SUBSTRATE SPECIFICITY AND
REGULATION OF NEDD4 PROTEINS
by
Mary Christine Bruce
A thesis submitted in conformity with the requirements for the Degree of Doctor of Philosophy Graduate Department of Biochemistry University of Toronto
© Copyright by Mary Christine Bruce 2009
Substrate Specificity and Regulation of Nedd4 proteins
Doctor of Philosophy, 2009
Mary Christine Bruce, Department of Biochemistry, University of Toronto
Abstract
Nedd4-1 and Nedd4-2 are closely related E3 ubiquitin protein ligases that contain a C2 domain,
3-4 WW domains, and a catalytic ubiquitin ligase HECT domain. The WW domains of Nedd4
proteins recognize substrates for ubiquitination by binding the sequence L/PPxY (the PY-motif)
found in target proteins. Nedd4-2 functions as a suppressor of the epithelial Na+ channel
(ENaC), which interacts with Nedd4-2 WW domains via PY-motifs located at its C-terminus.
The importance of Nedd4-2 mediated ENaC regulation is highlighted by the fact that mutations affecting the ENaC PY-motifs cause Liddle syndrome, a hereditary hypertension.
Since all Nedd4 family members recognize the same core sequence in their target proteins, the question was raised of how substrate specificity for Nedd4 family members is achieved. Using intrinsic tryptophan florescence to measure the binding affinity of Nedd4-1/-2
WW domains for their substrate PY-motifs, we demonstrate the importance of both PY-motif and WW domain residues, outside the core binding residues, in determining the specificity of
WW domain-ligand interactions.
Little was known about regulation of catalytic activity for this family of E3 ligases, and hence was the second focus of my work. Notably, Nedd4-2 contains a PY-motif within its
HECT domain, raising the possibility that its catalytic activity is regulated by an interaction between its WW domains and HECT domain. Here I present evidence supporting a model in which a low-affinity interaction between the Nedd4-2 WW domains and its HECT domain
ii
regulate Nedd4-2 stability by preventing self-ubiquitination and subsequent degradation.
Furthermore, evidence is presented suggesting that interaction between Nedd4-2 and the RING-
E3 ligase Rnf11, a Nedd4-2 substrate, may also serve to regulate Nedd4-2 stability, as this
interaction leads to decreased Nedd4-2 self-ubiquitination.
Collectively, the studies presented here further our understanding of the substrate specificity and regulation of Nedd4-1 and Nedd4-2.
iii
Acknowledgments
First of all, I would like to thank my supervisor Daniela Rotin for her support and guidance over
the past 7 years. Daniela’s passion for science first inspired me while I was a summer student in her lab and is ultimately what led to my decision to pursue graduate studies. Thank you Daniela for allowing me to pursue my interests, for providing me the opportunity to work on multiple collaborations, and most of all for always believing in me.
Thank you to all the members of the Rotin Lab, past and present, without whom this ride wouldn’t have been as enjoyable. Special thanks to Chris Fladd and Pauline Henry for teaching me the basics when I was ‘green’. Thanks to Voula Kanelis for helping me appreciate (and understand) structural biology, but more importantly for her inspirational passion for science and life. Thanks to Chong Jiang for doing an awesome job at running the lab and for fixing everything (or at least trying!) I would especially like to thank Wioletta Glowacka, my bench mate and my best friend, for her endless support, both scientific and personal, over the last three
years. I will miss working beside you!
I would also like to thank my graduate committee (Dr. Christine Bear and Dr. Sean Egan)
for their guidance over the years and for always asking the tough questions. Special thanks to
Carrie Harber in the Department of Biochemistry for being amazing at her job and for making
life as a graduate student much easier.
Finally I would like to thank my family. Thanks to my sisters Jenny, Lauren and
Catherine for laughter, love and friendship. Thanks to my parents Linda and Bill for always
believing in me. Your love, patience and encouragement made this PhD possible. Thanks to my
second family the Bruces, especially Bob and Susan, for your continuous love and support.
Above all, thanks to my husband Neil, for always being there, for your encouragement and
confidence in me and for your endless love.
iv
Table of Contents
Abstract ii Acknowledgments iv Table of Contents v List of Figures ix List of Tables xi Abbreviations xii
Chapter 1: Introduction 1 I) Thesis Overview 2 II) The Ubiquitination Pathway 4 A) Components of the Pathway 5 i) Ubiquitin and Ubiquitin-like proteins 5 ii) Ubiquitin-activating enzymes (E1s) 11 iii) Ubiquitin-conjugating enzymes (E2s) 12 iv) Ubiquitin protein ligases (E3s) 13 1) RING E3s 14 2) HECT E3s 18 v) E4 enzymes 22 vi) Deubiquitinating enzymes (DUBs) 23 vii) Ubiquitin-binding domains (UBDs) 24 B) Fates of Ubiquitinated Proteins 28 i) Protein Degradation: The Ubiquitin-Proteasome Pathway 28 ii) Role of Ubiquitination in trafficking, endocytosis and protein sorting 30 1) Endocytosis of plasma membrane proteins 30 2) Protein sorting at the MVB/Late Endosome 32 3) Protein sorting at the Trans-Golgi Network (TGN) 34 4) Viral Budding 35 III) Nedd4 family of E3 ligases 37 A) Function of the different domains of Nedd4 proteins 37
v
i) C2 domain 37 ii) WW domains 40 iii) HECT domain 41 B) Members of the Nedd4 family 41 C) Cellular targets of Nedd4-1 and Nedd4-2 42 i) Nedd4-2 43 ii) Nedd4-1 44 IV) Project Goals and Rationale 46
Chapter 2: Affinity and specificity of Nedd4-1 and -2 WW domains 47 for substrate PY motifs I) Summary 48 II) Introduction 49 III) Experimental Procedures 55 IV) Results 58 A) Contribution of WW3* domain to the Nedd4-ENaC interaction 58 i) Comparison of binding affinity of WW3* and WW4 toward the PY 58 motif of βENaC ii) Suppression of ENaC activity by the WW3*-containing dNedd4-1 62 iii) Homology modeling and comparison of WW3* and WW4 complexes 65 iv) Mutation analysis to test the role of WW3* Ala-504/Pro-505 in 66 conferring high affinity binding to PY motifs B) Identification of residues involved in high affinity interaction between 69 Drosophila Nedd4 WW3* and the Commissureless PY-motif i) Contributions of LPSY peptide residues to binding affinity 74 ii) Contributions of dNedd4 WW3* domain residues to binding affinity 75 V) Discussion 78 A) Importance of the WW3* domain in Nedd4-ENaC interactions 78 B) Role of the WW3* residues in high affinity binding with substrate PY motifs 79 C) Biological significance of Nedd4 WW domain specificity 80
vi
Chapter 3: Regulation of Nedd4-2 self-ubiquitination and stability by a 83 PY-motif located within its HECT-domain I) Summary 84 II) Introduction 85 III) Experimental Procedures 86 IV) Results 88 A) The Nedd4-2 WW domains bind to the HECT-PY motif 88 and regulate catalytic activity of the HECT domain B) Nedd4-2 stability is regulated by self-ubiquitination 95 and subsequent degradation C) The HECT-PY motif regulates self-, but not substrate, 100 ubiquitination and stability V) Discussion 108
Chapter 4: Characterization of the Nedd4-2 interacting protein Rnf11 115 I) Summary 116 II) Introduction 117 III) Experimental Procedures 124 IV) Results 127 A) Expression and localization of Rnf11 127 B) Rnf11 binds both Nedd4-1 and Nedd4-2 132 C) Rnf11 exhibits in vitro ubiquitin ligase activity 135 D) Role of the Nedd4-2/Rnf11 interaction 144 i) Ubiquitination of Rnf11 by Nedd4-2 144 ii) Effect of Rnf11 on Nedd4-2 ubiquitination 148 V) Discussion 151
Chapter 5: Thesis Summary and Future Directions 158 I) Molecular determinants of WW domain specificity of Nedd4 family members 159 II) Regulation of Nedd4 E3 catalytic activity 161
vii
III) Nedd4-2 binding protein Rnf11 163
References 167
Appendix I: Interactions between the three CIN85 SH3 Domains and Ubiquitin: 182 Implications for CIN85 Ubiquitination
Appendix II: Transport of LAPTM5 to lysosomes requires association with 184 the ubiquitin ligase Nedd4, but not LAPTM5 ubiquitination
Appendix III: Molecular determinants of voltage-gated sodium channel 186 regulation by the Nedd4/Nedd4-like proteins
viii
List of Figures
Chapter 1
Figure 1-1 The ubiquitination cascade 7
Figure 1-2 Schematic view of opposite faces of ubiquitin 10
Figure 1-3 RING-finger E3s 17
Figure 1-4 The HECT domain 21
Figure 1-5 The Nedd4 family 39
Chapter 2
Figure 2-1 Alignment of Nedd4-1 and Nedd4-2 proteins 54
Figure 2-2 Representative curves of fluorescence emission from rNedd4-1 WW4, 60 x/m/hNedd4-2 WW3*, or hNedd4-1 WW3* binding to αPY, βPY, or γPY peptides of ENaC
Figure 2-3 Suppression of ENaC by dNedd4-1 64
Figure 2-4 Homology model of Nedd4-2 WW3*·βENaC PY motif complex 68
LPSY Peptide Complex 71-כFigure 2-5 Solution Structure of the dNedd4 WW3
Domain-LPSY Peptide and rNedd4 73 כFigure 2-6 Comparison of the dNedd4 WW3 WW4 Domain-βENaC Complexes
Chapter 3
Figure 3-1 A PY motif is present within the HECT domain 90
Figure 3-2 The Nedd4-2 HECT PY motif binds its WW domains and regulates 94 catalytic activity
Figure 3-3 Both Nedd4-2(WT) and the HECT-PY mutant, Nedd4-2(YA), display 97 ubiquitination activity in an in vitro assay
Figure 3-4 Mutation of the HECT PY motif affects Nedd4-2 stability 99
ix
Figure 3-5 Nedd4-2(YA) ubiquitinates and regulates substrate as effectively as 102 Nedd4-2 (WT)
Figure 3-6 Substrate ubiquitination promotes Nedd4-2 self-ubiquitination 105
Figure 3-7 Homology model of a Nedd4-2 WW3 domain–HECT PY motif complex 107
Figure 3-8 Nedd4-2(FL) (phenylalanine to leucine residue mutant) displays an 110 increased affinity for the HECT PY motif but exhibits the same stability as Nedd4-2(WT)
Figure 3-9 Model depicting regulation of Nedd4-2 catalytic activity by the HECT 114 PY motif
Chapter 4
Figure 4-1 Schematic and protein sequence of the human Rnf11 protein 119
Figure 4-2 Comparison of Rnf11 orthologues 122
Figure 4-3 Rnf11 transcripts are found in HEK 293T, HeLa, and 3T3 cell lines 129
Figure 4-4 Rnf11 colocalizes with the plasma membrane 131
Figure 4-5 Rnf11 binds endogenous Nedd4 134
Figure 4-6 Rnf11 interacts preferentially with Nedd4-2 over Nedd4-1 137
Figure 4-7 Rnf11 has E2-dependent ubiquitin ligase activity in vitro 141
Figure 4-8 Rnf11 and Nedd4-2 make ubiquitin chains with different lysine linkages 143 in vitro
Figure 4-9 Nedd4-2 overexpression increases Rnf11 ubiquitination 146
Figure 4-10 Rnf11 ubiquitination is dependent on its UIM and PY motif 150
Figure 4-11 Rnf11 decreases Nedd4-2 ubiquitination 153
x
List of Tables
Chapter 1
Table 1-I Ubiquitin-binding domains 26
Chapter 2
Table 2-I Dissociation constants (Kd) of rNedd4-1 WW4, x/m/h Nedd4-2 WW3*, 61 or hNedd4-1 WW3* domains binding to ENaC α, β, and γ-PY motif peptides
Table 2-II Dissociation Constants for dNedd4 WW Domain-Comm Peptide 76 Interactions
xi
Abbreviations
ABC – ATP-Binding Cassette AMSH – Associated Molecule with SH3 domain of STAM APC – Anaphase-Promoting Complex ARF – ADP-ribosylation factor C2 domain – Conserved region 2 of Ca2+-dependent isoforms of PKC Comm – Commissureless DUB – Deubiquitinating enzyme E1 – Ubiquitin activating enzyme, the first enzyme in the ubiquitination cascade E2 – Ubiquitin conjugating enzyme, the second enzyme in the ubiquitination cascade E3 – Ubiquitin protein ligase, the third enzyme in the ubiquitination cascade E6-AP – E6 Associated Protein EGF – Epidermal Growth Factor EGFR – Epidermal Growth Factor Receptor ENaC – Epithelial Na+ Channel Eps15 – Epidermal growth factor receptor pathway substrate 15 ER – Endoplasmic Reticulum ESCRT – Endosomal Sorting Complex Required for Transport Gap – General amino acid permease GGA – Golgi-localizing, γ-adaptin ear domain homology, ARF-binding proteins GHR – Growth Hormone Receptor GPCR – G-protein Coupled Receptor GST – Glutathione S-Transferase HECT – Homologous to E6-AP C-terminus HEK 293T – Human Embryonic Kidney cells expressing the large T-antigen of simian virus 40 HGF – Hepatocyte Growth Factor Hrs – Hepatocyte growth factor–regulated tyrosine kinase substrate HTLV – Human T-cell Leukemia Virus IFN – Interferon IGF-IR – Insulin-like Growth Factor-I Receptor IP – Immunoprecipitation
xii
ISG – Interferon Stimulated Gene JAMM – Jab1/MPN domain metalloenzyme KO – knockout LAPTM5 – Lysosomal Associated Protein Transmembrane 5 LLnL – N-Acetyl-L-leucyl-L-leucyl-L-norleucinal MFS – Multi Facilitators Superfamily MHC – Major Histocompatibility Complex MJD – Machado-Joseph disease protease MVB – Multivesicular Body Nedd – Neuronal Precursor Cell-Expressed Developmentally Downregulated Ni-NTA – Nickel-nitrilotriacetic acid OTU – ovarian tumour protease PAGE – Polyacrylamide Gel Electrophoresis PBS – Phosphate Buffered Saline PDGFR – Platelet-Derived Growth Factor Receptor PKC – Protein Kinase C PMSF – phenyl methyl sulfonyl fluoride PPII – polyproline type II RING – Really Interesting New Gene Robo – Roundabout RTK – Receptor Tyrosine Kinase SCF – Skp1/cullin-1/F-box protein complex SDS – Sodium Dodecyl Sulfate Sgk – Serum and glucocorticoid-regulated kinase SH3 – Src Homology 3 Smurf – Smad ubiquitin regulatory factor STAM – Signal-Transducing Adaptor Molecule Su(dx) – Suppressor of Deltex SUMO – Small Ubiquitin-like Modifier TGFβ – Transforming Growth Factor β TGN – Trans-Golgi Network
xiii
TKB – Tyrosine Kinase Binding domain Ub – ubiquitin UBA – Ubiquitin Associated domain Ubc – Ubiquitin conjugating enzyme (E2) UBD – Ubiquitin-Binding Domain Ubl – ubiquitin-like UCH – Ubiquitin C-terminal Hydrolase UEV – Ubiquitin E2 Variant UIM – Ubiquitin Interacting Motif USP – Ubiquitin Specific Protease VCB – von Hippel-Lindau-Cul2/elongin B/elongin C Vps – Vacuolar protein sorting WT – wild-type WW domain – protein–protein interaction domain containing two conserved tryptophan residues ZnF – Zinc-finger
xiv
CHAPTER 1
Introduction
1 Chapter 1
I) Thesis Overview
This doctoral thesis presents a study on the Nedd4 ubiquitin protein ligases, Nedd4-1 and Nedd4-
2. Here I examine, in detail, the molecular determinants of Nedd4-1/-2 substrate specificity and explore potential mechanisms for regulation of Nedd4-2 catalytic activity.
An introductory literature review is reported in Chapter 1. The first part of this chapter provides an introduction to the ubiquitination pathway, with a focus on components of the pathway, and possible cellular fates of proteins that are recognized and targeted for ubiquitination. The second part of this chapter focuses on the Nedd4 family of E3 enzymes, and more specifically Nedd4-1 and Nedd4-2. Here I cover the functional domains of the protein, and briefly discuss the Nedd4 family of enzymes and the proteins they regulate.
Chapter 2 is a collaborative effort, examining the molecular determinants involved in WW domain-ligand recognition by Nedd4-1 and Nedd4-2. Using the dissociation constant as a reflection of binding affinity, we measured the contribution of specific amino acid residues to the integrity of the interaction between Nedd4 WW domains and target protein PY motifs through mutational analysis. Portions of the work presented in this chapter were published in the Journal of Biological Chemistry and Structure.
In Chapter 3, I present my work investigating possible inter- or intramolecular interactions between Nedd4-2 WW domains and a PY motif within its own HECT domain. Using Nedd4-2
HECT domain PY mutants, I determine if this motif plays a role in regulation of Nedd4-2 catalytic activity. To close the chapter, I propose a model in which inter- or intramolecular
Nedd4-2 interactions regulate its own stability, but not substrate stability, via the HECT domain
PY motif. The work presented in this chapter was published in Biochemical Journal.
2 Chapter 1
In Chapter 4, I present my preliminary characterization of the Nedd4-2 interacting protein
Rnf11, a RING-finger protein. Based on the findings reported here, and elsewhere, this chapter discusses possible biological functions of Rnf11 and implications of the Nedd4-2/Rnf11 interaction.
Finally, in Chapter 5, I describe how the data presented in the previous chapters contribute to our knowledge of Nedd4 WW domain specificity and regulation of HECT domain containing proteins. I conclude by suggesting future experiments aimed at exploring new questions raised by information described in this thesis.
3 Chapter 1
II) The Ubiquitination Pathway
Protein modification by ubiquitin – a process known as ubiquitination or ubiquitylation – is a
post-translational modification involved in regulating a host of critical functions. While the first
characterized role of ubiquitination was to target proteins for degradation by the proteasome, its
involvement in regulating many diverse cellular processes has been subsequently documented.
These processes include protein quality control, cell-cycle progression, membrane protein trafficking and sorting, gene transcription, immune responses and viral infection. Part of the
versatile nature of ubiquitin comes from its ability to modify substrate proteins in its monomeric
form (monoubiquitination), or to tag proteins with a polymer of ubiquitins (polyubiquitination),
giving rise to many types of ubiquitin chains. Given the capacity of ubiquitin modification to
influence stability, function and localization of many different proteins, it is of no surprise that
dysregulation of the ubiquitination pathway has been implicated in numerous diseases, including
cancer and several inherited disorders.
Ubiquitin is a small, 76 amino acid, polypeptide which becomes covalently attached to
target protein lysine residues by the sequential action of three enzymes: the ubiquitin-activating
enzyme (E1), the ubiquitin-conjugating enzymes (E2s), and the ubiquitin protein ligases (E3s).
The E1 enzyme activates the C-terminus of ubiquitin in an ATP-dependent manner, resulting in
the formation of a thiol-ester bond between ubiquitin and the E1 active site Cys residue. The E2
enzyme carries the activated ubiquitin, as a thiol-ester at its active site Cys residue, from the E1
to the E3. The E3 enzyme determines substrate specificity by binding to target substrates, and
catalyzes the transfer of activated ubiquitin from E2 to substrate lysine residues, or to the
growing end of a protein-bound polyubiquitin chain, resulting in formation of a covalent
isopeptide bond between the ubiquitin C-terminal Gly (G76) and the ε-amino group of a
4 Chapter 1
substrate (or ubiquitin) lysine residue (Fig. 1-1A) (Pickart 2001). Notably, this three-step
mechanism initiates all known ubiquitination reactions, regardless of the ultimate fate of the
ubiquitinated protein.
Additional factors further increase the complexity of the ubiquitination process. In some
cases, polyubiquitination requires the additional activity of ubiquitin-chain elongation factors,
termed E4 enzymes. Moreover, much like phosphorylation and dephosphorylation, the process of ubiquitination is reversible, with the cleavage of ubiquitin from substrates catalyzed by specific deubiquitinating enzymes (DUBs). The more recently discovered ubiquitin-binding domains (UBDs), which bind non-covalently to ubiquitin, further demonstrate the broad nature of ubiquitin’s signalling capability (Hicke et al. 2005).
The ubiquitination cascade is hierarchical in nature, with one (or a few) E1 enzyme(s) transferring ubiquitin to a significant, but limited number of E2s. A much larger number of E3s
each cooperate with one or a few E2s (Fig. 1-1B). The following section outlines the enzymes
involved in ubiquitination, what is known about the mechanisms of catalysis, and the protein
domains important for recognizing and propagating the signals coded by ubiquitin.
A) Components of the Pathway
i) Ubiquitin and Ubiquitin-like proteins
Ubiquitin is highly conserved from yeast to humans, and as suggested by its name, is ubiquitous
in biology. The crystal structure of ubiquitin reveals an extremely compact arrangement with a
pronounced hydrophobic core and distinct surface residues formed between several α-helix and
β-sheet secondary structures (Vijay-Kumar et al. 1987). This structure is often referred to as the
ubiquitin or β-grasp fold. Several key features of ubiquitin include: the C-terminal glycine
5 Figure 1-1. The ubiquitination cascade. (A) Free ubiquitin (Ub) is activated by the ubiquitin activating enzyme (E1) in a reaction requiring ATP, resulting in formation of a thioester bond between the C-terminal Gly of Ub and the active site Cys of E1. Activated Ub is then transferred to the active site Cys of the ubiquitin conjugating enzyme (E2). Subsequently, Ub is transferred from E2 to the ubiquitin ligase (E3) bound target forming an isopeptide linkage between Ub and a Lys ε-amino acid group in the target. For HECT E3s, the final step involves an E3-ubiquitin thiol ester. Polyubiquitin chains are formed when ubiquitin is ligated to previously added ubiquitins (ie. K29, K48 or K63 linkages). (B) Hierarchy of the ubiquitination cascade.
(Modified from Hicke et al. 2005).
6 Figure 1-1
A. Mono‐ or Polyubiquitination O O O = Ub = = HNC‐ Ub Ub S‐C‐ Ub S‐C‐ Ub Ub O K = ATP + Ub CO‐ E1 E2 substrate
NH2 K AMP + PPi E1 E2 E3 substrate SH SH SH
B.
E1 Yeast 1 Human ~10 E2 Yeast 11 Human ~100 E3 Yeast 54 Human ~600
7 Chapter 1
(G76), forming an isopeptide bond with substrate (or ubiquitin) lysine residues; a hydrophobic
patch surrounding Ile44, required for interaction of ubiquitin with ubiquitin-binding domains
(Hicke et al. 2005); and seven lysine residues (K6, K11, K27, K29, K33, K48, K63), all of which
can be used for chain formation (Kim et al. 2007), with different, and in some cases undefined,
functional implications (Fig. 1-2A,B). Of note, while most substrates are found to have lysine
conjugated ubiquitin, examples in which the free α-NH2 group of an N-terminal residue of a
protein serves as the primary ligation site have been reported (Ciechanover and Ben-Saadon
2004). More recently, ubiquitin modification of a cysteine side chain by a viral E3 ligase has
been described (Cadwell and Coscoy 2005).
Since the discovery of ubiquitin, several ubiquitin-like proteins (Ubls) have also been identified, with more than 10 Ubls described to date. These proteins do not share much sequence
similarity with ubiquitin or each other, but rather share the same three-dimensional structure, and
a C-terminal glycine residue required for formation of the isopeptide bond with substrate lysine
residues. These proteins use an analogous, but distinct cascade of enzymes for activation and
conjugation from those used for ubiquitination, although some overlap does exist (Kerscher et al.
2006). SUMO1 (small ubiquitin-like modifier, also SUMO2, 3), Nedd8 (neuronal precursor cell
developmentally downregulated 8), and ISG15 (Interferon Stimulated gene 15) are among the most well characterized Ubls. In general, these proteins function different from ubiquitin, and in
some cases, antagonize ubiquitination. A large proportion of SUMOylated proteins are localized
to the nucleus or at the nuclear envelope, suggesting a role for SUMOylation in
nucleocytoplasmic transport. Ubiquitin ligase activity of the anaphase-promoting complex
(APC) is stimulated by SUMOylation (Gutierrez and Ronai 2006), while conjugation of Nedd8,
or neddylation, to members of the cullin family of proteins, which function as scaffolding
8 Figure 1-2. (A) Schematic view of opposite faces of ubiquitin. Ubiquitin has seven Lys residues (in blue), all of which can be used for chain formation. Ile44 is also highlighted (in yellow), which is involved in binding to ubiquitin binding domains. (Modified from Staub and
Rotin 2006). (B) Ubiquitin modifications. Monoubiquitination is the modification of a protein with a single ubiquitin molecule on a single Lys residue. Multimonoubiquitination refers to the modification of several Lys residues with one ubiquitin each. Polyubiquitination is the modification of a protein, on one or more Lys residues, with a ubiquitin chain.
9 Figure 1-2
A.
B. Ub
K Monoubiquitination
Ub Ub Ub Multi‐ K K K monoubiquitination
Ub Ub Ub Ub
K Polyubiquitination
10 Chapter 1
subunits of the SCF (SKP1-CUL1-F-box) ubiquitin-ligases, is important in sustaining SCF
activity (Liu et al. 2002). An interesting example of a protein regulated by ubiquitin-like proteins is the tumour suppressor p53, which undergoes ubiquitination, SUMOylation, and neddylation. While mono- and poly-ubiquitination of p53 result in its nuclear export and degradation, respectively, SUMOylation functionally activates p53 and stabilizes the protein by preventing ubiquitination, and neddylation acts as a inhibitor of transcriptional activity
(Rodriguez et al. 1999; Li et al. 2003; Xirodimas et al. 2004). ISG15, whose expression is induced by interferons (IFN), has been implicated as a positive regulator of IFN-related immune responses (Ritchie and Zhang 2004), and more recently was shown to negatively regulate activity of the E3 enzyme Nedd4 (Malakhova and Zhang 2008). Notably, other than ubiquitin, the only
Ubls known to form chains are certain members of the SUMO family (Bylebyl et al. 2003), and to date, few domains capable of interacting with Ubls have been described, suggesting the signalling capabilities of Ubls may be limited compared with ubiquitin.
ii) Ubiquitin-activating enzymes (E1s)
The E1 enzyme reaction, activating the ubiquitin C-terminus for nucleophilic attack, begins with the ordered binding of MgATP and then ubiquitin, leading to the formation of a ubiquitin adenylate intermediate. This intermediate serves as the donor of ubiquitin to the E1 active site cysteine residue, with each fully loaded E1 carrying two molecules of activated ubiquitin – one as a thiol-ester, the other as an adenylate. E1 is an efficient enzyme allowing for production of sufficient activated ubiquitin for all downstream conjugation reactions, with reported catalytic rates 10- to 100-fold higher than those reported for substrate ubiquitination (Pickart 2001).
11 Chapter 1 iii) Ubiquitin-conjugating enzymes (E2s)
All E2s share a conserved core (~150 aa), consisting of a central β-sheet with flanking helices, and some members have N- or C-terminal extensions that may facilitate specific interactions.
Apart from the E2 cysteine, few catalytic residues have been identified, with the exception of an asparagine residue proposed to play a role in stabilizing the oxyanion intermediate during attack by the acceptor lysine (Wu et al. 2003).
The two main types of E3s for ubiquitin, the RING class and the HECT class, are mechanistically and structurally distinct. Interestingly, the crystal structures of UbcH7 E2 protein complexed to each type revealed that both E3 types interact with the E2 in the same manner, involving many of the same E2 residues in both cases (Huang et al. 1999; Zheng et al.
2000). Recent evidence suggests that, in the case of RING E3s, E2 binding triggers subtle conformational changes (in the E2), stimulating release of ubiquitin from the E2 cysteine and transfer to substrate (Ozkan et al. 2005). Moreover, E2 cannot bind to E1 and E3 at the same time, requiring release of E1 after transfer of ubiquitin to the E2 active site Cys before ubiquitin- thioester linked E2 can bind its cognate E3 (Eletr et al. 2005).
Curiously, if the molecular character of the E2-E3 interaction is conserved across many protein pairs, then why does each E3 need to have its own E2? Having many E2s likely helps to ensure that the pool of activated ubiquitin is appropriately distributed among different E3s, and in addition, accumulating evidence suggests that E2 enzymes play an important role in determining lysine linkages of polyubiquitin chains, affecting substrate fates. This topic will be discussed in more detail below.
12 Chapter 1
iv) Ubiquitin protein ligases (E3s)
E3 ligases impart selectivity on the process of ubiquitination by mediating the transfer of E2-
conjugated ubiquitin to substrate proteins specifically recognized by the E3. These enzymes
interact with target proteins through recognition of motifs, or domains, by regions of the E3
enzyme located outside of the catalytic domain. For example, the E3 Ubr1 (also known as E3α), is a RING E3 that recognizes particular substrates based on the properties of their N-terminal amino acids. This is known as the N-end rule and requires a sterically accessible lysine residue on the target protein to serve as the ubiquitin acceptor site (Bachmair et al. 1986; Bartel et al.
1990). Often, E3 binding sites are short regions of primary sequence, such as the destruction box
(RxALGxIxN), found in substrates of the APC ubiquitin ligase, or the PY-motif (L/PPxY) found in substrates of the Nedd4-family of E3 ligases. In some cases, E3 enzymes or substrate proteins must be covalently modified (i.e. phosphorylated) before an interaction can occur, or require additional factors or adaptor proteins for binding, providing regulation of substrate ubiquitination. Unlike Ubls SUMO and Nedd8, ubiquitin does not seem to modify lysine residues within a specific consensus sequence. Additionally, most E3s appear nonselective with respect to which lysine of a substrate is ubiquitinated, and although some substrates are ubiquitinated at a particular lysine residue, this site specificity likely reflects steric blockade (or lack) of other lysine residues.
Ubiquitin E3s typically utilize one of two catalytic domains - a HECT domain or a RING finger. These domains are very distinct, both structurally and mechanistically. As discussed below, differences in the mechanisms of HECT and RING finger E3 ligases suggest that whereas
HECT E3 ligases also provide a catalytic contribution, the primary role of a RING finger E3 ligase is to function as a molecular scaffold to bridge the E2 and substrate.
13 Chapter 1
1) RING E3s
The RING finger is a ~70 amino acid domain featuring a set of eight conserved cysteine and
histidine residues with distinctive spacing, allowing for coordination of two zinc ions in a unique cross-brace arrangement (Fig. 1-3A). Interestingly, it is the spacing of zinc ligands, rather than
the primary sequence, that is conserved in RING finger family proteins. The structurally related
U-box proteins, which also behave as ubiquitin ligases, possess similar folds to RING fingers,
however, the stabilizing zinc ions are replaced by residues forming buried salt bridges and
hydrogen bonds (Fig. 1-3B) (Aravind and Koonin 2000).
RING finger proteins are thought to act as molecular scaffolds, bringing together the E2-
conjugated ubiquitin and substrate lysine residue. Crystal structures of two RING E3s (c-Cbl
and SCF) reveal substrate binding sites that are quite distant from the E2 active site cysteine
(60Å and 50Å respectively), raising questions on the mechanism of ubiquitin transfer (Zheng et
al. 2000; Zheng et al. 2002). More recently, Ozkan et al. (2005) demonstrated that binding of a
RING finger to E2 UbcH5b triggered subtle long-range conformation changes in the E2,
enhancing its ability to transfer ubiquitin from the thioester to substrate. In the case of RING
E3s, it is likely that ubiquitin transfer occurs spontaneously when the highly labile E2-ubiquitin
thioester bond is presented to a substrate lysine in a favourable conformation.
While E3 enzymes had been thought to dictate whether a substrate becomes
monoubiquitinated or polyubiquitinated, in the case of RING E3 catalyzed ubiquitination, the E2 enzymes appear to play a key role. Recent work demonstrated that the heterodimeric RING E3
BRCA1-BARD1 could catalyze both mono- and polyubiquitination depending on the E2 used, where the ability of a particular E2 to catalyze polyubiquitination correlated with its ability to bind ubiquitin non-covalently, at a site distal to the active site (Christensen et al. 2007). In
14 Chapter 1 complex with RING E3s, the E2 UbcH5b appears to be quite promiscuous, binding to several E3 enzymes, and catalyzing the formation of substrate-bound polyubiquitin chains lacking specificity for any particular lysine residue of ubiquitin, even synthesizing forked ubiquitin chains containing all possible isopeptide linkages (Kim et al. 2007; Windheim et al. 2008).
Conversely, the Ubc13-Mms2 heterodimer and E2-25K E2s direct synthesis of K63- or K48- linked polyubiquitin chains, respectively. The crystal structure of the Ubc13-Mms2 complex revealed that Mms2, a UEV (ubiquitin E2 variant) capable of binding ubiquitin non-covalently, binds an ‘acceptor’ ubiquitin in such a fashion that its K63 residue is in close proximity to the thiol-ester bound ubiquitin at the Ubc13 active site, resulting in ubiquitin transfer from the
Ubc13 active site to K63 of the acceptor ubiquitin (Eddins et al. 2006). In some cases, E2s appear to act sequentially, one E2 acts first to monoubiquitinate the substrate, and a second E2 to extend a polyubiquitin chain (Christensen et al. 2007; Rodrigo-Brenni and Morgan 2007;
Windheim et al. 2008). Some E2s have also been shown to assemble polyubiquitin chains at their active site cysteine which can be transferred to substrate lysine residues (Li et al. 2007;
Ravid and Hochstrasser 2007).
RING finger E3s are further sub-classified as either single-subunit E3s, meaning that no partner besides the E2 is required for substrate ubiquitination, or multi-subunit E3s, in which the
RING finger protein is one subunit of a multiprotein complex. One of the most well characterized single-subunit RING finger proteins is c-Cbl, involved in ubiquitin dependent downregulation of activated receptor protein tyrosine kinases (RTKs), including the epidermal growth factor (EGF) receptor, platelet-derived growth factor (PDGF) receptor, and hepatocyte growth factor (HGF) receptor (Levkowitz et al. 1998; Joazeiro et al. 1999; Waterman et al.
1999; Peschard et al. 2004). Activation of EGFR, inducing its auto-phosphorylation, allows
15 Figure 1-3. RING-finger E3s. (A and B) Schematic of RING-finger and U-box domains. (A)
Numbered residues represent metal coordinating amino acids. The canonical RING finger has
His in position 4, and Cys in positions 1-3, and 6-8. RING-fingers are classified as RING-H2 or
RING-HC depending on whether position 5 is occupied by His or Cys, respectively. (B) In the
U-box the predicted conformation is conferred by hydrogen bonding and salt bridges indicated schematically by dashed lines. (A and B modified from Fang et al. 2003). (C, D, E and F)
Multisubunit RING E3s. (C) The SCF E3 complex is composed of a RING finger protein Rbx1, which together with the scaffold protein Cul1, constitutes the core catalytic activity. Cul1 is associates with an F-box protein through the adaptor protein Skp1, and substrate selection is carried out by the F-box protein. Rbx1 binds to an E2 enzyme and ubiquitin is transferred directly from the E2 to the F-box bound substrate. (D) The ECS (ElonginC-Cul2-SOCS box) complex: The Elongin B/C complex bridges interaction between Cul2 and SOCS-box containing proteins, and substrate ubiquitination is carried out as described in C. (E) Cul3-BTB: BTB proteins incorporate features of Skp1/F-box or ElonginC/SOCS-box dimers, and are thought to bridge the cullin and the substrate in a single polypeptide. (F) The APC is composed of many subunits. Apc2 is a scaffold subunit similar to the cullin proteins, and binds the RING finger protein Apc11. Substrates are recognized by an APC co-activator, such as Cdh1 (shown here) or
Cdc20 (not shown), which interacts with the complex through two subunits, Cdc27 and Apc2 (E modified from Peters 2006).
16 Figure 1-3
A. B.
2 6 3 7 Zn2+ Zn2+ 1 5 4 8
RING finger U-box
C. D. F‐box protein SOCS‐box protein Fbox SOCS C Skp1 Rbx1 B Rbx1 Cul1 Cl2Cul2
Cdc26 E. F. Apc9 Swm1 Cdc16 BTB Cdc27 Cdc23 Apc5 protein Apc4 Cdh1
BTB Rbx1 Apc11 Apc1 Cul3 Apc2
17 Chapter 1 binding of the c-Cbl N-terminal tyrosine kinase binding (TKB) domain, and results in c-Cbl mediated multiple monoubiquitination of the receptor (Levkowitz et al. 1999; Haglund et al.
