The Tansmembrane Associations of HIV-1 Vpu and Its Human Targets

The transmembrane associations of HIV-1 Vpu and its human targets

Gregory Cole

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Department of Biochemistry University of Toronto

Gregory Cole

Doctor of Philosophy

Department of Biochemistry University of Toronto

2019 Abstract

Many viruses evade the innate immune response through the expression of accessory proteins that modulate the activity of immune system proteins. The HIV-1 protein Vpu functions to downregulate a number of host cell surface proteins within infected cells. For several target proteins, including the NTB-A and PVR receptors and the restriction factor tetherin, downregulation occurs through direct interaction of the proteins’ transmembrane

(TM) domains. The mechanism through which a single TM domain can interact with a wide range of targets that share no obvious interaction motifs is poorly understood. In this thesis we characterize the TM domain homo and heterooligomerization of Vpu and its human targets, tetherin, NTB-A and PVR. By using a series of fluorophore labeled TM domain peptides we used

Förster resonance energy transfer (FRET) to characterize the interactions of Vpu with itself as well as its human targets. Our data show target TM domains compete for interaction with Vpu, suggesting that all targets bind the same helical face, also that formation of each heterooligomer has a similar free energy of association to the Vpu homooligomer. This leads to a model in which Vpu monomers, Vpu homooligomers and Vpu-target heterooligomers coexist

and suggests that the binding surface of the Vpu TM domain has been selected for weak binding to multiple targets. We also document the role of the lipid bilayer, specifically lipid phase and hydrophobic thickness, on TM domain assembly within the Vpu-target system. We show that thicker membranes and gel phase lipids non-specifically promote associations of Vpu and its targets without affecting secondary structure. The properties of the lipid bilayer are an important factor in Vpu-target oligomerization and may explain differences in Vpu-target oligomerization in different subcellular compartments. This work provides new insight into the molecular mechanism of Vpu and furthers our knowledge of membrane protein assembly

iii

Acknowledgments

This work was possible thanks to guidance and support from many people.

I’ll begin by thanking my supervisor, Dr. Simon Sharpe, who graciously gave me the opportunity to work in the lab. I’m very grateful for his guidance and mentorship, for letting me have free reign over my projects, and for encouraging me to have a wide range of experiences.

I would also like to thank my committee members, Dr. Charles Deber and Dr. Jeffrey Lee whose thoughtful advice and careful critique of my work has helped me to become a better scientist. Thanks for teaching me that graduate school is more than just research and time in the lab, and also for making me feel less intimidated about answering the tough questions and striving for big goals.

A big thanks has to go out to the past and present members of the Sharpe lab. To Karen, for taking care of all of us. To Pat, Dave, and Jason, because work is easier when done with your good friends. Sean, Lisa, Aditi and Melody, thanks for all the scientific discussions (and for putting up with all my clutter).

Thank you to the other support groups in my life, my Waterloo friends, the volleyball team, and the MSF coffee club. As well as my colleagues in the Biochemistry Department and the Molecular Medicine Program – thank you.

On a more personal note, I would like to thank my parents for all their support and their unwavering faith in me. And lastly I would like to thank the biggest influence in my life during my PhD, Tracy Stone – my lab partner, my partner in life, my smizmar – thank you for everything.

Table of Contents

Acknowledgments...... iv

Table of Contents ...... v

List of Tables ...... ix

List of Figures ...... x

List of Abbreviations ...... xii

Chapter 1: Introduction ...... 1

1.1 Membrane Protein Folding and Oligomerization ...... 2

1.1.1 The Three Stage Model of Membrane Protein Folding ...... 2

1.1.2 Studying TM Domains in Isolation ...... 6

1.1.3 Forces Driving the Oligomerization of Membrane Proteins ...... 6

1.1.4 Helix-Helix Interaction Motifs ...... 8

1.1.5 Factors Affecting Helix-Helix Interaction Strength ...... 9

1.1.6 The Role of Helix-Helix Interactions in Innate Immunity ...... 9

1.2 Viral Immune Evasion ...... 10

1.2.1 The HIV-1 Pandemic ...... 12

1.2.2 HIV-1 Pathogenesis and Disease ...... 12

1.2.3 HIV-1 Genome and Proteome ...... 13

1.2.4 HIV-1 Virus Particle Structure ...... 15

1.3 HIV-1 Life Cycle ...... 16

1.4 HIV-1 Disease Pathology, Treatment and Prevention ...... 19

1.5 HIV-1 Accessory Proteins and Mechanisms of HIV-1 Immune Evasion ...... 19

1.5.1 HIV-1 Vpu ...... 21

1.5.2 Vpu Structure and Function ...... 21

1.5.3 Vpu Mediated Antagonism of Cell Surface Proteins ...... 23

1.5.4 Importance of HIV-1 Vpu to the Group M Pandemic ...... 25

1.6 Host Cell Proteins Directly Targeted by Vpu ...... 28

1.6.1 Non-TM Mediated Targets of Vpu ...... 28

1.6.2 TM Domain-mediated Targets of Vpu ...... 29

1.6.3 Vpu Target Antagonism is Conserved Across Other Viral Species ...... 35

1.7 Measuring Membrane Protein Oligomerization...... 36

1.7.1 Membrane Mimetics ...... 36

1.7.2 In vivo Methods for Measuring TM Helix-Helix Interactions ...... 39

1.7.3 Monitoring TM Helix-Helix Interactions In vitro ...... 42

1.8 Thesis and Hypothesis...... 50

Chapter 2: Dimerization of the Transmembrane Domain of Human Tetherin in Membrane Mimetic Environments...... 53

2.1 Introduction ...... 54

2.2 Materials and Methods ...... 55

2.2.1 Peptide Synthesis, Purification, and Labeling ...... 55

2.2.2 Tetherin TM Peptide Incorporation into Liposomes ...... 56

2.2.3 SDS-PAGE Analysis ...... 57

2.2.4 Circular Dichroism ...... 57

2.2.5 Fluorescence Spectroscopy ...... 57

2.2.6 Size exclusion chromatography ...... 58

2.3 Results ...... 59

2.3.1 Secondary Structure of TethTM ...... 59

2.3.2 Oligomerization of TethTM in SDS Micelles ...... 60

2.3.3 Dimerization of TethTM Probed by Pyrene Excimer Fluorescence ...... 64

2.4 Discussion...... 66

Chapter 3: FRET Analysis of the Promiscuous yet Specific Interactions of the HIV-1 Vpu Transmembrane Domain ...... 69

3.1 Introduction ...... 71

3.2 Materials and Methods ...... 73

3.2.1 Peptide Synthesis, Purification and Labeling ...... 73

3.2.2 TM Domain Peptide Incorporation into Liposomes ...... 74

3.2.3 Circular Dichroism Spectroscopy ...... 74

3.2.4 Fluorescence Spectroscopy ...... 75

3.2.5 Analysis of FRET Data ...... 75

3.3 Results ...... 82

3.3.1 TM Domain Peptides Retain Native-like Secondary Structure in POPC Liposomes ...... 82

3.3.2 Homooligomerization of Vpu and Tetherin ...... 84

3.3.3 VpuTM Binds Specifically to the TM Domains of Tetherin, NTB-A and PVR ...... 87

3.3.4 Target Peptides Compete for Binding with Vpu ...... 89

3.3.5 Energetics of Vpu-target Interactions ...... 91

3.4 Discussion...... 94

Chapter 4: Hydrophobic matching of HIV-1 Vpu transmembrane domain helix-helix interactions is optimized for its subcellular location ...... 96

4.1 Introduction ...... 97

4.2 Materials and Methods ...... 99

4.2.1 Estimation of TM Domain Lengths ...... 99

4.3 Results ...... 99

4.3.1 The Effect of Temperature and Lipid Phase on Peptide Secondary Structure and FRET Efficiency ...... 101 vii

4.3.2 Effect of Phospholipid Acyl Chain Length on Interactions with VpuTM and its Human Targets ...... 105

4.4 Discussion...... 108

Chapter 5: General Discussion and Future Directions ...... 111

5.1 Understanding Vpu-target oligomerization ...... 112

5.2 The Tetherin TM Domain Forms Parallel Homodimers ...... 113

5.2.1 The Vpu TM Domain Forms Similarly Weak Interactions with Itself and its Target TM Domains in POPC Membranes ...... 113

5.2.2 One Face to Find Them All and in the Membrane Bind Them ...... 114

5.3 The Influence of the Lipid Bilayer on Interactions Within the Vpu-target System ...... 115

5.4 Future Directions ...... 115

5.4.1 Interaction faces and structures of Vpu NTB-A/PVR heterodimers ...... 115

5.4.2 The role of Vpu homooligomerization ...... 117

References ...... 119

viii

List of Tables

Table 1.1 HIV-1 proteome. 15 Table 1.2 Examples of Vpu targets. 25 Table 1.3 Vpu sequence alignment. 27 Table 1.4 SIV antagonists of tetherin. 28 Table 1.5 Viral antagonists of tetherin. 36 Table 1.6 Studies of TM-TM helix association using FRET. 48 Table 2.1 Primary structures of the TM peptides. 55 Table 3.1 Vpu and target transmembrane domain peptides. 82 Table 4.1 Phospholipid bilayers used in this study and their properties. 100 Table 4.2 Homo and heterooligomer free energies of association of VpuTM with its human targets in PC membranes of increasing chain length. 108

List of Figures

Figure 1.1 Examples of α-helical and ß-barrel membrane protein structures. 3 Figure 1.2 Accommodation of hydrophobic mismatch in the lipid bilayer. 5 Figure 1.3 Mature HIV-1 virion structure. 16 Figure 1.4 HIV-1 life cycle overview. 18 Figure 1.5 Vpu structure. 22 Figure 1.6 Tetherin sequence and extracellular domain structure 30 Figure 1.7 NTB-A homo-dimer structure and domain architecture. 32 Figure 1.8 PVR structure and domain architecture. 34 Figure 1.9 Membrane mimetic properties. 38 Figure 1.10 Jablonski diagram. 45 Figure 1.11 Model FRET pair donor and acceptor emission/excitation spectra. 46 Figure 1.12 Quantifying FRET transfer. 49 Figure 2.1 Secondary structure of TethTM in membrane mimetic environments. 60 Figure 2.2 SDS-PAGE analysis of TethTM. 61 Figure 2.3 Size exclusion chromatography (SEC) analysis of TethTM in SDS micelles. 62 Figure 2.4 SEC of TethTM under reducing conditions. 63 Figure 2.5 Comparison of SEC mobility of GpATM and Teth(Scr) versus TethTM in SDS micelles. 64 Figure 2.6 SDS-PAGE analysis of TethTM peptides and GpATM. 64 Figure 2.7 Pyrene fluorescence emission spectra of TethTM. 66 Figure 3.1 Secondary structure of TM domain peptides reconstituted into POPC liposomes. 84 Figure 3.2 Fluorescence emission spectra of dansyl and dabsyl conjugated VpuTM in POPC vesicles. 85 Figure 3.3 Homooligomerization of Vpu and tetherin TM domain peptides. 86 Figure 3.4 Competition experiments with unlabeled target peptides confirm the interactions between Vpu and target TM domain peptides. 88 Figure 3.5 Hetero-oligomerization of VpuTM with its targets. 89 Figure 3.6 Target TM domain peptides compete for interaction with VpuTM. 90 Figure 3.7 Thermodynamic analysis of Vpu TM domain homo and heterooligomeric assemblies. 91 Figure 3.8 Estimating the VpuTM oligomer size from acceptor titrations and lipid dilution experiments. 92 Figure 3.9 Effect of VpuTM on TethTM homodimer formation. 93 Figure 3.10 Thermodynamic and kinetic analysis of the Vpu TM domain interactions with target TM domain peptides. 94 Figure 4.1 Secondary structure of TM domain peptides in fluid phase PC liposomes. 101 Figure 4.2 Secondary structure of Vpu TM peptide as a function of lipid phase and temperature. 102 Figure 4.3 Effect of temperature on peptide helicity in DOPC liposomes. 103

Figure 4.4 Effect of temperature on the FRET efficiencies of Vpu-target pairs. 104 Figure 4.5 The influence of lipid phase on Vpu-target interactions. 105 Figure 4.6 Homooligomerization of Vpu and tetherin TM domains in PC lipid membranes. 106 Figure 4.7 Heterooligomerization of Vpu and its human targets in fluid PC lipid membranes. 107 Figure 5.1 Putative Vpu-target complexes. 116

List of Abbreviations

AIDS acquired immune deficiency syndrome AP adapter protein AUC analytical ultracentrifugation BAX B-cell lymphoma-2 associated protein CCR5 C-C chemokine receptor 5 CD circular dichroism CMC critical micelle concentration CXCR4 C-X-C chemokine receptor type 4 DLPC 1,2-dilauryl-sn-glycero-3-phosphorylcholine DMPC 1,2-dimyristoyl-sn-glycero-3-phosphorylcholine DNAM-1 DNAX accessory molecule 1 DOPC 1,2 dioleoyl-sn-glycero-3-phosphorylcholine DPC dodecyl phosphocholine DPPC 1,2-dipalmitoyl-sn-glycero-3-phosphorylcholine DSPC 1,2 distearoyl-sn-glycero-3-phosphorylcholine DTT dithiothrietol EIAV equine infectious anemia virus ER endoplasmic reticulum EBP50 ezrin radixin moesin binding phosphoprotein 50 ESCRT endosomal sorting complexes required for transport FGFR fibroblast growth factor receptor FRET Förster resonance energy transfer GpA Glycophorin A GPI glycosylphosphatidyl-inositol HAART highly active antiretroviral therapy HFIP 1,1,1,3,3,3 hexafluoroisopropanol HIV human immunodeficiency virus IFN interferon KSHV Kaposi's sarcoma-associated herpesvirus LUV large unilamellar vesicle MHC major histocompatibility complex MLV multi lamellar vesicle MRE mean residue ellipticity Nef negative regulatory factor NF-κB nuclear factor kappa-light-chain-enhancer of activated B cells NK natural killer NLR nucleotide oligomerization and binding domain-like receptors NMR nuclear magnetic resonance NTB-A natural killer, T and B cell antigen PAGE polyacrylamide gel electrophoresis PMP plasma membrane profiling xii

POPC 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine PRR pattern recognition receptor PVR poliovirus receptor RICH2 Rho-GTPase-activating protein interacting with Cdc-42-interacting protein 4 homolog 2 RING really interesting new gene RLR retinoic acid inducible gene-I-like receptor RP-HPLC reverse phase HPLC SDS sodium dodecyl sulfate SEC size exclusion chromatography SIV simian immunodeficiency virus SUV small unilamellar vesicle TAB transforming growth factor activated kinase TAK transforming growth factor-β-activated kinase TCEP tris(2-carboxyethyl)phosphine hydrochloride TFA 1,1,1 trifluoroacetic acid TFE 1,1,1 trifluoroethanol TGN trans Golgi network TLR toll-like receptor TM transmembrane TMT tandem mass tag TRAF tumor necrosis factor associated factor Vpr virus protein r Vpu virus protein u β-TRCP beta transducin repeat containing protein

xiii

Chapter 1: Introduction .

1.1 Membrane Protein Folding and Oligomerization

Membrane proteins make up approximately one third of all proteins in the human genome and play critical roles in cell viability (1). Indeed these membrane spanning proteins have a diverse array of biologic activities including signal transduction, transport of molecules across the bilayer, cell adhesion and catalysis of various chemical reactions. Thus, increasing our understanding of how these proteins fold and function may significantly impact our understanding of human biology and disease; however, when compared to the study of soluble proteins, the study of membrane proteins has lagged behind. This is due in large part to the challenges in purifying a sufficient amount of protein in a functional form, as well as technical challenges in acquiring data for biophysical and structural studies. Most experimental difficulties arise from the intrinsic hydrophobicity of the membrane spanning regions and the need to accommodate them via a membrane mimetic.

1.1.1 The Three Stage Model of Membrane Protein Folding

The folding of membrane proteins has not been studied as extensively as the folding of their soluble counterparts; however, the fundamentals of membrane protein folding are thought to be less complicated. The transmembrane domains (TMs) of integral membranes can generally be categorized as α-helical or ß-barrel in structure (Figure 1.1). The formation of helical or sheet secondary structures is necessary to shield the polar peptide backbone from the non-polar environment of the lipid bilayer and to provide stability through hydrogen bond formation.

Figure 1.1 Examples of α-helical and ß-barrel membrane protein structures. (A) Rhodopsin (PDB 2I36) is an alpha-helical bundle that acts as a light driven proton pump. (B) OmpA (PDB 1BXW) is ß-barrel protein that serves a multitude of functions, including roles in membrane structural integrity, host immune evasion and cell adhesion.

The insertion of a membrane spanning TM domain into a lipid bilayer is dependent on the context of whether this event is taking place in vivo or in vitro. Although there are examples of genes that can be translated into soluble proteins that later insert into the bilayer, such as C. difficile toxins (2), or the Bcl-2 associated protein (BAX) (3), most TM domain insertion in vivo is mediated by complex protein machinery; the translocon for α-helices, or the BAM complex for β-barrels (as reviewed by (4, 5)). This is in contrast to an in vitro system, where this machinery will not be present and TM domains must be inserted into the membrane differently. TM domains can be added to the bilayer externally, with the goal of spontaneous insertion; however, unless this is a native characteristic of the membrane protein (as mentioned above for bacterial toxins, etc.) very little of the membrane protein is expected to become stably folded within the membrane. Another common method for in vitro peptide insertion includes co-dissolution of lipid and protein in organic solvent. Unfortunately, any folded soluble domains would not likely remain folded throughout this treatment, limiting the use of this method to small peptides of TM domains. Finally, detergents can be used to mediate the insertion of membrane proteins into the bilayer by adding the protein/detergent complex to bilayers followed by the subsequent removal of the detergent via dialysis.

Early analysis of α-helical TM domain folding by Popot and Engelman initially described a two-step folding process: firstly the insertion and formation of stable α-helices in the bilayer, followed by inter-helix packing to form the tertiary fold of the protein (6). This model was later revised to include a third stage in folding which encompassed a wide range of subsequent events, including the formation of quaternary contacts with other proteins, the binding of ligands, and the folding of loops/re-entrant loops. Soluble protein folding is largely driven by the hydrophobic effect, or the increase in water entropy that is observed after hydrophobic sections of a protein collapse into a hydrophobic core. Similar effects drive the initial insertion of a peptide into the membrane (Stage 1); however, as the membrane lacks water, the hydrophobic effect does not drive Stage 2. Interestingly, water molecules within the translocon may play a role during in vivo in folding of nascent TM domain peptide chains (7). As the growing polypeptide chain is synthesized by the ribosome, it passes into the translocon and eventually inserts/partitions into the lipid bilayer. The degree of helicity of the nascent polypeptide chain at this stage is debated, with some findings suggesting the helix can form either in the ribosome (8), in the translocon (9), or post-insertion into the bilayer.

1.1.1.1 Stage One

The insertion of the nascent peptide into the bilayer is the first step of membrane protein folding. This process is facilitated by the translocon, but what determines if a peptide will insert into the bilayer or not? Most data support a model in which a potential TM region must be of a sufficient length and hydrophobicity to partition into the bilayer. While analysis of α-helical TM segments puts the average length of a TM segment at 25 amino acids, structural analysis shows that this can vary greatly, from 14-39 amino acids (10). The resident membrane of a TM domain has also been generally linked to the final TM domain length, with proteins residing in the plasma membrane exhibiting average TM domain lengths of 25 amino acids vs trans-Golgi network (TGN) residing TM domains averaging only 20 amino acids in length. This may be due in part to hydrophobic mismatch, which occurs when the hydrophobic length of a TM domain does not correlate with the core length of a bilayer. TM domains that are shorter than the bilayer thickness induce disorder in the acyl chains of the bilayer, decreasing the

bilayer thickness. Helices that are too long are accommodated either by increasing their tilt in the membrane plane or by inducing acyl chain ordering, increasing the bilayer thickness (Figure 1.2).

Figure 1.2 Accommodation of hydrophobic mismatch in the lipid bilayer. (A) A hydrophobically matched helix in a lipid bilayer. The hydrophobic core of the TM domain (grey residues) matches the hydrophobic core thickness of the bilayer. Native TM domains usually follow the ‘positive inside rule’ where in positive residues (blue) are over represented at the intracellular termini. (B) Hydrophobic mismatch occurs when the hydrophobic core of a TM domain does not match the thickness of the lipid bilayer. (C) Lipid acyl chains extend (or order) to accommodate hydrophobic mismatch. (D) Helical tilt can be induced to accommodate hydrophobic mismatch. (E) Changes in oligomeric state can reduce lipid-helix contacts. (F) Acyl chains can compress (or disorder) to accommodate a shorter TM domain.

1.1.1.2 Stages Two and Three

The second stage of membrane folding involves the lateral packing of helices through helix-helix interactions. Helix packing in the bilayer is a balance between, 1) solvation of the helix by bilayer lipids (lipid protein interactions) and 2) protein-protein interactions. Packing interactions are largely dominated by van der Waals forces, but there are also electrostatic contributions that can contribute to interaction stability (discussed in 1.1.3). Contributions from van der Waals forces are frequently described as “knobs” into “holes” or “ridges into grooves” packing, in which large residues (knobs/ridges) pack into the spaces around small residues 5

(holes/grooves). In this case, contacting amino acids are present along extended lengths of a helical face, providing optimal van der Waals interactions, and also allowing specific electrostatic contributions to arise from pairs of polar/charged residues. The membrane environment itself can also influence how membrane proteins fold with the degree to which lipids interact preferentially with themselves or with protein structures, promoting or discouraging helix-helix contacts.

The third stage of membrane folding occurs after membrane protein helices have associated in the second stage, this can facilitate partitioning of additional protein segments, such as coils or short helices, into the membrane (11). The third stage can also involve the binding of cofactors and prosthetic groups after helix association. This is exemplified by bacteriorhodopsin, which kinetic experiments have shown binds retinal after helical associations (12).

1.1.2 Studying TM Domains in Isolation

Membrane protein assembly can be driven by a combination of the interactions between soluble domains as well as TM domain-TM domain interactions, making it difficult to deconvolute the contributions from the TM domains alone. Soluble domains also introduce technical challenges to the handling of membrane proteins as they can unfold and cause aggregation or loss of function. Thus it can be advantageous to study TM domains in isolation. Numerous studies have shown that the TM domain sequences of many integral membrane proteins possess all the information required to form higher order oligomers, including the TM domains of Glycpophorin A (GpA), M2, phospholamban (13–15). The seven TM domain membrane protein rhodopsin, one of the best studied members of the GPCR family, has been shown to assemble correctly even when expressed as individual transmembrane segments (16).

1.1.3 Forces Driving the Oligomerization of Membrane Proteins

The association of two or more helices in the bilayer must be more energetically favorable than their separation. The forces driving membrane protein oligomerization within

the context of the non-polar bilayer are inherently different than those driving soluble protein oligomerization. For example, the exclusion of water from the bilayer limits contributions from the hydrophobic effect and increases the energetic contributions from electrostatic interactions. Thus, TM-TM interactions can arise from factors intrinsic to the membrane protein sequence that favor the formation of protein-protein contacts over protein-lipid contacts.

1.1.3.1 van der Waals Forces

The main force driving the association of helices within a lipid bilayer is thought to be van der Waals forces, which are non-covalent interactions that result from the attraction between two complementary dipoles. Nonpolar molecules can form permanent or induced dipole moments, arising from fluctuations in their electron clouds, which can form weak electrostatic attractions. In membrane proteins, interactions between helices become stronger when the contact surface area of the two interacting helices is greater due to a larger number of electrostatic attractions. Thus motifs that allow for the close packing of peptide backbones over the length of the TM helix are said to facilitate helix-helix associations (17).

1.1.3.2 Electrostatic Interactions

Electrostatic interactions can also drive the association of membrane helices and can take the form of attractions between two polar residues, hydrogen bonds, or cation-π interactions. Perhaps the strongest of these interactions to occur in the bilayer, a hydrogen bond occurs when a hydrogen atom bonded to an electronegative atom also interacts with a second electronegative atom. Hydrogen bonds formed in the lipid bilayer are stronger than those formed in aqueous conditions for two reasons; there is no water to compete for hydrogen bonding, and the formation of the bond shields both polar groups from the nonpolar bilayer. Hydrogen bonding between two membrane helices can occur between two polar side chains, a polar side chain and the peptide backbone, or between the backbone Cα and the backbone carbonyl.

Electrostatics can also influence helix association through cation-π, π-π or salt bridge formation. The π character of aromatic residues (Phe, Tyr or Trp) can provide an electronegative surface to which positive residues (Arg, Lys, or His) on neighboring helices can interact with (18–21). Designed cation-π interactions introduced in a GpA background provide dimer stability in an E. Coli membrane environment (19). The pairwise contact of acidic and basic residues can form salt bridges that are particularly strong in the low dielectric environment of the lipid membrane. The role of interhelical salt bridges has been demonstrated for the T-cell and other immune receptors (22, 23).

1.1.4 Helix-Helix Interaction Motifs

The oligomerization of α-helical TM domains is largely dictated by the residues lying on the interaction face(s). These patterns of amino acids result in structural motifs that optimize the attractive forces between two helices. Interaction motifs described to date fall into three main classes: the GG4, heptad, and polar residue motifs (17). These amino acid patterns have been shown to favor the formation of membrane protein oligomers by having key residues placed in critical positions, as discussed below.

The first TM helix-helix interaction motif to be discovered was the GG4 motif found in the erythrocyte glycoprotein GpA. The packing of two GpA helices is mediated by two glycine residues separated by 3 residues in the primary sequence (GxxxG), thereby placing them on the same helical face (24). The glycine residues act as holes that larger hydrophobic residues, in this case Val residues, can pack into, stabilizing dimer formation through van der Waals interactions. It was later noted that the GG4 motif is over represented in predicted TM sequences and further studies have shown that glycine can be replaced by other so called “small” amino acids such as alanine and serine as long as they follow a i, i+4 separation of small residues (25). The alanine rich face of Vpu is similar to an extended GG4 type motif in which small amino acids occur on one helical face.

More diverse, yet less common, are heptad motifs in which both large and small amino acids can be found. Like the GG4 motif, heptad motifs are also largely dependent on an increase 8

in van der Waals forces to energetically drive the association of two helices. A notable example of this motif is found in the cardiac sarcoplasmic reticulum phospholamban which has a IxxxLxxIxxxLxxIxxxL motif that results in the large hydrophobic residues (Leu/Ile) lining the same helical face, and providing the surfaces for van der Waals interactions (26).

1.1.5 Factors Affecting Helix-Helix Interaction Strength

While the amino acid sequence motif driving the association of two helices does have an effect on how strong or weak the interaction will be, the membrane environment itself can influence the level of interaction. The hydrophobic thickness and phase of the bilayer can influence the level of TM-TM associations within the bilayer. Changes in the hydrophobic thickness of the bilayer can result in hydrophobic mismatch of helices, while the phase of the lipid can influence the favorability of lipid-protein interactions. Hydrophobic mismatch occurs when there is a difference between the length of the hydrophobic thickness of a membrane protein and the hydrophobic thickness of the bilayer it spans. The system can respond hydrophobic mismatch through 3 major mechanisms: by changing the tilt of the membrane protein, by changing the membrane protein oligomerization state, or by changing the order of the lipid chain length to relieve the mismatch (Figure 1.2).

1.1.6 The Role of Helix-Helix Interactions in Innate Immunity

The direct interaction of the TM domains has been shown to be critical for immunological related functions. Examples include major histocompatibility complex (MHC)- Class-II molecules, which homooligomerize through GxxxG motifs, as well as the T-cell receptor, which forms TM mediated heterooligomers with CD3 and is responsible for the recognition of peptide antigens bound to MHC molecules (22, 27). The human immunodeficiency virus 1 (HIV- 1) has also been shown to use direct TM-TM domain interactions in order to evade the human immune system, discussed in detail in section 1.6.2.

1.2 Viral Immune Evasion

Members of all domains of life are susceptible to viral invasion/infection. As such, hosts ranging in complexity from bacteria to higher-order vertebrates have developed immune systems to fight viral infection and viruses have in turn developed counter measures. These viral immune system neutralizers often come in the form of proteins and can be active in all stages of the viral life cycle; that is, when the virus exists as a free virion or as genomic DNA inside an infected cell (termed a provirus). The viral proteins that drive immune evasion strategies may also represent novel targets for development of antiviral therapies, making it important to understand their mechanisms of action and how they are inhibiting the host organism.

The mammalian immune system presents a multi-layered set of defenses to any invading pathogen. Immune responses can be categorized as either innate or adaptive, both of which can be mitigated by the activity of viral proteins. The responses of the innate immune system are more rapid and less specific than the adaptive immune system, which is highly specific and results in the creation of an immune memory. Aspects of the innate human immune system include barriers (physical, chemical and biological), inflammation, the complement system, and natural killer (NK) cells. The adaptive immune system also employs specific cell types (T and B cells) in its cell mediated responses as well as in the secretion of neutralizing antibodies.

Activation of any immune response first requires that the host senses the infection. This is accomplished by detecting the invading organism directly by sensing its constituents (i.e. uncapped viral RNA, or viral proteins) or indirectly by sensing the cellular stress caused by infection (28). This recognition is carried out by a diverse group of pattern recognition receptors (PRRs), which include toll like receptors (TLRs), retinoic acid inducible gene-I-like receptors (RLRs), nucleotide oligomerization and binding domain-like receptors (NLRs), and pyrin-HIN domain receptors (28–32). Each receptor family has multiple members, and each individual member recognizes pathogen-associated products, damage-associated molecular patterns (such as reactive oxygen species) or apoptotic/necrotic cells (33). 10

An important consequence of PRR engagement is the production of interferons (IFNs), cytokines responsible for regulating the immune system. Hallmarks of the IFN response include activation of immune cells (i.e. natural killer (NK) cells/macrophages), upregulation of antigen presentation (MHC bound antigens), stimulation of antiviral genes (i.e. tetherin), and induction of the adaptive immune response (34). Adaptive immune responses involve the recognition of specific, non-self, antigens and direct the removal of any cells/microbes presenting these antigens. Microbial and viral antigens include proteins, carbohydrates, lipids and nucleic acids that are recognized by antibodies as well as receptors on T and B cells. Recognition of antigens results in activation of an immune response and the eventual engulfment and destruction of the antigen/pathogen by macrophages. After being activated, some T and B cells will become memory T and B cells, which facilitate faster and stronger immune response upon subsequent exposure to the antigen.

Viruses use multiple strategies to evade the immune responses of host organisms including avoiding initial detection, interfering with chemical messaging, inhibiting apoptosis of infected cells and down-regulating immune system proteins. For many retroviruses, the first line of defense against the immune system arises from the low fidelity of their reverse transcriptase enzymes. A relatively high error rate in the reverse transcription of viral RNA to DNA drives the production of escape mutants that are able to avoid immune recognition (35). This high error rate is also responsible for the generation of mutants that cannot be targeted by therapeutics. Both the adaptive and innate immune systems are regulated through chemical messengers (IFNs/chemokines/cytokines). Some viruses can interfere with this chemical messaging process by expressing proteins that bind and inactivate chemical messengers, a notable example being poxviruses that produce soluble versions of native receptors that can bind interferons, interleukins, tumor necrosis factors and other chemokines and cytokines (36). Some viruses block apoptosis of infected cells by interacting with elements of the pathways regulating programed cell death (as reviewed in (37)). Finally, viruses can modulate the expression of cell surface proteins in infected cells, a strategy employed by the human immunodeficiency virus (HIV) via its accessory protein Vpu (38).

1.2.1 The HIV-1 Pandemic

During the early 1980s there was an unsettling rise in cases of opportunistic infections in young men, which was ultimately attributed to a new disease, acquired immune deficiency syndrome (AIDS). In 1983, Françoise Barré-Sinoussi and Luc Montagnier linked the cause of AIDS to a retrovirus now termed the HIV-1, a discovery for which they shared the 2008 Nobel Prize in Medicine (39–41). Subsequent research found closely related viruses in simian populations (simian immunodeficiency viruses (SIVs)) infecting non-human primate species, as well as a second antigenically distinct virus in humans (HIV-2) (42–46). Since apes, such as chimpanzees, are known to hunt other mammals, including monkeys, it is hypothesized that these eating habits led to the initial cross-species spread of the disease. Similarly, bush-meat hunting practices in Western Africa are thought to have been the initial cause of transmission to humans, although now the disease is primarily transmitted sexually (47, 48). This cross- species transmission to humans is thought to have occurred through at least four independent events, giving rise to four distinct lineages of HIV-1 termed Groups M, N, O, and P (49–53). HIV- 1 has since become one of the most devastating infectious diseases of the modern age infecting 36.7 million people worldwide, with 1.8 million new cases occurring in 2016 (UNAIDS, 2016).

1.2.2 HIV-1 Pathogenesis and Disease

HIV-1 exhibits a limited number of transmission routes, with most infections beginning at mucosal membranes during sexual contact. HIV-1 can also be transmitted by exposure to infected body fluids, from mother to child during delivery or breast-feeding, and via the sharing of intravenous drugs. Before the introduction of blood screening for HIV-1, patients and hospital staff were also susceptible to infection with HIV upon exposure to tainted blood products. However, modern testing techniques and practices have largely removed this mode of transmission.

The sexual transmission of HIV-1 is an inefficient process, with infection often resulting from only a single virus. This effect is a result of bottlenecks in viral transmission caused by selection of specific phenotypes in the genital tract of the donor as well as the inability of the

majority of viruses to penetrate the mucosa of the recipient (54). After the initial infection event the founder virus begins replication at the site of infection by infecting CD4+ memory T cells. However, the virus often evolves to infect a broader range of T cell types by switching its co-receptor usage from CCR5 (C-C chemokine receptor 5) to CXCR4 (C-X-C chemokine receptor type 4) via mutation in gp120 (55, 56). As a result, a major characteristic of ongoing infection is the depletion of host CD4+ cells and loss of immune response. The time course of an HIV-1 infection can be broken down into 4 distinct stages that are defined by two markers, the count of CD4+ T cells and virions (the viremia) in the blood (reviewed in (57)). After establishing an infection there is an acute phase where the host viral load rapidly increases and the virus begins to spread to other organs and tissues. This continues until the viral load in the host reaches its peak as HIV-1 specific immune responses start to be observed (i.e. gp41 antibodies, and HIV-1 specific CD8+ T cells). These immune responses result in an initial decrease in the host’s viral load. The infection then enters a chronic phase during which CD4+ T cell levels gradually fall while the viral load begins to steadily increase. The chronic phase can persist for years before the onset of more serious AIDS symptoms; fatigue, rapid weight loss, and opportunistic infections that are the primary cause of patient mortality (58).

1.2.3 HIV-1 Genome and Proteome

HIV-1 is a retrovirus with a single-stranded RNA genome of ~9.2kb which is reverse transcribed and integrated into the host cell genome. This occurs during a seven stage life cycle: attachment/binding, fusion, reverse transcription, integration, replication, assembly and budding. The HIV genome itself is organized into 9 open reading frames that encode for 16 proteins listed in Table 1.1. Most proteins are expressed as part of one of three polyproteins, Gag, Env, or Pol. The Gag and Pol polyproteins are cleaved into their final forms by the viral protease, while the Env polyprotein (gp160) is cleaved by the host protease furin (59). The Env and Gag proteins make up the structural core and outer envelope of the final virus particle, while the Pol products carry out critical enzymatic functions and are also packaged into the virion. The remaining viral proteins include Tat and Rev, regulatory proteins that control viral gene expression, and four accessory proteins, Vpr, Vif, Nef and Vpu. Although considered

dispensable in most in vitro experimental conditions, accessory proteins significantly enhance HIV-1 replication in vivo by taking on facilitative roles (i.e. immune evasion) (60).

