UMEÅ UNIVERSITY MEDICAL DISSERTATIONS

New Series No. 1739 ISSN 0346-6612 ISBN 978-91-7601-315-1

TRAF6, a key regulator of TGFβ-induced oncogenesis in prostate cancer

Reshma Sundar

Department of Medical Biosciences, Pathology Umeå University,

Umeå 2015

Responsible publisher under Swedish law: The Dean of the Medical Faculty

This work is protected by the Swedish Copyright legislation (Act 1960:729)

Copyrights © Reshma Sundar

New Series No: 1739

ISSN: 0346-6612

ISBN: 978-91-7601-315-1

E-version available at: http://umu.diva-portal.org/

Cover picture: TGFβ type I receptor nuclear translocation

Cover picture design: Harvinder Singh, ZD technologies

Cover design: Print and Media

Printed by: Print and Media, Umeå, Sweden, 2015

To my father, who gave me the greatest gift anyone could give—he believed in me.

Some people grumble that roses have thorns;

I am grateful that thorns have roses.

Abstract Prostate cancer is the most common cancer in men, with the incidence rapidly increasing in Europe over the past two decades. Reliable biomarkers for prostate cancer are currently unavailable. Thus, there is an urgent need for improved biomarkers to diagnose prostate cancer at an early stage and to determine the best treatment options. Higher expression of transforming growth factor-β (TGFβ) has been reported in patients with aggressive cancer. TGFβ is a multifunctional cytokine that acts as a tumor suppressor during early tumor development, and as a tumor promoter at later stages of cancer. TGFβ signals through the canonical Smad or non-Smad cascade via TGFβ type II and type I receptors. The TGFβ signaling cascade is regulated by various post-translational modifications of its key components. The present investigation aimed to identify a potential function of TRAF6 in TGFβ-induced responses in prostate cancer. The first two articles of this thesis unveil the proteolytic cleavage of TGFβ type I receptor (TβRI), and the biological importance of the liberated TβRI intracellular domain (TβRI-ICD) in the nucleus. We found that tumor necrosis factor receptor-associated factor 6 (TRAF6) polyubiquitinates TβRI, which leads to cleavage of TβRI by tumor necrosis factor alpha converting enzyme (TACE) in a protein kinase C zeta (PKCζ)-dependent manner. Following ectodomain shedding, TβRI undergoes a second cleavage by presenilin 1 (PS1), which liberates TβRI-ICD. TβRI-ICD translocates to the nucleus, where it regulates its own expression as well as expression of the pro-invasive gene Snail1, thereby promoting invasion. We further found that TβRI-ICD associates with Notch intracellular domain (NICD) to drive expression of the pro-invasive gene Snail1, as well as Notch1 ligand Jag1. The third article provides evidence that TRAF6 promotes Lys63-linked polyubiquitination of TβRI at Lys178 in a TGFβ-dependent manner. TβRI polyubiquitination was found to be a prerequisite for TβRI nuclear translocation, and thus for regulation of the genes involved in cell cycle, differentiation, and invasion of prostate cancer cells. In the fourth article we investigated the role of the pro-invasive gene Snail1 in TGFβ-induced epithelial-to-mesenchymal transition (EMT) in prostate cancer cells.

List of Papers This thesis is based on the following papers: 1. Yabing Mu*, Reshma Sundar*, Noopur Thakur*, Maria Ekman*, Shyam Kumar Gudey, Mariya Yakymovych, Helena Dimitrou, Annika Hermansson, Maria Teresa Bengoechea-Alonso, Johan Ericsson, Carl-Henrik Heldin, Marene Landström. TRAF6 ubiquitinates TGFβ type I receptor to promote its cleavage and nuclear translocation in cancer. Nature Communications. 2011;2:330. 2. Shyam Kumar Gudey, Reshma Sundar, Yabing Mu, Anders Wallenius, Guanxiang Zang, Anders Bergh, Carl-Henrik Heldin, Marene Landström. TRAF6 Stimulates the Tumor-Promoting Effects of TGFβ Type I Receptor Through Polyubiquitination and Activation of Presenilin 1. Science Signalling. 2014 Jan 7;7(307). 3. Reshma Sundar, Shyam Kumar Gudey, Carl-Henrik Heldin, Marene Landström. TRAF6 promotes TGFβ-induced invasion and cell-cycle regulation via Lys63-linked polyubiquitination of Lys178 in TGFβ type I receptor. Cell Cycle. 2015; 14(4):554–65 4. Shyam Kumar Gudey, Reshma Sundar, Carl-Henrik Heldin, Marene Landström. Proinvasive Snail1 targets TGFβ Type I receptor to promote epithelial-to-mesenchymal transition in prostate cancer. Manuscript * Indicates that these authors contributed equally to the work Reprints have been made with permission from the respective publishers.

Related Articles

1. Reshma Sundar, Marene Landström. Chapter: TRAF6 in book Encyclopedia of Signaling Molecules, Edition: 2013. XLVIII, Publisher: Springer, pp.1916-1921. 2. Anneli Jorsback, Maria Winkvist, Asa Hagner-McWhirter, Marene Landstrom, Reshma Sundar, Ulla Engstrom. Abstract LB-276: 2-D DIGE and siRNA to find new cancer targets. 2012. Cancer Research 3. Terese Karlsson, Reshma Sundar, Marene Landström, Emma Persson. Osteoblast-derived factors increase metastatic potential in human prostate cancer cells, an effect partially mediated by non-canonical, TRAF6-dependent TGFβ signaling. Manuscript

Contents Introduction 1 1. Prostate cancer 2 2. TGFβ 3 3. TGFβ superfamily 4 4. TGFβ isoforms and receptors 5

1. Smad proteins 6 2. TGFβ ligand activation 8 3. TGFβ canonical Smad signaling 8 4. Non-Smad TGFβ signaling 9

5. Non-Smad signaling via TRAF6 10 1. TRAF family and TRAF6 10 2. Tumor necrosis factor-α-converting enzyme (TACE) 12 3. Presenilins (PSs) 14 4. PKCζ 15 5. Snail1 16 6. Post-translational modification 18 1. Ubiquitination 18 2. Sumoylation 21 3. Phosphorylation 22 7. Notch signaling 23 Present Investigation 26 Aims 26 Materials and methods 27

Cell culture Transient transfection Protein analysis Immunoprecipitation Nuclear cytoplasmic fractionation assay

In vivo ubiquitination assay In vitro ubiquitination assay Western blotting Immunofluorescence Proximity ligation assay (PLA) Quantitative real-time RT-PCR Chromatin immunoprecipitation assay Site-directed mutagenesis Invasion assay FACS analysis Immunohistochemistry Animal studies Patient material Statistical analysis

Results and discussion 34 Paper I 34 Paper II 35 Paper III 36 Paper IV 37

Conclusions 38 Future perspectives 40 Acknowledgements 43 References 47

Abbreviations ADAM A disintegrin and metalloproteinase

ALK Activin receptor-like kinase

AMH Anti-mullerian hormone

AP-1 Activator protein-1

APH Anterior pharynx-defective

APP Amyloid precursor protein

AREG Amphiregulin

BMP Bone morphogenetic protein

CBP CREB-binding protein

CTF C-terminal fragment

CYLD Cylindromatosis

DNA Deoxyribonucleic acid

DUB Deubiquitinating enzyme

EMT Epithelial-mesenchymal transition

ER Endoplasmic reticulum

ERK Extracellular signal-regulated kinase

GDF Growth and differentiation factor

GS Glycine and serine

GTP Guanosine-5′-triphosphate

HAT Histone acetyltransferase

HECT Homologous to the E6-AP Carboxyl Terminus

HMGA High mobility group

HIF Hypoxia-inducible factor

ICAM Intercellular adhesion molecule

ICD Intracellular domain

IKK IkappaB kinase

JNK c-Jun N-terminal kinase

kDa Kilodalton

LAP Latency-associated peptide

LTBP Latent TGFβ binding protein

LUBAC Linear ubiquitin assembly complex

MAML Mastermind-like

MAPK Mitogen-activated protein kinase

MH Mad homology

MMP Matrix metalloproteinase

NES Nuclear Export signal

NFκB Nuclear factor kappa B

NTF N-terminal fragment

PC-3U Prostate cancer-3-Uppsala

PEN Presenilin enhancer

PS Presenilin

PTEN Phosphatase and tensin homolog

RING Really Interesting New Gene

RIP Regulated intramembrane proteolysis

SARA Smad anchor for receptor activation

SBE Smad binding element

siRNA Small interfering RNA

Smurf Smad ubiquitylation regulator factor

SNAG Snail/Gfi-1

STRAP Serine threonine kinase receptor associated protein

SUMO Small ubiquitin-like modifier

TACE TNFα-converting enzyme

TAK1 TGFβ-activated kinase 1

TGFβ Transforming growth factor beta

TNFα Tumor necrosis factor alpha

TRAF Tumor necrosis factor receptor-associated factor

TβR TGFβ receptor

TβRI TGFβ receptor type I

TβRII TGFβ receptor type II

UBP Ubiquitin-specific processing proteases

VCAM Vascular cell adhesion molecule

  1  

Introduction

Cells communicate with surrounding cells by sending and receiving various

signals. Chemical cellular signals include growth factors, neurotransmitters,

hormones, and extracellular matrix components. Sensory cells in the skin

respond to mechanical stimuli like touch. In such cells, signals are transmitted

across the cell membrane via cell surface-bound receptor proteins that send

signals from the outside to the inside of the cell. Subsequently, these signals

are transmitted to the nucleus or other cellular components via activation of

other proteins. Post-translational protein modification can regulate

intracellular signaling by controlling protein function. At any one time,

several different signal transduction cascades are operating in the cytoplasm,

and signaling pathways crosstalk with each other at multiple levels, allowing

cells to coordinate signals and regulate their growth, survival, differentiation,

and migration. Molecular interactions between different networks produce a

steady state of cellular integrity. Any interruption or over-activity of cell-to-

cell communication may cause disease conditions. For example, changes in

transforming growth factor-β (TGFβ) signaling pathways can alter cell

communication, resulting in uncontrolled cell growth that leads to cancer.

This thesis focuses on TGFβ and its downstream activation of non-Smad

signaling pathways in cancer. The present study aims to advance our

knowledge regarding the roles of the E3 ligase tumor necrosis factor receptor-

associated factor 6 (TRAF6) and the zinc finger transcriptional repressor

Snail1 in TGFβ-induced oncogenesis, with a particular focus on prostate

cancer.

  2  

1. Prostate cancer

Prostate cancer is the most commonly diagnosed cancer in men, with a

lifetime prevalence of one in six men. In 2012, over 1,111,000 new cases of

prostate cancer were diagnosed worldwide, and estimated 307,000 men died

of this disease (Esfahani, Ataei et al. 2015). Since 1977, the prostate cancer

incidence has increased three-fold, and the median age of diagnosis is

decreasing (Hussein, Satturwar et al. 2015). Prostate cancer is the most

common cancer among men in Europe, with highest world age-standardized

incidence rate reported in Norway and the lowest in Albania (Cancer

Research UK website). Prostate cancer development is influenced by a variety

of factors, including ethnicity, environment, lifestyle, genetic factors, and the

male sex hormone testosterone (Felgueiras, Silva et al. 2014).

Improving the life expectancy for patients with prostate cancer will require

better prognostic markers and treatment methods. Currently, the most

commonly used prognostic tools are the prostate-specific antigen (PSA) test

and histopathological grading, and the more commonly applied optional

treatments are surgical removal of the prostate gland, radiation,

chemotherapy, and hormonal therapy (androgen-deprivation therapy).

However, many tumors relapse within three years and develop androgen-

independent regrowth. It may be possible to prevent and treat recurrent

disease to increase patient survival, by targeting growth factors, thus more

research should be focused on identifying such biomarkers (Wikstrom,

Damber et al. 2001; Diener, Need et al. 2010).

TGFβ acts as a tumor suppressor in normal prostate epithelium and

premalignant lesions. However, TGFβ1 overexpression has been detected in

prostate tumor, and high TGFβ1 levels are associated with prostate cancer

  3  

progression (Steiner, Zhou et al. 1994). Advanced prostate cancers show

elevated TGFβ levels and decreased TGFβ receptor expression, which are

associated with bad prognosis (Jones, Pu et al. 2009). Prostate cancers also

reportedly show loss of heterozygosity on chromosome 18q, where the genes

encoding intracellular transducers for TGFβ signals (Smad2 and Smad4) are

localized (Padalecki, Weldon et al. 2003). The available data suggest that

molecular targeting of TGFβ signaling may be a promising therapeutic

approach for prostate cancer treatment (Jones, Pu et al. 2009).

2. TGFβ

Over three decades ago, TGFβ was first purified and characterized (Roberts,

Lamb et al. 1980; Moses, Branum et al. 1981). The TGFβ family now

includes 33 related isoforms that are potent cytokines involved in regulating

an array of cellular processes, such as cell proliferation, differentiation,

morphogenesis, tissue homeostasis, motility, adhesion, and death (Massague

1998; Moustakas and Heldin 2009). The human genome encodes over 42

TGFβ family members, while the Drosophila melanogaster and

Caenorhabditis elegans genomes encode 7 and 4, respectively. Remarkably,

TGFβ signaling is evolutionarily conserved in the ancestral Trichoplax

adhaerens.

TGFβ exerts different and opposing effects depending on the cellular context

and condition, and acts as a double-edged sword in carcinogenesis and tumor

progression. Initially, it suppresses cellular growth in normal epithelial and

premalignant cells. However, increased levels of locally secreted TGFβ can

induce pro-tumorigenic effects by switching the cell fate to invasion and

migration (Roberts and Wakefield 2003; Heldin, Vanlandewijck et al. 2012;

Papageorgis 2015). Numerous human cancers (e.g., colorectal, pancreatic,

  4  

gastric, and prostate cancer) have shown mutations in genes encoding TGFβ

pathway components, including TGFβ receptor I (TβRI), TGFβ receptor II

(TβRII), Smad2, and Smad4 (de Caestecker, Piek et al. 2000; Kim, Park et al.

2005).

3. TGFβ superfamily

The TGFβ superfamily encompasses TGFβs, bone morphogenetic proteins

(BMPs), growth and differentiation factors (GDFs), activins, nodal, and anti-

mullerian hormone (AMH) (Padua and Massague 2009). TGFβ superfamily

ligands signal through heterotetrameric complexes comprising two pairs of

type I and type II serine/threonine kinase receptors localized at the cell

surfaces. There are seven type I receptors (activin receptor-like kinase and

ALK 1–7) and five type II receptors (TβRII, ActR-IIA, ActR-IIB, AMHRII,

and BMPRII) (Moustakas and Heldin 2009). The TGFβ superfamily is

divided into four subfamilies with different receptor binding specifications

(Figure 1).