2003). The importance of c-Cbl in RTK downregulation is exemplified by the oncogenic nature of the retroviral encoded v-Cbl protein, which lacks the RING domain and hence cannot downregulate activated EGFR or other RTKs (Blake et al. 1991).
Multisubunit RING E3s include the APC/C (anaphase-promoting complex/cyclosome) and the cullin-RING finger ligases (Fig. 1-3C,D,E,F). The cullin-RING finger ligases, constituting the largest family of ubiquitin ligases in eukaryotes, share a common catalytic core, composed of a cullin and the RING-finger protein Rbx1, but use different substrate recognition modules with structural similarities. The prototypic cullin-RING E3 is the SCF (Skp1-Cul1-F-box) complex, in which Cul1 binds the adaptor Skp1, which interacts with a large number of F-box proteins acting as substrate receptors (Fig. 1-3C) (Deshaies 1999). At least 3 other SCF-like complexes exist, grouped based on the cullin subunit employed (Cul2, Cul3, Cul4)(Fig. 1-3D,E). In the case of the Cul3-BTB complex, BTB (Bric-a-brac, Tramtrack, and Broad) domain-containing proteins, which bind Cul3, integrate the functions of both adaptor and substrate receptors (Fig. 1-
3E)(Furukawa et al. 2003; Xu et al. 2003). Similar to cullin-RING E3s, the APC/C employs a cullin-like subunit Apc2 and RING-finger subunit Apc11 as the core catalytic complex, ubiquitinating numerous substrates and promoting exit from mitosis (Fig. 1-3F) (Peters 2006).
2) HECT E3s
The HECT (homologous to E6-AP carboxy terminus) domain is an approximately 350 residue region first characterized in the human ubiquitin protein ligase E6-AP (E6-associated protein)
(Huibregtse et al. 1995). Unlike RING fingers, HECT domains are always located C-terminally,
18 Chapter 1
and actively participate in catalysis through a conserved cysteine residue which forms a thiol-
ester intermediate with ubiquitin. HECT E3s, ranging in size from 80 kDa to over 500 kDa, are
modular, with the unique N-terminus of each family member dictating substrate specificity.
The crystal structure of the E6-AP HECT domain revealed an overall L-shaped structure
composed of an elongated N-terminal lobe connected to a globular C-terminal lobe through a
short hinge loop (Huang et al. 1999). Solved in complex with the E2 UbcH7, the structure resolved residues in the N-lobe required for E2 binding, but raised questions about the catalytic
mechanism of ubiquitin transfer, since the HECT and E2 active site cysteines are 40Å apart (Fig
1-4A). A more recent structure of the WWP1 HECT domain revealed a different orientation of the N- and C-lobe relative to one another, closing the gap between E2 and HECT catalytic
cysteine residues to 16Å (Verdecia et al. 2003). Mutational analysis of the hinge loop between
the two lobes revealed that flexibility of this loop is essential for E3 activity, suggesting that
large movement of the C-lobe is likely required for ubiquitin transfer (Verdecia et al. 2003) (Fig.
1-4B). Clearly, further structural and dynamic studies are required to define the catalytic
mechanism employed by the HECT domain.
As HECT E3s actively participate in catalysis, the factors determining polyubiquitin chain
synthesis by a HECT E3 differs from those of RING E3s. The mechanism responsible for
polyubiquitin chain synthesis has been extensively studied for two HECT E3s, E6-AP and
KIAA10, and these appear to employ distinct mechanisms for chain formation. For E6-AP,
chain synthesis occurs through the reaction of two covalently bound ubiquitins: one linked to the
E2 and the other linked to the E3, with the E3-linked ubiquitin K48 attacking G76 of the E2-
linked ubiquitin (Fig. 1-4C) (Wang and Pickart 2005). For KIAA10, which synthesizes both
K29- and K48-linked chains, the ubiquitin contributing the lysine residue, termed the acceptor
19 Figure 1-4. The HECT domain. (A and B) Crystal structures of HECT domains. HECT domain C lobes are coloured purple, N lobes are coloured blue, and hinge loops are coloured gold. Dashed red lines indicate distances between the HECT domain catalytic cysteine and the
E2-ubiquitin thioester bond. (A) E6AP-UbcH7 complex with UbcH7 coloured green and ubiquitin modeled in red. The E6AP-UbcH7 protein complex depicted is an experimentally determined structure while the ubiquitin chain shown is modeled. (B) WWP1/AIP5 HECT domain with modeled UbcH5b-ubiquitin complex, based on the E6AP-UbcH7 complex structure. UbcH5b is coloured in green and ubiquitin is coloured red. (A and B modified from
Verdecia et al. 2003). (C and D) Mechanism for polyubiquitin chain synthesis catalyzed by
HECT E3s. (C) As the acceptor Ub is covalently linked to the HECT cysteine, the chain is built up at the E3 active site. The lysine of the acceptor Ub attacks Ub-G76 in the E2-Ub thiol ester
(in an E2/E3 heterodimer). (D) Sequential model: because the acceptor Ub binds non- covalently, the chain is built up as a free entity, with the acceptor Ub contributing the lysine residue to the isopeptide bond. (C and D modified from Wang and Pickart 2005).
20 Figure 1-4
A. B.
C.
D.
21 Chapter 1 ubiquitin, binds non-covalently to KIAA10, with the donor ubiquitin covalently bound to the
KIAA10 active site cysteine (Fig. 1-4D). Therefore, while E6-AP preassembles polyubiquitin chains at its active site before transfer to substrate lysine residues, KIAA10 transfers a single ubiquitin at a time (Wang and Pickart 2005). Further studies with KIAA10 indicate that interactions with regions of the acceptor ubiquitin surface that are proximal to the targeted lysine residues play a major role in linkage determination (Wang et al. 2006). Interestingly, studies examining chain formation of both RING and HECT E3s have shown that the type of isopeptide linkages formed by RING E3s is determined by the E2, and a single E3 can form different types of chains depending on the E2. By contrast, different HECT domain E3s using the same E2, form homogeneous chains composed of different linkages (Kim et al. 2007).
The most well characterized family of HECT E3s is the Nedd4 family, which is the focus of this thesis and will be discussed in more detail throughout the remainder of the Introduction.
v) E4 enzymes
In some cases, polyubiquitination of substrates is believed to require the additional activity of certain ubiquitin-chain elongation factors, which have been termed E4 enzymes. The first described E4 was the yeast protein Ufd2, shown to elongate short ubiquitin chains attached to a substrate, an activity essential for its function in targeting misfolded endoplasmic reticulum (ER) proteins for degradation (Koegl et al. 1999; Richly et al. 2005). Ufd2 contains a U-box domain
(Fig. 1-3B), which as mentioned, adopts a similar fold to that of the RING finger. However, recent evidence demonstrates that Ufd2 behaves as a bona fide E3, based on its ability to bind directly to the E2 enzyme and catalyze the formation of unanchored di-ubiquitin chains, and to ubiquitinate itself (Tu et al. 2007). The mammalian U-box protein CHIP binds chaperones
22 Chapter 1
Hsp70 and Hsp90, and displays E3 activity towards a variety of chaperone-bound substrates,
leading to their proteasomal degradation (Cyr et al. 2002). However, CHIP was also shown to work in collaboration with a RING finger protein (Parkin), where only a combination of the two proteins (with E1 and E2) efficiently polyubiquitinated the substrate (Imai et al. 2002).
Collectively, this suggests that some E4s may be better described as specialized E3s, since they contain intrinsic ubiquitin ligase activity and may ubiquitinate substrates in the absence of additional E3s.
Other types of E4 enzymes have been described which do not contain a U-box or other domains with homology to the RING or HECT domains. The transcriptional cofactor p300 has been shown to recognize Mdm2-produced monoubiquitinated p53 as a substrate for further ubiquitin-chain elongation (Grossman et al. 2003). Polyubiquitination of the general amino acid permease of S. Cerevisiae, Gap1, requires a complex of two proteins, Bul1 (binds to ubiquitin ligase) and Bul2. While the HECT E3 ligase Rsp5 is sufficient for Gap1 monoubiquitination, the
Bul1-Bul2 complex is required for Gap1 polyubiquitination, targeting Gap1 to the vacuole, where it is degraded (Helliwell et al. 2001). Thus, p300 and the Bul1-Bul2 complex likely represent ‘true’ E4 enzymes, due to their inability to catalyze ubiquitination in the absence of an
E3.
vi) Deubiquitinating enzymes (DUBs)
Deubiquitinating enzymes (DUBs) are ubiquitin proteases, catalyzing cleavage of the isopeptide bond between substrate lysine residues and ubiquitin, or between two ubiquitin molecules. The human genome encodes ~80 DUBs, which are divided into five subfamilies based on their catalytic domain. Four classes – the ubiquitin-specific proteases (USPs), ubiquitin C-terminal
23 Chapter 1 hydrolases (UCHs), Machado-Joseph disease proteases (MJDs) and ovarian tumour proteases
(OTUs) – are cysteine proteases, whereas the JAMM motif proteases are zinc metalloproteases.
DUBs are critical for regulating the pool of free ubiquitin, through cleavage of ubiquitin precursor proteins which must be ‘activated’ by specific proteases to expose their C-terminal glycine residue, and through deubiquitination of proteins arriving at the proteasome, allowing ubiquitin recycling. Some DUBs show a preference for deconjugating ubiquitin chains with specific lysine linkages, such as the yeast DUB Ubp2, which reverses Rsp5-catalyzed K63- polyubiquitination of Rsp5 substrates (Kee et al. 2005). A remarkable variation on this is the protein A20, possessing both E3 ligase and DUB activity in a single protein, downregulating NF-
κB signalling through the cooperative activity of its two ubiquitin-editing domains (Wertz et al.
2004).
vii) Ubiquitin-binding domains (UBDs)
Ubiquitin-binding domains (UBDs) bind non-covalently to ubiquitin and are responsible for mediating most of the downstream effects of protein ubiquitination. Ubiquitin-binding proteins, often termed ‘ubiquitin receptors’, generally have small (20-150 aa), independently folded UBDs that can interact directly with monoubiquitin and/or polyubiquitin chains, and are found in many proteins with diverse functions and locations. The first UBD to be characterized was the ubiquitin-interacting motif (UIM), of the proteasome subunit S5a/Rpn10, which functions as a receptor for polyubiquitinated proteins that are destined for degradation (Young et al. 1998). At about the same time, the ubiquitin-associated (UBA) domain, identified as a sequence pattern common to a subset of proteins involved in ubiquitination and deubiquitination reactions, was shown to bind directly to ubiquitin (Hofmann and Bucher 1996; Bertolaet et al. 2001; Wilkinson
24 Chapter 1
et al. 2001). While the UIM consists of a single α-helix, UBA domains are compact three-helix
bundles and bind in the centre of a ‘sandwich’ between two ubiquitin moieties, explaining their
preference for binding polyubiquitin chains (Trempe et al. 2005; Varadan et al. 2005). UBDs are structurally diverse, and include α-helical domains, like the UIM and UBA domains, and zinc-binding domains [zinc finger (ZnF)], as well as domains previously characterized to bind other proteins or lipids (Table 1-1). For example, a subset of SH3 domains, including the three
SH3 domains of CIN85, were recently shown to bind ubiquitin, using the same surface to bind ubiquitin as they do to bind proline-rich ligands [(He et al. 2007; Stamenova et al. 2007),
Appendix I].
Despite the structural diversity exhibited by UBDs, most interact with the same hydrophobic surface on ubiquitin surrounding Ile44 (Fig. 1-2A) (Hicke et al. 2005). While most
UBDs bind both mono- and polyubiquitin, the majority show a preference for polyubiquitin, with some even binding preferentially to either K48- or K63-linked chains. The binding affinities of these domains for monoubiquitin span a wide range, but are most commonly weak with Kd>100
µM. Clustering of low-affinity, high-specificity interactions can be advantageous, resulting in protein networks with a built-in dynamic instability, which allows for rapid assembly and disassembly of complexes. Furthermore, when necessary, low-affinity interactions can be leveraged into higher-avidity interactions, through the presence of tandem UBDs, modification multiplicity (i.e. poly- or multi-monoubiquitination), oligomerization of ubiquitinated proteins and ubiquitin receptors, or by further contacts between the ubiquitin receptor and the ubiquitinated target (Hicke et al. 2005).
Many UBD containing proteins are also monoubiquitinated in a UBD-dependent manner, a process termed coupled monoubiquitination (Polo et al. 2002). Notably, the UBDs do not
25 Table 1-1 Ubiquitin-binding domains
Domain Description Selected References UIM Single α-helix centered around a conserved alanine (Fisher et al. 2003; Swanson (Ub interacting motif) residue et al. 2003)
DUIM UIM variant. Two UIMs interlaid on a single helix (Hirano et al. 2006) (double-sided UIM) with both faces capable of binding Ub
MIU UIM variant. Single α-helix interacting with Ub in (Lee et al. 2006; Penengo et (motif interacting with Ub) opposite orientation of UIM al. 2006)
UBA Compact three-helix bundle binding “sandwiched” (Trempe et al. 2005; Varadan (Ub associated) between two Ub moieties of a polyUb chain et al. 2005)
CUE Structurally related to UBA. (Kang et al. 2003; Prag et al. (coupling of Ub to ER degradation) 2003)
GAT Found in GGA and TOM1 and binds Ub through (Akutsu et al. 2005; (GGA and TOM) two distinct sites. Also interacts with other proteins Kawasaki et al. 2005) including GTP-bound ARFs. VHS Binds to an acidic di-leucine motif in the (Mizuno et al. 2003) (Vps27/Hrs/STAM) cytoplasmic domain of sorting receptors, as well as to Ub
UEV Structurally analogous to E2 catalytic core but (McKenna et al. 2003; (Ubc E2-variant) lacking the catalytic cysteine residue Sundquist et al. 2004)
NZF 30-residue domain built around a single zinc- (Alam et al. 2004) (Npl4 ZnF) binding site. Only NZF domains containing the ‘TFΦ fingerprint’ bind Ub
PAZ Single zinc-binding site in N-terminus which is (Reyes-Turcu et al. 2006) (polyUb-associated ZnF) fused to an α/β fold A20 ZnF Zinc finger first identified in A20 protein, also has (Lee et al. 2006; Penengo et ubiquitin ligase activity al. 2006) UBZ Zinc finger occuring in Y-family DNA (Bienko et al. 2005) (Ub-binding ZnF) polymerases, involved in DNA repair UBM 30-residue domain with predicted helical structure, (Bienko et al. 2005) (Ub-binding motif) also found in Y-family DNA polymerases
GLUE Has a PH domain fold and binds both Ub and (Slagsvold et al. 2005) (GRAM-like Ub binding in EAP45) phosphoinositides Pru PH domain fold, found in Rpn13 subunit of (Husnjak et al. 2008; (pleckstrin-like receptor for Ub) proteasome, binding K48-diubiquitin with 90 nM Schreiner et al. 2008) affinity
Jab1/MPN Variant of JAMM motif (found in some DUBs) but (Bellare et al. 2006) lack key residues for catalysis PLU Found in Doa1, but no significant sequence identity (Mullally et al. 2006) (PLAA family Ub binding) to other UBDs SH3 Subset of SH3 domains (containing an essential (He et al. 2007; Stamenova et (Src-homology 3) Phe residue) bind Ub with same suface used for al. 2007), Appendix I binding proline-rich ligands
26 Chapter 1 contain lysine residues that function as acceptors for ubiquitin, suggesting their requirement involves another aspect of the process. The work of Woelk et al. (2006) demonstrates that coupled monoubiquitination results from interaction between the UBD and a ubiquitin moiety covalently attached to an E3 enzyme, logically, as many E3s are known to undergo self- ubiquitination. Furthermore, monoubiquitination of ubiquitin receptors has been shown to result in formation of intramolecular interactions between ubiquitin and their UBDs, inhibiting their ability to bind to and control the functions of ubiquitinated targets (Hoeller et al. 2006).
However, this also provides an explanation for why these proteins become monoubiquitinated, as opposed to polyubiquitinated, since binding of ubiquitin to the UBD occludes ubiquitin lysine residues necessary for polyubiquitination (Shekhtman and Cowburn 2002). More recently, ubiquitin-binding proteins have been shown to undergo E3-independent monoubiquitination through direct recruitment of Ub-loaded E2s by the UBD (Hoeller et al. 2007).
Coupled monoubiquitination likely plays a regulatory role, at least in some cases. For example, monoubiquitination of the UIM-containing protein Epsin negatively influences its in vitro binding to some membranes and protein partners, but not others (Chen et al. 2003).
Additionally, coupled ubiquitin binding and ubiquitination might form a signal relay network, where ubiquitination of a ubiquitin receptor may lead to interaction with other downstream
UBD-containing proteins (Di Fiore et al. 2003).
27 Chapter 1
B) Fates of ubiquitinated proteins i) Protein degradation: The Ubiquitin-Proteasome Pathway
The first characterized function of protein ubiquitination was to target proteins for degradation by the 26S proteasome, a huge multisubunit protease consisting of a catalytic core particle (the
20S complex) and regulatory particle (the 19S complex). The 20S complex is made up of four seven-membered rings – 2 central β-rings bearing peptidase activity and two distal α-rings, forming the gate of the protease chamber. The 19S regulatory particle, made up of base- and lid- complex, can bind one or both ends of the 20S complex, and in addition to recognizing ubiquitinated substrates, also catalyzes the removal of ubiquitin chains from substrates. It also contains six ATPases implicated in substrate unfolding and translocation (reviewed in (Pickart and Cohen 2004)).
Proteins tagged with a chain of 4 or more ubiquitins are recognized by the proteasome subunit Rpn10/S5a, which contains a UIM (Deveraux et al. 1994), or Rpn13/ADRM1/ARM1, which interacts with ubiquitin through a novel UBD termed the pleckstrin-like receptor for ubiquitin (Pru) (Husnjak et al. 2008; Schreiner et al. 2008). However, it is likely that still- unidentified proteasomal ubiquitin receptors exist (Husnjak et al. 2008). In addition, a family of ubiquitin-like (UBL)-UBA proteins, which are not integral proteasome subunits, deliver ubiquitinated targets to the proteasome, including Rad23 (hHR23a/b in humans), Dsk2 (hPLIC-
1/2 in humans) and Ddi1 (Chen and Madura 2002; Verma et al. 2004; Kaplun et al. 2005). UBA domains of these proteins bind ubiquitin, whereas their UBL domains interact reversibly with the proteasome (Elsasser et al. 2002; Walters et al. 2002). It has been proposed that most proteolytic substrates are escorted to the proteasome by ubiquitin-binding proteins, in part
28 Chapter 1
protecting these ubiquitinated proteins against the activity of ubiquitin proteases (Richly et al.
2005).
Proteins are deubiquitinated prior to degradation by the proteasome, preventing ubiquitin
degradation, a rather slow process owing to its stability, and allowing ubiquitin to be recycled.
The lid subunit Rpn11/POH1, a JAMM ubiquitin protease, is the main DUB associated with the proteasome (Lam et al. 1997), however, two other DUBs are also known to associate with the proteasome, Upb6/USP14 and Uch37/UCHL5. Recently, Ubp6 was shown to antagonize the chain elongating activity of the E4 Hul5 (KIAA10 in humans), also a proteasome-associated protein, suggesting proteasomal control over the final, and only irreversible, step in protein breakdown (Crosas et al. 2006). Uch37 trims ubiquitins sequentially from the distal end of a
chain, and has been suggested to oppose, rather than facilitate, degradation by rescuing poorly
ubiquitinated proteins (Lam et al. 1997; Yao et al. 2006).
Recognition of ubiquitinated substrates by the proteasome typically involves conjugation
of polyubiquitin chains linked through K48 (Pickart 1997; Thrower et al. 2000). However, K29-
and K63-linked chains can also be recognized and degraded by the proteasome (Finley et al.
1994; Mastrandrea et al. 1999; Hofmann and Pickart 2001; Saeki et al. 2005; Kim et al. 2007).
Thus, while K29- and K63-linked chains have roles in non-proteolytic signalling, these chains represent proteasomal targeting signals for a subset of proteins. This, together with data demonstrating reduced proteasomal binding by these chain types (Hofmann and Pickart 2001), may indicate a physiologically relevant role for these chains in decelerated protein degradation.
29 Chapter 1 ii) Role of ubiquitination in trafficking, endocytosis and protein sorting
In addition to targeting proteins for proteasomal degradation, ubiquitin has also been implicated as a sorting signal regulating multiple steps of protein trafficking through the endocytic and biosynthetic pathways. The role of ubiquitin in endocytosis of plasma membrane proteins, protein sorting, and viral budding is discussed below.
1) Endocytosis of plasma membrane proteins
In mammalian cells, clathrin-mediated endocytosis is facilitated by internalization signals including tyrosine-based endocytic signals such as YxxΦ (Φ representing a bulky hydrophobic residue), and di-leucine based endocytic motifs, which direct endocytosis through binding the clathrin-adaptor AP-2. However, not all receptors possess these internalization signals, and over the last decade a role for ubiquitin in endocytosis of plasma membrane protein has been well established.
In yeast, the vast majority of plasma membrane proteins rely on ubiquitination as an internalization signal (Dupre et al. 2004). The first evidence for this came from studies on the yeast mating pheromone a-factor transport Ste6 (Kolling and Hollenberg 1994) and Ste2, a G- protein coupled receptor (GPCR) for α-factor (Hicke and Riezman 1996). Endocytosis defective yeast strains accumulate ubiquitinated forms of Ste6 and Ste2, and deletion of genes encoding proteins of the ubiquitin pathway is associated with stabilization of Ste2 and Ste6. Genetic studies in yeast demonstrated a requirement for the HECT domain containing ubiquitin ligase
Rsp5 for ubiquitination and internalization of the yeast transporters Fur4, a uracil permease, and the general amino acid permease Gap1 (Hein et al. 1995). Rsp5 was shown to mediate rapid
+ downregulation of Gap1 in response to NH4 , which is preferred as a nitrogen source over amino
30 Chapter 1 acids, targeting Gap1 for vacuolar degradation (Springael and Andre 1998). Similarly, Fur4 is ubiquitinated in an Rsp5-dependent manner and undergoes constitutive endocytosis, the rate of which is increased under stress conditions or in the presence of excess uracil (Galan et al. 1996;
Seron et al. 1999).
In mammals, ubiquitination has been implicated in the endocytosis of several types of transmembrane proteins, including receptor tyrosine kinases (RTKs), GPCRs, transporters, ion channels, and T-cell and cytokine receptors. However, here, the role for ubiquitin is less clear.
For example, some receptors, such as the growth hormone receptor (GHR) and β2 adrenergic receptor, require ubiquitination of associated adaptor proteins for internalization, but not of themselves (Govers et al. 1998; Govers et al. 1999; Shenoy et al. 2001). Ubiquitination of the epidermal growth factor receptor (EGFR) is also not required for its ligand-induced internalization, since in cells expressing dominant negative mutants of the EGFR E3 ligase Cbl, or in cells deficient in Cbl, where EGFR does not undergo ligand-induced ubiquitination, the rates of internalization are unaffected (Levkowitz et al. 1998; Duan et al. 2003). The same has been observed for the GPCR CXCR4, as mutation of the ubiquitin-acceptor lysine residues does not affect receptor internalization (Marchese and Benovic 2001). However, ubiquitination does appear to play an important role in targeting proteins for degradation, since in all of the cases mentioned above, the absence of receptor ubiquitination resulted in defects in endosomal sorting and lysosomal degradation (see below).
At least in some cases, the lack of a requirement for receptor ubiquitination may reflect the utilization of alternative pathways for receptor internalization. EGFR is endocytosed through both clathrin-dependent and clathrin-independent pathways. While clathrin-dependent internalization of the EGFR occurs at low doses of EGF and does not involve receptor
31 Chapter 1 ubiquitination, at high doses of EGF, the receptor is ubiquitinated and is endocytosed through clathrin-independent, lipid raft-dependent route (Sigismund et al. 2005). The transforming growth factor β (TGF-β) receptor also utilizes clathrin-dependent and -independent pathways for endocytosis (Di Guglielmo et al. 2003). In both of the above cases, segregation of receptors into distinct endocytic compartments likely serves to regulate receptor signalling, as clathrin- mediated endocytosis allows continued receptor signalling from the endosomal compartment, whereas lipid raft-dependent internalization leads to receptor degradation.
At the plasma membrane, Epsin, Eps15 and their yeast homologues mediate endocytic internalization of a variety of ubiquitinated membrane proteins (reviewed in (Traub and Lukacs
2007)). These proteins bind both ubiquitin and clathrin, linking ubiquitinated cargo to the endocytic machinery. Notably, these endocytic adaptor proteins are also monoubiquitinated, and such modification may be required for assembly of endocytic complexes with other UBD- containing proteins and/or be involved in regulation of these proteins through intramolecular ubiquitin-UBD interactions (Hoeller et al. 2006).
2) Protein Sorting at the MVB/Late Endosome
Internalized plasma membrane proteins are either recycled back to the plasma membrane or targeted to the endosomal system. At the sorting endosome, proteins destined for degradation are sorted into intralumenal vesicles that bud from the limiting (outer) membrane, giving rise to multivesicular bodies (MVBs). Fusion of MVBs with the lysosome (or vacuole in yeast), initiates degradation of internal vesicles and their contents by hydrolytic enzymes. Membrane proteins remaining on the MVB limiting membrane are spared degradation and can be recycled back to the plasma membrane, or transported to other sites in the cell. Mechanisms involved in
32 Chapter 1
targeting of protein cargo into the MVB pathway is highly conserved between yeast and
mammalian cells, and involves ubiquitin modification and recognition of ubiquitin by different
ubiquitin-binding domains. As mentioned above, for many plasma membrane proteins ubiquitin
serves as an essential signal for entry into MVB vesicles at the endosome, and resulting
lysosomal degradation.
More than 50 gene products that are involved in vacuolar protein sorting (Vps) have been
found to function at distinct stages of protein transport to the vacuole; 17 of them are required for targeting to the MVB pathway, and are termed class E Vps proteins due to the malformed late endosome/MVB, known as the ‘class E compartment’, resulting from the loss of function of these proteins (Bankaitis et al. 1986; Rieder et al. 1996). Many of these class E Vps proteins form part of endosomal sorting complexes required for transport (ESCRTs). Four distinct
ESCRTs, ESCRT-0, -I, -II, and -III, are recruited to endosomes through both protein and lipid interactions. ESCRT-0, -I, and -II are enlisted early in the MVB pathway and have ubiquitin-
interacting modules necessary for cargo sorting. The Hrs/STAM complex, making up ESCRT-0,
is thought to recruit clathrin to endosomes and sequesters ubiquitinated cargo in clathrin-coated
microdomains (Clague 2002). Indeed, mutations that inactivate ubiquitin binding by this
complex result in specific defects in sorting ubiquitinated cargo into the MVB interior (Bilodeau
et al. 2002). Interestingly, Hrs binds both Eps15 (Sorkina et al. 1999), which acts at the
internalization step, and Tsg101 (Clague and Urbe 2003), a component of ESCRT-I, suggesting
that Hrs may also play a role in bridging receptor internalization and protein sorting. ESCRT-
III, the final complex in the pathway, does not seem to bind ubiquitin, but instead recruits
deubiquitinating enzymes (DUBs) to remove ubiquitin from the cargo before incorporation into
MVBs. ESCRT-III also recruits machinery to facilitate disassembly of the ESCRT complexes.
33 Chapter 1
The details and protein components of the ESCRT pathway has been reviewed recently and will not be covered in further detail here (Williams and Urbe 2007).
3) Protein Sorting at the Trans-Golgi Network (TGN)
Ubiquitin has also been shown to direct proteins to the endosomal/MVB pathway from the TGN, in some cases serving as a quality control mechanism to divert damaged or functionally unnecessary proteins to lysosomes, preventing them from appearing at the cell surface. This has been shown in yeast for nutrient transporters such as Gap1 and Fur4, which are sorted directly from the TGN to the endosome/MVB pathway as a direct result of ubiquitination, occuring in the presence of excess transporter substrate (Helliwell et al. 2001; Blondel et al. 2004). Here, a family of monomeric clathrin-binding proteins, the GGAs [Golgi-localizing, γ-adaptin ear domain homology, ADP-ribosylation factor (ARF)-binding proteins], play a key role in sorting proteins from the TGN to endosomes (Bonifacino 2004). These proteins have previously been shown to mediate non-ubiquitin dependent transport of the mannose-6-phosphate receptor between the TGN and endosomes (Puertollano et al. 2001; Ghosh et al. 2003). The discovery that they could also bind ubiquitin via their GAT domains, led to the finding that GGAs are necessary for delivery of ubiquitinated cargo from both biosynthetic and endosomal pathways to the lysosome (Puertollano and Bonifacino 2004; Scott et al. 2004). Sorting of Gap1 from the
TGN requires GGA2, since in ∆gga2 yeast cells, or in cells expressing gga2 lacking a GAT domain, Gap1 is targeted to the plasma membrane even in the presence of a good nitrogen source
(Scott et al. 2004). Our lab recently demonstrated a requirement for GGA3 in proper sorting of the lysosomal membrane protein LAPTM5 from the TGN to lysosomes. Interestingly,
34 Chapter 1
ubiquitination of LAPTM5 is not required; rather LAPTM5 associates with the ubiquitin ligase
Nedd4 and interacts with ubiquitinated GGA3 through its UIM (Pak et al. 2006, Appendix II).
The efficiency of a single ubiquitin moiety for these three sorting events is unclear. In yeast,
monoubiquitination (Fig. 1-2B) appears to be sufficient for all three steps, however multi-
ubiquitination can accelerate endocytosis and sorting (Galan and Haguenauer-Tsapis 1997).
Proteins containing multiple lysine residues can undergo multi-monoubiquitination (Fig. 1-2B), like the EGF receptor (Haglund et al. 2003), while other proteins, such as MHC-I and TrkA, have a single lysine that may be modified by a K63-linked polyubiquitin chain (Geetha et al.
2005; Duncan et al. 2006). Accumulating evidence indicates that a single ubiquitin does not efficiently signal internalization in higher eukaryotes, a result which may be explained by the low affinity of ubiquitin receptors for monoubiquitin, as discussed earlier (Barriere et al. 2006;
Hawryluk et al. 2006).
4) Viral budding
The process of viral budding is analogous to budding of vesicles into the MVB, in that both
produce cytoplasm containing vesicles budding away from the cell. Indeed, many viruses utilize
components of the cellular sorting machinery to promote their egress from infected cells. Viral
proteins from several virus types, including retroviruses, filoviruses, rhabdoviruses and
arenaviruses all contain one or more short recognition motifs for proteins involved in MVB
cargo sorting (Demirov and Freed 2004; Morita and Sundquist 2004). These include P(T/S)AP
motifs that bind the UEV domain of Tsg101 and PY motifs (PPxY) that bind the WW domains
of Nedd4 family members. For example, binding of HIV-1 PTAP motif to Tsg101 mimics the
35 Chapter 1
normal binding of Hrs to Tsg101, allowing recruitment of the virus to the ESCRT-I complex, and promoting recruitment of ESCRT-II and -III complexes (Pornillos et al. 2003). Therefore association of the virus with the ESCRT complexes allows it to gain entry in the MVB, from where it can bud out of the cell via exocytosis, or alternatively it can bud directly from the plasma membrane by local assembly of the ESCRT complexes at the membrane.
The involvement of ubiquitin in viral budding is highlighted by the finding that proteasome inhibitors, which decrease the cellular pool of free ubiquitin, lead to inhibition of viral release (Schubert et al. 2000; Harty et al. 2001), and the fact that a substantial fraction of viral Gag proteins are monoubiquitinated (Ott et al. 2000). Furthermore, as mentioned, many
viral proteins (e.g. Ebola late domain) contain PY motifs that bind directly to ubiquitin ligases of
the Nedd4 family. The requirement for Nedd4 binding, as well as ubiquitin acceptor lysine
residues in viral proteins, suggests that ubiquitination of viral proteins by Nedd4 family members
is necessary for budding (Vana et al. 2004). Some viral proteins contain both PTAP and PY
motifs, in which case ubiquitination by Nedd4 family members is a necessary prerequisite for
Tsg101 recruitment and subsequent viral budding (Blot et al. 2004; Vana et al. 2004).
36 Chapter 1
III) Nedd4 family of E3 ligases
Nedd4 (neural-precursor cell-expressed developmentally downregulated), a HECT-type E3
ligase, was originally identified in a screen for genes developmentally downregulated in the early
embryonic mouse central nervous system (Kumar et al. 1992). Subsequently, related proteins
with similar structures have been characterized. Nedd4 and Nedd4-like proteins are found in
eukaryotes from yeast to mammals and are defined by a similar domain organization, namely a
C2 domain, 2-4 WW domains, and a C-terminal HECT domain (Fig. 1-5). Although the
common domain make-up of this family might suggest redundant functional roles, recent work
has ascribed specific functions for individual members and has proposed mechanisms underlying this specificity. This section provides an overview of functional properties of the modular
domains comprising the Nedd4 family and highlights physiological processes regulated by
Nedd4 proteins, with focus on mammalian Nedd4-1 and Nedd4-2 family members.
A) Function of the different domains of Nedd4 proteins
i) C2 domain
The C2 domain was first identified in classical protein kinase C isoforms as a Ca2+-dependent
phospholipid binding domain (Coussens et al. 1986; Knopf et al. 1986). This domain, forming
an eight-stranded β-sandwich structure, interacts with a variety of phospholipids and proteins,
and is involved in protein localization and trafficking. The C2 domain of Nedd4 mediates its
Ca2+-dependent apical localization in polarized epithelial cells, facilitated through interaction with the apical raft associated protein Annexin XIIIb (Plant et al. 1997; Plant et al. 2000). In yeast, the Rsp5 C2 domain is required for its localization to endosomal membranes. A mutant
Rsp5 lacking the C2 domain is deficient in ubiquitination and sorting of biosynthetic cargo,
37 Figure 1-5. The Nedd4 family. Nedd4/Nedd4-like family members from yeast, worm, fly, and mammalian species are shown. They all contain a C2 domain (which in some cases can be spliced out, dashed lines), WW domains, and a HECT domain. Diagrams are not to scale
(modified from Staub and Rotin 2006).
38 Figure 1-5
39 Chapter 1
whereas endocytic cargo is ubiquitinated and sorted efficiently by this mutant (Dunn et al.
2004). Recently, a subset of Nedd4 family members were shown to be autoinhibited by
intramolecular interactions between the C2 and HECT domains, protecting these E3s and their
substrates from degradation (Wiesner et al. 2007). This observation is consistent with earlier
work demonstrating the ability of Nedd4-1 to inhibit ENaC only upon removal of its C2 domain
(Kamynina et al. 2001c; Snyder et al. 2001).
ii) WW domains
WW domains are small (~40 amino acid) protein-protein interaction modules containing two highly conserved tryptophans and an invariant proline (Sudol et al. 1995). WW domains, found
in a variety of different proteins, bind various proline-containing peptides within their target
molecules, and have been classified based on these recognition sequences. Class I WW
domains, which include the Nedd4 family WW domains, bind PY motifs (L/PPxY), Class II
domains bind PxxP sequences, Class III domains interact with regions rich in Pro and Arg, or
containing Pro, Met, and Gly, and Class IV domains bind phosphoSer- or phosphoThr-Pro
(pS/T-P) sequences (Sudol and Hunter 2000). NMR and crystallography studies have shown that
WW domains adopt a compact three-stranded, antiparallel β-sheet structure, forming a
hydrophobic ligand-binding groove together with a conserved binding pocket, termed the XP
groove (Huang et al. 2000; Verdecia et al. 2000; Kanelis et al. 2001; Pires et al. 2001; Toepert et
al. 2003; Pires et al. 2005; Kanelis et al. 2006). These solved structures reveal that core binding
residues in ligands, regardless of type, adopt a PPII helical conformation and contact
homologous residues in the WW domain. Additional binding energy and specificity is provided
by contacts outside these core regions, as seen most notably in the Nedd4 WW4-βENaC complex
40 Chapter 1
and the Nedd4 WW3-Comm PY motif complex. Both of these complexes are studied in Chapter
2.
iii) HECT domain
The HECT domain confers E3 ubiquitin ligase activity to Nedd4 family members as was
discussed earlier. Details of the HECT domain structure and function were covered in Section
II.A.iv.2.
