A Dissertation

entitled

The Regulatory Role of Mixed Lineage 4 Beta in MAPK Signaling and Ovarian

Cancer Cell Invasion

by

Widian F. Abi Saab

Submitted to the Graduate Faculty as partial fulfillment of the requirements for the Doctor of Philosophy Degree in Biology

______Dr. Deborah Chadee, Committee Chair ______Dr. Douglas Leaman, Committee Member ______Dr. Fan Dong, Committee Member ______Dr. John Bellizzi, Committee Member ______Dr. Max Funk, Committee Member ______Dr. Robert Steven, Committee Member ______Dr. William Taylor, Committee Member ______Dr. Patricia R. Komuniecki, Dean College of Graduate Studies

The University of Toledo

May 2013

Copyright 2013, Widian Fouad Abi Saab

This document is copyrighted material. Under copyright law, no parts of this document may be reproduced without the expressed permission of the author. An Abstract of

The Regulatory Role of Mixed Lineage Kinase 4 Beta in MAPK Signaling and Ovarian Cancer Cell Invasion

by

Widian F. Abi Saab

Submitted to the Graduate Faculty as partial fulfillment of the requirements for the Doctor of Philosophy Degree in Biology

The University of Toledo

May 2013

Mixed lineage kinase 4 (MLK4) is a member of the MLK family of - activated kinase (MAP3Ks). As components of a three-tiered

signaling cascade, MAP3Ks promote activation of mitogen-activated protein kinase

(MAPK), which in turn regulates different cellular processes including proliferation and invasion. Here, we show that the beta form of MLK4 (MLK4β), unlike its close relative,

MLK3, and other known MAP3Ks, negatively regulates the activities of the MAPKs,

p38, ERK and JNK, even in response to stimuli such as sorbitol or TNFα. MLK4β also

negatively regulates basal, but not TNFα-induced, NF-κB activity. Moreover, MLK4β

undergoes autophosphorylation and has kinase activity towards histone H2A, but has no

kinase activity towards the MAP2K, MEK4/SEK1, a known substrate for MLK3 and other MAP3Ks. Furthermore, MLK4β interacts with MLK3 and inhibits MLK3 activation. In addition, MLK4 blocks matrix metalloproteinase-9 gelatinase activity and invasion in SKOV3 ovarian cancer cells, both of which are cellular responses that require

MLK3. Collectively, our data establish MLK4β as a novel suppressor of MLK3 activation, MAPK signaling and cell invasion.

iii

This work is dedicated to my dad, Fouad Abi Saab, and mom, Nabila Abi Saab, who sacrificed a lot to provide a good education for my brother and me. I most certainly would not be where I am today if it wasn’t for them.

I also dedicate this work to my brother (Rawad), my grandmas (Fayza and

Samia), my aunts (Thouraya, Feryal, Noha and Sonia) and all my cousins (Yara, Ziad,

Lama, Wahid, Tamara and Faisal). However, a special dedication goes to my beloved

Grandma, Fayza Darweesh, who is my role model. She is my inspiration and the source of my strength and has always been my number one supporter. Her words and constant encouragement are my driving force to move forward in life.

I would also like to grab this opportunity to thank my dearest friends (Alexis, Ani,

Celia, Hadil, Hashem, Meenakshi, Mirella, Nancy and Natalya) who’ve been extremely encouraging and supportive throughout my Ph.D. program.

Acknowledgements

First, I would like to thank my advisor, Dr. Chadee, who had given me the chance

to be here and who taught me most of what I currently know in this field. Dr. Chadee is a

very supportive and positive person and creates a very amiable environment for her

students. In addition to being successful in her field, she is also extremely compassionate

and understanding. She was very supportive especially during hard times and for that I’ll

be forever grateful. Not only is Dr. Chadee successful in her career, but she also has an

exemplary sense of humanity which makes her a great role model for me.

I would also like to thank my committee members Dr. Douglas Leaman, Dr. Fan

Dong, Dr. John Bellizzi, Dr. Max Funk, Dr. Robert Steven and Dr. William Taylor for their constant input and guidance. I especially thank Dr. Taylor, Dr. Leaman and Dr.

Dong, for their technical support in a number of experiments.

I would like to especially thank Cathy (Dr. Yu Zhan) for teaching me most of the

techniques in lab and for being a good friend. Special thanks to Natalya Blessing for

being a wonderful lab mate and friend. I would also like to thank Meenakshi Bhansali for

her amazing friendship and support. Last but not least, I would like to thank Dr. Leah

Rider, Jenny, Alan, Peter, Sneha, April and Kyoung for being such good friends and for

adding a joyful and pleasant atmosphere to our working environment.

v

Table of Contents

Abstract ...... iii

Acknowledgements ...... v

Table of Contents ...... vi

List of Figures ...... ix

List of Abbreviations ...... xi

1 Introduction………………………………………………………………………..1

1.1 The Mitogen-activated protein kinase signaling cascade ...... 1

1.2 Characteristics and functions of MAPK proteins…………………………….3

1.2.1 The ERK1/2 pathway………………………………………………..3

1.2.2 The JNK pathway…………………………………………………...7

1.2.3 The p38 pathway…………………………………………………...11

1.2.4 The ERK5 pathway………………………………………………...14

1.3 The matrix metalloproteinases……………………………………………….15

1.4 The MAP2Ks………………………………………………………………...18

1.5 The MAP3Ks………………………………………………………………...20

1.5.1 The MEKK group………………………………………………….21

1.5.2 The Raf MAP3Ks………………………………………………….23

1.5.3 The TAK1 MAP3K group…………………………………………25

1.5.4 The TAO/Tpl2 and Mos MAP3K groups………………………….27

1.5.5 The MLK family of MAP3Ks……………………………………..27

vi 1.5.5.1 The DLK subgroup ...... 28

1.5.5.2 The ZAK subgroup………………………………………30

1.6 The MLK subfamily…………………………………………………………31

1.6.1 MLK1 and MLK2………………………………………………….32

1.6.2 MLK3 activation…………………………………………………...33

1.7 MLK3 signaling……………………………………………………………...36

1.7.1 MLK3 signaling in cancer…………………………………………38

1.8 MLK4: characteristics and function………………………………………….39

1.9 Significance…………………………………………………………………..40

2 Materials and Methods ...... 42

2.1 Cell culture…………………………………………………………………...42

2.2 Expression vectors…………………………………………………………...43

2.3 Plasmids and siRNA transfections…………………………………………...43

2.4 Immunoblotting………………………………………………………………45

2.5 Preparation of whole cell extracts and treatments…………………………...47

2.6 Immunoprecipitation…………………………………………………………47

2.7 MLK4β kinase assay…………………………………………………………48

2.8 Cell proliferation assay………………………………………………………49

2.9 Luciferase assay……………………………………………………………...50

2.10 Invasion assay………………………………………………………………50

2.11 Gelatin zymography………………………………………………………...51

3 Results……………………………………………………………………………52

3.1 The role of MLK4β in p38 signaling………………………………………...52

vii 3.1.1 The effect of ectopic expression of MLK4β on p38 activation…...52

3.1.2 The effect of endogenous MLK4 on the activation of p38. ………54

3.2 The effect of MLK4 on MEK3/MEK6 activation….………………………..56

3.3 The role of MLK4β in NF-κB signaling……………………………………..57

3.4 Comparison of the effects of MLK4β and MLK3 on p38 activation ……….60

3.5 The effects of MLK3 and MLK4 on ERK and JNK activation……………..62

3.6 MLK4β is not an upstream activator of MEK4……………………………...65

3.7 MLK4β kinase activity……………………………...……………………….67

3.8 The effect of MLK4β on MLK3 activation………………………………….69

3.9 The correlation between MLK4β expression and active MLK3 in different cell

lines……………………………………………………………...……………….72

3.10 MLK4β associates with MLK3……………………………………………..75

3.11 The effect of MLK4 on cell proliferation…………………………………..77

3.12 MLK3 is required for cell invasion in ovarian cancer cells………………...79

3.13 MLK4β inhibits SKOV3 cell invasion……………………………………..81

3.14 MLK3 regulates MMP-2 and MMP-9 activity…………………….82

3.15 MLK4β reduces MMP-9 activity in SKOV3 cells…………………………85

4 Discussion………………………………………………………………………..86

References……………………………………………………………………………….95

viii

List of Figures

1-1 The MAPK signaling cascade ...... 2

1-2 The Ras/Raf/ERK1/2 signaling pathway ...... 6

1-3 JNK-mediated …………………………………………………………10

1-4 The p38 MAPK signaling pathway………………………………………………13

1-5 MMP-2 and -9: structure and activation…………………………………………17

1-6 The NF-κB pathway……………………………………………………………...26

1-7 Signaling of the DLK family of MAP3Ks……………………………………….29

1-8 The structural domains of MLKs………………………………………………...31

1-9 Model mechanism of MLK3 activation by Cdc42………………………………35

3-10 MLK4β expression inhibits basal and stimulus-induced p38 activation ...... 54

3-11 Elevated active p38 in MLK4 knockdown cells…………………………………55

3-12 MLK4 negatively regulates MEK3/MEK6 activation…………………………...57

3-13 MLK4β negatively regulates basal NF-κB activation but has no effect on TNFα- induced NF-κB signaling……………………………………..………………………….59

3-14 MLK4β, unlike MLK3, inhibits activation of p38 ………………………………61

3-15 MLK3 promotes the activation of ERK and JNK in SKOV3 and HEY1B cells...62

3-16 MLK4 negatively regulates ERK and JNK activation…………………………...64

3-17 MLK4β does not phosphorylate Thr261 on GST-SEK1-KR……………………66

3-18 MLK4β: autophosphorylation and substrate specificity…………………………68

3-19 MLK4β inhibits induced MLK3 activation……………………………………...69

ix 3-20 MLK4 inhibits the basal activation of MLK3 in SKOV3 cells………………….71

3-21 Correlation between MLK4β expression and active MLK3.. …………...... 74

3-22 MLK4β associates with MLK3………………………………………………...... 76

3-23 MLK4 has no effect on HCT116 cell proliferation……………………………...78

3-24 MLK3 is essential for SKOV3 and HEY1B cell invasion……………………….80

3-25 MLK4β reduces the invasion of SKOV3 cells..…………………………………82

3-26 MLK3 mediates MMP-2 and MMP-9 activation in SKOV3 and HEY1B cells by a mechanism that involves ERK and JNK………………….……………………………..84

