarly -3891.07 deg cm2 dmol-1 in N-tail-S608. In contrast to non- phosphorylated N-tail-S608, increasing concentrations of TFE do not induce helical propensity in phosphorylated N-tail-S608 (compare Figures 22B and 33B), a trend that is consistent with our other N-tail construct.

6000

4000

2000

0 190 210 230 250 -2000 wavelength (nm) A cm²dmol¯¹) -4000 0%TFE -6000 20%TFE

Mean Residue Ellipticity (deg Ellipticity Residue Mean 30% TFE -8000 40%TFE 50%TFE -10000

6000

4000

2000

0 B 190 200 210 220 230 240 250 260 -2000 wavelength (nm)

cm²dmol¯¹) -4000 0%TFE -6000 20%TFE

Mean Residue Ellipticity (deg Ellipticity Residue Mean 30% TFE -8000 40%TFE 50%TFE -10000

Figure 33: Circular dichroism spectra of (A) non-phosphorylated and (B) phosphorylated N-tail-S608 with increasing concentrations of TFE.

3.5.1 Changes in the interaction between N-tail-S608 and NBD1 with phosphorylation

Like our N-tail-S615 construct, N-tail-S608 does appear to interact with NBD1 in its non- phosphorylated state. However, these changes are even less significant than they are with the N- tail-S615 shown in Figure 34A and B. However, our phosphorylated N-tail-S608 results provided us with a conflicting observation. It appeared that the phosphorylated N-tail-S608 interacts to a greater extent than its non-phosphorylated partner (Figure 35A and B). Because we see greater chemical shift changes and decreased intensity in spectra of phosphorylated N-tail-

S608 when NBD1 is present compared to spectra of non-phosphorylated N-tail-S608, it is possible that the weak association and thus these minor changes may be insignificant. Also note that the observed changes in N-tail-608 are significantly smaller than the changes observed in the N-tail-S615 interaction experiments with NBD1.

110.0

115.0

(ppm) 120.0

125.0

130.0 10.00 9.50 9.00 8.50 8.00 7.50 7.00 1H (ppm)

122.5 120.0

B N (ppm) N

(ppm) N 123.5 15

121.0 15

8.550 8.450 1H (ppm) 8.200 8.000 1H (ppm)

Figure 34: Interaction experiment comparing non-phosphorylated N-tail-S608 with/without NBD1. (A) A 10:1 ratio of 14N-ΔN-NBD1 and non-phosphorylated 15N-tail-S608, and corresponding control NMR spectra were recorded and compared. In the background, the black spectrum is APO-non-phosphorylated N-tail-S608 (i.e. without NBD1), and the blue spectrum in the foreground is a sample containing a 10:1 ratio of 14N-ΔN-NBD1 and 15N-tail-S608. (B) There are minimal shifts and some reduced intensities. However, since some peaks shift while others do not suggests that N-tail-S608 is in contact with NBD1 via these corresponding residues.

APO phosphorylated 110.0 N-tail-S608

NBD1 phosphorylated N-tail-S608 115.0

A 120.0

(ppm) N 15

125.0

130.0 10.00 9.50 9.00 8.50 8.00 7.50 7.00 1H (ppm)

B 120.0 122.5

(ppm) N 121.0

N (ppm) N 15

123.5 15

8.550 8.450 1H (ppm) 8.200 8.000 1 H (ppm)

Figure 35: Interaction experiment comparing phosphorylated N-tail-S608 with/without NBD1. (A) A 10:1 ratio of 14N-ΔN-NBD1 and phosphorylated 15N-tail-S608, and corresponding control NMR spectra were recorded and compared. In the background, the black spectrum is APO-phosphorylated N-tail-S608 (i.e. without NBD1), and the blue spectrum in the foreground is a sample containing a 10:1 ratio of 14N- ΔN-NBD1 and 15N-tail-S608. (B) There are minimal shifts and reduced intensities. However, since some peaks shift while others do not suggests that the N-tail-S608 is in contact with NBD1 via these corresponding residues.

The weak interaction between N-tail and NBD1 is difficult to characterize by NMR. With minimal chemical shifts and slight reductions in peak intensity it is difficult to probe this interaction. Additional experiments with N-tail in TFE may help provide tighter binding to NBD1 and thus exhibit greater differences in the non-phosphorylated versus phosphorylated N- tail interaction experiments with NBD1 by NMR. In addition to the structural studies presented thus far, probing the interaction of NBD1 to N-tail under conditions that induce helical propensity of N-tail would provide us with a broader understanding of NBD1 within the channel.

3.6 Resonance assignment of non-phosphorylated N-tail-S615 One of the greatest strengths of using NMR spectroscopy to study proteins is its ability to provide residue level information. In order to obtain such information and extend our knowledge about the N-tail, we required resonance assignments. We also determined from our CD/TFE and analytical size exclusion chromatography data that the N-tail possesses some residual compact structure. With resonance assignments we will be able to confirm the identity of the residues that are close to the phosphorylation sites in space and localize the helical propensity to the primary sequence, providing more detailed structural data. With the use of 3-dimensional triple resonance experiments, we can map out and assign each peak from our 15N1H HSQC experiment to an amino acid from the N-tail’s protein sequence. In Figure 36 the C-terminus of the N-tail is assigned and exemplifies the how the triple resonance data were used to assign the 15N1H HSQC spectrum. Peaks in each triple resonance dataset within common 15N1H frequencies are linked or clustered and each cluster corresponds to an HSQC peak. Figure 37 shows the assigned 15N1H HSQC of the non-phosphorylated N-tail- S615. To date, I have assigned >90% of the resonances in the 15N1H HSQC to a specific N-tail- S615 resonance.