Ligand Type I

receptor

Type II

receptor

R-Smad

Co-Smad

TGFβ1, 2, 3 TβRI (ALK5) TβRII Smad2

Smad3

Smad4

Activin ALK4, 7 ActRII-A, B

BMP/GDF ALK1, 2, 3, 6 BMPRII

ActRII-A, B

Smad1

Smad5

Smad8 AMH ALK2, 3, 6 AMHRII

Figure 1. Schematic representation of the transforming growth factor-β

(TGFβ) superfamily components

  5  

4. TGFβ isoforms and receptors

The TGFβ subfamily comprises TGFβ1, 2, and 3. These proteins signal

through the 53-kDa TβRI (ALK5) and the 70-kDa TβRII, which are both

transmembrane protein serine/threonine kinases. TβRI and TβRII each have a

transmembrane domain, an intracellular kinase domain, and a relatively short

extracellular domain of approximately 150 amino acids that is N-glycosylated

and contains cysteine-rich regions. TβRI includes a highly conserved, 30-

amino acid, serine/glycine-rich sequence called the GS domain, which is

localized just upstream of the serine/threonine protein kinase domain (Figure

2). Although the TβRI kinase is referred to as a serine/threonine kinase, it can

also phosphorylate tyrosine residues (Lawler, Feng et al. 1997). In both TβRII

and TβRI, the kinase domain is followed by a short C-terminal extension,

which is larger in TβRII compared to in TβRI.

Figure 2. Transforming growth factor-β receptor I (TβRI) structure

  6  

TGFβ1 binding creates a heterotetrameric complex of two TβRI and two

TβRII. Compared to TβRII, TβRI has a very weak affinity for the ligand.

Ligand binding to TβRII results in a 300-fold increase of the binding affinity

of TβRII and TβRI (Yamashita, ten Dijke et al. 1994). In the ligand-induced

receptor complex, TβRII transphosphorylates TβRI in the TTSGSGSGLP

sequence of the GS domain and activates downstream signaling through Smad

proteins (Okadome, Yamashita et al. 1994; Wrana, Attisano et al. 1994;

Franzen, Heldin et al. 1995).

4.1. Smad proteins

Smad proteins are intracellular mediators of TGFβ signaling. The term Smad

was derived from the names of two proteins: small body size (Sma) in

Caenorhabditis elegans and mothers against decapentaplegic (Mad) in

Drosophila melanogaster (Sekelsky, Newfeld et al. 1995; Savage, Das et al.

1996). Based on structure and function, Smads are classified into three

subfamilies: the receptor-associated Smads (R-Smads) Smad1, 2, 3, 5, and 8;

the inhibitory Smads (I-Smads) Smad6 and 7; and the common mediator

Smad (Co-Smad) Smad4. Smad2 and Smad3 are direct substrates of TβRI.

Smad7 is an antagonistic Smad that inhibits the TGFβ signaling cascade and

is transcriptionally induced in response to TGFβ (Hayashi, Abdollah et al.

1997; Nakao, Afrakhte et al. 1997).

Smad proteins (Figure 3) contain a highly conserved N-terminal Mad

homology (MH1) domain, a C-terminal Mad homology (MH2) domain, and

an intervening linker domain of variable sequence and length. The R-Smads

and Co-Smad have a conserved MH1 domain that contains a nuclear

localization signal and a DNA binding domain, and that facilitates protein-

protein interactions. I-Smads lack the MH1 domain. In the basal state, the

  7  

MH1 and MH2 domains inhibit each other. The MH2 domain is involved in

protein–protein interactions through hetero- and homo-oligomerization. This

region binds to TβRI and contains the receptor phosphorylation motif Ser-X-

Ser (SXS) at the C-terminal end in R-Smads. The MH2 region is bound by

several transcription factors, such as androgen receptor (AR), LEF1/TCF,

FAST Fos, etc. (Abdollah, Macias-Silva et al. 1997; Kretzschmar, Liu et al.

1997; Souchelnytskyi, Tamaki et al. 1997; Moustakas, Souchelnytskyi et al.

2001). The variable linker region contains phosphorylation sites for mitogen-

activated protein kinases (MAPKs) and other kinases. Phosphorylation in this

region inhibits Smad nuclear translocation and promotes Smad degradation.

The linker region also contains a PPXY (PY) motif that acts as a binding

motif for the WW domain of the ubiquitin ligase Smurf1/2. The linker region

of the Co-Smad contains a nuclear export signal (NES) but no PY motif.

Figure 3. Structure of Smad family proteins

  8  

4.2. TGFβ ligand activation

TGFβ ligands are secreted as latent complexes comprising an N-terminal pro-

peptide and a C-terminal mature TGFβ polypeptide. The propeptide is termed

the latency-associated peptide (LAP) and it noncovalently binds to the TGFβ

polypeptide to form the small latent TGFβ complex (SLC). The SLC then

interacts with a large secretory glycoprotein called the latent TGFβ-binding

protein (LTBP), forming a large latent TGFβ complex (LLC). LTBP is

required for secretion and storage in the extracellular matrix (ECM). The

latent TGFβ ligands are activated by several proteases and interact with

integrin and other proteins. LAP can be cleaved by matrix metalloproteases-2

(MMP-2), releasing mature TGFβ (Koli, Saharinen et al. 2001; Annes,

Munger et al. 2003; Keski-Oja, Koli et al. 2004).

4.3. TGFβ canonical Smad signaling

TGFβ first binds to the extracellular domain of the constitutively active

TβRII. This enhances TβRI binding affinity, leading to the formation of a

symmetric heterotetrameric complex comprising two TβRII and two TβRI

molecules. The constitutively active TβRII is autophosphorylated at Ser231

and Ser409, triggering its kinase activity. The kinase-active TβRII

phosphorylates serine and threonine residues in the GS domain, thus

transactivating TβRI. In turn, activated TβRI phosphorylates downstream

effectors (the R-Smads Smad2 and Smad3) at their conserved C-terminal

serine residues (SXS). The L3 loops of the R-Smads interact with the L45

loop in the TβRI kinase domain. R-Smad phosphorylation is inhibited by the

I-Smad Smad7, which competitively interacts with TβRI (Hayashi, Abdollah

et al. 1997; Imamura, Takase et al. 1997; Hata, Lagna et al. 1998). The

receptor-phosphorylated R-Smads associate with the Co-Smad Smad4,

  9  

forming a trimeric complex comprising phosphorylated homo- or heteromeric

R-Smads and Smad4. Smad7 can also inhibit trimeric complex formation

(Hata, Lagna et al. 1998).

The trimeric complex navigates to the nucleus and regulates transcription in a

context-dependent manner. R-Smads and Co-Smads shuttle between the

cytoplasm and the nucleus. Phosphorylation of MAPK sites in the R-Smad

linker region inhibits Smad nuclear transport, leading to proteasomal

degradation, and thus down-regulating TGFβ signaling (Massague 1998). In

the nucleus, R-Smads regulate transcriptional responses by binding to certain

gene promoters for DNA binding transcription factors (Pierreux, Nicolas et al.

2000; Xu, Kang et al. 2002). Smad3 and Smad4 predominantly bind to the

CAGAC DNA motif known as canonical Smad-binding element (SBE), while

Smad2 does not bind directly to DNA. The canonical TGFβ–Smad signaling

pathway regulates genes controlling proliferation, differentiation, migration,

and apoptosis through associations with transcriptional coactivators (CBP,

HATs, and p300) and corepressors (Ski and SnoN) (Roberts 1999; Ross and

Hill 2008; Massague 2012).

4.4. Non-Smad TGFβ signaling

TGFβ also transmits signals through other pathways, including those

involving extracellular signal-regulated kinase1/2 (ERK1/2), c-Jun N-terminal

kinase (JNK), and p38 mitogen-activated kinase (MAPK) (Derynck and

Zhang 2003; Moustakas and Heldin 2009; Landstrom 2010). TGFβ can

activate phosphatidylinositol-3-kinase (PI3K) and the downstream Akt

through RhoA, as well as the tyrosine kinase Src (Bakin, Tomlinson et al.

2000; Vinals and Pouyssegur 2001; Derynck and Zhang 2003).

  10  

TβRI can phosphorylate serine and tyrosine residues in the adaptor protein

ShcA. This leads ShcA to recruits the adaptor protein Grb2 to form a complex

with Ras guanine exchange factor and son of sevenless (SOS), thereby

activating Ras and the downstream Erk1/2 MAP kinase (Lee, Pardoux et al.

2007; Moustakas and Heldin 2009). Moreover, TβRII can phosphorylate the

polarity protein Par6, promoting its interaction with the E3 ubiquitin ligase

Smurf1. In turn, Smurf1 regulates the degradation of Rho small GTPase,

resulting in loss of intercellular tight junction (Ozdamar, Bose et al. 2005).

This thesis primarily focuses on non-Smad signaling pathways in which the

ubiquitin ligase TRAF6 is a key protein. The TGFβ-induced TRAF6–TAK1

signaling pathway can induce both apoptosis and epithelial–mesenchymal

transition (EMT) (Sorrentino, Thakur et al. 2008; Yamashita, Fatyol et al.

2008; Landstrom 2010; Mu, Gudey et al. 2012).

5. Non-Smad signaling via TRAF6

5.1. TRAF family and TRAF6

The tumor necrosis factor receptor-associated factor (TRAF) family is a class

of adapter proteins that activate various signaling pathways, including those

involving nuclear kappa-light-chain-enhancer of activated B cells (NF-κB),

JNK p38, and mitogen-activated protein kinases (MAPKs). TRAFs associate

with the tumor necrosis factor receptors (TNFR) superfamily as well as other

receptors (Inoue, Ishida et al. 2000; Kobayashi, Walsh et al. 2004; Landstrom

2010; Mu, Gudey et al. 2012). The TRAFs are genetically conserved in

Dictyostelium discoideum (Regnier, Tomasetto et al. 1995), Caenorhabditis

elegans (Wajant, Muhlenbeck et al. 1998), Drosophila melanogaster (Liu, Su

et al. 1999), and mammals (Arch, Gedrich et al. 1998).

  11  

Seven different types of TRAFs have been found in mammals (TRAF1–7).

TRAF1 and TRAF2 were initially identified as adaptor proteins of TNFR2

(Cao, Xiong et al. 1996; Ishida, Mizushima et al. 1996). TRAF2, TRAF3, and

TRAF6 are ubiquitously expressed, while TRAF1, TRAF4, and TRAF5

expressions are more restricted. TRAF1 is generally expressed in the spleen,

lung, and testis and TRAF4 in the spleen, lung, and thymus. TRAF5 is highly

expressed during embryogenesis, and in the hippocampus and the olfactory

bulb of adults (Rothe, Wong et al. 1994; Mosialos, Birkenbach et al. 1995;

Arch, Gedrich et al. 1998).

All TRAF family members contain a conserved TRAF domain comprising a

highly conserved carboxyl-terminal region (C-domain) and a coiled-coil

amino-terminal region (Figure 4). This TRAF domain mediates homo- and

hetero-dimerization of TRAF family members. Except for TRAF1, all TRAFs

also contain an amino-terminal really interesting new gene (RING) finger

motif, which is rich in cysteines and histidines that coordinate Zn2+ ion

binding (Borden, Lally et al. 1995; Arch, Gedrich et al. 1998). TRAF6 was

independently identified as an IL-1 signal transducer using yeast two-hybrid

systems with an EST expression library (Cao, Xiong et al. 1996; Ishida,

Mizushima et al. 1996). TRAF6 is a cytosolic protein that includes a cysteine-

containing C3HC3D-type RING finger, and five zinc finger repeats. TRAF6-

null mice have an abnormal phenotype with defective bone formation. They

die at early age due to severe osteoporosis with defects in tooth eruption and

bone remodeling, resulting from impaired osteoclast formation (Lomaga, Yeh

et al. 1999; Naito, Azuma et al. 1999). The role of TRAF6 in IL-17 signaling

was confirmed using embryonic fibroblasts from TRAF6 knockout mice

(Bradley and Pober 2001).

  12  

Figure 4. Tumor necrosis factor receptor-associated factor 6 (TRAF6)

structure

In its function as an E3 ubiquitin ligase, TRAF6 interacts with the E2-

conjugating enzyme Msm2 (comprised of Ubc13 and Uev1A domains) to

synthesize Lys63-linked polyubiquitin chains on target proteins (Deng, Wang

et al. 2000; Wang, Deng et al. 2001). Recent studies have reported that

TRAF6 binds to the conserved consensus motif basic residue-X-P-X-E-X-X-

aromatic or acidic residue in TβRI (Sorrentino, Thakur et al. 2008). This

TβRI–TRAF6 interaction leads to TRAF6 autoubiquitination and subsequent

Lys63-linked polyubiquitination of TAK1 and, in turn, the activated TAK1

activates the downstream p38 MAPK pathway (Sorrentino, Thakur et al.

2008; Yamashita, Fatyol et al. 2008; Mu, Gudey et al. 2012; Gudey,

Wallenius et al. 2014; Sundar, Gudey et al. 2015).

5.2. Tumor necrosis factor-α-converting enzyme (TACE)

Tumor necrosis factor-α-converting enzyme (TACE)—also known as A

disintegrin and metalloprotease (ADAM) 17—was initially identified as a

protease that cleaves the transmembrane molecule TNF-α from the plasma

membrane (Blobel 1997; Black 2002). TACE is a ubiquitously expressed,

multi-domain, type I transmembrane protein comprising a cysteine-rich pro-

  13  

domain, a zinc-dependent catalytic domain, a cysteine-rich disintegrin region,

and a cytoplasmic tail (Figure 5).

Figure 5. Tumor necrosis factor-α-converting enzyme (TACE) structure

Most TACE-null mice die at birth, but a few survive with hair and skin

defects and with eyelid fusion failure (Peschon, Slack et al. 1998; Schlondorff

and Blobel 1999; Black 2002). Ectodomain shedding by TACE leads to the

release of several transmembrane molecules from the cell surface, including

TGFα, notch1, CSF1, AREG, VCAM1, sIL6R, ICAM1, ErbB2, and EGFR

(Kenny and Bissell 2007; Talmadge, Donkor et al. 2007; Scheller, Chalaris et

al. 2011; Gudey, Wallenius et al. 2014). On the other hand, under in vitro

conditions, recombinant TACE leads to little or no cleavage of synthetic

peptides of HER4, APP, TNFR-I, RANKl, Notch, and IL-6R (Mohan, Seaton

et al. 2002).

TGFβ reportedly activates TACE via the ERK1/2 MAPK signaling pathway.

Activated TACE cleaves TβRI, thus decreasing TGFβ-induced growth

inhibition due to reduced TβRI levels at the cell membrane (Liu, Xu et al.

2009). Mu et al. (2011) also report that TACE cleaves TβRI at the

  14  

transmembrane region, liberating the TβRI intracellular domain, which enters

the nucleus and regulates EMT-associated genes, thereby promoting cancer

cell invasion.

5.3. Presenilins (PSs)

Presenilins (PSs) are polytransmembrane aspartyl proteases that were initially

identified through linkage analysis of Alzheimer’s disease (Medina and Dotti

2003), and that are highly conserved in both invertebrates and vertebrates.

Mammalian PSs include two ubiquitously expressed homologous proteins:

presenilin 1 (PS1) and presenilin 2 (PS2). PS1-null mice die immediately after

birth and exhibit retarded embryonic growth, while PS2 knockout mice are

viable and fertile, showing only mild pulmonary fibrosis and hemorrhage with

increasing age (Brunkan and Goate 2005). PS1 localizes in the endoplasmic

reticulum (ER), Golgi apparatus, endosomes, lysosomes, phagosomes, plasma

membrane, and mitochondria of cells (De Strooper, Beullens et al. 1997;

Brunkan and Goate 2005; Vetrivel, Zhang et al. 2006).