B) Members of the Nedd4 family
While Saccharomyces cerevisiae has only one Nedd4 family member, Rsp5, other eukaryotes
have several, which appear to posses both redundant and specialized functions (Fig. 1-5). As
discussed throughout the Introduction, Rsp5 regulates trafficking of numerous yeast plasma
membrane proteins, including G protein-coupled receptors for pheromones such as Ste2, ABC-
transporters (ATP-binding cassette) like Ste6, and transporters of the MFS (Multi Facilitators
Superfamily) such as Gap1 and Fur4. Rsp5 also has a role in regulating non-membrane proteins,
such as the large subunit of RNA polymerase II (Rpb1), which becomes ubiquitinated and
degraded in response to DNA damage (Beaudenon et al. 1999). Interestingly, Rpb1 and several
plasma membrane Rsp5 substrates do not contain PY motifs, suggesting that binding may occur
through alternate motifs or that the interaction between these proteins and Rsp5 is indirect. Rsp5
is a critical gene, as rsp5 loss-of-function mutations are lethal (Hein et al. 1995). In contrast,
genetic disruption of any one of the three Schizosaccharomyces pombe family members, pub1, pub2, and pub3, results in a viable phenotype, suggesting there may be some overlapping functions (Nefsky and Beach 1996; Tamai and Shimoda 2002).
41 Chapter 1
Functions of other Nedd4-like family members have been delineated in a number of signalling pathways. DNedd4, the Drosophila Nedd4-1 orthologue, plays a critical role in regulating axon guidance at the central nervous system midline, as well as in neuromuscular synaptogenesis through ubiquitination and downregulation of its target protein Commisureless
(Comm) (Myat et al. 2002; Ing et al. 2007). Smurf1 and 2 are involved in regulating the TGF-β signalling pathways by inducing ubiquitination and degradation of specific pathway components.
DSmurf also plays an essential role in the analogous decapentaplegic (DPP) pathway, which regulates dorsoventral patterning in the developing embryo. Disruption of mouse Itch causes immunological disorders, mainly attributed to the role of Itch in downregulation of Notch receptors. In accord, Suppressor of deltex, the Drosophila Itch orthologue, also regulates Notch signalling. Furthermore, Nedd4 family members Nedd4-1, Nedd4-2, WWP2 and AIP4 all bind
PY motifs of the Epstein-Barr virus latent protein 2A (LMP2A), leading to enhanced degradation of LMP2A associated kinases Lyn and Syk, inhibiting B-cell signalling in the host and promoting viral survival (reviewed in (Ingham et al. 2004; Staub and Rotin 2006)).
C) Cellular targets of Nedd4-1 and Nedd4-2
In the mammalian genome, two closely related Nedd4 proteins have been identified, Nedd4-1 and Nedd4-2. The Drosophila Nedd4 protein is closely related to both human Nedd4-1 (45% identity) and human Nedd4-2 (46% identity), suggesting a duplication event early in evolution before amphibian and mammalian diversification (Kamynina et al. 2001c). These two Nedd4 family members share 50-60% similarity depending on species and display distinct tissues expression profiles. While Nedd4-1 is ubiquitously expressed, Nedd4-2 expression is more restricted, being high in liver and kidney, and low in heart, brain, and lung (Kamynina et al.
42 Chapter 1
2001a). Other distinguishing features of Nedd4-2 are Sgk (serum glucocorticoid-inducible kinase) phosphorylation sites and the existence of multiple splice forms, with or without a C2 domain and with a variable number of WW domains (Debonneville et al. 2001; Itani et al. 2003).
Based on phenotypes of Nedd4-1 and Nedd4-2 knockout mice, it is clear that these proteins have non-redundant functions, which are discussed below.
i) Nedd4-2
The most well characterized Nedd4-2 substrate is the amiloride-sensitive epithelial sodium channel (ENaC). ENaC is an apically localized channel found in absorptive epithelia of organs involved in fluid and electrolyte homeostasis, such as the kidney, lung and colon (Garty and
Palmer 1997). In the kidney, ENaC is expressed in the distal part of the nephron and plays an important role in regulating sodium homeostasis and blood pressure in response to hormonal signalling, particularly aldosterone and vasopressin (Rossier et al. 2002). ENaC is composed of three homologous subunits (α, β, and γ), each consisting of two transmembrane domains separated by a large extracellular loop and short intracellular N- and C-termini (Canessa et al.
1994a; Canessa et al. 1994b). Recent structural studies of a closely related family member suggest ENaC functions as a heterotrimer (Jasti et al. 2007). Each ENaC subunit has a PY motif at its C-terminus which serve as binding sites for Nedd4-2 WW domains (Staub et al. 1996).
Nedd4-2 binding and subsequent ENaC ubiquitination suppress ENaC activity by decreasing its cell surface stability, primarily through induction of channel internalization, but also through sorting and targeting for lysosomal degradation (Staub et al. 1997; Abriel et al. 1999; Kamynina et al. 2001c; Lu et al. 2007). The physiological importance of ENaC regulation by Nedd4-2 is underlined by the finding that mutations in the PY motifs of β- and γ-ENaC cause Liddle’s
43 Chapter 1 syndrome, a hereditary form of arterial hypertension caused by elevated ENaC activity
(Shimkets et al. 1994; Hansson et al. 1995; Schild et al. 1996; Abriel et al. 1999).
As mentioned, Nedd4-2 also has two sgk phosphorylation sites ((Rx)RxRxxS/T). Sgk is an aldosterone inducible gene that activates ENaC. It has been suggested that sgk1, which possesses a PY motif, binds, and phosphorylates Nedd4-2, reducing its ability to suppress ENaC in Xenopus oocytes (Debonneville et al. 2001; Snyder et al. 2002). 14-3-3 proteins participate in aldosterone-mediated stabilization of ENaC by associating with phosphorylated Nedd4-2, and inhibiting its binding to ENaC PY motifs (Bhalla et al. 2005; Ichimura et al. 2005). Notably,
Nedd4-2 knockout (KO) mice survive normally and have the predicted Mendelian distribution at birth, but exhibit salt-sensitive hypertension (Shi et al. 2008). These mice also show higher expression of all three ENaC subunits in the kidney, impaired downregulation of ENaC function in the colon, and salt-sensitive hypertension that was substantially reduced in the presence of amiloride, a specific inhibitor of ENaC (Shi et al. 2008). While Nedd4-2 has also been implicated in regulation of a number of other ion channels and transporters (see Appendix III), the phenotype of Nedd4-2 knockout mice, characteristic of ENaC overactivity, suggests the major role of Nedd4-2 is downregulation of ENaC.
ii) Nedd4-1
Several Nedd4-1 specific targets have been identified. For example, the guanine-nucleotide exchange factor CNrasGEF is destabilized via Nedd4-1 mediated ubiquitination (Pham and
Rotin 2001). CNrasGEF binds the β1-adrenergic receptor and activates Ras in response to agonist-stimulation of the receptor, implicating Nedd4-1 in the Ras signalling pathway (Pham et al. 2000; Pak et al. 2002). Nedd4-1 is also required for proper trafficking of the lysosomal
44 Chapter 1
transmembrane protein LAPTM5 from the TGN to lysosomes, as discussed earlier. Although
LAPTM5 does undergo Nedd4-1 mediated ubiquitination, its ubiquitination does not appear to
be required for proper sorting ((Pak et al. 2006), Appendix II). Nedd4-1 also plays a role in viral
budding, as ubiquitination of the human T-cell leukemia virus type 1 (HTLV-1) Gag protein by
Nedd4-1 is required for efficient release of HTLV-1 particles (Blot et al. 2004). Recently,
Nedd4-1 was implicated in agonist-dependent ubiquitination and lysosomal degradation of the
β2-adrenergic receptor, with β-arrestin2 functioning as an adaptor to recruit Nedd4-1 to the
activated receptor (Shenoy et al. 2008). Nedd4-1 is also involved in Grb10-dependent downregulation of the insulin-like growth factor-I receptor (IGF-IR). Grb10, an adaptor protein,
binds to the Nedd4-1 C2 domain and IGF-IR, resulting in ligand-dependent ubiquitination of the
receptor and receptor internalization (Vecchione et al. 2003; Monami et al. 2008). Furthermore,
it was recently suggested that stability and nuclear localization of the tumour suppressor PTEN is
regulated by Nedd4-1 mediated ubiquitination (Trotman et al. 2007; Wang et al. 2007).
However, using cells and tissues derived from Nedd4-1 KO mice, our lab has generated data
which failed to support a role for Nedd4-1 as the E3 ligase regulating PTEN stability and
subcellular localization (Fouladkou et al. 2008). Unlike Nedd4-2 KO mice, Nedd4-1 KO mice
die at mid to late gestation and exhibit heart defects (Fouladkou et al. submitted), suggesting that
Nedd4-1 plays an important role in development. It is likely that there are several Nedd4-1
targets crucial for mammalian development, including Grb10, thrombospondin-I and Cbl, as well
as other targets that are yet to be identified.
45 Chapter 1
IV) Project Goals and Rationale
Ubiquitination is a highly versatile modification controlling all types of cellular events, with the
Nedd4 family of E3 ligases playing a critical role in the ubiquitination of a large number of
cellular targets. Despite commonalities exhibited by members of the Nedd4 family, specific
functions and cellular substrates have been ascribed to its members. Factors determining
specificity of interaction between Nedd4-like proteins and their substrates are not well defined.
Furthermore, given the implications of protein ubiquitination, one would expect that catalytic
activity of E3 ligases is tightly regulated, preventing untimely ubiquitination of cellular targets,
or of themselves. However, at the start of this project, little was known about what regulates the catalytic activity of E3 ubiquitin ligases. Therefore, for my PhD thesis, I set out to study the specificity and regulation of the Nedd4-1 and Nedd4-2 proteins. My goals were:
1) To define the specific elements and requirements for interactions between Nedd4-1/-2
WW domains and target PY motifs, using ENaC and Commissureless PY motifs as
models.
2) To investigate possible regulation of Nedd4-2 by interactions between its HECT domain
PY motif and its WW domains.
3) To characterize the novel Nedd4 WW binding protein Rnf11, an E3 ligase potentially
involved in regulating Nedd4-1/-2.
46
CHAPTER 2
Affinity and specificity of Nedd4-1 and -2 WW domains for substrate PY motifs
Portions of the work presented in this chapter were published in Journal of Biological Chemistry and Structure.
Henry P.C., Kanelis V., O’Brien [Bruce] M.C., Kim B., Gautschi I., Forman-Kay J., Schild L., and Rotin D. (2003) Affinity and specificity of interactions between Nedd4 isoforms and the Epithelial Na+ Channel. J. Biol. Chem. 278 (22): 20019-28.
Kanelis V., Bruce M.C., Skrynnikov N.R., Rotin D., and Forman-Kay J.D. (2006) Structural determinants for high-affinity binding in a Nedd4 WW3* domain-Comm PY motif complex. Structure 14: 543-553.
My contribution to this work was making several of the wild-type and mutant constructs used for protein expression, expression and purification of some proteins, as well as carrying out many of the binding experiments along with P. Henry and V. Kanelis. The homology modeling was carried out by V. Kanelis and the oocyte experiments by L. Schild.
47 Chapter 2
I) Summary
WW domains are small, protein-protein interaction domains found in a variety of proteins, and are capable of binding proline-rich sequences. Nedd4-1 and Nedd4-2 are HECT family ubiquitin ligase proteins containing 3-4 WW domains and usually a C2 domain. The
Nedd4 proteins are responsible for recognizing substrates for ubiquitination, occurring when Nedd4 WW domains bind the PY motif (L/PPxY) found in target proteins. Here we examine the molecular nature of PY motif binding by Nedd4 WW domains. Specifically, we have studied the interaction between Nedd4-2 (and Nedd4-1) and the epithelial sodium channel (ENaC), an interaction required to regulate ENaC cell surface stability, thus regulating salt and fluid homeostasis, as well as blood pressure. Our results show that it is the presence of a unique WW domain (WW3*) in Nedd4-2 (and some Nedd4-1 proteins) which establishes a high affinity interaction and the ability to suppress ENaC. In
Drosophila, interaction between dNedd4 and Commisureless (Comm) PY motifs promote axon crossing at the CNS midline and muscle synaptogenesis. Mutagenesis and binding studies of the dNedd4 WW3*-Comm PY motif highlights the importance of residues outside the core PY motif and identify a variable loop in WW3* responsible for its high- affinity interaction. Collectively, these studies expand out general understanding of the molecular determinants involved in WW domain-ligand recognition.
48 Chapter 2
II) Introduction
The process of ubiquitination requires the sequential action of three enzymes; the ubiquitin-
activating enzyme (E1), the ubiquitin-conjugating enzyme (E2) and the ubiquitin protein ligase
(E3). The E3 ligases have the critical role of selecting specific proteins for ubiquitination.
Nedd4 (Nedd4-1) and its close relative Nedd4-2 are E3 ubiquitin ligases comprised of a C2
domain, 3 or 4 WW domains, and an ubiquitin ligase HECT domain. While the C2 domain is
involved in membrane targeting (Plant et al. 1997; Plant et al. 2000) and the HECT domain
provides the catalytic E3 activity (Huibregtse et al. 1995), the WW domains are involved in
substrate recognition and binding.
WW domains are small modules of ~40 residues in length containing two highly
conserved tryptophan residues and an invariant proline, that bind to proline-rich sequences
(Sudol et al. 1995). Nedd4 WW domains have been grouped as Class I WW domains due to
their specificity for binding the PY motif, most recently defined as (L/P)PXY (Kasanov et al.
2001).
The most well characterized Nedd4-2 substrate, the epithelial Na+ channel (ENaC), is an
apically localized sodium channel found in absorptive epithelia of organs involved in fluid and
electrolyte homeostasis such as the kidney, lung, and colon (Garty and Palmer 1997; Rossier et
al. 2002). The three homologous subunits of ENaC (α, β, and γ) each consist of two transmembrane domains separated by a large extracellular loop and short N- and C- termini
(Canessa et al. 1994a; Canessa et al. 1994b). The C-terminus of each subunit contains a PY
motif which serve as binding sites for Nedd4 proteins. Nedd4-2 binding and ubiquitination of
the channel has been shown to suppress ENaC activity by decreasing its cell-surface stability
(Goulet et al. 1998; Abriel et al. 1999; Lu et al. 2007). Regulation of ENaC by Nedd4-2 is
49 Chapter 2
critical since deletion or mutation of the PY motif of β or γENaC cause Liddle syndrome, a
hereditary hypertension characterized by increased numbers and activity of ENaC at the plasma
membrane (Shimkets et al. 1994; Snyder et al. 1995; Firsov et al. 1996).
Solution structure of the third WW domain of rat Nedd4 (homologous to WW4 of human
Nedd4s, hereafter called rNedd4-1 WW4) in complex with the PY motif-containing region of
βENaC (614’PPPNYDSL621’, peptide residues are indicated with a “ ' ” to distinguish from WW domain residues) revealed a three-stranded anti-parallel β-sheet with a hydrophobic binding surface similar to other WW domains. The N-terminal βENaC PY residues (P614’-N617’) form
a polyproline type II (PPII) helix, which bind in the XP groove, a pocket found in all WW and
SH3 domains (Kanelis et al. 2001). Unique to the Nedd4 WW domain-βENaC PY complex is
the helical turn adopted by residues C-terminal to the PPII helix (Y618’-L621’), stabilized by
intrapeptide and peptide-domain interactions, suggesting the involvement of an extended PY
motif sequence (PPxYxxL), which is conserved in all ENaC chains (Kanelis et al. 2001). While
P616’ and Y618’ of the βENaC PY motif are essential for binding to WW4, mutation of P615’
or L618’ (to Ala) results in a 6-fold increase in Kd (i.e. a reduced affinity of binding) (Henry et
al. 2003).
There are several Nedd4 family members, and in particular Nedd4-1 and Nedd4-2 have
been studied with regard to ENaC binding and regulation. Recent work has proposed that
Nedd4-2 can bind to and regulate ENaC activity much better than Nedd4-1 (Harvey et al. 2001;
Kamynina et al. 2001a; Kamynina et al. 2001c), and moreover, Nedd4-2 contains sgk1
phosphorylation sites (missing from Nedd4-1), possibly involved in regulation of ENaC by
aldosterone (Debonneville et al. 2001; Snyder et al. 2002). However, the observation that
human Nedd4-1 (hNedd4-1) is a potent inhibitor of ENaC when lacking its N-terminal region
50 Chapter 2
(Kamynina et al. 2001c; Snyder et al. 2001) suggests that suppression of ENaC by different
Nedd4 family members may be more complicated than originally anticipated and requires further
investigation.
All Nedd4-2 proteins (including those from Xenopus, mouse, and human, i.e.
x/m/hNedd4-2) contain an “extra” WW domain, WW3*, located between WW2 and WW4,
which is lacking from rat and mouse Nedd4-1 but is found in hNedd4-1 and Drosophila Nedd4-1
(dNedd4-1) (Fig. 2-1A). Both WW3* and WW4 appear to play key roles in regulating ENaC
activity (Kamynina et al. 2001a; Kamynina et al. 2001c), and although we and others have
quantified binding between the Nedd4 WW domains and the ENaC PY motifs (Asher et al.
2001; Kanelis et al. 2001; Lott et al. 2002), a comparison between the affinity of the third and
fourth WW domains (from Nedd4-1 and Nedd4-2) for ENaC PY-motifs and their role in ENaC
suppression, have been lacking, and hence has been the first focus of our work.
To further understand the molecular basis for high-affinity binding by WW3*, our group recently examined the structure of WW3* (from dNedd4) in complex with a PY motif ligand
(Kanelis et al. 2006). Commissureless (Comm) is a target for dNedd4, where dNedd4 contains three WW domains (WW1, WW3*, and WW4) (Fig. 2-1A). Comm is a single-pass transmembrane proteins with two PY motifs (218’ESPPCYTIATGLPSYDEALH237’, PY motifs in bold) spaced closely together in its cytoplasmic C-terminus. Binding of dNedd4 WW domains to these PY motifs results in ubiquitination of Comm and subsequent internalization and
sorting of the Comm/Roundabout (Robo) complex, promoting axon crossing at the CNS midline
(Myat et al. 2002). Comm internalization is also important in the muscle, and our lab has
recently demonstrated that binding of dNedd4 to Comm, and its subsequent ubiquitination, is
51 Chapter 2 responsible for removal of Comm from the muscle cell surface, a necessary step for motoneuron innervation (Ing et al. 2007).
The solution structure of dNedd4 WW3* domain bound to the second PY motif in
Comm, 227’TGLPSYDEALH237’ (LPSY peptide), revealed that dNedd4 WW3* is very similar to previously solved WW domain structures, including the rNedd4 WW4 domain in complex with the βENaC mentioned earlier. However, contacts are observed between WW3* and LPSY peptide residues both N- and C-terminal to the PY motif, comprising residues T227’-L236’, a feature not previously reported for a WW domain-PY motif complex (Fig. 2-5) (Kanelis et al.
2006). Here we examine, using binding experiments with mutant peptides, the importance of extensive interactions observed in the dNedd4-Comm complex. Additionally, we demonstrate that mutation of the β1/β2 loop in dNedd4 WW4 to conserved residues present in WW3* confers high-affinity binding to LPxY and PPxY peptides.
There exists numerous Nedd4 family members sharing the same common modular architecture, in fact, the human genome encodes nine Nedd4/Nedd4-like proteins. While this common architecture and core WW domain binding specificity suggest redundant cellular roles for Nedd4 proteins, specific functions have been ascribed to family members based on recent knockout experiments. Taken together, our results provide new insight into the mechanism of sequence-specific recognition, highlighting the importance of interactions outside the core binding region in WW domain-ligand recognition, and sheding new light on how WW domain interactions contribute to the biological specificity of Nedd4 proteins.
52 Figure 2-1. Alignment of Nedd4-1 and Nedd4-2 proteins. (A) Schematic representation of the
alignment of Nedd4-1 and Nedd4-2 proteins from different species, illustrating schematically the
Nedd4 WW1, WW2, WW3*, and WW4 domains (not drawn to scale). All splice variants of
hNedd4-2 are not shown. (B) Sequence alignment of Nedd4 WW domains with amino acids of
the third WW domain of rat Nedd4-1 (WW4) that form key contact points with ENaC βPY highlighted in bold, and amino acids implicated in conferring higher PY motif binding affinity to
WW3* domains are shaded. Alignment was performed using CLUSTAL_X package; asterisks denote identical amino acids, colons denote conserved amino acid substitutions.
53 Figure 2-1
54 Chapter 2
III) Experimental Procedures
Peptides—Peptides representing wild-type or mutant sequences of rat ENaC subunits α, β, and γ,
and two PY motif-containing regions of Comm were synthesized by the Hospital for Sick
Children/Advanced Protein Technology Centre (Toronto, Canada). All peptides were purified by
reverse-phase high performance liquid chromatography using a C18 column with an acetonitrile
gradient. Mass and purity of the peptides were confirmed by electrospray mass spectrometry.
Wild-type peptide sequences from ENaC were: α, MTPPLALTAPPPAYATLG (residues 660–
677); β, TLPIPGTPPPNYDSL (residues 607–621); and γ, GSTVPGTPPPRYNTLR (residues
617–632). Wild-type peptide sequences from Comm were: ESPPCYTIAT (residues 218-227),
TGLPSYDEALH (residues 227-237), and IATGLPSYDEALHHQ (residues 225-239).
Lyophilized peptides were re-suspended in 150 mM KCl, 10 mM K+ phosphate, pH 6.5, or where indicated, with 150 mM NaCl, 10 mM Na+ phosphate, pH 6.5. Peptide concentrations were
measured in 6.0 M guanidine HCl at A280 (Pace et al. 1995).
Expression and Purification of Proteins—The third WW domain of rat Nedd4-1 (renamed WW4)
(accession number AAB48949, residues 451–498), Xenopus Nedd4-2 WW3* (accession number
CAA03915, residues 489–528), human Nedd4-1 WW3* (accession number BAA07655, residues
362–411), and Drosophila Nedd4-1 WW3* (residues 527-571) and WW4 (residues 577-612)
were sub-cloned into pQE-30 and expressed as N-terminal MRGS His6-tagged proteins in
Escherichia coli M15 pREP4 at 37 °C in LB (Sigma). Bacteria were induced with 1 mM
isopropyl-1-thio- -D-galactopyranoside (Promega) at an A600 of 0.6 for an additional 3 h at 37
°C, and cells were harvested by centrifugation at 6000 x g for 10 min. Each poly-His-tagged
protein was purified after lysing cells by sonication and applying the soluble supernatant (cleared
by 10,000 x g centrifugation) over a Ni2+-nitrilotriacetic acid-charged resin column (as described
55 Chapter 2
by the manufacturer, Qiagen). The protein was dialyzed overnight into 10 mM K+ phosphate, pH
6.5, plus 0.1 mM EDTA, 0.15 µg/ml aprotinin, 0.15 µg/ml leupeptin, 0.1 mM benzamidine at
4°C before concentrating and purification on a Superdex 75 gel filtration column (Amersham
Biosciences) in 150 mM KCl, 10 mM K+ phosphate, pH 6.5 (or Na+ replacing K+, where
indicated). Protein concentrations were measured in 6.0 M guanidine HCl at A280 (Pace et al.
1995). rNedd4-1 WW4 His-Thr→Ala-Pro (H470A/T471P) and hNedd4-1 WW3* Ala-Pro→His-
Thr (A381H/P382T) and dNedd4-1 WW4 His-Thr-Asp→Ala-Pro-Asn (H591A/T592P/D539N) substitutions were generated using the QuikChange site-directed mutagenesis kit (Stratagene).
His-470 and Thr-471 in rNedd4-1 WW4 and Ala-381 and Pro-382 in hNedd4-1 WW3* are
equivalent to residues Ala-504 and Pro-505, respectively, in xNedd4-2 WW3* (Fig. 2-1B).
Equilibrium Dissociation Constant (Kd) Measurement—Intrinsic tryptophan fluorescence of each
WW domain was used to monitor peptide binding (Viguera et al. 1994; Kanelis et al. 2001).
Fluorescence measurements were obtained using a Hitachi F-2500 fluorescence
spectrophotometer at 25 °C with excitation and emission wavelengths of 298 and 333 nm,
respectively, and slit widths of 2.5 nm, except for dNedd4 binding experiments, which were
performed on AVIV ATS105 fluorometer equipped with an automatic titrator, at 20°C, with
excitation and emission wavelengths of 298 and 330 nm and slit widths of 2 and 4 nm,
respectively. Experiments were measured in 150 mM KCl, 10 mM K+ phosphate, pH 6.5, or 150 mM NaCl, 10 mM Na+ phosphate, pH 6.5, with WW domain concentrations kept constant at 1
µM. Peptides were added at concentrations ranging from 0 to 2 mM. Kd measurements for n
experiments, where n is indicated in Table 2-I, were log-transformed such that residuals were
normally distributed. A one-way analysis of variance was performed, and pair-wise differences
were examined using the Tukey adjustment to determine adjusted p values (Selvin 2001).
56 Chapter 2
Electrophysiological Measurement of Mutant ENaC Activity—Site-directed mutagenesis was
performed on rat ENaC cDNA as described previously (Schild et al. 1996). Complementary
RNAs of each subunit (WT or mutants) and of xNedd4-2 (WT or CS mutant) or dNedd4-1 (WT
or CA mutant) were synthesized in vitro. Healthy stage V and VI Xenopus oocytes were pressure-
injected with 100 nl of a solution containing equal amounts of α, β, and γENaC complementary
RNA at a total concentration of 100 ng/µl. The oocytes were kept in modified low sodium
Barth’s saline containing 10 mM NaCl, 2 mM KCl, 80 mM n-methyl-D-glucamine chloride, 0.4
mM CaCl2, 0.3 mM CaNO3, 0.8 mM MgSO4, 5 mM n-methyl-D-glucamine-Hepes, pH 7.4.
Standard electrophysiological measurements were taken 16–20h after injection. Macroscopic
amiloride-sensitive Na+ currents, defined as the difference between Na+ currents obtained in the
presence and absence of 5 µM amiloride (Sigma) in the bath, were recorded using the two-
electrode voltage-clamp technique. For current measurements, oocytes were voltage-clamped to -
100 mV. The bath solution was a standard oocyte Ringer solution containing 120 mM NaCl, 2.5
mM KCl, 1.8 mM CaCl2, and 10 mM Hepes. Oocytes were initially placed in a bath solution
containing amiloride (10-6 M) to prevent changes in intracellular Na+ concentration, and current
was measured after washout of amiloride. Currents were recorded with a Dagan TEV-200
amplifier (Minneapolis, MN).
Homology Modeling of Nedd4-WW3* Domain Complexed to βP2 ENaC Peptide—Homology
models of complexes between xNedd4-2 or hNedd4-1 WW3* domains and the βENaC PY motif
were obtained using the program Modeler (Sali and Blundell 1993). Note that the WW3* domain
in Nedd4-2 (xNedd4-2, mNedd4-2, and hNedd4-2) is identical between species and differs by
four residues from that in hNedd4-1. WW domains from Nedd4 and other proteins were aligned
based on previously determined structures of WW domains from rNedd4-1 (PDB code 1I5H)
57 Chapter 2
(Kanelis et al. 2001), hYAP65 (PDB code 1JMQ) (Pires et al. 2001), h-dystrophin (PDB code
1EG4) (Huang et al. 2000), hPin1 (PDB code 1F8A) (Verdecia et al. 2000), mFBP28 (PDB code
1E0I), and yeast YJQ8 (PDB code 1E0N) (Macias et al. 2000). A structural alignment of ligands
within Modeler was performed for PY motif-containing peptides bound to rNedd4-1, hYAP65,
and h-dystrophin WW domains. The Nedd4, dystrophin and YAP65 WW domain complexes, the
Pin1 WW domain in the context of the complex, and the free FBP28 and YJQ8 WW domains
were used as templates. A total of 100 models were generated from which the 15 lowest energy
models were selected for analysis.
IV) Results
A) Contribution of WW3* domain to the Nedd4-ENaC interaction
i) Comparison of Binding Affinity of WW3* and WW4 toward the PY motif of βENaC
The differential ability of Nedd4-1 and -2 to function as ENaC suppressors, with Nedd4-2
proposed to bind better to ENaC and to more potently suppress channel activity than Nedd4-1
(Kamynina et al. 2001a; Kamynina et al. 2001c), is an observation that was not fully explained.
An obvious explanation for this effect is the presence of an extra WW domain, WW3*, in
Nedd4-2 (such as xNedd4-2, mNedd4-2, and hNedd4-2) but not in most Nedd4-1 proteins (e.g. rNedd4-1 and mNedd4-1) (Fig. 2-1A). To assess the role of this divergent WW3* domain, we
measured the binding affinity of WW3* from x/m/hNedd4-2 toward the three ENaC PY motifs
and compared it to that of the WW4 domain of rNedd4-1, found in both Nedd4-1 and -2. We
chose to compare these two domains because previous work has demonstrated that WW3* and
WW4 of Nedd4-2 play the most pivotal roles in regulating ENaC activity (Kamynina et al.
2001a; Kamynina et al. 2001c). The Kd was measured by monitoring the intrinsic tryptophan
58 Figure 2-2. Representative curves of fluorescence emission from rNedd4-1 WW4,
x/m/hNedd4-2 WW3*, or hNedd4-1 WW3* binding to αPY, βPY, or γPY peptides of
ENaC. Fluorescent emission spectra measured at 333 nm from purified His6-Nedd4 WW3* or
WW4 domains excited at 298 nm during the titration of ENaC αPY, βPY, or γPY peptides.
Dissociation constants were calculated as described under "Experimental Procedures" and summarized in Table 2-I.
59 Figure 2-2
60 Table 2-I
Dissociation constants (Kd) of rNedd4-1 WW4, x/m/h Nedd4-2 WW3*, or hNedd4-1 WW3* domains binding to ENaC α, β, and γ-PY motif peptides αPY βPY γPY Protein (MTPPLALTAPPPAYATLG) (TLPIPGTPPPNYDSL) (GSTTVPGTPPPRYNTLR) rNedd4-1 WW4 38 + 3 (4) 44 + 3 (4) 141 + 33 (5) x/m/hNedd4-2 WW3* 12 + 3 (5) 9 + 1 (8) 24 + 3 (5) hNedd4-1 WW3* 10 + 2 (4) 14 + 1 (4) 26 + 4 (4) rNedd4-1 WW4 (HT→AP) ND 65 + 6 (4) ND hNedd4-1 WW3* (AP→HT) ND 38 + 3 (6) ND
Calculated Kd values (µM) from (n ) experiments are expressed as the means + S.E. The mutants used were rNedd4-1 WW4 (HT→AP = H470A, T471P) and hNedd4-1 WW3* (AP→HT = A504H, P505T). For these mutants (bottom 2 rows), values represent the means + S.E. of untransformed data. Residue numbering for the mutants is described under "Experimental Procedures." All measurements were performed in 10 mM K+ phosphate, pH 6.5, 150 mM KCl. ND, not determined.
61 Chapter 2
fluorescence, which is highly sensitive due to the location of the second conserved Trp found in
the binding pocket of the WW domain. The emission spectra of this Trp residue, excited at 298
nm, changes as it shifts from an aqueous, unbound environment, to a hydrophobic, bound
environment.
Our results show that WW3* binds the PY motif of each ENaC subunit with a 3–6-fold
higher affinity than WW4 (Fig. 2-2 and Table 2-I). The human Nedd4-1, however, also contains
a WW3* domain. We thus compared its binding affinity to that of rNedd4-1 WW4. As seen in
Fig. 2-2 and Table 2-I, the WW3* of hNedd4-1 shows the same higher affinity interactions with
the PY motifs of ENaC as that seen with the WW3* of Nedd4-2. Moreover, the WW3* of
dNedd4-1 (see below) also exhibited this high affinity binding to α and βENaC-PY motifs.
These data suggest that it is the presence of WW3* that imparts a high affinity interaction
between Nedd4 and ENaC.
ii) Suppression of ENaC Activity by the WW3*-containing dNedd4-1
To test whether the presence of WW3* in a Nedd4-1 protein could impart on it the ability to
suppress ENaC, we took advantage of our observation that in dNedd4-1, which most closely
resembles rNedd4-1 (its likely orthologue), the WW2 is naturally replaced with WW3* (Fig. 2-
1A). DNedd4 was chosen here as a model Nedd4-1 (full-length) protein, although its natural
substrate in flies is unlikely to be ENaC since fly ENaC homologues (Pickpocket, Ripped pocket)
do not have PY motifs (Adams et al. 1998). We thus injected complementary RNA of dNedd4-1
into Xenopus oocytes together with αβγENaC. As seen in Fig. 2-3, dNedd4-1 was a strong suppressor of ENaC ( 90% suppression), similar to xNedd4-2. This suppression did not occur
when a catalytically inactive mutant dNedd4-1, dNedd4-1(CA), was used (Fig. 2-3). Moreover,
dNedd4-1, like xNedd4-2, was unable to inhibit activity of ENaC lacking all three PY motifs,
62 Figure 2-3. Suppression of ENaC by dNedd4-1. Wild-type ENaC (ENaC wt)- or ENaC-
bearing mutations in the PY motif of the α, β, and γ subunits that abolish binding to the Nedd4
WW domains (ENaC ∆PY) were co-expressed in Xenopus oocytes with H2O (controls), dNedd4-
1 (hatched bars) or xNedd4-2 (filled black bars) and their respective catalytically inactive forms
(xNedd4-2(CS) or dNedd4-1(CA)). ENaC activity was measured as the amiloride-sensitive INa.
Bars represents the mean of 6–32 oocytes from different frogs, and the asterisks denote significance at p < 0.05 relative to either one of the controls (open bars).
63 Figure 2-3
64 Chapter 2
which serve as binding sites for Nedd4-WW domains (Fig. 2-3). Thus, dNedd4-1, possessing a
WW3*, can effectively suppress ENaC activity, similar to Nedd4-2 proteins, and in sharp
contrast to its orthologue, rNedd4-1, which lacks WW3*.
iii) Homology Modeling and Comparison of WW3* and WW4 Complexes
To probe the molecular basis of the high-affinity interaction observed between WW3* and ENaC
PY motifs, we have generated homology models of the x/m/hNedd4-2 and hNedd4-1 WW3* domains in complex with the βPY peptide of ENaC using the program Modeler (Sali and
Blundell 1993). As expected from the high degree of sequence identity (88.6%) between the xNedd4-2 and hNedd4-1 WW3* domains, these models are very similar, and hence, only the
xNedd4-2 WW3*-βENaC model will be referred to here. The xNedd4-2 WW3*-βENaC models
were examined with PROCHECK (Laskowski et al. 1996), and were shown to have good
geometry and sampled favourable regions of Ramachandran space with 91, 7, and 1% of residues
in the favourable, allowed, and disallowed regions, respectively. These models are well defined
with a pairwise root mean square deviation of 0.20 ± 0.04 Å and 0.72 ± 0.10 Å for backbone and
all heavy atoms, respectively. This degree of precision in the models indicates that the same
conclusions are obtained for all members of the ensemble. The generated models are very similar
to the previously determined solution structure of the rNedd4-1 WW4 domain-βPY peptide
complex (Kanelis et al. 2001), with backbone root mean square deviation values of 0.33 ± 0.03
Å. This is expected given the sequence identity between rNedd4-1 WW4 and xNedd4-2 WW3*
domains of 65.7% for all residues and 77.8% for residues involved in ligand binding (Fig. 2-1B).
Furthermore, because of the high degree of sequence conservation between these domains, there is a high degree of confidence in the model of the x/m/hNedd4-2 WW3* domain-ENaC βPY
peptide complex, supporting the validity of conclusions obtained from analysis of this model.