Table 1.1 HIV-1 proteome. Protein Class Protein Gene Polyprotein Matrix (MA) Gag Gag polyprotein Capsid (CA) Gag Gag polyprotein Nucleocapsid Gag Gag polyprotein P6 Gag Gag polyprotein Spacer peptide 1 Gag Gag polyprotein Spacer peptide 2 Gag Gag polyprotein Structural Gp41 Env Gp160 Gp120 Env Gp160 Reverse Transcriptase (RT) Pol Pol Polyprotein RNase H Pol Pol Polyprotein Integrase Pol Pol Polyprotein Protease Pol Pol Polyprotein Tat Tat N/A Regulatory Rev Tat N/A Vpu Vpu N/A Nef Nef N/A Accessory Vif Vif N/A Vpr Vpr N/A

1.2.4 HIV-1 Virus Particle Structure

The HIV-1 virus particle is spherical in shape, with a membrane coat that surrounds and protects the internal proteins and genome (Figure 1.3). Construction of the viral particle begins in the host cell when proteins that are to be packaged into the virion assemble at the cell surface. During trafficking to the cell surface, the Env precursor (gp160) is cleaved by cellular proteases (59). The result is a heterodimer made up of the transmembrane protein (gp41) and the glycoprotein gp120 (61–63). These Env transmembrane heterodimers form trimers (called spikes) and stud the surface of the nascent viral particle. The lipid bilayer surrounding the viral core is taken from the host cell when the virus particle buds off of the plasma membrane. An average of ~14 Env spikes are found on the HIV-1 virion surface, along with cellular membrane proteins that happened to be present during the budding process (64, 65). The internal core of the virus is comprised of several layers that begin with the HIV-1 RNA genome bound to the nucleocapsid at its center (66). The genome is surrounded by a cone-shaped capsid shell made up of copies of CA that encapsulate the genome and viral proteins, including reverse 15

transcriptase, integrase, Tat, Vif, Vpr and Nef (66–68). The capsid is surrounded by a matrix composed of MA protein which is in turn surrounded by the lipid membrane.

Figure 1.3 Mature HIV-1 virion structure. The mature HIV-1 virion is approximately 110nm in diameter when imaged using cryoelectron microscopy (64). The virion is spherical in shape and encapsulates 15 HIV-1 derived proteins as well as a lipid membrane taken from the infected cell during budding.

1.3 HIV-1 Life Cycle

After entering the body, HIV-1 initially infects CD4+ T cells and replicates through a series of steps summarized in Figure 1.4, culminating in the production and release of new virus particles (69). To gain access to the cytosolic cellular machinery, the viral particle must bind the host cell and undergo membrane fusion. The envelope glycoprotein (Env) on the surface of the virion binds to the host cell through a CD4 specific binding site on gp120 (63, 70). Env binding to CD4 results in conformational changes in the viral protein, allowing subsequent binding of a co- receptor, most commonly CCR5 or CXCR4 (71). Binding to the co-receptor induces conformational changes in the TM region of gp41, which drives its fusion peptides into the host cell membrane (61, 62). This brings the host and viral membranes into close proximity, allowing them to fuse, creating a pore through which the encapsulated viral core enters the cytoplasm. Once inside the cell the viral capsid coat is removed, a process called uncoating, releasing its

RNA and proteins into the cytoplasm, after which reverse transcription of its single stranded RNA genome into a DNA-RNA hybrid is initiated. The viral RNase H degrades the RNA portion of the DNA-RNA hybrid as DNA synthesis proceeds, facilitating the synthesis of the final linear double-stranded DNA HIV-1 genome (72). The nascent DNA travels along the microtubule network, passes through the nuclear pore complex, and is ultimately integrated into the host cell genome by the HIV-1 integrase (73, 74). The integration catalyzes the removal of two nucleotides from the 3’ end of the linear viral DNA, creating a new 3’ end that attacks the host target DNA. This joins the 3’ end of the viral DNA to the 5’ end of the host target DNA, creating an integration intermediate. Subsequent steps (i.e., overhang removal and ligation) are then catalyzed by host cell DNA repair enzymes (75). Integration can occur at different places in the host cell genome, but tends to prefer transcriptionally active sites (76).The integrated HIV-1 genome, termed a provirus, is transcribed into mRNA by cellular host cell RNA polymerase II, which begins transcription from promoter elements located in the 5’ long terminal repeat region. Initially, most transcripts are terminated early, due to the lack of processivity of RNA Pol I. The viral regulatory protein Tat stimulates RNA Pol II processivity, resulting in an increase in the amount of full length transcripts (77, 78).Transcribed viral mRNAs are processed by the host cell spliceosome into several splice variants, including one fully spliced variant containing the genes for Tat, Rev and Nef that is exported and translated using endogenous pathways. The Rev protein facilitates the export and protection of any incompletely spliced or unspliced transcripts. Viral mRNAs are translated by the cellular machinery into the various HIV-1 proteins and polyproteins. Copies of the HIV-1 RNA genome and virus structural proteins are orchestrated to assemble at the plasma membrane by the Gag polyprotein (79–84). As viral proteins are being produced the accessory proteins Vpu and Nef downregulate host cell proteins in order to evade the immune response and facilitate viral replication (85). When the immature virion is finally assembled at the cell surface the Gag recruits the host ESCRT machinery and directs the budding off process (86, 87). At this stage the host restriction factor tetherin can interfere with virus release, as described in detail in section 1.6.2.1. Nascent virions then mature into their final form as the HIV-1 protease becomes active and cleaves the Gag and Pol polyproteins into individual units. The free Gag proteins undergo a series of

conformational changes and multimerization steps that lead to the formation of the final mature virion (88, 89).

Figure 1.4 HIV-1 life cycle overview. The cycle begins when the invading virus recognizes and binds the CD4 receptor on the surface of the target cell through gp120 (Step 1: Binding). The Env/CD4 interaction induces a conformational change that induces (Step 2: Fusion). Once inside the cell the viral core is uncoated and reverse transcription of the viral RNA genome commences (Step 3: Reverse Transcription). This creates a double stranded DNA genome that is shuttled into the nucleus and integrated into the host’s genome by the viral integrase and host DNA repair enzymes (Step 4: Integration) Host RNA Pol II transcribes the genome into mRNA molecules (Step 5a: Transcription). These mRNA molecules are exported to the cytoplasm where they are translated by host ribosomes into the encoded proteins/polyproteins (Step 5b: Translation). The Gag polyprotein multimerizes at the plasma membrane and organizes the assembly the viral particle (Step 6: Assembly). Completed viral particles then bud off from the host cell through the use of host ESCRT machinery. (Step 7: Budding). After this the viral protease in the nascent virion becomes active, cleaving Gag and Pol polyproteins, allowing the virus to reorganize into a mature infectious virion.

1.4 HIV-1 Disease Pathology, Treatment and Prevention

Since the initial discovery that HIV-1 is the causative agent of AIDS extensive efforts have been made to prevent infection and to treat or cure the disease. Today HIV-1 is treated with >20 FDA approved antiretroviral drugs that target viral entry, reverse transcription, integration and protease activity. The use of a drug cocktail acting on different viral targets has been dubbed highly active antiretroviral therapy (HAART), and can suppress HIV-1 replication sufficiently that patients exhibit undetectable levels of virus (90). Patients undergoing HAART have decreased viral loads, functional immune systems, and in many developed parts of the world, progression to AIDS is increasingly rare. Despite the fact that viral loads in patients undergoing HAART treatment decrease to undetectable levels, stopping treatment results in the return of high viral loads of the virus since the virus can persist in latent reservoirs (91–93). Thus, current treatments are not curative and require maintenance of HAART for the lifetime of the patient. However, there has been documented case of a functional HIV-1 cure (i.e., the so- called Berlin patient). An HIV-1 infected patient received a bone marrow transplant (to treat leukemia) from a donor that was naturally immune to HIV-1 due to a ∆32 mutation in the HIV-1 CCR5 co-receptor (94, 95).

Alongside the development of HAART, there has been considerable effort in the prevention of HIV through the design of an HIV vaccine (96). However, these efforts have so far been in vain. Many experimental vaccines have been tested but none have been capable of raising broadly neutralizing antibodies capable of impacting HIV infection. Several aspects of HIV make the induction of broadly neutralizing antibodies difficult, including the large diversity in global strains, the high mutation rate caused by a low fidelity reverse transcriptase as well as the high glycosylation level and low density of Env on the virion surface.

1.5 HIV-1 Accessory Proteins and Mechanisms of HIV-1 Immune Evasion

In addition to the structural and regulatory genes, HIV-1 also expresses 4 accessory proteins: Nef, Vpu, Vpr and Vif. Accessory proteins are not always necessary for viral replication 19

in in vitro experiments but play important roles as virulence factors in vivo. Both Nef and Vpu function to modulate the function and expression of a range of cellular proteins, creating a better cellular environment for the production of virus particles, and facilitating the immune evasion of the infected cell. Focusing on Nef, as Vpu will be discussed in detail later, the importance of Nef to HIV-1 pathogenicity was demonstrated in a study in which defective mutants of the nef gene were found in long-term nonprogressors (97). Acting as an adapter protein, Nef binds to CD4 and recruits the clathrin adapter AP-2, driving the downregulation of CD4, via the ESCRT pathway. Loss of CD4 on the cell surface prevents superinfection while loss of intracellular CD4 ensures that it cannot bind to Env intracellularly and that the virions can assemble properly. In a similar fashion, Nef can bind the cytoplasmic tails of the MHC-class I molecules HLA-A and HLA-B during the secretory pathway. Subsequent recruitment of the AP-1 complex directs the MHC molecules to endosomes, thereby protecting the infective cells from the adaptive immune response (98–101). Nef has also been implicated in affecting the activities of MHC-Class II molecules, PAK(102), DC-SIGN(103), and PVR (104). Also, SIV Nef proteins are capable of the downregulation of SIV tetherins, an ability that HIV-1 Nef lacks, instead human tetherin is downregulated by HIV-1 Vpu.

Nef and Vpu remain in the host-cell and are not actively packaged into budding virions. This is different from Vif and Vpr, whose roles take place shortly after membrane fusion between the virion and newly infected cell, requiring that they are packaged into the virion. Vif modulates the function of APOBEC3G, a cytidine deaminase that exerts innate antiretroviral activity by installing C to U mutations in the viral DNA (105). This hypermutation of the HIV-1 genome results in non-functional viral proteins upon their translation. Vif decreases the levels of active APOBEC3G by targeting it for proteasomal degradation (106, 107). Vpr has been shown to function in genome integration and cell cycle arrest leading to infected individuals having more CD4+ T cells in the G2 phase than uninfected controls (108). Although the benefits of this cell cycle arrest are not well understood, it has been shown that HIV-1 genome replication is enhanced under these conditions.

1.5.1 HIV-1 Vpu

The small integral membrane protein Vpu is encoded by HIV-1 and is expressed in infected cells (109, 110). The current consensus is that the main functions of Vpu (and Nef) are to reduce the cell surface expression of various proteins by altering their subcellular trafficking or directing their degradation. A recent proteomic screen using plasma membrane profiling (PMP) with the tandem mass tag (TMT) strategy has shown that >100 cell surface membrane proteins were downregulated during HIV-1 infection (111). The Vpu-mediated downregulation of these cell surface proteins is accomplished either by ubiquitin-mediated degradation or by sequestration to internal compartments by altered trafficking. Overall, this activity has a profound effect on the pathogenesis of HIV-1 as is discussed in detail later. Vpu has also been reported to display ion channel activity that is susceptible to channel blocking drugs (112, 113).

1.5.2 Vpu Structure and Function

Structurally, Vpu is an 81 residue membrane associated protein comprising a short N- terminal domain followed by a single α-helical TM domain and two cytosolic helices separated by a short loop (110, 114) (Figure 1.5). The TM domain of Vpu is of critical importance as it mediates interactions with a subset of Vpu targets and also mediates homooligomerization of the protein into higher order oligomers. Historically, Vpu has been thought to oligomerize into pentameric structures, although other multimeric arrangements (tetramer to hexamer) have been modeled computationally (109, 115, 116). More recent studies, using either computational approaches or chemical cross-linking paired with ultracentrifugation, have provided evidence for the existence of a heterogeneous mixture of Vpu oligomers ranging from dimers to 9mers (117, 118).

Figure 1.5 Vpu structure. (A) Structure of Vpu of Vpu solved by solution NMR in DHPC micelles (PDB 2N28) (119). The Vpu TM domain is colored green, conserved phosphoserine residues are shown in blue, and the membrane anchoring W76 is shown in red. (B) Amino acid sequence of Vpu from the NL4-3 HIV-1 strain with annotated structural features, including underlined helical elements. The conserved ß-TrCP binding motif with phosphoserine sites S52/56 is indicated in blue, while W76 is highlighted in red.

Within the Vpu cytoplasmic domain, the first helix displays amphipathic characteristics and is thought to lie on the membrane surface. Also, W76 has been proposed to act as a membrane anchor for the C-terminus of Vpu, which may lead to preferential localization of Vpu within lipid microdomains or rafts (120). The amphipathic helix is followed by a loop containing a highly conserved DSGNES motif that includes two phosphoserine sites (S52/S56) that are phosphorylated by casein kinase II and are a requirement for β-TrCP recruitment (121). Solution NMR structural studies of the cytoplasmic domain show that phosphorylation of the two sites only modestly alters the structure of Vpu (122).

Vpu is translated late in the HIV-1 replication cycle from a bicistronic mRNA that also encodes the Env polyprotein. The Vpu from HIV-1 subtype B (the most commonly studied subtype) localizes primarily to internal membranes, the trans Golgi network (TGN), the endoplasmic reticulum (ER) and the endosomal system, while only small numbers of Vpu molecules have been observed at the plasma membrane (123, 124). Vpu carries out its primary functions in the ER, endosomal compartments and the TGN, where Vpu induces the rapid degradation of CD4, preventing superinfection and allowing for proper virion assembly. Vpu also enhances viral particle release by reducing the amount of tetherin on the host cell surface. Finally, Vpu protects infected cells from NK and CTL cell mediated lysis by reducing the cell surface levels of NTB-A, PVR, CD1d, CD74, and MHC-Class I molecules. 22

Vpu-mediated downregulation of CD4 has been shown to require only the cytosolic domain of Vpu, with the particular need for two conserved phosphoserines for the binding of the E3 ligase β-TrCP (125). Activity against CD1d, CD74, and MHC-Class I molecules have also been linked solely to the cytoplasmic domain of Vpu. However, down-modulation of other proteins (NTB-A, tetherin, PVR) has been shown to require the Vpu TM domain. In the case of tetherin, this was supported by mutational analysis in vivo as well as recent NMR studies in vitro (126–129). Vpu causes internalization and ubiquitin-mediated degradation of tetherin through recruitment of the cellular E3 ligase β-TrCP (130–132).

Initial experimental studies of Vpu homooligomerization along with its structural similarity to the influenza M2 proton channel, led to the hypothesis that Vpu may function as an ion channel. This hypothesis has been confirmed both in cell culture (xenopus oocytes and E. coli) and in planar lipid bilayers (113, 133, 134). In those experiments, Vpu was able to increase the permeability of membranes to potassium and sodium, an activity dependent on its TM domain, as channel activity was abolished if the TM domain sequence was scrambled. However, Vpu ion channel activity has never been clearly linked to a biological activity, bringing into question whether ion channel activity is a function that is being actively selected for, or just an artifact of Vpu homooligomerization under certain experimental conditions. Nevertheless, measured Vpu ion channel activity has been shown to be weakly selective for monovalent cations. Blocking this activity with various amiloride derivatives has been shown to reduce virus release in certain cell models, suggesting that this may have some relevance in vivo (112, 134).

1.5.3 Vpu Mediated Antagonism of Cell Surface Proteins

Since the discovery that Vpu can influence the cell surface presence of cell surface proteins there has been significant interest in understanding the mechanism by which Vpu removes proteins from the cell surface. The general model for this activity requires the direct interaction of Vpu and a target protein, followed by the subsequent recruitment of a cellular protein co-factor: AP-1/2 or βTRCP (131). The binding of βTRCP2 results in the formation of a ternary complex that drives the ubiquitination and degradation of the target protein via the

proteasomal or lysosomal degradation pathways (130, 131, 135) . An alternative pathway exists in the form of AP-1 binding, which results in the aberrant trafficking of the target protein and its sequestration in a perinuclear compartment (136).

Table 1.2 Examples of Vpu targets. The listed targets of Vpu represent a subset of Vpu targets bound via their cytoplasmic or transmembrane domains Downmodulated Mechanism Targeting Biological Effects Protein Mechanism CD4 -Ubiquitination by SCF Cytoplasmic -Prevents superinfection E3 ligase via β-TrCP domain -Release of mature virions resulting in degradation by ERAD pathway CD1d -Suppression of Cytoplasmic -Interference with NK cell recycling domain activation -No degradation or enhanced endocytosis CD74 -unknown Cytoplasmic -Reduced T cell activation domain SNAT-1 - Β-TrCP dependent Cytoplasmic -Impaired T cell mitogenisis degradation domain Tetherin -Β-TrCP dependent TM -Enhancement of virion degradation release

NTB-A -No degradation or TM -Protection from NK cell enhanced endocytosis mediated lysis by prevention -Sequestration and of NK cell degranulation altered trafficking? PVR -No degradation TM -Prevents NK cell mediated -Sequestration to lysis of infected cells perinuclear compartment CCR7 -No degradation-or TM -Impaired CD4 cell migration enhance endocytosis

1.5.4 Importance of HIV-1 Vpu to the Group M Pandemic

Analysis of Vpu sequences across Group M, N, O, and P strains show that Vpu is highly variable but does have some conserved regions that retain the overall structure of Vpu: an N- terminal α-helical TM domain followed by a helix-loop-helix domain. However there are major differences in the activities of the Vpu proteins from different HIV-1 Groups. The pandemic Group M Vpu is the only Vpu capable of downregulating both tetherin and CD4 efficiently. In

contrast, the non-pandemic strains can only target tetherin (Group N) or only CD4 (Groups O and P), although some Group N strains have been shown to have activity against both tetherin and CD4 (137–140). Thus, only the pandemic strains of HIV-1 are said to have a “fully functional” Vpu. The alanine face of the Vpu TM domain required for tetherin downregulation is extremely well conserved across Group M and N strains of HIV-1 but is not present in Groups O and P Table 1.3. As a result only Group M and N strains are capable of downregulating tetherin, and presumably the other targets requiring this face of the TM domain. Additionally, Groups N and O have severe mutations in their E3 ligase binding site, affecting their ability to downregulate CD4 and presumably other targets.

Table 1.3 Vpu sequence alignment. NL4-3 Vpu sequence is shown on top for comparison. Representative sequences from Group M subtypes (recombinant sequences were omitted from this alignment) as well as Group N, O, and P. The alanine rich face required for activity against tetherin, PVR and NTB-A is highlighted as well as the β-TrCP binding site. HIV-1 Strain Vpu Amino Acid Sequence NL4-3 -MQPIQIAIAALVVAIIIAIVVWSIVIIEYRKILRQRKIDRLIDRLIERAEDSGNESEGEISALVEMGVEMGHH---APWDIDDL- A1.AU.03.PS1044_Day0.DQ676872 MTPLEIWSIVGLVVALIIAIVVWTIVGIEYKKLLKQRKIDRLIERIRERAEDSGNESEGDTDEL-ALLIEMGNY---DLGNDINL- A1.UG.92.92UG037.AB253429 MQLLEICAVVGLVVALIIAIVVWTIVGIEYKKLLKQRKIDRLVDRIRERAEDSGNESDGDREEL-SLLVDMGDY---DLGDDNNL- A2.CD.97.97CDKTB48.AF286238 MSPLAILSIVGLVVASILAIVVWTVVFIEYRKIKKQRKIDWLLERISERAEDSGNESDGDTEEL-SKMVGMGNL---GFWDDNDV- A2.CY.94.94CY017_41.AF286237 MLPLVILAIVGLIVALILAIVVWTIVFIEYKKIKKQRKIDWLIKRISERAEDSGNESDGDTEEL-SALVERGHL---DFGDVNNV- B.NL.00.671_00T36.AY423387 MQSLTIVAIVALVVAAIIAIVVWTIVFIEYRKILRQRKIDRLIDRIRERAEDSGNESEGDQEEL-SALVEMGHH---APWDVDDL- B.FR.83.HXB2_LAI_IIIB_BRU.K03455 -QPIPIVAIVALVVAIIIAIVVWSIVIIEYRKILRQRKIDRLIDRLIERAEDSGNESEGEISALVEMGVEMGHH---APWDVDDL- B.TH.90.BK132.AY173951 MHSLQILGIVALVVAGIIAIVVWSIVIIEYRKILRQRKIDRLIDRIIERAEDSGNESEGDQEEL-SALVEMGHL---APWDINDX- B.US.98.1058_11.AY331295 ---LYILGIVALVVAAILAIVVWSIVLIEYRKILKQKKIDKIIDRIRERAEDSGNESEGDQEEL-SALVEMGHYAHHAPWNIDDL- C.BR.92.BR025_d.U52953 GRIDYRLGVGALIVALIIVIIVWTIAYIEYRKLVRQRRIDWLVKRIKERAEDSGNESGGDTEEL-ETMVDMGHL---RLLDGNDL- C.ET.86.ETH2220.U46016 AKVDYRIVIVAFIVALIIAIVVWTIAYIEYRKLLRQRRIDRLIKRTRERAEDSGNESDGDTEEL-STMVDMGNL---RLLDVNDL- C.IN.95.95IN21068.AF067155 VNLDYKLGVGALIVALIIAIVVWTIVYIEYRKLVQQRKIDWLIKRIRERAEDSGNESEGDTEEL-STMVDMGRL---RLLDVNDL- C.ZA.04.04ZASK146.AY772699 ASVDYRLGVGALIIALILAIVVWIIVYIEYRKLLRQRKINKLIDRIRDREEDSGNESEGDIEEL-ATMVDMGHL---RLLDDNNL- D.CD.83.ELI.K03454 MQPLGIIAIAALVVAIILAIVVWTIVFIEYRRIKKQRRIDCLLDRITERAEDSGNESEGDREKL-SKLVEMGHH---APWDIDDL- D.CM.01.01CM_4412HAL.AY371157 -QSLEILAIVALVAALIIAIVVWTIVYIEYRKIKKQRQIDRLIDRIRERAEDSGNESDGDREEL-STLMEMGHH---APWDVADDI D.TZ.01.A280.AY253311 MTPLEITAIAALVIASILAIVVWTIVYIEYRRIKKQKKIDWLIDRIRERAEDSGNESDGDTEEL-STLVEMGHH---APWNVDDI- D.UG.94.94UG114.U88824 MQPLEILAIVALVVALILAIVVWTIVFIECKKLRRQRKIDWLIDRIRERAEDSGNESEGDKEEL-SALVEMGHD---APWDADDM- F1.BE.93.VI850.AF077336 MSYLLAIGIAALIVALIIAIVVWTIVYIEYKKLVRQRKINKLYKRIRERAEDSGNESEGDAEEL-AALGEMGPF---IPGDINNL- F1.BR.93.93BR020_1.AF005494 MSNLLAIGIAALIVALIITIVVWTIAYIEYKKLVRQRKINRLYKRISERAEDSGNESEGDAEEL-AALGEVGPF---IPGDINNL- Group M F1.FI.93.FIN9363.AF075703 MSDLLAITIVAFIVALIIVIVVWTIVFIEYKKLVRQRKINRLYIRIRERAEDSGNESEGDAEEL-AALGKMGPF---IPGDVNNL- F1.FR.96.96FR_MP411.AJ249238 MSNLYVLSIVAFIIALIIAIVVWTIVFIEYKKLLRQRKINRLYERIRERAEDSGNESEGDAEEL-AALGEMGSF---ISGDINNL- F2.CM.02.02CM_0016BBY.AY371158 MSYLIILVIVAFIVALIAAIIVWTIVYIEYKKQLRQKRINRLYERIRERAEDSGNESEGDAEEL-AALGEVGLF---IPGNINNL- F2.CM.95.95CM_MP255.AJ249236 MPSLLAVGISALIVALIITIIVWTIVYLEYKKLLRQKRINRLYERIRERAEDSGNESEGDAEEL-AALGEVGPF---IPGDINNL- F2.CM.95.95CM_MP257.AJ249237 MSLSLIVVIAAYIVVLILAIIVWTIVYIEYKKILRQKRINRLYERIIERAEDSGNESEGDAEEL-AALGEVGPL---IPGDINNL- F2.CM.97.CM53657.AF377956 MSSLLTIAIVAYIVAILIAIIVWTIVYIEYKKLLRQKRINKLYKRIRERAEDSGNESEGDAEEL-AALGEMGPF---IPGDINNL- G.BE.96.DRCBL.AF084936 MQPLEISAIVGLIVASIAAIVVWTIVFIEYRKIRKQKRIEKLLDRIRERAEDSGNESEGDTEEL-ATLMELGDF---DPWVGDNL- G.KE.93.HH8793_12_1.AF061641 MQSLEISAIVGLIVAFIAAIIVWTIVLVQYREIRKQRKVERLIDRIRERAEDSGNESEGDREEL-ATLMEMGDF---DPWVGDNL- G.NG.92.92NG083.U88826 ---MQALEISXLIVAFIAATIVWSIVFIEYRKIRKQKKIEKLLDRIRERAEDSGNESEGDTEEL-ATLMEMGDF---DPWVGNNL- G.PT.x.PT2695.AY612637 MQSLGIFAIVGLIVAGIAAIIVWIIVFIQYKEIRKQNKIQKILDRIRERAEDSGNESEGDTDEL-ATLVEMGDF---DPWIGDNL- H.BE.93.VI991.AF190127 -MNILGIGIGALVVAFIIAIVVWTIAYIEYRKLXKQRKIDRLIERIRERAEDSGNESDGDTEEL-SKLVEMGHL---NLGYVADL- H.BE.93.VI997.AF190128 -MYIIGIGIGALIVAFIIAIVVWTIVYIEYRKLVKQKKIDRLIQRIIEGAEDSGNESD---EEL-STMVERGHL---TFGYVADL- H.CF.90.056.AF005496 -MYILGLGIGALVVTFIIAVIVWTIVYIEYKKLVRQKKIDRLIERIGERAEDSGNESDGDTEEL-SKLMEMGHL---NLGYVADL- H.GB.00.00GBAC4001.FJ711703 -MYILGIGIGALIVAGILCIIVWTIVYLEYRKLVQQKRIDRLLERIRERAEDSGNESDGDTEEL-STLVERGHL---NLGYVADV- J.CD.97.J_97DC_KTB147.EF614151 MQPLDIAAIVGLIVAVIIAVIVWTIVFIEYRKVLRQRKIDRLINRIRERAEDSGNESDGDTDEL-EKLVEMGPH---DLWNVNDL- J.CM.04.04CMU11421.GU237072 MNSLQIASIVAXVVAFFLAVXAWXVAYIEYXKLVRQRRIDKLXDRIRXREEDSGNDSDGDTEEL-AKLVEMGPH---DLWNVDDL- J.SE.93.SE9280_7887.AF082394 MIPLQIAAIVAFIVAIFLAIGMWTIVYIEYKKLLRQRKIDKLIDRIRERAEDSGNESDGDTEEL-ADLVERGPH---DLWNVNDL- K.CD.97.97ZR_EQTB11.AJ249235 -MVPLTVGIIALVAALILAIIVWTIAYLEYRKVVRQKRINWLFDRIRERAEDSGNESEGDTEEL-AALGETGHL---ILGDINNL- K.CM.96.96CM_MP535.AJ249239 -MVSLAISIVALVVALILAIIVWTIVYIEYRKLVKQKRINWLIDRIRERAEDSGNESEGDAEEL-ADIGELGHL---ILGNIDNL- Group N N.CM.97.YBF106.AJ271370 ---MLWLGFIALGVAIIIAAIIWVLLYKEYKKIKLQEKIEQIRQKIRDRTEDRGKESDGDAEWL-AILLSPDKLD------N.CM.02.DJO0131.AY532635 ---MLEWGFAALGVAIIIAVIIWVLLYKEYKKLKLQEKIEQIRQRIRDRTEDSGSENDGDADWL-AILLSHDK------N.CM.95.YBF30.AJ006022 ---MLSLGFIALGAAVSIAVIVWALLYREYKKIKLQEKIKHIRQRIREREEDSGNESDGDAEWL-DGDEEWLVT---LL------Group O O.SN.99.99SE_MP1300.AJ302647 HHRDLLTLIIISALLLTNVILWAFILRQYLRQKKQDRREREILERLRRIRQIEDDSDYESDGKE-EQEVRDLVH---GYGFDNPM- Group P P.CM.06.U14788.HQ179987 --MHPRDEAVLIIAGVLLLCCIVXWGKVXLLVLKEREXXISLXKGXARWREGQEDEGYESNEEE-EEQLRELGN---LLGFDHVL- P.FR.09.RBF168.GU111555 DEAVLIIAGVLLICIIVVWGXRSCYLCXKRERKGXSLYKGXVRWRERQEDGGYESNEEEEEQLR-ELGNLLGFG---NILXCYXL-

The acquisition of Vpu is thought to have overcome a major hurdle in the transmission from non-human primates to humans (141). Many primate SIVs use the Nef protein to downregulate tetherin (Table 1.4), while some SIV strains, as well as HIV-1, require a Vpu protein (142). Functional SIV Nef proteins are able to antagonize the tetherin of their host species via an extended tetherin N-terminus. This extension is not present in human tetherin, preventing HIV-1 Nef activity, and requiring the function of Vpu for tetherin downregulation.

Table 1.4 SIV antagonists of tetherin. Species SIV Antagonist Chimpanzee Nef Gorilla Nef African Green Nef Blue Monkey Nef Rhesus Nef Moustached Monkey Vpu Mona Monkey Vpu Greater Spot-nose Vpu

1.6 Host Cell Proteins Directly Targeted by Vpu

1.6.1 Non-TM Mediated Targets of Vpu

The Vpu-mediated down regulation of cell surface proteins has been well characterized for two targets, CD4 and tetherin. The mechanism by which Vpu targets these two proteins differs, with Vpu targeting tetherin through its TM domain, specifically an AxxxAxxxA motif that has been shown to be dispensable for the targeting of CD4. In contrast, the factors required for CD4 targeting appear to be entirely present in the cytoplasmic domain of Vpu (143). Mutational analysis of the Vpu cytoplasmic domain showed that the amphipathic helix mediates interaction with CD4 (144), likely through residues 414 to 419 in the CD4 cytoplasmic domain (LSEKKT) (145–149). Both cytoplasmic domains were found to interact in the absence of their membrane domains in a yeast-two hybrid assay, and while one study suggested that Trp 22 in the Vpu transmembrane domain was important in CD4 downregulation, randomization of the Vpu TM domain had no effect on the levels of CD4 and was comparable to wild-type (143).

The downregulation of CD4 from the cell surface is a property that seems to be critical for HIV-1 as it employs two proteins to carry out this function: Vpu and Nef. Unlike Nef, Vpu is translated in the later stages of the infection in a Rev dependent manner, targets newly synthesized CD4 in the ER, and directs its degradation via the proteasome. The biological importance of CD4 downregulation seems to be two-fold: the prevention of superinfection and the prevention of CD4-Env complex formation. Super infected cells show increased levels of apoptosis, so by reducing CD4 levels Vpu prevents premature cell death, and also prevents CD4 from interfering with the production of assembling virus particles (150–155).

More recent studies of Vpu have shown that the Vpu cytoplasmic domain is sufficient for the down regulation of other cellular targets including CD74 and CD1d. Mutational analysis has linked CD1d downregulation to the APW motif on the C-terminus of Vpu, while the binding site on CD74 is suggested to be within residues 54-58 located near the C-terminal end of the Vpu amphipathic helix (156–158).

1.6.2 TM Domain-mediated Targets of Vpu

1.6.2.1 Tetherin

Tetherin (BST-2/CD317) is a type-II, interferon-α induced, cell surface membrane protein that has been shown to restrict the release of several enveloped viruses (114, 159–161), as well as having roles in both Nf-κB signaling (162–164), and structural organization of the cell(165). Tetherin is a comprised of a short cytoplasmic N-terminal tail, a single α-helical transmembrane (TM) domain, an extracellular coiled coil domain, and a C-terminal glycosylphosphatidylinositol (GPI) anchor linking the protein back to the cell membrane (Figure 1.6) (166, 167). The TM domain and GPI moiety act as membrane anchors, with the GPI increasing tetherin localization to lipid rafts, the sites of viral budding (166). The extracellular coiled-coil domain has been implicated in dimerization of tetherin and also contains two N-linked glycosylation sites (N65 and N92) that are important for tetherin trafficking to the cell membrane (166, 168, 169). An elegant study from Perez-Caballero et al supports the concept that the overall configuration and structural features, not primary sequence, are tetherin is paramount for its anti-viral 29

activity (170). Here, an artificial, chimeric tetherin made up of the three required domains (TM, coiled-coil, and GPI anchor) from three heterologous proteins was still able to inhibit virion release, albeit at 1/3 the capacity. Further mutational analysis of tetherin has shown that both membrane anchors are required for anti-HIV function (114, 159, 170, 171), as well as three conserved cysteine residues (C53, C63 and C91) that mediate homo-dimerization of tetherin (168, 170).

Figure 1.6 Tetherin sequence and extracellular domain structure (A) The tertiary structure of the extracellular domain of human tetherin (residues 80-147) solved by x-ray crystallography (PDB 2X7A)(172). The disulfide bridge at C91 is shown in yellow. (B) Domain architecture of tetherin, disulfide bound forming residues C53, C63 and C91 are shown as yellow bars, glycosylation sites N65 and N92 are indicated in red and an endocytic dityrosine motif (Y6 and Y8) is shown in black.

In 2010 two groups reported high resolution crystal structures of the coiled-coil domain of tetherin (172, 173). These structures strongly support a model in which tetherin exists as a parallel disulfide-linked dimer with a degree of conformational flexibility. Coiled-coil interactions paired with disulfide bonds may indicate a dynamic structure within tetherin, in which disassembly and reassembly of the coiled-coil domain occurs during function (172, 173). Two structural components are required for tetherin activity against viruses: the presence of both the TM domain and GPI membrane anchors (114, 159, 170, 171), and the formation of tetherin homodimers mediated by cytoplasmic domain disulfide bonds (168, 170). It is also important to note the evidence for higher order oligomers of tetherin. While in a reduced state, the N-terminal third of the extra-cellular domain has been crystallized as an anti-parallel four helix bundle (173), and super-resolution microscopy has observed the presence of tetherin

clusters at the sites of viral budding (174). The biological relevance of such higher order assemblies has yet to be determined, although mutations in the putative tetramerization motif can abrogate the Nf-κB signaling activity of tetherin.

The proposed mechanism of viral restriction by tetherin is through direct tethering of newly formed virus particles to the surface of the host cell, thus preventing the normal budding process (170, 175). The tethered virions remain linked to the host cell surface by a protease sensitive bridge comprised of one or more tetherin molecules that span the gap between the host plasma membrane and the membrane of the viral particle (170, 176). These tethered virions are subsequently internalized by the cell and degraded (177).

There is evidence that tetherin moonlights in other cellular functions. Over-expression of tetherin or the presence of budding virions has been shown to activate NF-κB(162–164). This is carried out by the recruitment of TRAF1 and or TRAF6 as well as TAK and TAB, activating the canonical NF-κB signaling pathway. Knockdown experiments have also revealed that tetherin may play a role in the maintenance of the actin network (165). A di-tyrosine motif in the cytoplasmic tail of tetherin binds to an adapter protein complex (RICH2/EBP50/Ezrin), linking tetherin to the actin network (178). Here tetherin seems to have multiple possible functions, acting as a membrane anchor for the cytoskeleton and also as a stabilizer of the actin network. As the GPI anchor of tetherin targets the protein to lipid rafts, tetherin may also function to link the cytoskeleton to lipid rafts, thereby playing a role in cell surface organization of lipid microdomains (165).

1.6.2.2 NTB-A

Human NK, B and T cells express the NTB-A antigen on their cell surface. A 60kDa glycoprotein, NTB-A is a co-activating receptor of NK cell lysis. The extracellular portion of NTB- A is made up of 2 immunoglobin (Ig)-like domains (an N-terminal IgV and a more C-terminal IgC2) that contains 8 glycosylation sites, followed by a single α-helical TM domain and a cytoplasmic domain containing 2 phosphotyrosines and 1 phosphoserine site (Figure 1.7).

Figure 1.7 NTB-A homo-dimer structure and domain architecture. (A) Structure of NTB-A homodimer solved by x-ray crystallography (PDB 2IF7) (179). In vivo an NTB-A monomer would be found on both an NK-cell and a target cell, with dimerization being a trigger signal for NK-cell degranulation. (B) The domain architecture of NTB-A.

Rather than hunt for pathogens directly, NK cells survey the cellular landscape for compromised host cells including cancerous and virally infected cells (reviewed in (180)). NK cells destroy their targets through release of cytoplasmic lytic granules (degranulation) that contain perforins and granules. Degranulation events are triggered by receptors on NK cells binding ligands on target cells. The receptor-ligand interactions can be categorized into three types: inhibiting, activating and co-activating. For example the activating receptor NKG2D is capable of recognizes stress-induced molecules such as ULBP-1/2. Activating receptor engagement is a required first step in NK cell mediated lysis, while degranulation is triggered by the simultaneous activation of co-receptors such as homotypic co-activator NTB-A. The NK mediated lysis of HIV-1 infected cells has been shown to be weak. The Vpu mediated downmodulation of NTB-A, as well as PVR (discussed 1.6.2.3 in below), from the surface of infected cells protects HIV-1 infected cells from NK cell mediated lysis (181, 182).