1 2 3 4 5 CβC α CβC α (CO) (CO) CCC- HNCαCβ NNH HNCαCβ NNH TOCSY

20.0 20.0

30.0 30.0 Cβ

Cγ

40.0 40.0 C

(ppm) (ppm) 50.0 50.0 Cα 60.0 60.0

70.0 70.0 8.60 8.40 8.20 8.20 8.20 8.00 8.20 1H (ppm) 1H (ppm) i-1 i-1 i i+1 i-1

Cluster33 Cluster42 Cluster6 E662 T663 E664 1 2 3 4 CβC α CβC α (CO) (CO) HNCαCβ NNH HNCαCβ NNH

20.0

30.0

40.0

C 13

(ppm) 50.0

60.0

70.0 8.60 8.40 8.60 8.40 8.60 8.40 8.20 8.00 1H (ppm) i-1 i-1 i i+1

Cluster10 Cluster33 Cluster42 A661 E662 T663

Figure 36: Resonance assignment of N-tail-S615 C-terminal residues. Clusters are linked based on 13 13 matching Cα and Cβ chemical shift between the HNCαCβ and CβCα(CO)NNH experiments. Because the CCC-TOCSY provides resonances for all alipathic 13C atoms in the i-1 position and because 13C chemical shifts are indicative of amino acid type, the CCC-TOCSY can be used to resolve ambiguities.

Using the approach described in Figure 36 and in section 1.9, over 90% of the 15N1H HSQC spectrum of N-tail-S615 has been assigned and is illustrated in Figure 37A-C. Selectively unlabeled lysine experiments were also used to assist in the assignment and is described below in 3.6.1. G645 110.0 G5 G621 G619, G2 G633 N*/Q* N*/Q* T622 T663 T632

115.0 T618 S615 S3

S627 Y647 A S636 N651 F625, Q654 R656 E662 N (ppm) N V634

120.0 E620 R646 15 D650 I639, E653 R657 M4 Y652 Q643, K, K L658R617 N640 A661 K L649 Q655 125.0 K637 W616 L623 A655 R659 H648 R641, E626 H631 E664 W616* 130.0 10.00 9.50 9.00 8.50 8.00 7.50 7.00 1 H (ppm)

G645 G5 110.0 G621

G619, G2 G633

T622 T663 T618 115.0 T632 S615 S3

S627 S636 N651 B R656 V634 N (ppm) N E662 F625, Q654 15 120.0 D650 I639, E620

R646 R657 E653 Y647 Y652 M4 Q643, K, K H648 L658 R617 W616 N640 E626, R659 H631 A655 A661 R641 L649 Q655 K, C628

125.0 K637 L623

E664 8.60 8.40 8.20 8.00 1 H (ppm) Tag TGSMG- 615 SWRTGEGTLPFESCKKHTGV C 635 QSKPINRKQPGRYHLDNYEQ 655 ARRLRPAETE Figure 37: (A) Assigned 15N1H HSQC spectrum of non-phosphorylated N-tail-S615. Amino acids with a (*) represent peaks generated from the corresponding residue’s side chain 15N1H groups. In red are the phosphorylation sites. Highlighted in green are peaks that have only the residue type assigned using a selectively unlabeling approach discussed below. (B) Expanded area of overlapped region of the assigned

N-tail-S615 (C) Primary sequence of N-tail-S615. The underlined residues could not be assigned to specific peaks in the 15N1H HSQC. In red are the phosphorylation site amino acids and in green are the residues that although could not be assigned to the 15N1H HSQC the amino acid type (i.e. Lys) was identified. Cysteine is highlighted in orange because we have data supporting the resonance assignment in 15N1H HSQC.

3.6.1 Selective unlabeling approach to assist in resonance assignment The lysine residues provided unclear 3D data, making those resonances difficult to assign. This may have been due to the extensive overlap caused by an increased flexibility amongst these residues. In order to assist in the assignment of the N-tail-S615 lysine residues, a selectively unlabeled lysine sample was prepared according to 2.1 in the Materials and Method. The only deviation was that the 15N culture was grown with 1 gram of 14N lysine 30 minutes prior to induction. With an excess of unlabeled lysine, the N-tail-S615 protein expressed was expected to contain 14N-labelled lysine and the remaining 19 amino acids as 15N-labelled. Since the unlabeled lysine is invisible in NMR, only the peaks corresponding to lysine should disappear. As shown in Figures 38A and B, an almost perfect overlay of 100% 15N N-tail-S615 and the selectively unlabeled (14N) lysine N-tail-S615 is apparent. The assigned K637 disappears in addition to possibly two other peaks located in the overlapped portion of the spectrum (circled blue in Figure 38B). Analysis of the 3D datasets also suggested that these resonances are derived from lysine residues. However, the triple resonance data is not sufficient quality to determine which resonance belongs to which lysine.

15 110.0 100% N non- phosphorylated N-tail-S615 14N-Lys 15N non- phosphorylated N-tail-S615 115.0

A 120.0

N (ppm) N 15

125.0

130.0

10.00 9.50 9.00 8.50 8.00 7.50 7.00 1H (ppm)

115.0

120.0

N (ppm) N 15 B

125.0 K637

8.60 8.40 8.20 8.00 1H (ppm) Figure 38: (A) Overlay of 15N1H HSQC spectra of 15N -labeled N-tail-S615 (black) and 14N-Lys labeled N-tail-S615 (green). The pink arrow shows that the peaks in the spectrum were folded in order to reduce the length of the 3D experiments. Therefore, these peaks that appear at the bottom of the 15N-labeled N- tail-S615 (black) spectrum are the same peaks shown at the top in the lysine unlabeled sample. (B) Circled in red is K637 that disappears and confirms its identity as a correctly assigned lysine as well as another lysine residue. The blue circle highlights the overlapped region that appears to lose two peaks – suggesting that these must be the remaining 2 lysine residues.

To assist in the assignment of the only Cys residue, iodoacetamide was reacted with a purified sample of N-tail-S615. Iodoacetamide is a cysteine alkylation reagent that can assist in peptide mapping (59). By incubating a sample of 15N N-tail-S615 with iodoacetamide, free cysteine residues can become modified into a corresponding thioether shown in Figure 39. Since our NMR buffer contains the reducing reagent TCEP, we expect that the cysteine residue is reduced, and thus free to react with iodoacetamide. With a labelled N-tail sample, we can monitor any changes to the cysteine residue by NMR. The modification on Cys will alter its chemical environment, thus causing the peak corresponding to cysteine to shift. Resonance corresponding to residues close to the Cys will likely also shift. Using this method we expect to be able to identify the HSQC peak that originates from the cysteine residue.