The most widely accepted PS1 structure contains eight transmembrane

domains, the N-terminus, the C-terminus, and a cytosolic hydrophilic loop

located between transmembrane segments 6 and 7 (Figure 6). A newly

synthesized PS1 holoprotein undergoes endoproteolytic cleavage by an

unknown presenilinase, generating an N-terminal fragment (NTF) of 27–28

kDa and a C-terminal fragment (CTF) of 16–17 kDa (Esler, Kimberly et al.

2000; Li, Xu et al. 2000). The PS1 holoprotein, PS1-NTF, and PS1-CTF each

undergo ubiquitination and proteasomal degradation (Li, Pauley et al. 2002).

  15  

Figure 6. Presenilin 1 (PS1) structure

PS1 associates with TRAF6, leading to TRAF6 autoubiquitination and further

ubiquitination of the p75 neurotrophin receptor in an NGF-dependent manner

(Powell, Twomey et al. 2009). PS1, nicastrin, anterior pharynx defective-1

(Aph-1), and presenilin enhancer 2 (pen-2) together form the active γ-

secretase complex, and co-expression of these components leads to increased

presenilin heterodimerization and γ-secretase activity. The γ-secretase

complex cleaves various type I transmembrane receptors, including amyloid

precursor protein (APP), Notch, and CD44 (Edbauer, Winkler et al. 2003;

Wakabayashi and De Strooper 2008; Gudey, Wallenius et al. 2014). PS1 is

also implicated in TβRI proteolytic cleavage (Gudey, Sundar et al. 2014).

5.4. PKCζ

Protein kinase C (PKC) represents a family of protein serine/threonine kinases

that serve as transducers of various signaling networks. PKCs play important

roles in cancer progression and metastasis (Koivunen, Aaltonen et al. 2006).

PKC isozymes are grouped into three subfamilies based on their structure and

  16  

function: classic or conventional PKCs (α, βΙ, βΙΙ, and γ), novel PKCs (δ, ε, η,

and θ), and atypical PKCs (ζ and λ/ι). PKCζ and PKCλ/ι lack a Ca2+-binding

domain and autoinhibitory region. PKCδ contains an N-terminal PB domain,

a pseudosubstrate (PS), a cysteine-rich zinc finger motif containing a C1

domain, and a C-terminal kinase domain.

Figure 7. Protein kinase C-ζ (PKCζ) structure

PKCζ is involved in the ERK1/2 MAPK signaling pathway and functions as

an adaptor in the ERK5–MAPK pathway. Its adaptor function is important in

regulating NF-κB and the innate immune response. PKCζ also promotes EGF-

stimulated chemotaxis of human breast cancer cells (Berra, Diaz-Meco et al.

1995; Griner and Kazanietz 2007; Xiao and Liu 2013). We have also found

that TGFβ can cause PKCζ activation (Mu, Sundar et al. 2011)

5.5. Snail1

Snail1 is a C2H2 zinc finger protein belonging to a transcription factor family

that also includes Slug and Scratch proteins. Snail1 was first characterized in

1984 through a mutagenesis screen of Drosophila melanogaster. The 2-kb

Snail1 gene is localized on chromosome 20q.13.2 and contains three exons.

SNAI1P is a Snail1 retrogene. Snail1 is a 264-amino acid protein with a

molecular mass of 29.1 kDa. Its structure includes four zinc finger domains,

an NES domain, a serine-rich domain (SRD), a destruction box (DB), and a

SNAG domain (Figure 8).

  17  

Figure 9. Snail1 structure

Transcriptional regulators—including Notch-ICD, NF-κB, Smad, HMGA2,

Erg-1, Prap-1, Stat-3, MTA3, IKK-α, and HIF-1-α—affect Snail1 expression

by directly binding to its promoter (Thuault, Tan et al. 2008; Kaufhold and

Bonavida 2014). Snail1 also binds its own promoter to upregulate its own

expression. Yin Yang 1 (YY1) induces Snail1, whereas p53 represses Snail1

expression through induction of miR-34a, miR-34b, and miR-34c (Kim, Kim

et al. 2011; Kaufhold and Bonavida 2014). In prostate cancer, Snail1

expression is correlated with high Gleason score (Poblete, Fulla et al. 2014).

Snail1-null mutations are lethal in mice due to a lack of gastrulation. EMT is

an important part of normal embryonal development, and also occurs in

metastatic cancer. EMT requires Snail1 activation by extracellular molecules,

including EGFR ligands, TGFβ, and Wnts (Heldin, Landstrom et al. 2009;

Fuxe, Vincent et al. 2010; Sou, Delic et al. 2010; Mu, Sundar et al. 2011).

Snail1 directly binds to the E-boxes (5′-CACCTG-3′ motifs) in the proximal

promoter region of E-cadherin, repressing E-cadherin expression (Peinado,

Olmeda et al. 2007). Snail1 also represses other epithelial markers (e.g.,

occludin, claudins, and mucin-1) and upregulates the expressions of

mesenchymal markers (e.g., vimentin, fibronectin, MMP-2, and MMP-9).

  18  

Additionally, Snail1 represses the tumor suppressors RKIP and PTEN

(Kaufhold and Bonavida 2014). Snail1 is overexpressed in many cancer types

and it promotes tumor cell migration, invasion, and metastasis. In breast

carcinoma cells, Snail1 overexpression induces TβRII expression

(Dhasarathy, Phadke et al. 2011). In lung cancer cells, Snail1 is repressed by

TTF1, thus reversing EMT (Saito, Watabe et al. 2009; Moustakas and Heldin

2012).

6. Post-translational modification

Post-translational modification involves the covalent addition of a functional

group to a protein after its biosynthesis. TGFβ receptors undergo multiple

types of post-translational modifications, including the well-characterized

processes of phosphorylation, sumoylation, and ubiquitination (Kang, Liu et

al. 2009).

6.1. Ubiquitination

Ubiquitination is the reversible labeling of a target protein with a small

modifier called ubiquitin. Ubiquitin is a 76-amino acid protein with a

molecular mass of 8.5 kDa. Ubiquitination influences different cellular

processes, including cell cycle regulation, DNA repair, endocytosis,

autophagy, protein degradation, and gene regulation (Emmerich, Schmukle et

al. 2011). The ubiquitination process involves three different enzymes: the E1

ubiquitin-activating enzyme, the E2 ubiquitin-conjugating enzyme, and the E3

ubiquitin-ligating enzyme (Figure 9).

E1 activates ubiquitin through a two-step ATP-dependent process. First, the

E1 activating enzyme binds both ATP and ubiquitin, forming an ubiquitin-

  19  

adenylate intermediate. In the next step, ubiquitin is transferred to the

sulfhydryl group of the cysteine residue of E1 through a thio-ester linkage.

The E2 conjugating enzyme then transfers the activated ubiquitin from E1 to a

cysteine of E2 via a trans-thio-esterification reaction. The E3 ubiquitin-

ligating enzyme acts as a substrate recognition module, catalyzing the transfer

of ubiquitin from E2 to the target protein. This enzymatic process creates an

isopeptide or peptide bond between the C-terminal glycine of ubiquitin and a

lysine of the target protein (Pickart 2001; Groettrup, Pelzer et al. 2008;

Schulman and Harper 2009; Dikic and Robertson 2012). Ubiquitin can also

bind the target protein through the N-terminus or methionine, cysteine, serine,

and threonine residues (Hochstrasser 2009). Moreover, ubiquitin molecules

can covalently bind to each other, forming polyubiquitin chains. The human

genome encodes two E1 activating enzymes, 37 E2 conjugating enzymes, and

over 600 E3 ligases (Komander 2009).

Figure 8. The ubiquitination process

  20  

Different types of ubiquitination control a variety of cellular processes by

regulating protein localization, protein–protein interaction, protein activation,

and proteasomal or lysosomal degradation of proteins. Marking a protein with

a single ubiquitin on a single lysine residue of a target protein is termed

monoubiquitination, while the addition of multiple ubiquitin moieties on

multiple lysine residues is called multi-monoubiquitination. These types of

ubiquitination trigger receptor endocytosis and DNA repair (Komander 2009).

Polyubiquitination is the formation of a ubiquitin chain on a single lysine

residue of the target protein. Ubiquitin has seven lysine residues at positions

6, 11, 27, 29, 33, 48, and 63, each of which can mediate a ubiquitin chain.

Lys48-linked polyubiquitin chains target proteins for proteasomal

degradation. On the other hand, Lys63-linked polyubiquitin chains target

proteins for activation of signal transduction, endocytic trafficking, or

translation (Dikic, Wakatsuki et al. 2009; Komander 2009; Kravtsova-

Ivantsiv and Ciechanover 2012). Linear ubiquitination or M1 linkage is a

recently discovered process in which a ubiquitin chain is linked by a

methionine residue created by the linear ubiquitin assembly complex

(LUBAC) (Rieser, Cordier et al. 2013).

In canonical TGFβ signaling, basal Smad protein basal levels are regulated by

ubiquitination and proteasomal degradation. Smurfs are E3 ligases of the

homologous to the E6-AP carboxyl terminus (HECT) family, which contain

an N-terminal C2 domain, WW domains, and a C-terminal HECT domain.

Smurf1 regulates Smad1 and Smad2 abundance. SCF/Roc1 E3 ligase targets

phosphorylated Smad3 for proteasomal degradation in the cytoplasm

(Fukuchi, Imamura et al. 2001; Moustakas, Souchelnytskyi et al. 2001).

Smad4 transcriptional activity is enhanced by monoubiquitination at lysine

507 in the MH2 domain (Moren, Hellman et al. 2003). Smurfs also target the

TGFβ receptor complex and the transcriptional co-repressor SnoN. Smurf1

  21  

and Smurf2 form a complex with nuclear Smad7, which is exported to the

cytoplasm where it targets the TβR for proteasomal and lysosomal

degradation (Moustakas, Souchelnytskyi et al. 2001). STRAP is a WD

domain protein that promotes Smurf-mediated TβRI ubiquitination (Liu, Elia

et al. 2000). The E3 ligase TIF1γ disrupts the nuclear–Smad complex through

Smad4 ubiquitination (Hesling, Fattet et al. 2011; Moustakas and Heldin

2012). Moreover, the E3 ligase Arkadia stimulates TGFβ signaling by

targeting the repressors Smad7, c-Ski, and SnoN. Fbxw also positively

regulates TGFβ signaling by promoting the degradation of phosphorylated

TGFβ-induced factor 1 (TGIF1) (De Boeck and ten Dijke 2012).

In non-canonical TGFβ signaling pathways, the E3 ligase TRAF6 associates

with TβRI at a conserved consensus motif, activating downstream TAK1–

p38/JNK MAPK pathways (Sorrentino, Thakur et al. 2008; Yamashita, Fatyol

et al. 2008). In this context, we have identified Lys178 as an acceptor lysine

for TGFβ- and TRAF6-induced polyubiquitination of TβRI (Sundar, Gudey et

al. 2015).

Deubiquitinating enzymes (DUBs) remove ubiquitin or ubiquitin chains from

the substrate. UCH37 is a DUB that deubiquitinates TβRI by binding Smad7,

and thereby promoting TGFβ-induced migration (Cutts, Soond et al. 2011).

Another DUB, cylindromatosis (CYLD), inhibits TAK1-p38 activation by

deubiquitinating Smad7 in T cells (Zhao, Thornton et al. 2011).

6.2. Sumoylation

Small ubiquitin-related modifier (SUMO) is a reversible post-translational

protein modifier that is covalently attached to acceptor lysines of target

proteins via sumoylation. SUMO proteins are approximately 10 kDa in size

  22  

and share 18% amino acid sequence identity with ubiquitin. The human

genome encodes four SUMO proteins: SUMO1, SUMO2, SUMO3, and

SUMO4. Sumoylation involves formation of an isopeptide bond between the

C-terminal glycine of SUMO and a lysine residue of the target protein. Unlike

ubiquitin, SUMO proteins have a consensus acceptor site: ΨKxE, where Ψ is

a hydrophobic amino acid, K is lysine, x is any amino acid, and E is glutamic

acid (Hannoun, Greenhough et al. 2010).

Sumoylation can alter target protein localization, stability, and activity

(Bayer, Arndt et al. 1998; Mossessova and Lima 2000; Geiss-Friedlander and

Melchior 2007). The majority of SUMO targets are nuclear proteins, but

sumoylation can also reportedly regulate cell surface receptors and channels

at the plasma membrane. TβRI is sumoylated by Ubc9 on a lysine residue that

is not a classical SUMO consensus acceptor site. TβRI sumoylation increases

TβRI kinase activation and enhances Smad2 and Smad3 activation (Kang, Liu

et al. 2009). Moreover, TGFβ reportedly down-regulates the sumo-specific

E3-ligase PIAS1, in turn, destabilizing SnoN by reducing sumoylation

(Netherton and Bonni 2010; Moustakas and Heldin 2012).

6.3. Phosphorylation

Phosphorylation is a post-translational modification involved in many cellular

processes. The covalent binding of a phosphate group to a target protein can

alter protein function by changing the protein conformation or by promoting

protein–protein interactions. TGFβ binding to TβRII induces phosphorylation

of serine and threonine residues in the TβRI GS domain, causing a TβRI

conformational change. TβRI phosphorylation activates its kinase activity,

resulting in phosphorylation of downstream R-Smads and activating the Smad

signaling cascade. Phosphorylation of TβRI also prevents binding of the

  23  

TGFβ signaling-inhibitor FKBP12 (Huse, Chen et al. 1999; Moustakas,

Souchelnytskyi et al. 2001; Xu, Liu et al. 2012). TβRI and TβRII are dual-

specificity kinases that can also phosphorylate tyrosines. Such

phosphorylation activates the Erk1/2–MAPK signaling pathway through

association between TβRI and the adaptor protein Shc (Lee, Pardoux et al.

2007). Dephosphorylation also regulates the TGFβ cascade, in which Smad7

interacts with GADD34 (a regulatory subunit of protein phosphatase 1),

resulting in TβRI dephosphorylation (Xu, Liu et al. 2012).

7. Notch signaling

In 1919, the laboratory of Thomas Hunt Morgan characterized a notched wing

phenotype in Drosophila melanogaster with loss of wing margin tissue from

the distal tip of the wing (Mohr 1919), which was determined to be caused by

a mutation in the newly discovered gene encoding the Notch protein

(Yamamoto, Schulze et al. 2014). The Notch signaling pathway has since

been determined to be important for cell–cell communication and cell-fate

determination during metazoan development, embryogenesis, cell lineage

progression, and differentiation (Andersson, Sandberg et al. 2011).

Notch is a single-pass transmembrane protein comprising two non-covalently

bound polypeptides. The mammalian genome encodes four Notch molecules

(Notch1–4). Canonical Notch signaling is initiated by the binding of one of

the five cell-bound Notch ligands: Jagged1, Jagged2, delta-like ligand (DLL)

1, DLL-3, or DLL-4. Ligand-bound Notch undergoes a sequential proteolytic

cleavage called regulated intramembrane proteolysis (RIP) (Andersson,

Sandberg et al. 2011; Ayaz and Osborne 2014). RIP reportedly regulates a

wide variety of cellular processes, including protein activation, intercellular

  24  

signaling, intracellular signaling, and protein degradation (McCarthy,

Twomey et al. 2009; De Strooper and Annaert 2010).