The mechanism of Vpu mediated NTB-A downmodulation is distinct from that observed for CD4 and tetherin. Vpu does not affect the intracellular levels of NTB-A, and does not require the recruitment of β-TrCP to reduce cell surface levels of NTB-A. Thus Vpu may instead interfere with NTB-A trafficking and/or recycling in order to sequester NTB-As in an intracellular compartment. Similar to tetherin antagonism, NTB-A and Vpu physically associate through their TM domains (181). The TM domain of Vpu, specifically the alanine rich face of the TM domain, is a requirement for this association, although the corresponding requirements on the NTB-A TM domain remain unknown.

1.6.2.3 PVR

A relatively recent addition to the list of Vpu targets, PVR is referred to in the literature by several names, due to its role in multiple viral infections as well as its own natural activities. These names include CD155, the poliovirus sensitivity gene, the poliovirus receptor, herpes virus mediator D, nectin-like molecule 5, and tumor associated glycoprotein E4 (183–187). However, the poliovirus receptor (PVR) remains the most widely used name and will be used in this thesis. The identity of the PVR protein was not discovered until 1989 (188) and it was eventually shown to facilitate the binding and entry of the polio virus into cells (189–191). PVR is a transmembrane glycoprotein which is transcribed into a 20kb mRNA containing 8 exons that can be differentially sliced into 4 different isoforms: PVRα, PVRβ, PVRγ, and PVRδ.

The largest PVR isoform, PVRα, contains three N-terminal Ig-like domains, a single helical TM domain, and a cytoplasmic domain. All splice variants have all 3 Ig-like domains (Figure 1.8) (190, 192, 193). The two soluble isoforms of PVR, PVRβ and PVRγ, lack either some or all of the transmembrane domain and are secreted from the cell. These soluble PVR isoforms have been shown to drastically reduce the level of infectivity poliovirus in tissue culture and increased levels of soluble PVR have also been detected in the serum of cancer patients (194). This behavior may be a cancer-specific defense mechanism against the innate immune system. By blocking activating receptors in NK and cytotoxic T cells, increased levels of soluble PVR mask natural receptors and may allow the cancer to evade a cell mediated immune response. A

similar strategy of immune evasion is likely being employed by HIV-1 by Vpu and Nef mediated down-regulation of transmembrane PVR.

Figure 1.8 PVR structure and domain architecture. (A) Structure of PVR extracellular domain solved by x-ray crystallography (PDB 4FQP)(195). (B) Domain architecture of PVR-α.

The transmembrane isoforms of PVR, PVRα and PVRδ, have identical extracellular domains (exons 1 to 5), while alternative splicing results in the last 32 amino acids in the full construct, PVRα, being replaced by 8 different amino acids in PVRδ (189). The additional residues on PVRα encode a tyrosine based inhibitory motif (ITIM) that PVRδ lacks which interacts with a clathrin adapter protein complex subunit, resulting in different subcellular locations of the two isoforms (196).

As its name implies, PVR functions as the entry receptor for the polio virus, although it also plays a role in the pathology of some cancers (194). Expression of PVR, specifically the soluble isoforms, is increased in lung, colorectal, gastric, breast, ovarian, melanoma, and glioblastoma cancers (184, 194, 197–201). This increase in PVR expression also correlates with tumor size and cancer stage (194).

Native activities of PVR include roles in cell adhesion, transendothilium migration (a process in which leukocytes migrate through cell-cell junctions), and innate immunity(202, 203). PVR functions in the innate immune response by interacting with DNAX-Accessory Molecule 1 (DNAM-1/CD 226), a cytotoxic receptor found on a number of lymphocytes including natural killer cells, CD8+ T cells and cytokine induced killer cells (CIKs) (202, 203). 34

DNAM-1 interaction leads to the release of cytokines which results in the apoptosis of the infected target cell. The decrease in PVR expression in HIV-1 infected cells reduces the probability that the infected cell will interact with a DNAM-1+ CTL, thus facilitating viral immune evasion.

1.6.3 Vpu Target Antagonism is Conserved Across Other Viral Species

The innate immune system proteins targeted by Vpu are also antagonized by a wide range of other viruses. This is best illustrated by tetherin, which displays a broad range of antiviral activity towards a diverse group of enveloped viruses (reviewed in (204)). These viruses have in turn evolved measures to counter-act tetherin activity using a variety of different mechanisms (Table 1.5). Although viral antagonism of tetherin has been well documented, cell surface expression of PVR and NTB-A has also been shown to be downregulated during cytomegalovirus infection (205, 206).

Table 1.5 Viral antagonists of tetherin. Virus Viral Antagonist Mechanism Reference HIV-1 Vpu Degradation and (159) Sequestration HIV-2 Env Sequestration (207– 210) Ebola GP Unknown (211, 212) KSHV (HHV-8) K5 Degradation (213, 214) Influenza Neuraminidase (N1 (215, 216) and N2) HSV-1 Glycoprotein M Reduced Surface (217) Expression Chikungunya Virus Non-structural protein Reduced Surface (218) 1 (Nsp1) Expression Equine infectious EIAV Env Reduced Surface (219) anemia virus (EIAV) Expression Sendavirus Fusion Protein (F) and Degradation (220) hemagglutinin- neuraminidase (HN)

1.7 Measuring Membrane Protein Oligomerization

The measurement of membrane protein oligomerization is complicated by several factors: obtaining large amounts of full-length (or even truncated segments of) membrane proteins from recombinant expression is difficult, and the proteins must be studied in a membrane mimetic (detergent micelles, or lipid bilayers). We have chosen to focus our study of Vpu-target oligomerization exclusively to that of the TM domains. This allows for the study of only the contributions of the TM domains to oligomerization, removing any complications soluble portions of the proteins may provide.

1.7.1 Membrane Mimetics

The choice of membrane mimetic used to study a membrane protein is nontrivial. There are some methods that can be used to study membrane protein oligomerization in live cell membranes (see 1.7.2.1). However, in vitro study requires membrane proteins to be solubilized in a membrane mimetic such as detergent micelles or liposomes.

1.7.1.1 Detergent Micelles

Detergent micelles are a more convenient, but less biologically relevant, system than liposomes for the study of membrane proteins. Upon reaching a critical micelle concentration (CMC), detergents aggregate into micelles in which the hydrophobic tails of the detergent face inward leaving the polar head groups in contact with the external aqueous environment. The internal sequestering of detergent hydrocarbon chains in a micelle is reminiscent to the formation of the lipid bilayer. This sequestration provides a hydrophobic environment suitable for TM domains, which adopt α-helical structure to satisfy their amide backbone hydrogen bonding requirements, just as in lipid bilayers. There have been several examples of membrane proteins retaining proper structural folds and oligomeric organization in detergent micelles, including GpA (24). Overall, micellar solutions offer easy handling, have high solubility, and beneficial optical properties make them ideal for spectroscopic studies such as fluorescence, Förster resonance energy transfer (FRET), and circular dichroism.

1.7.1.2 Liposomes

Lipid bilayers provide a more native environment to study membrane proteins but can be more technically demanding than detergent micelles. Rather than forming spherical micelles, upon exposure to an aqueous environment, lipid molecules assemble into bilayers, with the lipid tails pointing inward and the polar headgroups outward. The native membranes in mammalian cells are complex environments which display a wide range of heterogeneity in their lipid make-up, the formation of microdomains, as well as membrane asymmetry. The membranes of different cellular compartments (i.e., organelles) can differ as can the membranes of specialized cell types. Lipid heterogeneity can result in differences in membrane properties including lipid ordering (phase), hydrophobic thickness, charge, and membrane curvature (Figure 1.9).

Figure 1.9 Membrane mimetic properties. (A) A detergent micelle. (B) Membrane hydrophobic core thickness is modulated largely by lipid chain length and unsaturations. (C) Membrane curvature can be induced by discrepancies in the relative size of the lipid headgroup to the acyl chains. Smaller lipid headgroups, such as phosphatidylethanolamine, induce negative curvature, whereas lipids with shorter acyl chains (or lyso lipids) induce positive curvature. (D) The lipid chain spatial arrangement and freedoms of motion result in solid and fluid phases of membrane bilayers. Different phases result from different lipid chain lengths and degrees of unsaturation as well as the presence of sterols. Other phases not shown, such as hexagonal and cubic phases, are also thought to be present in cells but only during transient events such as membrane fission and fusion.

The use of native bilayers for study is largely limited to in vivo experiments, leaving high resolution structural studies to more artificial lipid environments like liposomes. Liposomes more strongly represent the native membrane environment than micelles by having a more defined hydrophobic interface, lateral pressure, and can be customized by the enrichment of particular lipid types. There are aspects of in vivo lipid bilayers that are easily reproduced (liposome size, unilamellar state, lipid make up), although some aspects are more difficult to replicate in vitro (lipid asymmetry and absolute protein orientation).

1.7.2 In vivo Methods for Measuring TM Helix-Helix Interactions

There are several approaches which have been used to allow measurement of TM helix associations in the native inner membrane of E. coli. In comparison to eukaryotic cells, a system developed in bacterial cells has several advantages, including a shorter experimental timeframe and better overall accessibility and reproducibility. However, the chemical makeup and asymmetry of the bacterial inner membrane varies significantly from that of eukaryotic membranes, and therefore, the results of any experiments on eukaryotic membrane proteins must be considered carefully. Most bacterial systems involve the fusion of the TM domain(s) of interest to a DNA binding domain which dimerizes in the E. coli cytoplasm upon association of the fused TM domains. The dimeric DNA binding domain subsequently binds the promoter of a reporter gene. These assays are useful in that the level of expression of the reporter gene can be linked to the affinity of the TM oligomers. Additionally, the assays facilitate the use of mutagenesis to determine faces of interaction as they will be sensitive to differences in affinity arising from mutant sequences. Methods with various features and reporter genes have been developed, including ToxR, TOXCAT, POSSYCCAT, GALLEX and BATCH (221–226). A general overview of their general features, along with their advantages and disadvantages, are discussed below.

1.7.2.1 The ToxR in vivo Based Systems: ToxR, TOXCAT and POSSYCCAT

The first system to monitor TM associations in E. coli membranes was the ToxR system developed by Langosch and coworkers (221). Here, a TM of interest is fused to the DNA binding domain of the ToxR protein from Vibrio cholera. ToxR is an integral membrane protein made up of an N-terminal cytoplasmic DNA binding domain, a single TM domain and periplasmic C- terminal domain. The periplasmic domain dimerizes as a response to external stimuli, this results in the dimerization of the DNA binding domain in the cytoplasm, which then binds and activates ctx promoters in the cell, inducing virulence factors. The TM and periplasmic domains are not necessarily required for ctx promoter activation and thus in the ToxR system they have been replaced in chimeric proteins used in the assay (221). The periplasmic domain is replaced by the maltose binding protein MalE, and the native TM by the TM of interest to be studied. 39

The MalE binding protein has been shown to facilitate membrane insertion and also allows overall protein detection as well as determination of protein membrane topology (227). Since MalE is monomeric, only dimerization of the TM domains can cause the dimerization of the DNA binding domain and activation of the reporter gene lacZ (encoding β-galactosidase) fused to a ctx promoter. The level of β-galactosidase activity present in the cell lysate is proportional to the strength of the TM association being probed as stronger TM domain associations result in higher expression levels of the reporter. Variants of the ToxR system have been developed to express the ToxR fusion proteins under arabinose inducible promoters rather than the native ToxR promoter which is constitutively active, allowing for controlled expression (228–232).

Engelman and co-workers later modified the ToxR system to use cat rather than lacZ as a reporter gene to create the TOXCAT system (222). The cat reporter gene here encodes for a chloramphenicol-acetyl-transferase protein fused to a ctx promoter. Interaction of TM domains in this system will result in the expression of the CAT protein, which confers resistance to the antibiotic. Increased TM domain interaction results in increased CAT expression and enables E. coli cells to survive in increasing chloramphenicol concentrations. This method of observation is much less labour intensive than testing of cell lysates for β-galactosidase activity, as used in the classical ToxR system. The final ToxR based system, POSSYCCAT, is closely related to TOXCAT, with the exception that the reporter gene is located on the bacterial genome rather than a plasmid (223).

Although originally designed to probe for homotypic interactions, the ToxR-based systems have also been used to measure the association of two different TMs in systems where formation of any homooligomers is significantly weaker than heterooligomerization (25). Additionally, a ToxR-dominate-negative system, in which two different TMs are expressed, one fused with wild type ToxR DNA binding domain and the other with an inactive ToxR S87H, has been used to probe TM domain heterooligomerization (233).

1.7.2.2 The GALLEX System

The major difference between the GALLEX system and the ToxR-based systems is that the ToxR DNA binding domain has been replaced by the LexA DNA binding domain, a transcription repressor involved in the bacterial SOS response (224). The dimeric wild type LexA protein recognizes the op promoter. While a triple mutant of LexA (LexA 408) recognizes an altered promoter sequence (op408). Therefore, a heterodimer composed of one wild type and one mutant LexA will activate the op/op408 promoter. Thus, GALLEX can be used to monitor homo and heterooligomeric TM domain interactions.

1.7.2.3 The BATCH System

A newer method for probing TM interactions is the bacterial adenylate cyclase two hybrid (BATCH) system, which is based on the restoration of adenylate cyclase activity(225, 226). Here the adenylate cyclase enzyme from Bordetella pertussis is split into two fragments, T18 and T25, and fused to TM segment(s) of interest and expressed in E. coli. Enzyme activity is restored when the fused TMs, and the T18/T25 fragments, dimerize. This results in the production of cyclic AMP (cAMP) which then complexes the activator protein CAP, which can control various reporter genes (i.e. lacZ). Unlike the ToxR-based and GALLEX systems described above, the adenylate cyclase fragments can be genetically fused to either terminus of a TM helix of interest. Therefore, it is possible for the BATCH method to test for the formation of homo and heterooligomerization in either parallel or anti-parallel orientations.

The ToxR experiment, its derivatives, and later genetic methods were designed to assess the extent of dimerization of membrane proteins in E. coli biological membranes. The various assays involve the conjugation of TM domains of interest to a transcriptional activator that is only functional in a dimeric from (i.e. ToxR, LexA). Dimerization of the TM domains results in oligomerization of the transcriptional activator and subsequent activation of the reporter gene. These assays are useful in that the level of expression of the reporter gene can be linked to the affinity of the TM oligomers. Additionally, the assays facilitate the use of mutagenesis to

determine faces of oligomerization as they will be sensitive to differences in affinity arising from mutant sequences.

1.7.3 Monitoring TM Helix-Helix Interactions In vitro

1.7.3.1 SDS-PAGE

Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS PAGE) is a ubiquitous biochemical technique used to determine the molecular weight and purity of protein samples (234). SDS-PAGE analysis is relatively simple, fast, affordable, and requires small amounts of sample. During analysis, a protein sample is diluted in buffer containing a running dye and SDS and is loaded at the top of a gel of cross-linked acrylamide fibers. Detergent molecules bind to hydrophobic segments of protein chains, resulting in the formation of an anionic protein/detergent complex. A potential difference is applied across the gel to drive the migration of particles along its length. Several factors influence the rate of migration on SDS PAGE: the size and shape of the protein/detergent particle, its net charge, the voltage across the gel, and the acrylamide concentration. The meshwork of the gel creates spaces accessible to proteins in a network of fibers which proteins to diffuse around/through to continue migrating down the gel. Thus larger proteins migrate slower than smaller proteins that can more easily diffuse through the gel mesh.

When solubilized in SDS micelles, globular proteins become denatured and lose their overall shape, as a result of a combination of hydrophobic interaction with the detergent and detergent-detergent charge repulsion. This reduces structural differences, and promotes binding of SDS in a protein mass dependent manner, such that the rate of migration becomes dependent mostly on the protein molecular weight. The migration of membrane proteins and TM domain peptides however, remains more unpredictable. Membrane proteins can display reduced, equivalent or increased rates of migration when compared to globular protein standards (235, 236).

This anomalous behavior of membrane protein migration can be explained by two linked phenomena: the retention of protein structure, and the amount of bound detergent in the protein/detergent complex. Firstly, SDS micelles are sufficiently similar to a native membrane that many integral membrane proteins, and especially isolated TM domains, retain their secondary structure and, in some cases, tertiary and quaternary contacts within the micelle. The shape of the protein then influences migration through the PAGE gel. Secondly, the relatively high hydrophobicity of membrane proteins results in the binding of more detergent molecules in the protein/detergent complex. However, retention of secondary/tertiary structure can affect the amount of bound SDS; the bacterial protein OmpA binds less SDS as a folded β-barrel then as a denatured species (237, 238). In general, retention of compact structure results in faster apparent migration, while the binding of more SDS than soluble proteins results in slower apparent migration on SDS-PAGE. The trade-offs of the two phenomena are described by Rath and Deber (235, 239). The strength of SDS-PAGE is in the analysis of higher affinity oligomers, as TM domains and membrane proteins may retain quaternary structure in SDS micelles, giving rise to discrete bands of monomer, dimer, and other oligomers (240, 241). In some cases, weaker associations between TM domains can manifest as a smeared band ranging in size from oligomer to monomer.

1.7.3.2 Sedimentation Equilibrium by Analytical Ultracentrifugation

Analytical ultracentrifugation (AUC) is used to analyze the oligomerization of membrane proteins solubilized in detergents through the use of sedimentation equilibrium experiments. Here, a sample is spun in an ultracentrifuge using a centrifugal force small enough so that macromolecules in solution reach an equilibrium where the movement of macromolecules due to sedimentation equals the movement due to diffusion. Thus, the migration of proteins becomes dependent on the molecular weight only. This method provides a very accurate way of determining the mass of the macromolecule being studied, and consequently, its oligomeric state.

When carrying out sedimentation equilibrium experiments on a membrane protein, the sedimenting particle will have contributions from the protein itself, as well as any bound lipids or detergents. Since monomeric and oligomeric forms of membrane proteins are likely to bind different amounts of detergent, steps must be taken to reduce the contributions of the detergent to the “apparent mass” of the particle. To ensure that the bound detergent molecules do not contribute to the molecular mass of the complex, solution conditions are adjusted to render the bound detergent gravitationally invisible. This is carried out by adjusting the density of the solvent to the point where it equals the density of the bound detergent. This is most easily accomplished by using the appropriate mixture of light (H2O) and heavy (D2O or

18 D2 O) water; however, more dense detergents may require the addition of agents such as sucrose or glycerol (242, 243). Several classes of neutrally buoyant detergents have been identified that are easily density matched, and that are frequently used for sedimentation equilibrium studies of TM proteins (reviewed in (244)). The requirement to use certain detergents can prevent reconstitution of many proteins in their active or correctly folded states, providing a limit for the use of this technique.

1.7.3.3 Förster Resonance Energy Transfer (FRET)

Förster’s initial discovery of the FRET phenomenon in the 1940s led to the use of FRET in the study of the assembly of soluble proteins and has since been applied to the study of membrane spanning proteins and peptides in vitro and in vivo (245). The energy transfer in FRET can arise through fluorescence or phosphorescence, with the former being more widely employed in biochemical studies. The process of fluorescence begins when a molecule absorbs a photon of the appropriate wavelength to create a transient singlet excited state. Herein, an orbital electron is excited from its ground state, from which there are several competing pathways for relaxation back to the ground state. The excited molecule can under-go non radiative decay, losing energy through molecular vibrations (heat) or through collisional transfer to a fluorescence quencher (such as O2 or acrylamide). Alternatively, the excited state can relax to a vibronic ground state through the emission of a photon (fluorescence) or transfer to a triplet state and undergo subsequent photon emission (phosphorescence). In most cases,

the excited state will exist for a given amount of time (the fluorescence lifetime), during which some energy will be lost to vibrational relaxation. Thus the final emitted photon will be of lower energy and have a longer wavelength than the photon used for excitation. The overall process of fluorescence can be illustrated using a Jablonski diagram (Figure 1.10).

Figure 1.10 Jablonski diagram. The Jablonski diagram illustrates possible vibrational and electronic transitions that occur during fluorescence.

FRET occurs when an excited donor fluorophore is quenched by non-radiative energy transfer to an acceptor chromophore that is in close proximity. The acceptor chromophore does not necessarily have to be fluorescent but there does have to be significant overlap in the emission spectrum of the donor and the absorption spectrum of the acceptor (Figure 1.11).

Figure 1.11 Model FRET pair donor and acceptor emission/excitation spectra. The overlap of donor emission (solid green) and acceptor excitation (dashed green) results in FRET. A good FRET pair will have a donor with a high quantum yield and a large spectral overlap between donor emission and acceptor excitation.

If the acceptor molecule is not fluorescent, FRET must be detected by the loss of donor fluorescence (donor quenching). However, if the acceptor is fluorescent, FRET can be measured either by using donor quenching or acceptor sensitized emission. The range of proximity for FRET is specific to each individual FRET pair, commonly on the order of 10 to 100 Å, and is dependent on a distance parameter known as the Förster radius (R0), defined as the distance between the donor and acceptor molecules at which the efficiency of transfer (i.e. the FRET efficiency) is 50%:

6 푅0 퐹푅퐸푇 퐸푓푓 = [ ] , (푅0 + 푟) where r is the distance between the donor and acceptor molecules. The Förster radius can also be defined by:

−4 2 1/6 3 푅0 = [(푄0퐽푛 퐾 ) × 9.7 × 10 ]

where Q0 is the quantum yield (the ratio of absorbed photons to emitted photons) of the donor fluorophore, J is the spectral overlap integral, n is the refractive index of the medium and K2 is a

dipole-dipole orientation factor. In most TM-based FRET experiments the orientation factor will be 2/3, a value consistent with the free rotation of a FRET label conjugated to the terminus of a TM peptide.

FRET can be used in straightforward experiments to qualitatively show protein/peptide oligomerization. Rigorous analysis of FRET data can also be used to determine oligomer size and even association energetics. In 1994, Adair and Engelman derived a formalism to distinguish between various oligomeric states and showed that GpA is dimeric in lipid membranes (13). Their approach, along with others, has been applied to many biological systems including extensive work on the GpA dimer as a model system (241, 246, 247). This work has been expanded upon by Hristova and co-workers, who have made significant contributions to the methodologies and formalisms used for interpretation of membrane:peptide FRET in their studies of receptor tyrosine kinases (248–250). Similarly, the Deber group has used natural and model systems in FRET studies to explore the underlying mechanisms of membrane protein folding (241, 251, 252). Table 1.6 highlights the work of these researchers and others in the development of FRET as a tool to measure TM helix associations.

Table 1.6 Studies of TM-TM helix association using FRET. Peptide System Oligomer Type Lipid Mimetic(s) Donor Acceptor FRET Method Reference Glycophorin A Homodimer DPC, SDS, and DDMAB Pyrene 7-dimethylaminocoumarin Acceptor sensitization (253) micelles emission Glycophorin A Homodimer C12 sulfate, C12 DMAB, Pyrene 7-dimethylaminocoumarin Acceptor sensitization (254) C10, C11 maltoside emission C10, C11, C12 DAO Glycophorin A Homodimer DMPC vesicles Dansyl Dabsyl Donor quenching (13) Phospholamban Higher order SDS, C12E8, and OG AMCA Dabsyl Donor quenching (15) oligomer micelles. DOPC vesicles Dansyl Dabsyl HER1, 2, 3, and 4 Homo and LDAO micelles Pyrene 7-dimethylaminocoumarin Acceptor sensitization (255) heterodimers emission FGFR3 and Homo and POPC vesicles Fluorescein Rhodamine Donor quenching (248, mutants heterodimers 249) FGFR3 Homodimer POPC vesicles Fluorescein rhodamine EmEx method (256) Ff Bacteriophage Homodimer SDS micelles Dansyl Dabsyl Donor quenching (251) Major Coat Protein FtsB and FtsL Homo and DPC micelles 7- FAM/FITC Donor quenching (257) hetero dimers POPC vesicles Hydroxycoumarin M2 Homotetramer DLPC and POPC BODIPY-fluorescein Rhodamine Donor quenching (14) Liposomes WALP23 model Homodimer C18:1PC, C14:1PC, Tryptophan Pyrene Donor quenching (258) peptides C22PC liposomes

In the simplest case, FRET data can be used to qualitatively show the oligomerization of two or more TM helices. However, a more robust analysis of FRET data can allow for the calculation of the association energetics of the studied system. This is because observed FRET efficiency is proportional to the amount of oligomeric species being formed. Several formalisms exist for calculation of the FRET efficiency: donor quenching and acceptor sensitized emission (Figure 1.12), as well as the EmEx method.

Figure 1.12 Quantifying FRET transfer. FRET and FRET efficiency can be monitored by (A) donor quenching or (B) acceptor excitation sensitization. The solid curves represent samples containing only donor labeled peptides while dashed curves represent samples containing increasing amounts of acceptor labeled peptides. Observing FRET by donor quenching is done by monitoring the decrease in donor emission while observation by acceptor excitation sensitization monitors the increase in acceptor emission.

Donor quenching is the most widely used and most straightforward method for analysis of experimental FRET data. In this approach, FRET efficiency is calculated by:

푭(흀푫) 푬 = ퟏ − ( ), 푭ퟎ(흀푫)

where F is the fluorescence of the donor in the presence of the acceptor and F0 is the fluorescence of the donor in the absence of the acceptor. The fluorescence emission can be measured either as the intensity at a given wavelength or as an integration across a range of wavelengths (λD). The acceptor sensitization method can result in easier detection of energy transfer using acceptors with high quantum yields, but requires characterization of the donor

and acceptor extinction coefficients in the solvent. The calculation of the FRET efficiency from acceptor sensitized emission is thus more complicated:

퐹(휆2) 휀퐴(휆2) 휀퐴(휆1) 퐸 = [ − ] × . 퐹(휆1) 휀퐴(휆1) 휀퐷(휆2)

Here the acceptor excitation fluorescence (F) is measured at two wavelengths, (λ1) where the donor has minimal absorption and (λ2) where the extinction coefficient of the donor (εD) is large relative to the extinction coefficient of the acceptor (εA). The EmEx method factors in precise measurements of the donor and acceptor concentrations in the sample, as well as the quantum yields of the donor and acceptor in its calculations. This significantly reduces experimental uncertainties. Overall, donor quenching is the most practical method to execute but lacks the elegant precision of the EmEx method needed for more in depth studies on TM association energetics.

Regardless of the methods used to analyze FRET data (donor quenching, acceptor sensitization, EmEx) various parameters can be calculated depending on the experimental conditions used. A titration with an increasing concentrations of acceptor can be used to determine the oligomeric state of the complex being formed. A linear response to an acceptor titration is indicative of dimer formation whereas a higher order oligomers result in a non-linear response (13). The strength of interaction of a TM oligomer can be quantified by measuring the FRET between donor and acceptor labeled constructs in decreasing peptide to lipid ratios (248, 259).

1.8 Thesis and Hypothesis

The bulk of research on Vpu has been bimodal, with researchers focusing on Vpu as either a homooligomeric ion channel, or as a monomeric transmembrane protein responsible for binding and facilitating the downregulation of cell surface proteins. As a result, current interpretations of Vpu mechanisms of action remain incomplete This thesis will treat Vpu as both a protein capable of both homo and heterooligomerization and provide a more thorough explanation of the promiscuous binding within the transmembrane Vpu-target system as well 50

as provide a greater understanding of Vpu’s multifunctionality. Developing our understanding of how Vpu-target helix-helix interactions are mediated in the lipid bilayer will provide information on the mechanism of HIV-1 viral immune evasion, and provide new insights into the fundamental biophysics underlying TM domain assembly and membrane protein folding.

Chapter 2 examines the homodimerization of the TM domain of the Vpu target tetherin. We found that, even in the absence of the coiled-coil cytoplasmic domain, the tetherin TM domain was able to dimerize in various membrane mimetics. This result may describe a step along the folding pathway of full length tetherin, which forms disulfide linked dimers via its coiled-coiled domain. The dimerization of the coiled-coiled domain is thought to provide tensile strength to the virion-cell membrane bridge and the dimerization of the TM domain may also perform a similar function. Similarly, the propensity of the tetherin TM domain to form homodimers has implications for the function of this protein in vivo, and for the interactions of the TM domain with HIV-1 Vpu.

In Chapter 3 we examine the promiscuous yet specific interactions of Vpu and its transmembrane targets tetherin, NTB-A and PVR. We confirm our methods by measuring the homooligomerization of both Vpu and tetherin, while demonstrating that NTB-A and PVR TM domains do not form homooligomers. The tetherin homooligomer was found to be dimeric, consistent with our Chapter 2 results, while the results for the Vpu TM domain are consistent with recent reports that this protein may exist as a mixture of several oligomeric states. Nevertheless, our steady-state FRET measurements suggest an averaged size close to a trimer. With the homooligomerization of tetherin and Vpu in mind, we were able to measure the binding of Vpu to all three target TM domains, confirm a 1:1 binding ratio (possibly 1:2 for tetherin) in each case, and show that all targets compete with each other for interaction with Vpu. Furthermore we calculated the relative binding affinity for each interaction in the system. The free energy of association of each helix-helix interaction is relatively weak, and there appears to be no preferred Vpu interaction partner, suggesting that conservation of the ability to form multiple weak interactions is an important aspect of Vpu structure and function.

We expand on this work in Chapter 4 by exploring the effects of the lipid bilayer make up on the interactions within the Vpu-target system. We show that temperature does not have a profound effect on the helix-helix associations within the system unless there is an accompanying lipid phase transition, in which case the gel phase seemed to increase the strength of all helix-helix associations. Interactions between Vpu and its target TM domains were modulated by variations in lipid chain length (hydrophobic thickness of the bilayer). The variation in membrane hydrophobic thickness also had little effect on secondary structure, as determined by CD.

Chapter 2: Dimerization of the Transmembrane Domain of Human Tetherin in Membrane Mimetic Environments.

References:

Sections of this chapter were originally published from Cole, G., Simonetti K., Ademi, I., and Sharpe, S. Dimerization of the transmembrane domain of human tetherin in membrane mimetic environments. Biochemistry. 2012. 51: 5033–40 and reprinted with permission. Copyright 2012 American Chemical Society.

Contributions:

The work is this Chapter was aided by Karen Simoneti and Irsa Ademi who contributed to the initial project design as well development of peptide synthesis and purification protocols. KS also permorfmed pyrene fluorescence and SDS PAGE experiments.

2.1 Introduction

It has been proposed that the direct association of HIV-1 Vpu and human tetherin leads to the subsequent internalization and ubiquitin-mediated degradation of tetherin (130–132, 260–262). This interaction is mediated between the TM domains of the two proteins. The importance of the tetherin TM domain for recognition by Vpu has been demonstrated in mutagenesis experiments where the alteration of several residues in tetherin can rescue tetherin from Vpu antagonism. In particular, either mutation of T45 or deletion of both G25 and I26 is capable of impeding virion release (127, 263). Likewise, a native Vpu TM domain is required for downregulation of tetherin. When these data are combined with fluorescence microscopy showing colocalization of Vpu and tetherin within internal membranes, they suggest a direct interaction of these two proteins, mediated by their TM domains, and resulting in a reduction of cell surface tetherin and enhanced virus release (127, 128, 130, 171, 263–265). This has recently been supported by a combined biophysical and cell biology study in which NMR was used to demonstrate the formation of transmembrane Vpu−tetherin heterodimers (266).

To obtain further insight into the membrane bound architecture of tetherin, the physical mechanisms underlying viral tethering, and the interplay of this protein with Vpu, it is important to extend structural studies of tetherin to include its TM domain. In the present study we show that a peptide corresponding to the TM domain of human tetherin, TethTM, forms stable homodimers in membrane mimetic environments, including detergent micelles and phospholipid liposomes. By performing SDS-PAGE and gel filtration under both reducing and non-reducing conditions, we show that dimer formation is independent of disulfide formation. Furthermore, the use of pyrene (Pyr) excimer fluorescence provides evidence that dimers are formed in a head-to-head orientation. These data support a parallel arrangement of tetherin dimers, as suggested by recent crystal structures of the coiled coil domains, with implications for the architecture of the tether formed between the host cell and viral membranes (172, 173).

2.2 Materials and Methods

2.2.1 Peptide Synthesis, Purification, and Labeling

The putative transmembrane segment for human tetherin was identified using the program TM-Finder (The Hospital for Sick Children, Toronto, Ontario) (267). Based on this analysis TethTM, a tetherin TM domain peptide corresponding to residues 19−49 (Table 2.1) of the full length human protein, was synthesized via automated solid phase Fmoc chemistry using PalPEG resin (Applied Biosystems). In order to minimize hydrophobic interactions and prevent aggregation or truncation, 50 mg of lithium chloride was added to each of the synthesis vessels during the coupling of residues 19-46. Two tags were introduced to facilitate solubility and aid in purification, as described by Melnyk et al (251, 268). A polar tag comprised of five histidines was added to the N-terminus, and the C-terminus was capped with four lysines. A scrambled tetherin TM domain sequence, Teth(Scr), was synthesized as a negative control, while a peptide containing the TM domain of glycophorin A (GpATM) was synthesized as a positive control for dimer formation for use in SEC and SDS-PAGE experiments (Table 2.1). GpA has been shown to form SDS-resistant dimers through specific transmembrane-domain interactions (269, 270).

Table 2.1 Primary structures of the TM peptides. N- and C-terminal solubility tags added as shown. The predicted transmembrane domain of tetherin, as determined by TM Finder, is underlined. Peptide Sequence TethTM HHHHH19RCKLLLGIGILVLLIIVILGVPLIIFTIKAN49KKKK TethScr HHHHHRCKTGLLVFLVIIPIIGVILLLILIILGKANKKKK GpATM HHHH70EPEITLIIFGVMAGVIGTILLISYGIRRLI99KKKK

Peptide concentration was estimated in 2,2,2 trifluoroethanol (TFE) using a Bradford assay. Purification was achieved by reverse-phase high performance liquid chromatography (HPLC) on an 11 × 300 mm C4 column (Vydac), using a 10 to 100% gradient of HPLC solvent A (10% acetonitrile, 0.1% TFA) to solvent B (90% acetonitrile, 0.1% TFA). HPLC fractions were lyophilized and peptide identity was confirmed by MALDI mass spectrometry at the APTC (The Advanced Protein Technology Centre, The Hospital for Sick Children, Toronto, ON). Labeling with pyrene−maleimide (Invitrogen) was achieved by reconstitution of TethTM in TFE, followed

by the addition of an equal volume of 100 mM Tris pH 7.5, giving a final protein concentration of 4 mg/mL. Tris(2-carboxyethyl)-phosphine (TCEP) was added at a final concentration of 3mM. A 20 mM stock of pyrene−maleimide was prepared in DMF, and a total of 250 μL added for each 1 mg of protein to be labeled as follows: 100 μL additions of dye were done dropwise every half an hour, and the reaction was then topped off with nitrogen gas and stirred in the dark; these steps were repeated up to complete addition of the desired volume. The reaction was then allowed to stir overnight while in the dark. Upon completion of incubation, quenching was achieved by the addition of β-mercaptoethanol, followed by purification of labeled peptide on a C4 column as described for unlabeled peptide. Peptide labeling was confirmed by mass spectrometry. The degree of labeling was estimated based on the molar extinction coefficient

−1 −1 (ε) of the dye (40 000 M cm )(271), the absorbance of pyrene at 338 nm (A338), and protein mass (4541.7 Da).

퐴338 푀푊 표푓 푝푒푝푡𝑖푑푒 푚표푙푒푠 표푓 푑푦푒 푋 = ε 푚푔 표푓 푝푟표푡푒𝑖푛/푚퐿 푚표푙푒푠 표푓 푝푟표푡푒𝑖푛

2.2.2 Tetherin TM Peptide Incorporation into Liposomes

Lyophilized TethTM peptide was dissolved in a 2:1 (v/v) mixture of methanol:TFE. Palmitoyl-2-oleoylphosphatidylcholine (POPC) in chloroform was added to the peptide, giving a final protein concentration of 4 mol % relative to phospholipid. The mixture was incubated to ensure complete dissolution and then dried under N2(g). To ensure complete removal of solvent, the sample was resuspended in water, frozen, and lyophilized. The final peptide-containing liposomes were reconstituted in 50 mM Tris, pH 7.5, followed by four cycles of freeze−thawing. To reduce light scattering in spectroscopic measurements, SUVs (small unilamellar vesicles) were produced by 12 cycles of sonication, 1 s each, with sample cooling. The resulting SUV samples were clarified via centrifugation to remove any residual large liposomes prior to experiments.