O O I S Cys-PEPTIDE + HS Cys-PEPTIDE H N H2N 2 Figure 39: Reaction of iodoacetamide with a peptide.

As shown in Figure 40, most peaks do not change and thus are not affected by the iodoacetamide chemical modification. These residues must be both far in space and sequence from C628. New peaks in the modified N-tail-S615 spectrum could be the modified C628 residue (circled in blue in Figure 40). The largest change occurs to the peak circled in green (Figure 40). This peak was not assigned from the previous 3D experiments. And according to the

Cα/Cβ chemical shifts, it is possible that this peak is C628. With such a large shift compared to the modified N-tail-S615 spectrum, it is likely that this peak is the cysteine residue. However, there is a discrepancy between spectra of the iodoacetamide-labelled and the selectively unlabeled lysine samples (Figure 38B). The additional missing peak (circled in red in Figure 38B) is the same peak that shifts in the iodoacetamide modified N-tail-S615 spectrum (Figure 40). Thus this peak may be a lysine residue or C628. It is also possible that since the unassigned lysine residues are close in sequence to C628 (i.e. K629, K630), their chemical environment could be altered by the modified cysteine. Alternatively, it is possible that there is overlap in this region of the spectrum and that the HSQC peak corresponds to two residues. Additional trials of the iodoacetamide reaction or N-tail-S615 proteins with mutated C628 may assist in the assignment of the C628 HSQC peak.

15 110.0 N non-phosphorylated N-tail-S615 15 Cys modified N non- phosphorylated N-tail-S615

115.0

N (ppm) N

15 120.0

125.0

130.0 10.00 9.50 9.00 8.50 8.00 7.50 7.00 1 H (ppm) Figure 40: In black is the N-tail-S615 15N1H HSQC in black overlayed with the iodoacetamide alkylated N-tail-S615 in orange. Circled in blue are new peaks that appear and circled in green is the largest shift which may signify that this peak corresponds to C628.

3.6.2 Analysis of assigned N-tail-S615 Chemical shifts of specific nuclei can be used to determine secondary structure from NMR (47, 69). That is to say, deviations from expected resonance in turn describes the secondary structure of the residue. Using their 15N1H and 15N chemical shifts, NMRView’s CSI (chemical shift index) function can be used to elucidate secondary structure as shown in Figure 42 (47-49, 69). The CSI tool confirms that the N-tail-S615 has some α-helical structure near the C-terminus, as predicted by AGADIR in 3.1 Figure 11. We can use assigned chemical shifts to confirm structure because different secondary structural elements position the peptide bond in different orientations with respect to one another. Thus, backbone and side chain Cβ atoms are in different electronic environments. Therefore, resonances of these nuclei have chemical shifts that differ from the expected if the residues were in a random coil conformation. And thus, chemical shifts can be used to determine the secondary structure of a protein with residue-level resolution. These data are consistent with our CD/TFE experiments.

Tag TGSMG- 615 SWRTGEGTLPFESCKKHTGV 635 QSKPINRKQPGRYHLDNYEQ 655 ARRLRPAETE Figure 41: Chemical Shift Index (CSI) calculation using NMRView (47, 69). In orange are the residues that are predicted to be in an extended β conformation, in black have no secondary structural propensity and in blue are the residues that are in a helix based the secondary shifts that were calculated based on its 1H and 15N chemical shifts. The underlined residues distinguish the residues that AGADIR predicted to have helical propensities. CSI confirms the AGADIR prediction.

Resonance assignment of N-tail-S615 also suggests that while phosphorylation causes some chemical shifts, it appears to have the greatest effect at the N-terminus of the N-tail-S615 (Figure 42), specifically at residues W616 to G621, and the few amino acids from the tag. Because the phosphorylation sites, S632 and T636 are 16 to 20 residues away, these spectral change confirm our hypothesis that there must be some residual structure in the N-tail bringing the phosphorylation sites closer in space to the N-terminal residues. The C-terminal residues also appear to be affected by phosphorylation, but the effect is not as great as that observed for the N- terminal residues. G5 110.0 G621

G619, Non-phosphorylated N-tail -S615G2 Phosphorylated N-tail-S615 G633

T632 115.0 T618 S615 S627

S636

120.0 E620 (ppm) N

15 R617 M4 W616 125.0

W*616 130.0 10.00 9.50 9.00 8.50 8.00 7.50 7.00 1 H (ppm) Figure 42: Overlay of 15N1H HSQC spectra of non-phosphorylated (black) and phosphorylated (red) N- tail-S615. Residues that display the largest changes in chemical shift are labeled. It appears that the most significant changes caused by phosphorylation occur at the N-terminus of N-tail-S615. This suggests that although the phosphorylation sites are far away, residual structure present in N-tail-S615 must bring these sites closer in space to the N-terminus.

Since the N terminus appears to be affected by phosphorylation where the C-terminus exhibits residual helical secondary structure, we wanted to determine if these factors could play a 55 role in the N-tail’s interaction with NBD1. We probed these changes using NMR spectroscopy and map interaction sites to specific N-tail resonances using resonance assignments of N-tail- S615. We compared the changes in chemical shift, peak intensity and volume from the interaction experiments described in Figures 23 and 24 to localize the N-tail-S615 residues involved in an NBD1 interaction (Figure 43).