3-27 MLK4β reduces MMP-9 activity in SKOV3 cells………………………………85

4-28 Schematic diagram illustrating the role of MLK4β in MAPK signaling………..94

x

List of Abbreviations

Akt1...... Rac-alpha serine/ kinase APS ...... Ammonium persulfate AP-1 ...... Activator protein 1 ASK...... Apoptosis signal-regulating kinase ATF2 ...... Activating transcription factor 2 ATM………………..Ataxia telangiectasia ATP ...... Adenosine triphosphate

Bax ...... Bcl2-associated X Bcl-2 ...... B-cell lymphoma 2 BSA ...... Bovine serum albumin

CBD ...... Collagen binding domain CR ...... Conserved region CRIB ...... Cdc42/Rac interactive binding protein

DTT ...... Dithiothreitol DLK ...... Dual leucine zipper-bearing kinase DNA ...... Deoxyribonucleic acid DSP ...... Dual specificity phosphatases

ECM ...... Extracellular matrix EMT ...... Epithelial-mesenchymal transition EGF ...... Epidermal growth factor ERK ...... Extracellular signal-regulated kinase

FADD.……….……..Fas-associated death domain protein FasL...... FBS ...... Fetal bovine serum FFA…………………Free fatty acids FGD1……………….FYVE, RhoGEF and PH domain-containing protein 1

GADD45 ...... Growth arrest and DNA damage-inducible 45 GDP………………...Guanosine diphosphate GEF ...... Guanosine nucleotide exchange factor

xi Grb2 ...... Growth factor receptor binding protein 2 GST ...... Glutathione S- GTP ...... Guanosine triphosphate

h……………………..hours HBx…………………Hepatitis B x antigen HPK1...... Hematopoietic protein kinase 1

IκBα ...... Inhibitor of kappa B alpha IKK ...... IκB kinase IKKK...... IκB kinase kinase IP ...... Immunoprecipitation

JIP1 ...... JNK-interacting protein 1 JNK ...... c-Jun N-terminal kinase

LPS ...... Lipopolysaccharide LZ ...... Leucine zipper

KSR ...... Kinase suppressor of ras

MAPK ...... Mitogen activated protein kinase MAPKAP-K ...... MAPK-activated protein kinase MAP2K ...... MAPK kinase MAP3K ...... MAPK kinase kinase MEK ...... MAPK/ERK kinase MEKK ...... MEK kinase Met ...... Methionine MLK ...... Mixed lineage kinase MLTKα ...... Mitogen-activated protein triple kinase alpha MKP ...... MAPK phosphatase MMP ...... Matrix metalloproteinase MNK ...... MAPK interacting kinase MP-1 ...... MEK partner 1 MSK ...... Mitogen and stress activated kinase

ND………………….Neuro-D NF ...... Neurofibromatosis NF-κB ...... Nuclear factor kappa-light-chain enhancer of activated B cells NGF...... Neuron growth factor

PAK1...... p21-GTPase activated kinase 1 PARP……………….poly (ADP-ribose) polymerase (PARP) PB1 ...... Phox/Bem1P PBS ...... Phosphate buffered saline PHD………………...Plextrin-homology domain

xii PMSF ...... Phenylmethylsulphonyl fluoride PP2A ...... Serine/threonine protein phosphatase 2A PP5 ...... Protein phosphatase 5 Pro ...... Proline PTP ...... protein phosphatase PVDF ...... Immobilon-P Polyvinylidene Flouride

ROS………………...Reactive oxygen species RSK ...... p90 RTK......

SAP1………………..Sodium-associated protein 1 SAM ...... Sterile-alpha-motif SAPK ...... Stress-activated protein kinase SCG………………...Superior cervical ganglion Ser ...... Serine SH ...... Src homology siRNA ...... small interfering RNA SOS ...... Son of sevenless STAT3...... Signal transducer and activator of transcription 3

TAB1...... TAK1-binding protein 1 TAK1 ...... Transforming growth factor β-activted protein 1 TAO ...... Thousand and one amino acid TCR…………………T cell antigen receptor (TCR) TGFβ………………..Transforming growth factor beta TIMP ...... Tissue inhibitor of metalloproteinases Thr ...... Threonine TNFα ...... alpha TLR4 ...... Toll-like receptor 4 Tpl2 ...... Tumor progression 2 TRAF4 ...... TNF receptor-associated factor 4 Tyr ...... Tyrosine

ZAK ...... Zipper-sterile-alpha motif kinase

xiii Chapter 1

Introduction

1.1 The Mitogen-activated protein kinase signaling cascade

The mitogen-activated protein kinase (MAPK) signaling pathway is a three-tiered

signaling cascade that is conserved from yeast to higher mammals including humans

(Widmann, et al., 1999). The MAPK pathway is activated by a wide range of stimuli such

as stress, cytokines and growth factors and leads to different cellular responses including

proliferation, inflammation, invasion and apoptosis (Figure 1) (Kyriakis and Avruch,

2001; Pearson, et al., 2001; Uhlik, et al., 2004). The MAPK kinase kinases, or MAP3Ks, form the top tier of the signaling cascade (Dhanasekaran and Johnson, 2007). Once

MAP3Ks are activated, they phosphorylate and activate their immediate downstream targets, the MAPK kinases (MAP2Ks or MEKs) that in turn phosphorylate and activate

MAPKs, the cascade’s executor kinases (Figure 1) (Johnson and Lapadat, 2002; Kyriakis and Avruch, 2001; Lawler, et al., 1998; Raingeaud, et al., 1996). The mammalian extracellular signal-regulated kinases 1 and 2 (ERK1/2), c-Jun N-terminal kinase (JNK), p38 kinase and ERK5 are four major MAPKs involved in this signaling cascade, which upon stimulation, activate cytosolic or nuclear-localized effectors and thereby translate

1 the stimulus into a corresponding cellular response (Figure 1) (Ben-Levy, et al., 1998;

Raingeaud, et al., 1996; Uhlik, et al., 2004).

Figure 1. The MAPK signaling cascade. MAP3Ks are activated in response to stress, cytokines or growth factors. Active MAP3Ks phosphorylate and activate MAP2Ks that in turn phosphorylate and activate MAPKs. Active MAPKs activate cytosolic targets or activate transcription factors that regulate the expression of that control different cellular processes like proliferation, invasion, apoptosis and inflammation.

2 1.2 Characteristics and functions of MAPK proteins

MAPKs are proteins that are ubiquitously expressed in all eukaryotic cell types

but yet regulate different cellular responses in a stimulus- and cell type-specific manner

(Dhanasekaran and Johnson, 2007; Uhlik, et al., 2004; Widmann, et al., 1999). MAPKs are proline directed serine/threonine kinases that activate nuclear or cytosolic substrates, by phosphorylation at serine or threonine residues found within the Pro-X-Ser/Thr-Pro consensus sequence (Alvarez, et al., 1991; Maeda and Firtel, 1997). MAPK proteins are activated upon dual phosphorylation, by specific MAP2Ks, on both the threonine and tyrosine residues of the Thr-X-Tyr motif present in the activation loop, where the amino acid X varies with different MAPKs (Ahn, et al., 1991; D'Mello, et al., 1993; DiDonato, et al., 1996; Estus, et al., 1994; Faris, et al., 1998; Frandsen and Schousboe, 1990).

MAPKs undergo an ordered phosphorylation mechanism, whereby the tyrosine residue of

the Thr-X-Tyr motif is phosphorylated first resulting in an increase in the affinity

between MAPKs and their specific MAP2Ks, a step that allows the subsequent

phosphorylation of the threonine residue and full MAPK activation (Haystead, et al.,

1992). Dual phosphorylation of MAPKs triggers a series of conformational changes in

the activation loop and surrounding sequences that ultimately result in the activation of

these proteins (Canagarajah, et al., 1997). Of the MAPK pathways, ERK, JNK and p38

signaling pathways are the best characterized.

1.2.1 The ERK1/2 pathway

ERK1 (44 kDa) and ERK2 (42 kDa), often referred to as ERK1/2, are two main

ubiquitously expressed isoforms of ERK that share more than 85% sequence identity

3 (Boulton, et al., 1991; Chen, et al., 2001; Seger and Krebs, 1995). Activation of ERK1/2

occurs upon specific recognition and subsequent phosphorylation of the Thr and Tyr

residues in the Thr-X-Tyr motif (Thr183 and Tyr185 in human ERK2 and Thr202 and

Tyr204 in human ERK1) by the upstream MAP2Ks, MEK1 and MEK2 (Crews, et al.,

1992; Zheng and Guan, 1993). Activity of ERK1/2 is also regulated by phosphatases,

including the MAPK phosphatases or MKPs which are dual-specificity phosphatases

(DSPs) that dephosphorylate both phospho-tyrosine and phospho-threonine residues

(Owens and Keyse, 2007; Raman, et al., 2007). Of the different MKPs, MKP3 shows higher specificity towards ERK1/2 than other MAPKs (Zhang, et al., 2003). The serine/threonine protein phosphatase 2A, or PP2A, also functions as a regulator of

ERK1/2 activity by dephosphorylating the threonine residue of the Thr-X-Tyr motif in

ERK1/2 (Anderson, et al., 1990).

Growth factors and are the primary activators of the ERK1/2 pathway, however, cytokines, activators of G protein-coupled receptors and different stresses have also been reported to activate this pathway (Johnson and Lapadat, 2002; Yoon and Seger,

2006). Early studies revealed a key role for Ras GTPases and B-Raf in ERK activation

(depicted in Figure 2 below). Briefly, receptor tyrosine kinases (RTKs), upon binding to their ligands such as growth factors, undergo dimerization and cytoplasmic domain transphosphorylation. Another transphosphorylation process then follows on specific tyrosine residues in the cytoplasmic region of the RTK, which leads to full activation of the receptor. The phospho- create docking sites for the Src homology 2 (SH2) domain of adaptor proteins, such as the growth factor receptor binding protein 2, or Grb2.