2.2.3 SDS-PAGE Analysis

A stock solution of peptide was made at 1 mg/mL in Laemmli buffer (62.5 mM Tris-HCl, pH 6.8, 25% glycerol, 2% SDS, 0.01% Bromophenol Blue). Subsequent dilutions for SDS-PAGE were done in Laemmli buffer. For samples containing 5% SDS, a stock of 10% SDS was titrated into the sample in order to achieve the desired detergent concentration. Samples were heated at 45 °C for 20 min and then loaded onto a 12% Bis-Tris gel (Invitrogen) and electrophoresis performed at 175V. PinkPlus (GeneDirex), Precision Plus (BioRad), andMark12 (Invitrogen) molecular weight markers were used.

2.2.4 Circular Dichroism

All CD measurements were performed with a Jasco J-810 spectropolarimeter, using a 1 mm path length quartz cuvette (Hellma Analytics). Spectra were measured from 190 to 250 nm at a scan rate of 100 nm/min and were recorded as an average of three measurements. CD spectra were recorded for samples containing 55 μM peptide in TFE, 1% sodium dodecyl sulfate (SDS), or 1-palmitoyl-2-oleoylphosphatidylcholine (POPC) (at a 4 mol% peptide relative to lipid).

2.2.5 Fluorescence Spectroscopy

Pyrene fluorescence emission measurements were performed on a Photon Technology International C60 spectrofluorimeter scanning emission from 370 and 550 nm and using an excitation wavelength of 340 nm. Each emission spectrum was recorded as an average of three scans, acquired with integration set to 0.3 s, at a step size of 1 nm, and excitation and emission slit widths of 5 nm. Experimental values were corrected for solvent background using appropriate blanks, and the area under the monomer (371−440 nm) and excimer (441−549 nm) peaks was calculated. These values were then used to calculate the excimer-to-monomer ratios, which was corrected based on the degree of TethTM labeling obtained as described above.

2.2.6 Size exclusion chromatography

Size exclusion chromatography (SEC) samples were prepared by dissolving lyophilized protein in TFE to create a protein stock. Concentration was estimated using a Bradford assay. TFE was evaporated from a volume containing 1 mg of protein, which was then reconstituted in 1 mL of running buffer (10 mM Tris pH 7.5, 1% SDS) to create a 1 mg/mL stock. Final samples were then made by serial dilution of the 1 mg/mL stock. Reduced samples were prepared by incubating overnight in running buffer containing 1 mM DTT or 1 mM TCEP. SEC was performed on a Phenomenex Biosep-SEC-S-2000 column. The column was equilibrated with 3 column volumes of running buffer. Samples were loaded at a volume of 100 μL, and the column was run at a flow rate of 1 mL/min. Protein was detected based on UV absorbance at 214 nm. Calibration of the SEC-2000S column was carried out using a low molecular weight gel filtration calibration kit purchased from GE Healthcare, which contained the following standards: Blue dextran (>2000 kDa), used for void volume; conalbumin (75 kDa); ovalbumin (43 kDa); carbonic anhydrase (29 kDa); ribonuclease A (13.7 kDa); aprotinin (6.5 kDa). Insulin chain B(3.5 kDa from Sigma) was also used to provide a lower molecular weight calibration. Each protein standard was dissolved in 10 mM Tris buffer (pH 7.5) containing 0.3% SDS and 5% β-mercaptoethanol, and boiled for 5 min, then incubated overnight at room temperature. After incubation, 0.1mg (100 μL) of each was applied individually to the equilibrated SEC-2000S column. A calibration curve was created by plotting the Kav (gel-phase distribution coefficient) versus log(MW) for each protein. Kav was calculated using the equation

퐾푎푣 = (푉푒 − 푉0)/(푉푐 − 푉0) where Ve is the elution volume of a protein, V0 is the column void volume, and Vc is the geometric column volume. Three replicate runs were used to obtain each data point.

2.3 Results

2.3.1 Secondary Structure of TethTM

While the C-terminal coiled-coil domain of human tetherin has been shown to form dimers, little is known about the potential role of the TM domain in oligomerization of this protein. In order to provide further insight into the modes of assembly accessible to tetherin in cellular membranes, the ability of its TM domain to self-assemble in the absence of the coiled- coil domain was examined. Further motivation for examining the structure and assembly of the TM domain alone has been provided by several studies that demonstrated a critical role for this region in mediating interactions with the HIV-1 accessory protein Vpu (127, 128, 130, 263, 266). The synthetic peptide TethTM, containing the TM domain of human tetherin (Table 2.1), was incorporated into several membrane mimetic environments. CD spectra recorded for this peptide in TFE solution, SDS micelles, and POPC liposomes all show the characteristic minima at 208 and 222 nm, indicating that the peptide adopts a predominantly helical structure in all cases (Figure 2.1)

Figure 2.1 Secondary structure of TethTM in membrane mimetic environments. CD spectra are shown for TethTM at 55 μM dissolved in TFE, incorporated into SDS micelles, or reconstituted into POPC liposomes. The CD spectrum for TethTM labeled at Cys20 with pyrene (Pyr) maleimide is also shown. For samples in POPC, the peptide to lipid ratio was 4 mol %. All spectra were recorded at 25 °C and are an average of three scans.

The molar ellipticity at these wavelengths is larger in TFE. This solvent is known to induce helix formation (272), and thus it is likely that the helical structure extends closer to the ends of the TethTM peptide under these conditions than in micellar or liposomal environments. These results demonstrate that TethTM is likely to adopt a native conformation in SDS or POPC, validating the use of the detergents as suitable membrane mimetics for studying the oligomerization of this peptide.

2.3.2 Oligomerization of TethTM in SDS Micelles

SDS-PAGE analysis was used as an initial screen for oligomerization of the tetherin TM domain. As shown in Figure 2.2A, B, the mobility of TethTM on SDS-PAGE gels strongly suggests the formation of dimers by this peptide in SDS. As has been previously reported for several

transmembrane peptides (273), TethTM migrates as a relatively diffuse band, although this is mitigated at lower concentrations and in the presence of higher concentrations of SDS. For instance, the silver stained gel in Figure 2.2B shows a relatively narrow band of approximately 8−10 kDa molecular weight for this peptide at 50 μg/mL in 5% SDS which is twice the expected monomer weight of 4.5 kDa.

Figure 2.2 SDS-PAGE analysis of TethTM. (A) SDS-PAGE of peptide loaded at indicated concentrations in 2 or 5% SDS. Stained with Coomassie Blue. The arrow points to the expected migration of the monomer, which is in equilibrium with the oligomer. Three different sets of molecular weight markers are shown to illustrate variations in estimated migration of similarly sized proteins under the same conditions. All lanes, including the markers, were run on the same gel and have only been cut and moved for clarity. (B) SDS-PAGE gel run under the same conditions but stained with silver stain.

Determination of molecular weights using SDS-PAGE is not necessarily accurate, especially at the lower mass ranges required for this study. As can be seen in Figure 2.2, three sets of molecular weight standards exhibit different relative mobilities for the lower molecular weight proteins, highlighting this effect. Similarly, it has been reported that hydrophobic proteins, especially those that retain helical structure in detergent micelles, do not bind quantitatively the same amount of SDS as denatured globular proteins (235). This can lead to significant deviation from ideal electrophoretic separation. Therefore, SEC was used to confirm the oligomerization of TethTM in SDS micelles (Figure 2.3). As shown in Figure 2.3A, TethTM in SDS micelles migrates as a single peak on an S-2000 column at 100 µg/ml. Lowering the peptide

concentration, to the lower limit of detection for this method (25 µg/ml), results in the appearance of a TethTM monomer peak that manifests as a shoulder off the main dimer peak. Comparison of the main peak retention time with that of a standard curve, prepared using proteins of known molecular weight dissolved in an SDS containing buffer, exhibited a 9955 Da molecular weight for TethTM. This is consistent with the formation of a strong dimer by TethTM in SDS micelles. Note that incubation with the reducing agents DTT or TCEP did not alter the mobility of TethTM (Figure 2.4), demonstrating that formation of the TM domain dimer does not require a disulfide linkage, although the presence of such a modification may have a stabilizing effect on dimers. Likewise, matching the SDS concentration used for the standard curve (0.3% SDS) had no effect on the mobility of TethTM (Figure 2.4); therefore, 1% SDS was used in all other experiments to avoid artificially concentrating the hydrophobic peptides and inducing nonspecific self-association.

Figure 2.3 Size exclusion chromatography (SEC) analysis of TethTM in SDS micelles. (A) Chromatograms of TethTM in 10 mM Tris (pH 7.5) containing 1% SDS are shown for three different concentrations of peptide. The peak of interest is indicated by an arrow, while other minor peaks are associated with small molecule contaminants in the sample and are present in solvent blanks. Note that the latter peaks do not change in volume with varied peptide concentration. (B) A molecular weight calibration curve for the S-2000 column, obtained by running standard proteins in the presence of SDS, is shown (line, black squares). The relative retention time of TethTM is also indicated on the plot (red square). Each point is an average of three runs of purified protein, with error reported as standard deviation.

Figure 2.4 SEC of TethTM under reducing conditions. (A) SEC of TethTM in 10 mM Tris (pH 7.5) containing 0.3% SDS and 1 mM DTT allowing direct comparison with the protein standard curve used to calibrate the SEC column. No significant change in sample mobility for the peptide peak (arrow) was observed relative to that in 1% SDS (Figure 2.3), and all other SEC experiments were carried out in 1% SDS to reduce any non-specific peptide interactions. (B) SEC of TethTM in 10 mM Tris (pH 7.5) containing 1% SDS and 1 mM DTT. Again, no change in the mobility of TethTM is observed relative to the non-reducing conditions shown in Figure 2.3. In both cases there are small molecule/buffer components that elute at the end of the column (denoted by an asterisk).

As an additional control, a peptide containing a scrambled tetherin TM domain, Teth(Scr), was used for SEC experiments. As shown in Figure 2.5, this peptide had a dramatically reduced self-association in SDS micelles relative to TethTM. This supports the presence of specific self-association of the tetherin TM domain, although a portion of the observed dimer may be due to nonspecific association of hydrophobic helices in detergent micelles. Further support for specific dimers of TethTM in SDS micelles comes from comparison with the SEC elution profile of a peptide containing the GpA TM domain (Figure 2.5). These results are consistent with some non-specific association of Teth(Scr) in SDS micelles, but nearly 100% dimerization of TethTM is indicative of a specific sequence motif for self-association, as observed for GpATM. As expected, this GpATM peptide is a constitutive dimer on SDS-PAGE (Figure 2.6A) and elutes from the SEC column at the same position as TethTM.

Figure 2.5 Comparison of SEC mobility of GpATM and Teth(Scr) versus TethTM in SDS micelles. Chromatograms are shown for TethTM, and two concentrations of GpATM and Teth(Scr), in 10mM Tris (pH 7.5) containing 1% SDS. The GpA TM domain peptide, which is known to form stable dimers in SDS micelles, elutes at the same position as the TethTM peptide. For the peptide containing a scrambled tetherin sequence, the dimer peak on the left is significantly reduced relative to the wild-type sequence, and decreases in relative intensity with reduced peptide concentration.

Figure 2.6 SDS-PAGE analysis of TethTM peptides and GpATM. A) Samples contain peptides at the concentrations indicated. TethTM runs as a smeared band with an upper edge near the molecular weight expected for a dimer. While still exhibiting significant smearing, the scrambled Teth(Scr) peptide runs significantly faster on SDS-PAGE, suggesting reduced self- association. GpATM, on the other hand, runs predominantly as a single band of dimer molecular weight. B) Attachment of pyrene maleimide to TethTM (i.e. TethTM(Pyr)) does not change peptide migration on SDS-PAGE.

2.3.3 Dimerization of TethTM Probed by Pyrene Excimer Fluorescence

In order to test the potential of TethTM dimerization in a bilayer membrane, a pyrene moiety was attached to Cys20 of TethTM. When two pyrene molecules are within 3.5 Å of each 64

other, they form an excited dimer, or excimer, giving rise to a characteristic fluorescence emission spectrum (274). The emission maximum of the excimer at 470 nm is readily distinguished from that of the monomeric fluorophore, which occurs at 376 and 396 nm (274). Pyrene excimer fluorescence has been used to probe assembly and conformational changes in both soluble and membrane embedded proteins (251, 275, 276). When incorporated into POPC liposomes, the pyrene labeled TethTM gave a CD spectrum that closely matched that of the unlabeled peptide (Figure 2.1B), indicating similar secondary structure in a bilayer membrane environment. Likewise, the labeled peptide exhibited identical mobility by SDS-PAGE and SEC (Figure 2.6B) validating the use of this labeled peptide to investigate the oligomeric state of the tetherin TM domain and also supporting the argument against a role for Cys20 in dimerization.

Fluorescence emission spectra recorded for pyrene-labeled TethTM incorporated into membrane mimetic environments are shown in Figure 2.7, along with the calculated excimer:monomer (E/M) ratios for each sample. In each case, the E/M ratio is 0.6−0.8, indicating that approximately 30−40% tetherin TM domain dimer is present at these concentrations in TFE/methanol solution, SDS micelles, and POPC liposomes. This observation agrees with the SDS-PAGE and SEC data and supports a role for the TM domain of tetherin in oligomerization. It is important to note, however, that SEC data obtained for samples containing 10−100 μM TethTM indicated 100% dimerization of this peptide in 1% SDS, while the E/M ratios obtained by fluorescence indicate a significantly lower dimer population. Since the degree of peptide labeling is already accounted for when calculating the E/M ratios, and only a small change in E/M is observed when the peptide concentration is increased to 13 μM (shown for POPC in Figure 2.7), the lower than anticipated excimer fluorescence likely stems from geometric considerations. The formation of excimers by pyrene requires extremely close approach of the two fluorophores, with high intensity excimer fluorescence only being observed for molecules within 3.5 Å of each other. Pyrene molecules separated by more than that distance will give rise to an apparent monomer signal. Several possibilities could therefore account for the fluorescence observed for the tetherin TM domain. A poorly defined dimer interface could allow flexible association of two monomers or permit several slightly differing conformations to exist. Alternatively, Cys20 from each monomer may either not be optimally 65

positioned relative to the dimer interface. Similarly, rotation of the Cys20 χ1 or χ2 bonds could place some or all of the pyrene moieties outside of the optimal distance for excimer formation. The presence of some antiparallel dimers may explain this result, although this has rarely been observed for homo-oligomeric membrane proteins.

Figure 2.7 Pyrene fluorescence emission spectra of TethTM. Fluorescence emission spectra for TethTM labeled with pyrene at Cys20, recorded in different membrane-mimetic environments. The excitation wavelength was 340 nm, with monomer and excimer emission maxima measured at 376/396 and 470 nm, respectively. Emission spectra recorded for blank samples containing only buffer and membrane mimetics are indicated by dashed lines and have been subtracted from the experimental values plotted. The excimer to monomer ratio (E/M) indicated for each experiment was calculated as described in the Experimental Procedures section.

2.4 Discussion

The role of tetherin in preventing the release of newly formed virions from the surface of an infected host cell requires that it make strong contacts with both the host cell membrane and the nascent virus particle. Several recent crystal structures of the dimeric C-terminal ectodomain from human and murine tetherin support the formation of an extended coiled-coil formed by two head-to-head monomers (172, 173, 175, 277). Structures of the reduced form of the ectodomain show a more relaxed structure, allowing formation of a tetrameric assembly, although the biological relevance of the putative tetramer has not been established (173, 175). Here, we have shown that the TM domain of tetherin forms parallel homodimers in the absence of the ectodomain and that dimerization does not rely on disulfide bond formation.

The observation that the isolated TM domain of human tetherin is capable of forming strong dimers in several membrane mimetic environments, including liposomal membranes, suggests that this domain may not solely act as an anchor but may also play a role in formation of tetherin oligomers in vivo. This interaction likely acts to further stabilize the parallel dimer formed by the ectodomain or could play a role early in folding and assembly of the mature protein. On the basis of our current data, we cannot exclude the possibility that the TM domain could serve to link two dimers, forming a tetramer, although this seems unlikely. Increased dimer stability in the absence of disulfide bonds may in part account for the nonessential nature of the ectodomain cysteines in tethering of Lassa and Marburg viruses (278). It is important to note that the tetramer packing observed in two recent crystal structures required removal of the Cys53 and Cys63 disulfide bonds or a reducing environment, which leads to a fraying of the coiled-coil structure and allowing interleaving of two adjacent dimers (173, 175). While high-resolution structural details of the TM domain dimer have not yet been reported, close packing of these domains would require a close approach of the N-terminal ends of the coiled-coil ectodomains in the dimer and would likely inhibit formation of the tetramer as previously reported. Likewise, our data suggest that models of tetherin architecture that place the two TM domains in different membranes are highly unlikely.

The self-association of the tetherin TM domain also has implications for the interaction of this domain with HIV-1 Vpu. Based on mutagenesis, the native TM domains of both proteins are required for Vpu to antagonize tetherin and prevent tethering of nascent HIV-1 virions (127, 128, 130, 131, 263, 266, 279–281). This has been supported by a recent NMR study which provided evidence for a direct and specific association of these two proteins in liposomes and in detergent micelles (266). While the complex was modeled as a heterodimer, in vivo tetherin is likely to interact with Vpu as a dimer given the extent of the tetherin homodimer interface, which we have shown to extend through the TM domain. While Vpu itself forms an oligomeric helical bundle in model membranes (109, 113, 117, 282–284), it is likely that the Vpu pentamer dissociates into free monomer prior to binding the tetherin TM domain, in a similar manner to the proposed interaction of phospholamban with the sarcoplasmic reticulum Ca2+ ATPase

(285–287). Further study will be required to elucidate both the structure of the full-length tetherin dimer and its role in virus tethering and the interplay with viral antagonists.

Chapter 3: FRET Analysis of the Promiscuous yet Specific Interactions of the HIV-1 Vpu Transmembrane Domain

References:

Sections of this chapter were originally published from Cole, G., Reichheld, S., and Sharpe S., FRET analysis of the promiscuous yet specific interactions of the HIV-1 Vpu transmembrane domain. Biophysical Journal. (2017) 113: 1992-2000 and reprinted with permission. Copyright 2017 Elsevier

Contributions:

The work in this Chapter was aided by Dr. Sean Reichheld who assisted in the generation of model fits. 69

3.1 Introduction

The TM domain of Vpu plays multiple roles in its function; the homooligomerization of the Vpu TM domain has been shown to be sufficient for channel activity, while a subset of the cell surface proteins targeted by Vpu are recognized via TM domain-TM domain interactions (129, 181, 288). Host cell proteins targeted by the Vpu TM domain include the co-activating natural killer (NK) cell receptor NTB-A, the activating NK cell receptor PVR, and the host restriction factor tetherin (129, 181, 288). The reduction in cell surface levels of these proteins provides HIV-1 with several biological advantages. A reduction in tetherin at the cell surface results in a dramatic increase in release of new virions from infected cells and thus an increased viral load within the host (159). Removal of PVR from the cell surface protects the infected host cell from being targeted by DNAM-1 NK cells (104). Finally, cells lacking NTB-A remain capable of activating NK cells but are unable to trigger sufficient NK cell degranulation (181). The fitness advantages of these Vpu activities are reflected in the pathogenicity of the different HIV-1 groups found in nature. Only the pandemic Group M viruses, and not the non-pandemic Group N, O or P viruses, encode a fully functional Vpu that is capable of efficiently antagonizing both tetherin and the CD4 receptor (289).

The Vpu mediated antagonism of NTB-A, PVR, and tetherin has been shown to depend on the TM domain of Vpu. Mutagenesis experiments performed in conjunction with flow cytometry and viral release assays have identified a conserved small-xxx-small amino acid motif (AxxxAxxxA) in the Vpu TM domain as a putative site for all TM domain interactions of Vpu (129, 266, 290). Based on proposed models of the Vpu homooligomer structure (109, 117), this alanine-rich face of the Vpu TM helix also mediates Vpu self-association, leading to a model in which Vpu oligomers must dissociate prior to interaction with human targets such as NTB-A, PVR and tetherin. To our knowledge there is no evidence for homooligomerization of the PVR or NTB-A TM domains, although both full-length tetherin and the isolated tetherin TM domain have been shown to exist as a parallel homodimer (168, 291). The full length tetherin homodimer is stabilized by three conserved disulfide bonds in its extracellular coiled-coiled

domain (172, 173), while the tetherin TM domain homodimer forms in the absence of any obvious interaction motifs (291).

Although a structural model has been proposed for the heterooligomeric interaction between the TM domains of Vpu and tetherin (266), no such details have been reported for other Vpu TM domain targets, including NTB-A and PVR. In fact, a direct TM interaction between the isolated TM domains of Vpu and NTB-A/PVR has not been previously demonstrated. In addition, the underlying energetics defining TM domain interactions have not been explored for any of these systems, raising the question of how a single TM domain is capable of mediating multiple homo and heterooligomeric interactions while retaining target specificity. To address this, we have designed a Förster resonance energy transfer (FRET) based assay to measure the binding of the Vpu transmembrane domain with the transmembrane domains of tetherin, PVR, and NTB-A. The association of TM domain peptides in liposomal environments has been previously measured using FRET in systems of varying complexity, with the best characterized being TM domain homodimers such as ErbB1, fibroblast growth factor receptor-3 (FGFR3), or glycophorin A (GpA) (13, 292, 293). More recent studies have investigated heterodimerization of transmembrane helices in the FtsB/FtsL complex of the E. coli divisome (257), and several larger complexes such as the influenza A M2 channel tetramer and pentamers of phospholamban have also been studied by FRET (14, 15). Likewise, heterodimeric interactions of ErbB family TM domains in detergent micelles have been studied by FRET, showing a hierarchy of interactions between family members, despite the presence of putative interaction motifs in each peptide (255). Although ErbB and FGFR display promiscuous binding within their respective protein families, there is significant sequence homology between the various interaction partners in each case (250, 255). Thus, neither of these systems display the level of promiscuity attributed to the Vpu TM domain in terms of TM domain recognition of non-homologous binding partners.

In this Chapter we provide evidence that dissociation of Vpu TM domain homooligomers occurs prior to interaction with the TM domain of each targeted host cell protein, such that the active form of Vpu is a monomer. Despite a lack of sequence homology, the target TM domains

compete in our FRET studies for Vpu binding, suggesting that they share a common binding surface on Vpu. From lipid titration experiments we have calculated the dissociation constants

(Kd) as well as the free energies of association (ΔG) for each TM domain interaction within our system, with similar values for each oligomerization reaction. This supports a model in which Vpu does not have a clearly preferred target, allowing promiscuous interaction with multiple targets. These results provide new insight into the molecular mechanism of Vpu-mediated immune evasion.

3.2 Materials and Methods

3.2.1 Peptide Synthesis, Purification and Labeling

The transmembrane segments for Vpu, PVR, NTB-A and tetherin were identified and peptides containing these TM domains (Table 3.1) were prepared using the Fmoc solid phase syntheses, as described in 2.2.1. TethTM was purified using the same conditions as in 2.2.1 while VpuTM, PVR TM and NTB-A TM were purified by reverse-phase (HPLC) on an 11x300mm C8 column (Vydac), using a 30 to 100% gradient of solvent A (10% acetonitrile, 0.1% TFA) to solvent B (isopropanol, 0.1% TFA). HPLC fractions were lyophilized and peptide identity was confirmed by electrospray ionization mass spectrometry at the SPARC Biocentre (The Hospital for Sick Children, Toronto, ON).

N-terminal labeling of peptides with dansyl or dabsyl chloride (Invitrogen) was carried out on resin by adding 10 mg of dye and 50 µL of diisopropylethylamine to 100 mgs of resin in 2 mL of dimethylformamide. The reaction was capped with nitrogen and left rocking overnight, after which the resin was washed with methanol and cleaved normally. Purification of labeled peptide was carried out as described for unlabeled peptide. Labeling was confirmed by both HPLC and mass spectrometry and quantified using methods modified from Rath et al (239). Briefly, the concentrations of dabsyl-labeled peptides in TFE were calculated from their absorbance at 470 nm using a molar extinction coefficient of 2.96 x 104 cm-1 M-1. Circular dichroism spectra were then recorded and the mean residue ellipticity (MRE) at 222 nm was calculated for each peptide. This MRE was then used to back calculate the concentrations of 73

dansyl, dabsyl or unlabeled peptides. This method was in good agreement with standard Trp absorbance methods where applicable (VpuTM and PVR TM). Fortunately N-terminal labeling of peptides resulted in a significant increase in peptide retention during HPLC purification. This allowed for the complete separation of labeled peptide from unlabeled peptide and free fluorophore, simplifying FRET analysis.

3.2.2 TM Domain Peptide Incorporation into Liposomes

TM domain peptides in a 1:2 TFE:methanol solution were added to 1-palmitoyl-2- oleoylphosphatidylcholine (POPC) in chloroform. Final peptide concentrations of 1 mol% relative to phospholipid (peptide:lipid ratio of 0.01) were used for acceptor titration and competition experiments. The peptide to lipid ratios for lipid dilution experiments ranged from 0.67 mol% to 0.02 mol%. The mixture was incubated to ensure complete dissolution and then dried under N2(g). To ensure complete removal of solvent, the sample was resuspended in water, frozen and lyophilized. The final peptide-containing liposomes were reconstituted in 50 mM sodium phosphate, pH 7.5, followed by 5 cycles of freeze-thawing. To reduce light- scattering in spectroscopic measurements, small unilamellar vesicles were produced by 10 cycles of mild probe sonication, 1 sec each, with sample cooling. Samples were left to equilibrate overnight in the dark at room temperature prior to fluorescence measurement.

3.2.3 Circular Dichroism Spectroscopy

All CD measurements were performed using a Jasco J-810 spectropolarimeter and a 1 mm path length quartz cuvette (Hellma Analytics). Spectra were measured from 190-250 nm at a scan rate of 50 nm/min, and were recorded as an average of three measurements. Liposome samples contained 55 µM peptide in POPC at 4 mol % peptide relative to lipid, in 50mM sodium phosphate buffer (pH 7.5). Spectra of blank samples containing appropriate amounts of lipid and buffer were subtracted from spectra of samples containing peptides. All CD measurements were made at room temperature.

3.2.4 Fluorescence Spectroscopy

Dabsyl fluorescence emission measurements were performed on a Photon Technology International C60 spectrofluorimeter scanning from 450 to 600 nm, using an excitation wavelength of 341 nm. Each emission spectrum was recorded as an average of three scans, acquired with integration set to 0.5 seconds, at a step size of 2 nm and with excitation and emission slit widths of 4 nm. The area under the resultant curve (450-600 nm) was calculated for each measurement.

3.2.5 Analysis of FRET Data

For unlabeled peptide competition experiments, the data are reported as normalized total fluorescence, with the integrated spectra of donor alone used to obtain a baseline fluorescence (F0) to which all samples containing donor and acceptor (F) were normalized:

퐹 퐹푅퐸푇 = ( ) (Eq. 1) 퐹0

Acceptor and lipid titration experiments are reported in FRET efficiency (E) which is calculated by:

퐹 퐸 = 1 − ( ) (Eq. 2) 퐹0

The efficiency observed directly from experiments (Eexpt) is a combination of FRET observed from specific oligomerization (Eoligo) as well as FRET due to the random colocalization of peptides (Eprox). In a lipid bilayer, donor and acceptor labeled peptides are limited to diffusion within the plane of the membrane, increasing the likelihood of FRET due to random proximity of donor and acceptor molecules. The estimated FRET due to proximity was subtracted from the experimental FRET efficiency for each peptide to lipid ratio tested in the lipid dilution experiments, resulting in only the FRET efficiency due to specific oligomerization:

퐸표푙𝑖𝑔표 = 퐸푒푥푝푡 − 퐸푝푟표푥 (Eq. 3)

Values for Eprox were obtained using the method of Wolber and Hudson which are described in detail below (294).

The FRET efficiency of an oligomer can also be presented as:

퐸표푙𝑖𝑔표 = 푓표푙𝑖𝑔표푝표푙𝑖𝑔표퐸푟 (Eq. 4)

Where foligo is the fraction of molecules in an oligomeric state, poligo is the probability of FRET occurring in an oligomer, and Er is the FRET efficiency within the oligomer. The fraction of molecules in an oligomeric state can be computed as:

푛[푂푙𝑖𝑔표] 푓 = (Eq. 5) 표푙𝑖𝑔표 [푇표푡푎푙] where [Oligo] is the concentration of the oligomeric species, [Total] is the total target peptide concentration and n is the number of monomers making up the oligomer. The probability of FRET occurring in an oligomer is dependent on whether or not oligomers are formed exclusively by donor or acceptor-labeled peptides, or by a mixture of the two. The probability of FRET occurring in a homooligomer can be calculated by the binomial distribution:

∑푛−1 푛 푘 푛−푘 푝표푙𝑖𝑔표 = 푘=1(푘) 푓퐷 푓퐴 (Eq. 6)

Here n is the size of the oligomer, k is the number of donors in the oligomer, fD and fA are the mole fractions of donor and acceptor respectively. This sum excludes the terms in which the oligomer will be made up of all acceptors, when k=0, and when the oligomer is made up of all donors, when k=n. When probing the formation of Vpu-target heterooligomers poligo=1 as any oligomers that form must have both a donor and an acceptor present. If the distance between the acceptor and donor dyes in an oligomer is smaller than the Forster radius of the FRET pair

(i.e. in a parallel oligomer) then Er = ~1. Vpu and its target peptides have known orientations in cellular compartments and are all predicted to form parallel oligomers with the exception of

Vpu-tetherin, which has been shown to be anti-parallel. As the R0 of the dansyl/dabsyl FRET pair approaches the bilayer thickness, an Er of 0.5 was used for tetherin.

Combining Eq. 3 and Eq. 4 relates oligomer concentration to FRET efficiency:

퐸 [Total] [푂푙𝑖푔표] = 푂푙𝑖𝑔표 (Eq. 7) 푛푝표푙𝑖𝑔표퐸푟

The oligomer fraction can then be used to calculate a dissociation constant given by:

[푀표푛표푚푒푟]푛 퐾푑푂푙𝑖𝑔표 = (Eq. 8) [푂푙𝑖𝑔표푚푒푟푛]

In which the monomer concentration [M] = [Total]-n[Oligo]. Finally, the free energy of oligomerization is defined as:

훥퐺 = RT ln(Kd) (Eq. 9)

3.2.5.1 Analysis of Tetherin Homodimerization

The equilibrium of tetherin homodimer formation in the absence of Vpu can be shown as:

퐾푑푇푒푡ℎ

2푇푒푡ℎ ⇆ 푇푒푡ℎ2

The dissociation constant of this equilibrium can be computed as:

[푇푒푡ℎ]2 퐾푑푇푒푡ℎ = ,(Eq. 10) [푇푒푡ℎ2]

where [Teth] and [Teth2] are the tetherin monomer and dimer concentrations respectively. The total tetherin concentration can be written as [TethTotal] = [Teth] + 2[Teth2], which can be rearranged and substituted into Eq. 9 to give:

2 (푇푒푡ℎ푇표푡−2푇푒푡ℎ2) 퐾푑푇푒푡ℎ = (Eq. 11) 푇푒푡ℎ2

The solution for this quadratic is:

2 2 4푇푒푡ℎ푇표푡+퐾푑푇푒푡ℎ−√(4푇푒푡ℎ푇표푡+퐾푑푇푒푡ℎ) −16푇푒푡ℎ푇표푡 [푇푒푡ℎ ] = (Eq. 12) 2 8 77

and can be related to FRET efficiency by:

[푇푒푡ℎ2] = 퐸푇푒푡ℎ[푇푒푡ℎ푇표푡] (Eq. 13)

Eq. 12 and Eq. 13 were used to fit the FRET data in Figure 3.7A in order to calculate KdTeth which was subsequently used in Eq. 9 to calculate the free energy of association, ΔGTeth, for the tetherin homodimer.

3.2.5.2 Analysis of Vpu Homooligomerization Like tetherin homooligomerization, Vpu homooligomer formation can be represented as:

퐾푑푉푝푢 푛푉푝푢 ⇆ 푉푝푢푛

푛 (푉푝푢푇표푡−푛푉푝푢푛) 퐾푑푉푝푢 = (Eq. 14) 푉푝푢푛

퐸푉푉 [푉푝푢푡표푡] [푉푝푢푛] = (Eq. 15) 푛(푝표푙𝑖𝑔표)

Eq. 14 and Eq. 15 were used to separately model Vpu homooligomerization for all possible oligomeric states proposed. This ranged from dimer (n=2) to pentamer (n=5). Unlike the quadratic function that results in tetherin homodimerization, Eq. 14 becomes quintic when n=5, which cannot be solved algebraically. Therefore we calculated the concentration of all the Vpu oligomers, [Vpun], for each individual sample using Eq. 15. The dissociation constant, KdVpu, was then computed for each sample according to Eq. 14, and was repeated for n=2 to n=5. All Kds for each value of n were averaged to give a value of KdVpu for each oligomeric state, which was then used in Eq. 9 to give ΔGVpu.

3.2.5.3 Analysis of Vpu NTB-A/PVR Heterooligomerization

Heterooligomerization of Vpu with its human targets NTB-A and PVR is convoluted by its own homooligomerization as both interactions will occur simultaneously in any sample. 78

However, conducting FRET experiments with exclusively donor labeled Vpu and acceptor labeled target allows for the direct observation of heterooligomerization. A bilayer containing Vpu as well as NTB-A or PVR (NP) will result in the formation of three species of Vpu: homooligomers (Vpu3), Vpu NTB-A/PVR heterodimers (VpuNP), and Vpu monomers (Vpu), the equilibria of which can be expressed as:

퐾푑푉푝푢 퐾푑푉푝푢푁푃 3푉푝푢 ⇆ 푉푝푢3, 푉푝푢 + 푁푃 ⇆ 푉푝푢푁푃

[푉푝푢][푁푃] 퐾푑 = (Eq. 16) 푉푝푢푁푃 [푉푝푢푁푃]

Where [NP] is the concentration of free NTB-A or PVR and [VpuNP] is the concentration of Vpu dimerized with either PVR or NTB-A. As tetherin also undergoes homooligomerization we will treat it separately below. Similar to tetherin homodimerization, the concentration of heterodimer, VpuNP, formed relates to FRET efficiency and total Vpu concentration as follows.

[푉푝푢푁푃] = 퐸푉푝푢푁푃[푉푝푢푇표푡] (Eq. 17)

In this context, there are only 2 species of NTB-A or PVR that can be formed, monomeric NTB- A/PVR, or Vpu bound NTB-A/PVR. Eq. 17 can be used to calculate the concentration of [VpuNP] and from this [NP] can be calculated by [NP]=[NPTot] – [VpuNP]. However, Vpu is also involved in homooligomerization, thus [Vpu] ≠ [VpuTot] – [VpuNP]. Instead [VpuTot] – [VpuNP] = [Vpuunbound], where [Vpuunbound] is a mixture of Vpu monomer and trimeric Vpu. The fit from Figure 3.7A was then used to back calculate the concentration of trimer and monomer concentrations of Vpu from [Vpuunbound]]. The Vpu monomer concentration was then used in Eq. 16 to calculate KdVpuNP and subsequently ΔGVpuNP.

3.2.5.4 Analysis of Vpu Tetherin Heterooligomerization

Vpu-tetherin heterooligomerization is treated in a similar fashion to Vpu NTB-A/PVR heterooligomerization with the addition of parameters describing tetherin homooligomerization. As discussed in the Results and Discussion below, we have proposed two 79

possible models that describe Vpu-tetherin heterooligomerization. Model 1: Only the tetherin homodimer binds to Vpu, forming a heterotrimer. Model 2: Both the tetherin homodimer and tetherin monomer bind to Vpu resulting in a mixture of Vpu-tetherin heterodimers and heterotrimers. The equilibria of both models are as follows:

Model 1

3푉푝푢 ⇆ 푉푝푢3, 푉푝푢 + 푇푒푡ℎ2 ⇆ 푉푝푢푇푒푡ℎ2 2푇푒푡ℎ ⇆ 푇푒푡ℎ2

Model 2

3푉푝푢 ⇆ 푉푝푢3, 2푉푝푢 + 푇푒푡ℎ2 + Teth ⇆ 푉푝푢푇푒푡ℎ2 + 푉푝푢푇푒푡ℎ

2푇푒푡ℎ ⇆ 푇푒푡ℎ2

Model 1 Model 2

퐸푉푝푢푇푒푡ℎ [푉푝푢푇표푡] 퐸푉푝푢푇푒푡ℎ2[푉푝푢푇표푡] 1/2 [푉푝푢푇푒푡ℎ2] = (Eq. 18A) [푉푝푢푇푒푡ℎ1/2] = (Eq. 18B) 퐸푟 퐸푟

2 [푉푝푢][푇푒푡ℎ2] [푉푝푢] [푇푒푡ℎ2][푇푒푡ℎ] 퐾푑푉푝푢푇푒푡ℎ2 = (Eq. 19A) 퐾푑푉푝푢푇푒푡ℎ = (Eq. 19B) [푉푝푢푇푒푡ℎ2] 1/2 [푉푝푢푇푒푡ℎ2][푉푝푢푇푒푡ℎ]

In model 1, the concentration of the VpuTeth2 heterotrimer, [VpuTeth2], was calculated using Eq. 18A, while [Vpu] was found using the same methods as above for Vpu-NTB-A/PVR heterooligomer formation. Tetherin homodimer and monomer concentrations were then calculated using [Tethunbound] = [TethTot] -2[VpuTeth2] and the tetherin fit from Figure 3.7A. All component concentrations were then used in Eq. 19A to calculate KdVpuTeth2. In model 2

[VpuTeth1/2], calculated by Eq. 18B, is a mixture of Vpu bound to either tetherin monomer,

[VpuTeth], or tetherin dimer [VpuTeth2]. As Vpu did not have a notable effect on tetherin homodimerization (Figure 3.9) we assumed that the tetherin dimer fraction remained constant whether or not Vpu was present. Therefore the ratio of tetherin dimer to monomer in isolation would be equivalent to the ratio of Vpu bound to tetherin dimer to Vpu bound to tetherin monomer. Eq. 18B and Eq. 19B were then used to calculate the parameters in model 2.