0.0100 Non-phosphorylated N-tail-S615 Phosphorylated N-tail-S615

0.0050

0.0000

P P P P

G2 G5

Shift (ppm) Shift

I639

S615 F625 S627 S636

T618 E620 T622 L623 E626 T632 L649 E653 L658 E662 T663 E664

R617 C628 R641 R646 R656 R657 R659

G619 G621 K629 K630 H631 G633 V634 Q635 K637 N640 K642 Q643 G645 Y647 H648 D650 N651 Y652 Q654 A655 A661 W616

-0.0050 W616*

Change inChemical Change Residue

-0.0100

100.00 Non-phosphorylated N-tail-S615 Phosphorylated N-tail-S615 50.00

0.00

(%)

P P P P

G2 G5

I639

S615 F625 S627 S636

T618 E620 T622 L623 E626 T632 L649 E653 L658 E662 T663 E664

R617 C628 R641 R646 R656 R657 R659

G619 G621 K629 K630 H631 G633 V634 Q635 K637 N640 K642 Q643 G645 Y647 H648 D650 N651 Y652 Q654 A655 A661 -50.00 W616

W616* Residue Change inIntensity Change -100.00

100.00 Non-phosphorylated N-tail-S615 Phosphorylated N-tail-S615 50.00

0.00

(%)

P P P P

G2 G5

I639

S615 F625 S627 S636

T618 E620 T622 L623 E626 T632 L649 E653 L658 E662 T663 E664

R617 C628 R641 R646 R656 R657 R659

G619 G621 K629 K630 H631 G633 V634 Q635 K637 N640 K642 Q643 G645 Y647 H648 D650 N651 Y652 Q654 A655 A661

-50.00 W616 W616*

Change inVolume Change Residue -100.00

Figure 43: Comparison of interaction experiments by chemical shift changes and changes in peak volume and peak intensity. It appears that there are less changes with phosphorylation and this suggests that the phosphorylated N-tail-S615 has a weaker interaction with NBD1 compared to its non-phosphorylated form.

Although the changes are small, it is apparent that there are more changes in chemical shift, intensity and volume when non-phosphorylated N-tail-S615 are in the presence of NBD1. These changes are even less apparent in the phosphorylated N-tail-S615 interaction experiments. This indicates that phosphorylation reduces the N-tail/NBD1 interaction and thus supports the role of post-translational modifications to regulate nucleotide binding at the NBDs (10). However, the changes range across the primary sequence of N-tail-S615 making it difficult to probe the NBD1 interaction on N-tail. In addition, since these changes are so small some may be insignificant contributing to the ambiguity in identifying which N-tail residues are involved in the interaction. Improving the interaction experiments with the use of TFE as described above, may allow us to confidently probe and explain the changes in NBD1 binding to N-tail. Figure 44 schematically describes the role phosphorylation plays in the regulation of NBD1 according to our studies on N-tail. Non-phosphorylated N-tail is more compact than its phosphorylated partner. However, it is predominantly in a transient association with NBD1. With phosphorylation (by PKA), N-tail becomes slightly more extended, with reduced (but not eliminated) structural tendencies. The altered structure in phosphorylated N-tail also reduces its interaction with NBD1 which again is not completely eliminated. The results presented in this thesis provide an extensive study on the structure of the N-terminal tail of NBD1 and how the structure of N-tail changes with phosphorylation. These data can further our understanding of the role of phosphorylation in altering N-tail’s interaction and regulation of NBD1 and ultimately

KATP channels.

non-phosphorylated NBD1 NBD1 N-tail-S615

PKA +2 MgATP -2 ADP P phosphorylated N-tail-S615 P NBD1 NBD1

P P

Figure 44: Schematic of N-tail-S615 and NBD1 interaction with phosphorylation. Non-phosphorylated N- tail-S615 is more compact and interacts with NBD1 slightly more than its phosphorylated partner. Thus, there is a weak interaction between N-tail-S615 and NBD1 that is reduced, but not eliminated with phosphorylation. This change may be attributed to reduced helical structure in phosphorylated N-tail.

4. Discussion

This thesis presented studies of the N-terminal tail of SUR2A NBD1, an intrinsically, disordered phospho-regulatory region. Intrinsically disordered proteins (IDPs) are prevalent throughout eukaryotic regulatory signaling pathways (70). Although they are not well understood, IDPs are excellent signal transduction candidates because they can control a large degree of protein motion between domains, they can disguise binding sites, they are easily accessible to post-translational modifications and can transiently interact with multiple complexes due to their flexibility (70). Through the structural characterization of disordered proteins we can improve our understanding of these key regulators in healthy systems and thus their contributing roles in diseases. With the use of a variety of techniques we have shown that although the non- phosphorylated N-tail is predominantly disordered it does have some residual compact structure that is reduced with phosphorylation. Further, phosphorylation of the N-tail decreases its interactions with the core NBD1 (10). From our three N-tail constructs we have demonstrated that the non-phosphorylated N-tail possesses helical propensity resulting in a compact, but still disordered protein. The presence of helical structure in the N-tail is predicted to be at the C- terminal residues of N-tail by AGADIR and was confirmed by secondary chemical shifts of non- phosphorylated N-tail-S615. In its non-phosphorylated form, N-tail-S615 transiently and weakly binds to the core NBD1, similar to a common trend that the compact conformers of disordered proteins serve as binding partners to structured proteins and mediate important subsequent protein interactions (71). The compact structure and interactions of N-tail are important for function, as mutations of N-tail residues (R663C and A665T in human SUR2A) cause early repolarization syndromes (ERS) (72). ERS is associated with life threatening arrhythmias (72). Other intrinsically disordered proteins that participate in the regulation of structured proteins are also known to exhibit transient structural propensities that are also controlled by phosphorylation post-translational modifications. For example, the cyclin-dependent kinase (CDK) inhibitor Sic1 is known to be compact, containing some transient helical structure like the N-tail (70). These transient helices are necessary for Sic1’s interaction with its receptor Cdc4, a process involved in yeast cell cycle progression (71, 73). Like the N-tail, the IDP Sic1 is known to participate in the regulation of structured proteins and similarly is controlled by phosphorylation, which alters its compact structure (70). Phosphorylation of Sic1 instead