Grb2 then recruits the guanine nucleotide exchange factor (GEF), son of sevenless (SOS),

4 via its Src homology 3 (SH3) domain. SOS then activates Ras by promoting the switch

from an inactive GDP-bound to an active GTP-bound form (Buday and Downward,

2008; Downward, 1996; Wittinghofer, et al., 1997). Upon activation, Ras interacts with

and promotes the activation of members of the Raf family of MAP3Ks, Raf-1, B-Raf and

A-Raf. Once activated, Rafs phosphorylate and activate MEK1 and MEK2 that in turn activate ERK1/2 (Chadee and Kyriakis, 2004; Dhillon, et al., 2007; Dunn, et al., 2005).

Active ERK1/2 will then undergo dimerization and either activate cytoplasmic substrates such as p90 ribosomal S6 kinases (RSKs), mitogen and stress activated kinases (MSKs) and MAPK interacting kinase (MNK), or translocate to the nucleus and regulate the expression of certain genes by directly activating several transcription factors including

AP-1, c-Myc and c-Fos (Buday and Downward, 2008; Chen, et al., 2001; Dunn, et al.,

2005; Raman, et al., 2007).

5

Figure 2. The Ras/Raf/ERK1/2 signaling pathway. In response to interaction with growth factors (GF), RTKs undergo dimerization and activation. Grb2 binds to the active

RTK and recruits SOS which in turn activates Ras. Active Ras interacts with and activates the Raf members. Active Rafs phosphorylate and activate MEK1/2 that in turn phosphorylate and activate ERK1/2. Once activated, ERK1/2 can trigger a cellular response either by activating cytoplasmic targets or by inducing transcriptional activation by translocating to the nucleus and activating transcription factors.

6 ERK1/2 is one of the main regulators of cell proliferation and transformation and

the ERK signaling pathway is deregulated at a high frequency in human cancers

(Dhillon, et al., 2007). ERK1/2, however, is implicated in numerous other functions including differentiation, cell motility, cell migration, cytoskeletal polymerization and golgi fragmentation (Ishibe, et al., 2004; Jesch, et al., 2001; Reszka, et al., 1995; Yoon and Seger, 2006). The ERKs have different subcellular localizations. In addition to the cytoplasm and nucleus, ERK1/2 were found to associate with membrane receptors and transporters, intracellular membrane compartments, microtubules, adherens junctions and focal adhesions (Furuchi and Anderson, 1998; Ishibe, et al., 2004; Jesch, et al., 2001;

Reszka, et al., 1995). Moreover, continuous ERK1/2 nucleo-cytoplasmic trafficking is required for proper signaling of ERK1/2 (Costa, et al., 2006; Marchi, et al., 2008).

1.2.2 The JNK pathway

JNK is a proline-directed kinase that is activated by stresses and TNFα (Kyriakis, et al., 1994). JNK1/SAPKβ, JNK2/SAPKα and JNK3/SAPKγ (ranging between 46 and

54kDa) are three JNK isoforms that share more than 85% sequence identity (Johnson and

Lapadat, 2002; Kyriakis and Avruch, 2001). While JNK3 expression is mostly restricted to the brain and , JNK1 and JNK2 are ubiquitously expressed (Dhillon, et al., 2007).

JNK is activated upon the dual phosphorylation of the threonine and tyrosine residues of the Thr-X-Tyr motif by the MAP2Ks, MEK4 and MEK7 (Lawler, et al., 1998). The dual specificity phosphatases MKP7 and VH5 also function to regulate JNK activity by dephosphorylating the active JNK sites (Alonso, et al., 2004). The phosphoserine/threonine protein phosphatase 5 (PP5) is yet another phosphatase that

7 deregulates the JNK pathway. PP5, however, does so by acting upstream of JNK,

primarily by inactivating MAP3Ks such as the apoptosis signal-regulating kinase 1, or

ASK1 (Chinkers, 2001; Zhou, et al., 2004).

The JNK pathway is mainly activated by cytokines and stresses but is also activated to a lesser extent by growth factors and inhibition of DNA and protein synthesis

(Kyriakis and Avruch, 2001). Once activated, JNK translocates to the nucleus where it activates numerous transcription factors, including c-Jun. Activation of c-Jun by JNK was found to mediate Ras-induced tumorigenesis and cellular transformation in vitro

(Johnson, et al., 1996; Smeal, et al., 1991). Other transcription factors activated by JNK proteins include p53, STAT3, ATF2, ELK1 and nuclear factor of activated T cells (Chen, et al., 2001; Ip and Davis, 1998). Through expression driven by one or more of these transcription factors, JNK can mediate differentiation, cytokine production, actin reorganization, inflammatory responses and apoptosis (Chen, et al., 2001). Of these different cellular responses, the ability of JNK to promote apoptosis is one of the best characterized.

Although the exact mechanism by which JNK triggers cell death is not well known, several lines of evidence support a role for JNK in mitochondrial apoptotic cell death (Dhanasekaran and Reddy, 2008). Mitochondrial apoptosis or the intrinsic apoptotic pathway is characterized by the permeabilization of the mitochondrial membrane, the release of , and the activation of 9 and 3 (Baliga and Kumar, 2003). This process is regulated by the Bcl-2 family of proteins (Kim, 2005).

JNK can activate mitochondrial cell death through phosphorylation and inhibition of the anti-apoptotic protein, Bcl-2 (Figure 3) (Dhanasekaran and Reddy, 2008). Bcl-2 normally

8 sequesters and inactivates the pro-apoptotic protein Bax (Breckenridge and Xue, 2004).

Upon phosphorylation, Bcl-2 dissociates from Bax rendering Bax active. Bax then

undergoes multimer formation that inserts in the mitochondrial membrane, forming

channels that allow the release of cytochrome c and other apoptotic proteins leading to

-3 dependent cell death (Figure 3) (Breckenridge and Xue, 2004). JNK can also

induce apoptosis in a p53-dependent manner. p53, upon phosphorylation and activation by JNK, can induce apoptosis by mediating the expression of pro-apoptotic genes such as

Bax (Figure 3) (Miyashita and Reed, 1995). p53 can also induce apoptosis in a

transcription-independent manner (Figure 3) (Caelles, et al., 1994). JNK can also trigger

apoptosis by the induction of the extrinsic apoptotic pathway (Figure 3) (Dhanasekaran

and Reddy, 2008; Tang, et al., 2012). The extrinsic apoptotic pathway is mediated by

FasL that upon binding to Fas and the adaptor protein FADD, activates which

then triggers apoptosis in a caspase 3-dependent manner (Guicciardi and Gores, 2009;

Strasser, et al., 2009). JNK was shown to activate this pathway by inducing the

expression of FasL (Figure 3) (Dhanasekaran and Reddy, 2008). For instance, JNK

upregulates FasL protein expression in HepG2 cells in response to the hepatitis B virus X

protein (HBx) in a mechanism that depends on MLK3 and MEK7 (Tang, et al., 2012).

9 Bcl-2 P JNK Bax Cyt c Cyt c Cyt c

Expression of FasL P Bcl-2

P Bax Bax

p53 Bax Cyt c Extrinsic pathway Cyt c

Bax Cyt c Bax Bax Transcription of pro- Cyt c Cyt c Activation of apoptotic genes Casp 9 caspase 8 Apaf- 1

Activation of caspase 3 Activation of caspase 3

APOPTOSIS

Figure 3. JNK-mediated apoptosis. JNK can induce mitochondrial apoptotic cell death through the phosphorylation and inactivation of Bcl-2 either directly or through Bad and

Bim. This leads to the activation of Bax which forms mitochondrial channels through which cytochrome c is released. Cytochrome c then activates the caspase cascade. JNK can also trigger the expression of pro-apoptotic genes by activating transcription factors such as p53. JNK activates the extrinsic apoptotic pathway by increasing the expression of the pro-apoptotic FasL gene.

10 1.2.3 The p38 pathway

p38 is a family of MAPK proteins that consists of four main members in

vertebrates (α, β, γ and δ). The p38 isoforms are encoded by different genes with

variations in the (Jiang, et al., 1996; Lechner, et al., 1996). The

variations in the sequence homology and in the interactions with other molecules could

explain the diverse functions of p38 (Cuenda and Rousseau, 2007). Moreover, of the four

p38 isoforms, only p38α and p38β are ubiquitously expressed while the other forms show

more limited expression, with p38γ mainly found in skeletal muscle and p38δ in

pancreas, kidney and small intestine (Goedert, et al., 1997; Raman, et al., 2007).

Thr-Gly-Tyr is the conserved motif in the activation loop of the p38 family, and

dual phosphorylation of the threonine and tyrosine residues lead to p38 activation

(Kyriakis and Avruch, 2001). MEK3 and MEK6 are the MAP2Ks that selectively

activate the p38 family of MAPKs (Cuenda, et al., 1997; Raingeaud, et al., 1996). The

JNK activator MEK4 can also activate p38 (Brancho, et al., 2003). Although p38 is mainly activated by MEKs, activation of p38α can also occur in a MEK-independent manner. In this respect, p38α, upon binding to TAB1 (Transforming growth factor β- activated protein 1 (TAK1)-binding protein 1), undergoes autophosphorylation at Thr180 and Tyr182 and subsequent activation (Ge, et al., 2002). Moreover, in T cells, p38α was proposed to be activated by undergoing conformational change and autophosphorylation at the activation sites (Thr180 and Try182) following the phosphorylation of Tyr323 by tyrosine kinases in a mechanism that is mediated by the T cell antigen receptor (TCR)

(Salvador, et al., 2005). Similar to other MAPKs, the activity of p38 is also regulated by multiple phosphatases. Some of these phospatases that interact with and inhibit p38

11 include MKP1, MKP5, MKP7 and MKP8. These phosphatases can also act on JNK but

show low affinity towards ERK1/2 (Keyse, 2000; Raman, et al., 2007; Tanoue and

Nishida, 2003). MKP1-null mice show enhanced activity of p38 and production of

proinflammatory cytokines in response to low exposures to LPS as compared to wild

types, suggesting MKP1 as an important regulator of LPS-induced p38 activity (Chi, et

al., 2006). The tyrosine phosphatase PTP and the serine/threonine phosphatases PP2C,

PP2A and PP2B have also been shown to inactivate p38 (Raman, et al., 2007; Takekawa,

et al., 2000; Takekawa, et al., 1998).