3.2.5.5 Proximity Corrections

A general approximation of FRET between donors and acceptors distributed randomly in a 2D plane, such as a lipid bilayer, has been previously described by Wolber and Hudson (294). Here we use their method to remove FRET arising from the random colocalization of peptides from our experimental data. As our experimental conditions do not allow for the control of peptide orientation in the bilayer, any estimation of FRET due solely to proximity must include parameters that account for random FRET that has occurred on the same side of the membrane as well as cross-bilayer interactions. The calculations require the two-dimensional concentration of acceptors (per unit area), R0, and the distance of closest approach Re, which was varied to account for cross-bilayer FRET. The two dimensional concentration of acceptors was calculated using the acceptor to lipid ratio and the area of a POPC headgroup, 62.7 Å2

(295). The R0 of the dansyl/dabsyl FRET pair is 33 Å (15). The distance of closest approach for random FRET occurring within the same plane was 0, (as we assumed the space taken up by the dye itself was negligible), where the distance of closest approach for random FRET occurring across the bilayer was the bilayer’s hydrophobic thickness 29.2 Å2 (295). The values for random FRET from both sides of the bilayer were summed and subtracted from experimental data for each lipid to peptide ratio used.

3.2.5.6 Statistical Analysis

Statistical analysis was carried out using GraphPad Prism 6.0 or Sigma Plot 11.0 software. In all cases, error bars represent the standard error of the mean. Where appropriate, confidence intervals were calculated using unpaired t tests and are reported as: ns p >0.05, * p≤0.05, ** p≤0.01, *** p≤0.001, **** and p≤0.0001.

3.3 Results

3.3.1 TM Domain Peptides Retain Native-like Secondary Structure in POPC Liposomes

The interaction between isolated TM domains of Vpu and tetherin has been documented previously using both in cell assays and in vitro NMR, although the energetics of this binding event have not been characterized (266, 290). Similarly, little is known about the TM domain interactions between Vpu and NTB-A/PVR. In order to investigate the energy landscape of Vpu-target interactions within the membrane, we have synthesized peptides containing the transmembrane domains of all 4 proteins, as well as a peptide composed of a scrambled Vpu TM domain sequence (VpuRD) previously shown to be incapable of antagonizing target proteins (Table 3.1) (288). Polar tags have been added to the termini of some of the TM domain peptides. This strategy has been shown to improve solubility and facilitate the purification of TM domain peptides without affecting their ability to form their native oligomeric states (268).

Table 3.1 Vpu and target transmembrane domain peptides. Predicted TM domains are underlined; N- and C-terminal polar tags are indicated in bold, where used.

Peptide name Amino acid sequence TethTM HHHHH19RCKLLLGIGILVLLIIVILGVPLIIFTIKAN49KKKK VpuTM 1 MQPIQIAIVALVVAIIIAIVVWSIVIIEYRK31 VpuRD MIPIIIAVILAVAVQAIVIVIVSWIIQEYRK NTB-A TM KK222TDTKMILFMVSGISIVFGFIILLLLVLRKRR252 PVR TM KKK339MSRNAIIFLVLGILVFLILLGIGIYFYW368KKK

Before probing the peptides for potential TM domain interactions, it was necessary to first determine if they adopted the expected secondary structure upon reconstitution into liposomes. In the present study POPC liposomes were used since this an abundant phospholipid in human membranes and has favorable physiochemical properties, including zwitterionic charge and a low phase transition temperature. Isolated synthetic peptides were incorporated into POPC liposomes by co-dissolution in organic solvent and CD spectra were recorded (Figure

3.1). As expected, all peptides display the characteristic minima at 208 and 222 nm, indicating that all TM domain peptides adopt a native-like α-helical structure in model membranes.

Figure 3.1 Secondary structure of TM domain peptides reconstituted into POPC liposomes. Circular dichroism spectra were recorded for each peptide at a peptide to lipid ratio of 0.04 and are reported as mean residue ellipticity (MRE). All spectra were recorded at 25oC in 50 mM sodium phosphate buffer (pH 7.5). The peptide concentration was 55 µM in each case.

3.3.2 Homooligomerization of Vpu and Tetherin

We have previously demonstrated that the tetherin TM domain forms homodimers in various membrane mimetics (see Chapter 2), and self-assembly of the Vpu TM domain has been demonstrated by several groups (115, 282, 291). While the size estimates of the Vpu homooligomer have varied (109, 115, 282), a pentameric structure is consistent with most experimental data. However, some studies have suggested the presence of multiple oligomeric states of Vpu in lipid bilayers and detergent micelles (117, 118). We carried out FRET experiments, which involve the non-radiative transfer of energy from a fluorescent donor to an acceptor, in order to first confirm the formation of tetherin and Vpu TM domain homooligomers and to probe for potential PVR TM/NTB-A TM homooligomerization. The dansyl fluorophore can be selectively excited in the presence of dabsyl labeled peptides (Figure 3.2) and this FRET pair has been previously used in TM oligomerization studies (15, 252).

Figure 3.2 Fluorescence emission spectra of dansyl and dabsyl conjugated VpuTM in POPC vesicles. The excitation wavelength was fixed at 341 nm while emission was scanned from 450 to 600 nm. The peptide to lipid ratio was 0.01, with a total POPC concentration of 200 µM. Spectra are the average of three independently made samples as described in section 3.2.2.

Competition experiments using unlabeled TM domain peptides were carried out to ensure that the FRET data reflect specific oligomerization, rather than non-specific peptide interactions caused by proximity in the lipid bilayer. The results for Vpu (Figure 3.3A) and tetherin (Figure 3.3B) both show a significant reduction of FRET following the addition of unlabeled peptide, indicating that the interactions observed are specific. Both NTB-A and PVR TM domains were also tested for self-association as it was unknown if either TM domain would self-associate. Neither of the NTB-A or PVR TM domains showed significant FRET in the presence of acceptor-labeled peptide, or any change in FRET signal during unlabeled competition experiments, indicating that the transmembrane domains of both NTB-A (Figure 3.3C) and PVR (Figure 3.3D) do not form homooligomers.

Figure 3.3 Homooligomerization of Vpu and tetherin TM domain peptides. Competition experiments with unlabeled peptide are shown for VpuTM (A), TethTM (B), NTB-A TM (C) and PVR TM (D). Each competition experiment consists of three samples: donor peptide only, 1:1 donor:acceptor, and 1:1:2 donor:acceptor:unlabeled. Donor quenching of dansyl-labeled VpuTM (E), or TethTM (F) in the presence of increasing mole fractions of dabsyl labeled acceptor peptide. The peptide to lipid ratio was maintained at 0.01 in each sample by addition of unlabeled peptide, the total concentration of POPC was 1 mM. All experiments were carried out in 50 mM sodium phosphate buffer (pH 7.5) and liposomes were incubated overnight prior to measurements. Data is fit to the equation FRET Eff=K(1-(1-X)n-1) where K is a constant and X Is the acceptor mole ratio. The oligomer size, n, was equal to 2 for TethTM. An n of 3 was used for VpuTM, as this value resulted in the best fit of lipid titration experiments discussed later (Figure 3.8B), suggesting that this represents the average or most prevalent oligomeric species.

After confirming the specific homo-interactions of VpuTM and TethTM peptides, the quenching of fluorescence emission from donor labeled peptides was measured in the presence of increasing amounts of acceptor labeled peptide, while maintaining a constant peptide to lipid ratio through addition of unlabeled peptide (Figure 3.2E, F). A linear response in this titration would be indicative of dimer formation, while departure from linearity would correspond to the formation of higher order oligomers (13, 15, 296). The linear fit for the TethTM titration supports our previous observation of a tetherin TM domain dimer, whereas

the results of the Vpu titration are best fit to a non-linear curve, consistent with formation of higher order oligomers from trimer to hexamer in size. Here, the curves describing trimer, tetramer, and pentamer formation provide equally good fits (discussed later in Figure 3.8A) to the acceptor titration data, preventing unambiguous determination of Vpu oligomer size. Therefore each oligomer size was treated separately in analysis of lipid dilution FRET experiments (Figure 3.8B) to determine the best fit.

3.3.3 VpuTM Binds Specifically to the TM Domains of Tetherin, NTB-A and PVR

To characterize the association between VpuTM and the target TM domains, FRET experiments were carried out in which the donor quenching of labeled VpuTM was measured in the presence of increasing amounts of acceptor labeled target peptides (Figure 3.4). FRET was observed between donor labeled VpuTM and all three acceptor labeled targets. Again, the specificity of these interactions was validated with unlabeled control experiments (Figure 3.4 A- C). The specificity of these interactions was further demonstrated using a scrambled Vpu peptide (VpuRD), which was unable to bind specifically to any of the target peptides (Figure 3.4 D-F). It is important to note that this study does not differentiate between parallel and anti- parallel interactions, as the R0 of the dansyl/dabsyl FRET pair is close to the thickness of the lipid bilayer (15). Previous studies have confirmed that Vpu-Vpu and tetherin-tetherin self- interactions are parallel in nature, while the Vpu/tetherin heterooligomer is anti-parallel. When accounting for the known topologies of Vpu, PVR, and NTB-A in cellular membranes, it is expected that these hetero-associations will be parallel (188, 297).

Figure 3.4 Competition experiments with unlabeled target peptides confirm the interactions between Vpu and target TM domain peptides. Relative fluorescence yields are shown for samples containing donor-labeled Vpu with acceptor labeled and/or unlabeled (A) TethTM, (B) NTB-A TM and (C) PVR TM. Negative controls using Vpu RD are shown for (D) TethTM, (E) NTB-A TM, and (F) PVR TM. Competition experiments were completed using the same donor:acceptor:unlabeled ratios described in Figure 3.3.

Acceptor titration experiments for Vpu binding to PVR and to NTB-A (Figure 3.5) show a linear response, suggesting that a heterodimer is formed in each case. However, the titration curve for the Vpu-tetherin interaction is non-linear, signifying the formation of a higher order heterooligomeric species. McNatt et al. observed the binding of Vpu monomers to tetherin dimers in cross-linking experiments performed using the full length proteins in vivo (170), suggesting that the Vpu binding face on the tetherin TM domain is distinct from the tetherin homodimer interface. Thus, while our data cannot rule out the possibility of simultaneous formation of both 1:1 and 1:2 VpuTM:TethTM complexes, the nonlinear response in the Vpu tetherin acceptor titration experiment demonstrates that a heterotrimer is likely to be present in our system.

Figure 3.5 Hetero-oligomerization of VpuTM with its targets. Donor quenching of dansyl- labeled VpuTM is reported in the presence of increasing mole fractions of dabsyl labeled target peptides (acceptor). The total peptide concentration was maintained at 10 µM by addition of unlabeled target peptide. Experiments were carried out in 50 mM sodium phosphate buffer (pH 7.5) and liposomes were incubated overnight prior to measurements. As in Figure 3.3E, F, experimental data were fit to FRET Eff=K(1-(1-X)n-1 where, n is equal to 2 for NTB-A and PVR, and 3 TethTM.

3.3.4 Target Peptides Compete for Binding with Vpu

Establishing that NTB-A, PVR and tetherin each bind to the Vpu TM domain raised the question of whether or not all three target peptides share the same binding face on Vpu. To address this, unlabeled competition experiments were carried out in which dansyl fluorescence was measured for samples containing donor labeled Vpu and acceptor labeled tetherin in the presence and absence of unlabeled NTB-A or PVR (Figure 3.6). In each case, fluorescence was restored in the presence of unlabeled peptide, indicating that NTB-A and PVR TM domain peptides are able to compete with the tetherin TM domain for association with Vpu.

Figure 3.6 Target TM domain peptides compete for interaction with VpuTM. Unlabeled competition experiments were carried out using donor labeled VpuTM in the presence of acceptor labeled TethTM and equal molar amounts of each unlabeled target peptide. Competition experiments were completed under identical conditions to the VpuTM/TethTM homooligomer experiments (Figure 3.3).

Currently, there are no proposed Vpu binding surfaces for the NTB-A or PVR TM domains. However, the ability of these peptides to compete with tetherin for binding to Vpu in liposomes suggests that these proteins are likely to also compete for binding in cell membranes. This supports the hypothesis that all three peptides share the same binding face on VpuTM, or that there is significant overlap between binding sites. This is supported by previously reported in vivo mutagenesis of the alanine face of Vpu (A10/A14/A18), in which loss of the small-XXX-small motifs was shown to reduce Vpu mediated antagonism of tetherin and PVR (129, 288). Remarkably, the TM domains of the three Vpu targets studied here do not share any identifiable sequence motifs that could account for their affinity for the same face of the Vpu TM domain. This is particularly intriguing since the Vpu TM domain is able to form both antiparallel interactions with the tetherin TM domain and parallel heterodimers with PVR or NTB-A, using the same binding surface in each case. Since this face of the Vpu helix is also proposed to be critical for Vpu homooligomerization (109), our data support a model in which the Vpu monomer interacts with target proteins. This could potentially result in the disassembly of higher order Vpu oligomers capable of conducting ions, a functionality that is been highly conserved across the pandemic strains of HIV-1, although it remains of unknown biological significance (298). 90

3.3.5 Energetics of Vpu-target Interactions

To determine if different binding energetics for each interaction could play a role in target specificity, we calculated the relative affinities of VpuTM – target oligomers, using FRET data recorded at varied lipid to peptide ratios, an approach validated in previous studies (257, 299). As transmembrane peptides are solubilized in a lipid bilayer rather than bulk solvent, the lipid to peptide ratio is the key factor in determining the fraction of oligomeric species (248). While the unlabeled FRET competition experiments described above demonstrate the presence of a specific interaction in each case, determination of accurate dissociation constants (Kd) and free energies of association (∆G) requires subtraction of proximity effects from the lipid titration data. The method of Wolber and Hudson (294), described in detail in section 3.2.5.5, was used to estimate the FRET efficiency caused by proximity effects and subtract it from our lipid dilution experiments. The corrected FRET efficiency can then be used to calculate the amount of oligomerization, and from this the Kd for each TM domain interaction. Figure 3.7A and Figure 3.7B show the proximity corrected FRET data used in the calculation of Kd and ∆G for TM domain homo and heterooligomers respectively.

Figure 3.7 Thermodynamic analysis of Vpu TM domain homo and heterooligomeric assemblies. (A) Dilution of VpuTM (solid line) or TethTM (dashed line) in POPC liposomes, showing loss of FRET efficiency due to dissociation of the homooligomers. The TethTM data were fit to a monomer-dimer equilibrium using Eq. 11-13. VpuTM was modeled as a trimer using Eq. 14 and Eq. 15. (B) Dilution of the VpuTM –hetero-oligomers in POPC liposomes. Fit lines were generated from the Kd values calculated from Eq. 16-18B.

Determination of these parameters for the VpuTM homooligomer was initially impeded since the acceptor titration data for VpuTM could not be fit to a single oligomeric state with any

confidence (Figure 3.8A). Thus, the lipid titration data were fit to curves calculated assuming that VpuTM existed as a single species ranging from dimer to pentamer in size. The results in Figure 3.8B show a best fit a VpuTM trimer. However, this likely represents the average FRET efficiency of an equilibrium mixture containing Vpu oligomers of various sizes. Such a mixture of species has previously been observed in chemical crosslinking and analytical ultracentrifugation studies in which dimers and trimers were the dominant species, along with smaller numbers of tetramers, pentamers, and higher order oligomers (117). In order to simplify the following Vpu- target calculations we assumed a model in which Vpu is trimeric.

Figure 3.8 Estimating the VpuTM oligomer size from acceptor titrations and lipid dilution experiments. (A) The FRET efficiency of VpuTM in POPC vesicles was measured as a function of acceptor-labeled protein fraction. The peptide to lipid ratio was kept constant at 0.01, and unlabeled peptide was replaced with increasing concentrations of dabsyl-labeled peptide. The data (dark circles) were fit to the equation FRET Eff=K(1-(1-X)n-1) where X is the acceptor mole fraction and n is the number of monomers in the oligomer. (B) The FRET efficiencies of VpuTM (black circles) at various peptide concentrations. Fit lines are modeled using the Kd values calculated from Eq. 14 and Eq. 15.

The self-association of VpuTM adds another dimension to determining the Kd and ∆G for Vpu-target interactions since three possible Vpu containing species can be formed; monomeric Vpu, homooligomeric Vpu, and target associated Vpu. Merzlyakov et al. have used an elegant method for FRET data analysis to calculate the energetics of heterodimerization of two peptides that also form homodimers (249). Here, we have modified this approach to account for the higher order homooligomerization of VpuTM. Additionally, while it was clear from the acceptor titration experiments that Vpu formed heterodimers with PVR and NTB-A, there was evidence of Vpu-tetherin heterotrimers, as described above. The addition of excess VpuTM to samples

containing both donor and acceptor labeled TethTM did not have an effect on tetherin TM domain dimer formation (Figure 3.9), and thus two possible models can explain Vpu-tetherin oligomer formation: 1) The Vpu monomer binds only to tetherin dimers; or 2) Vpu binds to both monomeric and dimeric tetherin.

Figure 3.9 Effect of VpuTM on TethTM homodimer formation. Unlabeled competition experiments were carried out using donor (D) labeled TethTM in the presence or absence of acceptor (A) labeled TethTM and two concentrations unlabeled (U) VpuTM. The presence of VpuTM has no statistically significant effect on FRET arising from TethTM homodimerization.

We have calculated the Kd and ΔG values for all VpuTM-target interactions, including both possible VpuTM-TethTM models. These are summarized in Figure 3.10, and show that the promiscuous interactions observed for Vpu are relatively weak relative to other TM domain interactions. In each case, the free energy of association for two helices forming a homo or heterodimer is from -1.7 to -2.6 kcal/mol. For VpuTM homooligomers or Vpu-TethTM heterooligomers, this value is shown in parentheses in Figure 3.10. Formation of VpuTM or VpuTM-TethTM trimers is more energetically favorable due to the additive effects of multiple TM domain-TM domain interactions. While the association of VpuTM with PVR-TM is slightly weaker than binding to TethTM or NTB-A TM, all of these are relatively weak interactions when compared with known strong TM helix dimers such as glycophorin A, which has a ΔG of dimerization from -5 to -5.7 kcal/mol in phosphatidylcholine membranes (247). The ΔG values 93

reported here for Vpu-target interactions (and for individual Vpu monomer contributions to the ΔG of homotrimers) are closer in magnitude to those reported for receptor tyrosine kinases, such as ErbB1 and FGFR3, which have ΔG values for dimerization of approximately -2 kcal/mol (292, 293, 300).

Figure 3.10 Thermodynamic and kinetic analysis of the Vpu TM domain interactions with target TM domain peptides. Free energies (kcal mol-1) and dissociation constants are shown for homooligomeric and heterooligomeric interactions of Vpu and target transmembrane domain peptides. Parameters were calculated from the data shown in Figures 3.7A and 3.7B. Values in parentheses are for the contribution of individual monomers to the Vpu homotrimer, or the Vpu-tetherin heterotrimer.

3.4 Discussion

The observation of uniformly weak Vpu TM domain interactions with itself and its targets may have several possible explanations. Firstly, there may be a requirement for release of Vpu prior to degradation of the target proteins, maintaining Vpu levels in the infected cell. Secondly, it may not be possible to design a TM domain that is capable of strong interactions with all possible targets, especially given the capacity for Vpu to engage in both parallel and antiparallel helical interactions. Finally, strong interactions would reduce the free monomer present, possibly preventing interaction with multiple partners. It is not known if an increase in TM affinity will correlate with an increased level of Vpu-mediated antagonism in vivo. Thus,

weak interactions that result in antagonism of multiple targets are likely to provide a greater fitness advantage to HIV-1 than strong binding to a single target. This may explain why the alanine-rich motif in the TM domain of Group M Vpu has been strongly conserved despite the high HIV-1 mutation rate. The impact of heterodimer formation on the putative ion channel activity of Vpu remains unclear, although our observations may suggest a careful balance between the ability to target host cell proteins through their TM domains while still maintaining the capacity for homooligomerization.

The subtle variations in affinity for each binding event suggest the possibility that different intermolecular interactions define the binding of Vpu to different targets, even if the same face of VpuTM is required in each case. A broad range of mechanisms through which transmembrane helix associations can be stabilized have been identified: the small-xxx-small motif, cation-π interactions, π-π stacking, ILV packing, and Cα-H-backbone hydrogen bonds can all contribute to helix-helix dimer stabilization (301). In the case of Vpu, the highly conserved small-xxx-small motif is very likely to play a key role in target TM association. The current model of Vpu-tetherin TM domain association involves the 30VxxxIxxLxxxL41 face of tetherin packing into the small-xxx-small motif of Vpu (266). However, without atomic level structures for each complex, it is not known what contacts this motif makes with each target helix. Both PVR TM and NTB-A TM contain multiple faces that display large hydrophobic residues, which could pack against the alanine-rich face of VpuTM. Studies of GpA have demonstrated that relatively conservative mutations in beta-branched amino acids (i.e. Val to Ile) can significantly enhance dimerization levels (241). Thus small changes in the binding face displayed by target TM domains could explain any slight variations in affinity towards VpuTM. To our knowledge, this is the first study showing that a single TM domain can bind with similar affinity to highly varied targets that do not share amino acid sequence motifs or a common membrane topology (i.e. Vpu binds to TM targets through parallel and anti-parallel interactions). Further work will be necessary to identify the precise mechanism of helix-helix association between Vpu and its different targets, and to determine whether or not this variance plays a biological role in preferential target selection.

Chapter 4: Hydrophobic matching of HIV-1 Vpu transmembrane domain helix-helix interactions is optimized for its subcellular location

4.1 Introduction

Variations in lipid head group and aliphatic chain group chemistry and functionalization give rise to >1000 different lipid species that make up the cellular membranes of mammalian cells (302). Molecular level variations in the lipids making up the lipid bilayer can result in changes in global bilayer properties such as hydrophobic thickness, surface charge, lateral pressure, lipid phase behaviour (acyl chain ordering) and membrane curvature (see section 1.7). Similarly, the presence of varying concentrations of sterols and proteins can significantly influence the physical nature of cellular membranes (303). The physical and chemical properties of lipid bilayers can have a significant impact on the folding and activity of integral membrane proteins (304–308). There is a heterogeneous mixture of membrane systems within a eukaryotic cell and membrane proteins synthesized in the ER are sorted and trafficked to their final subcellular location. Isolated membrane systems and organelles can have different membrane thicknesses and even different regions within the same membrane system can have different compositions and properties, such as in cholesterol rich microdomains (303, 309– 311). Among other parameters, several studies have focused on the role of bilayer thickness in membrane protein folding and assembly, specifically the influence of mismatch between the hydrophobic thickness of the lipid bilayer and the hydrophobic length of membrane-spanning protein segments (304, 305, 307, 308).

Responses to hydrophobic mismatch, (as shown in Figure 1.2) include changes in helical tilt, changes in lipid order/disorder, and also the non-specific oligomerization/aggregation of α- helical TM domains in the absence of specific helix-helix interaction motifs (258, 312). The influence of mismatch on otherwise stable TM oligomers has been studied using TM peptide- based FRET experiments for the GpA model system, as well as the tetrameric TM domain of influenza A M2 (14, 247). A FRET study of the M2 TM domain showed that its oligomerization was particularly susceptible to hydrophobic mismatch, with M2 being unable to oligomerize in thinner membranes (14). Conversely, FRET studies performed on a GpA TM domain peptide displayed the formation of dimers in bilayers ranging in hydrophobic thickness from 20 to 34Å (247). However, the bilayer thickness had a significant influence on the GpA TM domain

monomer-dimer equilibrium, with thicker membranes favoring dimerization. In the same study, an increase in aliphatic chain order, either in gel phase lipid membranes, or in fluid membranes with the addition of cholesterol, was shown to promote the self-association of the GpA TM domain.

These examples only explore the response of a single homotypic TM domain interaction to hydrophobic mismatch and do not predict the behavior of a promiscuous TM interactor like Vpu. In Chapter 3, we have shown that the HIV-1 Vpu TM domain forms weak, promiscuous yet specific interactions with itself and with the TM domains of NTB-A, PVR and tetherin in POPC membranes. It is unclear what effects changes in the bilayer environment, such as lipid phase and hydrophobic thickness, will have on protein-protein interactions within the Vpu-target system. Membrane thickness is of particular biological relevance as the different membrane compartments along the secretory pathway increase in thickness from the endoplasmic reticulum (ER) to the plasma membrane (303). All TM mediated targets of Vpu, as well as Vpu itself, are synthesized in the ER and travel along the secretory pathway to the plasma membrane, during which they are exposed to Vpu in a variety of different lipid bilayer environments (e.g. the ER, the Golgi, and endosomal compartment membranes). It is unknown if individual interactions will be favored under certain conditions or if all interactions will be simultaneously weakened or strengthened in different membrane environments.

In this Chapter 4, we use the FRET approach developed in Chapter 3 to show that changes in bilayer hydrophobic thickness and acyl chain ordering do not have a large effect on the secondary structure of these TM peptides, as measured by CD. However, both the hydrophobic thickness and the phase of the lipid bilayer have a significant effect on the levels of oligomerization observed between Vpu and its targets. These changes in the monomer- oligomer equilibria appeared to be global and independent of the Vpu target identity. Our results show that while the hydrophobic thickness of the bilayer can modulate the strength of each interaction, amino acid sequence is the dominant determinant of TM domain oligomer formation in this system. These results not only inform general studies on membrane protein folding and assembly, but also have impact on functional studies of Vpu, in that the

oligomerization of Vpu with its human targets may be optimized to occur primarily in specific subcellular compartments.

4.2 Materials and Methods

Methods for peptide synthesis, purification, and preparation and analysis of FRET samples were identical to those used in Chapter 3. Lipids used in this study, 1,2-dilauroyl-sn- glycero-3-phosphocholine (DLPC) ,1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC), 1,2- dioleoyl- sn-glycero-3-phosphocholine (DOPC), 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC), 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC), were purchased from Avanti Polar Lipids. Unless otherwise stated, fluid phase DMPC, DPPC and DSPC membranes were held at 5oC over their respective melting temperatures for all measurements.

4.2.1 Estimation of TM Domain Lengths

To estimate the length of the helical TM domains for each protein, the TM-spanning segments were predicted from the full length sequences of Vpu, tetherin, NTB-A and PVR, using

11 TM prediction algorithms: DAS (313), HMMTOP (314), ΔGapp predictor (315), OCOTPUS (316), TMHMM 2.0 (317), PHOBIUS (318), SCAMPI (319), SOSUI (320), SPLIT (321), TM Finder (267), and TM PRED (322). The resulting predictions were treated as ideal α-helical segments, with a helical rise of 1.5Å per residue, and the lengths of the resulting helices were averaged to give a final estimate.

4.3 Results

As discussed in section 1.6.4, the sequence dependent oligomerization of TM helices has been extensively explored in the literature but the overall influence of the membrane environment on these interactions remains a point of contention (323). Since studying the role of the lipid bilayer on membrane protein folding is complicated by the sheer number of possible lipid environments that can be tested, we have limited our current study to a range of lipids that vary in chain length and unsaturation levels but that have a consistent head group,

phosphatidylcholine (PC) (Table 4.1). The PC head group is the most abundant head group in mammalian cell membranes and PC liposomes are frequently used to represent a standard cell membrane (324). The lipids used in this study vary in hydrophobic thickness from 22 to 32Å and have phase transition temperatures ranging from -17 to 55oC, allowing for the comparison of Vpu interactions in as a function of mismatch and lipid phase transition.

Table 4.1 Phospholipid bilayers used in this study and their properties. Literature values for most bilayer thickness were measured by Kučerka et al while DOPC values were taken from the work of Tristram-Nagle and Nagle(295, 325). Here 2DC is the hydrophobic thickness and DHH is the phosphate to phosphate distance. Literature values for hydrophobic thickness were measured by small angle X-ray scattering.

Lipid Acyl chain length 2DC (Å) DHH (Å) Transition and saturation Temperature DLPC 12:0 21.9 30 -2 oC DMPC 14:0 25.7 34.5 24 oC DOPC 18:1 27 35.3 -17 oC DPPC 16:0 27.9 37.4 41 oC POPC 18:1,16:0 29.2 38.6 -2 oC DSPC 18:0 31.9 43.3 55oC

It was first necessary to establish that the Vpu and target TM domain peptides adopt their native secondary structures in each lipid environment. As can be seen in Figure 4.1, all peptides in fluid phase bilayers display the expected α-helical CD spectral features, regardless of the lipid used, with the exception of the thickest membrane used (DSPC) in which peptides exhibit a slightly reduced helicity, likely indicating that a hydrophobic mismatch threshold has been reached. Peptide CD spectra are similar in most lipids with the exception of the thickest membrane used in this study (DSPC) which display slightly less helical character, likely indicating that a hydrophobic mismatch threshold has been reached. Similar trends have been observed for pH low inserted peptides (pHLIP), that insert across a lipid bilayer in response to low pH, and which displayed lower helicity values when inserted into thicker membranes (326).

100

Figure 4.1 Secondary structure of TM domain peptides in fluid phase PC liposomes. CD spectra are shown for (A) VpuTM, (B) TethTM, (C) NTB-A TM, (D) PVR TM. All spectra are an average of 3 scans and were carried out in 50 mM sodium phosphate buffer (pH 7.5) at a 1:500 peptide:lipid ratio. Phospholipids with phase transitions above room temperature (DMPC, DPPC, and DSPC) were analyzed at 5oC above their respective phase transitions temperatures. CD spectra in DLPC and DOPC were measured at 21oC as these lipids were already in the fluid phase at room temperature.

4.3.1 The Effect of Temperature and Lipid Phase on Peptide Secondary Structure and FRET Efficiency

Probing Vpu-target interactions in our selected lipid environments is complicated by lipid gel to fluid phase transitions occurring across a wide temperature range. The secondary structure of all peptides was thus assessed in both the gel and fluid phases for DMPC, DPPC and DSPC lipids. To ensure that there were no temperature-dependent changes in helical character we also examined peptides in DOPC, which has the lowest phase transition temperature, at the highest temperatures used in the study (60oC). There were no significant differences in helicity 101

observed in CD data obtained in gel versus fluid phase membranes or at elevated temperatures well above the phase transition temperature, as can be seen from the representative spectra of VpuTM in Figure 4.2. The same trends were observed for PVR TM, TethTM, and NTB-A TM, (Figure 4.3), suggesting that peptide helicity is determined by lipid type and not by phase or temperature.

Figure 4.2 Secondary structure of Vpu TM peptide as a function of lipid phase and temperature. CD spectra are shown for VpuTM in (A) DMPC, (B) DPPC, (C) DSPC, and, (D) DOPC liposomes at the indicated temperatures. All spectra are an average of 3 scans and were carried out in 50 mM sodium phosphate buffer (pH 7.5).

102

Figure 4.3 Effect of temperature on peptide helicity in DOPC liposomes. The CD at 222nm was measured as an effect of temperature. Measurements were begun at 20 oC and increased at 1 oC/min until the maximum temperature 60 oC was reached. Each curve is an average of 2 melts from 2 separate samples and were carried out in 50 mM sodium phosphate buffer (pH 7.5).

After observing that TM peptide helicity was not significantly influenced by changes in lipid phase or temperature, we next tested the effects of these factors on the relative strength of Vpu-target interactions. The FRET efficiency of each Vpu-target interaction was measured as function of both temperature and lipid phase in DPPC liposomes (Figure 4.3A). In the fluid phase (above 41oC) a significant decrease in FRET efficiency was observed for each Vpu-target interaction, indicating increased oligomerization of all TM domain pairs in gel phase DPPC. Additionally, changes in temperature within either the fluid or gel phase did not significantly affect the FRET efficiency of any Vpu-target interaction. To further ensure that these results were caused by the change in bilayer properties due to the gel-fluid phase transition and not simply the result of a temperature dependent disruption of the Vpu-target interactions, we also measured the FRET efficiency for each TM domain pair across the same temperature range (22- 60oC) in DOPC membranes and DPC micelles (Figure 4.4B and C, respectively), neither of which undergo a phase transition within the temperature range studied. All Vpu-target interactions were unaffected by temperature changes in fluid DOPC membranes, and showed only a modest loss of FRET efficiency at higher temperatures in DPC micelles. The latter was far smaller than the loss of efficiency due to the change in FRET efficiency across the DPPC phase transition.

103

Figure 4.4 Effect of temperature on the FRET efficiencies of Vpu-target pairs. Donor quenching of the FRET efficiency for equimolar dansyl labeled Vpu in the presence of dabsyl labeled target, recorded as a function of temperature in (A) DPPC liposomes, (B) DOPC liposomes, and (C) DPC micelles. Samples were incubated at each target temperature for 5 minutes between reads. ∆FRET efficiency was measured as the difference in efficiency at a given temperature from the FRET efficiency at 22oC.

To further determine the impact of a phospholipid gel to fluid phase transition on the behavior of peptide-peptide interactions, we measured the FRET efficiencies of Vpu-target pairs at increasing lipid to peptide ratios in both the gel and fluid phase. Figure 4.5 shows the FRET efficiency of Vpu-target FRET pairs in DPPC in either the gel (Figure 4.5A) or fluid (Figure 4.5B) phase. FRET efficiencies for all Vpu-target pairs were significantly higher in the gel phase. The decrease in FRET efficiency in the fluid phase appears to be similar for all peptides (Figure 4.5C). This decrease also seems to be larger at higher peptide to lipid ratios, which may be indicative of some non-specific associations occurring at higher peptide concentrations in gel phase membranes. These results show that while temperature itself does not have a significant effect on Vpu-target oligomerization, there is a clear dependence on lipid phase behavior, with gel phase membranes favoring oligomerization. This is not unexpected, as the chain ordering occurring in gel phase lipids will favor lipid-lipid contacts over peptide-lipid contacts and potentially lead to an increase in non-specific TM domain peptide oligomerization (312, 327).

104

Figure 4.5 The influence of lipid phase on Vpu-target interactions. FRET efficiencies of Vpu- target interactions in DPPC (A) gel and (B) fluid phases at peptide to lipid ratios of 0.002 to 0.0002. The gel and fluid phases were maintained at 35 and 45oC respectively. (C) The difference in FRET efficiency in the gel phase from the fluid phase.

4.3.2 Effect of Phospholipid Acyl Chain Length on Interactions with VpuTM and its Human Targets

To determine the effect of hydrophobic thickness on oligomerization within the Vpu target system, FRET experiments were performed in PC bilayers with hydrophobic thicknesses ranging from 22 to 32Å. It is important to note that hydrophobic mismatch can change the tilt of a helical TM domain within the bilayer, with thinner bilayers potentially resulting in an increased crossing angle for any TM-TM interactions (328). While this increase in crossing angle will increase the distance between the fluorophores on the TM domain peptide termini, the expected variation is only on the order of a few angstroms, which is well within the Förster radius of the FRET pair used in this study (33 Å). Thus any observed changes in FRET efficiency will arise solely from changes in the monomer-oligomer equilibrium, and not from changes in the geometry of interhelical association.

As described in Chapter 3, FRET efficiency was measured for equimolar ratios of donor and acceptor labeled peptides at decreasing peptide to lipid ratios, allowing quantification of both VpuTM/TethTM homooligomerization (Figure 4.5) and VpuTM heterooligomerization (Figure 4.6) in liposomes composed of lipids with different acyl chain lengths. All membranes

105

were maintained in the fluid phase by equilibrating samples at 5oC above their phase transition temperature prior to measurement of fluorescence emission spectra.