In its initial proteolysis, known as S1 cleavage at trans-Golgi, Notch is

cleaved by furin-like convertases, forming a bipartite receptor that

translocates to the cell membrane. Notch RIP then starts with the binding of a

ligand that is presented to Notch by a neighboring cell, triggering ectodomain

shedding by ADAM17/TACE, termed S2 cleavage. This forms the Notch

extracellular truncation fragment (NEXT), also known as notch extracellular

domain (NECD). Ectodomain shedding depends on the distance of the

substrate from the plasma membrane and on conformational changes of the

cleavage site (Hayashida, Bartlett et al. 2010). Subsequently, S3 cleavage of

NEXT/NECD is mediated by γ-secretase, which releases the Notch

intracellular domain (NICD). PS1—the catalytic core of the γ-secretase

complex—cleaves NEXT/NECD in the transmembrane region (Brou, Logeat

et al. 2000; Hayashida, Bartlett et al. 2010; Andersson, Sandberg et al. 2011;

Bi and Kuang 2015). The NICD then translocates to the nucleus, where it

interacts with the transcriptional repressor RBPJκ/CSL, converting it into a

transcriptional activator of downstream target genes, including hairy/enhancer

of split 1 (HES1) and HES-related with YRPW motif families (HEY1) (Nam,

Weng et al. 2003; Ayaz and Osborne 2014; Teodorczyk and Schmidt 2014).

The growing list of Notch targets includes HES7, HEY2, PPAR, c-Myc,

Sox2, Pax6, cyclinD1, NFκB, and p21/waf1. Nuclear NICD later undergoes

proteasomal degradation by the E3 ligase F-box and WD repeat domain-

containing 7 (FBW7). HES5 reportedly represses FBW7 transcription and

creates a positive feedback loop in Notch signaling (O'Neil, Grim et al. 2007;

Sancho, Blake et al. 2013; Teodorczyk and Schmidt 2014).

  25  

The Notch signaling pathway is implicated in the progression of breast, lung,

prostate, and cervical cancers. The TGFβ and Notch signaling pathways are

not completely independent of each other, as TβRI-ICD and NICD form a

complex in TGFβ-stimulated prostate cancer cells (Gudey, Sundar et al.

2014). Moreover, TGFβ induces Jagged-1 expression, and Jagged-1

knockdown reduces TGFβ-induced p21 expression (Gudey, Sundar et al.

2014; Lindsey and Langhans 2014). TGFβ also induces HES1 and HEY1

expressions (Blokzijl, Dahlqvist et al. 2003; Zavadil, Cermak et al. 2004).

  26  

Present Investigation

Aims

The aim of this thesis was to investigate the functional role of TGFβ-induced

and TRAF6-mediated signal transduction in cancer, with the main focus on

prostate cancer. The specific aims of each paper were as follows:

Paper I: To investigate the functional role of TRAF6 on the TβRI, and its

biological significance.

Paper II: To elucidate the molecular mechanism underlying the

transmembrane cleavage of TβRI by PS1, and its role in cancer.

Paper III: To identify the ubiquitin acceptor lysine in TβRI, as well as the

biological significance of TRAF6-induced TβRI ubiquitination.

Paper IV: To demonstrate the functional role of the zinc finger protein Snail1

in TGFβ-induced tumor progression.

  27  

Materials and methods

Cell culture

All cells were cultured at 37°C in 5% CO2 in a humidified incubator.

Human prostate cells (including LnCaP cells and PC3U cells originated from

PC3 cells) and human lung carcinoma cells (A549) were cultured in Roswell

Park Memorial Institute medium (RPMI-1640) supplemented with 10% fetal

bovine serum (FBS), 1% L-glutamine, and 1% penicillin and streptomycin

(PEST). Cells were starved for 16–18 hours in RPMI with 1% FBS, 1% L-

glutamine, and 1% PEST.

Human breast cancer cells (MDA-MB-231), human embryonic kidney (HEK)

cells containing SV40-T antigen (293T), TRAF6 knockout (TRAF6−/−) mouse

embryonic fibroblasts (MEFs), PS1 knockout (PS1−/−) MEFs, and wild-type

MEFs were grown in Dulbecco’s modified essential medium (DMEM)

supplemented with 10% FBS, 1% L-glutamine, and 1% PEST. MDA-MB-231

and 293T cells were starved for 16–18 hours in DMEM containing 1% FBS,

1% L-glutamine, and 1% PEST. MEFs were starved for 16–18 hours in

DMEM containing 0.5% FBS, 1% L-glutamine, and 1% PEST.

Mouse prostate epithelial cells (TRAMPC2) derived from C57BL/6 mice

were cultured in DMEM with 5% FBS, 5% Nu-serum IV, 1% PEST, and 4

mM L-glutamine; adjusted to contain 1.5 g/L sodium bicarbonate and 4.5 g/L

glucose; and supplemented with 0.005 mg/ml bovine insulin and 10 nM

dehydroisoandrosterone. Cells were starved for 16–18 hours in DMEM

containing 1% FBS, 1% Nu-serum IV, 1% PEST, and 4 mM L-glutamine;

adjusted to contain 1.5 g/L sodium bicarbonate and 4.5 g/L glucose; and

  28  

supplemented with 0.005 mg/mL bovine insulin and 10 nM

dehydroisoandrosterone. After starvation, cells were stimulated with TGFβ1

(10 ng/mL) for the indicated time periods.

Transient transfection

PC3U and HEK 293T cells were transfected with plasmids using Fugene6

transfection reagent following the manufacturer’s instructions. PC3U cells

were transfected with a non-targeting control RNA or a specific short

interfering RNA (siRNA) using oligofectamine transfection reagent following

the manufacturer’s instructions.

Protein analysis

After TGFβ1 stimulation, cells were harvested and washed with ice-cold

phosphate-buffered saline (PBS). For protein extraction, cells were lysed in

radioimmunoprecipitation assay (RIPA) buffer. Protein concentration was

measured using a BCA kit and samples were subjected to SDS-PAGE on

NuPAGE gels, followed by western blotting.

Immunoprecipitation

Cell lysates were incubated overnight at 4°C with the indicated antibody.

Next, Protein G Sepharose beads were added. The Protein G Sepharose beads

with protein complexes attached were washed in RIPA buffer using

centrifugation. After addition of SDS sample buffer with reducing agent, the

samples were heated at 95°C for 10 min.

Nuclear cytoplasmic fractionation assay

TGFβ1-stimulated cells were harvested, washed with ice-cold PBS, and

centrifuged. To the pellet, we added Buffer I containing 20 mM Tris HCl (pH

7), 10 mM KCl, 2 mM MgCl2, 0.5% NP40, 1 mM aprotinin, and 1 mM

  29  

Pefabloc. The cells were then sheared with a needle and syringe, and then the

samples were centrifuged and the supernatant was collected as the

cytoplasmic fraction. The pellet was washed with Buffer II (Buffer I + 0.5 M

NaCl) and centrifuged again. The supernatants were collected as the nuclear

fraction. The cytoplasmic and nuclear fractions were subjected to protein

measurement and western blotting.

In vivo ubiquitination assay

PBS containing 1% SDS was added to the harvested cells, which were then

washed with ice-cold PBS. Samples were heated at 95°C for 10 minutes,

followed by the addition of PBS containing 0.5% NP40 and protease

inhibitors. After the samples were centrifuged, the thick slimy layer was

discarded and the remaining parts of the samples were incubated overnight

with a specific antibody. Next, Protein G Sepharose beads were added, the

samples were centrifuged, and the Protein G Sepharose beads were washed in

PBS containing 0.5% NP40. We next added SDS sample buffer containing 1

mM dithiothreitol (DTT) reducing agent to the samples, which then were

heated at 95°C. Samples were subjected to SDS-PAGE followed by western

blotting.

In vitro ubiquitination assay

An in vitro ubiquitination assay was performed with E1, E2 (Ubc13-Uev1A),

and GST-TβRI or GST-K178R-TβRI proteins with or without GST-TRAF6.

The reaction mixture also contained 10 mM MgCl2, 50 mM NaCl, 20 mM

Tris (pH 7.4), 10 mM ATP, 10 mM DTT, 2.5 µM ubiquitin, and 100 µM

MG132. The reaction was stopped with addition of SDS sample buffer. Then

the samples were subjected to SDS-PAGE, followed by western blotting.

  30  

Western blotting

Samples were subjected to SDS-gel electrophoresis using NuPAGE Novex 4

12% or 10% gels, 12% Bis-Tris gels, or 3–5% or 7% Tris-Acetate gels. After

electrophoresis, proteins were transferred to a nitrocellulose membrane (iBlot

transfer stack) using an iBlot gel transfer device. The membranes were probed

using specific primary and secondary antibodies, and developed using

chemiluminescence and autoradiography.

Immunofluorescence

Harvested cells were rinsed in PBS, and then fixed in 4% paraformaldehyde

for 10 minutes, followed by permeabilization in 2% Triton X-100. PBS with

5% BSA was used for blocking, and then the cells were incubated with

specific primary and secondary antibodies. Subsequently, the slides were

mounted with mounting medium containing 4′,6-diamidino-2-phenylindole

(DAPI).

Proximity ligation assay (PLA)

Following TGFβ1 stimulation, cells were harvested. The cells were then fixed

in 4% paraformaldehyde, permeabilized in 2% Triton X-100, blocked in PBS

with 5% BSA, and incubated with specific primary antibodies. The proximity

ligation assay (PLA) was performed following the manufacturer’s

instructions. Confocal microscopy was used to visualize the slides and

pictures were taken. Signals were quantified using Blob-Finder software.

Quantitative real-time RT-PCR

After TGFβ1 stimulation, cells were harvested and RNA was isolated using

the Qiagen RNA isolation kit. Next, cDNA was synthesized using the

Thermoscript RT-PCR cDNA synthesis kit. Real-time RT-PCR was

  31  

performed using the Power SYBR Green PCR Master Mix and appropriate

forward and reverse primers.

Chromatin immunoprecipitation assay

The chromatin immunoprecipitation (ChIP) assay was performed using the

Abcam ChIP kit following the manufacturer’s protocol. DNA and proteins

were cross-linked in 4% formaldehyde, and the cells were sheared. Then the

chromatin was precipitated using specific antibodies and was reverse cross-

linked. After purification, DNA was amplified using quantitative RT-PCR

with the Power SYBR PCR Master Mix and specific forward and reverse

primers.

Site-directed mutagenesis

Following the manufacturer’s instructions, we used the QuikChange Site-

Directed Mutagenesis kit to create point mutants of different constitutively

active TβRI constructs. Mutagenesis was confirmed by sequencing the

plasmids.

Invasion assay

Invasion assays were performed using CytoSelect Cell Invasion assay kits

following the manufacturer’s instructions. Serum-free RPMI-1640 was used

to rehydrate the upper basement membrane. Cells were seeded in the upper

layer of the chamber in medium with or without TGFβ1. The lower chamber

was filled with RPMI 1640 with 10% FBS. Non-invasive cells were removed,

while invasive cells were stained with crystal violet staining solution and then

photographed. Colorimetric quantification was performed by measuring the

optical density at 560 nm using a spectrophotometer.

  32  

FACS analysis

TGFβ1-stimulated cells were harvested, fixed, permeabilized, and blocked.

The cells were then stained with primary antibody and incubated with

secondary antibody. Next the cells were incubated with propidium iodide

solution to label RNA, and with Hoechst dye to label DNA. Finally, the cells

were sorted using FACS LSR II, and the data were analyzed using BD

bioscience software.

Immunohistochemistry

Tissue sections and a tissue microarray (TMA) containing multiple samples

were deparaffinized, pre-treated, and blocked. The sections were incubated

with primary antibody, and then HRP polymer. After counter-staining with

hematoxylin, the tissue sections were analyzed using a bright-field

microscope.

Animal studies

This study was approved by the local animal review board in Umeå, Sweden

(Approval ID: A110-12; Date: 21 August, 2012). TRAMPC2 cells were

subcutaneously injected into C57BL/6 mice. Upon reaching palpable tumor

volume, the mice were intraperitoneally injected with DMSO (control) or

DBZ (a γ-secretase inhibitor) for 10 days. Then the mice were sacrificed and

tumors were collected. Tissues were frozen or processed for RNA and protein

analysis.

Patient material

We used human tumor tissues from prostate, kidney, and urinary bladder

cancers. These studies were approved by the Uppsala Ethical Review Board.

  33  

Statistical analysis

Statistical analyses were performed using SPSS statistics. The Mann-Whitney

U test was used to calculate p values in Papers II and III. One-way ANOVA

and Student’s two-tailed t-test were used to determine statistical significance

in Papers I, III and IV.

  34  

Results and discussion

Paper I

TRAF6 ubiquitinates the TGFβ type I receptor to promote its cleavage

and nuclear translocation in cancer

An increasing number of transmembrane receptors reportedly undergo

ectodomain shedding (Hayashida, Bartlett et al. 2010). TβRI ectodomain

cleavage is mediated by TACE in an Erk1/2 MAP-kinase-dependent manner,

resulting in reduced TGFβ signaling (Liu, Xu et al. 2009). We found that

TGFβ activation caused Lys63-linked polyubiquitination of TβRI by TRAF6,

as well as PKCζ activation, which promotes TβRI cleavage by TACE.

Nuclear TβRI-ICD accumulation in human prostate cancer cells was

confirmed using ectopically expressed HA-TβRI or GFP-TβRI. TGFβ

stimulation activates TACE, which then cleaves TβRI between glycine 120

and leucine 121, releasing a 34-kDa TβRI-ICD.

We found that TβRI-ICD translocated to the nucleus, where it associated with

the transcriptional co-activator p300 in nuclear promyelocytic leukemia

(PML) bodies, promoting invasion in various types of cancer cells.

Moreover, ChIP assays revealed TGFβ-induced binding of TβRI-ICD to the

Snail1 promoter. Nuclear TβRI-ICD accumulation was observed only in

malignant prostate cells, not in primary human epithelial cells. In conclusion,

our data suggested that TRAF6-dependent non-canonical TGFβ signaling

plays a specific role in tumor invasion.

  35  

Paper II

TRAF6 stimulates the tumor-promoting effects of the TGFβ type I

receptor through polyubiquitination and activation of presenilin1

Regulated intramembrane proteolysis is a series of regulated events targeting

transmembrane proteins to release their intracellular domains (ICDs). The γ-

secretase complex, of which PS1 is the catalytic core, is involved in the

transmembrane cleavage of various receptors, including Notch (McCarthy,

Twomey et al. 2009). We investigated whether γ-secretase is involved in

TRAF6-dependent TβRI-ICD formation by PS1 (Mu, Sundar et al. 2011; Mu,

Gudey et al. 2012).

Our results suggested that TGFβ enhances PS1 expression and activation.