59 enhances folding facilitating Sic1’s interaction with the Cdc4 receptor, a common trend that phosphorylation alters protein structure (71). The cystic fibrosis transmembrane conductance regulator (CFTR) contains an intrinsically disordered regulatory region (R region). R region which is located between NBD1and MSD2, interacts with the core NBD1 via multiple transient helices (76). Phosphorylation reduces the compact structure of R region resulting in the loss of helical structure, and diminishes interaction with NBD1 (76). This also suggests the role of compact conformers and phosphorylation in IDP regulation. The data on CFTR, Sic1 and N-tail demonstrate that disordered proteins can have some structural tendencies that are necessary to facilitate an interaction with structured binding partners. Post-translational modifications like phosphorylation are employed to change the net charge of the IDP, thereby altering protein structure and subsequent interactions that in turn regulate complex protein pathways (71, 73). Protein phosphorylation is one of the most common regulatory mechanisms utilized by the cell (74). ABC transporters are multimeric protein complexes that move a variety of substrates across cell membranes (4). ABC protein function is regulated by phosphorylation and mutations can alter their regulatory mechanisms contributing to many human diseases (4). In particular, the SUR proteins are unlike most ABC proteins in that they do not possess transporter activity, but instead regulate the gating of KATP channels (10). Activity of the SUR proteins are controlled by the NBDs where MgATP binding at NBDs promotes channel opening and K+ conductance (10). The action of the SUR NBDs, in particular the activity of NBD1, is regulated by the phosphorylation of its intrinsically disordered N-terminal extension (N-tail). Phosphorylation at the N-terminus is known to enhance channel activity (75). Phosphorylation of N-tail enables ATP binding to NBD1, enhancing the formation of the NBD1/NBD2 complex which is required for KATP channel opening (10). We have shown that the significant structural changes associated with phosphorylating the disordered N-tail can alter its interaction with NBD1, and consequently promote ATP binding at the NBDs (10). The CFTR chloride channel is mutated in patients with cystic fibrosis and also has an intrinsically disordered regulatory (R) region that must be phosphorylated at multiple sites to activate the CFTR channel (76). Like the N-tail of SUR2A NBD1, the R region exhibits reduced helical propensity upon phosphorylation, which explains its reduced interactions with its respective NBD1 core (76). The reduced interaction with NBD1 upon phosphorylation allows for

60 dimerization with its binding partner NBD2 leading to channel activation (76). Mutations that affect phosphorylation will affect NBD dimerization compromising chloride channel regulation in cystic fibrosis patients. This emphasizes the importance of understanding phosphorylation as a regulatory mechanism employed by ABC proteins. Another ABC protein, Ste6 is involved in the secretion of mating pheromones in Saccharomyces cerevisiae yeast cells (77). Ubiquitination of the ABC-transporter Ste6 is known to target Ste6 for degradation in the yeast vacuole and suggests that ubiquitination of other cell surface proteins is a trigger for subsequent degradation in the yeast vacuole (77). Like SUR2A and CFTR, phosphorylation regulates the activity of this Ste6 transporter by controlling the ubiquitin attachment site (78). Ubiquitination is mediated by a linker region known as the degradation-box (D-box) and when removed, Ste6 is no longer ubiquinated. When part of the D- box is deleted (A-box), phosphorylation is completely inhibited resulting in no ubiquitination causing an accumulation of Ste6 in the cell membrane (78). Removal of the phosphorylation mechanism again illustrates the inability to control the yeast ABC protein Ste6. Rv1747 is an ABC exporter required for the growth of Mycobacterium tuberculosis (79, 80). Rv1747 has two forkhead-associated (FHA) domains that are phosphorylated at two threonine residues (80). Mutating these residues to prevent phosphorylation revealed that phosphorylation positively regulates Rv1747 function (80). Phosphorylation of Rv1747 is required for a fully functioning transporter activity (80). It is possible that phosphorylation causes conformational changes in the FHA domains to enable protein-protein interactions (80). Since phosphorylation activates Rv1747, it suggests that phosphorylation may also play a role in regulating the growth of M. tuberculosis (80). With an understanding of the phosphorylation- dependent regulatory mechanism of Rv1747 or other ABC proteins, phosphorylation could serve at a target for tuberculosis (TB) therapeutics. Although the role of phosphorylation in other ABC proteins is not completely understood, our data on phosphorylation in SUR2A may provide insight into understanding phosphorylation as a regulatory mechanism in other related proteins such as ABCB1 that is a multidrug resistant transporter overexpressed in cancer cells, ABCC2 responsible for Dubin-Johnson Syndrome, ABCC8 responsible for hyperinsuliemic hypoglycemia, among many others (81). Understanding the role of phosphorylation regulation and structural changes associated with this post- translational modification could be a turning point in the development of ABC related human disease therapeutics.

5. References

1. Altenberg, G.A. (2003) The engine of ABC proteins. News Physiol Sci 18, 191-195.

2. Burke, M. A., Mutharasan, R. K., and Ardehali, H. (2008) The Sulfonylurea Receptor, an Atypical ATP-Binding Cassette Protein and its Regulation of the KATP Channel. Circulation Research 103, 1458-1465.

3. Rees, D.C., Johnson, E., and Lewinson, O. (2009) ABC transporters: the power to change. Nat Rev Mol Cell Biol 10, 218-27.

4. Stolarczyk, E. I., Reiling, C. J., and Paumi, C. M. (2011) Regulation of ABC Transporter Function Via Phosphorylation by Protein Kinases. Curr Pharm Biotechnol 12, 621-635.

5. Zhoa, Q. F., Yu, J. T. and Tan, M. S. (2015) ABCA7 in Alzheimer’s Disease. Molecular Neurobiology 51, 1008-1016.

6. Joyce, H., McCann, A., Clynes, M., and Larkin, A. (2015) Influence of multidrug resistance and drug transport proteins on chemotherapy drug metabolism. Expert Opinion on Drug Metabolism & Toxicology 11, 795-809.

7. Miller, D. S. (2015) Regulation of ABC Transporters at the Blood-Brain Barrier. Clinical Pharmacology & Therapeutics 97, 395-403.

8. de Araujo, E. D., and Kanelis, V. (2014) Successful development and use of a thermodynamic stability screen for optimizing the yield of nucleotide binding domains. Protein Expression and Purification 103, 38-47.