No FRET was observed for any peptide pairs in the thinnest bilayers (DLPC), likely due to the high degree of hydrophobic mismatch resulting in increased peptide tilt and unfavorable helix-helix crossing angles, preventing oligomerization. For the other lipids, an overall trend in both the homooligomeric and heterooligomeric data is observed. FRET efficiencies for each peptide pair increase with bilayer thickness, until a maximum is reached in DPPC membranes, after which the FRET efficiency begins to decrease. Based on previous reported data for Vpu and tetherin TM domains (329), the homooligomeric TethTM data were fit to a monomer-dimer equilibrium (Figure 4.5A), while the homooligomeric VpuTM data were fit initially to a monomer-trimer equilibrium Figure 4.5B (solid lines). This trimeric model fit poorly with the data of VpuTM in DSPC (Figure 4.5B solid purple line), and a monomer-dimer equilibrium (Figure 4.5B, dashed purple line) resulted in an improved fit. This may reflect a reduced self- association of VpuTM in DSPC membranes, rather than dimer formation, as several lines of evidence suggest that Vpu is capable of forming multiple oligomeric species, such that FRET data represent an average (117, 118).

Figure 4.6 Homooligomerization of Vpu and tetherin TM domains in PC lipid membranes. Dilution of TethTM (A) or VpuTM (B) homooligomers in PC lipids of varying chain length. TethTM was fit as a monomer-monomer dimer equilibrium using Eqs 11, 12 and 13 from Chapter 3. VpuTM was modeled as a trimer using Eqs 14 and 15, with the exception of DSPC, which was modeled either as a trimer (solid purple lines) or a dimer (dashed purple lines).

106

Figure 4.7 Heterooligomerization of Vpu and its human targets in fluid PC lipid membranes. Lipid dilution of equimolar amounts of donor labeled VpuTM with acceptor labeled (A) TethTM, (B) NTB-A, (C) PVR TM in PC lipids of varying chain length. Fit lines were generated from Kd values calculated from Eqs 16-18B. Fit lines for VpuTM modeled as a homodimer (rather than trimer) in DSPC are shown as dashed lines.

Promiscuous binding of Vpu to its human TM targets was observed in all lipid environments used here, with the exception of DLPC membranes, in which no TM domain interactions were observed for any combination of peptides (Figures 4.5 and 4.6). These results are similar to those of the homooligomeric measurements in that FRET efficiency increases with chain length to a maximum threshold in DPPC membranes. We assumed that the stoichiometry of Vpu-target oligomers matched those observed in POPC membranes (see Chapter 3), with Vpu forming heterodimers with PVR and NTB-A, while being capable of interacting with monomers or dimers of TethTM. Unbound VpuTM was also assumed to homooligomerize in the presence of target peptides, similar to VpuTM in isolation. However, improved fits were obtained for the DSPC data using a Vpu monomer-dimer equilibrium, rather than monomer- trimer (Figures 4.5 and 4.6, dashed lines).

4.3.3 Energetics of Vpu-target Interactions in PC Liposomes

The lipid dilution experiments shown in Figures 4.5 and 4.6 were used to calculate the dissociation constants (Kd) and the free energies of association (∆G) for homotypic) and heterotypic (Table 4.2) interactions of Vpu. The lipid chain length, or hydrophobic thickness of the membrane, significantly affects the monomer-oligomer equilibrium of all helix-helix interactions present in the Vpu-target system, with all pairwise TM domain interactions having the largest free energy of association in DPPC membranes.

107

Table 4.2 Homo and heterooligomer free energies of association of VpuTM with its human targets in PC membranes of increasing chain length. Free energies of association of homo and hetero association within the Vpu target system in membranes of increasing hydrophobic thickness (2Dc) Values in parentheses were calculated using a VpuTM monomer-dimer model, rather than monomer-trimer. All ∆G values are reported in (kcal/mol).

Homooligomers Vpu-target heterooligomers ∆G (kcal/mol) ∆G (kcal/mol) Lipid 2DC (Å) VpuTM TethTM TethTM NTB-A TM PVR TM DLPC 21.9 n/a n/a n/a n/a n/a DMPC 25.7 -4.51 -2.54 -2.08 -2.65 -2.01 DOPC 27 -4.88 -2.72 -2.82 -2.60 -2.8 POPCa 29.2 -5.1 -2.2 -2.6 -2.6 -2.3 DPPC 27.9 -7.77 -6.84 -4.64 -4.05 -3.42 DSPC 31.9 -6.04 (-3.48) -4.98 -3.49 (-3.51) -3.3 (-3.28) -2.91 (-3.0) a POPC values are as measured in Chapter 3.

4.4 Discussion

When considering membrane protein folding in a bilayer, the lipid environment will play an important role in determining the relative contributions of peptide-lipid interactions that may influence peptide oligomerization. These effects are in addition to peptide-peptide interactions that stem from the peptide/protein amino acid sequences and structuring. In this Chapter, we have provided an initial investigation into the influence of the lipid bilayer on the interaction of HIV-1 Vpu with target TM proteins, focusing on changes in the hydrophobic thickness and phase behavior of the membrane.

Using a FRET approach similar to that used in Chapter 3, we show that the VpuTM peptide is able to form heterooligomeric assemblies with its target peptides in lipid bilayers that vary in hydrophobic thickness from 25 to 32Å. In the thinnest membrane tested (DLPC, 22Å), Vpu did not form any homo- or heterooligomeric interactions, suggesting that there is a limit for the degree of helical tilt that will permit favorable helix-helix interactions. The length of a TM domains hydrophobic core will largely determine the degree of its helical tilt. We have estimated the average hydrophobic core lengths Vpu and all of its targets using 11 TM helix prediction algorithms (see 4.2 for details) to find that they ranged from 34 to 34.8 Å, suggesting

108

that any hydrophobic mismatch with the bilayer may result in similar changes in helical tilt angle for all peptides. Previous studies have shown that the Vpu TM domain can adopt a wide range of helical tilt angles (15o to 51o) based on the level of hydrophobic mismatch with the residing membrane (266, 279, 330). The helical tilt angles of Vpu targets have not been as thoroughly studied, with tetherin in DMPC membranes (24o) being the only prior measurement (266).

The effect of helical tilt angle on membrane protein oligomerization is still a relatively unexplored field. Increasing helical tilt angle through negative hydrophobic mismatch has been shown to induce non-specific oligomerization in synthetic peptides that lack specific oligomerization motifs (258). In other cases, changes in hydrophobic mismatch reduced the level of association of peptides that form specific oligomers (14, 247). Our data is similar to the latter studies in which departure from the optimized membrane thickness results in a lower level of specific association.

Additionally, we show the effect of lipid bilayer phase behavior on Vpu interactions, that is the gel phase seems to promote the non-specific formation of Vpu homo and heterooligomers. Although this finding may not be directly relevant in a biological context, it instead reflects a scenario in which less favourable lipid-protein interactions have the potential to drive protein-protein assembly regardless of sequence or interaction motifs.

Although the TM domain interactions are retained in all but the thinnest lipid bilayers (DLPC), differences in the overall oligomer fraction (and the corresponding strength of interaction) were observed as a function of bilayer thickness. There appears to be less VpuTM homooligomerization in thinner membranes, which may stem from hydrophobic matching conditions between TM domains and the residing bilayer. This finding also supports the work of Hussain et al., who used in cell FRET to show that Vpu was unable to oligomerize in 37.5 Å thick ER membranes (measured as phosphate to phosphate distance), but was able to form homooligomers in the other compartments of the secretory pathway including 39.5 Å thick Golgi membranes (282, 303). Functional studies of Vpu mediated downregulation of tetherin/NTB-A/PVR suggest post ER-interference that occurs in perinuclear compartments 109

such as the Golgi apparatus (182, 265, 331). This is in contrast to Vpu mediated CD4 downregulation, which occurs in the ER but does not require direct TM-TM association (135).

Overall, our data support a model in which the effects of hydrophobic mismatch on membrane protein oligomerization has an upper and a lower threshold. This is likely to arise from geometric considerations, with changes in TM domain tilt significantly impacting the helix packing interface. The thicknesses of the synthetic membranes where the highest Vpu-target FRET efficiencies are observed roughly correlate with the natural membranes Vpu-target complexes are found (i.e. the Golgi network and endosomal compartments) (210, 265, 295, 303). This apparent optimization of hydrophobic matching in specific membranes may have functional implications for Vpu. With ideal Vpu function occurring in the thicker Golgi and endosomal membranes, rather than the ER membrane, Vpu can readily target both newly synthesized proteins, as well as those being recycled from the cell surface, rather than only nascent proteins from the ER.

110

Chapter 5: General Discussion and Future Directions

111

5.1 Understanding Vpu-target oligomerization

The importance of Vpu is illustrated by the fact that only the HIV-1 strains encoding a more functional Vpu (the Group M strains) are pandemic. Overall, Vpu allows HIV-1 infected cells to escape the innate immune system, creates a more optimal environment for viral replication, and mediates the release of budding viral particles. The Vpu proteins from Groups N, O, and P display attenuated Vpu activity through either a narrower range of targets or by showing weaker levels of target downregulation. It is remarkable that Vpu can interact with such a broad range of cellular targets with little sequence homology while being such a small protein (81 amino acids). Studying the mechanisms through which Vpu and its targets oligomerize within the membrane is a powerful tool in furthering our understanding of membrane protein folding and also reveals key details in HIV-1:host interactions.

Understanding the role of Vpu in HIV-1 infection remains challenging, especially considering that several of its targets are bound via contacts in the membrane spanning regions. Difficulties often arise in the expression and purification of sufficient amounts of membrane protein, as well as in achieving stable sample conditions, suitable for biophysical and structure characterization. A peptide based approach allows for the isolation of highly pure, site specifically labeled TM domains in sufficient yield for biophysical characterization. By omitting the cytosolic portions of membrane proteins, we are able to unambiguously study the contributions of the TM domains to protein assembly. An in vitro approach also facilitates the use of highly controlled membrane environments to conduct biophysical experiments and avoids the stochastic nature of in cell research.

We have used a combination of biophysical techniques to investigate Vpu and its transmembrane mediated interaction with targeted TM domains. FRET was chosen as the major technique for the work presented in this thesis since it allowed us to observe and quantify individual TM domain assemblies in complex systems containing many lipid-lipid, protein-lipid, and protein-protein interactions. It is sensitive to oligomeric size, something critical for studies with Vpu, where the stoichiometry of assembly has remained an unresolved issue in previous studies. Finally, in this work, we wanted to apply FRET to a challenging multi- 112

protein equilibrium that required expanding on current methods available for analyzing this type of data.

5.2 The Tetherin TM Domain Forms Parallel Homodimers

In Chapter 2, we demonstrated the formation of parallel tetherin TM domain homodimers in model membranes. This evidence implies that the tetherin TM domain may have a role beyond that of just a membrane anchor. In the context of full length tetherin, dimerization of the TM domain may facilitate dimerization of the ectodomains - a potential third stage of membrane protein folding. Certainly dimerization of the TM domain limits the potential topologies for tetherin when tetherin budding viruses, and may also influence heterooligomeric interactions with viral antagonists such as Vpu. The overall dimerization of tetherin has also been shown to provide tensile strength and stiffness to the proteins steered molecular dynamics simulations (332), suggesting that dimerization of the TM domain may contribute to the overall strength of the membrane tether. Nevertheless, the dimerization of the tetherin TM domain is likely to be an important factor in any efforts to selectively inhibit the Vpu-tetherin interaction.

5.2.1 The Vpu TM Domain Forms Similarly Weak Interactions with Itself and its Target TM Domains in POPC Membranes

Preliminary data have supported a direct TM-TM association of Vpu with a subset of its human targets, mostly through in vivo studies showing the sensitivity of co- immunoprecipitation experiments to mutations in the TM domain of Vpu (104, 129, 181, 288). In Chapter 3, we demonstrated the presence of direct interactions between the Vpu TM domain and the TM domains of human tetherin, NTB-A and PVR. Furthermore, we quantified the strengths of these interactions, as well as the homooligomeric assemblies of Vpu and TethTM, in hopes of finding a preferential Vpu target. Remarkably, Vpu was able to bind all of its membrane targets with similar affinities, perhaps because the downregulation of several unique targets is evolutionarily more favorable than the tight binding to one major target. Alternatively, weak promiscuous binding may more readily allow Vpu to dissociate from its 113

target protein before target degradation, thereby maintaining constant cellular levels of Vpu. This work provides a rationale for the high degree of transmembrane domain sequence conservation in Vpu from pandemic strains of HIV-1, and may suggest a role for Vpu in increased virulence of these.

Furthermore, we have showcased that relatively simple FRET experiments can be used to describe a complex system made up of multiple homo- and heterooligomeric interactions. Through the use of discrete labeling schemes and subsequent analysis, we were able to quantify each of the individual interactions within the Vpu-target system, while taking the stoichiometries of each type of assembly into account.

5.2.2 One Face to Find Them All and in the Membrane Bind Them

In further work in Chapter 3, we used FRET competition experiments to show that all three Vpu targets studied compete for interaction with Vpu. This finding implies that all Vpu targets are binding to the same face of the Vpu TM domain helix. This is further supported by in vivo mutagenesis experiments performed by other groups, in which mutations in the alanine rich (10AXXXAXXXAXXXW22 ) face of Vpu attenuated Vpu binding and downregulation of tetherin and PVR (129, 288). Competition for interaction in our in vitro studies implies that this competition is also occurring in the cell membranes of HIV-1 infected cells. The ability of a single TM domain to bind to multiple TM domains has been documented within families of receptor tyrosine kinases and it is thought that changes in the ratios of homo/heterodimers formed can have profound biological effects (250). However, the proteins within the FGFR network of receptor tyrosine kinases are homologs, where Vpu is unique in that it binds different TM domains that do not share common sequence motifs, a phenomenon which is not well understood and which may require high-resolution structural studies of each TM domain interaction.

114

5.3 The Influence of the Lipid Bilayer on Interactions Within the Vpu- target System

Our initial study of the interactions with Vpu and its TM targets was limited to liposomes composed of POPC. While this a commonly used mimetic of mammalian cell membranes, it lacks the diversity of chemical and physical properties found within biological membranes in vivo. To shed light on the role of the lipid bilayer in the folding and oligomerization of Vpu and its TM targets, we conducted biophysical experiments in lipids that differed in hydrophobic thickness and acyl chain order, but maintained a zwitterionic phosphatidylcholine headgroup.

Temperature only had a modest effect on peptide oligomerization, suggesting that the melting temperature for each TM domain oligomer was beyond the range of our experiments. Both acyl chain length and lipid phase (acyl chain order) had significant impact on the level of Vpu-target oligomerization. Peptide interactions were stronger in the gel phase which is largely expected: the ordering of the lipid chains results in an intolerance of the peptide impurity. That is, the presence of the peptide in the lipid membrane interferes with lipid-lipid packing, and as lipid-protein interactions become less favorable, protein oligomer formation becomes more energetically favorable. This latter result may suggest there may be a temporal or subcellular localization selection occurring in Vpu, i.e. the sequence of Vpu may be driven by evolution to not only bind many TM partners, but also to bind these partners in a specific membrane environment.

5.4 Future Directions

5.4.1 Interaction faces and structures of Vpu NTB-A/PVR heterodimers

The helical interaction faces of Vpu and tetherin have been established through in vivo mutagenesis studies as well as solution state NMR (129, 263, 266). Large hydrophobic residues in the tetherin TM are proposed to pack into grooves on the Vpu TM domain created by an alanine rich face on Vpu. The same alanine-rich face has been implicated in downregulation of both NTB-A and PVR, albeit the packing residues on the human targets remain unknown (181,

115

288). We have run Vpu and its target sequences through the PREDDIMER algorithm described by Polyansky (333, 334) in order to predict possible Vpu target structures (Figure 5.1). In each case, the algorithm was able to generate highly ranked structures in which the alanine rich face of Vpu was the interaction face and was able to predict correctly predict the known interaction face of tetherin (Figure 5.1A). Although the algorithm does not predict with certainty whether or not these TM domains will dimerize or not, a difficult problem using only in silico techniques, it does provide putative dimer conformations that can be confirmed using more biological approaches such as TOXCAT or through mutational analysis in functional/activity assays. Alternatively, full structural elucidation of Vpu-target heterodimers, using solution and/or solid state NMR techniques, would not only identify the target interaction residues but also map the knobs into holes packing surface. These methods will require the production of isotopically labeled TM peptides and the homooligomerization of Vpu in the presence of Vpu-target heterodimers may make analysis challenging. Identifying the interacting residues on PVR/NTB-A will allow for comparison of the interaction face on tetherin and the homo-interaction face on Vpu. This data will give insight into how Vpu can bind such a wide range of host cell proteins, furthering our knowledge of membrane protein assembly.

Figure 5.1 Putative Vpu-target complexes. Spatial structures of generated using the PREDDIMER algorithm of (A) VpuTM-TethTM, (B) VpuTM-NTB-A TM and (C) VpuTM-PVR TM.

116

5.4.2 The role of Vpu homooligomerization

The major model in the literature that attempts to explain the biological role of Vpu homooligomerization is the Vpu ion channel model. Several studies have been able to document the Vpu mediated transport of ions across planar bilayers in purified systems(109, 112). Some in vivo studies have also measured the (change in) ion flow in the presence of Vpu in living systems, such as E. coli, yeast, and xenopus oocytes (133, 335, 336). However, this model is incomplete as Vpu ion channel activity has never been formally linked to a biological function. Early studies attributed the enhanced virus release to the Vpu ion channel, although this is now well established to be due to antagonism of cell-surface tetherin (159). HIV-1 can control the cell cycle to its benefit through the activity of Vpr (108), so there is a possibility that Vpu ion channel activity may help facilitate this control of the cell cycle by perturbing ion homeostasis within the cell. However, this hypothesis has yet to be been tested.

Finally, Vpu has been well documented as a modulator of the host innate immune response, an activity that we (and others) have shown involves binding of Vpu to target proteins as a monomer. Ergo it is unlikely that Vpu is acting simultaneously as an ion channel and as an immune modulator. Our attempts at reproducing ion channel activity in a liposome based assay were ultimately unsuccessful, leading us to postulate that Vpu ion channel activity may be an artifact of its peculiar homooligomerization properties. In this hypothesis, which is also suggested by recent coarse-grained MD simulations of Vpu assembly (118), the dynamic and constant assembly/disassembly of Vpu into multiple oligomeric states causes the formation of a channel-like conformation causing the occasional ‘slip’ of ions through holes made by Vpu. Thus the monomer is the functional Vpu unit in our model, although homooligomer formation may serve to modulate the concentration of monomer in the membrane or to mediate the various heterooligomeric interactions available to the Vpu TM domain. Alternatively, the homooligomerization of Vpu may simply be an unintended consequence of having a promiscuous interaction face.

The very presence of multiple self-assembled states for Vpu has made study of its oligomerization challenging. Most techniques, including our FRET study, can only observe an 117

average signal that is a weighted representation of the entire population of Vpu in the membrane. Therefore future studies of Vpu homo/heterooligomerization should be carried out using methods that can observe TM domain interactions at the single molecule level. Two possible methods include single-molecule photobleaching, or single molecule localization microscopy, which employs super resolution techniques to calculate the stoichiometry of labeled complexes (337, 338). This approach would allow testing of the hypothesis that Vpu exists in multiple oligomeric states, and provide new evidence for or against a biological role for Vpu homooligomers.

118

References

1. Stevens, T.J., and I.T. Arkin. 2000. Do more complex organisms have a greater proportion of membrane proteins in their genomes? Proteins Struct. Funct. Genet. 39: 417–420.

2. Aktories, K., C. Schwan, and T. Jank. 2017. Clostridium difficile toxin biology. Annu. Rev. Microbiol. 71: 281– 307.

3. Bleicken, S., M. Classen, P.V.L. Padmavathi, T. Ishikawa, K. Zeth, H.J. Steinhoff, and E. Bordignon. 2010. Molecular details of Bax activation, oligomerization, and membrane insertion. J. Biol. Chem. 285: 6636– 6647.

4. Cymer, F., G. Von Heijne, and S.H. White. 2015. Mechanisms of integral membrane protein insertion and folding. J. Mol. Biol. 427: 999–1022.

5. Selkrig, J., D.L. Leyton, C.T. Webb, and T. Lithgow. 2014. Assembly of β-barrel proteins into bacterial outer membranes. Biochim. Biophys. Acta - Mol. Cell Res. 1843: 1542–1550.

6. Popot, J.L., and D.M. Engelman. 1990. Membrane protein folding and oligomerization: the two-stage model. Biochemistry. 29: 4031–7.

7. Capponi, S., M. Heyden, A.-N. Bondar, D.J. Tobias, and S.H. White. 2015. Anomalous behavior of water inside the SecY translocon. Proc. Natl. Acad. Sci. 112: 9016–9021.

8. Woolhead, C.A., P.J. McCormick, and A.E. Johnson. 2004. Nascent membrane and secretory proteins differ in FRET-detected folding far inside the ribosome and in their exposure to ribosomal proteins. Cell. 116: 725–736.

9. Gogala, M., T. Becker, B. Beatrix, J.P. Armache, C. Barrio-Garcia, O. Berninghausen, and R. Beckmann. 2014. Structures of the Sec61 complex engaged in nascent peptide translocation or membrane insertion. Nature. 506: 107–110.

10. Bowie, J.U. 1997. Helix packing in membrane proteins. J. Mol. Biol. 272: 780–789.

11. Engelman, D.M., Y. Chen, C.N. Chin, A.R. Curran, A.M. Dixon, A.D. Dupuy, A.S. Lee, U. Lehnert, E.E. Matthews, Y.K. Reshetnyak, A. Senes, and J.L. Popot. 2003. Membrane protein folding: beyond the two stage model. FEBS Lett. 555: 122–125.

12. Popot, J.L., S.E. Gerchman, and D.M. Engelman. 1987. Refolding of bacteriorhodopsin in lipid bilayers. A thermodynamically controlled two-stage process. J. Mol. Biol. 198: 655–676.

13. Adair, B.D., and D.M. Engelman. 1994. Glycophorin A helical transmembrane domains dimerize in phospholipid bilayers: a resonance energy transfer study. Biochemistry. 33: 5539–44.

14. Schick, S., L. Chen, E. Li, J. Lin, I. Köper, and K. Hristova. 2010. Assembly of the M2 tetramer is strongly modulated by lipid chain length. Biophys. J. 99: 1810–1817.

15. Li, M., L.G. Reddy, R. Bennett, N.D. Silva, L.R. Jones, and D.D. Thomas. 1999. A fluorescence energy transfer method for analyzing protein oligomeric structure: application to phospholamban. Biophys. J. 76: 2587–99.

16. Ridge, K.D., S.S. Lee, and L.L. Yao. 1995. In vivo assembly of rhodopsin from expressed polypeptide fragments. Proc. Natl. Acad. Sci. U. S. A. 92: 3204–3208.

119

17. Rath, A., D. V Tulumello, and C.M. Deber. 2009. Peptide models of membrane protein folding. Biochemistry. 48: 3036–45.

18. Hong, H., S. Park, R.H.F. Jiménez, D. Rinehart, and L.K. Tamm. 2007. Role of aromatic side chains in the folding and thermodynamic stability of integral membrane proteins. J. Am. Chem. Soc. 129: 8320–8327.

19. Johnson, R.M., K. Hecht, and C.M. Deber. 2007. Aromatic and cation-pi interactions enhance helix-helix association in a membrane environment. Biochemistry. 46: 9208–14.

20. Sal-Man, N., D. Gerber, and Y. Shai. 2005. The identification of a minimal dimerization motif QXXS that enables homo- and hetero-association of transmembrane helices in vivo. J. Biol. Chem. 280: 27449–27457.

21. Bocharov, E. V., K.S. Mineev, P.E. Volynsky, Y.S. Ermolyuk, E.N. Tkach, A.G. Sobol, V. V. Chupin, M.P. Kirpichnikov, R.G. Efremov, and A.S. Arseniev. 2008. Spatial structure of the dimeric transmembrane domain of the growth factor receptor ErbB2 presumably corresponding to the receptor active state. J. Biol. Chem. 283: 6950–6956.

22. Call, M.E., J. Pyrdol, M. Wiedmann, and K.W. Wucherpfennig. 2002. The organizing principle in the formation of the T cell receptor-CD3 complex. Cell. 111: 967–979.

23. Call, M.E., K.W. Wucherpfennig, and J.J. Chou. 2010. The structural basis for intramembrane assembly of an activating immunoreceptor complex. Nat. Immunol. 11: 1023–1029.

24. MacKenzie, K.R., J.H. Prestegard, and D.M. Engelman. 1997. A transmembrane helix dimer: structure and implications. Science. 276: 131–3.

25. Herrmann, J.R., A. Fuchs, J.C. Panitz, T. Eckert, S. Unterreitmeier, D. Frishman, and D. Langosch. 2010. Ionic interactions promote transmembrane helix-helix association depending on sequence context. J. Mol. Biol. 396: 452–461.

26. Simmerman, H.K.B., Y.M. Kobayashi, J.M. Autry, and L.R. Jones. 1996. A leucine zipper stabilizes the pentameric membrane domain of phospholamban and forms a coiled-coil pore structure. J. Biol. Chem. 271: 5941–5946.

27. King, G., and A.M. Dixon. 2010. Evidence for role of transmembrane helix–helix interactions in the assembly of the Class II major histocompatibility complex. Mol. Biosyst. 6: 1650.

28. Wilkins, C., and M. Gale. 2010. Recognition of viruses by cytoplasmic sensors. Curr. Opin. Immunol. 22: 41– 47.

29. Janeway, C.A., and R. Medzhitov. 2002. Innate immune recognition. Annu. Rev. Immunol. 20: 197–216.

30. Sato, S., C. St-Pierre, P. Bhaumik, and J. Nieminen. 2009. Galectins in innate immunity: Dual functions of host soluble β-galactoside-binding lectins as damage-associated molecular patterns (DAMPs) and as receptors for pathogen-associated molecular patterns (PAMPs). Immunol. Rev. 230: 172–187.

31. Aoshi, T., S. Koyama, K. Kobiyama, S. Akira, and K.J. Ishii. 2011. Innate and adaptive immune responses to viral infection and vaccination. Curr. Opin. Virol. 1: 226–232.

32. Anand, P.K., R.K. Subbarao Malireddi, J.R. Lukens, P. Vogel, J. Bertin, M. Lamkanfi, and T.D. Kanneganti. 2012. NLRP6 negatively regulates innate immunity and host defence against bacterial pathogens. Nature. 488: 389–393.

120

33. Braciale, T.J., and Y.S. Hahn. 2013. Immunity to viruses. Immunol. Rev. 255: 5–12.

34. Schneider, W.M., M.D. Chevillotte, and C.M. Rice. 2014. Interferon-Stimulated Genes: A Complex Web of Host Defenses. Annu. Rev. Immunol. 32: 513–545.

35. Domingo, E., J. Sheldon, and C. Perales. 2012. Viral Quasispecies Evolution. Microbiol. Mol. Biol. Rev. 76: 159–216.

36. Alcamí, A., J.A. Symons, A. Khanna, and G.L. Smith. 1998. Poxviruses: Capturing cytokines and chemokines. Semin. Virol. 8: 419–427.

37. Teodoro, J.G., and P.E. Branton. 1997. Regulation of apoptosis by viral gene products. J. Virol. 71: 1739– 1746.

38. González, M. 2015. Vpu Protein: The Viroporin Encoded by HIV-1. Viruses. 7: 4352–4368.

39. Barré-Sinoussi, F., J.C. Chermann, F. Rey, M.T. Nugeyre, S. Chamaret, J. Gruest, C. Dauguet, C. Axler-Blin, F. Vézinet-Brun, C. Rouzioux, W. Rozenbaum, and L. Montagnier. 1983. Isolation of a T-lymphotropic retrovirus from a patient at risk for acquired immune deficiency syndrome (AIDS). Science. 220: 868–71.

40. Gallo, R.C., S.Z. Salahuddin, M. Popovic, G.M. Shearer, M. Kaplan, B.F. Haynes, T.J. Palker, R. Redfield, J. Oleske, and B. Safai. 1984. Frequent detection and isolation of cytopathic retroviruses (HTLV-III) from patients with AIDS and at risk for AIDS. Science. 224: 500–3.

41. Popovic, M., M.G. Sarngadharan, E. Read, and R.C. Gallo. 1984. Detection, isolation, and continuous production of cytopathic retroviruses (HTLV-III) from patients with AIDS and pre-AIDS. Science. 224: 497– 500.

42. Clavel, F., D. Guétard, F. Brun-Vézinet, S. Chamaret, M.A. Rey, M.O. Santos-Ferreira, A.G. Laurent, C. Dauguet, C. Katlama, and C. Rouzioux. 1986. Isolation of a new human retrovirus from West African patients with AIDS. Science. 233: 343–6.

43. Huet, T., R. Cheynier, A. Meyerhans, G. Roelants, and S. Wain-Hobson. 1990. Genetic organization of a chimpanzee lentivirus related to HIV-1. Nature. 345: 356–359.

44. Gao, F., E. Balles, D.L. Robertson, Y. Chen, C.M. Rodenburg, S.F. Michael, L.B. Cummins, L.O. Arthur, M. Peeters, G.M. Shaw, P.M. Sharp, and B.H. Hahn. 1999. Origin of HIV-1 in the chimpanzee Pan troglodytes troglodytes. Nature. 397: 436–441.

45. Santiago, M.L., C.M. Rodenburg, S. Kamenya, F. Bibollet-Ruche, F. Gao, E. Bailes, S. Meleth, S.-J. Soong, J.M. Kilby, Z. Moldoveanu, B. Fahey, M.N. Muller, A. Ayouba, E. Nerrienet, H.M. McClure, J.L. Heeney, A.E. Pusey, D.A. Collins, C. Boesch, R.W. Wrangham, J. Goodall, P.M. Sharp, G.M. Shaw, and B.H. Hahn. 2002. SIVcpz in wild chimpanzees. Science. 295: 465.

46. Keele, B.F., F. Van Heuverswyn, Y. Li, E. Bailes, J. Takehisa, M.L. Santiago, F. Bibollet-Ruche, Y. Chen, L. V. Wain, F. Liegeois, S. Loul, E.M. Ngole, Y. Bienvenue, E. Delaporte, J.F.Y. Brookfield, P.M. Sharp, G.M. Shaw, M. Peeters, and B.H. Hahn. 2006. Chimpanzee reservoirs of pandemic and nonpandemic HIV-1. Science. 313: 523–6.

47. J., H. 2012. The origin and diversity of the HIV-1 pandemic. Trends Mol. Med. 18: 182–192.

48. Hahn, B.H., G.M. Shaw, K.M. De Cock, and P.M. Sharp. 2000. AIDS as a zoonosis: scientific and public health implications. Science. 287: 607–14. 121 49. De Leys, R., B. Vanderborght, M. Vanden Haeselvelde, L. Heyndrickx, A. Van Geel, C. Wauters, R. Bernaerts, E. Saman, P. Nijs, B. Willems, H. Taelman, G. van der Groen, P. Piot, T. Tersmette, J.G. Huisman, and H. van Heuverswyn. 1990. Isolation and partial characterization of an unusual human immunodeficiency retrovirus from two persons of west-central African Isolation and Partial Characterization of an Unusual Human Immunodeficiency Retrovirus from Two Persons of West-Central African. J. Virol. 64: 1207–1216.

50. Gürtler, L.G., P.H. Hauser, J. Eberle, A. von Brunn, S. Knapp, L. Zekeng, J.M. Tsague, and L. Kaptue. 1994. A new subtype of human immunodeficiency virus type 1 (MVP-5180) from Cameroon. J. Virol. 68: 1581–5.

51. Mauclère, P., I. Loussert-Ajaka, F. Damond, P. Fagot, S. Souquières, M. Monny Lobe, F.X. Mbopi Keou, F. Barré-Sinoussi, S. Saragosti, F. Brun-Vézinet, and F. Simon. 1997. Serological and virological characterization of HIV-1 group O infection in Cameroon. AIDS. 11: 445–53.

52. Simon, F., P. Mauclère, P. Roques, I. Loussert-Ajaka, M.C. Müller-Trutwin, S. Saragosti, M.C. Georges- Courbot, F. Barré-Sinoussi, and F. Brun-Vézinet. 1998. Identification of a new human immunodeficiency virus type 1 distinct from group M and group O. Nat. Med. 4: 1032–7.

53. Plantier, J.-C., M. Leoz, J.E. Dickerson, F. De Oliveira, F. Cordonnier, V. Lemée, F. Damond, D.L. Robertson, and F. Simon. 2009. A new human immunodeficiency virus derived from gorillas. Nat. Med. 15: 871–872.

54. Joseph, S.B., R. Swanstrom, A.D.M. Kashuba, and M.S. Cohen. 2015. Bottlenecks in HIV-1 transmission: insights from the study of founder viruses. Nat. Rev. Microbiol. 13: 414–25.

55. Schuitemaker, H., M. Koot, N.A. Kootstra, M.W. Dercksen, R.E. de Goede, R.P. van Steenwijk, J.M. Lange, J.K. Schattenkerk, F. Miedema, and M. Tersmette. 1992. Biological phenotype of human immunodeficiency virus type 1 clones at different stages of infection: progression of disease is associated with a shift from monocytotropic to T-cell-tropic virus population. J. Virol. 66: 1354–60.

56. Ho, S., N. Trunova, A. Gettie, J. Blanchard, and C. Cheng-Mayer. 2008. Different mutational pathways to CXCR4 coreceptor switch of CCR5-using simian-human immunodeficiency virus. J. Virol. 82: 5653–6.

57. Douek, D.C., L.J. Picker, and R.A. Koup. 2003. T cell dynamics in HIV-1 infection. Annu. Rev. Immunol. 21: 265–304.

58. Bellan, S.E., J. Dushoff, A.P. Galvani, L.A. Meyers, M. Markowitz, M. Vesanen, K. Tenner-Racz, Y. Cao, J.M. Binley, A. Talal, A. Hurley, X. Jin, M.R. Chaudhry, M. Yaman, S. Frankel, M. Heath-Chiozzi, J.M. Leonard, J.P. Moore, P. Racz, D.F. Nixon, D.D. Ho, E.W. Fiebig, D.J. Wright, B.D. Rawal, P.E. Garrett, R.T. Schumacher, L. Peddada, C. Heldebrant, R. Smith, A. Conrad, S.H. Kleinman, M.P. Busch, B.F. Keele, E.E. Giorgi, J.F. Salazar- Gonzalez, J.M. Decker, K.T. Pham, M.G. Salazar, C. Sun, T. Grayson, S. Wang, H. Li, X. Wei, C. Jiang, J.L. Kirchherr, F. Gao, J.A. Anderson, L.-H. Ping, R. Swanstrom, G.D. Tomaras, W.A. Blattner, P.A. Goepfert, J.M. Kilby, M.S. Saag, E.L. Delwart, M.P. Busch, M.S. Cohen, D.C. Montefiori, B.F. Haynes, B. Gaschen, G.S. Athreya, H.Y. Lee, N. Wood, C. Seoighe, A.S. Perelson, T. Bhattacharya, B.T. Korber, B.H. Hahn, G.M. Shaw, A. Schuetz, C. Deleage, I. Sereti, R. Rerknimitr, N. Phanuphak, Y. Phuang-Ngern, J.D. Estes, N.G. Sandler, S. Sukhumvittaya, M. Marovich, S. Jongrakthaitae, S. Akapirat, J.L.K. Fletscher, E. Kroon, R. Dewar, R. Trichavaroj, N. Chomchey, D.C. Douek, R.J. O???Connell, V. Ngauy, M.L. Robb, P. Phanuphak, N.L. Michael, J.L. Excler, J.H. Kim, M.S. de Souza, J. Ananworanich, M.S. Cohen, C.L. Gay, et al. 2003. Dynamics of HIV viremia and antibody seroconversion in plasma donors: implications for diagnosis and staging of primary HIV infection. Aids. 17: 1871–1879.

59. Moulard, M., S. Hallenberger, W. Garten, and H.D. Klenk. 1999. Processing and routage of HIV glycoproteins by furin to the cell surface. Virus Res. 60: 55–65.

60. Malim, M.H., and M. Emerman. 2008. HIV-1 accessory proteins--ensuring viral survival in a hostile 122 environment. Cell Host Microbe. 3: 388–98.

61. Chan, D.C., D. Fass, J.M. Berger, and P.S. Kim. 1997. Core structure of gp41 from the HIV envelope glycoprotein. Cell. 89: 263–273.

62. Weissenhorn, W., A. Dessen, S.C. Harrison, J.J. Skehel, and D.C. Wiley. 1997. Atomic structure of the ectodomain from HIV-1 gp41. Nature. 387: 426–430.

63. Kwong, P.D., R. Wyatt, J. Robinson, R.W. Sweet, J. Sodroski, and W.A. Hendrickson. 1998. Structure of an HIV gp 120 envelope glycoprotein in complex with the CD4 receptor and a neutralizing human antibody. Nature. 393: 648–659.