65 Chapter 2
Examination of the peptide binding site reveals an amino acid substitution (Fig. 2-4); H470 in
rNedd4-1 WW4 (which forms the back of the XP groove and makes extensive contacts to the
bound peptide through its aromatic ring) is an Ala (A504) in WW3*. The result is a larger
pocket to accommodate P616’ and P615’. In addition, the presence of a Pro (P505) residue in
WW3* instead of a Thr (T471), as in WW4, provides additional interactions with P616’. The
side chain of P505 is oriented toward the binding interface, allowing for interactions with the
βPY peptide (Fig. 2-4A,B), in contrast to T471 in the rNedd4 WW4-βENaC PY complex. The
combination of a larger XP groove to accommodate the ligand and additional contacts conferred
by the Pro residue likely contributes to the greater binding affinity of WW3* for ENaC PY
motifs.
iv) Mutation Analysis to Test the Role of WW3* Ala-504/Pro-505 in Conferring High Affinity
Binding to PY Motifs
To experimentally test the role of A504/P505 in conferring high-affinity binding to WW3* we
mutated A504/P505 in WW3* to H504/T505 to mimic the equivalent residues in WW4 (and
numerous other WW domains), which do not bind as tightly to the ENaC PY motifs. Table 2-I shows that such substitutions indeed reduced binding affinity to the βENaC PY motif 2–3-fold, in
support of the model (Fig. 2-4). However, the reciprocal substitutions in WW4 (H470A/T471P)
did not lead to gain of high affinity binding toward the βENaC PY motif (Table 2-I). Taken
together, these results suggest that the A504/P505 in WW3* are necessary but not sufficient to
impart high affinity interactions of this domain toward the PY motifs of ENaC and that another
region(s) in the domain is likely involved as well.
66 Figure 2-4. Homology model of Nedd4-2 WW3*·βENaC PY motif complex. (A) Ribbon diagrams of the solution structure of the rNedd4-1 WW4·βENaC and homology model x/m/hNedd4-2 WW3*·βENaC complexes. The backbone of each WW domain is coloured blue for β-strands and gray for loops and termini. Side chains involved in peptide binding are shown in cyan. Peptide backbones are in orange, and side chains of Pro-615', Pro-616', Tyr-618', and
Leu-621' are shown in yellow. Lines from Pro-505 in xNedd4-2 WW3* indicate contacts to the
ENaC peptide. (B) Surface representations of the solution structure of rNedd4-1 WW4·βENaC and homology model of x/m/hNedd4-2 WW3*·βENaC complexes. Molecular surfaces of WW domains are shown colored with red and blue for negative and positive electrostatic potential, respectively. Ligands (PY peptides) are shown as ribbons with the backbone as a thick yellow tube and side chains as thin yellow lines. Numbering of WW3* residues refers to the xNedd4-2 sequence.
67 Figure 2-4
68 Chapter 2
B) Identification of residues involved in high affinity interaction between Drosophila Nedd4
WW3* and the Commissureless PY-motif
To further examine the molecular basis for the high-affinity binding observed for WW3* we
turned our attention to our lab’s recently solved structure of dNedd4 WW3* in complex with the
second Comm PY-motif (227’TGLPSYDEALH237’) (Kanelis et al. 2006). This was chosen based
on preliminary experiments which indicated that the affinity of dNedd4 WW3* for Comm PY-
motifs was high in comparison with other WW domain PY-motif complexes. The dNedd4
WW3* domain structure is very similar to previously solved WW domain structures, forming a
three-stranded anti-parallel β-sheet, stabilized by a hydrophobic cluster formed in part by the first
conserved Trp (W535) and the invariant Pro (P560) (Fig. 2-5 and 2-6A). The LPSY sequence of
the ligand adopts the same conformation as seen for PPxY ligands bound to WW domains, a
PPII helical conformation, interacting with corresponding WW domain residues (Huang et al.
2000; Kanelis et al. 2001; Pires et al. 2001).
Surprisingly, residues both N- and C-terminal to the PY motif interact with the WW3*
domain. As was observed in the rNedd4 WW4 domain-βENaC complex (Kanelis et al. 2001),
C-terminal residues adopt a helical turn conformation starting at the PY motif Tyr, however here this turn is extended by one residue comprising Y232’-L236’, where A235’ in Comm is structurally analogous to L621’ in βENaC (Fig. 2-6). Interactions are also seen between the
β1/β2 loop in WW3*, most notably residues A540/P541/N542, and residues N-terminal to the
LPSY sequence, while analogous interactions were not observed in the rNedd4 WW4 domain-
βENaC complex (Fig. 2-6). Markedly, our homology model of the x/m/hNedd4-2 WW3*-
βENaC PY-motif complex (Fig. 2-6) and mutational analysis (Table 2-I) identified analogous residues, A504/P505, as necessary for the high affinity binding observed with WW3*.
69 LPSY Peptide Complex. Schematic- כFigure 2-5. Solution Structure of the dNedd4 WW3
ribbon diagram of the lowest energy structure. The backbone of the WW domain is in blue. The
side chains of the canonical WW domain Trp and Pro residues are shown in pink. Residues involved in packing the hydrophobic cluster and in peptide binding are shown in cyan. The
LPSY peptide backbone is in red and selected side chains are shown in yellow, with the exception of Gly 228′ which is shown as a gray sphere for the Cα atom. Residue numbers and the C terminus for the LPSY peptide are denoted with a prime symbol (′).
70 Figure 2-5
71 Domain-LPSY Peptide and rNedd4 WW4 כFigure 2-6. Comparison of the dNedd4 WW3
(domain and (B כDomain-βENaC Complexes. Molecular surfaces of the (A) dNedd4 WW3 rNedd4 WW4 domain (PDB code 1I5H) are shown with blue and red, representing positive and negative electrostatic potential, respectively. The Cα traces of the bound ligands are shown in green and the side chains of residues involved in binding in yellow. The dNedd4 WW3* domain residues of the β1/β2 loop important for high-affinity binding are circled.
72 Figure 2-6
73 Chapter 2 i) Contributions of LPSY Peptide Residues to Binding Affinity
As mentioned, the affinity of the dNedd4 WW3* domain for the Comm LPSY peptide is higher than observed for other WW domain-PY motif complexes, having a Kd value of ~3µM (Table 2-
II). Here we used a mutagenesis approach to further investigate the contribution of LPSY peptide residues to the affinity of binding to dNedd4 WW3*.
While the canonical PY motif has a Pro in the first position, the Comm ligand contains a
Leu in place of Pro. Interestingly, mutation of L229’ to Pro, thereby creating a canonical PY motif, results in a ~2-fold reduction in binding affinity (Table 2-II). Although having a Leu instead of a Pro is expected to decrease the PPII helix conformation of the unbound LPSY peptide, and hence its affinity for the WW domain, the greater enthalpic contributions from the larger Leu residue compensate for the loss of PPII helical conformation. Studies of SH3 domain binding indicate that other hydrophobic residues, such as Leu and Val, are tolerated at this position, and in some cases preferred, since they can achieve as good, or better, packing than a
Pro (Lim et al. 1994; Yu et al. 1994).
The contribution of C-terminal peptide residues D233’ and E234’ were not tested, since these residues do not interact with the WW3* domain and are not expected to alter binding.
Mutation of H237’ had little effect on binding affinity, in agreement with the solution structure revealing that this residue has little interaction with the WW3* domain. However, mutation of
L236’ resulted in a 2-3 fold decrease in binding (Table 2-II). This modest decrease is likely explained by similar contacts made by the Ala substitution, compensating for removal of Leu.
Mutation of N-terminal T227’, whose backbone is nestled in a groove that extends from the XP pocket, results in a 4-fold decrease in binding affinity. This was expected since it is unlikely that conformational changes occur in the peptide to promote analogous interactions with
74 Chapter 2
the shorter Ala side chain. Not surprisingly, mutation of invariant PY motif residues P230’ and
Y232’ completely disrupt binding to WW3*. Binding experiments with a peptide containing two additional residues on each of the N- and C-termini, 225’IATGLPSYDEALHHQ239’, gave identical Kd values, for binding to WW3*, as observed with the shorter peptide, indicating that
key interactions were not missed by truncating the peptide (Table 2-II). Removal of all N- and
C-terminal residues, or their replacement by Ala, decreases binding by ~50- to 100-fold,
illustrating the importance of specific residues outside the PY motif core to the high-affinity
dNedd4 WW3* domain-Comm LPSY interaction (Table 2-II). Finally, as mentioned, Comm has
2 PY motifs spaced closely together. Surprisingly, binding was not observed between dNedd4
WW3* and a peptide derived from the first Comm PY motif, 218’ESPPCYTIAT227’, even though
it has a canonical PY motif (Table 2-II).
ii) Contributions of dNedd4 WW3* Domain Residues to Binding Affinity
To further probe interactions responsible for high-affinity binding, mutations were made in the
dNedd4 WW3* domain, and binding to the LPSY peptide measured. While previous studies
have examined the role of WW domain residues involved in binding core PY motif residues
(Chen et al. 1997; Espanel and Sudol 1999; Toepert et al. 2003; Kato et al. 2004), little has been
done to investigate contributions of WW domain residues that bind sequences outside the core
PY motif. Here we focused on the contribution of the β1/β2 loop, due to its importance in
binding to the LPSY peptide.
The dNedd4 WW3* domain complex structure illustrated the involvement of β1/β2 loop
residues, A540, P541 and N542, in binding the LPSY peptide. While corresponding residues in
the rNedd4 WW4 domain, conserved in dNedd4 WW4, are not involved in βENaC binding, our
homology model of the Nedd4-2 WW3*-βENaC PY motif complex (Fig. 2-4) predicted the
75 Table 2-II Dissociation Constants for dNedd4 WW Domain-Comm Peptide Interactions H591A/T592P/D593N Peptide Sequencea WW3* WW4 WW4 227'TGLPSYDEALH236' 3.1 + 0.5 (3) NB 2.9 + 0.8 (2) AGLPSYDEALH 12.0 + 3.3 (4) TGPPSYDEALH 6.7 + 0.9 (4) 22.2 + 2.3 (2) 5.9 + 1.0 (3) TGLPSYDEAAH 8.0 + 3.5 (3) TGLPSYDEALA 4.1 + 0.4 (3) TGAPSYDEALH NB TGLPSADEALH NB 225'IATGLPSYDEALHHQ238' 3.2 + 0.7 (3) 218'ESPPCYTIAT227' NB
Ac-LPSY-NH2 158.8 + 34.8 (3) AALPSYAAAAA 210.5 + 48.2 (4)
The equilibrium dissociation constants (Kd, µM) were determined using intrinsic tryptophan fluorescence and are reported as averages + standard deviations. The number in parentheses indicates number of experiments performed for each binding analysis. NB, no observable binding. aUnderlined residues denote mutation.
76 Chapter 2
involvement of analogous Nedd4-2 WW3* residues A504/P505 in contributing to high-affinity
βENaC PY motif binding. This result, along with conservation of the sequence APN in the
β1/β2 loop of all high-affinity binding Nedd4 WW3* domains (Fig. 2-1B), suggested APN may
be involved in conferring high-affinity interaction with ligand.
To test this hypothesis, we created a chimera between the dNedd4 WW3* and WW4
domains by replacing the sequence HTD in WW4 with APN. The Kd for the interaction between
the WW4 domain mutant and the LPSY peptide is ~3 µM, identical to what is observed for
binding to the WW3* domain (Table 2-II). A gain of function is also observed for mutant WW4
binding to peptide in which L229’ is mutated to Pro, producing a canonical PY motif, with a Kd value of 6 µM – a ~3- to 4-fold increase in affinity over the wild-type WW4 domain and similar to the WW3* domain (Table 2-II). Earlier, we observed that mutation of A540 and P541 residues in WW3* to corresponding His and Thr in WW4 decreased binding by ~4-fold, but that the reverse mutations did not impart high-affinity binding to the Nedd4 WW4 domain (Table 2-
I). Taken together, these results demonstrate that all three residues of the β1/β2 loop, A540,
P541, and N542, are necessary and sufficient to confer higher affinity binding to the PY-motif.
Surprisingly, the dNedd4 WW4 domain binds very weakly to the wild-type LPSY peptide, although it does bind the L229’P mutant peptide, with a Kd value of ~22 µM, similar to what is observed for the rNedd4 WW4 domain binding to βENaC (Kanelis et al. 2001).
Preference for the smaller Pro instead of the Leu in the variable binding position of the XP groove is likely due to the bulky His in the β1/β2 loop of WW4.
77 Chapter 2
V) Discussion
The binding studies carried out here emphasize the importance of residues outside the core
L/PPxY binding motif for affinity and specificity in WW domain interactions. This ability to
preferentially bind distinct PY motif ligands with differing affinities is one mechanism that
underlies different Nedd4 cellular specificities.
A) Importance of the WW3* domain in Nedd4-ENaC interactions
Recent work suggesting that Nedd4-2 (e.g. xNedd4-2, mNedd4-2), but not Nedd4-1 (e.g.
rNedd4-1, mNedd4-1), can bind to and regulate ENaC activity (Abriel et al. 1999; Kamynina et al. 2001a; Kamynina et al. 2001c) is explained by our observation here of a high affinity interaction between the WW3* domain of Nedd4-2 and ENaC PY motifs. We believe that the presence of WW3* determines the ability to interact with and downregulate ENaC. Indeed,
Nedd4-1 proteins possessing this WW3* domain (i.e. hNedd4-1 lacking its C2 domain
(Kamynina et al. 2001c; Snyder et al. 2001) or dNedd4 (Fig. 2-3)) are equally potent suppressors of ENaC as Nedd4-2 proteins. It is unlikely that WW3* alone is sufficient for ENaC regulation
(Snyder et al. 2001), however it is clear that the presence of this high affinity binder alongside
WW4 (which binds ENaC with moderate affinity) is sufficient to suppress ENaC function, at least in Xenopus oocytes. The observation that hNedd4-1 lacking its C2 domain, but not full- length hNedd4-1, suppresses ENaC suggests that selectivity of Nedd4 proteins for ENaC may be partially dictated by N-terminal regions of the protein. This also suggests that Nedd4-2, rather than Nedd4-1, is the natural ENaC E3 ligase. Obviously, other regions of Nedd4-2 are likely to be important for regulation of substrates, including ENaC, in native tissues/cells.
78 Chapter 2
B) Role of the WW3* residues in high affinity binding with substrate PY motifs
Through comparison of the solution structure of the rNedd4-1 WW4-βENaC PY motif complex with our modeled structure of the WW3* domain complex, we were able to identify two residues, A504 and P505, in WW3* facilitating a more stable interaction with the PY motif.
However, our binding data revealed that while these residues are necessary to achieve high affinity interactions with ENaC PY motifs, they are not sufficient since substituting Ala and Pro at the equivalent position in WW4 does not impart high affinity binding to this domain. Upon examination of the structure of WW3* from dNedd4-1 in complex with a PY motif from
Commissureless, we identified corresponding residues in dNedd4-1 WW3*, A540 and P541, as playing key roles in conferring high affinity binding to this domain. Moreover, we identified
N542 as playing a role as well. These residues, located in the β1/β2 loop of WW3*, are conserved in all WW3* domains and are involved in binding N-terminal PY motif residues.
Comparison of rNedd4 WW4-βENaC PY motif and dNedd4 WW3*-Comm PY motif solution structures reveals why the APN sequence found in WW3* allows a higher affinity interaction with selected PY motifs. In the case of the Comm LPSY peptide, the small Ala in
WW3* creates a groove that accommodates the large L229’ side chain, while the corresponding residue in WW4, His, sterically inhibits binding of L229’, explaining why binding was not observed between the LPSY peptide and WW4. Furthermore, P541 and N542 in the β1/β2 loop of WW3* provide favourable groups for binding the hydrophobic LPSY peptide compared with a Thr and a negatively charged Asp, respectively, in WW4. Additionally, due to N-alkyl substitution of proline residues, P541 in WW3* likely restricts conformation of the β1/β2 loop in the free WW3* domain, resulting a in a smaller entropic cost upon binding, which ultimately leads to a higher affinity interaction. Similarly, in the case of the βENaC PY motif peptide, the
79 Chapter 2
smaller Ala in WW3* creates a larger XP groove that may better accommodate the PY motif Pro
residues and other aliphatic residues found in the X position. However, our binding data
suggests the WW3* domain preferentially binds PY motifs with a Leu in the first position, likely
since the larger Leu can achieve better packing than Pro in the larger XP groove.
A comparison of the two complexes also reveals a displacement of the backbone of the
C-terminal helical turn of the bound ligand, so that the methyl groups of A235’ in the LPSY
peptide and L621’ in βENaC are in identical positions and make conserved interactions with
their respective WW domain. This flexibility suggests that other methyl-containing residues
could be accommodated C-terminal to the PY motif, raising the possibility that this is a
conserved interaction in Nedd4 WW domain-ligand complexes, particularly given the sequence
of Nedd4 ligands. For example, the sequence PPxYxxL is conserved in all ENaC subunits
across all species. It is likely that the sequence PPxYxxΦ (where Φ is a hydrophobic residue) is found in other Nedd4 targets.
C) Biological significance of Nedd4 WW domain specificity
Although dNedd4 WW3* and WW4 domains have ~52% sequence identity, and 5 of 8 binding
site residues are identical, they have very different binding properties. Our binding data
demonstrates that the WW3* domain has a much higher binding affinity for the Comm LPSY
sequence than WW4. Surprisingly, dNedd4 WW3* does not interact with the ESPPCYTIAT
peptide, even though it contains a canonical PY motif. However, preliminary binding
experiments indicate that dNedd4 WW4 does interact with this peptide, albeit weakly (data not
shown). These results suggest a discriminatory aspect to Comm PY motif binding by the
dNedd4 WW domains. WW1 is very similar in sequence to other Nedd4 WW1 domains, which
we have shown has very low affinity for PY motifs (Kanelis et al. 2001; Henry et al. 2003).
80 Chapter 2
Therefore, it is likely that only WW3* binds the LPSY ligand in vivo, whereas WW4 may bind
the PPCY ligand. However, since both PY motifs are very close to each other, steric restraints
between WW domain complexes would likely prevent simultaneous binding of Comm ligands
by the dNedd4 WW3* and WW4 domains.
Comm, in complex with Robo, is involved in regulating axon crossing at the CNS
midline during Drosophila development (Kidd et al. 1998). This process, requiring sequestration
of Robo away from the axon surface, is carried out by dNedd4 mediated ubiquitination of the
Comm/Robo complex, resulting in endocytosis (Myat et al. 2002) and/or sorting of the complex
to endosomes (Keleman et al. 2002). In addition, dNedd4 is involved in muscle synaptogenesis
through removal of Comm from muscle surface prior to motoneuron innervation (Ing et al.
2007). Since high-affinity interaction with the Comm LPSY motif required the presence of APN
sequences within the WW3* domain, and as the WW3* domain is only present in dNedd4 and
not other Nedd4 family members in flies, dNedd4 is required for regulating Comm.
ENaC plays a key role in regulating salt and fluid absorption in several tissues and in the
regulation of blood pressure. The importance of Nedd4-2 mediated ENaC regulation is
underscored by the deleterious effect caused by mutation of ENaC PY motifs in Liddle syndrome. The presence of the high-affinity WW3* domain in mammalian Nedd4-2 (and some
Nedd4-1) proteins provides an explanation for the observation that only these WW3* containing proteins can effectively suppress ENaC activity by regulating cell surface stability of the channel
(Lu et al. 2007). Finally, as mentioned, the presence of WW3* alone in Nedd4-2 (or Nedd4-1) is
not sufficient to suppress ENaC, a process requiring at a minimum WW3*, WW4 and HECT
(Snyder et al. 2001). The presence of a C-terminal PY motif in each ENaC subunit, capable of
81 Chapter 2 binding WW3* and WW4, suggests simultaneous binding of WW3* and WW4 to separate
ENaC subunits may be involved in Nedd4-2 mediated ENaC suppression.
In summary, the work presented here illustrates the molecular basis for high-affinity binding of Nedd4 WW3* domains to their cognate ligands and expands our understanding of the specificity of Nedd4 proteins.
82
CHAPTER 3
Regulation of Nedd4-2 self-ubiquitination and stability by a PY- motif located within its HECT-domain
The work presented in this chapter was published in Biochemical Journal:
Bruce M.C., Kanelis V., Fouladkou F., Debonneville A., Staub O., and Rotin D. (2008) Regulation of Nedd4-2 self-ubiquitination and stability by a PY motif located within its HECT- domain. Biochem. J. 415(1):155-164.
I completed the design, execution and analysis of the majority of the experiments presented in this chapter. V. Kanelis carried out the homology modeling and F. Fouladkou performed the electrophysiological experiments in the lab of O. Staub.
83 Chapter 3
I) Summary
Ubiquitin ligases play a pivotal role in substrate recognition and ubiquitin transfer, yet
little is known about the regulation of their catalytic activity. Nedd4-2 is an E3 ubiquitin-
ligase comprised of a C2-domain, 4 WW domains that bind PY-motifs (L/PPxY) and a ubiquitin-ligase HECT domain. Here we show that the WW domains of Nedd4-2 bind
(weakly) to a PY-motif (LPxY) located within its own HECT domain and inhibit auto- ubiquitination. Pulse-chase experiments demonstrate that mutation of the HECT-PY motif decreases stability of Nedd4-2, suggesting that it is involved in stabilization of this E3 ligase. Interestingly, the HECT-PY motif mutation does not affect ubiquitination or downregulation of a known Nedd4-2 substrate, ENaC. ENaC ubiquitination, in turn, appears to promote Nedd4-2 self-ubiquitination. These results support a model whereby the inter/intramolecular WW-domain : HECT-PY motif interaction stabilizes Nedd4-2 by preventing self-ubiquitination. Substrate binding disrupts this interaction, allowing
Nedd4-2 self-ubiquitination and subsequent degradation, serving to downregulate Nedd4-2 once it has ubiquitinated a target. These findings also point to a novel mechanism for ubiquitin ligases to differentially regulate self versus substrate ubiquitination and stability.
84 Chapter 3
II) Introduction
Ubiquitination plays a critical role in many cellular processes, primarily regulating protein stability and trafficking/endocytosis (Glickman and Ciechanover 2002; Hicke et al. 2005; Staub and Rotin 2006). The ubiquitination cascade is carried through the sequential action of E1, E2 and E3 enzymes. E3 ubiquitin ligases, belonging primarily to the RING or HECT families, impart selectivity on the ubiquitination process by recognizing target substrates and promoting transfer of activated ubiquitin from the E2 (RING family) or themselves (HECT family) to substrate lysine residues (Glickman and Ciechanover 2002).
The Nedd4 family of HECT E3 ligases (including Nedd4-1 and -2) contain a C2 domain,
3-4 WW domains that interact with PY motifs (L/PPxY) of target proteins, and a C-terminal ubiquitin ligase HECT domain (Staub and Rotin 2006). The best characterized Nedd4-2 substrate is the epithelial Na+ channel (ENaC). Nedd4-2 WW domains bind to ENaC-PY motifs, regulating cell surface stability of the channel by targeting it for endocytosis and lysosomal degradation (Staub et al. 1996; Staub et al. 1997; Kamynina et al. 2001a; Snyder et al. 2001).
Nedd4-2-mediated downregulation of the channel is critical, since mutations in the PY motif of
ENaC that disrupt Nedd4-2 binding cause Liddle’s syndrome, a hereditary hypertension caused by elevated ENaC stability/activity (Lifton et al. 2001).
Despite their importance, little is known about the regulation of E3 ligase catalytic activity. Indeed, Nedd4-2 was previously thought to be constitutively active. Here we show that
Nedd4-2 regulates its own stability through self-ubiquitination, which is inhibited by interaction between Nedd4 WW domains and a PY motif located within its own HECT domain.
85 Chapter 3
III) Experimental Procedures
Plasmids - Human (h) Nedd4-2 (isoform lacking the C2 domain) was shuttled into the V5
expression vector (pCDNA3.1-nV5), and hNedd4-2 (isoform with C2 domain) was shuttled into
GST vector for bacterial expression (pDEST15 with engineered PreScission Protease site), using
Gateway Technology (Invitrogen). The bacterial expression plasmids pGEX-6P1-HECT, pQE-
30-WW1, pQE-30-WW2, pQE-30-WW3, and pQE-30-WW4 were constructed by standard PCR
and restriction enzyme cloning procedures using Xenopus (x) Nedd4-2 as a template, and pGEX-
KG-βPY using rat βENaC as a template. Construction of pSDeasy plasmids encoding ENaC
subunits and xNedd4-2 was described previously (Abriel et al. 1999). All Nedd4-2 and HECT
mutations were made using the QuikChange mutagenesis kit (Stratagene).
Transfections and Antibodies - Transient transfections in HEK293T cells were performed using
the calcium phosphate precipitation method, with a total of 20 µg DNA per 10-cm dish. The
antibodies used for immunoblotting were anti-RGS-His (Qiagen), anti-ubiquitin (Covance), anti-
Flag (Sigma), anti-HA (Covance), anti-myc (Chemicon), anti-V5 (Invitrogen), and anti-Nedd4-
HECT antibodies (Pak et al. 2006).
Far Westerns - GST fusion proteins were produced in bacteria and purified on glutathione-
agarose beads (Sigma). MRGS-His6 tagged WW domains were expressed and purified as previously described (Henry et al. 2003). GST fusions (50 µg) were separated by SDS-PAGE, transferred to nitrocellulose (GE Health), incubated with 50 µg of WW domain, and analyzed by
immunoblotting with anti-His antibodies.
86 Chapter 3
Ubiquitination assays - For in vitro ubiquitination assays, bacterially-expressed xNedd4-2-HECT or full-length hNedd4-2 (~1 μg) domain cleaved of its GST tag (using PreScission Protease, GE
Health), or hNedd4-2 immunopurified from transfected HEK293T cells, were incubated in
reactions containing 250 ng yeast E1 (Boston Biochem), 250 ng E2 (UbcH7), 2 μg ubiquitin
(Sigma) and 4 mM ATP in reaction buffer (25 mM Tris (pH 7.5), 50 mM NaCl, 0.1 μM DTT, 4 mM MgCl2). Reactions were incubated for 1 h at room temperature (unless otherwise noted) and analyzed by immunoblotting with anti-ubiquitin antibodies. For hNedd4-2 and ENaC ubiquitination in HEK293T cells, cells were treated with 20 μM MG101 (Boston Biochem) for
12 hours before lysis, or with 10 μM clasto-lactacystin-β-lactone (Boston Biochem) for 3 hours before lysis. Cells were lysed (48-hours post-transfection) in lysis buffer supplemented with 50
μM LLnL and 0.4 mM chloroquine (Sigma). Cell lysates were denatured with 2% SDS and boiled for 5 min, then diluted 11 times with lysis buffer to dilute the SDS prior to immunoprecipitation.
Pulse-chase Experiments - Cells were washed (3X) with Met/Cys-deficient medium (Invitrogen) within 30 min and then incubated in the same media supplemented with 0.1 mCi/mL
[35S]Met/Cys (Promix, GE Health) for 2 hrs. Cells were washed (3X) with chase medium
(containing 10 mM unlabeled Met and Cys) and then incubated further as indicated. Before lysis
(as above), cells were washed (3X) with phosphate-buffered saline. Nedd4-2 was immunoprecipitated using V5 antibodies from denatured lysates (as described above). Gels were dried and exposed to x-ray film. The amount of Nedd4-2 was quantified from x-ray films using an Alpha Imager and the spot density analysis function of AlphaEase software (Alpha Innotech
Corp.).
87 Chapter 3
Electrophysiological measurements of ENaC activity in Xenopus oocytes - Plasmids encoding all
3 ENaC subunits (3.3 ng each) and xNedd4-2, either WT or the YA mutant (2.5 ng each), were
linearized, in vitro transcribed and injected into X. laevis oocytes. Electrophysiological measurements for amiloride (5 μM)-sensitive Na channel activity were carried out after
overnight incubation, as previously described (Schild et al. 1996).
IV) Results
A) The Nedd4-2 WW domains bind to the HECT-PY motif and regulate catalytic activity
of the HECT domain
The HECT domain of Nedd4/Nedd4-2 and other HECT containing proteins contain a PY (LPxY)
motif (Fig. 3-1A and (Kasanov et al. 2001) ) located adjacent to the conserved catalytic Cys
residue (Fig. 3-1B). This suggests that the catalytic activity of Nedd4 family members may be
regulated through intra- or inter-molecular interactions between the WW domain(s) and the
HECT domain. To test for such binding, we performed an in vitro binding assay with GST-
tagged-HECT and His(x6)-tagged individual WW domains of Nedd4-2, but could not detect an
interaction between the two proteins (data not shown). Since the LPxY motif is in a turn, rather
than the normally extended conformation (Fig. 3-1B), it may be inaccessible to the WW domains
until a conformational change occurs, which is expected upon ubiquitin transfer from E2 to the
HECT active site Cys (Huang et al. 1999; Verdecia et al. 2003). To ensure accessibility of the
HECT-PY motif, we performed a Far-Western analysis, in which the GST-HECT was denatured
prior to incubation with the His6-WW domains. Strips of nitrocellulose (NC) bound with
denatured GST-HECT(WT), or GST-HECT(YA), in which Tyr971 is mutated to Ala, were
88 Figure 3-1. A PY motif is present within the HECT domain. (A) Alignment of C-terminal sequences of various human HECT domains, some belonging to the Nedd4 family with the conserved PY (LPXY) motif (orange) and active-site cysteine residue (yellow) highlighted. (B)
Ribbon diagram of the tertiary structure of the HECT domain of WWP1/AIP4 (a Nedd4-family member) (PDB code 1ND7) highlighting the PY motif found in a turn within the C lobe of the
HECT domain, with the N lobe coloured blue, the C lobe in coral and the linker hinge in green.
The side chains of the PY motif (L896–Y899) and the catalytic cysteine residue (C890) are coloured yellow.
89 Figure 3-1
90 Chapter 3
incubated with individual purified His6-Nedd4-2 WW domains. Interactions were detected by
probing the nitrocellulose with anti-His6 antibodies. Fig. 3-2A shows that His6-Nedd4-2 WW3, and to a lesser extent WW4 and WW2, bound the denatured HECT(WT) but not the HECT(YA) mutant. WW1 bound the denatured HECT domain very weakly (seen in only some of our experiments). Overall, binding to the HECT-PY motif appeared much weaker than binding to a
PY motif from βENaC (βPY) (Fig. 3-2A), a known Nedd4-2 substrate. The βPY peptide binds
Nedd4-2 WW3 and WW4 domains with Kd of ~10 μM and ~50 μM, respectively (Henry et al.
2003). Our measurements of the affinity of interaction between Nedd4-2 WW domains and a
Nedd4-2 HECT-PY peptide using intrinsic tryptophan fluorescence (the same method used to measure affinity between Nedd4-2 WW domains and βPY (Henry et al. 2003)) revealed un- saturable binding for up to 1 mM HECT-PY motif peptide incubated with 1 µM WW domains, suggesting that binding is very weak (Kd>400 μM). Note, however, that this interaction is
measured with Nedd4-2 WW domains and HECT PY motif peptides as separate molecules. In
the natural protein, the WW domains and the HECT domain are part of the same polypeptide
chain. The Nedd4-2 WW domains and HECT-PY motif could be involved in an intramolecular
interaction within Nedd4-2 and, therefore, a high affinity interaction may not be necessary for
binding.
We next investigated whether this Nedd4-WW domain : HECT-PY motif interaction affects catalytic activity of the HECT domain. Using purified recombinant Nedd4-2 HECT
(after removal of its GST tag), we set up an in vitro ubiquitination assay in which Nedd4-2
HECT catalytic activity was measured by self(auto)-ubiquitination and detected using anti- ubiquitin (anti-Ub) antibodies. Self-ubiquitination was then measured in the presence of increasing concentrations of individual Nedd4-2 WW domains (His6-WW1,-2,-3,-4). Our results
91 Chapter 3
show that increasing amounts of Nedd4-2 WW1, WW2, and WW3 significantly inhibited
Nedd4-2 HECT self-ubiquitination (Fig. 3-2B), however WW4 and GST alone (used as negative
control) did not affect ubiquitination. We also observed inhibition of HECT catalytic activity by
WW3 in the context of full-length Nedd4-2 (Fig. 3-2C). These results suggest that interaction
between Nedd4-2 WW domain and HECT domain may play a role in regulating the catalytic activity of Nedd4-2. The discrepancy between WW4 domain binding and inhibition may be a result of the method used to test binding. We had previously demonstrated that WW4 does not bind well to PY motifs with a Leu in the first position (i.e. LPxY) due to steric restraints between this residue and a bulky His in the WW4 domain (Kanelis et al. 2006), therefore, it is not
surprising that WW4 was not able to inhibit HECT activity. In the Far Western binding
experiment (Fig. 3-2A), the HECT domain is denatured prior to binding to the WW domains,
which may have allowed for weak binding that is not observed when both proteins are in their native state.
To demonstrate dependence of WW domain-mediated inhibition on the HECT PY motif
we attempted to repeat this experiment using the HECT(YA) mutant, with the expectation that
HECT(YA) activity would not be inhibited by WW domains since they cannot bind. However,
we observed that the HECT(YA) mutant was not active in the presence or absence of WW
domains (data not shown). While CD spectroscopy revealed that the HECT(YA) mutant was
folded, the CD spectra was not identical to HECT(WT) (data not shown), suggesting there may be slight structural changes affecting activity. Interestingly, the YA mutation in the context of full-length Nedd4-2 (Nedd4-2(YA)) was active, but its activity was somewhat reduced relative to
Nedd4-2(WT) (see below). This difference in catalytic activity precluded quantitative comparison of WW inhibition of Nedd4-2 WT and YA.
92 Figure 3-2. The Nedd4-2 HECT PY motif binds its WW domains and regulates catalytic
activity. (A) Weak binding of HECT PY motif to the WW domains of Nedd4-2. Upper panel:
Strips of nitrocellulose containing denatured GST–HECT(WT) or the YA mutant domain
(arrow) were incubated with purified His6–WW domains (His–WW1, His–WW2, His–WW3 and
His–WW4) from Nedd4-2, and immunoblotted with an anti-His6 antibody to detect binding of
WW domains to the HECT domain. The nitrocellulose strips also contained GST–PY motif from
as a positive control. Lower panel: Coomassie staining to demonstrate the (כ ,βENaC (βPY
amount of WW domain inputs (50% of the total protein incubated with nitrocellulose strips). (B)
In vitro ubiquitination of the Nedd4-2 HECT domain is inhibited by its WW domains. Purified
Nedd4-2 HECT was incubated with increasing amounts of His6–Nedd4-2 WW1, WW2, WW3 or
WW4 domains (0, 2.5, 5 and 10 μg), or GST alone, together with E1 and E2 (UbcH7) enzymes
and ATP. Reaction mixtures were analysed by immunoblotting with an anti-ubiquitin antibody.
(C) In vitro ubiquitination of full-length Nedd4-2 is also inhibited by its WW domains. Purified
Nedd4-2 was incubated with increasing amounts of His6–Nedd4-2 WW3 (0, 2.5, 7.5 and 15 μg), together with E1 and E2 (UbcH7) enzymes and ATP, and analysed by blotting with an anti- ubiquitin (anti-Ub) antibody. A Western blot using an anti-Nedd4-2 antibody shows that an equal amount of Nedd4-2 was present in all lanes (lower panel).
93 Figure 3-2
94 Chapter 3
B) Nedd4-2 stability is regulated by self-ubiquitination and subsequent degradation
We have frequently observed higher expression levels of a catalytically inactive Nedd4-2
(Nedd4-2(CS)) compared to Nedd4-2(WT) transfected in HEK293T cells, suggesting that
Nedd4-2 may destabilize itself through self-ubiquitination. To examine this in more detail, we
studied ubiquitination levels and stability of Nedd4-2 in vitro and in cells. First, we compared the
ability of full length V5-tagged Nedd4-2 (WT, CS or YA) expressed and immunopurified from
HEK293T cells, to carry out ubiquitination in an in vitro assay. As seen in Fig. 3-3, both Nedd4-
2(WT) and Nedd4-2(YA), but not Nedd4-2(CS), exhibited ubiquitination of Nedd4-2 and
possibly other putative associated proteins that co-precipitated with it.