Moreover, TRAF6 polyubiquitinates PS1 and recruits it to TβRI to cleave

TβRI at the transmembrane region. Thus, TβRI is first cleaved by TACE in

the ectodomain, and then by PS1 at the transmembrane region. After TGFβ

stimulation, TβRI-ICD associated with NICD, driving the expressions of

genes including TβRI, Jag1, and Snail1. In vitro treatment of cells with a γ-

secretase inhibitor prevented TGFβ-induced invasion in various cancer cells.

Moreover, using xenograft prostate cancer mice as an in vivo model, we found

that γ-secretase inhibitor treatment prevented TβRI cleavage and reduced

tumor growth.

  36  

Paper III

TRAF6 promotes TGFβ-induced invasion and cell-cycle regulation via

Lys63-linked polyubiquitination of Lys178 in TGFβ type I receptor.

Ubiquitination is a post-translational protein modification that can alter

protein function as well as control cellular processes, such as apoptosis, cell-

cycle regulation, and gene regulation. We recently reported the association

between TRAF6 and TβRI. Furthermore, TGFβ induces Lys63-dependent

TβRI ubiquitination and prompts TβRI-ICD formation by promoting

proteolytic cleavage by TACE and PS1 (Sorrentino, Thakur et al. 2008; Mu,

Sundar et al. 2011; Gudey, Sundar et al. 2014).

In this study, we identified Lys178 of TβRI as the acceptor lysine for TRAF6-

induced TβRI ubiquitination following TGFβ stimulation. Site-directed

mutagenesis of Lys178 to arginine inhibited TβRI ubiquitination in vivo and

in vitro. Moreover, a TβRI point mutant with Lys178 changed to arginine

failed to form the ICD and to translocate to the nucleus. Our findings

suggested that Lys63-linked TβRI polyubiquitination promoted TβRI-ICD

formation and induced the expressions of EMT markers, such as Vimentin,

Twist, N-cadherin, and the cell cycle regulator cyclin D1. Moreover, TβRI

polyubiquitination promoted invasion in prostate cancer cells.

  37  

Paper IV

Proinvasive Snail1 targets TGFβ Type I receptor to promote epithelial to

mesenchymal transition in prostate cancer

Epithelial-to-mesenchymal transition (EMT) leads epithelial cells to lose their

polarity and cell–cell adhesion, and causes cancer cells to gain migratory and

invasive properties (Hanahan and Weinberg 2000; Hanahan and Weinberg

2011). TGFβ plays an important role in promoting tumor growth by triggering

EMT. The zinc finger protein Snail1 also plays a key role in the EMT process

(Padua and Massague 2009; Heldin, Vanlandewijck et al. 2012).

In this study, we found that the transcriptional regulator Snail1 regulated

TβRI expression at both the mRNA and protein levels. We also found that

Snail1 regulated HES1 expression. Our results indicated that TβRI down-

regulation reduced HES1 expression, suggesting that Snail1 also promotes

EMT through TβRI in a TGFβ-dependent manner.

  38  

Conclusions

The following conclusions describe the molecular mechanism of TRAF6-

induced TGFβ signal transduction in prostate cancer (Figure 9):

• TGFβ causes TRAF6 activation.

• TGFβ-induced TRAF6 activation causes Lys63-polyubiquitination of

TβRI.

• TGFβ induces activation of PKCζ, which in turn activates TACE.

• TRAF6 polyubiquitinates and activates PS1.

• TβRI is cleaved at the ectodomain region by TACE and at the

transmembrane region by PS1, forming TβRI-ICD.

• TβRI-ICD translocates to the nucleus, where it binds p300 and interacts

with NICD.

• TβRI-ICD promotes the expressions of Snail1, Jag1, Notch1, MMP2,

p300, and TβRI.

• TRAF6 ubiquitinates TβRI at lysine 178.

• Lys63-linked polyubiquitination of TβRI promotes cell cycle regulation

in human prostate cancer cells.

• TβRI ubiquitination and TβRI-ICD nuclear translocation promotes

invasion in prostate cancer cells, lung carcinoma, and breast carcinoma

cells.

• TβRI-ICD nuclear translocation occurs in various human cancers, but

not in normal prostate epithelial cells.

• In vivo experiments in a prostate cancer xenograft mouse model

showed that γ-secretase inhibitor treatment reduced tumor growth and

inhibited TβRI-ICD formation.

• Snail1 regulates the expressions of TβRI and HES1.

  39  

Figure 9: Illustrated summary of Papers I, II, and III

  40  

Future perspectives

Prostate cancer is the most prevalent non-cutaneous cancer among men. Its

multistage oncogenesis includes sustained proliferative signaling, evasion of

apoptosis, growth suppressor resistance, uncontrolled replicative potential,

sustained angiogenesis, and activation of tissue invasion and metastasis.

These cancer hallmarks are caused by genetic or epigenetic modifications in

tumor cells. Moreover, the tumor microenvironment contributes to prostate

cancer tumorigenesis. Heterotypic cellular signaling also plays a predominant

role in the oncogenic process (Hanahan and Weinberg 2000; Hanahan and

Weinberg 2011; Scheller, Chalaris et al. 2011; Felgueiras, Silva et al. 2014).

TGFβ acts as both a tumor promoter and a tumor suppressor. The present

thesis elucidates a molecular mechanism whereby TGFβ promotes prostate

cancer oncogenesis. We identified a novel TGFβ-induced non-Smad signaling

pathway in cancerous cells in which TRAF6 causes Lys63-linked TβRI

polyubiquitination, promoting TβRI proteolytic cleavage by TACE and PS1,

which releases TβRI-ICD. This liberated TβRI-ICD translocates to the

nucleus in cancer cells, where it induces expressions of its own gene and the

EMT markers Snail1 and MMP2, thus promoting invasion in cancer cells.

Moreover, we report that TβRI-ICD associates with NICD in a TGFβ-

dependent manner. Site-directed mutagenesis of TβRI at the TACE and/or

PS1 cleavage site inhibited its nuclear translocation. Experiments in a prostate

cancer xenograft mouse model further revealed that treatment with a γ-

secretase inhibitor (DBZ) reduced tumor growth and TβRI-ICD formation.

We also identified Lys178 as the acceptor lysine in TβRI for TGFβ-induced

Lys63-linked polyubiquitination by TRAF6. Mutagenesis of lysine 178 in

  41  

TβRI to arginine resulted in failure to form TβRI-ICD and also prevented

invasion in prostate cancer cells (Mu, Sundar et al. 2011; Gudey, Sundar et al.

2014; Sundar, Gudey et al. 2015).

The non-Smad TGFβ signaling pathway that leads to TβRI-ICD formation

may be a positive feed-back mechanism in cancer cells, in parallel with the

classical Smad signaling cascade. Our findings suggest that this occurs in a

TβR kinase-independent manner, which also promotes autoregulation of TβRI

expression. This may enable cancer cells to be hyper-responsive to specific

TGFβ signals. TβRI-ICD is not a classical transcription factor, and it remains

unclear whether it directly binds to DNA or if it acts as a co-transcription

factor. It will be important to identify any possible direct transcriptional

targets of TβRI-ICD, and to determine whether TβRI-ICD regulates genes

involved in regulating cell proliferation. It will also be important to determine

where TβRI-ICD binds to the Snail1 and TβRI promoters.

Although we have detected TβRI-ICD nuclear translocation in cancer cells, it

is remains unknown how TβRI-ICD translocates to the nucleus? Full-length

TβRI reportedly translocates to the nucleus with the help of importin β1

(Chandra, Zang et al. 2012). Future investigations should examine whether

TβRI-ICD uses a similar mode of translocation to the nucleus.

Histopathological examination of cancer specimens is a specific and

conventional diagnostic method for grading prostate cancer (Felgueiras, Silva

et al. 2014). It would be interesting to investigate a possible correlation

between TβRI-ICD expression and Gleason score in prostate cancer tissue.

A major challenge in prostate cancer diagnosis is the difficulty of predicting

the disease outcome prior to metastasis, and the existence of only limited

  42  

clinical symptoms. Elucidation of better predictive biomarkers will help to

provide better prognostic information and the best treatment options for

individuals.

Finally, it will also be interesting to analyze the possible effects of targeting

TRAF6 to prevent TβRI ubiquitination and cleavage. Future studies should

also examine tumor tissues for possible TβRI mutations—for example, on the

amino acid residues  G120, VI129, and K178, which we found to be important

for activation of non-canonical signals.

  43  

Acknowledgements

This thesis is not only a culmination of my work at the keyboard—it

represents a milestone of five years of my work at the Department of

Medical Biosciences, Umeå University. During this time, my studies have

been financially supported by grants from Umeå University, The Swedish

Cancer Society, ALF-VLL, The Kempe Foundation, The Knut and Alice

Wallenberg Foundation, and The Cancer Research Foundation in Northern

Sweden: Lion’s Cancer Research Foundation.

I have not travelled alone on this journey to obtain my Ph.D. Numerous others

have supported and encouraged me along the way—including my family,

colleagues, and friends. I am grateful to everyone who made this thesis

possible, and I would especially like to thank the following people.

First and foremost, I give my heartfelt thanks to my supervisor, Prof. Marene

Landström. You have been motivating, encouraging, and enlightening. Your

patience, flexibility, care, and faith in me have enabled my completion of this

dissertation journey. You have been a tremendous mentor to me, and your

scientific advice has been priceless. I am forever grateful to you for helping to

make my Ph.D. studies enjoyable.

I would also like to give sincere thanks to my co-supervisor, Prof. Carl-

Henrik Heldin. You have supported me since the day I began as a student at

the Ludwig Institute of Cancer Research, Uppsala. You are a true inspiration

to all aspiring scientists. Thank you particularly for your encouragement

during our scientific discussions.

  44  

My special thanks to all of the people at the Ludwig Institute for Cancer

Research, the Rudbeck lab, and the Department of Medical Biosciences.

Thank you, Noopur, for your initial guidance in the lab when I first arrived as

a student. Ola, thank you for teaching me mutagenesis and for taking such

good care of me.

I cannot complete this thesis without expressing my gratitude to Annika

Hermansson. Thank you for your great support. You are one of the nicest

people I have met in Uppsala.

I have also been very happy to get to know and collaborate with many

members of the Landström group at Ludwig and Umeå. Thanks to Yabing,

Maria Ekman, Mariya, Marianna, Verena, Ihor, Linda, Guangxiang,

Stina, Anders, Pramod, Anahita, Alexei, Chunyan Li, and Lena.

Song Jie, Gizem, Maria, and Kartik, I have greatly enjoyed sharing an

office with you. Thank you for the many nice conversations.

I would also like to acknowledge Prof. Aris Moustakas and his group

members for our interactions during journal club and scientific presentations

at Ludwig.

Special thanks to Prof. Anders Bergh, Pernilla Wikström, Anders

Widmark, Emma Persson, and all others who participated in prostate cancer

meetings.

I am grateful to Prof. Karin Nylander and Prof. Gunilla Olivercrona for

their encouragement.

  45  

I also thank the members of Bengt Westermark, and Karin Forsberg Nilsson’s

group at Rudbeck. Thank you, Umash, for helping with GFP plasmid

construction. Demet and Soumi, I give special thanks to you both for the

friendly chats and assistance.

Thanks to Maria Brattsand, Maria Nilson, Susanne Gidlund, Pernilla

Andersson, Hinnerk, Erik, Sofia, Xingru, Camilla, Zhara, Ela, Lee Ann,

and Isil for all of the nice conversations.

I also recognize the support and help from Karin Boden, Carina Ahlgren,

Clara, and Clas Wikström. Terry Persson and Åsa Lundsten, I have been

privileged to get to know you. Thank you for all your great help.

Anika Jahns, there are no words that can sufficiently express my gratitude

and appreciation to you. You have been there whenever I have needed any

help. It was my greatest pleasure to meet you. I would also like to thank your

mom, dad, and grandmother.

Thanks to our special friends Ravi, Vidya, and the little cutie Kanasu for the

wonderful time that we spent together. I give my heartfelt acknowledgement

for the friendly help that you have given at all times.

Thank to the Umeå malayalees, Sajna, Sinosh, Remya, Sujith, Nisha and

Philge, for the great times we’ve had together. Maria and Edvin, I give my

heartfelt thanks for your caring and help.

I further wish to acknowledge a friendship that began at Uppsala. Navya,

you’re a special person. Thanks also to Nisha, Rajesh, Nimesh, and

  46  

Sandeep. Smitha, I also appreciate the time that you spent helping me with

cloning.

I would also like to extend warm thanks to Sharvani, Munender, and

Aadyom, Aunbhav, Dinesh, Kamla, Vijay, Uma, and Bala.

I also recognize the support and help from Shyam’s childhood buddies and

their family: Jyothishwar, Harika, Karthik, Diana, Harish, Ratna-Sree,

Uday, Saroja, and Kushru.

Thanks to all of my teachers from school to university. Especially, I must say

thank you to Manoj Narayanan, Santhosh Kumar, and Gopa Kumar for

your unconditional support. I am not sure that I would be writing this thesis

without your encouragement.

Finally, I thank my family: my Mother, Sister, Brother, Brother-in-law,

Kashy, Bhagya, Uncle, Aunty, Bhaskar, and Sandeep. You have always

been beside me offering your support during both the happy and hard

moments.

I have no adequate words to express my thanks for the support, care, and love

I have recieved from my Father. You gave me the strength to reach for the

stars and to chase my dream, but today I must look for you in the stars.

A journey is easier when you travel together. I owe everything to you,

Shyam. During my time with you, I have been happier, stronger, and more

content. Through your love, support, patience, and unwavering belief in me, I

have been able to complete this Ph.D. journey. To our sweet little angel,

Ameyaa, our love, thank you for being an adaptable daughter.