Similar to JNK, p38 is activated by stresses and cytokines, but can also be

activated to a much lesser extent by growth factors (Zarubin and Han, 2005). Activated

p38 then either activates cytoplasmic substrates or translocates to the nucleus where it

activates transcription factors that regulate the expression of genes involved in a number

of cellular responses such as apoptosis, growth, migration, differentiation and

inflammation, the deregulation of which can lead to certain pathologies (Figure 4) (Ben-

Levy, et al., 1998; Krens, et al., 2006; Zarubin and Han, 2005). The wide range of p38

substrates is responsible for the induction of these different biological functions.

Although all p38 isoforms have many substrates in common including ATF2, ELK-1, and

SAP1, differences in substrate specificity exist between p38α and p38β and between p38γ

and p38δ. For instance, while MAPKAP-K2 and MAPKAP-K3 are substrates for p38α

and p38β, they cannot be activated by p38γ and p38δ (Cuenda, et al., 1997; Goedert, et al., 1997). Some substrates are also specific to either p38γ or p38δ. p38γ binds to and phosphorylates many PDZ domain-containing proteins via its KETXL C-terminal sequence that is unique among the p38 isoforms (Hasegawa, et al., 1999; Sabio, et al.,

12 2005; Sabio, et al., 2004). p38δ has stathmin and tau as substrates, both of which play a role in regulating microtubule dynamics (Feijoo, et al., 2005; Parker, et al., 1998).

Figure 4. The p38 MAPK signaling pathway (adapted from (Cuenda and Rousseau,

2007)). p38 in response to stimulating agents, is activated by MEK3/MEK6, also known as MKK3/MKK6. Once activated, p38 phosphorylates and activates numerous cytosolic proteins and transcription factors to regulate many biological functions. Deregulation of p38 signaling can contribute to the development of several pathologies.

13 p38 has many targets in different subcellular localizations. In some cell lines, p38

was found in the nucleus in resting cells and in response to stimuli, p38 was

phosphorylated and formed a complex with MAPKAP-K2 or MAPKAP-K5 (Ben-Levy, et al., 1998; Seternes, et al., 2002). p38-MAPKAP-K2 or p38-MAPKAP-K5 complexes then are exported to the cytoplasm where p38 further activates cytosolic substrtaes (Ben-

Levy, et al., 1998; Seternes, et al., 2002). The intracellular localization of p38α can also

be regulated by TAK1 and TAB1 (Lu, et al., 2006). The localization of p38 can also vary

with isoforms. For instance, in cardiac myocytes, p38α and p38β shuttle between the

nucleus and the and are retained in the nucleus in response to the nuclear export

inhibitor, leptomycin B. This inhibitor, however, did not affect the localization of p38γ

which was mainly associated with cytosolic foci (Court, et al., 2002).

1.2.4 The ERK5 pathway

The ERK5 pathway is the least characterized among the different MAPK

pathways. ERK5 signaling is triggered in response to EGF and TNFα (Kato, et al., 1998;

Weldon, et al., 2002). ERK5 is activated upon phosphorylation of Thr218 and Tyr220 of

the Thr-X-Tyr motif by the MAP2K, MEK5 (Weldon, et al., 2002). MEKK2 and

MEKK3 are the only MAP3Ks shown to activate MEK5 and hence ERK5 (Sun, et al.,

2001). A role for ERK5 in cell proliferation, transformation and metastasis has been

demonstrated (Buschbeck, et al., 2005; Kato, et al., 1998; Kesavan, et al., 2004). For

instance, ERK5 promoted cell proliferation in response to EGF (Kato, et al., 1998).

Moreover, ERK5 was shown to mediate transformation downstream of the oncogene,

BCR/Abl (Buschbeck, et al., 2005). In addition to regulating MMP-9 and IL-6, ERK5

14 was found to be overexpressed in human breast cancer patients with poor prognosis.

These findings support a role for ERK5 in cancer (Kesavan, et al., 2004; Mehta, et al.,

2003; Montero, et al., 2009).

1.3 The matrix metalloproteinases

Matrix metalloproteinases (MMPs) are zinc-dependent proteolytic that

can function as downstream targets of MAPKs (Ispanovic and Haas, 2006; Spallarossa, et

al., 2006; Westermarck and Kahari, 1999). MMPs play a key role in cancer cell invasion but are also involved in other biological functions including development, morphogenesis, angiogenesis and inflammation (Hua, et al., 2011). Of the 28 different

MMPs, members of the gelatinase group (MMP-2 and MMP-9) are implicated the most in cancer (Bauvois, 2012). The expression and activation of these MMPs can be mediated by p38, ERK1/2 and/or JNK (Cho, et al., 2000; Ispanovic and Haas, 2006; Spallarossa, et al., 2006). MMPs can thus execute signals triggered by MAP3Ks. MMP-2 and MMP-9 consist of the following structural domains: the N-terminal signal peptide domain, the propeptide region, the catalytic domain containing the highly conserved zinc binding region, and the C-terminus that has a hemopexin-like domain that is connected to the catalytic domain by a small hinge region (Figure 5) (Bauvois, 2012; Nagase and

Woessner, 1999). The catalytic domain of gelatinases is characterized by the presence of a collagen-binding domain (CBD), composed of three fibronectin type II modules that function in enhancing the efficiency of collagen and gelatin degradation (Figure 5)

(Morgunova, et al., 1999).

15 MMPs are first synthesized as inactive pre-proenzymes that undergo two main

steps of activation (Benjamin and Khalil, 2012; Overall and Lopez-Otin, 2002). The first

step involves the cleavage of the signal peptide during translation to yield proMMPs

(Figure 5) (Benjamin and Khalil, 2012). ProMMPs are then maintained inactive by a

conserved cysteine residue in the prodomain that acts as a “cysteine switch”, whereby, it

forms a coordination bond with the zinc ion in the catalytic domain (Figure 5)

(Springman, et al., 1990). Full activation of MMPs occurs upon disruption of this bond often by the removal of the propeptide domain by an autoproteolytic event that can also be induced by chemicals (Figure 5) (Springman, et al., 1990).

MMP-2 and -9 are considered as biomarkers for malignant tumors with poor prognosis (Bauvois, 2012). MMP-2 and/or -9 were reported to be overexpressed in a number of cancers including, breast, colorectal, ovarian, prostate and lung (Roy, et al.,

2009; Turpeenniemi-Hujanen, 2005). MMP-2 and -9, through their ability to degrade the basement membrane and different other extracellular-matrix (ECM) components, were found to promote cell invasion in several types of cancer (Hua, et al., 2011). MMP-2 and

MMP-9 can also mediate the epithelilal-mesenchymal transition (EMT), a process

required for tumor invasion, which further supports a role for these MMPs in cell

invasion (Tester, et al., 2007; Thiery, et al., 2009; Xu, et al., 2009).

16

Figure 5. MMP-2 and -9: structure and activation. MMP-2 and MMP-9 generally contain several structural domains that sequentially are: the signal peptide domain (SPD), the propeptide domain (Pro), the catalytic domain containing the Zn2+-binding region and the collagen-binding domain (CBD), a hemopexin-like C-terminus connected to the catalytic domain by a hinge linker. Activation of these MMPs is a two-step process that first involves cleavage of the SPD then followed by the proteolytic cleavage of the pro- pepetide domain. C stands for cysteine and Zn2+ for zinc.

17 1.4 The MAP2Ks

MAP2Ks, as mentioned earlier, represent the second tier of kinases in the MAPK

signaling cascade. The MAP2Ks are dual specificity kinases where they act as both

tyrosine and serine/threonine kinases, and phosphorylate the tyrosine and threonine

residues of the Thr-X-Tyr motif within the MAPKs (Ashworth, et al., 1992). In addition

to the Thr-X-Tyr sequence containing the motif, the tertiary structure of the MAPK

protein is also very important for MAP2K binding (Widmann, et al., 1999). Thus, each

MAP2K interacts with and regulates a limited number of MAPKs (Widmann, et al.,

1999). MAP2Ks are much fewer in number than MAP3Ks (Widmann, et al., 1999). Since

numerous MAP3Ks are activated in response to each stimuli, the multiple signals

triggered are thus funneled down by the MAP2Ks, which specifically direct those signals

to one or more of the MAPK pathways (Widmann, et al., 1999).

MAP2Ks have two domains, a docking (D) domain and a domain for versatile docking (DVD), that are essential for binding to MAPKs and MAP3Ks, respectively

(Sharrocks, et al., 2000; Takekawa, et al., 2005). The DVD domain of MAP2Ks lies 20 residues downstream of the catalytic core (Takekawa, et al., 2005). In addition to the D and DVD domains, MEK5, an activator of ERK5, contains a Phox/Bem1P (PB1) domain that plays a role in protein-protein interactions and associates with other PB1-containing proteins such as the MAP3Ks, MEKK2/MEKK3 (Lamark, et al., 2003; Nakamura, et al.,

2006). The N-terminal and C-terminal regions of the PB1 domain of MEK5 also bind to

ERK5, allowing MEK5 to act as a that couples MEKK2/MEKK3 to

ERK5 (Nakamura, et al., 2006).

18 Seven members of MAP2Ks, MEK1-7, exist with MEK1 and MEK2 being

specific activators of ERK1/2, MEK3 and MEK6 of p38, MEK4 and MEK7 of JNK and

MEK5 of ERK5 (Dhanasekaran and Johnson, 2007; Dhillon, et al., 2007; Raman, et al.,

2007). MEK4, however, also phosphorylates and activates p38 (Brancho, et al., 2003).