To test specifically for Nedd4-2 ubiquitination, HEK293T cells were transfected with human V5-tagged Nedd4-2 (WT, CS, or YA). Cell lysates were then boiled in SDS to dissociate
Nedd4-2 interacting proteins. Nedd4-2 was immunoprecipitated (IP) from the boiled lysates using anti-V5 antibodies, and immunoblotted with anti-ubiquitin antibodies to detect conjugation of ubiquitin. As shown in Fig. 3-4A, a high molecular weight smear representing self- ubiquitinated Nedd4-2 is observed for Nedd4-2(WT), but not for Nedd4-2(CS), demonstrating that the ubiquitination observed is indeed self-ubiquitination. Notably, Nedd4-2(YA) was also able to self-ubiquitinate.
To determine whether self-ubiquitination results in protein degradation we carried out pulse-chase experiments and compared stability of Nedd4-2(WT) with that of its mutants. As seen in Fig. 3-4B,C Nedd4-2(CS), which cannot ubiquitinate itself, is much more stable than
Nedd4-2(WT) (half-life of ~13 hrs relative to ~7 hrs for the WT), suggesting that self-
ubiquitination results in protein degradation. Notably, despite the somewhat decreased catalytic
activity exhibited by Nedd4-2(YA), it nevertheless displayed reduced stability relative to Nedd4-
95 Figure 3-3. Both Nedd4-2(WT) and the HECT-PY mutant, Nedd4-2(YA), display ubiquitination activity in an in vitro assay. V5-tagged Nedd4-2 (WT, CS or YA) were immunopurified from transfected HEK-293T cell lysates and incubated with ATP, ubiquitin, and
E1 enzyme, with (+) or without (−) E2 (UbcH7) enzyme, as indicated. Reaction mixtures were analysed by immunoblotting with an anti-ubiquitin antibody (anti-Ub) (upper panel). Lower panel: Western blot analysis with an anti-V5 antibody to show that equal amounts of V5–Nedd4-
2 (WT, CS and YA) lysates [and immunoprecipitates (IP)] were used in all treatments.
96 Figure 3-3
97 Figure 3-4. Mutation of the HECT PY motif affects Nedd4-2 stability. (A) Nedd4-2(YA)
displays similar self-ubiquitination to Nedd4-2(WT). V5–Nedd4-2 (WT, CS or YA) constructs
were transfected into HEK-293T cells. Cells were lysed and lysates boiled in SDS to remove
Nedd4-2-interacting proteins. Lysates were then immunoprecipitated (IP) with an anti-V5
antibody (to precipitate Nedd4-2) and immunoblotted with an anti-ubiquitin antibody to detect
ubiquitinated Nedd4-2. (B) HEK-293T cells expressing V5–Nedd4-2 (WT, CS or YA) were
labelled with [35S]methionine/cysteine (pulse) and chased with non-radioactive methionine/cysteine for indicated times. Cell lysates (500 μg from each sample) were immunoprecipitated with an anti-V5 antibody, followed by SDS/PAGE. The experiment was performed in triplicate, with a representative experiment shown. (C) Half-life (t½) of Nedd4-2
WT and mutants (CS and YA). The amount of Nedd4-2 was quantified from film by spot-density
analysis. Results are means + S.D. (n =3, except n =2 for Nedd4-2 CS). The t½ of the YA mutant
.(P<0.01, unpaired t testככ) was significantly shorter than that of Nedd4-2 WT
98 Figure 3-4
99 Chapter 3
2(WT) (half-life of 5 hrs). These data suggest that the WW domain : HECT interaction stabilizes
Nedd4-2 by inhibiting self-ubiquitination, in agreement with the observed inhibition of HECT
activity upon WW domain binding (Fig. 3-2B,C).
C) The HECT-PY motif regulates self-, but not substrate, ubiquitination and stability
We next examined the affects of the Nedd4-2 HECT-PY mutation, in the context of the full-
length protein, on the ability of Nedd4-2 to regulate a known substrate in a physiological assay.
Using a Xenopus oocyte expression system, we measured the ability of Nedd4-2(YA) to
downregulate ENaC. Voltage-clamp experiments were performed in which amiloride-sensitive
+ Na currents (INa) were measured in Xenopus oocytes expressing ENaC in the absence or
presence of co-expressed (Xenopus) Nedd4-2(WT) or Nedd4-2(YA). Surprisingly, we observed
that Nedd4-2(YA) could downregulate ENaC similarly to Nedd4-2(WT) (Fig. 3-5A), suggesting
that the small reduction in activity exhibited by Nedd4-2(YA) is not sufficient to affect ENaC
downregulation, at least in this system.
We then investigated the ability of Nedd4-2(YA) to specifically ubiquitinate its known
substrate, ENaC. Fig. 3-5B demonstrates that this Nedd4-2(YA) mutant can ubiquitinate ENaC
(and destabilize the channel) almost as effectively as Nedd4-2(WT), unlike the catalytically-
inactive Nedd4-2(CS) that was used as a negative control. Fig. 3-5B depicts ubiquitination of
γENaC, the subunit previously shown to be most important for ENaC ubiquitination (Staub et al.
1997) (we also found Nedd4-2-mediated ubiquitination of αENaC (not shown), which is also
known to contribute to channel ubiquitination (Staub et al. 1997)). These results suggest that
mutation of the HECT-PY motif does not affect the ability of Nedd4-2 to ubiquitinate its
substrates, in agreement with the ability of this mutant to properly downregulate ENaC.
100 Figure 3-5. Nedd4-2(YA) ubiquitinates and regulates substrate as effectively as Nedd4-2
(WT). (A) Inhibition of ENaC by Nedd4-2(YA). Xenopus oocytes were injected with cRNA encoding ENaC α, β and γ subunits (3.3 ng per subunit) together with 2.5 ng of Xenopus Nedd4-
+ 2(WT) or Nedd4-2(YA). After incubating for 24 h at 19◦C, INa (amiloride-sensitive Na currents) were measured and normalized to control cells which had been injected only with rENaC.
Results are means + S.E.M. (n =19) (B) Ubiquitination of ENaC by the Nedd4-2(YA) mutant.
Epitope-tagged α, β and γ subunits of ENaC (with each subunit differently tagged as indicated) were transfected into HEK-293T cells together with V5-tagged human Nedd4-2 (WT, CS or
YA). γENaC (the subunit that undergoes most prominent ubiquitination) was then immunoprecipitated (IP) with an anti-FLAG antibody and immunoprecipitates blotted with anti- ubiquitin (first panel) or anti-FLAG antibodies (second panel). The lower four panels depict expression of each ENaC subunit (arrows) and Nedd4-2 (WT, CS or YA) in the lysates of transfected cells. Note the reduced stability of ENaC subunits in cells transfected with Nedd4-
2(WT) or Nedd4-2(YA), but not with Nedd4-2(CS).
101 Figure 3-5
102 Chapter 3
Since substrate binding likely disrupts any inhibitory interaction between Nedd4-2 WW
domains and the HECT domain, we hypothesized that substrate binding might also promote
Nedd4-2 self-ubiquitination. To test this hypothesis, we repeated the Nedd4-2 ubiquitination
experiment comparing ubiquitination of Nedd4-2 immunoprecipitated from HEK293T cells that were transfected with Nedd4-2 alone (WT or mutants) or Nedd4-2 plus ENaC. In the presence of ENaC, Nedd4-2(WT) reveals greater self-ubiquitination than when expressed alone (Fig. 3-6, top panel). In contrast, the YA mutant, which as mentioned is intrinsically less active, displays a
less drastic increase in self-ubiquitination upon ENaC co-expression compared with WT. This is
likely because substrate binding is not required to relieve an inhibitory interaction and promote
self-ubiquitination in this case.
A model depicting the binding surface between Nedd4-2 WW3 and the HECT-PY motif
based on previously solved structures of WW domain – PY motif complexes is shown in Fig. 3-
7A,B. Although the core PY motif residues (LPPY) adopt a conformation seen with other PY
motif-containing peptides bound to WW domains, unexpectedly, the C-terminal Phe residue does
not contact the WW domain (Fig. 3-7A). This is in contrast to our previously solved Nedd4
WW4 domain : βENaC PY motif and Nedd4 WW3 domain : Commissureless PY motif
complexes in which the analogous residue bound its cognate WW domain (Fig. 3-7B). Based on
this model and our previous work we predicted that replacing the HECT-PY motif with the
higher affinity PY motif from βENaC (PPNYDSL) would strengthen intra/inter-molecular interactions with the Nedd4-2 WW domains and prevent substrate ubiquitination. While such substitution indeed strengthened the interaction (Fig. 3-8A), unfortunately, it also abolished
HECT activity (not shown). A less drastic mutation was then employed, replacing Phe at the
Tyr+3 position with Leu (LPPYDSF to LPPYDSL). According to our WW3:HECT PY-motif
103 Figure 3-6. Substrate ubiquitination promotes Nedd4-2 self-ubiquitination. V5-tagged
hNedd4-2 (WT or YA) was transfected into HEK-293T cells in the presence (+) or absence (−) of α, β and γ subunits of ENaC. Cells were lysed and lysates boiled in SDS to remove Nedd4-2- interacting proteins. Lysates were then immunoprecipitated (IP) with an anti-V5 antibody (to precipitate Nedd4-2) and immunoblotted with an anti-ubiquitin (anti-Ub) antibody to detect ubiquitinated Nedd4-2 (first panel) or an anti-V5 antibody (second panel). The lower three panels depict expression of each ENaC subunit (arrows) in lysates of transfected cells.
104 Figure 3-6
105 Figure 3-7. Homology model of a Nedd4-2 WW3 domain–HECT PY motif complex. (A) An ensemble of the 100 lowest-energy models (from a total of 500). (B) Comparison of the human
Nedd4-2 WW3 domain–HECT PY motif complex with the Nedd4 WW4 domain–βENaC PY motif complex. The WW domain is shown in blue for the backbone and cyan for the side chains.
For clarity, only side chains of canonical tryptophan (W523 and W545) and proline (P548) residues are shown. The HECT PY motif is shown in orange for backbone and yellow for side chains. The canonical PY motif residues (L968, P969 and Y971) are shown, as well as F974.
106 Figure 3-7
107 Chapter 3 model (Fig. 3-7A,B), the bulky Phe residue at the Tyr+3 position was likely responsible, at least in part, for the low affinity interaction observed with the WW domain; indeed, mutation of this
Phe residue to Leu led to a modest increase in strength of interaction (Fig. 3-8A). Although this
Phe->Leu (FL) substitution did not affect activity of the full-length protein, as evidenced by its ability to self-ubiquitinate (Fig. 3-8B), it also did not affect its stability (data not shown), suggesting that the modest increase in affinity of the FL mutant was not sufficient to compete with substrate binding.
V) Discussion
Our work suggests a novel mode of regulation of catalytic activity of the E3 ligase Nedd4-2, via an intra- or inter-molecular interaction between its WW domains and an internal PY motif within its HECT domain, leading to regulation of self-ubiquitination, but not substrate ubiquitination
(Fig. 3-9). This report differs from very recent reports describing regulation of other HECT- containing E3s. Smurf2, a member of the Nedd4 family involved in TGF-β signalling, is regulated through binding of an adaptor protein Smad7, promoting E2 binding to the HECT domain (Ogunjimi et al. 2005). Similar to what we observed for Nedd4-2, Smurf2 self- ubiquitination results in degradation and is enhanced by Smad7 binding. More recently, Smurf2 catalytic activity was also shown to be regulated by an auto-inhibitory interaction between the
C2 and HECT domains (Wiesner et al. 2007). Itch, another Nedd4 family member, is regulated by JNK-mediated phosphorylation, which enhances catalytic activity by disrupting an inhibitory interaction between the central region (including a Pro-rich region and Itch WW domains) and the HECT domain (Gallagher et al. 2006); however, they did not identify or implicate the
HECT-PY motif in the observed intramolecular interaction. Interestingly, in this case,
108 Figure 3-8. Nedd4-2(FL) (phenylalanine to leucine residue mutant) displays an increased
affinity for the HECT PY motif but exhibits the same stability as Nedd4-2(WT). (A)
Increased binding of the mutant HECT PY motif to Nedd4-2 WW domains. Strips of nitrocellulose containing denatured GST–HECT domain [WT, YA mutant, βENaC PY motif
(PPPNYDSL mutant) or FL mutant (LPPYDSF mutated to LPPYDSL)] were incubated with purified His6–WW domains (WW1–WW4 domains) from Nedd4-2, and immunoblotted with an
anti-His antibody to detect binding of WW domains to the mutated HECT domain. Note the
strong increase in binding of WW2, WW3 and WW4 domains to the βENaC PY motif, and the modest increase in binding to the FL mutant, relative to HECT(WT). (B) Nedd4-2(FL) mutant self-ubiquitinated similarly to Nedd4-2(WT) and Nedd4-2(YA). V5–Nedd4-2 (WT, CS, YA or
FL) was immunoprecipitated (IP) from transfected HEK-293T cell lysates using an anti-V5
antibody, boiled in SDS to remove Nedd4-2-interacting proteins. Subsequently, ubiquitinated
Nedd4-2 was observed by immunoblotting with an anti-ubiquitin (anti-Ub) antibody, with
blotting with an anti-V5 antibody performed as a control.
109 Figure 3-8
110 Chapter 3
phosphorylation was required to disrupt the inhibitory interaction. Contrary to what was
observed with Smurf2 and Nedd4-2, there was no evidence to suggest that activated
(phosphorylated) Itch was less stable than the unphosphorylated form. Activity of the yeast
Nedd4 homologue, Rsp5, was recently shown to be modulated by association with a deubiquitylating enzyme complex (Kee et al. 2005). Notably, this association affects only ubiquitination of Rsp5 substrates, since Rsp5 is not known to undergo auto-ubiquitination in vivo
and has the same half-life as that of the catalytically inactive Rsp5. Additionally, Mule/ARF-
BP1, a HECT E3 ligase for p53, is negatively regulated by its binding partner ARF, which binds
directly to the HECT domain (Chen et al. 2005). Our data and these recent reports collectively
suggest that while some similarities in HECT activity regulation exist, each HECT ligase seems to exhibit its own unique mode of regulation, tailored to fit the needs of the proteins they regulate.
For Nedd4-2 (and possibly other Nedd4 family members), the PY motif within the HECT domain regulates low affinity intra/inter-molecular interactions and self-ubiquitination. This likely maintains the HECT domain in an inactive state, preventing self-ubiquitination, thereby stabilizing the protein. When the enzyme encounters a real substrate (e.g. ENaC), to which it binds with higher affinity (Henry et al. 2003), the intra/inter-molecular WW domain: HECT-PY motif interaction is disrupted. Our data suggest that once substrate ubiquitination is complete self-ubiquitination becomes prominent, either subsequent to or simultaneously with substrate
ubiquitination (Fig. 3-9), leading to destabilization of Nedd4-2 itself.
Finally, it is curious that the HECT-PY motif is also conserved in HECT E3s that do not contain WW domains (e.g. E6-AP), raising the possibility that they may be regulated by inter-
111 Chapter 3 molecular interactions with Nedd4-family members in cells that express such family members in the same compartment.
112 Figure 3-9. Model depicting regulation of Nedd4-2 catalytic activity by the HECT PY motif. (A) The interaction between Nedd4-2 WW domains and its own HECT domain stabilizes the protein by preventing Nedd4-2 self-ubiquitination. (B) Upon binding to a substrate (e.g.
ENaC), the inhibitory WW–HECT interaction is competed off by the substrate PY motif, since this interaction is of a higher affinity. This leads to substrate ubiquitination and down-regulation.
(C) Following (or concurrent with) substrate ubiquitination and the disruption of the inhibitory
WW–HECT interaction, Nedd4-2 now becomes self-ubiquitinated, leading to its own de- stabilization.
113 Figure 3-9
114
CHAPTER 4
Characterization of the Nedd4-2 interacting protein Rnf11
The work in this chapter describes unpublished work. I performed the design, execution and analysis of all experiments presented in this chapter.
115 Chapter 4
I) Summary
Nedd4-1 and Nedd4-2 are E3 ubiquitin ligases belonging to a family of proteins sharing the
same domain makeup, namely, a C2 domain, 3-4 WW domains, and a ubiquitin ligase
HECT domain. Our lab previously identified Rnf11 in an expression screen looking for
proteins interacting with Nedd4 WW domains. Rnf11 is a highly conserved protein
containing a single PY-motif, through which it binds Nedd4 WW domains, a putative
ubiquitin-interacting motif (UIM), and a C-terminal RING-finger domain. RING-finger proteins are also known to have ubiquitin ligase activity, thus I studied the putative
reciprocal regulation of these two E3 ligases; the HECT-containing Nedd4 and the RING-
containing Rnf11. Here we demonstrate that Rnf11 exhibits intrinsic ubiquitin ligase activity, and that Rnf11 preferentially binds Nedd4-2, relative to Nedd4-1. Additionally, we show that Nedd4-2, but not Nedd4-1, mediates mainly mono- and di-ubiquitination of
Rnf11 and that interaction between the two E3s results in a decrease of Nedd4-2 mediated
self-ubiquitination. While the exact function of Rnf11 is unclear, possible roles for Rnf11,
and the Rnf11/Nedd4-2 interaction, in protein trafficking, cell signalling and transcription
are discussed.
116 Chapter 4
II) Introduction
Ubiquitination, referring to covalent attachment of the highly conserved 76-amino acid polypeptide ubiquitin to target protein lysine residues, is an essential process regulating a host of critical functions including protein quality control, protein stability, signal transduction, endocytosis and protein trafficking. Of the three enzymes that carry out this process; E1, E2 and
E3, the E3s, or ubiquitin protein ligases, are responsible for determining substrate specificity by binding to target proteins. There are two major classes of E3 enzymes; HECT-type E3 ligases and RING-finger E3s. Both HECT and RING domains bind E2 enzymes, however, the HECT domain forms a covalent intermediate with ubiquitin during catalysis, while the RING-finger acts as a molecular scaffold to transfer ubiquitin from E2 to substrate. RING-finger E3s can be further subdivided into two groups; single-subunit E3s such as c-Cbl and Mdm2, and those that function in a multi-subunit complex such as Roc1 (also known as Rbx1) and Apc11, which form the catalytic core of a multi-subunit E3 complex with Cullin proteins.
Nedd4-1, and its close relative Nedd4-2, are HECT-type E3 ubiquitin ligases belonging to a family of Nedd4-like proteins all sharing the same common domain architecture; an N-terminal
C2 domain, 3-4 WW domains, and a HECT domain. The WW domains of Nedd4-1/-2 mediate substrate recognition by binding to PY motifs (L/PPxY) found in target proteins (Kanelis et al.
2001; Kasanov et al. 2001). An expression screen of a 16-day mouse embryo library, using the second WW domain (WW2) of rat Nedd4-1 (rNedd4) as bait, was carried out by our lab in order to identify potential Nedd4 substrates. Among interacting proteins identified was the RING- finger protein Rnf11 (N. Pham – PhD Thesis).
The Rnf11 gene product encodes a 154-amino acid protein containing a single PY motif, a potential ubiquitin-interacting motif (UIM), and a C-terminal RING-finger (Fig. 4-1). Northern
117 Figure 4-1. Schematic and protein sequence of the human Rnf11 protein. Rnf11, a 154 amino-acid protein, contains an N-terminal myristoylation sequence (in green), a PY-motif (in
yellow), a potential ubiquitin-interaction motif (UIM) (in red), and a C-terminal RING-H2 domain (in blue). Rnf11 also contains four lysine (K) residues which could serve as potential ubiquitin acceptor sites (indicated with an arrow). (B) Schematic of Rnf11 mutants used in this study.
118 Figure 4-1
A. K K KK
PY UIM RING Myristoylation
MGNCLKSPTSDDISLLHESQSDRASFGEGTEPDQEPPPPYQEQVPVPVYHPTPSQTRLAT
QLTEEEQIRIAQRIGLIQHLPKGVYDPGRDGSEKKIRECVICMMDFVYGDPIRFLPCMHIY
HLDCIDDWLMRSFTCPSCMEPVDAALLSSYETN
B. PPPY‐>PPPA
YA PY UIM RING
C‐>S
CS PY UIM RING
STOP ∆R PY UIM
EEE‐>AAA
UIMm PY UIM RING
119 Chapter 4
blot analysis of mouse and human tissues reveals that Rnf11 is expressed in most tissues,
including heart, brain, lung, liver, kidney, skeletal muscle, spleen and testes (Seki et al. 1999;
Kitching et al. 2003). Rnf11 homologues are found in a number of vertebrate organisms
including frog, zebrafish, mice and human, as well as invertebrates, namely fruitflies and
nematodes (Fig. 4-2A). At the protein level, Rnf11 is virtually unchanged from frog to human, however, the N-terminal domain which contains the PY motif seems to diverge at the level of invertebrates, suggesting the Rnf11 N-terminus may serve a unique function in vertebrates (Fig.
4-2A,B). Additionally, the absence of any immediate Rnf11 family members with related structure suggests Rnf11 may have an integral and non-redundant role in cell physiology (Azmi and Seth 2005).
When this project was started nothing was published with regards to Rnf11, however, subsequent to our labs identification of interactions between Nedd4-1 and Rnf11 (N. Pham –
PhD Thesis), Rnf11 was shown to interact with other Nedd4 family members, including Smurf1 and 2, AIP4/Itch, and WWP1 and 2, in a PY-motif dependent manner (Jolliffe et al. 2000;
Colland et al. 2004; Azmi and Seth 2005). The literature to date has focused on the interaction between Rnf11 and Smurf2. Smurf2 is involved in downregulating TGFβ signalling through ubiquitination of the TGFβ receptor, a process requiring interaction between Smurf2 and Smad7
(Kavsak et al. 2000). It has been suggested that Rnf11 plays a role in upregulating TGFβ signalling by disrupting formation of the Smurf2/Smad7 complex, thereby inhibiting ubiquitination of the TGFβ receptor by Smurf2 (Subramaniam et al. 2003). Rnf11 was found to be highly overexpressed in breast and pancreatic cancers, and moderately overexpressed in head and neck, colon and lung cancers, which may be linked to its potential role in TGFβ signalling
120 Figure 4-2. Comparison of Rnf11 orthologues. (A) Amino acid alignment of Rnf11 from various species. Aligned using CLUSTALW package from the European Bioinformatics
Institute (EMBL-EBI). “ * ” denotes identical residues; “ : ” denotes conserved substitutions;
“ . ” denotes semi-conserved substitutions. (B) Phylogenetic tree analysis of Rnf11 homologues, also performed using the CLUSTALW program.
121 Figure 4-2 A.
Human MGNCLKSPT----SDDISLLHESQSDRASFGEGTEPDQEPPPPYQEQVPV-PVYHPTPSQTR 57 Mouse MGNCLKSPT----SDDISLLHESQSDRASFGEGTEPDQEPPPPYQEQVPV-PIYHPTPSQTR 57 X.laevis MGNCLKSPT----SDDISLLHESQSDRASYGEGNDGDHEPPPPYQEQAPI-PVYHPTPSQTR 57 Pufferfish MGNCLKSPT----SDDISLLHESQSDRASYGDGADPDQEPPPPYEEQIHI-PVYHPTPSQAR 57 Zebrafish MGNCLSSQG----ADDLSLLNES------EGASLPGEPPPPYQERAQV-PVYHPTPSQTR 49 Drosophila MGNCLKIST----SDDISLLRGN------DSQISGTQPVYHQGEHYQR-ELYPSTSSSTT 49 C.elegans MGNCLPSLFGFVRQHDETPLRRSRSS-----DAAMSFTNPSMMQQSSSYVNQLYQHNVIRQR 57 ***** .* : *. . :. :* . :* .
Human LAT-----QLTEEEQIRIAQRIGLIQHLPKGVYDPGRDGSEKKIRECVICMMDFVYGDPIR 113 Mouse LAT-----QLTEEEQIRIAQRIGLIQHLPKGVYDPGRDGSEKKIRECVICMMDFVYGDPIR 113 X.laevis LAT-----QLTEEEQIRIAQRIGLIQHLPKGVYDPGRDGSEKKIRECVICMMDFVYGDPIR 113 Pufferfish LAT-----QLTEEEQVRIAQRIGLIQHLPKGVYDPGRDGSEKKIRECVICMMDFVYGDPIR 113 Zebrafish LAT-----QLTEEEQVRIAQRIGLIQHLPRGIFDPGSEPSDKKIKECVICMMDFEYGDPIR 105 Drosophila LTPSSNNRQLSDENQVKIAKRIGLMQYLPIGTYD----GSSKKARECVICMAEFCVNEAVR 106 C.elegans QAQE----QGKELDEAKKNRIRGLLEQIPADVFR-----GDMTSNECAICMIDFEPGERIR 109 : * .: :: : : **:: :* . : .. . .**.*** :* .: :* Human FLPCMHIYHLDCIDDWLMRSFTCPSCMEPVDAALLSSYETN------154 Mouse FLPCMHIYHLDCIDDWLMRSFTCPSCMEPVDAALLSSYETN------154 X.laevis FLPCMHIYHMDCIDDWLMRSFTCPSCMEPVDAALLSSYETN------154 Pufferfish FLPCMHIYHVDCIDDWLMRSFTCPSCMEPVDAALLSSYETN------154 Zebrafish FLPCMHIYHVDCIDAWLMRSFTCPSCMEPVDAALLSSYETN------146 Drosophila YLPCMHIYHVNCIDDWLLRSLTCPSCLEPVDAALLTSYEST------147 C.elegans FLPCMHSFHQECVDEWLMKSFTCPSCLEPVDSTILSSLTAHNMQSLQQIVCSPTSSSTAKP 170 :***** :* :*:* **::*:*****:****:::*:* :
B.
122 Chapter 4
(Subramaniam et al. 2003), since this pathway is commonly deregulated in cancer (Levy and Hill
2006).
Here we demonstrate that Rnf11 has intrinsic ubiquitin ligase activity, suggesting it functions as a single subunit RING-E3 ligase. While the original screen carried out by our lab identified Rnf11 using the Nedd4-1 WW2 domain as bait (which is almost identical to the
Nedd4-2 WW2 domain), we show that Rnf11 preferentially binds Nedd4-2. Additionally, we demonstrate that Nedd4-2 is involved in mediating mainly mono- and di-ubiquitination of Rnf11.
Interestingly, we observe that interaction between Rnf11 and Nedd4-2 results in a decrease of
Nedd4-2 mediated self-ubiquitination.
In addition to binding members of the Nedd4 family, Rnf11 has also been shown to interact with a diverse array of proteins (Colland et al. 2004; Azmi and Seth 2005) . While the majority of Rnf11 interacting proteins identified (by others) are involved in the ubiquitin pathway, a large percentage (17%) are known to be involved in endocytosis and protein sorting.
Notably, Rnf11 contains a putative UIM and is monoubiquitinated, two characteristics shared by many adaptor proteins involved in protein trafficking, suggesting that Rnf11 may also play a role in this process. Here we discuss possible functions of Rnf11 and the role of the Rnf11/Nedd4-2 interaction.
123 Chapter 4
III) Experimental Procedures
Reverse-transcriptase PCR – RNA was isolated from confluent flasks (75 cm2) of HEK 293T,
3T3 and HeLa cells using the RNAeasy Mini Kit (Qiagen). Isolated RNA was then used to
generate cDNA using the cDNA synthesis field-test kit (Qiagen). The presence of Rnf11 mRNA
was determined by PCR using cDNA (using Platinum Pfx, Invitrogen) from 293T, 3T3, and
HeLa as template and the following Rnf11 primers: 5’ primer (GGG GAA CTG CCT CAA
ATC CCC), 3’ primer (CAT AGG ATG AAA GCA GTG CTG C). The same primers were
used for both mouse (3T3) and human (293T and HeLa) cell lines, since mouse and human
Rnf11 have the exact same DNA sequence in the regions to which our primers anneal. PCR
products were run on 1% agarose gels containing ethidium bromide and detected under UV light.
Plasmids – Mouse Rnf11 was shuttled from the Gateway entry vector (Invitrogen), into the Flag
expression vector (pcDNA3.1-nFlag), and into GST expression vectors for bacterial (pDEST15)
and mammalian expression (pDEST27). All Rnf11 mutations (i.e. CS, YA, ∆R, ∆RYA) were
made using the QuikChange mutagenesis kit (Stratagene). Human (h) Nedd4-1, and hNedd4-2
(isoform lacking the C2 domain) were shuttled into the V5 expression vector (pcDNA3.1-nV5), using Gateway Technology (Invitrogen). Nedd4-1/-2 CS mutations were also created using the
QuikChange mutagenesis kit. The bacterial expression plasmid pGEX-6P1-HECT, used for the in vitro ubiquitination assay was constructed by standard PCR and restriction enzyme cloning procedures using Xenopus (x) Nedd4-2 as template.
Transfections and Antibodies – Transient transfections in HEK 293T cells were performed using the calcium phosphate precipitation method, with a total of 20 µg DNA per 10-cm dish, or 4 µg
124 Chapter 4
DNA per well of a 6-well plate (for immunofluorescence). The antibodies used for
immunoblotting were anti-ubiquitin (Covance), anti-Flag (Sigma), anti-V5 (Invitrogen), anti-
GST (affinity purified rabbit polyclonal antibodies generated against GST) and anti-Nedd4
(rabbit polyclonal antibodies generated against rat Nedd4-1 WW2).
Immunofluorescent Confocal Microscopy – HEK 293T cells expressing Flag-Rnf11-WT or -YA
were stained with Alexa Fluor 488-conjugated concanavalin A (Invitrogen) to visualize the plasma membrane. Cells were then fixed with 4% paraformaldehyde in PBS and permeabilized with 0.1% Triton X-100. Fixed cells were incubated with appropriate sera to block nonspecific antibody binding. Flag-Rnf11 was visualized using anti-Flag antibodies (1:1000; Sigma) followed by Cy3-conjugated anti-mouse secondary antibodies (Jackson ImmunoResearch
Laboratories). Single-plane images were acquired with a LSM 510 confocal microscope under
63X, 1.4 NA, oil-immersion objective (Carl Zeiss MicroImaging, Inc.).
Pull-down and co-immunoprecipitation (co-IP) assays – Transfected HEK 293T cells were lysed
(48-hours post-transfection) with 1 ml (per 10-cm dish) of lysis buffer (50 mM Hepes, pH 7.4,
150 mM NaCl, 1% Triton C-100, 10% glycerol, 1.5 mM MgCl2, 1 mM EGTA), supplemented
with protease inhibitors (10 µg/ml leupeptin, 10 µg/ml aprotinin, 1 µg/ml pepstatin and 1 mM
PMSF). Lysates were cleared by centrifugation and used for pull-down and co-IP experiments.
For pull-down experiments, 1 mg of cell lysate was incubated with either 10 µL glutathione-
agarose beads (Sigma) for cells transfected with GST-Rnf11 (Fig. 4-5), or with 50 µg GST or
GST-Rnf11 bound to glutathione-agarose beads (Fig. 4-6A) for 2 h at 4°C. Beads were washed
twice with lysis buffer and twice with HNTG (20 mM Hepes, pH 7.4, 150 mM NaCl, 10%
125 Chapter 4
glycerol, and 0.1% Triton X-100). Bound proteins were eluted from the beads with 1 X SDS-
PAGE sample buffer, resolved by SDS-PAGE, and transferred to nitrocellulose. Bound proteins
were detected by immunoblotting with appropriate antibodies (see above), followed by secondary antibodies and ECL detection (GE Healthcare). For co-IP (Fig. 4-6B), 1.5 mg HEK
293T lysates transfected with Flag-Rnf11 (-WT, -YA, -∆R, or -∆RYA) and V5-Nedd4-1 or V5-
Nedd4-2 were incubated with 10 µL Flag-M2-agarose beads (Sigma) for 2 h at 4°C. Beads were
washed and proteins resolved and detected as above.
Ubiquitination assays – For in vitro ubiquitination assays, ~1 µg bacterially expressed GST-
Rnf11 or xNedd4-2-HECT domain, cleaved of its GST tag (using PreScission Protease, GE
Health) were incubated in reactions containing 100 nM yeast E1 (Boston Biochem), 500 nM E2
(bacterially purified His-UbcH5b or His-UbcH7), 2 µg wild-type ubiquitin (Sigma), or K29-,
K48-, K63-, or 3KR-Ub (Boston Biochem) and 4 mM ATP in reaction buffer (25 mM Tris (pH
7.5), 50 mM NaCl, 0.1 µM DTT, 4 mM MgCl2). Reactions were incubated for 1 h at room
temperature and analyzed by immunoblotting with anti-ubiquitin antibodies. For ubiquitination
assays in HEK 293T, cells were transfected with His-Ub, Flag-Rnf11 (WT or mutant) and V5-
Nedd4-1/-2, and then lysed 48-hours post-transfection in lysis buffer (as above) supplemented
with 50 µM LLnL and 0.4 mM chloroquine (both from Sigma). Cleared cell lysates were
denatured with 2% SDS and boiled for 5 min, then diluted 11 times with lysis buffer to dilute the
SDS prior to precipitation with Ni-NTA agarose beads (Qiagen). Beads were washed and
proteins resolved and detected same as for pull-down experiments. In Fig. 4-10 the amount of ubiquitinated Rnf11, and Rnf11 in cell lysates (non-ubiquitinated), was quantified from X-ray film using the spot-density-analysis function of AlphaEase software (Alpha Innotech). The
126 Chapter 4 relative Rnf11 ubiquitination was determined by dividing the amount of ubiquitinated Rnf11 by the total amount of Rnf11 (ubiquitinated plus non-ubiquitinated).
IV) Results
A) Expression and localization of Rnf11
In order to detect and localize endogenous Rnf11 in cultured cell lines we produced rabbit polyclonal antibodies against the N-terminus of Rnf11. While these antibodies efficiently detected overexpressed Rnf11 by western and immunofluorescence (not shown), we were unable to detect endogenous Rnf11 in several common cell lines (HEK 293T, HeLa, 3T3). Since previous studies have suggested that Rnf11 is ubiquitously expressed (Seki et al. 1999; Kitching et al. 2003), we verified the presence of Rnf11 mRNA transcripts in the indicated cell lines by
RT-PCR (Fig. 4-3).
Rnf11 was reported to be a predominantly cytoplasmic protein, with some nuclear accumulation, based on immunohistochemistry of tumour tissues and fractionation from breast cancer cell lines (Kitching et al. 2003; Subramaniam et al. 2003; Connor et al. 2005). To examine Rnf11 localization in a different cell line we carried out immunofluorescence on HEK
293T overexpressing Flag-tagged Rnf11 using confocal microscopy. While Fig. 4-4 clearly shows some cytoplasmic localization of Rnf11, a large portion of Rnf11 localizes to the plasma membrane, as determined by co-localization with the plasma membrane marker concanavalin A
(ConA). We did not observe any nuclear localization as previously reported, however it has been suggested that AKT-mediated phosphorylation of Rnf11 and subsequent 14-3-3 binding are required for translocation of Rnf11 to the nucleus (Connor et al. 2005). Additionally, we observed that interaction between Rnf11 and Nedd4-family members is not required for plasma
127 Figure 4-3. Rnf11 transcripts are found in HEK 293T, HeLa, and 3T3 cell lines. RT-PCR of
cDNA isolated from HEK 293T, HeLa and 3T3 cells. PCR products were run on 1% agarose gels containing ethidium bromide and detected under UV light.
128 Figure 4-3
129 Figure 4-4. Rnf11 colocalizes with the plasma membrane. Immunofluorescent (confocal) analysis of HEK 293T cells transfected with Flag-Rnf11-WT or -YA, immunostained with anti-
Flag antibodies (red), and Concanavalin A (ConA), used as a marker for the plasma membrane
(green). Colocalization of Rnf11 with the plasma membrane can be seen in the merged image
(yellow).
130 Figure 4-4
ConA Anti‐Flag, Cy3 Merge DIC
Flag‐Rnf11 WT
Flag‐Rnf11 YA
131 Chapter 4 membrane localization, as a Rnf11 mutant with the Tyr residue of the PY motif mutated to Ala
(Rnf11-YA) displays the same cellular localization as Rnf11-WT (Fig. 4-4).