  47  

References

http://www.cancerresearchuk.org/cancerinfo/cancerstats/types/prostate/incidence/#age

Abdollah,  S.,  M.  Macias-­‐Silva,  T.  Tsukazaki,  H.  Hayashi,  L.  Attisano  and  J.  L.  Wrana  (1997).  "TbetaRI  phosphorylation  of  Smad2  on  Ser465  and  Ser467  is  required  for  Smad2-­‐Smad4   complex   formation   and   signaling."   J   Biol   Chem  272(44):   27678-­‐27685.  Andersson,  E.  R.,  R.  Sandberg  and  U.  Lendahl  (2011).  "Notch  signaling:  simplicity  in  design,  versatility  in  function."  Development  138(17):  3593-­‐3612.  Annes,  J.  P.,  J.  S.  Munger  and  D.  B.  Rifkin  (2003).  "Making  sense  of  latent  TGFbeta  activation."  J  Cell  Sci  116(Pt  2):  217-­‐224.  Arch,   R.   H.,   R.   W.   Gedrich   and   C.   B.   Thompson   (1998).   "Tumor   necrosis   factor  receptor-­‐associated   factors   (TRAFs)-­‐-­‐a   family  of   adapter  proteins   that   regulates  life  and  death."  Genes  Dev  12(18):  2821-­‐2830.  Ayaz,  F.  and  B.  A.  Osborne  (2014).   "Non-­‐canonical  notch  signaling   in  cancer  and  immunity."  Front  Oncol  4:  345.  Bakin,  A.  V.,  A.  K.  Tomlinson,  N.  A.  Bhowmick,  H.  L.  Moses  and  C.  L.  Arteaga  (2000).  "Phosphatidylinositol   3-­‐kinase   function   is   required   for   transforming   growth  factor  beta-­‐mediated   epithelial   to  mesenchymal   transition   and   cell  migration."   J  Biol  Chem  275(47):  36803-­‐36810.  Bayer,  P.,  A.  Arndt,  S.  Metzger,  R.  Mahajan,  F.  Melchior,  R.   Jaenicke  and   J.  Becker  (1998).   "Structure   determination   of   the   small   ubiquitin-­‐related  modifier   SUMO-­‐1."  J  Mol  Biol  280(2):  275-­‐286.  Berra,  E.,  M.  T.  Diaz-­‐Meco,  J.  Lozano,  S.  Frutos,  M.  M.  Municio,  P.  Sanchez,  L.  Sanz  and   J.   Moscat   (1995).   "Evidence   for   a   role   of   MEK   and   MAPK   during   signal  transduction  by  protein  kinase  C  zeta."  EMBO  J  14(24):  6157-­‐6163.  Bi,  P.  and  S.  Kuang  (2015).  "Notch  signaling  as  a  novel  regulator  of  metabolism."  Trends  Endocrinol  Metab  26(5):  248-­‐255.  Black,   R.   A.   (2002).   "Tumor   necrosis   factor-­‐alpha   converting   enzyme."   Int   J  Biochem  Cell  Biol  34(1):  1-­‐5.  Blobel,   C.   P.   (1997).   "Metalloprotease-­‐disintegrins:   links   to   cell   adhesion   and  cleavage  of  TNF  alpha  and  Notch."  Cell  90(4):  589-­‐592.  Blokzijl,   A.,   C.   Dahlqvist,   E.   Reissmann,   A.   Falk,   A.  Moliner,   U.   Lendahl   and   C.   F.  Ibanez   (2003).   "Cross-­‐talk  between   the  Notch   and  TGF-­‐beta   signaling  pathways  mediated  by  interaction  of  the  Notch  intracellular  domain  with  Smad3."  J  Cell  Biol  163(4):  723-­‐728.  Borden,  K.  L.,  J.  M.  Lally,  S.  R.  Martin,  N.  J.  O'Reilly,  L.  D.  Etkin  and  P.  S.  Freemont  (1995).   "Novel   topology   of   a   zinc-­‐binding   domain   from   a   protein   involved   in  regulating  early  Xenopus  development."  EMBO  J  14(23):  5947-­‐5956.  Bradley,   J.  R.   and   J.   S.  Pober   (2001).   "Tumor  necrosis   factor   receptor-­‐associated  factors  (TRAFs)."  Oncogene  20(44):  6482-­‐6491.  Brou,  C.,  F.  Logeat,  N.  Gupta,  C.  Bessia,  O.  LeBail,  J.  R.  Doedens,  A.  Cumano,  P.  Roux,  R.  A.  Black  and  A.   Israel   (2000).   "A  novel  proteolytic  cleavage   involved   in  Notch  signaling:   the   role  of   the  disintegrin-­‐metalloprotease  TACE."  Mol  Cell  5(2):  207-­‐216.  

  48  

Brunkan,  A.  L.  and  A.  M.  Goate  (2005).  "Presenilin  function  and  gamma-­‐secretase  activity."  J  Neurochem  93(4):  769-­‐792.  Cao,   Z.,   J.   Xiong,  M.   Takeuchi,   T.   Kurama   and  D.   V.   Goeddel   (1996).   "TRAF6   is   a  signal  transducer  for  interleukin-­‐1."  Nature  383(6599):  443-­‐446.  Chandra,  M.,  S.  Zang,  H.  Li,  L.  J.  Zimmerman,  J.  Champer,  A.  Tsuyada,  A.  Chow,  W.  Zhou,  Y.  Yu,  H.  Gao,  X.  Ren,  R.  J.  Lin  and  S.  E.  Wang  (2012).  "Nuclear  translocation  of  type  I  transforming  growth  factor  beta  receptor  confers  a  novel  function  in  RNA  processing."  Mol  Cell  Biol  32(12):  2183-­‐2195.  Cutts,   A.   J.,   S.  M.   Soond,   S.   Powell   and  A.   Chantry   (2011).   "Early   phase   TGFbeta  receptor   signalling   dynamics   stabilised   by   the   deubiquitinase   UCH37   promotes  cell  migratory  responses."  Int  J  Biochem  Cell  Biol  43(4):  604-­‐612.  De  Boeck,  M.  and  P.  ten  Dijke  (2012).  "Key  role  for  ubiquitin  protein  modification  in  TGFbeta  signal  transduction."  Ups  J  Med  Sci  117(2):  153-­‐165.  de   Caestecker,   M.   P.,   E.   Piek   and   A.   B.   Roberts   (2000).   "Role   of   transforming  growth  factor-­‐beta  signaling  in  cancer."  J  Natl  Cancer  Inst  92(17):  1388-­‐1402.  De  Strooper,  B.  and  W.  Annaert  (2010).  "Novel  research  horizons  for  presenilins  and   gamma-­‐secretases   in   cell   biology   and   disease."   Annu  Rev   Cell   Dev   Biol  26:  235-­‐260.  De  Strooper,  B.,  M.  Beullens,  B.  Contreras,  L.  Levesque,  K.  Craessaerts,  B.  Cordell,  D.  Moechars,  M.  Bollen,  P.  Fraser,  P.  S.  George-­‐Hyslop  and  F.  Van  Leuven  (1997).  "Phosphorylation,   subcellular   localization,   and   membrane   orientation   of   the  Alzheimer's  disease-­‐associated  presenilins."  J  Biol  Chem  272(6):  3590-­‐3598.  Deng,  L.,  C.  Wang,  E.  Spencer,  L.  Yang,  A.  Braun,  J.  You,  C.  Slaughter,  C.  Pickart  and  Z.  J.  Chen  (2000).  "Activation  of  the  IkappaB  kinase  complex  by  TRAF6  requires  a  dimeric  ubiquitin-­‐conjugating  enzyme  complex  and  a  unique  polyubiquitin  chain."  Cell  103(2):  351-­‐361.  Derynck,   R.   and   Y.   E.   Zhang   (2003).   "Smad-­‐dependent   and   Smad-­‐independent  pathways  in  TGF-­‐beta  family  signalling."  Nature  425(6958):  577-­‐584.  Dhasarathy,   A.,   D.   Phadke,   D.   Mav,   R.   R.   Shah   and   P.   A.   Wade   (2011).   "The  transcription  factors  Snail  and  Slug  activate  the  transforming  growth  factor-­‐beta  signaling  pathway  in  breast  cancer."  PLoS  One  6(10):  e26514.  Diener,  K.  R.,  E.  F.  Need,  G.  Buchanan  and  J.  D.  Hayball  (2010).  "TGF-­‐beta  signalling  and  immunity  in  prostate  tumourigenesis."  Expert  Opin  Ther  Targets  14(2):  179-­‐192.  Dikic,   I.   and  M.  Robertson   (2012).   "Ubiquitin   ligases  and  beyond."  BMC  Biol  10:  22.  Dikic,  I.,  S.  Wakatsuki  and  K.  J.  Walters  (2009).  "Ubiquitin-­‐binding  domains  -­‐  from  structures  to  functions."  Nat  Rev  Mol  Cell  Biol  10(10):  659-­‐671.  Edbauer,   D.,   E.  Winkler,   J.   T.   Regula,   B.   Pesold,   H.   Steiner   and   C.   Haass   (2003).  "Reconstitution  of  gamma-­‐secretase  activity."  Nat  Cell  Biol  5(5):  486-­‐488.  Emmerich,   C.   H.,   A.   C.   Schmukle   and  H.  Walczak   (2011).   "The   emerging   role   of  linear  ubiquitination  in  cell  signaling."  Sci  Signal  4(204):  re5.  Esfahani,  M.,   N.   Ataei   and  M.   Panjehpour   (2015).   "Biomarkers   for   Evaluation   of  Prostate  Cancer  Prognosis."  Asian  Pac  J  Cancer  Prev  16(7):  2601-­‐2611.  Esler,  W.  P.,  W.  T.  Kimberly,  B.  L.  Ostaszewski,  T.  S.  Diehl,  C.  L.  Moore,  J.  Y.  Tsai,  T.  Rahmati,  W.  Xia,  D.   J.   Selkoe   and  M.   S.  Wolfe   (2000).   "Transition-­‐state   analogue  inhibitors   of   gamma-­‐secretase   bind   directly   to   presenilin-­‐1."   Nat   Cell   Biol  2(7):  428-­‐434.  

  49  

Felgueiras,   J.,   J.   V.   Silva   and  M.   Fardilha   (2014).   "Prostate   cancer:   the   need   for  biomarkers  and  new  therapeutic  targets."  J  Zhejiang  Univ  Sci  B  15(1):  16-­‐42.  Franzen,   P.,   C.   H.   Heldin   and   K.   Miyazono   (1995).   "The   GS   domain   of   the  transforming   growth   factor-­‐beta   type   I   receptor   is   important   in   signal  transduction."  Biochem  Biophys  Res  Commun  207(2):  682-­‐689.  Fukuchi,   M.,   T.   Imamura,   T.   Chiba,   T.   Ebisawa,   M.   Kawabata,   K.   Tanaka   and   K.  Miyazono  (2001).  "Ligand-­‐dependent  degradation  of  Smad3  by  a  ubiquitin  ligase  complex  of  ROC1  and  associated  proteins."  Mol  Biol  Cell  12(5):  1431-­‐1443.  Fuxe,   J.,   T.  Vincent   and  A.  Garcia  de  Herreros   (2010).   "Transcriptional   crosstalk  between   TGF-­‐beta   and   stem   cell   pathways   in   tumor   cell   invasion:   role   of   EMT  promoting  Smad  complexes."  Cell  Cycle  9(12):  2363-­‐2374.  Geiss-­‐Friedlander,  R.  and  F.  Melchior  (2007).  "Concepts  in  sumoylation:  a  decade  on."  Nat  Rev  Mol  Cell  Biol  8(12):  947-­‐956.  Griner,   E.   M.   and   M.   G.   Kazanietz   (2007).   "Protein   kinase   C   and   other  diacylglycerol  effectors  in  cancer."  Nat  Rev  Cancer  7(4):  281-­‐294.  Groettrup,   M.,   C.   Pelzer,   G.   Schmidtke   and   K.   Hofmann   (2008).   "Activating   the  ubiquitin  family:  UBA6  challenges  the  field."  Trends  Biochem  Sci  33(5):  230-­‐237.  Gudey,  S.  K.,  R.  Sundar,  Y.  Mu,  A.  Wallenius,  G.  Zang,  A.  Bergh,  C.  H.  Heldin  and  M.  Landstrom   (2014).   "TRAF6   stimulates   the   tumor-­‐promoting   effects   of   TGFbeta  type   I   receptor   through   polyubiquitination   and   activation   of   presenilin   1."   Sci  Signal  7(307):  ra2.  Gudey,   S.  K.,  A.  Wallenius  and  M.  Landstrom  (2014).   "Regulated   intramembrane  proteolysis   of   the   TGFbeta   type   I   receptor   conveys   oncogenic   signals."   Future  Oncol  10(11):  1853-­‐1861.  Hanahan,  D.  and  R.  A.  Weinberg   (2000).   "The  hallmarks  of   cancer."  Cell  100(1):  57-­‐70.  Hanahan,   D.   and   R.   A.   Weinberg   (2011).   "Hallmarks   of   cancer:   the   next  generation."  Cell  144(5):  646-­‐674.  Hannoun,   Z.,   S.   Greenhough,   E.   Jaffray,   R.   T.   Hay   and   D.   C.   Hay   (2010).   "Post-­‐translational  modification  by  SUMO."  Toxicology  278(3):  288-­‐293.  Hata,  A.,  G.  Lagna,  J.  Massague  and  A.  Hemmati-­‐Brivanlou  (1998).  "Smad6  inhibits  BMP/Smad1   signaling   by   specifically   competing   with   the   Smad4   tumor  suppressor."  Genes  Dev  12(2):  186-­‐197.  Hayashi,  H.,  S.  Abdollah,  Y.  Qiu,  J.  Cai,  Y.  Y.  Xu,  B.  W.  Grinnell,  M.  A.  Richardson,  J.  N.  Topper,   M.   A.   Gimbrone,   Jr.,   J.   L.  Wrana   and   D.   Falb   (1997).   "The  MAD-­‐related  protein   Smad7   associates   with   the   TGFbeta   receptor   and   functions   as   an  antagonist  of  TGFbeta  signaling."  Cell  89(7):  1165-­‐1173.  Hayashida,   K.,   A.   H.   Bartlett,   Y.   Chen   and   P.   W.   Park   (2010).   "Molecular   and  cellular  mechanisms  of  ectodomain  shedding."  Anat  Rec  (Hoboken)  293(6):  925-­‐937.  Heldin,   C.   H.,  M.   Landstrom   and   A.  Moustakas   (2009).   "Mechanism   of   TGF-­‐beta  signaling   to   growth   arrest,   apoptosis,   and   epithelial-­‐mesenchymal   transition."  Curr  Opin  Cell  Biol  21(2):  166-­‐176.  Heldin,  C.  H.,  M.  Vanlandewijck  and  A.  Moustakas  (2012).  "Regulation  of  EMT  by  TGFbeta  in  cancer."  FEBS  Lett  586(14):  1959-­‐1970.  Hesling,  C.,  L.  Fattet,  G.  Teyre,  D.  Jury,  P.  Gonzalo,  J.  Lopez,  C.  Vanbelle,  A.  P.  Morel,  G.  Gillet,   I.  Mikaelian  and  R.  Rimokh   (2011).   "Antagonistic   regulation  of  EMT  by  TIF1gamma  and  Smad4  in  mammary  epithelial  cells."  EMBO  Rep  12(7):  665-­‐672.  