While MEK7 was suggested to be a stronger activator of JNK than MEK4, full activation

of JNK was reported to require both MEK4 and MEK7 that preferentially phosphorylate

the tyrosine and threonine residues within the Thr-X-Tyr motif of JNK, respectively

(Lawler, et al., 1998). Another study reported that both of these residues can be

phosphorylated by either of the two MAP2Ks (Chen, et al., 2002). However, JNK

activation in the absence of either MEK4 or MEK7 was found to be lower than that in the

presence of both MAP2Ks (Chen, et al., 2002). MEK4 and MEK7 were both shown to be activated by stress to trigger the JNK pathway (Wang, et al., 2007). Pro-inflammatory cytokines on the other hand, induce JNK signaling primarily by activating MEK7

(Moriguchi, et al., 1997). MEK4 and MEK7 also exhibit some variations in the MAP3Ks that activate them (Wang, et al., 2007). For instance, while MEKK4 activates MEK4,

MEKK1 activates both MEK4 and MEK7 (Takekawa, et al., 2005).

The p38 activators, MEK3, MEK4 and MEK6 are activated in response to

stresses, proinflammatory cytokines and UV irradiation (Brancho, et al., 2003). p38α is

activated by all three MEKs (Brancho, et al., 2003). p38γ is activated in response to

environmental stresses by both MEK3 and MEK6, while activation by TNFα is mainly

mediated by MEK6 (Remy, et al., 2010). Similar to p38γ, stress-induced p38β activation is mediated by both MEK3 and MEK6 (Remy, et al., 2010). The activity of the p38δ, in

19 response to numerous stimuli, including UV radiation, stress and TNFα is primarily regulated by MEK3 (Remy, et al., 2010).

Mutations in MAP2Ks have also been implicated in cancers. For instance, MEK4, though at a low frequency, was reported to be inactivated in numerous types of cancer, including breast, colon and lung cancer (Cunningham, et al., 2006). This might suggest a suppressive role for MEK4 in cancer. In contrast, overexpressing MEK4 in breast and pancreatic cancer cells promoted proliferation and invasion which implicates a pro- oncogenic role for MEK4 (Wang, et al., 2004). Moreover, although, knocking out MEK4 in human pancreatic cancer cell lines did not yield any effect on proliferation, metastasis was suppressed (Cunningham, et al., 2006). The differences observed in the cellular responses triggered by MEK4 could be attributed to the degree of activation in response to stresses and environmental context of cells that favor the induction of either proliferation or apoptosis (Dhillon, et al., 2007).

1.5 The MAP3Ks

The MAP3K are serine/threonine kinases that are encoded by at least 20 genes, and represent the largest and most diverse components of the MAPK cascade. MAP3Ks function to direct signals generated by a wide range of stimuli to the MAPKs that ultimately lead to specific functional responses (Cuevas, et al., 2007; Uhlik, et al., 2004).

The differences in the domains and motifs between MAP3Ks, allow individual MAP3Ks to selectively regulate the localization and activation of specific MAP2Ks and MAPKs

(Dhanasekaran and Johnson, 2007). Many of the MAP3Ks however, can activate more than one MAP2K and thereby regulate multiple MAPK pathways (Cuevas, et al., 2007).

20 Based on protein homology, the different MAP3Ks are distributed into several main

clusters: the mitogen-activated protein/ERK kinase kinase (MEKKs), Rafs, thousand and

one amino acid (TAOs), tumor progression locus 2 (Tpl2) and apoptosis signal regulating

kinases (ASKs), TGF-β-activated kinase 1 (TAK1) with MOS, dual leucine zipper

kinases (DLKs) and zipper sterile alpha kinase (ZAKs), and the mixed lineage kinases

(MLK1-4), (Craig, et al., 2008; Cuevas, et al., 2007). However, this cluster distribution

is different if only the homology in the kinase domain is considered, with ASKs being

grouped with MEKKs, Tpl2 with the TAOs, and the DLKs and ZAKs with MLKs (Craig,

et al., 2008). Moreover, MAP3Ks, also have kinase-independent functions that are

mediated by protein-protein interactions (Mita, et al., 2002). In addition, MAP3Ks, can

regulate other signaling pathways as well, including the NF-κB and Akt/MTOR pathways

(Hirano, et al., 1996).

1.5.1 The MEKK group

The MEKK family includes six members, MEKK (1-4) as well as ASK1 and

ASK2. Each of the MEKK members can regulate multiple biological functions, some of which are not redundant, as evidenced by the lethal effects of knocking out MEKK3 or

MEKK4, but not MEKK1 or MEKK2, on mice embryos (Abell and Johnson, 2005;

Cuevas, et al., 2007; Yang, et al., 2000). MEKK1 was the first to be identified and is one of the best characterized MEKK members. MEKK1 promotes cell survival by activating

ERK1/2 and JNK pathways in response to cytokines and growth factors (Xia, et al., 2000;

Yujiri, et al., 1998). Moreover, MEKK1, upon association with and activation by small G proteins, including RhoA, regulates ERK1/2-mediated cell migration (Cuevas, et al.,

21 2003; Fanger, et al., 1997). Although MEKK1 is a potent activator of ERK1/2, it can also act to negatively regulate ERK1/2 via ubiquitination and subsequent degradation of

ERK1/2 (Lu, et al., 2002). ERK1/2 ubiquitination is mediated by the plextrin homology

(PH) domain of MEKK1 that functions as an E3 ubiquitin (Lu, et al., 2002).

MEKK1 also regulates AP-1-induced (Cuevas, et al., 2005). MEKK2 and MEKK3 are the only MAP3Ks known to regulate the unique MAPK, ERK5, by heterodimerization with and subsequent activation of MEK5 (Sun, et al., 2001).

Dimerization between MEKK2 or MEKK3 and MEK5 requires the PB1 domain found in these three kinases and is a prerequisite for the regulation of ERK5 signaling (Nakamura and Johnson, 2003).

MEKK2 and MEKK3 share around 95% sequence homology in the kinase domain, however, the overall sequence identity is only approximately 55% (Blank, et al.,

1996). Although, MEKK2 and MEKK3 both activate ERK5 and induce AP-1-mediated expression of inflammatory cytokines, MEKK3 also plays a role in regulating the p38 and the NF-κB pathways, while MEKK2 functions to activate the JNK pathway

(Garrington, et al., 2000; Huang, et al., 2004; Kesavan, et al., 2004; Uhlik, et al., 2003;

Xu, et al., 2004).

MEKK4, one of the least characterized MEKK, was shown to activate both JNK and p38 via the direct phosphorylation of MEK4/MEK7 and MEK3/MEK6, respectively

(Abell, et al., 2005; Chi, et al., 2005). The activity of MEKK4 seems to be modulated by numerous upstream proteins. Such proteins include Rac, Cdc42 and TRAF4, all of which trigger JNK signaling by binding to MEKK4 (Abell and Johnson, 2005; Fanger, et al.,

1997; Gerwins, et al., 1997). Other proteins also known to bind to and activate MEKK4

22 are the isoforms α, β, and γ of the small growth arrest and DNA damage-inducible

(GADD45) proteins (Takekawa and Saito, 1998).

ASK1, also known as MEKK5, is activated by numerous stimuli that can result in elevated intracellular levels of reactive oxygen species (ROS) such as stresses, cytokines and LPS (Hayakawa, et al., 2006). One of the first characterized functions of ASK1 is the regulation of stress-induced apoptosis (Tobiume, et al., 1997). ASK1 however, similar to

MEKK4, activates MEK3, MEK6, MEK4 and MEK7 that in turn activate p38 and JNK

(Ichijo, et al., 1997). Moreover, studies focused on the activation of ASK1 by LPS revealed an important role for ASK1 in regulation of the innate immune system

(Matsuzawa, et al., 2005). A role for ASK2, or MEKK6, in apoptosis has also been demonstrated, whereby ASK2 can form a heterodimer with ASK1 to trigger apoptosis

(Ortner and Moelling, 2007; Takeda, et al., 2007).

1.5.2 The Raf MAP3Ks

As mentioned earlier, the Rafs are MAP3Ks that mainly mediate Ras-induced activation of ERK1/2 by phosphorylating MEK1/MEK2 (Dunn, et al., 2005). Three main

Raf proteins, A-Raf, B-Raf and Raf-1 (or c-Raf), have three highly conserved regions

(CR1-3). The CR1, which is localized in the N-terminus and harbors the Ras-binding domain and the cysteine rich domain, mediates the binding between the Rafs and Ras.

The CR2 is serine/threonine rich and regulates Raf activity by undergoing phosphorylation, and the CR3 harbors the C-terminal kinase domain (Mott, et al., 1996;

Vojtek, et al., 1993; Wellbrock, et al., 2004). The Raf proteins have distinct expression patterns and are not functionally redundant (Cuevas, et al., 2007). Raf-1, but not A-Raf or

23 B-Raf, is ubiquitously expressed (Craig, et al., 2008). Both B-Raf and A-Raf are widely

expressed, but A-Raf is mainly found in urogenital and gastrointestinal tissues (Luckett,

et al., 2000). The Raf isoforms can also display differences in function. For instance,

while B-Raf and Raf-1 phosphorylate and activate both MEK1 and MEK2, A-Raf only

activates MEK1 (Sridhar, et al., 2005; Wu, et al., 1996). Moreover, B-Raf appears to be

involved in cellular differentiation through activation of the MAPK, ERK3 (Coulombe, et

al., 2003; Hoeflich, et al., 2006). Raf-1 phosphorylates and inactivates Bad, which

suggests that it has a role in regulating apoptosis (Kebache, et al., 2007). A role for Raf-1

in neuronal and cardiac apoptosis was also demonstrated in mice with cardiac-specific

deletion of Raf-1 (Kanamoto, et al., 2000; Yamaguchi, et al., 2004). A role for Raf-1 in

suppressing apoptosis was also demonstrated in Raf-1 deficient mice, whereby Raf-1-/-

mice die at the embryonic stage (Mikula, et al., 2001). Moreover, Raf-1-/- fibroblasts were

shown to undergo apoptosis more readily than the wild-type cells, which further supports

an anti-apoptotic function for Raf-1 (Mikula, et al., 2001). Activation of ERK1/2 by Raf

proteins can be facilitated by several scaffold proteins including β-arrestin, and the kinase suppressor of ras (KSR) (Morrison and Davis, 2003).