B) Rnf11 binds both Nedd4-1 and Nedd4-2
As mentioned, Rnf11 was identified as a Nedd4-1 binding partner in a screen carried out by our lab (N. Pham, Thesis). In order to verify the interaction, HEK 293T cells were transfected with either GST-Rnf11-WT or GST-Rnf11-CS, a catalytically-inactive form in which the first RING- finger Cys residue was mutated to Ser. Glutathione-agarose beads were used to precipitate GST fusion proteins from transfected lysates and binding to endogenous Nedd4 was detected by immunoblotting with anti-Nedd4 antibodies (capable of detecting both Nedd4-1 and Nedd4-2), which are expressed endogenously in these cells (Pak et al. 2006). We observed coprecipitation of endogenous Nedd4 with catalytically-inactive (but not wild-type) Rnf11 (Fig. 4-5). In fact, two Nedd4 bands are observed coprecipitating with GST-Rnf11-CS, likely representing Nedd4-1 and Nedd4-2, or different Nedd4-2 isoforms, since several are known to exist (Itani et al. 2003).
To further verify the interaction, and to compare binding of Rnf11 to Nedd4-1 versus
Nedd4-2, we used recombinant GST-Rnf11 to pulldown Nedd4-1 or -2 from HEK 293T expressing catalytically-inactive V5-tagged Nedd4-1 or -2, in which the HECT catalytic Cys residue in each case was mutated to Ser (CS). Coprecipitation of both V5-Nedd4-1(CS) and V5-
Nedd4-2(CS) with GST-Rnf11 was observed, but not with GST alone, used as a negative control
(Fig. 4-6A). Notably, Rnf11 seems to preferentially interact with Nedd4-2, since more Nedd4-2 coprecipitated with GST-Rnf11, relative to Nedd4-1, even though the lysates contained equal amounts of both proteins (Fig. 4-6A).
132 Figure 4-5. Rnf11 binds endogenous Nedd4. HEK 293T cells were transfected with GST- tagged Rnf11 (-WT or -YA) or GST alone. GST fusion proteins were then precipitated from transfected lysates with glutathione-agarose beads. The complex was washed, resolved on SDS-
PAGE and immunoblotted with anti-Nedd4 antibodies (capable of detecting both Nedd4-1 and
Nedd4-2), to detect precipitation of endogenous Nedd4 (top panel), or anti-GST antibodies to verify precipitation of transfected GST fusion proteins (bottom panel).
133 Figure 4-5
134 Chapter 4
Using several Rnf11 mutants, we further examined the Nedd4/Rnf11 interaction. In addition to the Rnf11 PY-mutant, Rnf11-YA, used for confocal studies, we created a RING- deleted Rnf11 mutant (Rnf11-∆R), and a double mutant, deleted for the PY-motif and for the
RING-finger (Rnf11-∆RYA). Flag-Rnf11 (-WT, -YA, -∆R, or -∆RYA) was immunoprecipitated from HEK 293T lysates cotransfected with V5-Nedd4-1 or V5-Nedd4-2 and binding was detected by blotting against V5. As expected, both Nedd4-1 and -2 did not bind to Rnf11 PY- mutants, Rnf11-YA and Rnf11-∆RYA (Fig. 4-6B). Additionally, interaction between Rnf11 and
Nedd4-1/-2 was not dependent on the RING-finger domain, since both could bind Rnf11-∆R.
Again, we observe preferential binding of Rnf11 to Nedd4-2, relative to Nedd4-1, as only weak binding was observed between Nedd4-1 and Rnf11 and only to the RING-deleted mutant (also catalytically-inactive, Fig. 4-6B). In parallel, we cotransfected Smurf1 or Smurf2 with Flag-
Rnf11 (-WT or mutants) to compare Rnf11 binding between Nedd4-1/-2 and Smurf1/2. We observed a similar Rnf11 binding pattern for Smurf1 and 2 to that observed for Nedd4-1/-2, with both Smurf1 and Smurf2 requiring an intact Rnf11 PY-motif for binding (Fig. 4-6C). Overall,
Rnf11 appears to bind Smurf1 and Smurf2 slightly better than Nedd4-1 and -2. In addition,
Smurf1, like Nedd4-1, interacted with catalytically-inactive Rnf11 (-∆R), but not Rnf11-WT
(Fig. 4-6C).
C) Rnf11 exhibits in vitro ubiquitin ligase activity
RING-finger E3s functioning as a single subunit can function without the addition of accessory proteins. On the other hand, RING E3s belonging to multi-subunit complexes require binding of
Cullin proteins for E3 activity. To determine whether Rnf11 could function as a single-subunit
E3, recombinant GST-Rnf11 was used in an in vitro ubiquitination assay. GST-Rnf11 fusion
135 Figure 4-6. Rnf11 interacts preferentially with Nedd4-2 over Nedd4-1. (A) Bacterially
expressed and purified GST-Rnf11, or GST alone, bound to glutathione-agarose beads, were
used in a pulldown from HEK 293T lysates transfected with V5-Nedd4-1 CS or V5-Nedd4-2 CS
(catalytically-inactive mutants). Binding of Nedd4-1 or -2 was detected by blotting with anti-V5
antibodies (top panel). Equal amounts of GST-Rnf11 or GST alone was used for each pulldown,
as verified by blotting with anti-GST antibodies (bottom panel). Black lines separate non-
adjacent lanes from the same gel. (B) Flag-Rnf11 (-WT, -YA, -∆R, or -∆RYA) was immunoprecipitated from HEK 293T lysates co-transfected with V5-Nedd4-1 or V5-Nedd4-2.
Binding of Nedd4 (-1 or -2) was determined by immunoblotting with anti-V5 antibodies (top
panel) and immunoprecipitation of Flag-Rnf11 was verified by anti-Flag antibodies (second
panel). Cell lysates were blotted with anti-V5 and anti-Flag antibodies to detect amounts of
transfected Nedd4-1 or -2 and Rnf11, respectively (bottom two panels). (C) Same as in B, except
Flag-Rnf11 (-WT or mutants) was cotransfected with V5-Smurf1 or V5-Smurf2. Experiments
depicted in B and C were performed at the same time, exposed on the same sheet of film, and the
same exposures are shown for each, allowing for direct comparison of results.
136 Figure 4-6
A.
B. C.
137 Chapter 4
proteins were incubated with ubiquitin, recombinant E1 and E2 (UbcH5b or UbcH7), and
ubiquitin ligase activity was measured by detection of ubiquitinated GST-Rnf11, or
ubiquitination of other components in the mixture (i.e. E1, E2). UbcH5b was chosen as an E2
for this assay because it has been successfully used in similar ubiquitination assays with other
RING containing proteins (Lorick et al. 1999; Nie et al. 2002), and UbcH7 was chosen because its crystal structure was solved in complex with the RING finger protein c-Cbl (Zheng et al.
2000). Indeed, we observed E2-dependent ubiquitination in reactions containing GST-Rnf11
WT and UbcH5b (Fig. 4-7), in contrast to a recent report, observing Rnf11 E3 ligase activity only in the presence of the Cullin protein Cul1 (Azmi and Seth 2005). Interestingly, we did not observe catalytic activity for GST-Rnf11 in a reaction with UbcH7, demonstrating E2 specificity for Rnf11. As expected, mutation of the first Cys residue of the Rnf11 RING-finger domain to
Ser (GST-Rnf11-CS) severely reduced Rnf11 activity (Fig. 4-7).
As ubiquitin itself possesses several acceptor lysine residues – notably Lys29, Lys48 and
Lys63 – polyubiquitin chains with different types of ubiquitin linkages are possible. While K48- linked (and sometimes K29-linked) ubiquitin chains containing at least 4 ubiquitins target proteins for proteasomal degradation (Finley et al. 1994), K63-linked ubiquitin chains are involved in other processes, namely DNA repair, protein kinase activation, and endocytosis
(Weissman 2001; Sun and Chen 2004). By substituting wild-type (WT) ubiquitin in the in vitro
ubiquitination assay described above, with a mutant ubiquitin in which all ubiquitin Lys residues,
except one, are mutated to Arg (K29-, K48-, K63-Ub), we can determine which types of
ubiquitin chains are formed by Rnf11, possibly providing further clues to its function. From Fig.
4-8A we observe Rnf11 is capable of catalyzing the formation of ubiquitin-chains with K29-,
K48-, and K63-linkages, but not with a ubiquitin mutant in which only K29, K48 and K63 are
138 Chapter 4
mutated to Arg (3KR-Ub), suggesting Rnf11 does not form ubiquitin chains with linkages using any other ubiquitin lysine residue. Interestingly, the length of chains formed by Rnf11 with the mutant ubiquitins differs, with Rnf11 catalyzing the formation of shorter chains with K29-Ub than observed for WT-Ub, K48-Ub or K63-Ub (Fig. 4-8A).
Testing for the type of modification catalyzed by the Nedd4-2 HECT domain in the in
vitro ubiquitination assay (with single-lysine ubiquitins), as above, allowed for comparison between the types of ubiquitin chains made by Nedd4-2 and Rnf11. Interestingly, the longest ubiquitin chains formed by Nedd4-2 HECT contain K63-linkages, with Nedd4-2 forming much shorter chains with K29- or K48-linkages (Fig. 4-8B). Similar results were obtained using full- length Nedd4-2 in the same assay (not shown), in-line with previous results demonstrating
formation of K63-linked chains by the yeast Nedd4 homologue Rsp5 (Galan and Haguenauer-
Tsapis 1997). The absence of ubiquitin conjugation with 3KR-Ub implies that Nedd4-2, like
Rnf11, does not catalyze formation of ubiquitin chains with linkages other than K29, K48 or
K63. Notably, a much stronger ubiquitin signal is observed when WT-Ub is used in the assay, relative to the single-Lys ubiquitins, likely due to either a reduced efficiency of ubiquitin ligation by Nedd4-2 with ubiquitin mutants, as previously reported for another ubiquitin mutant (Woelk
et al. 2006), or due to the formation of multiple chain-linkages with WT-Ub (Fig. 4-8B).
Conversely, the same phenomenon is not observed for Rnf11, with a similar amount of
ubiquitination observed when either wild-type or mutant ubiquitin is used (Fig. 4-8A), perhaps
reflecting differences in the catalytic mechanisms employed by RING-fingers and HECT
domains.
139 Figure 4-7. Rnf11 has E2-dependent ubiquitin ligase activity in vitro. GST fusion of Rnf11, wild-type (WT) or a catalytically-inactive mutant (CS), bound to glutathione-agarose beads, were incubated in an in vitro ubiquitination reaction mixture in the absence (-) or presence of E2
(UbcH5b or UbcH7). The reaction products were resolved by SDS-PAGE and ubiquitinated proteins detected by western blotting using anti-ubiquitin antibodies. The bottom panel depicts 1
µg of the GST or GST-Rnf11 proteins used in the in vitro assay, western blotted with anti-GST antibody.
140 Figure 4-7
141 Figure 4-8. Rnf11 and Nedd4-2 make ubiquitin chains with different lysine linkages in vitro. (A) GST-Rnf11 was incubated in an in vitro ubiquitination assay with E1, E2 (UbcH5b),
ATP and either wild-type (WT) ubiquitin or ubiquitin mutants containing a single lysine residue
(K29-, K48-, or K63-Ub). The reaction products were resolved by SDS-PAGE and the formation of ubiquitin chains detected by immunoblotting with anti-ubiquitin antibodies. (B) as in A, except the Nedd4-2 HECT domain was used as the E3 ligase in place of Rnf11.
142 Figure 4-8
A. B.
143 Chapter 4
D) Role of the Nedd4-2/Rnf11 interaction
i) Ubiquitination of Rnf11 by Nedd4-2
Given that both Rnf11 and Nedd4-2 are E3 ligases, it is foreseeable that either Nedd4-2
ubiquitinates Rnf11, or Rnf11 ubiquitinates Nedd4-2, or both. To examine the former, HEK
293T were transfected with histidine-tagged ubiquitin (His-Ub), Flag-Rnf11 (-WT or -YA), in
the presence, or absence, of either V5-Smurf2, V5-Nedd4-1, or V5-Nedd4-2. Ubiquitinated proteins were precipitated using Ni-NTA agarose beads which bind to the histidine-tag on ubiquitin, and ubiquitinated Rnf11 was detected by immunoblotting with anti-Flag antibodies. In the absence of overexpressed Nedd4, ubiquitinated Rnf11 appears as two distinct bands, with
sizes corresponding to mono- and di-ubiquitinated Rnf11 (Fig. 4-9A), representing either
monoubiquitination at two lysines (multi-monoubiquitination) or di-ubiquitination at a single lysine residue. The PY-motif of Rnf11 does not appear to be required for this observed ubiquitination, as Flag-Rnf11-YA also displays the same pattern of ubiquitination. Clearly,
Nedd4-2 is playing a role in Rnf11 ubiquitination, as coexpression of V5-Nedd4-2 with Rnf11 dramatically increased the amount of ubiquitinated Rnf11 (Fig. 4-9A). Here we also observed an additional Rnf11 band, representing the addition of three ubiquitin moieties to Rnf11 (Fig. 4-
9A). Nedd4-2-mediated Rnf11 ubiquitination is further supported by the observation that overexpression of a catalytically inactive Nedd4-2 (V5-Nedd4-2 (CS)) is not able to promote
Rnf11 ubiquitination (Fig. 4-9B).
Surprisingly, although no stable interaction was detected between Rnf11-YA and
Nedd4-2 (Fig. 4-6B), the Rnf11 PY-motif was not required for Nedd4-2 mediated ubiquitination, as Flag-Rnf11-YA also displays increased ubiquitination upon Nedd4-2 overexpression, although slightly less than what is observed for Flag-Rnf11-WT (Fig. 4-9A). On the other hand,
144 Figure 4-9. Nedd4-2 overexpression increases Rnf11 ubiquitination. (A) HEK 293T cells were transfected with Flag-Rnf11 (-WT or -YA), histidine-tagged ubiquitin (His-Ub), with or without coexpression of V5-Smurf2, V5-Nedd4-1 or V5-Nedd4-2. Ubiquitinated proteins were precipitated using Ni-NTA agarose beads and ubiquitinated Rnf11 was detected by immunoblotting with anti-Flag antibodies (top panel). Cell lysates were blotted with anti-V5 and anti-Flag antibodies to detect amounts of Smurf2, Nedd4-1 or -2 and Rnf11, respectively, in transfected lysates (bottom panels). (B) Same as in A, except Flag-Rnf11 was cotransfected with
V5-Nedd4-2 (WT) or a catalytically inactive Nedd4-2 (V5-Nedd4-2(CS)). Black line separates lanes from different gels exposed on the same film.
145 Figure 4-9
A.
B.
146 Chapter 4
despite an interaction between Rnf11 and Smurf2, overexpression of Smurf2 did not increase
Rnf11 ubiquitination. Although in Fig. 4-9A, a band corresponding to tri-ubiquitination of
Rnf11 is observed upon co-expression of Smurf2 with Rnf11-WT, this weak band was not
observed in all experiments, and may not represent a specific effect of Smurf2 overexpression.
Furthermore, overexpression of Nedd4-1 did not affect Rnf11 ubiquitination, as a similar amount
of ubiquitinated Rnf11 was observed in the absence or presence of overexpressed Nedd4-1 (Fig.
4-9A), consistent with our binding studies showing little binding between Nedd4-1 and Rnf11
(Fig. 4-6A,B).
As mentioned, Rnf11 contains a putative UIM (Fig. 4-1), however we were unable to
detect binding of Rnf11 to monoubiquitin or to polyubiquitin chains linked through either K48 or
K63 (not shown). It is possible that either this motif does not bind ubiquitin, as previously observed for one of the UIMs of Eps15 (Polo et al. 2002), or that the method we used (GST- pulldown) was not sensitive enough to detect binding, as the affinity of UIMs for ubiquitin is generally quite low, with Kds in the range of 100-400 µM (Hicke et al. 2005).
Several UIM containing proteins, including Eps15, Epsin, and Hrs, are monoubiquitinated in a UIM-dependent manner (Polo et al. 2002). In each case, the UIM is thought to recruit the ubiquitination machinery, by interacting with ubiquitinated E3 ligases through their UIMs. In accord, ubiquitination of Nedd4 proteins correlates with their ability to monoubiquitinate the UIM-containing protein Eps15 (Woelk et al. 2006). The ability of the
Rnf11 PY-mutant to become ubiquitinated in a Nedd4-2-dependent manner suggests these proteins may interact independent of the PY-motif, although likely only transiently, as no stable interaction was detected between Nedd4-2 and Rnf11-YA (Fig. 4-6B). As noted, we were not able to detect ubiquitin binding by the Rnf11 UIM. However, a UIM mutation affecting the
147 Chapter 4 ability of Rnf11 to become ubiquitinated would strongly suggest ubiquitin binding by this motif in vivo. We thus created an Rnf11 UIM-mutant (UIMm), in which the UIM N-terminal glutamic acid residues (EEE), making up part of the UIM-consensus motif, were mutated to alanine
(AAA), since this mutation was shown to severely reduce ubiquitin-binding by the Hrs UIM
(Fisher et al. 2003). This mutant was not affected in its ability to bind Nedd4-2, as seen in Fig.
4-10A. To determine the effect of this mutation on Rnf11 ubiquitination, HEK 293T were transfected with His-Ub, Flag-Rnf11 (-WT, -YA, or -UIMm), with or without cotransfection of
V5-Nedd4-2, and ubiquitinated Rnf11 was detected as above. In Fig. 4-10B we can see that while both Flag-Rnf11-YA and -UIMm display an increase in ubiquitination when coexpressed with V5-Nedd4-2, the extent of ubiquitination is significantly reduced compared to Flag-Rnf11-
WT. However, in this experiment, the total amount of Rnf11 expressed in each of the cell lysates was not the same (Fig. 4-10B, middle panel), preventing a direct comparison of Rnf11 ubiquitination levels. To compensate for this, we determined the relative amount of ubiquitinated Rnf11 by quantitating the amount of ubiquitinated Rnf11 and dividing it by the total amount of Rnf11 (ubiquitinated and non-ubiquitinated) (procedure described in
Experimental Methods). From this quantitation, shown in Fig. 4-10C, we can clearly see that mutation of the PY-motif or of the UIM results in a decrease in relative Rnf11 ubiquitination.
Collectively, this data suggests that Nedd4-2 mediates, mainly, mono- and di-ubiquitination of
Rnf11, and that both the PY-motif and UIM are involved in mediating Rnf11 ubiquitination.
ii) Effect of Rnf11 on Nedd4-2 ubiquitination
As discussed in Chapter 3, Nedd4-2 is capable of self-ubiquitination, as determined by the inability of catalytically-inactive Nedd4-2 (Nedd4-2 CS) to become ubiquitinated. To determine
148 Figure 4-10. Rnf11 ubiquitination is dependent on its UIM and PY motif. (A) Flag-Rnf11
(-WT, -YA or -UIMm) was immunoprecipitated from HEK 293T lysates cotransfected with V5-
Nedd4-2. Binding of Nedd4-2 was determined by immunoblotting with anti-V5 antibodies (top
panel) and immunoprecipitation of Flag-Rnf11 was verified by anti-Flag antibodies (second
panel). Cell lysates were blotted with anti-V5 and anti-Flag antibodies to detect amounts of
transfected Nedd4-2 and Rnf11, respectively (bottom two panels). (B) Ubiquitinated proteins
were precipitated, using Ni-NTA agarose beads, from HEK 293T lysates transfected with Flag-
Rnf11 (-WT, -YA, or -UIMm), histidine-tagged ubiquitin (His-Ub), with (+) or without (-) coexpression V5-Nedd4-2. Ubiquitinated Rnf11 was detected by immunoblotting with anti-Flag antibodies (top panel). Cell lysates were blotted with anti-Flag and anti-V5 antibodies to detect amounts of Rnf11 and Nedd4-2, respectively, in transfected lysates (bottom panels). Black lines in A and B separate non-adjacent lanes from the same, or different, gels exposed on the same sheet of film. (C) Quantitation of experiment represented in B. The amount of ubiquitinated
Rnf11 and non-ubiquitinated Rnf11 (in cell lysates) was quantified from the X-ray film. The relative Rnf11 ubiquitination was determined by dividing the amount of ubiquitinated Rnf11 by the total amount of Rnf11 (ubiquitinated + non-ubiquitinated), relative to Rnf11-WT transfected alone. The light and dark grey bars represent the relative Rnf11 ubiquitination when transfected alone or when co-transfected with Nedd4-2, respectively. Results are means + S.D. (n=2).
149 Figure 4-10
A. B.
C.
9.0 8.0 7.0 6.0
ubiquitination 5.0
1 4.0
Rnf1 3.0
2.0 1.0
Relative 0.0 Rnf11‐WT Rnf11‐YA Rnf11‐UIMm
Rnf11 alone Rnf11 with Nedd‐2
150 Chapter 4
whether Rnf11 also ubiquitinates Nedd4-2, ubiquitinated proteins were precipitated, using Ni-
NTA agarose beads, from HEK 293T lysates transfected with His-Ub and V5-Nedd4-2, with or without Flag-Rnf11-WT or -YA. Interestingly, detection of ubiquitinated Nedd4-2 with anti-V5 antibodies revealed a Rnf11-mediated decrease in Nedd4-2 self-ubiquitination (Fig. 4-11).
Moreover, this decrease in ubiquitinated Nedd4-2 requires a direct interaction between Rnf11 and Nedd4-2, as evidenced by the inability of Flag-Rnf11-YA to exert the same effect (Fig. 4-
11).
V) Discussion
While the Rnf11 literature to date has focused on its interaction with Smurf2 and its potential role in antagonizing Smurf2-mediated inhibition of TGFβ signalling (reviewed in (Azmi and
Seth 2005)), here we have clearly demonstrated a role for Rnf11/Nedd4-2 interaction in regulating ubiquitination of both proteins. The implications of this and other possible functions for the interaction between Rnf11 and Nedd4-2 are outlined below.
It has been suggested that Rnf11 functions as part of a multi-subunit E3, owing to its ability to catalyze ubiquitination in the presence of the Cullin family protein Cul1 (Azmi and
Seth 2005). However, here we demonstrate that Rnf11 also possesses intrinsic catalytic activity, exhibiting ubiquitin ligase activity in the absence of any accessory proteins, in the presence of its appropriate E2 (Fig. 4-7). Thus it is possible that Rnf11 may possess the unique ability to
function as both a single-subunit E3 and as part of a multi-subunit E3 complex. Notably, while
Fig. 4-7 demonstrates the ability of Rnf11 to self-ubiquitinate in vitro, the observed
ubiquitination of Rnf11 in cells is not self-mediated, as a catalytically-inactive Rnf11 displays
the same pattern of ubiquitination to that of Rnf11-WT (not shown).
151 Figure 4-11. Rnf11 decreases Nedd4-2 ubiquitination. Ubiquitinated proteins were precipitated, using Ni-NTA agarose beads, from HEK 293T lysates transfected with histidine- tagged ubiquitin (His-Ub), and V5-Nedd4-2, in the absence or presence of coexpressed Flag-
Rnf11 (-WT or -YA). Ubiquitinated Nedd4-2 was detected by immunoblotting with anti-V5 antibodies (top panel). Cell lysates were blotted with anti-V5 and anti-Flag antibodies to detect amounts of transfected Nedd4-2 and Rnf11, respectively (bottom panels).
152 Figure 4-11
153 Chapter 4
Recent studies have also described interactions between HECT and RING-finger E3s,
with Nedd4, and its relative AIP4/Itch, reported to bind to the Cbl family of RING E3 ligases, c-
Cbl, Cbl-b, and Cbl-c (Courbard et al. 2002; Magnifico et al. 2003). The Cbl RING E3 ligases, involved in ubiquitination and downregulation of the epidermal growth factor receptor (EGFR), are ubiquitinated by Nedd4 and targeted for proteasomal degradation, reversing Cbl’s effects on
EGFR (Magnifico et al. 2003). Additionally, AIP4/Itch binds and ubiquitinates the RING-H2 protein Deltex, targeting it for lysosomal degradation (Chastagner et al. 2006). However, the
purpose of the Nedd4-2/Rnf11 interaction is likely not for Nedd4-2 to target Rnf11 for
degradation, as the pattern of ubiquitinated Rnf11 is suggestive of mono-, di-, and tri-
ubiquitination (Fig. 4-9), which generally does not serve as a signal for degradation. Rather, monoubiquitination has been implicated in bringing about changes in enzymatic activity
(Haglund and Dikic 2005), binding properties (Chen et al. 2003; Chen and De Camilli 2005), intracellular localization (Hoeller et al. 2006), and in the case of proteins containing ubiquitin- binding domains (UBDs), may function to inhibit their ubiquitin-binding capacity through formation of intramolecular ubiquitin-UBD interactions (Hoeller et al. 2006).
Markedly, Nedd4-2 mediated ubiquitination of Rnf11 does not require a stable interaction between Nedd4-2 and Rnf11, as the Rnf11 PY-mutant (Rnf11-YA), which we demonstrated does not bind Nedd4-2 (Fig. 4-6B), is efficiently ubiquitinated by Nedd4-2 (Fig. 4-9A). However,
UIM-containing proteins have been shown to interact transiently with ubiquitinated E3s through their UIMs (Woelk et al. 2006), and the decrease in ubiquitination observed for the Rnf11 UIM mutant (Rnf11-UIMm) suggests this may also be the case for Rnf11 (Fig. 4-10B,C). Taken
together, this data suggest that Rnf11 interacts with Nedd4-2 in two ways; through its PY-motif,
which binds to Nedd4-2 WW domains, and through its UIM, which binds ubiquitinated Nedd4-2,
154 Chapter 4
with both interactions resulting in Rnf11 ubiquitination. A Rnf11 mutant in which both the PY- motif and UIM are mutated was not tested, but should be used to determine if these are the only factors contributing to Rnf11 ubiquitination.
Despite the fact that an interaction between Rnf11 and Smurf2 was detected (Fig. 4-6C),
Smurf2 does not appear promote Rnf11 ubiquitination, as was observed for Nedd4-2 (Fig. 4-9A).
In fact, the overall amount of ubiquitinated Rnf11 decreases in the presence of overexpressed
Smurf2, possibly due to the ability of Smurf2 to bind the E2 required for Rnf11 ubiquitination and sequester it from its natural ligase, likely Nedd4-2. Therefore, if Smurf2 is in fact a true in vivo binding partner of Nedd4-2, the purpose of this interaction is likely not for Smurf2 to act as a ubiquitin ligase for Rnf11. Alternatively, Smurf2 may ubiquitinate Rnf11 only under certain cellular conditions.
The ability of both Nedd4-1 and Smurf1 to preferentially interact with catalytically- inactive Rnf11 is indicative of Rnf11 regulating the stability of these proteins (Fig. 4-6B,C), as interaction of Nedd4-1/Smurf1 with Rnf11-WT may lead to their degradation, although this possibility was not investigated here.
In Chapter 3, the ability of Nedd4-2 to undergo self-ubiquitination was established, as was the inability of catalytically-inactive Nedd4-2 to do the same. Furthermore, we showed evidence suggesting that binding of Nedd4-2 to its substrate ENaC enhanced Nedd4-2 self- ubiquitination, due to disruption of an inhibitory intramolecular interaction. Here, we demonstrate that interaction between Rnf11 and Nedd4-2 results in a decrease in Nedd4-2 mediated self-ubiquitination (Fig. 4-11). Interestingly, while Nedd4-2 mediated ubiquitination of Rnf11 does not require a stable interaction between the two proteins, the ability of Rnf11 to decrease Nedd4-2 self-ubiquitination does, as Rnf11-YA was not able to exert the same effect
155 Chapter 4
(Fig. 4-11). There are several possible explanations for the Rnf11 mediated decrease in Nedd4-2
ubiquitination. Since both Nedd4-2 and Rnf11 are capable of binding the E2 UbcH5b,
interaction of Rnf11 with Nedd4-2 may sequester the E2 from the Nedd4-2 HECT domain,
preventing Nedd4-2 from ubiquitinating itself. Alternatively, this finding may be linked to
Rnf11’s reported association with the deubiquitinating (DUB) enzyme AMSH (associated
molecule with SH3 domain of STAM)(Li and Seth 2004). Noticeably, AMSH exhibits
specificity for K63- over K48-linked polyubiquitin chains (McCullough et al. 2006), and
therefore a role for AMSH in Nedd4-2 deubiquitination, through interaction with Rnf11, is consistent with our results indicating that Nedd4-2 preferentially generates ubiquitin chains with
K63-linkages (Fig. 4-8B). In either case, Rnf11 likely acts to stabilize Nedd4-2. In addition, as ubiquitination of Nedd4 family members may be required for binding, and subsequent monoubiquitination, of some endocytic adaptor proteins containing UBDs (Hoeller et al. 2006), deubiquitination of Nedd4-2 could alter its network of protein interactions, regulating receptor endocytosis and protein sorting.
Our results suggest that Nedd4-2 preferentially catalyzes the formation the K63-linked polyubiquitin chains, at least in vitro (Fig. 4-8B), and is in line with other reports demonstrating that both Nedd4-1 and the yeast Nedd4 homologue Rsp5 also generate K63-linked chains (Galan and Haguenauer-Tsapis 1997; Kim et al. 2007). Different ubiquitin chain linkages are linked to
distinct physiological functions in cells. While K48-linked ubiquitin chains target proteins for proteasomal degradation, K63-linked chains are generally thought to have non-proteolytic roles in endocytosis and DNA repair, although K63-linked chains have been reported to target some
proteins to the proteasome (Kim et al. 2007). The ability of Nedd4-2 to associate with Rnf11, an
E3 ligase catalyzing formation of K29-, K48-, and K63-linked ubiquitin chains (Fig. 4-8A),
156 Chapter 4
raises the possibility that Rnf11 and Nedd4-2 work together to alter the ubiquitin chain linkages
of Nedd4-2 substrates, ultimately altering their fate.
The observation that Rnf11 is monoubiquitinated and contains a UIM, like many
endocytic adaptor proteins (i.e. eps15, epsin), and proteins involved in protein sorting (ie. GGA1,
GGA3), and its ability to interact with these proteins (Colland et al. 2004; Li and Seth 2004), strongly suggest a role for Rnf11 in the ubiquitin signalling network regulating endocytosis and protein trafficking. This is supported by our observation that at least some Rnf11 localizes to the plasma membrane (Fig. 4-4), and by observations indicating that Rnf11 downregulates EGFR signalling (Azmi and Seth 2005). Additionally, at least 6% of the total number of characterized proteins that bound to Rnf11 are transcription factors, and as Rnf11 is reported to be localized to the nucleus under certain cellular conditions (Connor et al. 2005), Rnf11 may also play a role in regulating gene expression in the nucleus. Supportively, accumulating evidence suggests that many endocytic adaptor proteins play dual roles in endocytosis and nuclear transcriptional regulation (reviewed in (Pilecka et al. 2007)).
It is clear that Rnf11, being highly conserved throughout evolution, is an especially
interesting protein, with likely roles in protein trafficking, cell signalling and transcription.
While the work presented here provides further insights into Rnf11’s function, much work
remains to be done to unveil the diverse functions of this ubiquitin ligase.
157
CHAPTER 5
Thesis Summary and Future Directions
158 Chapter 5
Thesis Summary
This thesis aimed to decipher the molecular nature of substrate specificity of Nedd4-1 and -2
WW domains and to investigate the mechanisms through which catalytic activity of these proteins is regulated. With regards to WW domain specificity, the results presented here demonstrate the importance of amino acid residues outside the core binding residues in determining substrate selection by Nedd4-1 and -2. Furthermore, it has provided convincing evidence for a novel mode of regulation of Nedd4-2 catalytic activity, through inter- or intramolecular interactions of its domains. In addition, the preliminary evidence presented here also suggests further regulation of Nedd4-2 stability through interaction with the RING finger protein Rnf11. However, the information provided in this thesis has also raised new questions with regards to substrate specificity, regulation of catalytic activity and a possible connection between the two. These, and future experiments designed to answer these questions are outlined below.
I) Molecular determinants of WW domain specificity of Nedd4 family members
Structural studies have revealed that WW domains of Nedd4 family members adopt a common recognition mechanism for the L/PPXY core sequence. However, the prevalence of WW domain-type interactions in cellular protein-protein networks makes it difficult to understand how specificity is maintained. Although overlap in substrate specificities of Nedd4 proteins has been reported, distinct roles for Nedd4 family members exist, exemplified by the distinct phenotypes of Nedd4-1, Nedd4-2 and Nedd4-like knockout mice. Therefore the question remains, how is the potential promiscuity of interactions overcome?
159 Chapter 5
In Chapter 2, we studied the molecular nature of PY-motif binding by Nedd4-1 and -2
WW domains using the ENaC and Commissureless PY-motifs as model ligands. Taking a
mutational approach, and using intrinsic tryptophan fluorescence to measure binding affinities,
we were able to assess the contribution of both WW domain residues and PY-motif residues to
overall binding affinities. These results highlighted the importance of ligand residues outside the
core PY-motif, and WW domain residues of the β1-β2 loop, in determining substrate specificity.
Further supporting our findings are recent structural studies on Smurf2 and Drosophila
Suppressor of Deltex (Su(dx)) WW domains, in complex with their PY-motif ligands, also
demonstrating that interactions between WW domain residues of the β1-β2 loop and ligand
residues outside the core PY-motif are critical for determining substrate specificity (Chong et al.
2006; Jennings et al. 2007). The observation that WW domain residues of the β1-β2 loop exhibits more sequence variability among Class I WW domains than the PY motif binding pocket residues rationalizes their importance in determining substrate specificity.
Most of the studies to date looking at specificity of Nedd4 family WW domains have focused on individual WW domains binding PY-motif peptides in isolation. However, WW domains frequently exist in multiple copies and often interact with substrates having multiple PY motifs (i.e. Comm and ENaC), suggesting they may function by binding separate PY-motif of a single target, separate targets, or act together to increase specificity and avidity, as demonstrated with tandem SH2 and PDZ domains (Futterer et al. 1998; Ottinger et al. 1998; Kang et al. 2003;
Long et al. 2003). Interestingly, studies with Nedd4 family members, including Nedd4-1 and
Rsp5, fail to separate the function of two WW domains located in a pair just prior to the catalytic
HECT domain (Harvey et al. 1999; Wang et al. 1999; Kamynina et al. 2001b; Kamynina et al.
2001c; Snyder et al. 2001). Recently, it was shown for Su(dx) that ligand binding is regulated by
160 Chapter 5 association of tandemly arranged WW domains. Here, association of WW3 and WW4 impedes proper domain folding and prevents ligand binding by WW4, which is relieved upon binding of another ligand to WW3 (Jennings et al. 2007). The highly conserved tandem arrangement of
WW domains in Nedd4 proteins, and similar arrangements in more diverse proteins, suggests domain-domain communication may represent a universal regulatory mechanism for these domains. This cooperativity may be particularly important given the ability of isolated WW domains to bind more than one target PY-motif.
Structural studies on individual Nedd4 WW domains has significantly furthered our understanding of the molecular nature of substrate recognition by these proteins. However, the frequency with which WW domains appear in multiple copies suggests WW domain interplay may serve an important function. Therefore, future studies should focus on examining the structure of tandemly arranged Nedd4 WW domains, alone and in complex with substrate(s), which could provide valuable information regarding WW domain cooperativity.
II) Regulation of Nedd4 E3 catalytic activity
Despite their importance, little was known about regulation of E3 catalytic activity, particularly in the case of HECT E3s. Interestingly, E3 self-ubiquitination has been reported, raising the important question as to how premature self-ubiquitination and degradation of E3s is controlled.
In Chapter 3, I presented evidence for a model in which Nedd4-2 catalytic activity is regulated by a low-affinity interaction between Nedd4-2 WW domains and a PY-motif located within its
HECT domain. We demonstrated that binding of Nedd4-2 WW domains to the HECT domain disrupts HECT catalytic activity, and that mutation of the HECT domain PY-motif decreases the stability of Nedd4-2, suggesting that the interaction between Nedd4-2 WW domains and the
161 Chapter 5
HECT-PY motif stabilizes the protein by preventing Nedd4-2 self-ubiquitination. Notably, I found that this interaction affects only self-ubiquitination of Nedd4-2, and does not affect the
ability of Nedd4-2 to bind and ubiquitinate its substrate, ENaC.