64. Ofek, G.A., P. Zhu, J. Liu, J.B. Jr, E. Chertova, J.D. Lifson, H. Grise, K.A. Taylor, and K.H. Roux. 2006. Distribution and three-dimensional structure of AIDS virus envelope spikes. 441: 847–852.

65. Ott, D.E. 2008. Cellular proteins detected in HIV-1. Rev. Med. Virol. 18: 159–175.

66. Ganser, B.K., S. Li, V.Y. Klishko, J.T. Finch, and W.I. Sundquist. 1999. Assembly and analysis of conical models for the HIV-1 core. Science (80-. ). 283: 80–83.

67. Briggs, J.A.G., T. Wilk, R. Welker, H.G. Kräusslich, and S.D. Fuller. 2003. Structural organization of authentic, mature HIV-1 virions and cores. EMBO J. 22: 1707–1715.

68. Wilk, T., I. Gross, B.E. Gowen, and T. Rutten. 2001. Organization of immature human immunodeficiency virus type 1. J. Virol. 72: 759–771.

69. Engelman, A., and P. Cherepanov. 2012. The structural biology of HIV-1: Mechanistic and therapeutic insights. Nat. Rev. Microbiol. 10: 279–290.

70. McDougal, J.S., M.S. Kennedy, J.M. Sligh, S.P. Cort, A. Mawle, and J.K. Nicholson. 1986. Binding of HTLV- III/LAV to T4+ T cells by a complex of the 110K viral protein and the T4 molecule. Science. 231: 382–5.

71. Berger, E.A., R.W. Doms, E.-M. Fenyö, B.T.M. Korber, D.R. Littman, J.P. Moore, Q.J. Sattentau, H. Schuitemaker, J. Sodroski, and R.A. Weiss. 1998. A new classification for HIV-1. Nature. 391: 240–240.

72. Hu, W.-S., and S.H. Hughes. 2012. HIV-1 reverse transcription. Cold Spring Harb. Perspect. Med. 2: a006882.

73. Fassati, A. 2006. HIV infection of non-dividing cells: A divisive problem. Retrovirology. 3: 1–15.

74. Goff, S.P. 2001. Intracellular trafficking of retroviral genomes during the early phase of infection: Viral exploitation of cellular pathways. J. Gene Med. 3: 517–528.

75. Craigie, R., and F.D. Bushman. 2012. HIV DNA integration. Cold Spring Harb. Perspect. Med. 2: a006890.

76. Schröder, A.R.W., P. Shinn, H. Chen, C. Berry, J.R. Ecker, F. Bushman, J.D. Boeke, S.E. Devine, P.O. Brown, B. Bowerman, H.E. Varmus, J.M. Bishop, M.I. Bukrinsky, N. Sharova, T.L. McDonald, T. Pushkarskaya, G.W. Tarpley, M. Stevenson, W.D. Burke, D.G. Eickbush, Y. Xiong, J. Jakubczak, T.H. Eickbush, F.D. Bushman, F.D. Bushman, F.D. Bushman, R. Craigie, S. Butler, M. Hansen, F.D. Bushman, S. Carteau, C. Hoffmann, F.D. Bushman, J.M. Coffin, S.H. Hughes, H.E. Varmus, J. Corbeil, D. Sheeter, D. Genini, S. Rought, L. Leoni, P. Du, M. Ferguson, D.R. Masys, J.B. Welsh, J.L. Fink, E. Al., C.B. Davis, I. Dikic, D. Unutmaz, C.M. Hill, J. Arthos, M.A. Siani, D.A. Thompson, J. Schlessinger, D.R. Littman, V.H. Ellison, H. Abrams, T. Roe, J. Lifson, P.O. Brown, C.M. Farnet, W.A. Haseltine, C. Farnet, F.D. Bushman, A. Follenzi, L.E. Ailes, S. Bakovic, M. Gueuna, 123

L. Naldini, N. Gaiano, A. Amsterdam, K. Kawakami, M. Allende, T. Becker, N. Hopkins, P. Gallay, S. Swingler, J. Song, F. Bushman, D. Trono, G.K. Geiss, R.E. Bumgarner, M.C. An, M.B. Agy, A.B. van’t Wout, E. Hammersmark, V.S. Carter, D. Upchurch, J.I. Mullins, M.G. Katze, M.S.T. Hansen, G.J.I. Smith, T. Kafri, V. Molteni, J.S. Siegel, F.D. Bushman, S. Hartung, et al. 2002. HIV-1 integration in the human genome favors active genes and local hotspots. Cell. 110: 521–9.

77. Karn, J., and C.M. Stoltzfus. 2012. Transcriptional and posttranscriptional regulation of HIV-1 gene expression. Cold Spring Harb. Perspect. Med. 2: a006916.

78. Yankulov, K., and D. Bentley. 1998. Transcriptional control: Tat cofactors and transcriptional elongation. Curr. Biol. 8: R447-9.

79. Gheysen, D., E. Jacobs, F. de Foresta, C. Thiriart, M. Francotte, D. Thines, and M. De Wilde. 1989. Assembly and release of HIV-1 precursor Pr55gagvirus-like particles from recombinant baculovirus-infected insect cells. Cell. 59: 103–112.

80. Göttlinger, H.G. 2001. The HIV-1 assembly machine. AIDS. 15 Suppl 5: S13-20.

81. Jouvenet, N., P.D. Bieniasz, and S.M. Simon. 2008. Imaging the biogenesis of individual HIV-1 virions in live cells. Nature. 454: 236–240.

82. Jouvenet, N., S.J.D. Neil, C. Bess, M.C. Johnson, C.A. Virgen, S.M. Simon, and P.D. Bieniasz. 2006. Plasma membrane is the site of productive HIV-1 particle assembly. PLoS Biol. 4: 2296–2310.

83. Saad, J.S., J. Miller, J. Tai, A. Kim, R.H. Ghanam, and M.F. Summers. 2006. Structural basis for targeting HIV- 1 Gag proteins to the plasma membrane for virus assembly. Proc. Natl. Acad. Sci. U. S. A. 103: 11364–9.

84. Ono, A., S.D. Ablan, S.J. Lockett, K. Nagashima, and E.O. Freed. 2004. Phosphatidylinositol (4,5) bisphosphate regulates HIV-1 Gag targeting to the plasma membrane. Proc. Natl. Acad. Sci. 101: 14889– 14894.

85. Tokarev, A., and J. Guatelli. 2011. Misdirection of membrane trafficking by HIV-1 Vpu and Nef. 1: 90–102.

86. Garrus, J.E., U.K. Von Schwedler, O.W. Pornillos, S.G. Morham, K.H. Zavitz, H.E. Wang, D.A. Wettstein, K.M. Stray, M. Côté, R.L. Rich, D.G. Myszka, and W.I. Sundquist. 2001. Tsg101 and the vacuolar protein sorting pathway are essential for HIV-1 budding. Cell. 107: 55–65.

87. Martin-Serrano, J., T. Zang, and P.D. Bieniasz. 2001. HIV-1 and Ebola virus encode small peptide motifs that recruit Tsg101 to sites of particle assembly to facilitate egress. Nat. Med. 7: 1313–1319.

88. Pornillos, O., B.K. Ganser-Pornillos, and M. Yeager. 2011. Atomic-level modelling of the HIV capsid. Nature. 469: 424–427.

89. Yeager, M. 2011. Design of in vitro symmetric complexes and analysis by hybrid methods reveal mechanisms of HIV capsid assembly. J. Mol. Biol. 410: 534–552.

90. Piacenti, F.J. 2006. An update and review of antiretroviral therapy. Pharmacotherapy. 26: 1111–1133.

91. Autran, B., G. Carcelain, T.S. Li, C. Blanc, D. Mathez, R. Tubiana, C. Katlama, P. Debré, and J. Leibowitch. 1997. Positive effects of combined antiretroviral therapy on CD4+ T cell homeostasis and function in advanced HIV disease. Science. 277: 112–6.

92. Komanduri, K. V, M.N. Viswanathan, E.D. Wieder, D.K. Schmidt, B.M. Bredt, M.A. Jacobson, and J.M. 124

McCune. 1998. Restoration of cytomegalovirus-specific CD4+ T-lymphocyte responses after ganciclovir and highly active antiretroviral therapy in individuals infected with HIV-1. Nat. Med. 4: 953–6.

93. Siliciano, R.F., and W.C. Greene. 2011. HIV latency. Cold Spring Harb. Perspect. Med. 1: 1–20.

94. Hütter, G., D. Nowak, M. Mossner, S. Ganepola, A. Müssig, K. Allers, T. Schneider, J. Hofmann, C. Kücherer, O. Blau, I.W. Blau, W.K. Hofmann, and E. Thiel. 2009. Long-term control of HIV by CCR5 Delta32/Delta32 stem-cell transplantation. N. Engl. J. Med. 360: 692–8.

95. Hütter, G., and E. Thiel. 2011. Allogeneic transplantation of CCR5-deficient progenitor cells in a patient with HIV infection: An update after 3 years and the search for patient no. 2. Aids. 25: 273–274.

96. Shin, S.Y. 2016. Recent update in HIV vaccine development. Clin. Exp. Vaccine Res. 5: 6.

97. Kirchhoff, F., T.C. Greenough, D.B. Brettler, J.L. Sullivan, and R.C. Desrosiers. 1995. Brief report: absence of intact nef sequences in a long-term survivor with nonprogressive HIV-1 infection. N. Engl. J. Med. 332: 228–32.

98. Roeth, J.F., M.R. Kasper, T.M. Filzen, and K.L. Collins. 2004. HIV-1 Nef disrupts MHC-I trafficking by recruiting AP-1 to the MHC-I cytoplasmic tail. J. Cell Biol. 167: 903–913.

99. Noviello, C.M., S. Benichou, and J.C. Guatelli. 2008. Cooperative binding of the class I major histocompatibility complex cytoplasmic domain and human immunodeficiency virus type 1 Nef to the endosomal AP-1 complex via its mu subunit. J. Virol. 82: 1249–58.

100. Atkins, K.M., L. Thomas, R.T. Youker, M.J. Harriff, F. Pissani, H. You, and G. Thomas. 2008. HIV-1 Nef binds PACS-2 to assemble a multikinase cascade that triggers major histocompatibility complex class I (MHC-I) down-regulation: Analysis using short interfering RNA and knock-out mice. J. Biol. Chem. 283: 11772– 11784.

101. Lubben, N.B., D.A. Sahlender, A.M. Motley, P.J. Lehner, P. Benaroch, and M.S. Robinson. 2007. HIV-1 Nef- induced down-regulation of MHC class I requires AP-1 and clathrin but not PACS-1 and is impeded by AP-2. Mol. Biol. Cell. 18: 3351–65.

102. Wolf, D., V. Witte, B. Laffert, K. Blume, E. Stromer, S. Trapp, P. D’aloja, A. Schürmann, and A.S. Baur. 2001. HIV-1 Nef associated PAK and PI3-kinases stimulate Akt-independent bad-phosphorylation to induce anti- apoptotic signals. Nat. Med. 7: 1217–1224.

103. Sol-Foulon, N., A. Moris, C. Nobile, C. Boccaccio, A. Engering, J.P. Abastado, J.M. Heard, Y. Van Kooyk, and O. Schwartz. 2002. HIV-1 Nef-induced upregulation of DC-SIGN in dendritic cells promotes lymphocyte clustering and viral spread. Immunity. 16: 145–155.

104. Matusali, G., M. Potesta, a. Santoni, C. Cerboni, and M. Doria. 2012. The Human Immunodeficiency Virus Type 1 Nef and Vpu Proteins Downregulate the Natural Killer Cell-Activating Ligand PVR. J. Virol. 86: 4496– 4504.

105. Sheehy, A.M., N.C. Gaddis, J.D. Choi, and M.H. Malim. 2002. Isolation of a human gene that inhibits HIV-1 infection and is suppressed by the viral Vif protein. Nature. 418: 646–650.

106. Sheehy, A.M., N.C. Gaddis, and M.H. Malim. 2003. The antiretroviral enzyme APOBEC3G is degraded by the proteasome in response to HIV-1 Vif. Nat. Med. 9: 1404–1407.

107. Marin, M., K.M. Rose, S.L. Kozak, and D. Kabat. 2003. HIV-1 Vif protein binds the editing enzyme APOBEC3G 125 and induces its degradation. Nat. Med. 9: 1398–1403.

108. Zimmerman, E.S., M.P. Sherman, J.L. Blackett, J.A. Neidleman, C. Kreis, P. Mundt, S.A. Williams, M. Warmerdam, J. Kahn, F.M. Hecht, R.M. Grant, C.M. de Noronha, A.S. Weyrich, W.C. Greene, and V. Planelles. 2006. Human immunodeficiency virus type 1 Vpr induces DNA replication stress in vitro and in vivo. J Virol. 80: 10407–10418.

109. Park, S.H., A.A. Mrse, A.A. Nevzorov, M.F. Mesleh, M. Oblatt-Montal, M. Montal, and S.J. Opella. 2003. Three-dimensional structure of the channel-forming trans-membrane domain of virus protein “u” (Vpu) from HIV-1. J. Mol. Biol. 333: 409–424.

110. Maldarelli, F., M.Y. Chen, R.L. Willey, and K. Strebel. 1993. Human immunodeficiency virus type 1 Vpu protein is an oligomeric type I integral membrane protein. J. Virol. 67: 5056–5061.

111. Matheson, N., K. Wals, M. Weekes, R. Rapiteanu, R. Vigan, R. Antrobus, D. Smith, S. Neil, and P. Lehner. 2015. Antagonism of aminoacid transport in primary CD4 T cells by HIV-1 Vpu. Lancet. 385 Suppl: S66.

112. D. Ewart, K.M.G.B.C.G. 2002. Amiloride derivatives block ion channel activity and enhancement of virus-like particle budding caused by HIV-1 protein Vpu. Eur. Biophys. J. 31: 26–35.

113. Schubert, U., A. V. Ferrer-Montiel, M. Oblatt-Montal, P. Henklein, K. Strebel, and M. Montal. 1996. Identification of an ion channel activity of the Vpu transmembrane domain and its involvement in the regulation of virus release from HIV-1-infected cells. FEBS Lett. 398: 12–18.

114. Van Damme, N., D. Goff, C. Katsura, R.L. Jorgenson, R. Mitchell, M.C. Johnson, E.B. Stephens, and J. Guatelli. 2008. The interferon-induced protein BST-2 restricts HIV-1 release and is downregulated from the cell surface by the viral Vpu protein. Cell Host Microbe. 3: 245–52.

115. Moore, P.B., Q. Zhong, T. Husslein, and M.L. Klein. 1998. Simulation of the HIV-1 Vpu transmembrane domain as a pentameric bundle. FEBS Lett. 431: 143–148.

116. Lopez, C.F., M. Montal, J. Kent Blasie, M.L. Klein, and P.B. Moore. 2002. Molecular dynamics investigation of membrane-bound bundles of the channel-forming transmembrane domain of viral protein U from the human immunodeficiency virus HIV-1. Biophys. J. 83: 1259–1267.

117. Lu, J.X., S. Sharpe, R. Ghirlando, W.M. Yau, and R. Tycko. 2010. Oligomerization state and supramolecular structure of the HIV-1 Vpu protein transmembrane segment in phospholipid bilayers. Protein Sci. 19: 1877–1896.

118. Chen, C.-P., M.-H. Lin, Y.-T. Chan, L.-C. Chen, C. Ma, and W.B. Fischer. 2016. Membrane protein assembly: two cytoplasmic phosphorylated serine sites of Vpu from HIV-1 affect oligomerization. Sci. Rep. 6: 28866.

119. Zhang, H., E.C. Lin, B.B. Das, Y. Tian, and S.J. Opella. 2015. Structural determination of virus protein U from HIV-1 by NMR in membrane environments. Biochim. Biophys. Acta - Biomembr. 1848: 3007–3018.

120. Lewinski, M.K., M. Jafari, H. Zhang, S.J. Opella, and J. Guatelli. 2015. Membrane anchoring by a C-terminal tryptophan enables HIV-1 Vpu to displace bone marrow stromal antigen 2 (BST2) from sites of viral assembly. J. Biol. Chem. 290: 10919–33.

121. Singh, D.K., D.M. Griffin, E. Pacyniak, M. Jackson, M.J. Werle, B. Wisdom, F. Sun, D.R. Hout, D.M. Pinson, R.S. Gunderson, M.F. Powers, S.W. Wong, and E.B. Stephens. 2003. The presence of the casein kinase II phosphorylation sites of Vpu enhances the CD4+ T cell loss caused by the simian-human immunodeficiency virus SHIVKU-lbMC33 in pig-tailed macaques. Virology. 313: 435–451. 126

122. Willbold, D., S. Hoffmann, and P. Rosch. 1997. Secondary structure and tertiary fold of the human immunodeficiency virus protein U ( Vpu ) cytoplasmic domain in solution. 588: 581–588.

123. Varthakavi, V., R.M. Smith, K.L. Martin, A. Derdowski, L.A. Lapierre, J.R. Goldenring, and P. Spearman. 2006. The pericentriolar recycling endosome plays a key role in Vpu-mediated enhancement of HIV-1 particle release. Traffic. 7: 298–307.

124. Dubé, M., B.B. Roy, P. Guiot-Guillain, J. Mercier, J. Binette, G. Leung, and E.A. Cohen. 2009. Suppression of Tetherin-restricting activity upon human immunodeficiency virus type 1 particle release correlates with localization of Vpu in the trans-Golgi network. J. Virol. 83: 4574–90.

125. Paul, M., and M.A. Jabbar. 1997. Phosphorylation of both phosphoacceptor sites in the HIV-1 Vpu cytoplasmic domain is essential for Vpu-mediated ER degradation of CD4. Virology. 232: 207–216.

126. Kobayashi, T., H. Ode, T. Yoshida, K. Sato, P. Gee, S.P. Yamamoto, H. Ebina, K. Strebel, H. Sato, and Y. Koyanagi. 2011. Identification of amino acids in the human tetherin transmembrane domain responsible for HIV-1 Vpu interaction and susceptibility. J. Virol. 85: 932–45.

127. Rong, L., J. Zhang, J. Lu, Q. Pan, R.-P. Lorgeoux, C. Aloysius, F. Guo, S.-L. Liu, M. a Wainberg, and C. Liang. 2009. The transmembrane domain of BST-2 determines its sensitivity to down-modulation by human immunodeficiency virus type 1 Vpu. J. Virol. 83: 7536–46.

128. McNatt, M.W., T. Zang, T. Hatziioannou, M. Bartlett, I. Ben Fofana, W.E. Johnson, S.J.D. Neil, and P.D. Bieniasz. 2009. Species-specific activity of HIV-1 Vpu and positive selection of tetherin transmembrane domain variants. PLoS Pathog. 5: e1000300.

129. Vigan, R., and S.J.D. Neil. 2010. Determinants of tetherin antagonism in the transmembrane domain of the human immunodeficiency virus type 1 Vpu protein. J. Virol. 84: 12958–12970.

130. Douglas, J.L., K. Viswanathan, M.N. McCarroll, J.K. Gustin, K. Früh, and A. V Moses. 2009. Vpu directs the degradation of the human immunodeficiency virus restriction factor BST-2/Tetherin via a βTrCP-dependent mechanism. J. Virol. 83: 7931–47.

131. Mitchell, R.S., C. Katsura, M. a Skasko, K. Fitzpatrick, D. Lau, A. Ruiz, E.B. Stephens, F. Margottin-Goguet, R. Benarous, and J.C. Guatelli. 2009. Vpu antagonizes BST-2–mediated restriction of HIV-1 release via β-TrCP and endo-lysosomal trafficking. PLoS Pathog. 5: e1000450.

132. Mangeat, B., G. Gers-Huber, M. Lehmann, M. Zufferey, J. Luban, and V. Piguet. 2009. HIV-1 Vpu neutralizes the antiviral factor tetherin/BST-2 by binding it and directing its Beta-TrCP2-dependent degradation. PLoS Pathog. 5: e1000574.

133. Coady, M.J., N.G. Daniel, E. Tiganos, B. Allain, J. Friborg, J.Y. Lapointe, and E. a Cohen. 1998. Effects of Vpu expression on Xenopus oocyte membrane conductance. Virology. 244: 39–49.

134. Ewart, G.D., T. Sutherland, P.W. Gage, and G.B. Cox. 1996. The Vpu protein of human immunodeficiency virus type 1 forms cation-selective ion channels. J. Virol. 70: 7108–15.

135. Willey, R.L., F. Maldarelli, M. a Martin, and K. Strebel. 1992. Human immunodeficiency virus type 1 Vpu protein induces rapid degradation of CD4. J. Virol. 66: 7193–200.

136. Jia, X., E. Weber, A. Tokarev, M. Lewinski, M. Rizk, M. Suarez, J. Guatelli, and Y. Xiong. 2014. Structural basis of HIV-1 Vpu-mediated BST2 antagonism via hijacking of the clathrin adaptor protein complex 1. Elife. 3: 1–24. 127

137. Sauter, D., D. Unterweger, M. Vogl, S.M. Usmani, A. Heigele, S.F. Kluge, E. Hermkes, M. Moll, E. Barker, M. Peeters, G.H. Learn, F. Bibollet-Ruche, J. V Fritz, O.T. Fackler, B.H. Hahn, and F. Kirchhoff. 2012. Human tetherin exerts strong selection pressure on the HIV-1 group N Vpu protein. PLoS Pathog. 8: e1003093.

138. Vigan, R., and S.J.D. Neil. 2011. Separable determinants of subcellular localization and interaction account for the inability of group O HIV-1 Vpu to counteract tetherin. J. Virol. 85: 9737–48.

139. Sauter, D., S. Hué, S. Petit, J. Plantier, G. Towers, F. Kirchhoff, and R. Gupta. 2011. HIV-1 Group P is unable to antagonize human tetherin by Vpu, Env or Nef. Retrovirology. 8: 103.

140. Yang, S.J., L.A. Lopez, C.M. Exline, K.G. Haworth, and P.M. Cannon. 2011. Lack of adaptation to human tetherin in HIV-1 Group O and P. Retrovirology. 8: 1–18.

141. Kirchhoff, F. 2010. Immune evasion and counteraction of restriction factors by HIV-1 and other primate lentiviruses. Cell Host Microbe. 8: 55–67.

142. Douglas, J.L., J.K. Gustin, K. Viswanathan, M. Mansouri, A. V Moses, and K. Früh. 2010. The great escape: viral strategies to counter BST-2/tetherin. PLoS Pathog. 6: e1000913.

143. Schubert, U., S. Bour, A. V Ferrer-Montiel, M. Montal, F. Maldarell, and K. Strebel. 1996. The two biological activities of human immunodeficiency virus type 1 Vpu protein involve two separable structural domains. J. Virol. 70: 809–19.

144. Tiganos, E., X.J. Yao, J. Friborg, N. Daniel, and E.A. Cohen. 1997. Putative alpha-helical structures in the human immunodeficiency virus type 1 Vpu protein and CD4 are involved in binding and degradation of the CD4 molecule. J. Virol. 71: 4452–60.

145. Bour, S., U. Schubert, and K. Strebel. 1995. The human immunodeficiency virus type 1 Vpu protein specifically binds to the cytoplasmic domain oBour, S., Schubert, U. & Strebel, K., 1995. The human immunodeficiency virus type 1 Vpu protein specifically binds to the cytoplasmic domain of CD4: implicat. J. Virol. 69: 1510–20.

146. Lenburg, M.E., and N.R. Landau. 1993. Vpu-induced degradation of CD4: Requirement for specific amino acid residues in the cytoplasmic domain of CD4. J. Virol. 67: 7238–7245.

147. Vincent, M.J., N.U. Raja, and M.A. Jabbar. 1993. Human immunodeficiency virus type 1 Vpu protein induces degradation of chimeric envelope glycoproteins bearing the cytoplasmic and anchor domains of CD4: role of the cytoplasmic domain in Vpu-induced degradation in the endoplasmic reticulum. J. Virol. 67: 5538– 5549.

148. Willey, R.L., A. Buckler-White, and K. Strebel. 1994. Sequences present in the cytoplasmic domain of CD4 are necessary and sufficient to confer sensitivity to the human immunodeficiency virus type 1 Vpu protein. J. Virol. 68: 1207–12.

149. Yao, X.J., J. Friborg, F. Checroune, S. Gratton, F. Boisvert, R.P. Sékaly, and E.A. Cohen. 1995. Degradation of CD4 induced by human immunodeficiency virus type 1 Vpu protein: a predicted alpha-helix structure in the proximal cytoplasmic region of CD4 contributes to Vpu sensitivity. Virology. 209: 615–23.

150. Lama, J., A. Mangasarian, and D. Trono. 1999. Cell-surface expression of CD4 reduces HIV-1 infectivity by blocking Env incorporation in a Nef- and Vpu-inhibitable manner. Curr. Biol. 9: 622–631.

151. Wildum, S., M. Schindler, J. Munch, and F. Kirchhoff. 2006. Contribution of Vpu, Env, and Nef to CD4 downmodulation and resistance of human immunodeficiency virus type 1-infected T cells to superinfection. J. 128

Virol. 80: 8047–8059.

152. Cortés, M.J., F. Wong-Staal, and J. Lama. 2002. Cell surface CD4 interferes with the infectivity of HIV-1 particles released from T cells. J. Biol. Chem. 277: 1770–1779.

153. Levesque, K., Y. Sen Zhao, and É.A. Cohen. 2003. Vpu exerts a positive effect on HIV-1 infectivity by down- modulating CD4 receptor molecules at the surface of HIV-1-producing cells. J. Biol. Chem. 278: 28346– 28353.

154. Tanaka, M., T. Ueno, T. Nakahara, K. Sasaki, A. Ishimoto, and H. Sakai. 2003. Downregulation of CD4 is required for maintenance of viral infectivity of HIV-1. Virology. 311: 316–325.

155. Bour, S., C. Perrin, H. Akari, and K. Strebel. 2001. The human immunodeficiency virus type 1 Vpu protein inhibits NF-kappa B activation by interfering with beta TrCP-mediated degradation of Ikappa B. J. Biol. Chem. 276: 15920–8.

156. Bächle, S.M., D. Sauter, S. Sibitz, J.K. Sandberg, F. Kirchhoff, and M. Moll. 2015. Involvement of a C-terminal motif in the interference of primate lentiviral Vpu proteins with CD1d-mediated antigen presentation. Sci. Rep. 5: 9675.

157. Le Noury, D. a., S. Mosebi, M. a. Papathanasopoulos, and R. Hewer. 2015. Functional roles of HIV-1 Vpu and CD74: Details and implications of the Vpu–CD74 interaction. Cell. Immunol. 298: 25–32.

158. Hussain, A., C. Wesley, M. Khalid, A. Chaudhry, and S. Jameel. 2008. Human immunodeficiency virus type 1 Vpu protein interacts with CD74 and modulates major histocompatibility complex class II presentation. J. Virol. 82: 893–902.

159. Neil, S.J.D., T. Zang, and P.D. Bieniasz. 2008. Tetherin inhibits retrovirus release and is antagonized by HIV-1 Vpu. Nature. 451: 425–30.

160. Jouvenet, N., S.J.D. Neil, M. Zhadina, T. Zang, Z. Kratovac, Y. Lee, M. McNatt, T. Hatziioannou, and P.D. Bieniasz. 2009. Broad-spectrum inhibition of retroviral and filoviral particle release by tetherin. J. Virol. 83: 1837–44.

161. Sakuma, T., T. Noda, S. Urata, Y. Kawaoka, and J. Yasuda. 2009. Inhibition of Lassa and Marburg virus production by tetherin. J. Virol. 83: 2382–5.

162. Matsuda, A., Y. Suzuki, G. Honda, S. Muramatsu, O. Matsuzaki, Y. Nagano, T. Doi, K. Shimotohno, T. Harada, E. Nishida, H. Hayashi, and S. Sugano. 2003. Large-scale identification and characterization of human genes that activate NF-κB and MAPK signaling pathways. Oncogene. 22: 3307–3318.

163. Galão, R.P., A. Le Tortorec, S. Pickering, T. Kueck, and S.J.D. Neil. 2012. Innate sensing of HIV-1 assembly by tetherin induces NFκB-dependent proinflammatory responses. Cell Host Microbe. 12: 633–644.

164. Tokarev, A., M. Suarez, W. Kwan, K. Fitzpatrick, R. Singh, and J. Guatelli. 2013. Stimulation of NF-κB activity by the HIV restriction factor BST2. J. Virol. 87: 2046–57.

165. Rollason, R., V. Korolchuk, C. Hamilton, P. Schu, and G. Banting. 2007. Clathrin-mediated endocytosis of a lipid-raft-associated protein is mediated through a dual tyrosine motif. J. Cell Sci. 120: 3850–8.

166. Kupzig, S., V. Korolchuk, R. Rollason, A. Sugden, A. Wilde, and G. Banting. 2003. Bst-2/HM1.24 is a raft- associated apical membrane protein with an unusual topology. Traffic. 4: 694–709.

129

167. Ishikawa, J., T. Kaisho, H. Tomizawa, B.O. Lee, Y. Kobune, J. Inazawa, K. Oritani, M. Itoh, T. Ochi, and K. Ishihara. 1995. Molecular cloning and chromosomal mapping of a bone marrow stromal cell surface gene, BST2, that may be involved in pre-B-cell growth. Genomics. 26: 527–34.

168. Andrew, A.J., E. Miyagi, S. Kao, and K. Strebel. 2009. The formation of cysteine-linked dimers of BST- 2/tetherin is important for inhibition of HIV-1 virus release but not for sensitivity to Vpu. Retrovirology. 6: 80.

169. Ohtomo, T., Y. Sugamata, Y. Ozaki, K. Ono, Y. Yoshimura, S. Kawai, Y. Koishihara, S. Ozaki, M. Kosaka, T. Hirano, and M. Tsuchiya. 1999. Molecular cloning and characterization of a surface antigen preferentially overexpressed on multiple myeloma cells. Biochem. Biophys. Res. Commun. 258: 583–91.

170. Perez-Caballero, D., T. Zang, A. Ebrahimi, M.W. McNatt, D. a Gregory, M.C. Johnson, and P.D. Bieniasz. 2009. Tetherin inhibits HIV-1 release by directly tethering virions to cells. Cell. 139: 499–511.

171. Iwabu, Y., H. Fujita, M. Kinomoto, K. Kaneko, Y. Ishizaka, Y. Tanaka, T. Sata, and K. Tokunaga. 2009. HIV-1 accessory protein Vpu internalizes cell-surface BST-2/tetherin through transmembrane interactions leading to lysosomes. J. Biol. Chem. 284: 35060–72.

172. Hinz, A., N. Miguet, G. Natrajan, Y. Usami, H. Yamanaka, P. Renesto, B. Hartlieb, A. a McCarthy, J.-P. Simorre, H. Göttlinger, and W. Weissenhorn. 2010. Structural basis of HIV-1 tethering to membranes by the BST-2/tetherin ectodomain. Cell Host Microbe. 7: 314–23.

173. Schubert, H.L., Q. Zhai, V. Sandrin, D.M. Eckert, M. Garcia-Maya, L. Saul, W.I. Sundquist, R. a Steiner, and C.P. Hill. 2010. Structural and functional studies on the extracellular domain of BST2/tetherin in reduced and oxidized conformations. Proc. Natl. Acad. Sci. U. S. A. 107: 17951–6.

174. Lehmann, M., S. Rocha, B. Mangeat, F. Blanchet, H. Uji-I, J. Hofkens, and V. Piguet. 2011. Quantitative multicolor super-resolution microscopy reveals tetherin HIV-1 interaction. PLoS Pathog. 7: e1002456.

175. Yang, H., J. Wang, X. Jia, M.W. McNatt, T. Zang, B. Pan, W. Meng, H.-W. Wang, P.D. Bieniasz, and Y. Xiong. 2010. Structural insight into the mechanisms of enveloped virus tethering by tetherin. Proc. Natl. Acad. Sci. U. S. A. 107: 18428–32.

176. Morita, E., and W.I. Sundquist. 2004. Retrovirus budding. Annu. Rev. Cell Dev. Biol. 20: 395–425.

177. Miyakawa, K., A. Ryo, T. Murakami, K. Ohba, S. Yamaoka, M. Fukuda, J. Guatelli, and N. Yamamoto. 2009. BCA2/Rabring7 promotes tetherin-dependent HIV-1 restriction. PLoS Pathog. 5: e1000700.

178. Rollason, R., V. Korolchuk, C. Hamilton, M. Jepson, and G. Banting. 2009. A CD317/tetherin-RICH2 complex plays a critical role in the organization of the subapical actin cytoskeleton in polarized epithelial cells. J. Cell Biol. 184: 721–736.

179. Cao, E., U. a. Ramagopal, A. Fedorov, E. Fedorov, Q. Yan, J.W. Lary, J.L. Cole, S.G. Nathenson, and S.C. Almo. 2006. NTB-A receptor crystal structure: insights into homophilic interactions in the signaling lymphocytic activation molecule receptor family. Immunity. 25: 559–570.

180. Lanier, L.L. 2008. Evolutionary struggles between NK cells and viruses. Nat. Rev. Immunol. 8: 259–268.

181. Shah, A.H., B. Sowrirajan, Z.B. Davis, J.P. Ward, E.M. Campbell, V. Planelles, and E. Barker. 2010. Degranulation of natural killer cells following interaction with HIV-1-infected cells is hindered by downmodulation of NTB-A by Vpu. Cell Host Microbe. 8: 397–409.

130

182. Bolduan, S., P. Hubel, T. Reif, V. Lodermeyer, K. Höhne, J. V. Fritz, D. Sauter, F. Kirchhoff, O.T. Fackler, M. Schindler, and U. Schubert. 2013. HIV-1 Vpu affects the anterograde transport and the glycosylation pattern of NTB-A. Virology. 440: 190–203.

183. Baury, B., R.J. Geraghty, D. Masson, P. Lustenberger, P.G. Spear, and M.G. Denis. 2001. Organization of the rat Tage4 gene and herpesvirus entry activity of the encoded protein. Gene. 265: 185–194.

184. Masson, D. 2001. Overexpression of the CD155 gene in human colorectal carcinoma. Gut. 49: 236–240.

185. Nixdorf, R., J. Schmidt, A. Karger, and T.C. Mettenleiter. 1999. Infection of Chinese hamster ovary cells by pseudorabies virus. J. Virol. 73: 8019–8026.

186. Siddique, T., R. McKinney, W.Y. Hung, R.J. Bartlett, G. Bruns, T.K. Mohandas, H.H. Ropers, C. Wilfert, and A.D. Roses. 1988. The poliovirus sensitivity (PVS) gene is on chromosome 19q12----q13.2. Genomics. 3: 156–60.

187. Takai, Y., J. Miyoshi, W. Ikeda, and H. Ogita. 2008. Nectins and nectin-like molecules: roles in contact inhibition of cell movement and proliferation. Nat. Rev. Mol. Cell Biol. 9: 603–615.

188. Mendelsohn, C.L., E. Wimmer, and V.R. Racaniello. 1989. Cellular receptor for poliovirus: Molecular cloning, nucleotide sequence, and expression of a new member of the immunoglobulin superfamily. Cell. 56: 855–865.

189. Oda, T., S. Ohka, and A. Nomoto. 2004. Ligand stimulation of CD155alpha inhibits cell adhesion and enhances cell migration in fibroblasts. Biochem. Biophys. Res. Commun. 319: 1253–64.

190. Koike, S., I. Ise, and A. Nomoto. 1991. Functional domains of the poliovirus receptor. Proc. Natl. Acad. Sci. 88: 4104–4108.

191. Strauss, M., D.J. Filman, D.M. Belnap, N. Cheng, R.T. Noel, and J.M. Hogle. 2015. Nectin-like interactions between poliovirus and its receptor trigger conformational changes associated with cell entry. J. Virol. 89: 4143–4157.

192. Baury, B., D. Masson, B.M. McDermott, A. Jarry, H.M. Blottière, P. Blanchardie, C.L. Laboisse, P. Lustenberger, V.R. Racaniello, and M.G. Denis. 2003. Identification of secreted CD155 isoforms. Biochem. Biophys. Res. Commun. 309: 175–182.

193. Koike, S., H. Horie, I. Ise, A. Okitsu, M. Yoshida, N. Iizuka, K. Takeuchi, T. Takegami, and A. Nomoto. 1990. The poliovirus receptor protein is produced both as membrane-bound and secreted forms. EMBO J. 9: 3217–24.

194. Iguchi-Manaka, A., G. Okumura, H. Kojima, Y. Cho, R. Hirochika, H. Bando, T. Sato, H. Yoshikawa, H. Hara, A. Shibuya, and K. Shibuya. 2016. Increased soluble CD155 in the serum of cancer patients. PLoS One. 11: 1– 12.