  50  

Hochstrasser,  M.   (2009).   "Origin  and   function  of  ubiquitin-­‐like  proteins."  Nature  458(7237):  422-­‐429.  Huse,  M.,  Y.  G.  Chen,   J.  Massague  and  J.  Kuriyan  (1999).   "Crystal  structure  of   the  cytoplasmic  domain  of  the  type  I  TGF  beta  receptor  in  complex  with  FKBP12."  Cell  96(3):  425-­‐436.  Hussein,   S.,   S.   Satturwar   and   T.   Van   der   Kwast   (2015).   "Young-­‐age   prostate  cancer."  J  Clin  Pathol.  Imamura,   T.,   M.   Takase,   A.   Nishihara,   E.   Oeda,   J.   Hanai,   M.   Kawabata   and   K.  Miyazono   (1997).   "Smad6   inhibits   signalling   by   the   TGF-­‐beta   superfamily."  Nature  389(6651):  622-­‐626.  Inoue,   J.,   T.   Ishida,   N.   Tsukamoto,   N.   Kobayashi,   A.   Naito,   S.   Azuma   and   T.  Yamamoto   (2000).   "Tumor   necrosis   factor   receptor-­‐associated   factor   (TRAF)  family:   adapter   proteins   that  mediate   cytokine   signaling."   Exp   Cell   Res  254(1):  14-­‐24.  Ishida,  T.,  S.  Mizushima,  S.  Azuma,  N.  Kobayashi,  T.  Tojo,  K.  Suzuki,  S.  Aizawa,  T.  Watanabe,  G.  Mosialos,  E.  Kieff,  T.  Yamamoto  and  J.   Inoue  (1996).  "Identification  of   TRAF6,   a   novel   tumor   necrosis   factor   receptor-­‐associated   factor   protein   that  mediates   signaling   from   an   amino-­‐terminal   domain   of   the   CD40   cytoplasmic  region."  J  Biol  Chem  271(46):  28745-­‐28748.  Jones,  E.,  H.  Pu  and  N.  Kyprianou  (2009).  "Targeting  TGF-­‐beta  in  prostate  cancer:  therapeutic   possibilities   during   tumor   progression."   Expert   Opin   Ther   Targets  13(2):  227-­‐234.  Kang,  J.  S.,  C.  Liu  and  R.  Derynck  (2009).  "New  regulatory  mechanisms  of  TGF-­‐beta  receptor  function."  Trends  Cell  Biol  19(8):  385-­‐394.  Kaufhold,   S.   and  B.  Bonavida   (2014).   "Central   role  of   Snail1   in   the   regulation  of  EMT   and   resistance   in   cancer:   a   target   for   therapeutic   intervention."   J   Exp   Clin  Cancer  Res  33:  62.  Kenny,   P.   A.   and   M.   J.   Bissell   (2007).   "Targeting   TACE-­‐dependent   EGFR   ligand  shedding  in  breast  cancer."  J  Clin  Invest  117(2):  337-­‐345.  Keski-­‐Oja,   J.,   K.   Koli   and   H.   von   Melchner   (2004).   "TGF-­‐beta   activation   by  traction?"  Trends  Cell  Biol  14(12):  657-­‐659.  Kim,  J.  Y.,  D.  Y.  Park,  G.  H.  Kim,  K.  U.  Choi,  C.  H.  Lee,  G.  Y.  Huh,  M.  Y.  Sol,  G.  A.  Song,  T.  Y.  Jeon,  D.  H.  Kim  and  M.  S.  Sim  (2005).  "Smad4  expression  in  gastric  adenoma  and  adenocarcinoma:   frequent   loss   of   expression   in   diffuse   type   of   gastric  adenocarcinoma."  Histol  Histopathol  20(2):  543-­‐549.  Kim,  N.  H.,  H.  S.  Kim,  X.  Y.  Li,  I.  Lee,  H.  S.  Choi,  S.  E.  Kang,  S.  Y.  Cha,  J.  K.  Ryu,  D.  Yoon,  E.  R.  Fearon,  R.  G.  Rowe,  S.  Lee,  C.  A.  Maher,  S.   J.  Weiss  and   J.   I.  Yook  (2011).   "A  p53/miRNA-­‐34   axis   regulates   Snail1-­‐dependent   cancer   cell   epithelial-­‐mesenchymal  transition."  J  Cell  Biol  195(3):  417-­‐433.  Kobayashi,   T.,   M.   C.   Walsh   and   Y.   Choi   (2004).   "The   role   of   TRAF6   in   signal  transduction  and  the  immune  response."  Microbes  Infect  6(14):  1333-­‐1338.  Koivunen,  J.,  V.  Aaltonen  and  J.  Peltonen  (2006).  "Protein  kinase  C  (PKC)  family  in  cancer  progression."  Cancer  Lett  235(1):  1-­‐10.  Koli,  K.,  J.  Saharinen,  M.  Hyytiainen,  C.  Penttinen  and  J.  Keski-­‐Oja  (2001).  "Latency,  activation,  and  binding  proteins  of  TGF-­‐beta."  Microsc  Res  Tech  52(4):  354-­‐362.  Komander,   D.   (2009).   "The   emerging   complexity   of   protein   ubiquitination."  Biochem  Soc  Trans  37(Pt  5):  937-­‐953.  Kravtsova-­‐Ivantsiv,  Y.  and  A.  Ciechanover  (2012).  "Non-­‐canonical  ubiquitin-­‐based  signals  for  proteasomal  degradation."  J  Cell  Sci  125(Pt  3):  539-­‐548.  

  51  

Kretzschmar,  M.,  F.  Liu,  A.  Hata,   J.  Doody  and  J.  Massague  (1997).  "The  TGF-­‐beta  family  mediator  Smad1   is  phosphorylated  directly   and  activated   functionally  by  the  BMP  receptor  kinase."  Genes  Dev  11(8):  984-­‐995.  Landstrom,  M.  (2010).  "The  TAK1-­‐TRAF6  signalling  pathway."  Int  J  Biochem  Cell  Biol  42(5):  585-­‐589.  Lawler,  S.,  X.  H.  Feng,  R.  H.  Chen,  E.  M.  Maruoka,  C.  W.  Turck,  I.  Griswold-­‐Prenner  and   R.   Derynck   (1997).   "The   type   II   transforming   growth   factor-­‐beta   receptor  autophosphorylates   not   only   on   serine   and   threonine   but   also   on   tyrosine  residues."  J  Biol  Chem  272(23):  14850-­‐14859.  Lee,  M.  K.,  C.  Pardoux,  M.  C.  Hall,  P.  S.  Lee,  D.  Warburton,  J.  Qing,  S.  M.  Smith  and  R.  Derynck   (2007).   "TGF-­‐beta   activates   Erk   MAP   kinase   signalling   through   direct  phosphorylation  of  ShcA."  EMBO  J  26(17):  3957-­‐3967.  Li,  J.,  A.  M.  Pauley,  R.  L.  Myers,  R.  Shuang,  J.  R.  Brashler,  R.  Yan,  A.  E.  Buhl,  C.  Ruble  and   M.   E.   Gurney   (2002).   "SEL-­‐10   interacts   with   presenilin   1,   facilitates   its  ubiquitination,  and  alters  A-­‐beta  peptide  production."  J  Neurochem  82(6):  1540-­‐1548.  Li,  Y.  M.,  M.  Xu,  M.  T.  Lai,  Q.  Huang,  J.  L.  Castro,  J.  DiMuzio-­‐Mower,  T.  Harrison,  C.  Lellis,  A.  Nadin,  J.  G.  Neduvelil,  R.  B.  Register,  M.  K.  Sardana,  M.  S.  Shearman,  A.  L.  Smith,   X.   P.   Shi,   K.   C.   Yin,   J.   A.   Shafer   and   S.   J.   Gardell   (2000).   "Photoactivated  gamma-­‐secretase  inhibitors  directed  to  the  active  site  covalently  label  presenilin  1."  Nature  405(6787):  689-­‐694.  Lindsey,  S.  and  S.  A.  Langhans  (2014).  "Crosstalk  of  Oncogenic  Signaling  Pathways  during  Epithelial-­‐Mesenchymal  Transition."  Front  Oncol  4:  358.  Liu,   C.,   P.   Xu,   S.   Lamouille,   J.   Xu   and   R.   Derynck   (2009).   "TACE-­‐mediated  ectodomain   shedding   of   the   type   I   TGF-­‐beta   receptor   downregulates   TGF-­‐beta  signaling."  Mol  Cell  35(1):  26-­‐36.  Liu,  H.,  Y.  C.  Su,  E.  Becker,  J.  Treisman  and  E.  Y.  Skolnik  (1999).  "A  Drosophila  TNF-­‐receptor-­‐associated  factor  (TRAF)  binds  the  ste20  kinase  Misshapen  and  activates  Jun  kinase."  Curr  Biol  9(2):  101-­‐104.  Liu,  X.,  A.  E.  Elia,  S.  F.  Law,  E.  A.  Golemis,   J.  Farley  and  T.  Wang  (2000).  "A  novel  ability   of   Smad3   to   regulate   proteasomal   degradation   of   a   Cas   family   member  HEF1."  EMBO  J  19(24):  6759-­‐6769.  Lomaga,  M.  A.,  W.  C.  Yeh,  I.  Sarosi,  G.  S.  Duncan,  C.  Furlonger,  A.  Ho,  S.  Morony,  C.  Capparelli,  G.  Van,  S.  Kaufman,  A.  van  der  Heiden,  A.  Itie,  A.  Wakeham,  W.  Khoo,  T.  Sasaki,  Z.  Cao,  J.  M.  Penninger,  C.  J.  Paige,  D.  L.  Lacey,  C.  R.  Dunstan,  W.  J.  Boyle,  D.  V.  Goeddel  and  T.  W.  Mak  (1999).  "TRAF6  deficiency  results  in  osteopetrosis  and  defective  interleukin-­‐1,  CD40,  and  LPS  signaling."  Genes  Dev  13(8):  1015-­‐1024.  Massague,  J.  (1998).  "TGF-­‐beta  signal  transduction."  Annu  Rev  Biochem  67:  753-­‐791.  Massague,  J.  (2012).  "TGFbeta  signalling  in  context."  Nat  Rev  Mol  Cell  Biol  13(10):  616-­‐630.  McCarthy,  J.  V.,  C.  Twomey  and  P.  Wujek  (2009).  "Presenilin-­‐dependent  regulated  intramembrane   proteolysis   and   gamma-­‐secretase   activity."   Cell   Mol   Life   Sci  66(9):  1534-­‐1555.  Medina,  M.  and  C.  G.  Dotti  (2003).  "RIPped  out  by  presenilin-­‐dependent  gamma-­‐secretase."  Cell  Signal  15(9):  829-­‐841.  Mohan,  M.  J.,  T.  Seaton,  J.  Mitchell,  A.  Howe,  K.  Blackburn,  W.  Burkhart,  M.  Moyer,  I.  Patel,  G.  M.  Waitt,   J.  D.  Becherer,  M.  L.  Moss  and  M.  E.  Milla   (2002).   "The   tumor  

  52  

necrosis   factor-­‐alpha   converting   enzyme   (TACE):   a   unique   metalloproteinase  with  highly  defined  substrate  selectivity."  Biochemistry  41(30):  9462-­‐9469.  Mohr,  O.  L.  (1919).  "Character  Changes  Caused  by  Mutation  of  an  Entire  Region  of  a  Chromosome  in  Drosophila."  Genetics  4(3):  275-­‐282.  Moren,  A.,  U.  Hellman,  Y.  Inada,  T.  Imamura,  C.  H.  Heldin  and  A.  Moustakas  (2003).  "Differential  ubiquitination  defines  the  functional  status  of  the  tumor  suppressor  Smad4."  J  Biol  Chem  278(35):  33571-­‐33582.  Moses,  H.  L.,  E.  L.  Branum,  J.  A.  Proper  and  R.  A.  Robinson  (1981).  "Transforming  growth   factor   production   by   chemically   transformed   cells."   Cancer   Res   41(7):  2842-­‐2848.  Mosialos,  G.,  M.  Birkenbach,  R.  Yalamanchili,  T.  VanArsdale,  C.  Ware  and  E.  Kieff  (1995).   "The   Epstein-­‐Barr   virus   transforming   protein   LMP1   engages   signaling  proteins  for  the  tumor  necrosis  factor  receptor  family."  Cell  80(3):  389-­‐399.  Mossessova,  E.  and  C.  D.  Lima  (2000).  "Ulp1-­‐SUMO  crystal  structure  and  genetic  analysis  reveal  conserved  interactions  and  a  regulatory  element  essential  for  cell  growth  in  yeast."  Mol  Cell  5(5):  865-­‐876.  Moustakas,   A.   and   C.   H.   Heldin   (2009).   "The   regulation   of   TGFbeta   signal  transduction."  Development  136(22):  3699-­‐3714.  Moustakas,   A.   and   C.   H.   Heldin   (2012).   "Induction   of   epithelial-­‐mesenchymal  transition  by  transforming  growth  factor  beta."  Semin  Cancer  Biol  22(5-­‐6):  446-­‐454.  Moustakas,  A.,  S.  Souchelnytskyi  and  C.  H.  Heldin  (2001).  "Smad  regulation  in  TGF-­‐beta  signal  transduction."  J  Cell  Sci  114(Pt  24):  4359-­‐4369.  Mu,  Y.,  S.  K.  Gudey  and  M.  Landstrom  (2012).  "Non-­‐Smad  signaling  pathways."  Cell  Tissue  Res  347(1):  11-­‐20.  Mu,   Y.,   R.   Sundar,   N.   Thakur,   M.   Ekman,   S.   K.   Gudey,   M.   Yakymovych,   A.  Hermansson,  H.  Dimitriou,  M.  T.  Bengoechea-­‐Alonso,  J.  Ericsson,  C.  H.  Heldin  and  M.  Landstrom  (2011).  "TRAF6  ubiquitinates  TGFbeta  type  I  receptor  to  promote  its  cleavage  and  nuclear  translocation  in  cancer."  Nat  Commun  2:  330.  Naito,   A.,   S.   Azuma,   S.   Tanaka,   T.   Miyazaki,   S.   Takaki,   K.   Takatsu,   K.   Nakao,   K.  Nakamura,  M.  Katsuki,  T.   Yamamoto  and   J.   Inoue   (1999).   "Severe  osteopetrosis,  defective   interleukin-­‐1   signalling   and   lymph   node   organogenesis   in   TRAF6-­‐deficient  mice."  Genes  Cells  4(6):  353-­‐362.  Nakao,  A.,  M.  Afrakhte,  A.  Moren,  T.  Nakayama,  J.  L.  Christian,  R.  Heuchel,  S.  Itoh,  M.  Kawabata,  N.  E.  Heldin,  C.  H.  Heldin  and  P.  ten  Dijke  (1997).  "Identification  of  Smad7,   a   TGFbeta-­‐inducible   antagonist   of   TGF-­‐beta   signalling."   Nature  389(6651):  631-­‐635.  Nam,  Y.,  A.  P.  Weng,  J.  C.  Aster  and  S.  C.  Blacklow  (2003).  "Structural  requirements  for   assembly   of   the   CSL.intracellular   Notch1.Mastermind-­‐like   1   transcriptional  activation  complex."  J  Biol  Chem  278(23):  21232-­‐21239.  Netherton,  S.  J.  and  S.  Bonni  (2010).  "Suppression  of  TGFbeta-­‐induced  epithelial-­‐mesenchymal   transition   like   phenotype   by   a   PIAS1   regulated   sumoylation  pathway  in  NMuMG  epithelial  cells."  PLoS  One  5(11):  e13971.  O'Neil,  J.,  J.  Grim,  P.  Strack,  S.  Rao,  D.  Tibbitts,  C.  Winter,  J.  Hardwick,  M.  Welcker,  J.  P.  Meijerink,  R.  Pieters,  G.  Draetta,  R.  Sears,  B.  E.  Clurman  and  A.  T.  Look  (2007).  "FBW7   mutations   in   leukemic   cells   mediate   NOTCH   pathway   activation   and  resistance  to  gamma-­‐secretase  inhibitors."  J  Exp  Med  204(8):  1813-­‐1824.  Okadome,   T.,   H.   Yamashita,   P.   Franzen,   A.  Moren,   C.   H.   Heldin   and  K.  Miyazono  (1994).  "Distinct  roles  of  the  intracellular  domains  of  transforming  growth  factor-­‐