9. Dean, M., Rzhetsky, A., and Allikmets, R. (2001) The human ATP-binding cassette (ABC) transporter superfamily. Genome Res 11, 1156-1166.

10. De Araujo, E. D., Alvarez, C. P., Lopez-Alonso, J. P., Sooklal, C. R., Staglar, M., and Kanelis, V. (2015) Phosphorylation-dependent changes in protein dynamics and nucleotide binding of SUR2A NBD1. Journal of Biological Chemistry. 11. Rosenberg, M. F., ORyan, L. P., Hughes, G., Zhao, Z., Aleksandrov, L. A., Riordan, J. R., and Ford, R. C. (2011) The Cystic Fibrosis Transmembrane Conductance Regulatory (Cftr):3D Structure and Localisation of a Channel Gate. Journal of Biological Chemistry 286, 42647.

12. Dawson, R.J.P. and K.P. Locher (2007). Structure of the multidrug ABC transporter Sav1866 from Staphylococcus aureus in complex with AMP-PNP. Febs Letters 581, 935-938.

13. Dawson, R.J.P. and K.P. Locher (2006) Structure of a bacterial multidrug ABC transporter. Nature 443, 180-185.

14. Dean, M., Hamon, Y., and Chimini, G. (2001) The human ATP-binding cassette (ABC) transporter superfamily. Journal of Lipid Research 42, 1007-1017.

15. Roundhill, E. A., Fletcher, J. I., Haber, M., and Norris, M. D. Clinical Relevance of Multidrug-Resistance Proteins (MRPs) for Anticancer Drug Resistance and Prognosis Resistance to Targeted ABC Transporters in Cancer. Springer International, 2015.

16. Kemp, S., Pujol, A., Waterham, H. R., van Geel, B. M., Raymond, G. V., Cutting, G. R., Wanders, R. J., and Moser, H. W. (2001) ABCD1 mutations and the X-linked adrenoleukodystrophy mutation database: role in diagnosis and clinical correlations. Hum Mutat 18, 499-515.

17. Sturm, A., Cunningham, P., and Dean, M. (2009) The ABC transporter gene family of Daphniapulex. BMC Genomics 10, 170.

18. Deeley, R.G., Westlake, C., and Cole, S. P. C. (2006) Transmembrane Transport of Endo- and Xenobiotics by Mammalian ATP-Binding Cassette Multidrug Resistance Proteins. Physiological Reviews 86, 849-899.

19. Shi, NQ., Ye, B. and Makielski, JC. (2005) Function and distribution of the SUR isoforms and splice variants. J Mol Cell Cardiol 39, 51-61.

20. Seino, S., and Miki, T. (2003) Physiological and pathophysiological roles of ATP-sensitive K+ channels. Prog Biophys Mol Biol 81, 133-176.

21. Mikhailov, M. V., Campbell, J. D., de Wet, H., Shimomura, K., Zadek, B., Collins, R. F., Sansom, M. S. P., Ford, R. C., and Ashcroft, F. M. (2005) 3-D structural and functional characterization of the purified KATP channel complex Kir6.2-SUR1. EMBO Journal 24, 4166- 4175.

22. Zingman, L.V., Alekseev, A.E., Bienengraeber, M., Hodgson, D., Karger, A.B., Dzeja, P.P., and Terzic, A. (2001) Signalling in channel/enzyme multimers: ATPase transitions in SUR module gate ATP-sensitive K+ conductance. Neuron 31, 233-245.

23. Tucker, S. J., Gribble, F. M., Zhao, C., Trapp, S., and Ashcroft, F. M. (1998) Molecular determinants of KATP channel inhibition by ATP. EMBO Journal 17, 3290-3296.

24. Shi, Y., Chen, X., Wu, Z., Shi, W., Yang, Y., Cui, N., Jiang, C., and Harrison, R.W. (2008) J. Biol. Chem 283, 7523-7530.

25. Quinn, K.V.; Giblin, J.P.; and Tinker, A. (2004) Circ Res 94, 1359-1366.

26. Pasyk, E. A., Kang, Y., Huang, X., Cui, N., Sheu, L., and Gaisano, H. Y. (2004) Syntaxin-1A Binds the Nucleotide-binding Folds of Sulphonylurea Receptor 1 to Regulate the KATP Channel. Journal of Biological Chemistry 279, 4234-4240.

27. Narayanan, A., and Jacobson, M. P. (2009) Computational studies of protein regulation by post-translational phosphorylation. Current Opinions in Structural Biology 19, 156-163.

28. Nichols, C. G. (2006) KATP channels as molecular sensors of cellular metabolism. Nature 440, 470-476.

29. Dunker, A. K., and Obradovic, Z. (2001) Intrinsically Disordered Proteins. Journal of Molecular Graphics and Modelling 19, 26-59.

30. Habchi, J., Tompa, P., Longhi, S., and Uverskey, V. N. (2014) Chemical Reviews: Introducing Protein Intrinsic Disorder.

31. Oldfield, C. J., and Dunker, A. K. (2014) Intrinsically Disordered Proteins and Intrinsically Disordered Protein Regions: Annu. Rev. Biochem 83, 553-84.

32. Dogan, J., Gianni, S., and Jemth, P. (2014) The binding mechanisms of intrinsically disordered proteins. Phys. Chem 16, 6323.

33. de Araujo, E. D., Ikeda, L. K., Tzvetkova, S., and Kanelis, V. (2011) The First Nucleotide Binding Domain of the Sulfonylurea Receptor 2A Contains Regulatory Elements and is Folded and Functions as an Independent Module. Biochemistry 50, 6655-6666.

34. Quinn, K. V., Giblin, J. P., and Tinker, A. (2004) Multisite Phosphorylation Mechanism for Protein Kinase A Activation of the Smooth Muscle ATP-Sensitive K+ Channel. Circulation Research 94, 1359-1366.

35. Kanelis, V., Chong, P. A., Forman-Kay, J. D. (2011) NMR Spectroscopy to Study the Dynamics and Interactions of CFTR. Methods in Molecular Biology 741, 377-403.