195. Harrison, O.J., J. Vendome, J. Brasch, X. Jin, S. Hong, P.S. Katsamba, G. Ahlsen, R.B. Troyanovsky, S.M. Troyanovsky, B. Honig, and L. Shapiro. 2012. Nectin ectodomain structures reveal a canonical adhesive interface. Nat. Struct. Mol. Biol. 19: 906–915.

196. Ohka, S., H. Ohno, K. Tohyama, and A. Nomoto. 2001. Basolateral sorting of human poliovirus receptor α involves an interaction with the μ1B subunit of the clathrin adaptor complex in polarized epithelial cells. Biochem. Biophys. Res. Commun. 287: 941–948.

131

197. Casado, J.G., G. Pawelec, S. Morgado, B. Sanchez-Correa, E. Delgado, I. Gayoso, E. Duran, R. Solana, and R. Tarazona. 2009. Expression of adhesion molecules and ligands for activating and costimulatory receptors involved in cell-mediated cytotoxicity in a large panel of human melanoma cell lines. Cancer Immunol. Immunother. 58: 1517–1526.

198. Chan, C.J., D.M. Andrews, and M.J. Smyth. 2012. Receptors that interact with nectin and nectin-like proteins in the immunosurveillance and immunotherapy of cancer. Curr. Opin. Immunol. 24: 246–251.

199. Gromeier, M., S. Lachmann, M.R. Rosenfeld, P.H. Gutin, and E. Wimmer. 2000. Intergeneric poliovirus recombinants for the treatment of malignant glioma. Proc. Natl. Acad. Sci. 97: 6803–6808.

200. Nakai, R., Y. Maniwa, Y. Tanaka, W. Nishio, M. Yoshimura, Y. Okita, C. Ohbayashi, N. Satoh, H. Ogita, Y. Takai, and Y. Hayashi. 2010. Overexpression of Necl-5 correlates with unfavorable prognosis in patients with lung adenocarcinoma. Cancer Sci. 101: 1326–1330.

201. Ochiai, H., S.A. Moore, G.E. Archer, T. Okamura, T.A. Chewning, J.R. Marks, J.H. Sampson, and M. Gromeier. 2004. Treatment of intracerebral neoplasia and neoplastic meningitis with regional delivery of oncolytic recombinant poliovirus. Clin. Cancer Res. 10: 4831–4838.

202. Reymond, N., A.-M. Imbert, E. Devilard, S. Fabre, C. Chabannon, L. Xerri, C. Farnarier, C. Cantoni, C. Bottino, A. Moretta, P. Dubreuil, and M. Lopez. 2004. DNAM-1 and PVR Regulate Monocyte Migration through Endothelial Junctions. J. Exp. Med. 199: 1331–1341.

203. Shibuya, A., D. Campbell, C. Hannum, H. Yssel, K. Franz-Bacon, T. McClanashan, T. Kitamura, J. Nicholl, G.R. Sutherland, L.L. Lanier, and J.H. Phillips. 1996. DNAM-1, a novel adhesion molecule involved in the cytolytic function of T lymphocytes. Immunity. 4: 573–581.

204. Neil, S.J.D. 2013. The antiviral activities of tetherin. Curr. Top. Microbiol. Immunol. 371: 67–104.

205. Pignoloni, B., C. Fionda, V. Dell’Oste, A. Luganini, M. Cippitelli, A. Zingoni, S. Landolfo, G. Gribaudo, A. Santoni, and C. Cerboni. 2016. Distinct Roles for Human Cytomegalovirus Immediate Early Proteins IE1 and IE2 in the Transcriptional Regulation of MICA and PVR/CD155 Expression. J. Immunol. 197: 4066–4078.

206. Zarama, A., N. Pérez-Carmona, D. Farré, A. Tomic, E.M. Borst, M. Messerle, S. Jonjic, P. Engel, and A. Angulo. 2014. Cytomegalovirus m154 hinders CD48 cell-surface expression and promotes viral escape from host natural killer cell control. PLoS Pathog. 10: e1004000.

207. Bour, S., and K. Strebel. 1996. The human immunodeficiency virus (HIV) type 2 envelope protein is a functional complement to HIV type 1 Vpu that enhances particle release of heterologous retroviruses. J. Virol. 70: 8285–300.

208. Le Tortorec, A., and S.J.D. Neil. 2009. Antagonism to and intracellular sequestration of human tetherin by the human immunodeficiency virus type 2 envelope glycoprotein. J. Virol. 83: 11966–78.

209. Abada, P., B. Noble, and P.M. Cannon. 2005. Functional domains within the human immunodeficiency virus type 2 envelope protein required to enhance virus production. J. Virol. 79: 3627–38.

210. Hauser, H., L. a Lopez, S.J. Yang, J.E. Oldenburg, C.M. Exline, J.C. Guatelli, and P.M. Cannon. 2011. HIV-1 Vpu and HIV-2 Env counteract BST-2/tetherin by sequestration in a perinuclear compartment. Retrovirology. 8: 85.

211. Lopez, L. a, S.J. Yang, H. Hauser, C.M. Exline, K.G. Haworth, J. Oldenburg, and P.M. Cannon. 2010. Ebola virus glycoprotein counteracts BST-2/Tetherin restriction in a sequence-independent manner that does not 132 require tetherin surface removal. J. Virol. 84: 7243–55.

212. Kühl, A., C. Banning, A. Marzi, J. Votteler, I. Steffen, S. Bertram, I. Glowacka, A. Konrad, M. Stürzl, J.-T. Guo, U. Schubert, H. Feldmann, G. Behrens, M. Schindler, and S. Pöhlmann. 2011. The Ebola virus glycoprotein and HIV-1 Vpu employ different strategies to counteract the antiviral factor tetherin. J. Infect. Dis. 204 Suppl: S850-60.

213. Mansouri, M., K. Viswanathan, J.L. Douglas, J. Hines, J. Gustin, A. V Moses, and K. Früh. 2009. Molecular mechanism of BST2/tetherin downregulation by K5/MIR2 of Kaposi’s sarcoma-associated herpesvirus. J. Virol. 83: 9672–81.

214. Pardieu, C., R. Vigan, S.J. Wilson, A. Calvi, T. Zang, P. Bieniasz, P. Kellam, G.J. Towers, and S.J.D. Neil. 2010. The RING-CH ligase K5 antagonizes restriction of KSHV and HIV-1 particle release by mediating ubiquitin- dependent endosomal degradation of tetherin. PLoS Pathog. 6: e1000843.

215. Yondola, M. a, F. Fernandes, A. Belicha-Villanueva, M. Uccelini, Q. Gao, C. Carter, and P. Palese. 2011. Budding capability of the influenza virus neuraminidase can be modulated by tetherin. J. Virol. 85: 2480– 91.

216. Mangeat, B., L. Cavagliotti, M. Lehmann, G. Gers-Huber, I. Kaur, Y. Thomas, L. Kaiser, and V. Piguet. 2012. Influenza virus partially counteracts restriction imposed by tetherin/BST-2. J. Biol. Chem. 287: 22015–29.

217. Blondeau, C., A. Pelchen-Matthews, P. Mlcochova, M. Marsh, R.S.B. Milne, and G.J. Towers. 2013. Tetherin restricts herpes simplex virus 1 and is antagonized by glycoprotein m. J. Virol. 87: 13124–33.

218. Jones, P.H., M. Maric, M.N. Madison, W. Maury, R.J. Roller, and C.M. Okeoma. 2013. BST-2/tetherin- mediated restriction of chikungunya (CHIKV) VLP budding is counteracted by CHIKV non-structural protein 1 (nsP1). Virology. 438: 37–49.

219. Yin, X., Z. Hu, Q. Gu, X. Wu, Y.-H. Zheng, P. Wei, and X. Wang. 2014. Equine tetherin blocks retrovirus release and its activity is antagonized by equine infectious anemia virus envelope protein. J. Virol. 88: 1259–1270.

220. Bampi, C., L. Rasga, and L. Roux. 2013. Antagonism to human BST-2/tetherin by Sendai virus glycoproteins. J. Gen. Virol. 94: 1211–1219.

221. Langosch, D., B. Brosig, H. Kolmar, and H.J. Fritz. 1996. Dimerisation of the glycophorin A transmembrane segment in membranes probed with the ToxR transcription activator. J. Mol. Biol. 263: 525–530.

222. Russ, W.P., and D.M. Engelman. 1999. TOXCAT: a measure of transmembrane helix association in a biological membrane. Proc. Natl. Acad. Sci. U. S. A. 96: 863–868.

223. Gurezka, R., and D. Langosch. 2001. In Vitro selection of membrane-spanning leucine zipper protein- protein interaction motifs using POSSYCCAT. J. Biol. Chem. 276: 45580–45587.

224. Schneider, D., and D.M. Engelman. 2003. GALLEX, a measurement of heterologous association of transmembrane helices in a biological membrane. J. Biol. Chem. 278: 3105–3111.

225. Karimova, G., N. Dautin, and D. Ladant. 2005. Interaction network among Escherichia coli membrane proteins involved in cell division as revealed by bacterial two-hybrid analysis. J. Bacteriol. 187: 2233–43.

226. Karimova, G., C. Robichon, and D. Ladant. 2009. Characterization of YmgF, a 72-residue inner membrane protein that associates with the Escherichia coli cell division machinery. J. Bacteriol. 91: 333–346. 133 227. Kolmar, H., F. Hennecke, K. Götze, B. Janzer, B. Vogt, F. Mayer, and H.J. Fritz. 1995. Membrane insertion of the bacterial signal transduction protein ToxR and requirements of transcription activation studied by modular replacement of different protein substructures. EMBO J. 14: 3895–904.

228. Zhou, F.X., H.J. Merianos, A.T. Brunger, and D.M. Engelman. 2001. Polar residues drive association of polyleucine transmembrane helices. Proc. Natl. Acad. Sci. U. S. A. 98: 2250–5.

229. Kubatzky, K.F., W. Ruan, R. Gurezka, J. Cohen, R. Ketteler, S.S. Watowich, D. Neumann, D. Langosch, and U. Klingmüller. 2001. Self assembly of the transmembrane domain promotes signal transduction through the erythropoietin receptor. Curr. Biol. 11: 110–115.

230. Mendrola, J.M., M.B. Berger, M.C. King, and M.A. Lemmon. 2002. The single transmembrane domains of ErbB receptors self-associate in cell membranes. J. Biol. Chem. 277: 4704–4712.

231. Laage, R., and D. Langosch. 1997. Dimerization of the synaptic vesicle protein synaptobrevin (vesicle- associated membrane protein) II depends on specific residues within the transmembrane segment. Eur. J. Biochem. 249: 540–546.

232. Huber, O., R. Kemler, and D. Langosch. 1999. Mutations affecting transmembrane segment interactions impair adhesiveness of E-cadherin. J. Cell Sci. 112 ( Pt 2: 4415–23.

233. Lindner, E., and D. Langosch. 2006. A ToxR-based dominant-negative system to investigate heterotypic transmembrane domain interactions. Proteins. 65: 803–7.

234. Nielsen, T.B., and J.A. Reynolds. 1978. Measurements of molecular weights by gel electrophoresis. Methods Enzymol. 48: 3–10.

235. Rath, A., M. Glibowicka, V.G. Nadeau, G. Chen, and C.M. Deber. 2009. Detergent binding explains anomalous SDS-PAGE migration of membrane proteins. Proc. Natl. Acad. Sci. U. S. A. 106: 1760–5.

236. Rath, A., and C.M. Deber. 2013. Correction factors for membrane protein molecular weight readouts on sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Anal. Biochem. 434: 67–72.

237. Dornmair, K., H. Kiefer, and F. Jähnig. 1990. Refolding of an integral membrane protein. OmpA of Escherichia coli. J. Biol. Chem. 265: 18907–11.

238. Ohnishi, S., K. Kameyama, and T. Takagi. 1998. Characterization of a heat modifiable protein, Escherichia coli outer membrane protein OmpA in binary surfactant system of sodium dodecyl sulfate and octylglucoside. Biochim. Biophys. Acta - Biomembr. 1375: 101–109.

239. Rath, A., F. Cunningham, and C.M. Deber. 2013. Acrylamide concentration determines the direction and magnitude of helical membrane protein gel shifts. Proc. Natl. Acad. Sci. U. S. A. 110: 15668–73.

240. Rath, A., R.A. Melnyk, and C.M. Deber. 2006. Evidence for assembly of small multidrug resistance proteins by a “two-faced” transmembrane helix. J. Biol. Chem. 281: 15546–15553.

241. Cunningham, F., B.E. Poulsen, W. Ip, and C.M. Deber. 2011. Beta-branched residues adjacent to GG4 motifs promote the efficient association of glycophorin A transmembrane helices. Biopolymers. 96: 340–7.

242. Lustig, A., A. Engel, G. Tsiotis, E.M. Landau, and W. Baschong. 2000. Molecular weight determination of membrane proteins by sedimentation equilibrium at the sucrose or Nycodenz-adjusted density of the hydrated detergent micelle. Biochim. Biophys. Acta - Biomembr. 1464: 199–206.

134 243. Center, R.J., P. Schuck, R.D. Leapman, L.O. Arthur, P.L. Earl, B. Moss, and J. Lebowitz. 2001. Oligomeric structure of virion-associated and soluble forms of the simian immunodeficiency virus envelope protein in the prefusion activated conformation. Proc. Natl. Acad. Sci. U. S. A. 98: 14877–14882.

244. Burgess, N.K., A.M. Stanley, and K.G. Fleming. 2008. Determination of membrane protein molecular weights and association equilibrium constants using sedimentation equilibrium and sedimentation velocity. Methods Cell Biol. 84: 181–211.

245. Loura, L.M.S., and M. Prieto. 2011. FRET in Membrane Biophysics: An Overview. Front. Physiol. 2: 82.

246. Stangl, M., A. Veerappan, A. Kroeger, P. Vogel, and D. Schneider. 2012. Detergent properties influence the stability of the glycophorin a transmembrane helix dimer in lysophosphatidylcholine micelles. Biophys. J. 103: 2455–2464.

247. Anbazhagan, V., and D. Schneider. 2010. The membrane environment modulates self-association of the human GpA TM domain-Implications for membrane protein folding and transmembrane signaling. Biochim. Biophys. Acta - Biomembr. 1798: 1899–1907.

248. You, M., E. Li, W.C. Wimley, and K. Hristova. 2005. Förster resonance energy transfer in liposomes: Measurements of transmembrane helix dimerization in the native bilayer environment. Anal. Biochem. 340: 154–164.

249. Merzlyakov, M., M. You, E. Li, and K. Hristova. 2006. Transmembrane helix heterodimerization in lipid bilayers: Probing the energetics behind autosomal dominant growth disorders. J. Mol. Biol. 358: 1–7.

250. Del Piccolo, N., S. Sarabipour, and K. Hristova. 2017. A new method to study heterodimerization of membrane proteins and its application to fibroblast growth factor receptors. J. Biol. Chem. 292: 1288– 1301.

251. Melnyk, R. a, A.W. Partridge, and C.M. Deber. 2002. Transmembrane domain mediated self-assembly of major coat protein subunits from Ff bacteriophage. J. Mol. Biol. 315: 63–72.

252. Ng, D.P., and C.M. Deber. 2016. Modulating transmembrane α-Helix interactions through pH-sensitive boundary residues. Biochemistry. 55: 4306–4315.

253. Fisher, L.E., D.M. Engelman, and J.N. Sturgis. 1999. Detergents modulate dimerization, but not helicity, of the glycophorin A transmembrane domain. J. Mol. Biol. 293: 639–651.

254. Fisher, L.E., D.M. Engelman, and J.N. Sturgis. 2003. Effect of detergents on the association of the glycophorin a transmembrane helix. Biophys. J. 85: 3097–3105.

255. Duneau, J.-P., A.P. Vegh, and J.N. Sturgis. 2007. A dimerization hierarcy in the transmembrane domains of the {HER} receptor family. Biochemistry. 46: 2010–2019.

256. Merzlyakov, M., L. Chen, and K. Hristova. 2007. Studies of receptor tyrosine kinase transmembrane domain interactions: The EmEx-FRET method. J. Membr. Biol. 215: 93–103.

257. Khadria, A.S., A. Senes, and M. Wi. 2013. The transmembrane domains of the bacterial cell division proteins FtsB and FtsL form a stable high-order oligomer. Biochemistry. 52: 7542–7550.

258. Sparr, E., W.L. Ash, P. V Nazarov, D.T.S. Rijkers, M. a Hemminga, D.P. Tieleman, and J.A. Killian. 2005. Self- association of transmembrane alpha-helices in model membranes: importance of helix orientation and role of hydrophobic mismatch. J. Biol. Chem. 280: 39324–31. 135

259. Li, E., M. You, and K. Hristova. 2005. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Förster resonance energy transfer suggest weak interactions between fibroblast growth factor receptor 3 (FGFR3) transmembrane domains in the absence of extracellular domains and ligands. Biochemistry. 44: 352–60.

260. Goffinet, C., I. Allespach, S. Homann, H.-M. Tervo, A. Habermann, D. Rupp, L. Oberbremer, C. Kern, N. Tibroni, S. Welsch, J. Krijnse-Locker, G. Banting, H.-G. Kräusslich, O.T. Fackler, and O.T. Keppler. 2009. HIV-1 antagonism of CD317/tetherin is species-specific and involves Vpu-mediated proteasomal degradation of the intrinsic immunity factor. Retrovirology. 6: O10.

261. Andrew, A., and K. Strebel. 2010. HIV-1 Vpu targets cell surface markers CD4 and BST-2 through distinct mechanisms. Mol. Aspects Med. 31: 407–17.

262. Miyagi, E., A.J. Andrew, S. Kao, and K. Strebel. 2009. Vpu enhances HIV-1 virus release in the absence of Bst-2 cell surface down-modulation and intracellular depletion. Proc. Natl. Acad. Sci. U. S. A. 106: 2868–73.

263. Gupta, R.K., S. Hué, T. Schaller, E. Verschoor, D. Pillay, and G.J. Towers. 2009. Mutation of a single residue renders human tetherin resistant to HIV-1 Vpu-mediated depletion. PLoS Pathog. 5: e1000443.

264. Jia, B., R. Serra-Moreno, W. Neidermyer, A. Rahmberg, J. Mackey, I. Ben Fofana, W.E. Johnson, S. Westmoreland, and D.T. Evans. 2009. Species-specific activity of SIV Nef and HIV-1 Vpu in overcoming restriction by tetherin/BST2. PLoS Pathog. 5: e1000429.

265. Dubé, M., B.B. Roy, P. Guiot-Guillain, J. Binette, J. Mercier, A. Chiasson, and E. a. Cohen. 2010. Antagonism of tetherin restriction of HIV-1 release by Vpu involves binding and sequestration of the restriction factor in a perinuclear compartment. PLoS Pathog. 6: e1000856.

266. Skasko, M., Y. Wang, Y. Tian, A. Tokarev, J. Munguia, A. Ruiz, E.B. Stephens, S.J. Opella, and J. Guatelli. 2012. HIV-1 Vpu protein antagonizes innate restriction factor BST-2 via lipid-embedded helix-helix interactions. J. Biol. Chem. 287: 58–67.

267. Deber, C.M., C. Wang, L.P. Liu, A.S. Prior, S. Agrawal, B.L. Muskat, and A.J. Cuticchia. 2001. TM Finder: a prediction program for transmembrane protein segments using a combination of hydrophobicity and nonpolar phase helicity scales. Protein Sci. 10: 212–9.

268. Melnyk, R. a., A.W. Partridge, J. Yip, Y. Wu, N.K. Goto, and C.M. Deber. 2003. Polar residue tagging of transmembrane peptides. Biopolym. - Pept. Sci. Sect. 71: 675–85.

269. Lemmon, M. a, J.M. Flanagan, J.F. Hunt, B.D. Adair, B.J. Bormann, C.E. Dempsey, and D.M. Engelman. 1992. Glycophorin A dimerization is driven by specific interactions between transmembrane alpha-helices. J. Biol. Chem. 267: 7683–9.

270. Furthmayr, H., and V.T. Marchesi. 1976. Subunit structure of human erythrocyte glycophorin A. Biochemistry. 15: 1137–44.

271. Sabnis, R.W. 2015. Handbook of Fluorescent Dyes and Probes. Hoboken, NJ, USA: John Wiley & Sons, Inc.

272. Shiraki, K., K. Nishikawa, and Y. Goto. 1995. Trifluoroethanol-induced stabilization of the alpha-helical structure of beta-lactoglobulin: implication for non-hierarchical protein folding. J. Mol. Biol. 245: 180–94.

273. Rath, A., and C.M. Deber. 2012. Protein structure in membrane domains. Annu. Rev. Biophys. 41: 135–55.

274. Lehrer, S.S. 1997. Intramolecular pyrene excimer fluorescence: A probe of proximity and protein 136

conformational change. Methods Enzymol. 278: 286–295.

275. Liu, W., Y. Chen, H. Watrob, S.G. Bartlett, L. Jen-Jacobson, and M.D. Barkley. 1998. N-termini of EcoRi restriction endonuclease dimer are in close proximity on the protein surface. Biochemistry. 37: 15457– 15465.

276. Hammarström, P., M. Persson, P.O. Freskgård, L.G. Mårtensson, D. Andersson, B.H. Jonsson, and U. Carlsson. 1999. Structural mapping of an aggregation nucleation site in a molten globule intermediate. J. Biol. Chem. 274: 32897–32903.

277. Swiecki, M., S.M. Scheaffer, M. Allaire, D.H. Fremont, M. Colonna, and T.J. Brett. 2011. Structural and biophysical analysis of BST-2/tetherin ectodomains reveals an evolutionary conserved design to inhibit virus release. J. Biol. Chem. 286: 2987–97.

278. Sakuma, T., A. Sakurai, and J. Yasuda. 2009. Dimerization of tetherin is not essential for its antiviral activity against Lassa and Marburg viruses. PLoS One. 4: e6934.

279. Marassi, F.M., C. Ma, H. Gratkowski, S.K. Straus, K. Strebel, M. Oblatt-Montal, M. Montal, and S.J. Opella. 1999. Correlation of the structural and functional domains in the membrane protein Vpu from HIV-1. Proc. Natl. Acad. Sci. U. S. A. 96: 14336–14341.

280. Paul, M., S. Mazumder, N. Raja, and M. a Jabbar. 1998. Mutational analysis of the human immunodeficiency virus type 1 Vpu transmembrane domain that promotes the enhanced release of virus- like particles from the plasma membrane of mammalian cells. J. Virol. 72: 1270–9.

281. Hout, D.R., M.L. Gomez, E. Pacyniak, L.M. Gomez, S.H. Inbody, E.R. Mulcahy, N. Culley, D.M. Pinson, M.F. Powers, S.W. Wong, and E.B. Stephens. 2005. Scrambling of the amino acids within the transmembrane domain of Vpu results in a simian-human immunodeficiency virus (SHIVTM) that is less pathogenic for pig- tailed macaques. Virology. 339: 56–69.

282. Hussain, A., S.R. Das, C. Tanwar, and S. Jameel. 2007. Oligomerization of the human immunodeficiency virus type 1 (HIV-1) Vpu protein – a genetic, biochemical and biophysical analysis. Virol. J. 4: 81.

283. Bour, S., and K. Strebel. 2003. The HIV-1 Vpu protein: A multifunctional enhancer of viral particle release. Microbes Infect. 5: 1029–1039.

284. Fischer, W.B. 2003. Vpu from HIV-1 on an atomic scale: Experiments and computer simulations. FEBS Lett. 552: 39–46.

285. Becucci, L., A. Cembran, C.B. Karim, D.D. Thomas, R. Guidelli, J. Gao, and G. Veglia. 2009. On the function of pentameric phospholamban: Ion channel or storage form? Biophys. J. 96: L60–L62.

286. Cornea, R.L., J.M. Autry, Z. Chen, and L.R. Jones. 2000. Reexamination of the role of the leucine/isoleucine zipper residues of phospholamban in inhibition of the Ca2+ pump of cardiac sarcoplasmic reticulum. J. Biol. Chem. 275: 41487–94.

287. Kimura, Y., K. Kurzydlowski, M. Tada, and D.H. MacLennan. 1997. Phospholamban inhibitory function is activated by depolymerization. J. Biol. Chem. 272: 15061–4.

288. Bolduan, S., T. Reif, M. Schindler, and U. Schubert. 2014. HIV-1 Vpu mediated downregulation of CD155 requires alanine residues 10, 14 and 18 of the transmembrane domain. Virology. 464–465: 375–384.

289. Sauter, D., M. Schindler, A. Specht, W.N. Landford, J. Münch, K.-A. Kim, J. Votteler, U. Schubert, F. Bibollet- 137

Ruche, B.F. Keele, J. Takehisa, Y. Ogando, C. Ochsenbauer, J.C. Kappes, A. Ayouba, M. Peeters, G.H. Learn, G. Shaw, P.M. Sharp, P. Bieniasz, B.H. Hahn, T. Hatziioannou, and F. Kirchhoff. 2009. Tetherin-driven adaptation of Vpu and Nef function and the evolution of pandemic and nonpandemic HIV-1 strains. Cell Host Microbe. 6: 409–421.

290. McNatt, M.W., T. Zang, and P.D. Bieniasz. 2013. Vpu binds directly to tetherin and displaces it from nascent virions. PLoS Pathog. 9: e1003299.

291. Cole, G., K. Simonetti, I. Ademi, and S. Sharpe. 2012. Dimerization of the transmembrane domain of human tetherin in membrane mimetic environments. Biochemistry. 51: 5033–40.

292. Chen, L., M. Merzlyakov, T. Cohen, Y. Shai, and K. Hristova. 2009. Energetics of ErbB1 transmembrane domain dimerization in lipid bilayers. Biophys. J. 96: 4622–4630.

293. Li, E., M. You, and K. Hristova. 2006. FGFR3 dimer stabilization due to a single amino acid pathogenic mutation. J. Mol. Biol. 356: 600–612.

294. Wolber, and Hudson. 1979. An analytic solution to the Förster energy transfer problem in two dimensions. Biophys. J. 28: 197–210.

295. Kučerka, N., M.P. Nieh, and J. Katsaras. 2011. Fluid phase lipid areas and bilayer thicknesses of commonly used phosphatidylcholines as a function of temperature. Biochim. Biophys. Acta - Biomembr. 1808: 2761– 2771.

296. Veatch, W., and L. Stryer. 1977. The dimeric nature of the gramicidin A transmembrane channel: Conductance and fluorescence energy transfer studies of hybrid channels. J. Mol. Biol. 113: 89–102.

297. Bottino, C., M. Falco, S. Parolini, E. Marcenaro, R. Augugliaro, S. Sivori, E. Landi, R. Biassoni, L.D. Notarangelo, L. Moretta, and a Moretta. 2001. NTB-A, a novel SH2D1A-associated surface molecule contributing to the inability of natural killer cells to kill Epstein-Barr virus-infected B cells in X-linked lymphoproliferative disease. J. Exp. Med. 194: 235–246.

298. Greiner, T., S. Bolduan, B. Hertel, C. Groß, K. Hamacher, U. Schubert, A. Moroni, and G. Thiel. 2016. Ion channel activity of Vpu proteins is conserved throughout evolution of HIV-1 and SIV. 2: 1–11.

299. Merzlyakov, M., and K. Hristova. 2008. Förster resonance energy transfer measurements of transmembrane helix dimerization energetics. Methods Enzymol. 450: 107–27.

300. You, M., J. Spangler, E. Li, X. Han, P. Ghosh, and K. Hristova. 2007. Effect of pathogenic cysteine mutations on FGFR3 transmembrane domain dimerization in detergents and lipid bilayers. Biochemistry. 46: 11039– 11046.

301. Walters, R.F.S., and W.F. DeGrado. 2006. Helix-packing motifs in membrane proteins. Proc. Natl. Acad. Sci. U. S. A. 103: 13658–63.

302. van Meer, G., D.R. Voelker, and G.W. Feigenson. 2008. Membrane lipids: where they are and how they behave. Nat. Rev. Mol. Cell Biol. 9: 112–124.

303. Mitra, K., I. Ubarretxena-Belandia, T. Taguchi, G. Warren, and D.M. Engelman. 2004. Modulation of the bilayer thickness of exocytic pathway membranes by membrane proteins rather than cholesterol. Proc. Natl. Acad. Sci. U. S. A. 101: 4083–8.

304. Montecucco, C., G. a Smith, F. Dabbeni-sala, a Johannsson, Y.M. Galante, and R. Bisson. 1982. Bilayer 138

thickness and enzymatic activity in the mitochondrial cytochrome c oxidase and ATPase complex. FEBS Lett. 144: 145–148.

305. Froud, R.J., C.R. Earl, J.M. East, and A.G. Lee. 1986. Effects of lipid fatty acyl chain structure on the activity of the (Ca2+ + Mg2+)-ATPase. Biochim. Biophys. Acta. 860: 354–60.

306. Johannsson, A., G.A. Smith, and J.C. Metcalfe. 1981. The effect of bilayer thickness on the activity of (Na+ + K+)-ATPase. BBA - Biomembr. 641: 416–421.

307. Dumas, F., J.F. Tocanne, G. Leblanc, and M.C. Lebrun. 2000. Consequences of hydrophobic mismatch between lipids and melibiose permease on melibiose transport. Biochemistry. 39: 4846–4854.

308. Pilot, J.D., J.M. East, and A.G. Lee. 2001. Effects of bilayer thickness on the activity of diacylglycerol kinase of Escherichia coli. Biochemistry. 40: 8188–8195.

309. Meder, D., M.J. Moreno, P. Verkade, W.L.C. Vaz, and K. Simons. 2006. Phase coexistence and connectivity in the apical membrane of polarized epithelial cells. Proc. Natl. Acad. Sci. U. S. A. 103: 329–34.

310. Epand, R.M., and R.F. Epand. 2009. Lipid domains in bacterial membranes and the action of antimicrobial agents. Biochim. Biophys. Acta - Biomembr. 1788: 289–294.

311. van Meer, G. 2002. Cell biology. The different hues of lipid rafts. Science. 296: 855–7.

312. Mall, S., R. Broadbridge, R.P. Sharma, J.M. East, and a. G. Lee. 2001. Self-association of model transmembrane alpha-helices is modulated by lipid structure. Biochemistry. 40: 12379–86.

313. Cserzö, M., E. Wallin, I. Simon, G. von Heijne, and A. Elofsson. 1997. Prediction of transmembrane alpha- helices in prokaryotic membrane proteins: the dense alignment surface method. Protein Eng. 10: 673–6.

314. Tusnády, G.E., and I. Simon. 2001. The HMMTOP transmembrane topology prediction server. Bioinformatics. 17: 849–50.

315. Hessa, T., N.M. Meindl-Beinker, A. Bernsel, H. Kim, Y. Sato, M. Lerch-Bader, I. Nilsson, S.H. White, and G. Von Heijne. 2007. Molecular code for transmembrane-helix recognition by the Sec61 translocon. Nature. 450: 1026–1030.

316. Viklund, H., and A. Elofsson. 2008. OCTOPUS: Improving topology prediction by two-track ANN-based preference scores and an extended topological grammar. Bioinformatics. 24: 1662–1668.

317. Krogh, A., B. Larsson, G. Von Heijne, and E.L.L. Sonnhammer. 2001. Predicting transmembrane protein topology with a hidden Markov model: Application to complete genomes. J. Mol. Biol. 305: 567–580.

318. Käll, L., A. Krogh, and E.L.L. Sonnhammer. 2007. Advantages of combined transmembrane topology and signal peptide prediction-the Phobius web server. Nucleic Acids Res. 35: 429–432.

319. Bernsel, A., H. Viklund, J. Falk, E. Lindahl, G. von Heijne, and A. Elofsson. 2008. Prediction of membrane- protein topology from first principles. Proc. Natl. Acad. Sci. U. S. A. 105: 7177–81.

320. Hirokawa, T., S. Boon-Chieng, and S. Mitaku. 1998. SOSUI: classification and secondary structure prediction system for membrane proteins. Bioinformatics. 14: 378–9.

321. Juretić, D., L. Zoranić, and D. Zucić. Basic charge clusters and predictions of membrane protein topology. J. Chem. Inf. Comput. Sci. 42: 620–32. 139

322. K. Hofmann, W.S. 1993. TMBase - A database of membrane spaning protein segments. Biol. Chem. Hoppe- Seyler. 374.

323. Popot, J.-L., and D.M. Engelman. 2016. Membranes do not tell proteins how to fold. Biochemistry. 55: 5– 18.

324. Dowhan, W. 1997. Molecular basis for membrane phospholipid diversity: why are there so many lipids? Annu. Rev. Biochem. 66: 199–232.

325. Tristram-Nagle, S., H.I. Petrache, and J.F. Nagle. 1998. Structure and interactions of fully hydrated dioleoylphosphatidylcholine bilayers. Biophys. J. 75: 917–25.

326. Barrera, F.N., J. Fendos, and D.M. Engelman. 2012. Membrane physical properties influence transmembrane helix formation. Proc. Natl. Acad. Sci. U. S. A. 109: 14422–7.

327. Rinia, H.A., R.A. Kik, R.A. Demel, M.M.E. Snel, J.A. Killian, J.P.J.M. Van Der Eerden, and B. De Kruijff. 2000. Visualization of highly ordered striated domains induced by transmembrane peptides in supported phosphatidylcholine bilayers. Biochemistry. 39: 5852–5858.

328. Muhle-Goll, C., S. Hoffmann, S. Afonin, S.L. Grage, A.A. Polyansky, D. Windisch, M. Zeitler, J. Bürck, and A.S. Ulrich. 2012. Hydrophobic matching controls the tilt and stability of the dimeric platelet-derived growth factor receptor (PDGFR) α transmembrane segment. J. Biol. Chem. 287: 26178–26186.

329. Cole, G.B., S.E. Reichheld, and S. Sharpe. 2017. FRET analysis of the promiscuous yet specific interactions of the HIV-1 Vpu transmembrane domain. Biophys. J. 113: 1992–2003.

330. Sang, H.P., and S.J. Opella. 2005. Tilt angle of a trans-membrane helix is determined by hydrophobic mismatch. J. Mol. Biol. 350: 310–318.

331. Schmidt, S., J. V. Fritz, J. Bitzegeio, O.T. Fackler, and O.T. Keppler. 2011. HIV-1 Vpu blocks recycling and biosynthetic transport of the intrinsic immunity factor CD317/tetherin to overcome the virion release restriction. MBio. 2: e00036-11.

332. Du Pont, K.E., A.M. McKenzie, O. Kokhan, I. Sumner, and C.E. Berndsen. 2016. The Disulfide Bonds within BST-2 Enhance Tensile Strength during Viral Tethering. Biochemistry. 55: 940–947.

333. Polyansky, A.A., A.O. Chugunov, P.E. Volynsky, N.A. Krylov, D.E. Nolde, and R.G. Efremov. 2014. PREDDIMER: A web server for prediction of transmembrane helical dimers. Bioinformatics. 30: 889–890.

334. Polyansky, A.A., P.E. Volynsky, and R.G. Efremov. 2011. Structural, dynamic, and functional aspects of helix association in membranes: a computational view. Adv. Protein Chem. Struct. Biol. 83: 129–61.

335. Herrero, L., N. Monroy, and M.E. González. 2013. HIV-1 Vpu protein mediates the transport of potassium in Saccharomyces cerevisiae. Biochemistry. 52: 171–7.

336. Taube, R., R. Alhadeff, D. Assa, M. Krugliak, and I.T. Arkin. 2014. Bacteria-based analysis of HIV-1 Vpu channel activity. PLoS One. 9: e105387.

337. Fricke, F., J. Beaudouin, R. Eils, and M. Heilemann. 2015. One, two or three? Probing the stoichiometry of membrane proteins by single-molecule localization microscopy. Sci. Rep. 5: 14072.

338. Das, S.K., M. Darshi, S. Cheley, M.I. Wallace, and H. Bayley. 2007. Membrane protein stoichiometry determined from the step-wise photobleaching of dye-labelled subunits. ChemBioChem. 8: 994–999. 140 Copyright Acknowledgements

Chapter 2 has been, in part, adapted with permission from Cole, G., Simonetti K., Ademi, I., and Sharpe, S. Dimerization of the transmembrane domain of human tetherin in membrane mimetic environments. Biochemistry. 2012. 51: 5033–40. Copyright 2012 American Chemical Society.

Chapter 3 has been in part adapted with permission from Cole, G., Reichheld, S., and Sharpe S., FRET analysis of the promiscuous yet specific interactions of the HIV-1 Vpu transmembrane domain. Biophysical Journal. (2017) 113: 1992-2000. Copyright 2017 Elsevier

141