  53  

beta   type   I   and   type   II   receptors   in   signal   transduction."   J   Biol   Chem  269(49):  30753-­‐30756.  Ozdamar,   B.,   R.   Bose,   M.   Barrios-­‐Rodiles,   H.   R.  Wang,   Y.   Zhang   and   J.   L.  Wrana  (2005).   "Regulation   of   the   polarity   protein   Par6   by   TGFbeta   receptors   controls  epithelial  cell  plasticity."  Science  307(5715):  1603-­‐1609.  Padalecki,  S.  S.,  K.  S.  Weldon,  X.  T.  Reveles,  C.  L.  Buller,  B.  Grubbs,  Y.  Cui,  J.  J.  Yin,  D.  C.  Hall,  B.  T.  Hummer,  B.  E.  Weissman,  M.  Dallas,  T.  A.  Guise,  R.  J.  Leach  and  T.  L.  Johnson-­‐Pais   (2003).   "Chromosome   18   suppresses   prostate   cancer  metastases."  Urol  Oncol  21(5):  366-­‐373.  Padua,   D.   and   J.   Massague   (2009).   "Roles   of   TGFbeta   in   metastasis."   Cell   Res  19(1):  89-­‐102.  Papageorgis,   P.   (2015).   "TGFbeta   Signaling   in   Tumor   Initiation,   Epithelial-­‐to-­‐Mesenchymal  Transition,  and  Metastasis."  J  Oncol  2015:  587193.  Peinado,   H.,   D.   Olmeda   and   A.   Cano   (2007).   "Snail,   Zeb   and   bHLH   factors   in  tumour   progression:   an   alliance   against   the   epithelial   phenotype?"   Nat   Rev  Cancer  7(6):  415-­‐428.  Peschon,  J.  J.,  J.  L.  Slack,  P.  Reddy,  K.  L.  Stocking,  S.  W.  Sunnarborg,  D.  C.  Lee,  W.  E.  Russell,   B.   J.   Castner,   R.   S.   Johnson,   J.   N.   Fitzner,   R.   W.   Boyce,   N.   Nelson,   C.   J.  Kozlosky,  M.  F.  Wolfson,  C.  T.  Rauch,  D.  P.  Cerretti,  R.  J.  Paxton,  C.  J.  March  and  R.  A.  Black   (1998).   "An   essential   role   for   ectodomain   shedding   in   mammalian  development."  Science  282(5392):  1281-­‐1284.  Pickart,  C.  M.  (2001).  "Mechanisms  underlying  ubiquitination."  Annu  Rev  Biochem  70:  503-­‐533.  Pierreux,  C.  E.,  F.  J.  Nicolas  and  C.  S.  Hill  (2000).  "Transforming  growth  factor  beta-­‐independent   shuttling   of   Smad4   between   the   cytoplasm   and   nucleus."   Mol   Cell  Biol  20(23):  9041-­‐9054.  Poblete,  C.  E.,   J.  Fulla,  M.  Gallardo,  V.  Munoz,  E.  A.  Castellon,   I.  Gallegos  and  H.  R.  Contreras   (2014).   "Increased   SNAIL   expression   and   low   syndecan   levels   are  associated  with   high  Gleason   grade   in   prostate   cancer."   Int   J   Oncol  44(3):   647-­‐654.  Powell,   J.  C.,  C.  Twomey,  R.   Jain  and  J.  V.  McCarthy  (2009).  "Association  between  Presenilin-­‐1   and   TRAF6  modulates   regulated   intramembrane   proteolysis   of   the  p75NTR  neurotrophin  receptor."  J  Neurochem  108(1):  216-­‐230.  Regnier,  C.  H.,  C.  Tomasetto,  C.  Moog-­‐Lutz,  M.  P.  Chenard,  C.  Wendling,  P.  Basset  and  M.   C.   Rio   (1995).   "Presence   of   a   new   conserved   domain   in   CART1,   a   novel  member  of  the  tumor  necrosis  factor  receptor-­‐associated  protein  family,  which  is  expressed  in  breast  carcinoma."  J  Biol  Chem  270(43):  25715-­‐25721.  Rieser,   E.,   S.  M.   Cordier   and  H.  Walczak   (2013).   "Linear   ubiquitination:   a   newly  discovered  regulator  of  cell  signalling."  Trends  Biochem  Sci  38(2):  94-­‐102.  Roberts,   A.   B.   (1999).   "TGF-­‐beta   signaling   from   receptors   to   the   nucleus."  Microbes  Infect  1(15):  1265-­‐1273.  Roberts,  A.  B.,  L.  C.  Lamb,  D.  L.  Newton,  M.  B.  Sporn,  J.  E.  De  Larco  and  G.  J.  Todaro  (1980).  "Transforming  growth  factors:   isolation  of  polypeptides  from  virally  and  chemically  transformed  cells  by  acid/ethanol  extraction."  Proc  Natl  Acad  Sci  U  S  A  77(6):  3494-­‐3498.  Roberts,  A.  B.  and  L.  M.  Wakefield  (2003).  "The  two  faces  of  transforming  growth  factor  beta  in  carcinogenesis."  Proc  Natl  Acad  Sci  U  S  A  100(15):  8621-­‐8623.  Ross,   S.   and   C.   S.   Hill   (2008).   "How   the   Smads   regulate   transcription."   Int   J  Biochem  Cell  Biol  40(3):  383-­‐408.  

  54  

Rothe,  M.,   S.  C.  Wong,  W.   J.  Henzel   and  D.  V.  Goeddel   (1994).   "A  novel   family  of  putative  signal  transducers  associated  with  the  cytoplasmic  domain  of  the  75  kDa  tumor  necrosis  factor  receptor."  Cell  78(4):  681-­‐692.  Saito,   R.   A.,   T.   Watabe,   K.   Horiguchi,   T.   Kohyama,   M.   Saitoh,   T.   Nagase   and   K.  Miyazono   (2009).   "Thyroid   transcription   factor-­‐1   inhibits   transforming   growth  factor-­‐beta-­‐mediated   epithelial-­‐to-­‐mesenchymal   transition   in   lung  adenocarcinoma  cells."  Cancer  Res  69(7):  2783-­‐2791.  Sancho,  R.,  S.  M.  Blake,  C.  Tendeng,  B.  E.  Clurman,  J.  Lewis  and  A.  Behrens  (2013).  "Fbw7   repression   by   hes5   creates   a   feedback   loop   that   modulates   Notch-­‐mediated   intestinal   and   neural   stem   cell   fate   decisions."   PLoS   Biol   11(6):  e1001586.  Savage,   C.,   P.   Das,   A.   L.   Finelli,   S.   R.   Townsend,   C.   Y.   Sun,   S.   E.   Baird   and   R.  W.  Padgett  (1996).  "Caenorhabditis  elegans  genes  sma-­‐2,  sma-­‐3,  and  sma-­‐4  define  a  conserved  family  of  transforming  growth  factor  beta  pathway  components."  Proc  Natl  Acad  Sci  U  S  A  93(2):  790-­‐794.  Scheller,  J.,  A.  Chalaris,  C.  Garbers  and  S.  Rose-­‐John  (2011).  "ADAM17:  a  molecular  switch  to  control  inflammation  and  tissue  regeneration."  Trends  Immunol  32(8):  380-­‐387.  Schlondorff,   J.   and   C.   P.   Blobel   (1999).   "Metalloprotease-­‐disintegrins:   modular  proteins   capable   of   promoting   cell-­‐cell   interactions   and   triggering   signals   by  protein-­‐ectodomain  shedding."  J  Cell  Sci  112  (  Pt  21):  3603-­‐3617.  Schulman,  B.  A.  and  J.  W.  Harper  (2009).  "Ubiquitin-­‐like  protein  activation  by  E1  enzymes:   the   apex   for   downstream   signalling   pathways."   Nat   Rev  Mol   Cell   Biol  10(5):  319-­‐331.  Sekelsky,  J.  J.,  S.  J.  Newfeld,  L.  A.  Raftery,  E.  H.  Chartoff  and  W.  M.  Gelbart  (1995).  "Genetic  characterization  and  cloning  of  mothers  against  dpp,  a  gene  required  for  decapentaplegic   function   in   Drosophila   melanogaster."   Genetics   139(3):   1347-­‐1358.  Sorrentino,  A.,  N.  Thakur,  S.  Grimsby,  A.  Marcusson,  V.  von  Bulow,  N.  Schuster,  S.  Zhang,   C.   H.   Heldin   and   M.   Landstrom   (2008).   "The   type   I   TGF-­‐beta   receptor  engages  TRAF6  to  activate  TAK1  in  a  receptor  kinase-­‐independent  manner."  Nat  Cell  Biol  10(10):  1199-­‐1207.  Sou,  P.  W.,  N.  C.  Delic,  G.  M.  Halliday  and   J.  G.  Lyons   (2010).   "Snail   transcription  factors  in  keratinocytes:  Enough  to  make  your  skin  crawl."  Int  J  Biochem  Cell  Biol  42(12):  1940-­‐1944.  Souchelnytskyi,   S.,   K.   Tamaki,   U.   Engstrom,   C.  Wernstedt,   P.   ten   Dijke   and   C.   H.  Heldin   (1997).   "Phosphorylation   of   Ser465   and   Ser467   in   the   C   terminus   of  Smad2  mediates  interaction  with  Smad4  and  is  required  for  transforming  growth  factor-­‐beta  signaling."  J  Biol  Chem  272(44):  28107-­‐28115.  Steiner,   M.   S.,   Z.   Z.   Zhou,   D.   C.   Tonb   and   E.   R.   Barrack   (1994).   "Expression   of  transforming   growth   factor-­‐beta   1   in   prostate   cancer."   Endocrinology   135(5):  2240-­‐2247.  Sundar,  R.,  S.  K.  Gudey,  C.  H.  Heldin  and  M.  Landstrom  (2015).  "TRAF6  promotes  TGFbeta-­‐induced   invasion   and   cell-­‐cycle   regulation   via   Lys63-­‐linked  polyubiquitination  of  Lys178   in  TGFbeta   type   I  receptor."  Cell  Cycle  14(4):  554-­‐565.  Talmadge,  J.  E.,  M.  Donkor  and  E.  Scholar  (2007).  "Inflammatory  cell  infiltration  of  tumors:  Jekyll  or  Hyde."  Cancer  Metastasis  Rev  26(3-­‐4):  373-­‐400.  

  55  

Teodorczyk,   M.   and   M.   H.   Schmidt   (2014).   "Notching   on   Cancer's   Door:   Notch  Signaling  in  Brain  Tumors."  Front  Oncol  4:  341.  Thuault,   S.,  E.   J.  Tan,  H.  Peinado,  A.  Cano,  C.  H.  Heldin  and  A.  Moustakas   (2008).  "HMGA2  and  Smads  co-­‐regulate  SNAIL1  expression  during  induction  of  epithelial-­‐to-­‐mesenchymal  transition."  J  Biol  Chem  283(48):  33437-­‐33446.  Vetrivel,   K.   S.,   Y.  W.   Zhang,   H.   Xu   and   G.   Thinakaran   (2006).   "Pathological   and  physiological  functions  of  presenilins."  Mol  Neurodegener  1:  4.  Vinals,   F.   and   J.   Pouyssegur   (2001).   "Transforming   growth   factor   beta1   (TGF-­‐beta1)   promotes   endothelial   cell   survival   during   in   vitro   angiogenesis   via   an  autocrine   mechanism   implicating   TGF-­‐alpha   signaling."   Mol   Cell   Biol   21(21):  7218-­‐7230.  Wajant,  H.,  F.  Muhlenbeck  and  P.  Scheurich  (1998).  "Identification  of  a  TRAF  (TNF  receptor-­‐associated   factor)   gene   in   Caenorhabditis   elegans."   J   Mol   Evol   47(6):  656-­‐662.  Wakabayashi,  T.  and  B.  De  Strooper  (2008).  "Presenilins:  members  of  the  gamma-­‐secretase   quartets,   but   part-­‐time   soloists   too."   Physiology   (Bethesda)   23:   194-­‐204.  Wang,  C.,  L.  Deng,  M.  Hong,  G.  R.  Akkaraju,  J.  Inoue  and  Z.  J.  Chen  (2001).  "TAK1  is  a  ubiquitin-­‐dependent  kinase  of  MKK  and  IKK."  Nature  412(6844):  346-­‐351.  Wikstrom,  P.,  J.  Damber  and  A.  Bergh  (2001).  "Role  of  transforming  growth  factor-­‐beta1  in  prostate  cancer."  Microsc  Res  Tech  52(4):  411-­‐419.  Wrana,  J.  L.,  L.  Attisano,  R.  Wieser,  F.  Ventura  and  J.  Massague  (1994).  "Mechanism  of  activation  of  the  TGF-­‐beta  receptor."  Nature  370(6488):  341-­‐347.  Xiao,  H.  and  M.  Liu  (2013).  "Atypical  protein  kinase  C  in  cell  motility."  Cell  Mol  Life  Sci  70(17):  3057-­‐3066.  Xu,  L.,  Y.  Kang,  S.  Col  and  J.  Massague  (2002).  "Smad2  nucleocytoplasmic  shuttling  by  nucleoporins  CAN/Nup214  and  Nup153  feeds  TGFbeta  signaling  complexes  in  the  cytoplasm  and  nucleus."  Mol  Cell  10(2):  271-­‐282.  Xu,   P.,   J.   Liu   and   R.   Derynck   (2012).   "Post-­‐translational   regulation   of   TGF-­‐beta  receptor  and  Smad  signaling."  FEBS  Lett  586(14):  1871-­‐1884.  Yamamoto,   S.,   K.   L.   Schulze   and   H.   J.   Bellen   (2014).   "Introduction   to   Notch  signaling."  Methods  Mol  Biol  1187:  1-­‐14.  Yamashita,   H.,   P.   ten   Dijke,   P.   Franzen,   K.   Miyazono   and   C.   H.   Heldin   (1994).  "Formation   of   hetero-­‐oligomeric   complexes   of   type   I   and   type   II   receptors   for  transforming  growth  factor-­‐beta."  J  Biol  Chem  269(31):  20172-­‐20178.  Yamashita,  M.,   K.   Fatyol,   C.   Jin,   X.  Wang,   Z.   Liu   and   Y.   E.   Zhang   (2008).   "TRAF6  mediates   Smad-­‐independent   activation   of   JNK   and   p38   by   TGF-­‐beta."   Mol   Cell  31(6):  918-­‐924.  Zavadil,   J.,   L.   Cermak,  N.   Soto-­‐Nieves   and  E.   P.   Bottinger   (2004).   "Integration   of  TGF-­‐beta/Smad   and   Jagged1/Notch   signalling   in   epithelial-­‐to-­‐mesenchymal  transition."  EMBO  J  23(5):  1155-­‐1165.  Zhao,  Y.,  A.  M.  Thornton,  M.  C.  Kinney,  C.  A.  Ma,  J.  J.  Spinner,  I.  J.  Fuss,  E.  M.  Shevach  and  A.   Jain  (2011).   "The  deubiquitinase  CYLD  targets  Smad7  protein   to  regulate  transforming   growth   factor   beta   (TGF-­‐beta)   signaling   and   the   development   of  regulatory  T  cells."  J  Biol  Chem  286(47):  40520-­‐40530.