B-Raf is a human oncogene that is mutated at a high frequency (more than 60%) in melanomas (Davies, et al., 2002). b-raf gene mutations are found in a large number of cancers, including colon cancer and ovarian cancer (Dhillon, et al., 2007; Wan, et al.,

2004). The V600E in the activation loop of B-Raf is the most common among b-raf mutations (Wan, et al., 2004). This mutation induces constitutive activation of B-Raf and ERK signaling (Garnett and Marais, 2004). Interestingly, B-Raf, with mutations that don’t enhance activity, also increases ERK activity (Wan, et al., 2004).

24 Such mutations were proposed to indirectly promote ERK signaling by activating Raf-1,

a heterodimerization partner of B-Raf (Garnett and Marais, 2004; Garnett, et al., 2005;

Rushworth, et al., 2006).

1.5.3 The TAK1 MAP3K group

TAK1 is mainly activated by TGFβ but can also be activated by other

proinflammatory stimuli such as IL-1β, TNFα and LPS (Irie, et al., 2000; Takaesu, et al.,

2003; Wagner and Siddiqui, 2007). Three isoforms of TAK1, TAK1a, TAK1b and

TAK1c, exist with TAK1a being the most predominant (Craig, et al., 2008). TAK1

triggers JNK and p38 signaling by activating MEK4 and MEK6, respectively (Shim, et

al., 2005). p38 however, when present in high levels, functions as part of a negative

feedback loop to regulate TAK1 activity (Cheung, et al., 2003). TAK1 is essential for

vesicular development, most likely by inducing the expression of JNK- and p38- dependent genes involved in angiogenesis, and TAK1 deficiency is lethal to mouse embryos (Jadrich, et al., 2006). The most distinctive role of TAK1 that has been identified is the one involving the activation of the NF-κB pathway to regulate cellular responses including inflammation, apoptosis and cell cycle regulation (Ghosh, 1999;

Sakurai, et al., 1998). Activation of the NF-κB by TAK1 requires the formation of a complex between TAK1, TAB2, TAB3 and the E3 ubiquitin ligase/TRAF6 (Wang, et al.,

2001). The complex then phosphorylates and activates IKK in a ubiquitin-dependent manner, a step that is essential for IκBα degradation (Wang, et al., 2001). Briefly, the transcription factor NF-κB is sequestered in the cytosol in an inactive form by the inhibitor IκBα (Karin, et al., 2002). Stimulation with proinflammatory cytokines like

25 TNFα and IL-1, activates the IκB kinase (IKK) kinase (IKKK) that in turn phosphorylates and activates IKK (Figure 6) (Karin, et al., 2002). Once activated, IKK phosphorylates IκBα on two serine residues, Ser32 and Ser36 (Ghosh and Karin, 2002).

Phospho-IκBα is then targeted for ubiquitination and proteasomal degradation, thereby allowing NF-κB to translocate to the nucleus and activate the expression of genes involved in numerous cellular responses (Figure 6) (Karin, et al., 2002).

Figure 6. The NF-κB pathway. NF-κB is maintained in the cytoplasm by its inhibitor

IκBα. Proinflammatory stimuli activate the IKK activator, IKKK, which in turn phosphorylates IKK. Active IKK then phosphorylates IκBα on Ser32 and Ser36.

Phosphorylation on these residues leads to the ubiquitination and subsequent proteasomal degradation of IκBα which leads to the release of NF-κB. NF-κB then translocates to the nucleus where it activates gene expression.

26 1.5.4 The TAO/Tpl2 and Mos MAP3K groups

The TAO kinases are one of the least characterized MAP3Ks. Three TAO

members TAO1, TAO2 and TAO3 have so far been identified (Craig, et al., 2008).

TAO1, in response to stress, mainly activates the p38 pathway by associating with and

activating MEK3 (Hutchison, et al., 1998). DNA damage is another factor that acts as an

activator of TAO (Raman, et al., 2007). This process is mediated by Ataxia telangiectasia mutated (ATM), a cell-cycle protein that binds to and phosphorylates TAO resulting in

p38-dependent DNA repair (Raman, et al., 2007). TAO2, upon activation by osmotic

stress, also functions as a negative regulator of TAK1-induced NF-κB activation

(Huangfu, et al., 2006).

Tpl2 is a MAP3K that activates ERK and JNK signaling in response to LPS,

CD40 and TNF (Dumitru, et al., 2000). In response to these stimuli, Tpl2 is

phosphorylated and activated by IKKβ, an upstream activator of NF-κB, suggesting that

Tpl2 plays a role to coordinate both MAPK and NF-κB signaling (Cho, et al., 2005;

Eliopoulos, et al., 2003). As for c-Mos, it was found to activate ERK1/2 signaling via

direct association and activation of MEK1 (Chen and Cooper, 1995; Okazaki and Sagata,

1995). C-Mos-mediated activation of ERK1/2 stabilizes c-Fos, a main component of the

AP-1 transcription factor, and thus drives AP-1-dependent gene expression (Okazaki and

Sagata, 1995).

1.5.5 The MLK family of MAP3Ks

MLKs, or mixed-lineage kinases, are so named because they exhibit characteristics of both serine/threonine kinases and tyrosine kinases. Out of the eleven

27 protein kinase conserved regions in MLKs, the first seven have sequence similarity to

serine/threonine kinases while the remaining four are more similar to tyrosine kinases

(Gallo and Johnson, 2002). However, since early studies on several members of the

MLKs have demonstrated autophosphorylation on serine and threonine residues and also

phosphorylation of downstream MAP2Ks on serine residues, MLKs are considered to be

serine/threonine kinases (Gallo and Johnson, 2002; Gallo, et al., 1994). The MLK family

is further subdivided into the dual leucine zipper-bearing kinases (DLKs), the zipper-

sterile-α motif kinase (ZAK) and the mixed-lineage kinases (MLKs) subfamilies (Gallo

and Johnson, 2002).

1.5.5.1 The DLK subgroup

The DLK subfamily consists of two members, DLK and the leucine zipper kinase

(LZK) that are approximately 87% identical in the catalytic domains (Gallo and Johnson,

2002). This subfamily is characterized by the presence of two leucine zipper domains that

are 31 amino acids apart and localized downstream of the kinase domain (Gallo and

Johnson, 2002). The DLK members both undergo homodimerization via the leucine zipper domains, a step that is required for autophosphorylation and subsequent activation

of these proteins (Nihalani, et al., 2000). It is noteworthy that, homodimerization of DLK

and LZK occurs independently of both protein phosphorylation and kinase activity (Mata,

et al., 1996). Heterodimerization between DLK and LZK is also possible but involves the

amino terminal ends of both proteins rather than the leucine zipper domain (Nihalani, et al., 2000). However, the association between DLK and LZK is indirect and is thought to be mediated by the scaffold protein JNK-interacting protein 1 (JIP1) (Nihalani, et al.,

28 2000). JIP1 also complexes with and regulates the activities of DLK and LZK to induce

JNK signaling (Ikeda, et al., 2001; Nihalani, et al., 2001). DLK signaling is important in

the neuronal system, where it triggers JNK-dependent synaptic formation and migration

in the cerebral cortex development but induces neuronal cell apoptosis when overexpressed or when cells are deprived of trophic factors (Figure 7) (Harris, et al.,

2002; Hirai, et al., 2006; Mota, et al., 2001; Xu, et al., 2001). In addition to JNK, DLK

can also mediate differentiation and apoptosis by activating p38 (Figure 7) (Craig, et al.,

2008; Gallo and Johnson, 2002).

Figure 7. Signaling of the DLK subfamily of MAP3Ks (adapted from (Craig, et al.,

2008). While LZK mainly induces JNK signaling by phosphorylating MEK4/7, DLK can trigger migration, apoptosis and differentiation through the activation of both

MEK4/7/JNK and MEK3/6/p38 pathways.

29 1.5.5.2 The ZAK subgroup

The ZAK MAP3K, also known as MLTKα, is characterized by the presence of

the sterile-α-motif or (SAM) domain in the middle of the protein sequence just

downstream of the leucine zipper domain (Gallo and Johnson, 2002). A shorter splice

variant of ZAKα, ZAKβ, also commonly known as MLK7, that is highly expressed in

cardiomyocytes, was identified (Gotoh, et al., 2001). This isoform is identical to ZAKα

from the N-terminus to the leucine zipper domain but lacks the SAM domain (Gallo and

Johnson, 2002). Similar to other MAP3Ks, ZAKα activates p38, JNK and ERK

pathways. In lung cancer cells, overexpression of ZAKα suppresses proliferation and AP-

1 transcriptional activity in a JNK- and ERK-dependent manner (Yang, et al., 2010).

ZAKα was shown to be required for doxorubicin-induced pro-inflammatory and pro- apoptotic responses in HaCaT cells, a pseudo-normal keratinocyte cell line, in a mechanism that is mediated by p38 and JNK signaling (Sauter, et al., 2010). However, an oncogenic function for ZAKα was also demonstrated (Cho, et al., 2004). For instance, injecting athymic nude mice with cells stably expressing ZAKα results in the formation of tumors (Cho, et al., 2004). In addition, ZAKα induces transformation of neoplastic cells (Cho, et al., 2004). Thus, ZAKα, through the activation of MAPKs, can either induce tumorigensis or apoptosis. Similar to ZAKα, ZAKβ can induce p38- and JNK- dependent apoptosis upon activation by stresses such as UV irradiation and anisomycin

(Wang, et al., 2005). Moreover, a role for ZAKβ in regulating cardiac function has also been demonstrated. For instance, in transgenic mice that specifically overexpress ZAKβ in the cardiac tissue, stimulation of ZAKβ by beta-adrenergic stimuli activates both p38 and JNK and leads to cardiac myocyte hypertrophy as well as mice mortality (Christe, et

30 al., 2004). ZAKβ can also induce JNK activation in response to saturated free fatty acids

or FFA (Jaeschke and Davis, 2007).