In addition to the work described in Chapter 3, other reports have recently emerged
demonstrating regulation of the catalytic activity of Nedd4 family members by a variety of
mechanisms, including intramolecular interactions, phosphorylation, E2 recruitment, and through
association with accessory proteins, regulating self-ubiquitination, substrate ubiquitination or
both. The Nedd4-like protein Itch has been shown to be regulated by an intramolecular
interaction, phosphorylation and by association with a deubiquitinating enzyme (Gallagher et al.
2006; Mouchantaf et al. 2006). Other members have also been shown to be regulated by multiple mechanisms. Overall, it appears there is not one specific mode of regulation for Nedd4
family members, and it seems that regulation of each family member is specifically tailored to
the needs of the substrates they regulate. While some regulatory mechanisms likely represent
constitutive modes of regulation, such as C2-HECT interaction and the WW-HECT interaction,
others may represent substrate specific pathways, such as regulation by phosphorylation, in
which E3 phosphorylation is dependent on the activation of specific kinases.
Despite evidence presented in Chapter 3, several questions remain regarding regulation of
Nedd4-2 via the WW-HECT interaction. For example, is this an inter- or intra-molecular
interaction? What is the mechanism by which WW domain binding to the HECT domain
inhibits HECT catalytic activity? Furthermore, given the low affinity of binding of a single
Nedd4-2 WW domain for the HECT-PY motif, is there any cooperativity of WW domain
binding affecting affinity? Structural studies on a fragment of Nedd4-2 consisting of the WW domains and HECT domain could help answer these remaining questions.
162 Chapter 5
As mentioned, several Nedd4 family members are regulated by multiple mechanisms,
and it is possible that Nedd4-2 is as well. For Smurf2, regulation is achieved by an
intramolecular interaction between its C2 and HECT domain (Wiesner et al. 2007), as well as through E2 recruitment (Ogunjimi et al. 2005). In the latter case, the Smurf2 substrate Smad7 interacts with WW domains and with the HECT domain, which contains a suboptimal E2 binding pocket. Binding of the N-terminal domain of Smad7 to the HECT domain promotes E2 binding, enhancing both substrate and self-ubiquitination (Ogunjimi et al. 2005). Markedly, some of the Smurf2 HECT domain residues producing a suboptimal E2 binding pocket are conserved in Nedd4-2 and other Nedd4 family members. Therefore, substrate interaction with the HECT domain may represent a general mechanism for regulating Nedd4 family members and may explain the observation that one HECT domain cannot be substituted for another HECT
(Schwarz et al. 1998). Interestingly, this also suggests a link between substrate specificity and regulation of catalytic activity. Structural studies of full-length Nedd4, in complex with different
substrates, would provide key information on how the protein functions as a whole, potentially
regulating both substrate specificity and catalytic activity simultaneously.
III) Nedd4-2 binding protein Rnf11
Our lab previously identified Rnf11 in an expression library screen looking for binding partners of the second WW domain of Nedd4-1, however, I found that Rnf11 preferentially binds Nedd4-
2 (whose WW2 domain is almost identical to that of Nedd4-1). In Chapter 4, I set out to characterize the interaction between Nedd4-2 and its binding partner Rnf11. Here we showed that Rnf11 has E3 ligase activity, and that this activity is mediated by specific E2 enzyme availability. Given that both Rnf11 and Nedd4-2 exhibit ubiquitin ligase activity, I investigated
163 Chapter 5 whether Nedd4-2 ubiquitinates Rnf11 or vice versa. My results demonstrate that Rnf11 is a substrate for Nedd4-2 mediated ubiquitination, and suggest that its ubiquitination is dependent on both the Rnf11 PY-motif and its UIM. However, I did not observe Rnf11 mediated ubiquitination of Nedd4-2. In fact, interaction between Rnf11 and Nedd4-2 decreases Nedd4-2 self-ubiquitination through an unknown mechanism, but requires a stable interaction between both proteins.
Currently the biological function of Rnf11 is unknown. Although we have shown that
Rnf11 exhibits ubiquitin ligase activity, the cellular substrates for Rnf11 remain unknown. As mentioned in Chapter 4, a large number of Rnf11 interacting proteins have been identified, however the question of whether any of these interactors are substrates for Rnf11 needs to be explored. Finding substrates for Rnf11 mediated ubiquitination may provide further clues to unravelling its function. Moreover, many of the nearly 80 protein interactors reported by Seth and colleagues are proteins involved in endocytosis, membrane trafficking, and signal transduction (Azmi and Seth 2005). Rnf11’s domain structure and potential interactors highlight a possible role for Rnf11 in regulation of intracellular trafficking. This warrants further investigation.
While our results suggest a role for Nedd4-2 in ubiquitination of Rnf11, the purpose of this ubiquitination is unclear. Nedd4-2 mediates mono-, di-, and tri-ubiquitination of Rnf11 which generally does not serve as a signal for proteasomal degradation. However, some proteins modified in this manner are targeted for lysosomal degradation, therefore the effect of Rnf11 ubiquitination on its stability should be investigated. As Rnf11 contains a putative UIM, its ubiquitination may also serve a regulatory role, resulting in intramolecular ubiquitin-UIM interactions and inhibiting its binding to downstream targets. Alternatively, the role of Rnf11
164 Chapter 5
ubiquitination may be to provide binding sites for downstream effectors containing ubiquitin-
binding domains.
Our results also suggest that binding between Rnf11 and Nedd4-2 leads to decreased
Nedd4-2 self-ubiquitination, although the mechanism through which this occurs is unknown.
We have suggested two possibilities warranting further investigation. As the RING domain of
Rnf11 is capable of binding an E2, it is possible that binding of Rnf11 to Nedd4-2 affects E2
binding by the Nedd4-2 HECT domain, thereby preventing self-ubiquitination. This possibility
can easily be tested using Rnf11 RING domain mutants affecting E2 binding. However, this
could only regulate Nedd4-2 self-ubiquitination by a subset of E2s, given that Rnf11 does not
bind all of the same E2s as Nedd4-2. Alternatively, Rnf11’s ability to decrease Nedd4-2
ubiquitination may be related to its reported association with the deubiquitinating enzyme
AMSH. This possibility can also be tested, using RNAi directed against AMSH or by
overexpression of a catalytically-inactive AMSH. If AMSH is involved, Rnf11’s ability to decrease Nedd4-2 self-ubiquitination would be reduced or inhibited.
The question remains as to the role of the Rnf11 mediated decrease in Nedd4-2 ubiquitination. The results presented in Chapter 3 suggest that Nedd4-2 self-ubiquitination results in degradation, signifying that Rnf11 may stabilize Nedd4-2. This possibility should be tested experimentally. Also, does Rnf11 affect ubiquitination of Nedd4-2 substrates in the same manner? This is only feasible if Nedd4-2 binds both Rnf11 and its substrates at the same time, which is possible given that Rnf11 was identified as a binding partner of the second Nedd4 WW domain, and most Nedd4 substrates have been reported to bind the third and fourth WW domains.
165 Chapter 5
Overall, the work presented in this thesis expands our understanding of the molecular manner in which substrate selection is carried out by WW domains of Nedd4 proteins and provides insight into novel mechanisms employed by Nedd4-2 to regulate its catalytic activity.
166
References
Abriel, H., J. Loffing, J. F. Rebhun, J. H. Pratt, L. Schild, J. D. Horisberger, D. Rotin and O. Staub (1999). Defective regulation of the epithelial Na+ channel by Nedd4 in Liddle's syndrome. J Clin Invest 103(5): 667-73. Adams, C. M., M. G. Anderson, D. G. Motto, M. P. Price, W. A. Johnson and M. J. Welsh (1998). Ripped pocket and pickpocket, novel Drosophila DEG/ENaC subunits expressed in early development and in mechanosensory neurons. J Cell Biol 140(1): 143-52. Aravind, L. and E. V. Koonin (2000). The U box is a modified RING finger - a common domain in ubiquitination. Curr Biol 10(4): R132-4. Asher, C., A. Chigaev and H. Garty (2001). Characterization of interactions between Nedd4 and beta and gammaENaC using surface plasmon resonance. Biochem Biophys Res Commun 286(5): 1228-31. Azmi, P. and A. Seth (2005). RNF11 is a multifunctional modulator of growth factor receptor signalling and transcriptional regulation. Eur J Cancer 41(16): 2549-60. Bachmair, A., D. Finley and A. Varshavsky (1986). In vivo half-life of a protein is a function of its amino-terminal residue. Science 234(4773): 179-86. Bankaitis, V. A., L. M. Johnson and S. D. Emr (1986). Isolation of yeast mutants defective in protein targeting to the vacuole. Proc Natl Acad Sci U S A 83(23): 9075-9. Barriere, H., C. Nemes, D. Lechardeur, M. Khan-Mohammad, K. Fruh and G. L. Lukacs (2006). Molecular basis of oligoubiquitin-dependent internalization of membrane proteins in Mammalian cells. Traffic 7(3): 282-97. Bartel, B., I. Wunning and A. Varshavsky (1990). The recognition component of the N-end rule pathway. Embo J 9(10): 3179-89. Beaudenon, S. L., M. R. Huacani, G. Wang, D. P. McDonnell and J. M. Huibregtse (1999). Rsp5 ubiquitin-protein ligase mediates DNA damage-induced degradation of the large subunit of RNA polymerase II in Saccharomyces cerevisiae. Mol Cell Biol 19(10): 6972-9. Bertolaet, B. L., D. J. Clarke, M. Wolff, M. H. Watson, M. Henze, G. Divita and S. I. Reed (2001). UBA domains of DNA damage-inducible proteins interact with ubiquitin. Nat Struct Biol 8(5): 417-22. Bhalla, V., D. Daidie, H. Li, A. C. Pao, L. P. LaGrange, J. Wang, A. Vandewalle, J. D. Stockand, O. Staub and D. Pearce (2005). Serum- and glucocorticoid-regulated kinase 1 regulates ubiquitin ligase neural precursor cell-expressed, developmentally down-regulated protein 4-2 by inducing interaction with 14-3-3. Mol Endocrinol 19(12): 3073-84. Bilodeau, P. S., J. L. Urbanowski, S. C. Winistorfer and R. C. Piper (2002). The Vps27p Hse1p complex binds ubiquitin and mediates endosomal protein sorting. Nat Cell Biol 4(7): 534- 9. Blake, T. J., M. Shapiro, H. C. Morse, 3rd and W. Y. Langdon (1991). The sequences of the human and mouse c-cbl proto-oncogenes show v-cbl was generated by a large truncation encompassing a proline-rich domain and a leucine zipper-like motif. Oncogene 6(4): 653- 7. Blondel, M. O., J. Morvan, S. Dupre, D. Urban-Grimal, R. Haguenauer-Tsapis and C. Volland (2004). Direct sorting of the yeast uracil permease to the endosomal system is controlled by uracil binding and Rsp5p-dependent ubiquitylation. Mol Biol Cell 15(2): 883-95. Blot, V., F. Perugi, B. Gay, M. C. Prevost, L. Briant, F. Tangy, H. Abriel, O. Staub, M. C. Dokhelar and C. Pique (2004). Nedd4.1-mediated ubiquitination and subsequent
167
recruitment of Tsg101 ensure HTLV-1 Gag trafficking towards the multivesicular body pathway prior to virus budding. J Cell Sci 117(Pt 11): 2357-67. Bonifacino, J. S. (2004). The GGA proteins: adaptors on the move. Nat Rev Mol Cell Biol 5(1): 23-32. Bylebyl, G. R., I. Belichenko and E. S. Johnson (2003). The SUMO isopeptidase Ulp2 prevents accumulation of SUMO chains in yeast. J Biol Chem 278(45): 44113-20. Cadwell, K. and L. Coscoy (2005). Ubiquitination on nonlysine residues by a viral E3 ubiquitin ligase. Science 309(5731): 127-30. Canessa, C. M., A. M. Merillat and B. C. Rossier (1994a). Membrane topology of the epithelial sodium channel in intact cells. Am J Physiol 267(6 Pt 1): C1682-90. Canessa, C. M., L. Schild, G. Buell, B. Thorens, I. Gautschi, J. D. Horisberger and B. C. Rossier (1994b). Amiloride-sensitive epithelial Na+ channel is made of three homologous subunits. Nature 367(6462): 463-7. Chastagner, P., A. Israel and C. Brou (2006). Itch/AIP4 mediates Deltex degradation through the formation of K29-linked polyubiquitin chains. EMBO Rep 7(11): 1147-53. Chen, D., N. Kon, M. Li, W. Zhang, J. Qin and W. Gu (2005). ARF-BP1/Mule is a critical mediator of the ARF tumor suppressor. Cell 121(7): 1071-83. Chen, H. and P. De Camilli (2005). The association of epsin with ubiquitinated cargo along the endocytic pathway is negatively regulated by its interaction with clathrin. Proc Natl Acad Sci U S A 102(8): 2766-71. Chen, H., S. Polo, P. P. Di Fiore and P. V. De Camilli (2003). Rapid Ca2+-dependent decrease of protein ubiquitination at synapses. Proc Natl Acad Sci U S A 100(25): 14908-13. Chen, H. I., A. Einbond, S. J. Kwak, H. Linn, E. Koepf, S. Peterson, J. W. Kelly and M. Sudol (1997). Characterization of the WW domain of human yes-associated protein and its polyproline-containing ligands. J Biol Chem 272(27): 17070-7. Chen, L. and K. Madura (2002). Rad23 promotes the targeting of proteolytic substrates to the proteasome. Mol Cell Biol 22(13): 4902-13. Chong, P. A., H. Lin, J. L. Wrana and J. D. Forman-Kay (2006). An expanded WW domain recognition motif revealed by the interaction between Smad7 and the E3 ubiquitin ligase Smurf2. J Biol Chem 281(25): 17069-75. Christensen, D. E., P. S. Brzovic and R. E. Klevit (2007). E2-BRCA1 RING interactions dictate synthesis of mono- or specific polyubiquitin chain linkages. Nat Struct Mol Biol 14(10): 941-8. Ciechanover, A. and R. Ben-Saadon (2004). N-terminal ubiquitination: more protein substrates join in. Trends Cell Biol 14(3): 103-6. Clague, M. J. (2002). Membrane transport: a coat for ubiquitin. Curr Biol 12(15): R529-31. Clague, M. J. and S. Urbe (2003). Hrs function: viruses provide the clue. Trends Cell Biol 13(12): 603-6. Colland, F., X. Jacq, V. Trouplin, C. Mougin, C. Groizeleau, A. Hamburger, A. Meil, J. Wojcik, P. Legrain and J. M. Gauthier (2004). Functional proteomics mapping of a human signaling pathway. Genome Res 14(7): 1324-32. Connor, M. K., P. B. Azmi, V. Subramaniam, H. Li and A. Seth (2005). Molecular characterization of ring finger protein 11. Mol Cancer Res 3(8): 453-61. Courbard, J. R., F. Fiore, J. Adelaide, J. P. Borg, D. Birnbaum and V. Ollendorff (2002). Interaction between two ubiquitin-protein isopeptide ligases of different classes, CBLC and AIP4/ITCH. J Biol Chem 277(47): 45267-75.
168
Coussens, L., P. J. Parker, L. Rhee, T. L. Yang-Feng, E. Chen, M. D. Waterfield, U. Francke and A. Ullrich (1986). Multiple, distinct forms of bovine and human protein kinase C suggest diversity in cellular signaling pathways. Science 233(4766): 859-66. Crosas, B., J. Hanna, D. S. Kirkpatrick, D. P. Zhang, Y. Tone, N. A. Hathaway, C. Buecker, D. S. Leggett, M. Schmidt, R. W. King, S. P. Gygi and D. Finley (2006). Ubiquitin chains are remodeled at the proteasome by opposing ubiquitin ligase and deubiquitinating activities. Cell 127(7): 1401-13. Cyr, D. M., J. Hohfeld and C. Patterson (2002). Protein quality control: U-box-containing E3 ubiquitin ligases join the fold. Trends Biochem Sci 27(7): 368-75. Debonneville, C., S. Y. Flores, E. Kamynina, P. J. Plant, C. Tauxe, M. A. Thomas, C. Munster, A. Chraibi, J. H. Pratt, J. D. Horisberger, D. Pearce, J. Loffing and O. Staub (2001). Phosphorylation of Nedd4-2 by Sgk1 regulates epithelial Na(+) channel cell surface expression. Embo J 20(24): 7052-9. Demirov, D. G. and E. O. Freed (2004). Retrovirus budding. Virus Res 106(2): 87-102. Deshaies, R. J. (1999). SCF and Cullin/Ring H2-based ubiquitin ligases. Annu Rev Cell Dev Biol 15: 435-67. Deveraux, Q., V. Ustrell, C. Pickart and M. Rechsteiner (1994). A 26 S protease subunit that binds ubiquitin conjugates. J Biol Chem 269(10): 7059-61. Di Fiore, P. P., S. Polo and K. Hofmann (2003). When ubiquitin meets ubiquitin receptors: a signalling connection. Nat Rev Mol Cell Biol 4(6): 491-7. Di Guglielmo, G. M., C. Le Roy, A. F. Goodfellow and J. L. Wrana (2003). Distinct endocytic pathways regulate TGF-beta receptor signalling and turnover. Nat Cell Biol 5(5): 410-21. Duan, L., Y. Miura, M. Dimri, B. Majumder, I. L. Dodge, A. L. Reddi, A. Ghosh, N. Fernandes, P. Zhou, K. Mullane-Robinson, N. Rao, S. Donoghue, R. A. Rogers, D. Bowtell, M. Naramura, H. Gu, V. Band and H. Band (2003). Cbl-mediated ubiquitinylation is required for lysosomal sorting of epidermal growth factor receptor but is dispensable for endocytosis. J Biol Chem 278(31): 28950-60. Duncan, L. M., S. Piper, R. B. Dodd, M. K. Saville, C. M. Sanderson, J. P. Luzio and P. J. Lehner (2006). Lysine-63-linked ubiquitination is required for endolysosomal degradation of class I molecules. Embo J 25(8): 1635-45. Dunn, R., D. A. Klos, A. S. Adler and L. Hicke (2004). The C2 domain of the Rsp5 ubiquitin ligase binds membrane phosphoinositides and directs ubiquitination of endosomal cargo. J Cell Biol 165(1): 135-44. Dupre, S., D. Urban-Grimal and R. Haguenauer-Tsapis (2004). Ubiquitin and endocytic internalization in yeast and animal cells. Biochim Biophys Acta 1695(1-3): 89-111. Eddins, M. J., C. M. Carlile, K. M. Gomez, C. M. Pickart and C. Wolberger (2006). Mms2- Ubc13 covalently bound to ubiquitin reveals the structural basis of linkage-specific polyubiquitin chain formation. Nat Struct Mol Biol 13(10): 915-20. Eletr, Z. M., D. T. Huang, D. M. Duda, B. A. Schulman and B. Kuhlman (2005). E2 conjugating enzymes must disengage from their E1 enzymes before E3-dependent ubiquitin and ubiquitin-like transfer. Nat Struct Mol Biol 12(10): 933-4. Elsasser, S., R. R. Gali, M. Schwickart, C. N. Larsen, D. S. Leggett, B. Muller, M. T. Feng, F. Tubing, G. A. Dittmar and D. Finley (2002). Proteasome subunit Rpn1 binds ubiquitin- like protein domains. Nat Cell Biol 4(9): 725-30. Espanel, X. and M. Sudol (1999). A single point mutation in a group I WW domain shifts its specificity to that of group II WW domains. J Biol Chem 274(24): 17284-9.
169
Finley, D., S. Sadis, B. P. Monia, P. Boucher, D. J. Ecker, S. T. Crooke and V. Chau (1994). Inhibition of proteolysis and cell cycle progression in a multiubiquitination-deficient yeast mutant. Mol Cell Biol 14(8): 5501-9. Firsov, D., L. Schild, I. Gautschi, A. M. Merillat, E. Schneeberger and B. C. Rossier (1996). Cell surface expression of the epithelial Na channel and a mutant causing Liddle syndrome: a quantitative approach. Proc Natl Acad Sci U S A 93(26): 15370-5. Fisher, R. D., B. Wang, S. L. Alam, D. S. Higginson, H. Robinson, W. I. Sundquist and C. P. Hill (2003). Structure and ubiquitin binding of the ubiquitin-interacting motif. J Biol Chem 278(31): 28976-84. Fouladkou, F., T. Landry, H. Kawabe, A. Neeb, C. Lu, N. Brose, V. Stambolic and D. Rotin (2008). The ubiquitin ligase Nedd4-1 is dispensable for the regulation of PTEN stability and localization. Proc Natl Acad Sci U S A 105(25): 8585-90. Furukawa, M., Y. J. He, C. Borchers and Y. Xiong (2003). Targeting of protein ubiquitination by BTB-Cullin 3-Roc1 ubiquitin ligases. Nat Cell Biol 5(11): 1001-7. Futterer, K., J. Wong, R. A. Grucza, A. C. Chan and G. Waksman (1998). Structural basis for Syk tyrosine kinase ubiquity in signal transduction pathways revealed by the crystal structure of its regulatory SH2 domains bound to a dually phosphorylated ITAM peptide. J Mol Biol 281(3): 523-37. Galan, J. M. and R. Haguenauer-Tsapis (1997). Ubiquitin lys63 is involved in ubiquitination of a yeast plasma membrane protein. Embo J 16(19): 5847-54. Galan, J. M., V. Moreau, B. Andre, C. Volland and R. Haguenauer-Tsapis (1996). Ubiquitination mediated by the Npi1p/Rsp5p ubiquitin-protein ligase is required for endocytosis of the yeast uracil permease. J Biol Chem 271(18): 10946-52. Gallagher, E., M. Gao, Y. C. Liu and M. Karin (2006). Activation of the E3 ubiquitin ligase Itch through a phosphorylation-induced conformational change. Proc Natl Acad Sci U S A 103(6): 1717-22. Garty, H. and L. G. Palmer (1997). Epithelial sodium channels: function, structure, and regulation. Physiol Rev 77(2): 359-96. Geetha, T., J. Jiang and M. W. Wooten (2005). Lysine 63 polyubiquitination of the nerve growth factor receptor TrkA directs internalization and signaling. Mol Cell 20(2): 301-12. Ghosh, P., J. Griffith, H. J. Geuze and S. Kornfeld (2003). Mammalian GGAs act together to sort mannose 6-phosphate receptors. J Cell Biol 163(4): 755-66. Glickman, M. H. and A. Ciechanover (2002). The ubiquitin-proteasome proteolytic pathway: destruction for the sake of construction. Physiol Rev 82(2): 373-428. Goulet, C. C., K. A. Volk, C. M. Adams, L. S. Prince, J. B. Stokes and P. M. Snyder (1998). Inhibition of the epithelial Na+ channel by interaction of Nedd4 with a PY motif deleted in Liddle's syndrome. J Biol Chem 273(45): 30012-7. Govers, R., T. ten Broeke, P. van Kerkhof, A. L. Schwartz and G. J. Strous (1999). Identification of a novel ubiquitin conjugation motif, required for ligand-induced internalization of the growth hormone receptor. Embo J 18(1): 28-36. Govers, R., P. van Kerkhof, A. L. Schwartz and G. J. Strous (1998). Di-leucine-mediated internalization of ligand by a truncated growth hormone receptor is independent of the ubiquitin conjugation system. J Biol Chem 273(26): 16426-33. Grossman, S. R., M. E. Deato, C. Brignone, H. M. Chan, A. L. Kung, H. Tagami, Y. Nakatani and D. M. Livingston (2003). Polyubiquitination of p53 by a ubiquitin ligase activity of p300. Science 300(5617): 342-4.
170
Gutierrez, G. J. and Z. Ronai (2006). Ubiquitin and SUMO systems in the regulation of mitotic checkpoints. Trends Biochem Sci 31(6): 324-32. Haglund, K. and I. Dikic (2005). Ubiquitylation and cell signaling. Embo J 24(19): 3353-9. Haglund, K., S. Sigismund, S. Polo, I. Szymkiewicz, P. P. Di Fiore and I. Dikic (2003). Multiple monoubiquitination of RTKs is sufficient for their endocytosis and degradation. Nat Cell Biol 5(5): 461-6. Hansson, J. H., L. Schild, Y. Lu, T. A. Wilson, I. Gautschi, R. Shimkets, C. Nelson-Williams, B. C. Rossier and R. P. Lifton (1995). A de novo missense mutation of the beta subunit of the epithelial sodium channel causes hypertension and Liddle syndrome, identifying a proline-rich segment critical for regulation of channel activity. Proc Natl Acad Sci U S A 92(25): 11495-9. Harty, R. N., M. E. Brown, J. P. McGettigan, G. Wang, H. R. Jayakar, J. M. Huibregtse, M. A. Whitt and M. J. Schnell (2001). Rhabdoviruses and the cellular ubiquitin-proteasome system: a budding interaction. J Virol 75(22): 10623-9. Harvey, K. F., A. Dinudom, D. I. Cook and S. Kumar (2001). The Nedd4-like protein KIAA0439 is a potential regulator of the epithelial sodium channel. J Biol Chem 276(11): 8597-601. Harvey, K. F., A. Dinudom, P. Komwatana, C. N. Jolliffe, M. L. Day, G. Parasivam, D. I. Cook and S. Kumar (1999). All three WW domains of murine Nedd4 are involved in the regulation of epithelial sodium channels by intracellular Na+. J Biol Chem 274(18): 12525-30. Hawryluk, M. J., P. A. Keyel, S. K. Mishra, S. C. Watkins, J. E. Heuser and L. M. Traub (2006). Epsin 1 is a polyubiquitin-selective clathrin-associated sorting protein. Traffic 7(3): 262- 81. He, Y., L. Hicke and I. Radhakrishnan (2007). Structural basis for ubiquitin recognition by SH3 domains. J Mol Biol 373(1): 190-6. Hein, C., J. Y. Springael, C. Volland, R. Haguenauer-Tsapis and B. Andre (1995). NPl1, an essential yeast gene involved in induced degradation of Gap1 and Fur4 permeases, encodes the Rsp5 ubiquitin-protein ligase. Mol Microbiol 18(1): 77-87. Helliwell, S. B., S. Losko and C. A. Kaiser (2001). Components of a ubiquitin ligase complex specify polyubiquitination and intracellular trafficking of the general amino acid permease. J Cell Biol 153(4): 649-62. Henry, P. C., V. Kanelis, M. C. O'Brien, B. Kim, I. Gautschi, J. Forman-Kay, L. Schild and D. Rotin (2003). Affinity and specificity of interactions between Nedd4 isoforms and the epithelial Na+ channel. J Biol Chem 278(22): 20019-28. Hicke, L. and H. Riezman (1996). Ubiquitination of a yeast plasma membrane receptor signals its ligand-stimulated endocytosis. Cell 84(2): 277-87. Hicke, L., H. L. Schubert and C. P. Hill (2005). Ubiquitin-binding domains. Nat Rev Mol Cell Biol 6(8): 610-21. Hoeller, D., N. Crosetto, B. Blagoev, C. Raiborg, R. Tikkanen, S. Wagner, K. Kowanetz, R. Breitling, M. Mann, H. Stenmark and I. Dikic (2006). Regulation of ubiquitin-binding proteins by monoubiquitination. Nat Cell Biol 8(2): 163-9. Hoeller, D., C. M. Hecker, S. Wagner, V. Rogov, V. Dotsch and I. Dikic (2007). E3-independent monoubiquitination of ubiquitin-binding proteins. Mol Cell 26(6): 891-8. Hofmann, K. and P. Bucher (1996). The UBA domain: a sequence motif present in multiple enzyme classes of the ubiquitination pathway. Trends Biochem Sci 21(5): 172-3.
171
Hofmann, R. M. and C. M. Pickart (2001). In vitro assembly and recognition of Lys-63 polyubiquitin chains. J Biol Chem 276(30): 27936-43. Huang, L., E. Kinnucan, G. Wang, S. Beaudenon, P. M. Howley, J. M. Huibregtse and N. P. Pavletich (1999). Structure of an E6AP-UbcH7 complex: insights into ubiquitination by the E2-E3 enzyme cascade. Science 286(5443): 1321-6. Huang, X., F. Poy, R. Zhang, A. Joachimiak, M. Sudol and M. J. Eck (2000). Structure of a WW domain containing fragment of dystrophin in complex with beta-dystroglycan. Nat Struct Biol 7(8): 634-8. Huibregtse, J. M., M. Scheffner, S. Beaudenon and P. M. Howley (1995). A family of proteins structurally and functionally related to the E6-AP ubiquitin-protein ligase. Proc Natl Acad Sci U S A 92(7): 2563-7. Husnjak, K., S. Elsasser, N. Zhang, X. Chen, L. Randles, Y. Shi, K. Hofmann, K. J. Walters, D. Finley and I. Dikic (2008). Proteasome subunit Rpn13 is a novel ubiquitin receptor. Nature 453(7194): 481-8. Ichimura, T., H. Yamamura, K. Sasamoto, Y. Tominaga, M. Taoka, K. Kakiuchi, T. Shinkawa, N. Takahashi, S. Shimada and T. Isobe (2005). 14-3-3 proteins modulate the expression of epithelial Na+ channels by phosphorylation-dependent interaction with Nedd4-2 ubiquitin ligase. J Biol Chem 280(13): 13187-94. Imai, Y., M. Soda, S. Hatakeyama, T. Akagi, T. Hashikawa, K. I. Nakayama and R. Takahashi (2002). CHIP is associated with Parkin, a gene responsible for familial Parkinson's disease, and enhances its ubiquitin ligase activity. Mol Cell 10(1): 55-67. Ing, B., A. Shteiman-Kotler, M. Castelli, P. Henry, Y. Pak, B. Stewart, G. L. Boulianne and D. Rotin (2007). Regulation of Commissureless by the ubiquitin ligase DNedd4 is required for neuromuscular synaptogenesis in Drosophila melanogaster. Mol Cell Biol 27(2): 481- 96. Ingham, R. J., G. Gish and T. Pawson (2004). The Nedd4 family of E3 ubiquitin ligases: functional diversity within a common modular architecture. Oncogene 23(11): 1972-84. Itani, O. A., J. R. Campbell, J. Herrero, P. M. Snyder and C. P. Thomas (2003). Alternate promoters and variable splicing lead to hNedd4-2 isoforms with a C2 domain and varying number of WW domains. Am J Physiol Renal Physiol 285(5): F916-29. Jasti, J., H. Furukawa, E. B. Gonzales and E. Gouaux (2007). Structure of acid-sensing ion channel 1 at 1.9 A resolution and low pH. Nature 449(7160): 316-23. Jennings, M. D., R. T. Blankley, M. Baron, A. P. Golovanov and J. M. Avis (2007). Specificity and autoregulation of Notch binding by tandem WW domains in suppressor of Deltex. J Biol Chem 282(39): 29032-42. Joazeiro, C. A., S. S. Wing, H. Huang, J. D. Leverson, T. Hunter and Y. C. Liu (1999). The tyrosine kinase negative regulator c-Cbl as a RING-type, E2-dependent ubiquitin-protein ligase. Science 286(5438): 309-12. Jolliffe, C. N., K. F. Harvey, B. P. Haines, G. Parasivam and S. Kumar (2000). Identification of multiple proteins expressed in murine embryos as binding partners for the WW domains of the ubiquitin-protein ligase Nedd4. Biochem J 351 Pt 3: 557-65. Kamynina, E., C. Debonneville, M. Bens, A. Vandewalle and O. Staub (2001a). A novel mouse Nedd4 protein suppresses the activity of the epithelial Na+ channel. Faseb J 15(1): 204- 214.
172
Kamynina, E., C. Debonneville, R. P. Hirt and O. Staub (2001b). Liddle's syndrome: a novel mouse Nedd4 isoform regulates the activity of the epithelial Na(+) channel. Kidney Int 60(2): 466-71. Kamynina, E., C. Tauxe and O. Staub (2001c). Distinct characteristics of two human Nedd4 proteins with respect to epithelial Na(+) channel regulation. Am J Physiol Renal Physiol 281(3): F469-77. Kanelis, V., M. C. Bruce, N. R. Skrynnikov, D. Rotin and J. D. Forman-Kay (2006). Structural determinants for high-affinity binding in a Nedd4 WW3* domain-Comm PY motif complex. Structure 14(3): 543-53. Kanelis, V., D. Rotin and J. D. Forman-Kay (2001). Solution structure of a Nedd4 WW domain- ENaC peptide complex. Nat Struct Biol 8(5): 407-12. Kang, B. S., D. R. Cooper, F. Jelen, Y. Devedjiev, U. Derewenda, Z. Dauter, J. Otlewski and Z. S. Derewenda (2003). PDZ tandem of human syntenin: crystal structure and functional properties. Structure 11(4): 459-68. Kaplun, L., R. Tzirkin, A. Bakhrat, N. Shabek, Y. Ivantsiv and D. Raveh (2005). The DNA damage-inducible UbL-UbA protein Ddi1 participates in Mec1-mediated degradation of Ho endonuclease. Mol Cell Biol 25(13): 5355-62. Kasanov, J., G. Pirozzi, A. J. Uveges and B. K. Kay (2001). Characterizing Class I WW domains defines key specificity determinants and generates mutant domains with novel specificities. Chem Biol 8(3): 231-41. Kato, Y., K. Nagata, M. Takahashi, L. Lian, J. J. Herrero, M. Sudol and M. Tanokura (2004). Common mechanism of ligand recognition by group II/III WW domains: redefining their functional classification. J Biol Chem 279(30): 31833-41. Kavsak, P., R. K. Rasmussen, C. G. Causing, S. Bonni, H. Zhu, G. H. Thomsen and J. L. Wrana (2000). Smad7 binds to Smurf2 to form an E3 ubiquitin ligase that targets the TGF beta receptor for degradation. Mol Cell 6(6): 1365-75. Kee, Y., N. Lyon and J. M. Huibregtse (2005). The Rsp5 ubiquitin ligase is coupled to and antagonized by the Ubp2 deubiquitinating enzyme. Embo J 24(13): 2414-24. Keleman, K., S. Rajagopalan, D. Cleppien, D. Teis, K. Paiha, L. A. Huber, G. M. Technau and B. J. Dickson (2002). Comm sorts robo to control axon guidance at the Drosophila midline. Cell 110(4): 415-27. Kerscher, O., R. Felberbaum and M. Hochstrasser (2006). Modification of proteins by ubiquitin and ubiquitin-like proteins. Annu Rev Cell Dev Biol 22: 159-80. Kidd, T., K. Brose, K. J. Mitchell, R. D. Fetter, M. Tessier-Lavigne, C. S. Goodman and G. Tear (1998). Roundabout controls axon crossing of the CNS midline and defines a novel subfamily of evolutionarily conserved guidance receptors. Cell 92(2): 205-15. Kim, H. T., K. P. Kim, F. Lledias, A. F. Kisselev, K. M. Scaglione, D. Skowyra, S. P. Gygi and A. L. Goldberg (2007). Certain pairs of ubiquitin-conjugating enzymes (E2s) and ubiquitin-protein ligases (E3s) synthesize nondegradable forked ubiquitin chains containing all possible isopeptide linkages. J Biol Chem 282(24): 17375-86. Kitching, R., M. J. Wong, D. Koehler, A. M. Burger, G. Landberg, G. Gish and A. Seth (2003). The RING-H2 protein RNF11 is differentially expressed in breast tumours and interacts with HECT-type E3 ligases. Biochim Biophys Acta 1639(2): 104-12. Knopf, J. L., M. H. Lee, L. A. Sultzman, R. W. Kriz, C. R. Loomis, R. M. Hewick and R. M. Bell (1986). Cloning and expression of multiple protein kinase C cDNAs. Cell 46(4): 491-502.