36. Kanelis, V., Forman-Kay, J. D., and Kay, L. E. (2001) Multidimensional NMR Methods for Protein Structure Determination. IUBMB Life 52, 291–302.

37. Cavanagh, J., Fairbrother, W. J., Palmer III, A. G., Rance, M., and Skelton, N. J. Protein NMR Spectroscopy Principles and Practice. Elsevier Academic Press, 2007.

38. Kitevski-LeBlanc, J. L., Evanics, F., and Prosser, S. (2010) Optimizing 19F NMR protein spectroscopy by fractional biosynthetic labeling. Journal of Biomolecular NMR 48, 113-121.

39. Rule, G.S.; Hitchens, T.K. Fundamentals of Protein NMR Spectroscopy. Netherlands: Springer, 2006.

40. Keeler, J. Understanding NMR Spectroscopy. John Wiley & Sons, 2010.

41. Corrêa, D. H. A., and Ramos, C. H. I. (2009) The use if circular dichroism spectroscopy to study protein folding, form and function. African Journal of Biochemistry Research 3, 164-173.

42. Fandrich, M., and Dobson, C. M. (2002) The behavior of polyamino acids reveals an inverse side chain effect in amyloid structure formation. The EMBO Journal 21, 5682-5690.

43. Luo, P., and Baldwin, R. L. (1997) Mechanism of Helix Induction by Trifluoroethanol: A Framework of Extrapolating the Helix-Forming Properties of Peptides from Trifluoroethanol/Water Mixtures Back to Water. Biochemistry 36, 8413-8421.

44. (Malvern Instruments) Dynamic Light Scattering: An Introduction in 30 minutes.

45. Shaw, R. Dynamic Light Scattering Training, Achieving reliable nanoparticle sizing. Malvern instruction manual. ATA Scientific.

46. Pace, C. N., Vajdos, F., Fee, L., Grimsley, G., and Gray, T. (1995) How to measure and predict the molar absorption coefficient of a protein. Protein Science 4, 2411-2423.

47. Wishart, D. S., Bigam, C. G., Yao, J., Abildgaard, F., Dyson, H. J., Oldfield, E., Markley, J. L., and Sykes, B. D. (1995) 1H, 13C and 15N chemical shift referencing in biomolecular NMR. J. Biomol. NMR 6, 135-140.

48. Delaglio, F., Grzesiek, S., Vuister, G. W., Zhu, G., and Pfeifer, J. (1995) AD Bax. NMRPipe: a multidimensional spectral processing system based on UNIX pipes. Journal of Biomolecular NMR 6, 277-293.

49. Johnson, B. A., and Blevins, R. A. (1994) NMRView: a computer program for the visualization and analysis of NMR data. J. Biomol. NMR 4, 603-614.

50. Sattler, M., Schleucher, J., and Griesinger, C. (1999) Heteronuclear multidimensional NMR experiments for the structure determination of proteins in solution employing pulsed field gradients. Progress in Nuclear Magnetic Resonance Spectroscopy 34, 93-158.

51. Li X., Romero P., Rani M., Dunker A. K., and Obradovic Z. (1999) Predicting Protein Disorder for N-, C-, and Internal Regions. Genome Inform. Ser. Workshop 10, 30-40.

52. Romero P., Obradovic Z., and Dunker A. K. (1997) Sequence Data Analysis for Long Disordered Regions Prediction in the Calcineurin Family. Genome Inform. Ser. Workshop 8, 110- 124.

53. Romero P., Obradovic Z., Li X., Garner E. C., Brown C. J., and Dunker A. K. (2001) Sequence complexity of disordered protein. Proteins 42, 38-48.

54. Malakhov, M. P., Mattern, M. R., Malakhova, O. A., Drinker, M., Weeks, S. D. and Butt, T. R. (2004) SUMO fusions and SUMO-specific protease for efficient expression and purification of proteins. Journal of Structural and Functional Genomics 5, 75-86.

55. Kjaergaad, M., Brander, S., and Poulsen, F. M. (2011) Random coil chemical shift for intrinsically disordered proteins: effects of temperature and pH. Journal of Biomolecular NMR 49, 139-149.

56. Baker, K. A., Tzitzilonis, C., Kwiatkowski, W., Choe, S., and Riek, R. (2007) Conformational dynamics of the KcsA potassium channel governs gating properties. Nature Structural & Molecular Biology 14, 1089-1095.

57. Mittag, T., Orlicky, S., Choy, W. Y., Tang, X., Lin, H., Sicheri, F., Kay, L. E., Tyers, M., and Forman-Kay, J. D. (2006) Dynamic equilibrium engagement of polyvalent ligand with a single- site receptor. PNAS 105, 17772-17777.

58. López, J. P., de Araujo, E. D., Kanelis, V. (2012) NMR and fluorescence studies of drug binding to the first nucleotide binding domain of SUR2A. Biochemistry 45, 9211-9222.

59. Sechi, S. and Chait, B. T. (1998) Modification of cysteine residues by alkylation. A tool in peptide mapping and protein identification. Analytical Chemistry 70, 5150-5158.

60. Cuthbertson J. M., Doyle D. A., and Sansom M. S. (2005) Transmembrane helix prediction: a comparative evaluation and analysis. Protein Eng. Des. Sel. 18, 295-308.

61. Hudson R. P., Chong P. A., Prostasevich I. I., Vernon R., Noy E., Bihler H., An J. L., Kaild O., Sela-Culang I., Mense M., Senderowitz H., Brouillette C. G., and Forman-Kay J. D. (2012) Conformational Changes Relevant to Channel Activity and Folding within the first Nucleotide Binding Domain of CFTR. J. Biol. Chem. 287, 28480-28494.

62. Aller S. G., Yu J., Ward A., Weng Y., Chittaboina S., Zhuo R., Harrell P. M., Trinh Y. T., Zhang Q., Urbatsch I. L., and Chang G. (2009) Structure of P-glycoprotein reveals a molecular basis for poly-specific drug binding. Science 323, 1718-1722.