1.6 The MLK subfamily

Similar to other MAP3Ks, several members of the MLK subfamily activate multiple MAPK signaling pathways. The four MLK members (MLK1-MLK4) all share similar domain structures that consist of an N-terminus Src-homology-3 (SH3) domain, a kinase domain, a leucine zipper (LZ) domain, a Cdc42/Rac-interacting binding (CRIB) domain and a proline rich C-terminus (Figure 8) (Gallo and Johnson, 2002). In addition to these domains, MLK3 is also characterized by a Gly-Pro-rich N-terminus (Gallo and

Johnson, 2002). Unlike the different structural domains that have more than 60% sequence homology between the four members, the C-terminal domain is highly diverse

(Gallo and Johnson, 2002). The kinase domain of all MLK subfamily members contains a putative activation region with a TTXXS motif (Leung and Lassam, 2001).

H2N SH3 Kinase Zipper CRIB COOH

Figure 8. The structural domains of MLKs. The four MLK members share sequentially an N-terminal SH3 domain, a kinase domain, a leucine zipper domain, a

CRIB domain and a proline-rich C-terminal region.

31 1.6.1 MLK1 and MLK2

MLK1 and MLK2 show a more restricted expression than other MLKs, with

MLK1 being mainly found in epithelial cells and MLK2 in the brain as well as in skeletal

tissue (Gallo and Johnson, 2002). MLK2 strongly activates JNK by directly

phosphorylating and activating MEK4 (Hirai, et al., 1997). Moreover, MLK2 when

overexpressed in COS-1 cells slightly activates the p38 and ERK pathways as well (Hirai,

et al., 1997). In Xenopus, MLK2-induced JNK activation can be inhibited by the p21-

GTPase activated kinases, PAK1 (Poitras, et al., 2003). MLK2 was also implicated in

neuronal development and survival since it phosphorylates and activates NeuroD (ND), a

transcription factor that functions as a key regulator of these processes (Marcora, et al.,

2003). In a more recent study, MLK2 was found to directly bind to, phosphorylate and

inactivate the transcription factor E47 in a mechanism that is JNK-dependent (Pedraza, et

al., 2009). This leads to a reduction in the protein levels of the TrkB receptor, whose

expression is regulated by E47 (Pedraza, et al., 2009). As for MLK1, very little is known

except for a study that addressed the physiological functions of MLK1 and MLK2

subfamily members in vivo. In this aspect, disrupting the gene encoding MLK1 or MLK2 did not produce any phenotypical or developmental defects in mice (Bisson, et al., 2008).

Moreover, mice with null mutations in both MLK1 and MLK2 were also viable, fertile and healthy (Bisson, et al., 2008). These results suggest that MLK1 and MLK2 do not have a significant role in development and may have functions that are compensated for by other MAP3Ks.

32 1.6.2 MLK3 activation

MLK3 is the best characterized MLK subfamily member. This kinase is

ubiquitously expressed (Gallo and Johnson, 2002). In MLK3, the TTXXS motif (residues

276-281) undergoes autophosphorylation on T277 and S281, a step that is required for

MLK3 activity (Leung and Lassam, 2001). Phosphorylation of these sites by hematopoietic protein kinase 1 (HPK1), leads to the activation of MLK3 (Leung and

Lassam, 2001). This motif is conserved in the other MLK subfamily members as well, and it is believed that the kinase activity of other MLKs is also regulated by autophosphorylation. MLK3 dimerizes through the LZ domain (Leung and Lassam,

2001; Vacratsis and Gallo, 2000). Dimerization is not a prerequisite for

autophosphorylation but it is required for MLK3-mediated activation of JNK signaling

(Gallo and Johnson, 2002). In the absence of a stimulant, the N-terminal SH3 domain of

MLK3 binds to a single proline residue that lies between the LZ and the CRIB domain, and maintains MLK3 in an inactive state by preventing dimerization and autophosphorylation (Figure 9) (Zhang and Gallo, 2001). Members of the Rho family of

GTPases, such as Cdc42, interact with and activate MLK3 (Bock, et al., 2000; Teramoto, et al., 1996). Co-expressing active Cdc42 with MLK3 results in oligomerization and enhanced activation of MLK3 (Bock, et al., 2000; Leung and Lassam, 2001; Teramoto, et al., 1996). Moreover, both active Cdc42 and Rac were found to bind to the CRIB motif of

MLK3 (Burbelo, et al., 1995). Although the exact mechanism by which Cdc42 or Rac activate MLK3 is not well understood, the proposed model suggests that upon binding of active Cdc42 or Rac to the CRIB domain of MLK3, the autoinhibitory effect of the SH3 domain is relieved, thereby promoting MLK3 dimerization, autophosphorylation and

33 subsequent activation (Figure 9) (Gallo and Johnson, 2002). Active Cdc42 and Rac also

function to localize MLK3 to the plasma membrane and other cellular membrane

compartments where it mediates spatial MAPK signaling (Du, et al., 2005; Gallo and

Johnson, 2002). In addition to Cdc42 and Rac, the or Akt1 also plays a role in regulating MLK3 activity by phosphorylation of MLK3 on Ser674, which inhibits

MLK3 kinase activity (Barthwal, et al., 2003). Another physiological inhibitor of MLK3 is the tumor suppressor Neurofibromatosis-2 (NF-2) protein, merlin (Zhan, et al., 2011).

The proline-rich MLK C-terminus has a cluster of phosphorylation sites that are potential targets for the SH3-containing kinases, such as MAPKs and cyclin-dependent kinases (Phelan, et al., 2001; Vacratsis, et al., 2002). A consensus sequence for MAPK phosphorylation (PXS/TP) is present in the MLK3 C-terminus and is phosphorylated by

JNK (Schachter, et al., 2006). This phosphorylation triggers a positive feedback effect that leads to MLK3 activation and hence sustained JNK signaling (Schachter, et al.,

2006). Other than MLK3, JNK was also found to associate with and phosphorylate

MLK2 (Sakuma, et al., 1997).

34

Figure 9. Model mechanism of MLK3 activation by Cdc42 (adapted from (Gallo and Johnson, 2002)). The N-terminal SH3 domain of MLK3 has an autoinhibitory effect and maintains MLK3 in an inactive form. Binding of GTP-bound Cdc42 to the CRIB domain relieves this autoinhibitory effect which subsequently allows MLK3 dimerization, autophosphorylation and activation. P = Proline residue.

35 1.7 MLK3 signaling

Of the four MLK subfamily members, MLK3 has been best characterized in terms

of its biological and biochemical function. MLK3, like other MAP3Ks, is activated by

stress, cytokines, and ceramides, but, how these different stimuli activate MLK3 is still

unclear (Kyriakis and Avruch, 2001). MLK3, once activated by dimerization and

autophosphorylation, phosphorylates and activates the MAP2Ks, MEK4/7 and MEK3/6,

that in turn phosphorylate and activate JNK and p38, respectively (Tibbles, et al., 1996).

The activated MAPKs then activate various transcription factors involved in mediating

different cellular responses such as proliferation, inflammation and apoptosis (Gallo and

Johnson, 2002). For instance, MLK3 activates MEK7 and triggers JNK3-dependent cell

death in the rat hippocampus (Pan, et al., 2005). MLK3 overexpression induces JNK-

dependent apoptosis in superior cervical ganglion (SCG) neurons, and withdrawal of

neuron growth factor (NGF) activates and induces apoptosis (Mota, et al., 2001). In accordance with such a role, transient expression of wild-type MLK3 (MLK3-WT) or

kinase dead MLK3 (MLK3-KD) induced poly (ADP-ribose) polymerase (PARP)

cleavage and enhanced etoposide-induced cell death in HEK293 cells (Cole, et al., 2009).

However, MLK3-WT enhances etoposide-induced cell death to a greater extent than

MLK3-KD, suggesting an important function for the kinase activity of MLK3 in this mechanism (Cole, et al., 2009). Moreover, knocking down MLK3 elevated the level of active IKK in NIH3T3 and HEK293 cells, demonstrating an inhibitory role for MLK3 on

IKK activity (Cole, et al., 2009). A recent study also established a role for MLK3 in intestinal epithelial cell migration (Kovalenko, et al., 2012). This was demonstrated in the

epithelial intestinal Caco-2 cells, whereby, treating the cells with the MLK inhibitors,

36 CEP-1347 and CEP-11004, reduced ERK and JNK signaling and migration of cells across collagen substrates (Kovalenko, et al., 2012). Thus, both the stimulus and cell type dictate the signaling pathway activated by MLK3 and the resulting cellular response.

A few studies have also addressed the role of MLK3 in vivo. Similar to MLK1-/-

/MLK2-/- mice, mice deficient in MLK3 are viable and had a similar life span to wild-type mice and only showed minor morphological defects indicated by a reduction in the thickness of the dorsal epidermal tissues (Brancho, et al., 2005). MLK3-/- mice also displayed defects only in TNF-induced JNK activation suggesting a role for MLK3 in mediating this pathway (Brancho, et al., 2005). More recently, targeted deletion of MLK3 in mice resulted in reduced bone mass and several other skeletal defects (Zou, et al.,

2011). Similar effects were observed in mice expressing a mutant allele of mlk3 that is incapable of activation by FYVE, RhoGEF, and PH domain-containing 1 (FGD1), which was shown in vitro to act as an upstream activator of MLK3 (Zou, et al., 2011). These results demonstrate a role for MLK3 in mediating FGD1 regulatory functions in bone development (Zou, et al., 2011). Furthermore, MLK3 knockout mice revealed reduced jejunal mucosal ulcer healing, as compared to wild-type mice, that was also associated with a decrease in ERK and JNK signaling (Kovalenko, et al., 2012). These observations also implicate a role for MLK3 in regulating intestinal ulcer healing in vivo (Kovalenko, et al., 2012). In addition to all these functions, MLK3 was also shown to play important roles in cancer.