173
Koegl, M., T. Hoppe, S. Schlenker, H. D. Ulrich, T. U. Mayer and S. Jentsch (1999). A novel ubiquitination factor, E4, is involved in multiubiquitin chain assembly. Cell 96(5): 635- 44. Kolling, R. and C. P. Hollenberg (1994). The ABC-transporter Ste6 accumulates in the plasma membrane in a ubiquitinated form in endocytosis mutants. Embo J 13(14): 3261-71. Kumar, S., Y. Tomooka and M. Noda (1992). Identification of a set of genes with developmentally down-regulated expression in the mouse brain. Biochem Biophys Res Commun 185(3): 1155-61. Lam, Y. A., W. Xu, G. N. DeMartino and R. E. Cohen (1997). Editing of ubiquitin conjugates by an isopeptidase in the 26S proteasome. Nature 385(6618): 737-40. Laskowski, R. A., J. A. Rullmannn, M. W. MacArthur, R. Kaptein and J. M. Thornton (1996). AQUA and PROCHECK-NMR: programs for checking the quality of protein structures solved by NMR. J Biomol NMR 8(4): 477-86. Levkowitz, G., H. Waterman, S. A. Ettenberg, M. Katz, A. Y. Tsygankov, I. Alroy, S. Lavi, K. Iwai, Y. Reiss, A. Ciechanover, S. Lipkowitz and Y. Yarden (1999). Ubiquitin ligase activity and tyrosine phosphorylation underlie suppression of growth factor signaling by c-Cbl/Sli-1. Mol Cell 4(6): 1029-40. Levkowitz, G., H. Waterman, E. Zamir, Z. Kam, S. Oved, W. Y. Langdon, L. Beguinot, B. Geiger and Y. Yarden (1998). c-Cbl/Sli-1 regulates endocytic sorting and ubiquitination of the epidermal growth factor receptor. Genes Dev 12(23): 3663-74. Levy, L. and C. S. Hill (2006). Alterations in components of the TGF-beta superfamily signaling pathways in human cancer. Cytokine Growth Factor Rev 17(1-2): 41-58. Li, H. and A. Seth (2004). An RNF11: Smurf2 complex mediates ubiquitination of the AMSH protein. Oncogene 23(10): 1801-8. Li, M., C. L. Brooks, F. Wu-Baer, D. Chen, R. Baer and W. Gu (2003). Mono- versus polyubiquitination: differential control of p53 fate by Mdm2. Science 302(5652): 1972-5. Li, W., D. Tu, A. T. Brunger and Y. Ye (2007). A ubiquitin ligase transfers preformed polyubiquitin chains from a conjugating enzyme to a substrate. Nature 446(7133): 333-7. Lifton, R. P., A. G. Gharavi and D. S. Geller (2001). Molecular mechanisms of human hypertension. Cell 104(4): 545-56. Lim, W. A., F. M. Richards and R. O. Fox (1994). Structural determinants of peptide-binding orientation and of sequence specificity in SH3 domains. Nature 372(6504): 375-9. Liu, J., M. Furukawa, T. Matsumoto and Y. Xiong (2002). NEDD8 modification of CUL1 dissociates p120(CAND1), an inhibitor of CUL1-SKP1 binding and SCF ligases. Mol Cell 10(6): 1511-8. Long, J. F., H. Tochio, P. Wang, J. S. Fan, C. Sala, M. Niethammer, M. Sheng and M. Zhang (2003). Supramodular structure and synergistic target binding of the N-terminal tandem PDZ domains of PSD-95. J Mol Biol 327(1): 203-14. Lorick, K. L., J. P. Jensen, S. Fang, A. M. Ong, S. Hatakeyama and A. M. Weissman (1999). RING fingers mediate ubiquitin-conjugating enzyme (E2)-dependent ubiquitination. Proc Natl Acad Sci U S A 96(20): 11364-9. Lott, J. S., S. J. Coddington-Lawson, P. H. Teesdale-Spittle and F. J. McDonald (2002). A single WW domain is the predominant mediator of the interaction between the human ubiquitin- protein ligase Nedd4 and the human epithelial sodium channel. Biochem J 361(Pt 3): 481-8.
174
Lu, C., S. Pribanic, A. Debonneville, C. Jiang and D. Rotin (2007). The PY motif of ENaC, mutated in Liddle syndrome, regulates channel internalization, sorting and mobilization from subapical pool. Traffic 8(9): 1246-64. Macias, M. J., V. Gervais, C. Civera and H. Oschkinat (2000). Structural analysis of WW domains and design of a WW prototype. Nat Struct Biol 7(5): 375-9. Magnifico, A., S. Ettenberg, C. Yang, J. Mariano, S. Tiwari, S. Fang, S. Lipkowitz and A. M. Weissman (2003). WW domain HECT E3s target Cbl RING finger E3s for proteasomal degradation. J Biol Chem 278(44): 43169-77. Malakhova, O. A. and D. E. Zhang (2008). ISG15 inhibits Nedd4 ubiquitin E3 activity and enhances the innate antiviral response. J Biol Chem 283(14): 8783-7. Marchese, A. and J. L. Benovic (2001). Agonist-promoted ubiquitination of the G protein- coupled receptor CXCR4 mediates lysosomal sorting. J Biol Chem 276(49): 45509-12. Mastrandrea, L. D., J. You, E. G. Niles and C. M. Pickart (1999). E2/E3-mediated assembly of lysine 29-linked polyubiquitin chains. J Biol Chem 274(38): 27299-306. McCullough, J., P. E. Row, O. Lorenzo, M. Doherty, R. Beynon, M. J. Clague and S. Urbe (2006). Activation of the endosome-associated ubiquitin isopeptidase AMSH by STAM, a component of the multivesicular body-sorting machinery. Curr Biol 16(2): 160-5. Monami, G., V. Emiliozzi and A. Morrione (2008). Grb10/Nedd4-mediated multiubiquitination of the insulin-like growth factor receptor regulates receptor internalization. J Cell Physiol 216(2): 426-37. Morita, E. and W. I. Sundquist (2004). Retrovirus budding. Annu Rev Cell Dev Biol 20: 395-425. Mouchantaf, R., B. A. Azakir, P. S. McPherson, S. M. Millard, S. A. Wood and A. Angers (2006). The ubiquitin ligase itch is auto-ubiquitylated in vivo and in vitro but is protected from degradation by interacting with the deubiquitylating enzyme FAM/USP9X. J Biol Chem 281(50): 38738-47. Myat, A., P. Henry, V. McCabe, L. Flintoft, D. Rotin and G. Tear (2002). Drosophila Nedd4, a ubiquitin ligase, is recruited by Commissureless to control cell surface levels of the roundabout receptor. Neuron 35(3): 447-59. Nefsky, B. and D. Beach (1996). Pub1 acts as an E6-AP-like protein ubiquitiin ligase in the degradation of cdc25. Embo J 15(6): 1301-12. Nie, J., M. A. McGill, M. Dermer, S. E. Dho, C. D. Wolting and C. J. McGlade (2002). LNX functions as a RING type E3 ubiquitin ligase that targets the cell fate determinant Numb for ubiquitin-dependent degradation. Embo J 21(1-2): 93-102. Ogunjimi, A. A., D. J. Briant, N. Pece-Barbara, C. Le Roy, G. M. Di Guglielmo, P. Kavsak, R. K. Rasmussen, B. T. Seet, F. Sicheri and J. L. Wrana (2005). Regulation of Smurf2 ubiquitin ligase activity by anchoring the E2 to the HECT domain. Mol Cell 19(3): 297- 308. Ott, D. E., L. V. Coren, E. N. Chertova, T. D. Gagliardi and U. Schubert (2000). Ubiquitination of HIV-1 and MuLV Gag. Virology 278(1): 111-21. Ottinger, E. A., M. C. Botfield and S. E. Shoelson (1998). Tandem SH2 domains confer high specificity in tyrosine kinase signaling. J Biol Chem 273(2): 729-35. Ozkan, E., H. Yu and J. Deisenhofer (2005). Mechanistic insight into the allosteric activation of a ubiquitin-conjugating enzyme by RING-type ubiquitin ligases. Proc Natl Acad Sci U S A 102(52): 18890-5. Pace, C. N., F. Vajdos, L. Fee, G. Grimsley and T. Gray (1995). How to measure and predict the molar absorption coefficient of a protein. Protein Sci 4(11): 2411-23.
175
Pak, Y., W. K. Glowacka, M. C. Bruce, N. Pham and D. Rotin (2006). Transport of LAPTM5 to lysosomes requires association with the ubiquitin ligase Nedd4, but not LAPTM5 ubiquitination. J Cell Biol 175(4): 631-45. Pak, Y., N. Pham and D. Rotin (2002). Direct binding of the beta1 adrenergic receptor to the cyclic AMP-dependent guanine nucleotide exchange factor CNrasGEF leads to Ras activation. Mol Cell Biol 22(22): 7942-52. Peschard, P., N. Ishiyama, T. Lin, S. Lipkowitz and M. Park (2004). A conserved DpYR motif in the juxtamembrane domain of the Met receptor family forms an atypical c-Cbl/Cbl-b tyrosine kinase binding domain binding site required for suppression of oncogenic activation. J Biol Chem 279(28): 29565-71. Peters, J. M. (2006). The anaphase promoting complex/cyclosome: a machine designed to destroy. Nat Rev Mol Cell Biol 7(9): 644-56. Pham, N., I. Cheglakov, C. A. Koch, C. L. de Hoog, M. F. Moran and D. Rotin (2000). The guanine nucleotide exchange factor CNrasGEF activates ras in response to cAMP and cGMP. Curr Biol 10(9): 555-8. Pham, N. and D. Rotin (2001). Nedd4 regulates ubiquitination and stability of the guanine- nucleotide exchange factor CNrasGEF. J Biol Chem 276(50): 46995-7003. Pickart, C. M. (1997). Targeting of substrates to the 26S proteasome. Faseb J 11(13): 1055-66. Pickart, C. M. (2001). Mechanisms underlying ubiquitination. Annu Rev Biochem 70: 503-33. Pickart, C. M. and R. E. Cohen (2004). Proteasomes and their kin: proteases in the machine age. Nat Rev Mol Cell Biol 5(3): 177-87. Pilecka, I., M. Banach-Orlowska and M. Miaczynska (2007). Nuclear functions of endocytic proteins. Eur J Cell Biol 86(9): 533-47. Pires, J. R., C. Parthier, R. Aido-Machado, U. Wiedemann, L. Otte, G. Bohm, R. Rudolph and H. Oschkinat (2005). Structural basis for APPTPPPLPP peptide recognition by the FBP11WW1 domain. J Mol Biol 348(2): 399-408. Pires, J. R., F. Taha-Nejad, F. Toepert, T. Ast, U. Hoffmuller, J. Schneider-Mergener, R. Kuhne, M. J. Macias and H. Oschkinat (2001). Solution structures of the YAP65 WW domain and the variant L30 K in complex with the peptides GTPPPPYTVG, N-(n-octyl)-GPPPY and PLPPY and the application of peptide libraries reveal a minimal binding epitope. J Mol Biol 314(5): 1147-56. Plant, P. J., F. Lafont, S. Lecat, P. Verkade, K. Simons and D. Rotin (2000). Apical membrane targeting of Nedd4 is mediated by an association of its C2 domain with annexin XIIIb. J Cell Biol 149(7): 1473-84. Plant, P. J., H. Yeger, O. Staub, P. Howard and D. Rotin (1997). The C2 domain of the ubiquitin protein ligase Nedd4 mediates Ca2+-dependent plasma membrane localization. J Biol Chem 272(51): 32329-36. Polo, S., S. Sigismund, M. Faretta, M. Guidi, M. R. Capua, G. Bossi, H. Chen, P. De Camilli and P. P. Di Fiore (2002). A single motif responsible for ubiquitin recognition and monoubiquitination in endocytic proteins. Nature 416(6879): 451-5. Pornillos, O., D. S. Higginson, K. M. Stray, R. D. Fisher, J. E. Garrus, M. Payne, G. P. He, H. E. Wang, S. G. Morham and W. I. Sundquist (2003). HIV Gag mimics the Tsg101- recruiting activity of the human Hrs protein. J Cell Biol 162(3): 425-34. Puertollano, R., R. C. Aguilar, I. Gorshkova, R. J. Crouch and J. S. Bonifacino (2001). Sorting of mannose 6-phosphate receptors mediated by the GGAs. Science 292(5522): 1712-6.
176
Puertollano, R. and J. S. Bonifacino (2004). Interactions of GGA3 with the ubiquitin sorting machinery. Nat Cell Biol 6(3): 244-51. Ravid, T. and M. Hochstrasser (2007). Autoregulation of an E2 enzyme by ubiquitin-chain assembly on its catalytic residue. Nat Cell Biol 9(4): 422-7. Richly, H., M. Rape, S. Braun, S. Rumpf, C. Hoege and S. Jentsch (2005). A series of ubiquitin binding factors connects CDC48/p97 to substrate multiubiquitylation and proteasomal targeting. Cell 120(1): 73-84. Rieder, S. E., L. M. Banta, K. Kohrer, J. M. McCaffery and S. D. Emr (1996). Multilamellar endosome-like compartment accumulates in the yeast vps28 vacuolar protein sorting mutant. Mol Biol Cell 7(6): 985-99. Ritchie, K. J. and D. E. Zhang (2004). ISG15: the immunological kin of ubiquitin. Semin Cell Dev Biol 15(2): 237-46. Rodrigo-Brenni, M. C. and D. O. Morgan (2007). Sequential E2s drive polyubiquitin chain assembly on APC targets. Cell 130(1): 127-39. Rodriguez, M. S., J. M. Desterro, S. Lain, C. A. Midgley, D. P. Lane and R. T. Hay (1999). SUMO-1 modification activates the transcriptional response of p53. Embo J 18(22): 6455-61. Rossier, B. C., S. Pradervand, L. Schild and E. Hummler (2002). Epithelial sodium channel and the control of sodium balance: interaction between genetic and environmental factors. Annu Rev Physiol 64: 877-97. Saeki, Y., E. Isono and E. A. Toh (2005). Preparation of ubiquitinated substrates by the PY motif-insertion method for monitoring 26S proteasome activity. Methods Enzymol 399: 215-27. Sali, A. and T. L. Blundell (1993). Comparative protein modelling by satisfaction of spatial restraints. J Mol Biol 234(3): 779-815. Schild, L., Y. Lu, I. Gautschi, E. Schneeberger, R. P. Lifton and B. C. Rossier (1996). Identification of a PY motif in the epithelial Na channel subunits as a target sequence for mutations causing channel activation found in Liddle syndrome. Embo J 15(10): 2381-7. Schreiner, P., X. Chen, K. Husnjak, L. Randles, N. Zhang, S. Elsasser, D. Finley, I. Dikic, K. J. Walters and M. Groll (2008). Ubiquitin docking at the proteasome through a novel pleckstrin-homology domain interaction. Nature 453(7194): 548-52. Schubert, U., D. E. Ott, E. N. Chertova, R. Welker, U. Tessmer, M. F. Princiotta, J. R. Bennink, H. G. Krausslich and J. W. Yewdell (2000). Proteasome inhibition interferes with gag polyprotein processing, release, and maturation of HIV-1 and HIV-2. Proc Natl Acad Sci U S A 97(24): 13057-62. Schwarz, S. E., J. L. Rosa and M. Scheffner (1998). Characterization of human hect domain family members and their interaction with UbcH5 and UbcH7. J Biol Chem 273(20): 12148-54. Scott, P. M., P. S. Bilodeau, O. Zhdankina, S. C. Winistorfer, M. J. Hauglund, M. M. Allaman, W. R. Kearney, A. D. Robertson, A. L. Boman and R. C. Piper (2004). GGA proteins bind ubiquitin to facilitate sorting at the trans-Golgi network. Nat Cell Biol 6(3): 252-9. Seki, N., A. Hattori, A. Hayashi, S. Kozuma, M. Sasaki, Y. Suzuki, S. Sugano, M. A. Muramatsu and T. Saito (1999). Cloning and expression profile of mouse and human genes, Rnf11/RNF11, encoding a novel RING-H2 finger protein. Biochim Biophys Acta 1489(2- 3): 421-7.
177
Selvin, S. (2001). Epidemiologic analysis : a case-oriented approach. New York ; Oxford, Oxford University Press. Seron, K., M. O. Blondel, R. Haguenauer-Tsapis and C. Volland (1999). Uracil-induced down- regulation of the yeast uracil permease. J Bacteriol 181(6): 1793-800. Shekhtman, A. and D. Cowburn (2002). A ubiquitin-interacting motif from Hrs binds to and occludes the ubiquitin surface necessary for polyubiquitination in monoubiquitinated proteins. Biochem Biophys Res Commun 296(5): 1222-7. Shenoy, S. K., P. H. McDonald, T. A. Kohout and R. J. Lefkowitz (2001). Regulation of receptor fate by ubiquitination of activated beta 2-adrenergic receptor and beta-arrestin. Science 294(5545): 1307-13. Shenoy, S. K., K. Xiao, V. Venkataramanan, P. M. Snyder, N. J. Freedman and A. M. Weissman (2008). Nedd4 Mediates Agonist-dependent Ubiquitination, Lysosomal Targeting, and Degradation of the {beta}2-Adrenergic Receptor. J Biol Chem 283(32): 22166-76. Shi, P. P., X. R. Cao, E. M. Sweezer, T. S. Kinney, N. R. Williams, R. F. Husted, R. Nair, R. M. Weiss, R. A. Williamson, C. D. Sigmund, P. M. Snyder, O. Staub, J. B. Stokes and B. Yang (2008). Salt-sensitive hypertension and cardiac hypertrophy in mice deficient in the ubiquitin ligase Nedd4-2. Am J Physiol Renal Physiol 295(2): F462-70. Shimkets, R. A., D. G. Warnock, C. M. Bositis, C. Nelson-Williams, J. H. Hansson, M. Schambelan, J. R. Gill, Jr., S. Ulick, R. V. Milora, J. W. Findling and et al. (1994). Liddle's syndrome: heritable human hypertension caused by mutations in the beta subunit of the epithelial sodium channel. Cell 79(3): 407-14. Sigismund, S., T. Woelk, C. Puri, E. Maspero, C. Tacchetti, P. Transidico, P. P. Di Fiore and S. Polo (2005). Clathrin-independent endocytosis of ubiquitinated cargos. Proc Natl Acad Sci U S A 102(8): 2760-5. Snyder, P. M., D. R. Olson, F. J. McDonald and D. B. Bucher (2001). Multiple WW domains, but not the C2 domain, are required for inhibition of the epithelial Na+ channel by human Nedd4. J Biol Chem 276(30): 28321-6. Snyder, P. M., D. R. Olson and B. C. Thomas (2002). Serum and glucocorticoid-regulated kinase modulates Nedd4-2-mediated inhibition of the epithelial Na+ channel. J Biol Chem 277(1): 5-8. Snyder, P. M., M. P. Price, F. J. McDonald, C. M. Adams, K. A. Volk, B. G. Zeiher, J. B. Stokes and M. J. Welsh (1995). Mechanism by which Liddle's syndrome mutations increase activity of a human epithelial Na+ channel. Cell 83(6): 969-78. Sorkina, T., A. Bild, F. Tebar and A. Sorkin (1999). Clathrin, adaptors and eps15 in endosomes containing activated epidermal growth factor receptors. J Cell Sci 112 ( Pt 3): 317-27. Springael, J. Y. and B. Andre (1998). Nitrogen-regulated ubiquitination of the Gap1 permease of Saccharomyces cerevisiae. Mol Biol Cell 9(6): 1253-63. Stamenova, S. D., M. E. French, Y. He, S. A. Francis, Z. B. Kramer and L. Hicke (2007). Ubiquitin binds to and regulates a subset of SH3 domains. Mol Cell 25(2): 273-84. Staub, O., S. Dho, P. Henry, J. Correa, T. Ishikawa, J. McGlade and D. Rotin (1996). WW domains of Nedd4 bind to the proline-rich PY motifs in the epithelial Na+ channel deleted in Liddle's syndrome. Embo J 15(10): 2371-80. Staub, O., I. Gautschi, T. Ishikawa, K. Breitschopf, A. Ciechanover, L. Schild and D. Rotin (1997). Regulation of stability and function of the epithelial Na+ channel (ENaC) by ubiquitination. Embo J 16(21): 6325-36.
178
Staub, O. and D. Rotin (2006). Role of ubiquitylation in cellular membrane transport. Physiol Rev 86(2): 669-707. Subramaniam, V., H. Li, M. Wong, R. Kitching, L. Attisano, J. Wrana, J. Zubovits, A. M. Burger and A. Seth (2003). The RING-H2 protein RNF11 is overexpressed in breast cancer and is a target of Smurf2 E3 ligase. Br J Cancer 89(8): 1538-44. Sudol, M., H. I. Chen, C. Bougeret, A. Einbond and P. Bork (1995). Characterization of a novel protein-binding module--the WW domain. FEBS Lett 369(1): 67-71. Sudol, M. and T. Hunter (2000). NeW wrinkles for an old domain. Cell 103(7): 1001-4. Sun, L. and Z. J. Chen (2004). The novel functions of ubiquitination in signaling. Curr Opin Cell Biol 16(2): 119-26. Tamai, K. K. and C. Shimoda (2002). The novel HECT-type ubiquitin-protein ligase Pub2p shares partially overlapping function with Pub1p in Schizosaccharomyces pombe. J Cell Sci 115(Pt 9): 1847-57. Thrower, J. S., L. Hoffman, M. Rechsteiner and C. M. Pickart (2000). Recognition of the polyubiquitin proteolytic signal. Embo J 19(1): 94-102. Toepert, F., T. Knaute, S. Guffler, J. R. Pires, T. Matzdorf, H. Oschkinat and J. Schneider- Mergener (2003). Combining SPOT synthesis and native peptide ligation to create large arrays of WW protein domains. Angew Chem Int Ed Engl 42(10): 1136-40. Traub, L. M. and G. L. Lukacs (2007). Decoding ubiquitin sorting signals for clathrin-dependent endocytosis by CLASPs. J Cell Sci 120(Pt 4): 543-53. Trempe, J. F., N. R. Brown, E. D. Lowe, C. Gordon, I. D. Campbell, M. E. Noble and J. A. Endicott (2005). Mechanism of Lys48-linked polyubiquitin chain recognition by the Mud1 UBA domain. Embo J 24(18): 3178-89. Trotman, L. C., X. Wang, A. Alimonti, Z. Chen, J. Teruya-Feldstein, H. Yang, N. P. Pavletich, B. S. Carver, C. Cordon-Cardo, H. Erdjument-Bromage, P. Tempst, S. G. Chi, H. J. Kim, T. Misteli, X. Jiang and P. P. Pandolfi (2007). Ubiquitination regulates PTEN nuclear import and tumor suppression. Cell 128(1): 141-56. Tu, D., W. Li, Y. Ye and A. T. Brunger (2007). Inaugural Article: Structure and function of the yeast U-box-containing ubiquitin ligase Ufd2p. Proc Natl Acad Sci U S A 104(40): 15599-606. Vana, M. L., Y. Tang, A. Chen, G. Medina, C. Carter and J. Leis (2004). Role of Nedd4 and ubiquitination of Rous sarcoma virus Gag in budding of virus-like particles from cells. J Virol 78(24): 13943-53. Varadan, R., M. Assfalg, S. Raasi, C. Pickart and D. Fushman (2005). Structural determinants for selective recognition of a Lys48-linked polyubiquitin chain by a UBA domain. Mol Cell 18(6): 687-98. Vecchione, A., A. Marchese, P. Henry, D. Rotin and A. Morrione (2003). The Grb10/Nedd4 complex regulates ligand-induced ubiquitination and stability of the insulin-like growth factor I receptor. Mol Cell Biol 23(9): 3363-72. Verdecia, M. A., M. E. Bowman, K. P. Lu, T. Hunter and J. P. Noel (2000). Structural basis for phosphoserine-proline recognition by group IV WW domains. Nat Struct Biol 7(8): 639- 43. Verdecia, M. A., C. A. Joazeiro, N. J. Wells, J. L. Ferrer, M. E. Bowman, T. Hunter and J. P. Noel (2003). Conformational flexibility underlies ubiquitin ligation mediated by the WWP1 HECT domain E3 ligase. Mol Cell 11(1): 249-59.
179
Verma, R., R. Oania, J. Graumann and R. J. Deshaies (2004). Multiubiquitin chain receptors define a layer of substrate selectivity in the ubiquitin-proteasome system. Cell 118(1): 99- 110. Viguera, A. R., J. L. Arrondo, A. Musacchio, M. Saraste and L. Serrano (1994). Characterization of the interaction of natural proline-rich peptides with five different SH3 domains. Biochemistry 33(36): 10925-33. Vijay-Kumar, S., C. E. Bugg and W. J. Cook (1987). Structure of ubiquitin refined at 1.8 A resolution. J Mol Biol 194(3): 531-44. Walters, K. J., M. F. Kleijnen, A. M. Goh, G. Wagner and P. M. Howley (2002). Structural studies of the interaction between ubiquitin family proteins and proteasome subunit S5a. Biochemistry 41(6): 1767-77. Wang, G., J. Yang and J. M. Huibregtse (1999). Functional domains of the Rsp5 ubiquitin- protein ligase. Mol Cell Biol 19(1): 342-52. Wang, M., D. Cheng, J. Peng and C. M. Pickart (2006). Molecular determinants of polyubiquitin linkage selection by an HECT ubiquitin ligase. Embo J 25(8): 1710-9. Wang, M. and C. M. Pickart (2005). Different HECT domain ubiquitin ligases employ distinct mechanisms of polyubiquitin chain synthesis. Embo J 24(24): 4324-33. Wang, X., L. C. Trotman, T. Koppie, A. Alimonti, Z. Chen, Z. Gao, J. Wang, H. Erdjument- Bromage, P. Tempst, C. Cordon-Cardo, P. P. Pandolfi and X. Jiang (2007). NEDD4-1 is a proto-oncogenic ubiquitin ligase for PTEN. Cell 128(1): 129-39. Waterman, H., G. Levkowitz, I. Alroy and Y. Yarden (1999). The RING finger of c-Cbl mediates desensitization of the epidermal growth factor receptor. J Biol Chem 274(32): 22151-4. Weissman, A. M. (2001). Themes and variations on ubiquitylation. Nat Rev Mol Cell Biol 2(3): 169-78. Wertz, I. E., K. M. O'Rourke, H. Zhou, M. Eby, L. Aravind, S. Seshagiri, P. Wu, C. Wiesmann, R. Baker, D. L. Boone, A. Ma, E. V. Koonin and V. M. Dixit (2004). De-ubiquitination and ubiquitin ligase domains of A20 downregulate NF-kappaB signalling. Nature 430(7000): 694-9. Wiesner, S., A. A. Ogunjimi, H. R. Wang, D. Rotin, F. Sicheri, J. L. Wrana and J. D. Forman- Kay (2007). Autoinhibition of the HECT-type ubiquitin ligase Smurf2 through its C2 domain. Cell 130(4): 651-62. Wilkinson, C. R., M. Seeger, R. Hartmann-Petersen, M. Stone, M. Wallace, C. Semple and C. Gordon (2001). Proteins containing the UBA domain are able to bind to multi-ubiquitin chains. Nat Cell Biol 3(10): 939-43. Williams, R. L. and S. Urbe (2007). The emerging shape of the ESCRT machinery. Nat Rev Mol Cell Biol 8(5): 355-68. Windheim, M., M. Peggie and P. Cohen (2008). Two different classes of E2 ubiquitin- conjugating enzymes are required for the mono-ubiquitination of proteins and elongation by polyubiquitin chains with a specific topology. Biochem J 409(3): 723-9. Woelk, T., B. Oldrini, E. Maspero, S. Confalonieri, E. Cavallaro, P. P. Di Fiore and S. Polo (2006). Molecular mechanisms of coupled monoubiquitination. Nat Cell Biol 8(11): 1246-54. Wu, P. Y., M. Hanlon, M. Eddins, C. Tsui, R. S. Rogers, J. P. Jensen, M. J. Matunis, A. M. Weissman, C. Wolberger and C. M. Pickart (2003). A conserved catalytic residue in the ubiquitin-conjugating enzyme family. Embo J 22(19): 5241-50.
180
Xirodimas, D. P., M. K. Saville, J. C. Bourdon, R. T. Hay and D. P. Lane (2004). Mdm2- mediated NEDD8 conjugation of p53 inhibits its transcriptional activity. Cell 118(1): 83- 97. Xu, L., Y. Wei, J. Reboul, P. Vaglio, T. H. Shin, M. Vidal, S. J. Elledge and J. W. Harper (2003). BTB proteins are substrate-specific adaptors in an SCF-like modular ubiquitin ligase containing CUL-3. Nature 425(6955): 316-21. Yao, T., L. Song, W. Xu, G. N. DeMartino, L. Florens, S. K. Swanson, M. P. Washburn, R. C. Conaway, J. W. Conaway and R. E. Cohen (2006). Proteasome recruitment and activation of the Uch37 deubiquitinating enzyme by Adrm1. Nat Cell Biol 8(9): 994- 1002. Young, P., Q. Deveraux, R. E. Beal, C. M. Pickart and M. Rechsteiner (1998). Characterization of two polyubiquitin binding sites in the 26 S protease subunit 5a. J Biol Chem 273(10): 5461-7. Yu, H., J. K. Chen, S. Feng, D. C. Dalgarno, A. W. Brauer and S. L. Schreiber (1994). Structural basis for the binding of proline-rich peptides to SH3 domains. Cell 76(5): 933-45. Zheng, N., B. A. Schulman, L. Song, J. J. Miller, P. D. Jeffrey, P. Wang, C. Chu, D. M. Koepp, S. J. Elledge, M. Pagano, R. C. Conaway, J. W. Conaway, J. W. Harper and N. P. Pavletich (2002). Structure of the Cul1-Rbx1-Skp1-F boxSkp2 SCF ubiquitin ligase complex. Nature 416(6882): 703-9. Zheng, N., P. Wang, P. D. Jeffrey and N. P. Pavletich (2000). Structure of a c-Cbl-UbcH7 complex: RING domain function in ubiquitin-protein ligases. Cell 102(4): 533-9.
181
APPENDIX I
Interactions between the Three CIN85 SH3 Domains and Ubiquitin: Implications for CIN85 Ubiquitination
This work was published in Biochemistry:
Bezsonova, I., Bruce, M.C., Wiesner, S., Lin, H., Rotin, D. and Forman-Kay, J.D. (2008) Interactions between the three CIN85 SH3 domains and Ubiquitin: Implications for CIN85 Ubiquitination. Biochemistry 47: 8937–8949.
My contribution to this work were the CIN85 ubiquitination experiments depicted in Figure 7, as well as the construction of all mutant constructs associated with those experiments.
182 Interactions between the three CIN85 SH3 domains and ubiquitin: implications for CIN85 ubiquitination
Bezsonova I, Bruce MC, Wiesner S, Lin H, Rotin D, Forman-Kay JD.
Abstract
CIN85 is an adaptor protein linking the ubiquitin ligase Cbl and clathrin-binding proteins in clathrin-mediated receptor endocytosis. The SH3 domains of CIN85 bind to a proline-rich region of Cbl. Here we show that all three SH3 domains of CIN85 bind to ubiquitin. We also present a data-based structural model of the CIN85 SH3-C domain in complex with ubiquitin. In this complex, ubiquitin binds to the canonical interaction surface of the SH3 domain for proline-rich ligands and mimics the PPII helix, and we provide evidence that ubiquitin competes with these ligands for binding. We demonstrate that disruption of ubiquitin binding results in constitutive ubiquitination of CIN85 and an increased level of ubiquitination of EGFR in the absence of EGF stimulation. These results suggest that competition between Cbl and ubiquitin binding to CIN85 regulates Cbl function and EGFR endocytosis.
183
APPENDIX II
Transport of LAPTM5 to lysosomes requires association with the ubiquitin ligase Nedd4, but not LAPTM5 ubiquitination
This work was published in the Journal of Cell Biology:
Pak, Y., Glowacka, W.K., Bruce, M.C., Pham, N., and Rotin, D. (2006) Transport of LAPTM5 to lysosomes requires association with the ubiquitin ligase Nedd4, but not LAPTM5 ubiquitination. J. Cell Bio. 175(4): 631-645.
My contribution to this work was demonstrating Nedd4 mediated ubiquitination of GGA3, depicted in Figure 6E.
184 Transport of LAPTM5 to lysosomes requires association with the ubiquitin ligase Nedd4,
but not LAPTM5 ubiquitination
Pak Y, Glowacka WK, Bruce MC, Pham N, Rotin D.
Abstract
LAPTM5 is a lysosomal transmembrane protein expressed in immune cells. We show that
LAPTM5 binds the ubiquitin-ligase Nedd4 and GGA3 to promote LAPTM5 sorting from the
Golgi to the lysosome, an event that is independent of LAPTM5 ubiquitination. LAPTM5 contains three PY motifs (L/PPxY), which bind Nedd4-WW domains, and a ubiquitin-interacting motif (UIM) motif. The Nedd4-LAPTM5 complex recruits ubiquitinated GGA3, which binds the
LAPTM5-UIM; this interaction does not require the GGA3-GAT domain. LAPTM5 mutated in its Nedd4-binding sites (PY motifs) or its UIM is retained in the Golgi, as is LAPTM5 expressed in cells in which Nedd4 or GGA3 is knocked-down with RNAi. However, ubiquitination- impaired LAPTM5 can still traffic to the lysosome, suggesting that Nedd4 binding to LAPTM5, not LAPTM5 ubiquitination, is required for targeting. Interestingly, Nedd4 is also able to ubiquitinate GGA3. These results demonstrate a novel mechanism by which the ubiquitin-ligase
Nedd4, via interactions with GGA3 and cargo (LAPTM5), regulates cargo trafficking to the lysosome without requiring cargo ubiquitination.
185
APPENDIX III
Molecular determinants of voltage-gated sodium channel regulation by the Nedd4/Nedd4-like proteins
This work was published in the American Journal of Physiology, Cell Physiology:
Rougier, J-S., van Bemmelen, M.X., Bruce, M.C., Jespersen, T., Gavillet, B., Apotheloz, F., Cordonier, S., Staub, O., Rotin, D. and Abriel, H. (2005) Molecular determinants of voltage- gated sodium channel regulation by the Nedd4/Nedd4-like proteins. Am. J. Physiol. Cell Physiol. 288:C692-C701.
My contribution to this work was measuring of the binding affinity between Nedd4-1 and -2 WW domains and PY-motifs from the channels Nav1.5 and hERG, shown in Table 1.
186 Molecular determinants of voltage-gated sodium channel regulation by the Nedd4/Nedd4- like proteins
Rougier JS, van Bemmelen MX, Bruce MC, Jespersen T, Gavillet B, Apothéloz F, Cordonier S,
Staub O, Rotin D, Abriel H.
Abstract
The voltage-gated Na(+) channels (Na(v)) form a family composed of 10 genes. The COOH termini of Na(v) contain a cluster of amino acids that are nearly identical among 7 of the 10 members. This COOH-terminal sequence, PPSYDSV, is a PY motif known to bind to WW domains of E3 protein-ubiquitin ligases of the Nedd4 family. We recently reported that cardiac
Na(v)1.5 is regulated by Nedd4-2. In this study, we further investigated the molecular determinants of regulation of Na(v) proteins. When expressed in HEK-293 cells and studied using whole cell voltage clamping, the neuronal Na(v)1.2 and Na(v)1.3 were also downregulated by Nedd4-2. Pull-down experiments using fusion proteins bearing the PY motif of Na(v)1.2,
Na(v)1.3, and Na(v)1.5 indicated that mouse brain Nedd4-2 binds to the Na(v) PY motif. Using intrinsic tryptophan fluorescence imaging of WW domains, we found that Na(v)1.5 PY motif binds preferentially to the fourth WW domain of Nedd4-2 with a K(d) of approximately 55 µM.
We tested the binding properties and the ability to ubiquitinate and downregulate Na(v)1.5 of three Nedd4-like E3s: Nedd4-1, Nedd4-2, and WWP2. Despite the fact that along with Nedd4-2,
Nedd4-1 and WWP2 bind to Na(v)1.5 PY motif, only Nedd4-2 robustly ubiquitinated and downregulated Na(v)1.5. Interestingly, coexpression of WWP2 competed with the effect of
Nedd4-2. Finally, using brefeldin A, we found that Nedd4-2 accelerated internalization of
Na(v)1.5 stably expressed in HEK-293 cells. This study shows that Nedd4-dependent
187 ubiquitination of Na(v) channels may represent a general mechanism regulating the excitability of neurons and myocytes via modulation of channel density at the plasma membrane.
188