63. Dawson R. J., and Locher K. P. (2006) Structure of a bacterial multidrug ABC transporter. Nature 443, 180-185.

64. Hohl M., Briand C., Grutter M. G., and Seeger M. A. (2012) Crystal structure of a heterodimeric ABC transporter in its inwardfacing conformation. Nat. Struct. Mol. Biol. 19, 395- 402.

65. Hollenstein K., Dawson R. J., and Locher K. P. (2007) Structure and mechanism of ABC transporter proteins. Curr. Opin. Struct. Biol. 17, 412-418.

66. Jin M. S., Oldham M. L., Zhang Q., and Chen J. (2012) Crystal structure of the multidrug transporter P-glycoprotein from Caenorhabditis elegans. Nature 490, 566- 569.

67. Shintre C. A., Pike A. C., Li Q., Kim J. I., Barr A. J., Goubin S., Shrestha L., Yang J., Berridge G., Ross J., Stansfeld P. J., Sansom M. S., Edwards A. M., Bountra C., Marsden B. D., von Delft F., Bullock A. N., Gileadi O., Burgess-Brown N. A., and Carpenter E. P. (2013) Structures of ABCB10, a human ATP-binding cassette transporter in apo- and nucleotide-bound states. Proc. Natl. Acad. Sci. U. S. A. 110, 9710-9715.

68. Ward A., Reyes C. L., Yu J., Roth C. B., and Chang G. (2007) Flexibility in the ABC transporter MsbA: Alternating access with a twist. Proc. Natl. Acad. Sci. U. S. A. 104, 19005- 19010.

69. Wishart, D.S., Sykes, B.D., and Richards, F.M. (1991) Relationship between nuclear magnetic resonance chemical shift and protein secondary structure. Journal of Molecular Biology 222, 311–333.

70. Mittag, T., Kay, L. E., and Forman-Kay, J. D. (2010) Protein dynamics and conformational disorder in molecular recognition. Journal of Molecular Recognition 23, 105-116.

71. Mittag T., Orlicky, S., Choy, W. Y., Tang, X., Lin, H., Sicheri, F., Tyers, M and Forman-Kay, J. D. (2008) Dynamic equilibrium engagement of a polyvalent ligand with a single-site receptor. Proc Natl Acad Sci 105, 17772-17777.

72. Hu, D., Barajas-Martínez, H., Terzic, A., Park, S., Pfeiffer, R., Burashnikov, E., Wu, Y., Borggrefe, M., Veltmann, C., Schimpf, R., Cai, J. J., Nam, G., Deshmukh, P., Scheinman, M., Preminger, M., Steinberg, J., Lópex-Izquierdo, A., Ponce-Balbuena, D., Wolpert, C., Haïssaguerre, M., Sánchez-Chapula, J. A. and Antzelevitch, C. (2014) ABCC9 is a novel Brugada and early repolarization syndrome susceptibility gene. International Journal of Cardiology 171, 431-442.

73. Mittag, T., Marsh, J., Grishaev, A., Orlicky, S., Lin, H., Sicheri, F., Tyers, M., and Forman- Kay, J. D. (2010) Structure/function implications in a dynamic complex of the intrinsically disordered Sic1 with the Cdc4 subunit of an SCF ubiquitin ligase Structure 18, 494-506.

74. Philips, C. (2001) The role of protein phosphorylation in human health and disease. Eur. J. Biochem 268, 5001-5010.

75. Shi, Y., Chen, X., Wu, Z., Shi, W., Yang, Y., Cui, N., Jiang, C., and Harrison, R. W. (2008) cAMP-dependent Protein Kinase Phosphorylation Produces Interdomain Movement in SUR2B Leading to Activation of the Vascular KATP Channel. Journal of Biological Chemistry 283, 7523-7530.

76. Baker, J. M. R., Hudson, R. P., Kanelis, V., Choy, W. Y., Thibodeau, P. H., Thomas, P. J., and Forman-Kay, J. D. (2007) CFTR regulatory region interacts with NDB1 predominantly via multiple transient helices. Nature Structural and Molecular Biology 14, 738-745.

77. Losko, S., Kopp, F., Kranz, A., and Kolling, R. (2001) Uptake of the ATP-Binding Cassette (ABC) Transporter Ste6 into the Yeast Vacuole Is Blocked in the doa4 Mutant. Molecular Biology of the Cell 12, 1047–1059.

78. Kolling, R. (2002) Mutations a¡ecting phosphorylation, ubiquitination and turnover of the ABC-transporter Ste6. FEBS Letters 531, 548-552.

79. Spivey, V. L., Whalan, R. H., Hirst, E. M. A., Smerdon, S. J., and Buxton, R. S. (2013) An attenuated mutant of the Rv1747 ATP-binding cassette transporter of Mycobacterium tuberculosis and a mutant of its cognate kinase, PknF, show increased expression of the efflux pump-related iniBAC operon. FEMS Microbial Lett. 347, 107-115.

80. Spivey, V. L., Molle, V., Whalan, R. H., Rodgers, A., Leiba, J., Stach, L., Walker, K. B., Smerdon, S. J., and Buxton, R. S. (2001) Forkhead-associated (FHA) Domain Containing ABC Transporter Rv1747 Is Positively Regulated by Ser/Thr Phosphorylation in Mycobacterium tuberculosis. Journal of Biological Chemistry 286, 26198 –26209.

81. Luo, P., and Baldwin, R. L. (1997) Mechanism of Helix Induction by Trifluoroethanol: A Framework of Extrapolating the Helix-Forming Properties of Peptides from Trifluoroethanol/Water Mixtures Back to Water. Biochemistry 36, 8413-8421.

Figure 1 (10): JBC automatically allows permission for authors to use this article in a thesis and/or dissertation. I am co-author in this article.

Figure 2 (28): – Order License Id: 3705430445888

Figure 3, 4 (38): – Order License Id: 3682080358484

Figure 5 (41): – Order License Id: 3682000249292

Figure 14 (10): JBC automatically allows permission for authors to use this article in a thesis and/or dissertation. I am co-author in this article.