37 1.7.1 MLK3 signaling in cancer

Cancer is a multistep process that is initiated by alterations in one or more essential genes such as tumor suppressor genes or oncogenes leading to uncontrolled cell proliferation and formation of tumors (Hanahan and Weinberg, 2011). Additional genetic/epigenetic modifications promote tumor cells to undergo different processes such as transformation, migration and angiogenesis all of which allow tumors to gain metastatic potential (Hanahan and Weinberg, 2011; Umar, et al., 2002). Interestingly, a role for MLK3 in the induction of apoptosis was characterized in several tumor cell lines.

In FaO rat and Hep3B human hepatoma cells, MLK3 activates the p38 but not the JNK apoptotic pathway (Kim, et al., 2004). In the human ovarian cancer epithelial cells,

SKOV3, silencing MLK3 reduced the protein levels of the NF-κB inhibitor, the inhibitor of kappa B alpha (IκBα), implicating MLK3 as a negative regulator of the NF-κB cell survival pathway in these cells (Cole, et al., 2009).

A role for MLK3 in cell proliferation has also been documented in fibroblasts and tumor cells with mutations in the oncogene K-ras or loss of function mutations in neurofibromatosis-1 (NF1) or NF2, through activation of JNK or p38 and ERK, respectively (Brancho, et al., 2005; Chadee and Kyriakis, 2004; Hartkamp, et al., 1999) .

A hallmark of malignant tumors is their invasive capacity (Hanahan and Weinberg,

2011). A role for MLK3 in cell invasion was also implicated. In this respect, MLK3 was shown to be essential for JNK/AP-1-mediated breast cancer cell migration and invasion

(Chen, et al., 2010). A requirement for MLK3 in cell migration was also demonstrated in gastric tumor cells and the lung cell line, A459 (Mishra, et al., 2010; Swenson-Fields, et al., 2008). Recently, MLK3 was also found to promote tumorigenesis in vivo (Cronan, et

38 al., 2012). Orthotopic xenograft models generated from stable MLK3 shRNA-knockdown

MDA-MB-231 cells, exhibited a reduction in tumor growth and metastasis as compared

to control models (Cronan, et al., 2012). Thus, active MLK3, through activation of JNK,

p38 or ERK can induce proliferation, invasion and cell survival pathways in a stimulus

and cell type-specific manner.

1.8 MLK4: characteristics and function

MLK4 represents one of the least characterized members of the MLK family. In

one report, MLK4 was found to be mutated in 6.8% of the colorectal cancers analyzed

(Bardelli, et al., 2003). In half of these cases, the mutations were found in exon 1 or exon

2 that encode for the kinase domain (Bardelli, et al., 2003). However, another study

reported no observed MLK4 mutations in any of the 46 colorectal cancer samples tested

and out of the 24 different colorectal cancer cell lines screened, the mutation (R470C)

was present in only two of the cell lines, LoVo and CaR1 (Shao, et al., 2007). A recent

study showed that while wild-type MLK4 exhibits low kinase activity, the R470C mutation along with the H261Y, G291E and R555Stp mutations significantly enhanced the kinase activity (Martini et al. 2013). The frequency of these mutations, however, is

very low (Martini et al. 2013). Moreover, no MLK4 kinase domain mutations were

detected in gastric and hepatocellular carcinomas (Soung, et al., 2006). Therefore, the

relationship between MLK4 mutation and cancer development remains elusive.

MLK4 exists in two forms, α (580 aa) and β (1036 aa), that are generated by

alternative splicing. The two isoforms are identical in all structural domains except for

the C-terminal region (Kashuba, et al. 2011). Expression patterns of MLK4α and MLK4β

39 revealed that although the mRNA expression of both forms was detected in the pancreas,

kidney, liver, lung and brain, MLK4α exhibited a lower level of expression in comparison

to MLK4β (Kashuba, et al. 2011). However, skeletal muscle only expressed MLK4β

(Kashuba, et al. 2011). Overexpression experiments revealed differences in the

intracellular localization between the two isoforms with MLK4α being found only in the

cytoplasm and MLK4β in both the cytoplasm and the nucleus (Kashuba, et al. 2011).

Although transgenic mice expressing MLK4β showed no phenotypic abnormalities,

MCF7 cells with ectopic expression of either MLK4α or MLK4β exhibited reduced colony formation and cell proliferation (Kashuba, et al. 2011). Nonetheless, the biochemical and biological function of MLK4 remains largely unidentified. MLK4 shares >70% sequence homology with MLK3 in the kinase catalytic domain (Kashuba, et al. 2011). This might suggest that MLK4, like MLK3, functions as an upstream activator of MAPK signaling. However, in a recent study, MLK4 was found to negatively regulate

Toll-like receptor 4 (TLR4) signaling by inhibiting LPS-induced JNK and ERK activation (Seit-Nebi, et al., 2012). Therefore, further elucidating the signaling specificity of MLK4 is critical in unraveling its biological significance and determining whether it adopts similar functions to MLK3 and other MAP3Ks.

1.9 Significance

Many MAP3Ks, being activators of MAPK signaling, have pivotal roles in cancer and other pathologies. Increasing evidence implicates a role for MLK3 in different stages of cancer, yet very little is known about the other members of the MLK subfamily

40 members. Unraveling the biological functions of the other MLKs is essential to reveal any contributions to cancer and other diseases and can thus provide insights into new therapeutic approaches. Elucidating the role of MLK4, one of the least characterized

MAP3Ks, is in line with this goal. Therefore, this study is directed at identifying the role of MLK4β in MAPK signaling and ovarian cancer.

41 Chapter 2

Materials and Methods

2.1 Cell culture

Human colon cancer (HCT116 and HT-29), ovarian cancer (SKOV3) cells, human embryonic kidney (HEK293) and human glioblastoma (T98G) cells were obtained from the American Type Culture Collection, Manassas, VA, USA. The ovarian cancer cell line, HEY1B, was a gift from Dr. Douglas Leaman (The University of Toledo). T80 cells are immortalized human ovarian epithelial cells (Liu, et al., 2004). HCT116,

SKOV3 and HEY1B cells were cultured in growth media made from DMEM

(Mediatech, Herndon, VA, USA) and supplemented with 10% FBS (fetal bovine serum)

(Hyclone, Logan, UT, USA). T80 cells were cultured in medium 199 (Mediatech) with

10% MCDB 105 (Sigma-Aldrich, St. Louis, MO, USA) and 10% FBS. All tissue culture media were supplemented with 25 µg/ml streptomycin and 25 I.U. penicillin

(Mediatech). Cells were cultured in a humidified atmosphere with 5% CO2 at 37 °C. For treatments, HCT116 cells were treated with 0.5 M sorbitol (Fisher Scientific, Fairlawn,

NJ, USA).

42 2.2 Expression Vectors

The mammalian expression vectors used in this study are as follows:

pCMV5-FLAG empty vector, cDNA constructs for expression of human MAPKs:

pCMV-GST-MLK3, pCMV5-FLAG-MLK3, pCMV-FLAG-MLK4β, pCMV-HA-

MLK4β, pCDNA3-HA-ASK1, and pEBG-SEK-1-KR (GST fusion protein).

2.3 Plasmids and siRNA transfections

Transient transfections were performed using either Lipofectamine (Invitrogen,

Frederick, MD, USA) or PolyJetTM In Vitro DNA Transfection Reagent (SignaGen

Laboratories, Rockville, MD, USA). Cells were plated in 6 cm dishes and transiently

transfected when they reached 50-60% confluency. For Lipofectamine transfection,

depending on the plasmid used, (0.7-2) μg of DNA was added to 800 μl of serum-free

media. 8 μl of Lipofectamine was then mixed with the DNA suspension and incubated for

15 min. The cells were washed once with serum-free media and then 1.7 ml of serum-free

media was added to dish. The DNA-Lipofectamine mixture was then added to the cells.

The cells were returned to the incubator for 4-5 hours (h). After the incubation, 2.5 ml of

media containing 20% FBS was added to each dish and the cells were incubated for at

least 16 hours before harvesting. For 10 cm dishes, (1.4-4) μg of DNA and 16 μl of

Lipofectamine in 1.6 ml were used. Transfection using PolyJet was performed as

suggested by the manufacturer. Briefly, 0.7 – 2 μg and 1.4 – 4 μg of DNA was added to

100 μl (for 6 cm dishes) and 250 μl (for 10 cm dishes) of serum-free media, respectively.

In a separate tube containing these same volumes of serum-free media, 5 μl and 10 μl of

Polyjet was added, respectively. The Polyjet-media mix was then added to the DNA

43 suspension, mixed and incubated for 15 min. 2.8 ml and 5 ml of growth media were added to 6 cm and 10 cm plates, respectively. Two hundred μl (6 cm dish), or 500 μl (10 cm dish) of the Polyjet-DNA complex solution was then added dropwise to each dish.

Transfected cells were incubated for at 18 hours before harvesting.

SiRNA transfection was carried out using either Lipofectamine 2000 (Invitrogen,

Frederick, MD, USA) or GeneMute (SignaGen Laboratories, Rockville, MD, USA). For

Lipofectamine 2000 transfection, 100 nM of siRNA and 8 μl of Lipofectamine 2000 were used and transfection was performed as described earlier. For GeneMute transfection,

100 nM siRNA and 9 μl of the GeneMute reagent were added to 300 μl of 1X transfection buffer; For 10 cm dishes, 100 nM siRNA and 24 μl of GeneMute were added to 800 μl of the transfection buffer. The transfection mixtures were mixed well and incubated for 15 min. Cells were washed with 1X PBS and 3 ml and 8 ml of growth media were added to 6 cm and 10 cm plates, respectively. The transfection solution was then added to the cells and the cells were placed in the incubator for 15 min after which the transfection media was replaced with fresh full growth media. The cells were then incubated for at least 16 hours before harvesting. SiRNA in both transfection techniques was added to a final concentration of 100 nM. The siRNA oligonucleotide sequences are as follows: human MLK4 siRNA1: 5’-GGGCAGTGATGACTGAGAT-3’ corresponding to nucleotides 1469-1487; human MLK4 siRNA2 (Ambion, Life Technologies, Grand

Island, NY, USA): 5’-GGACCACCAAAATGAGCAC-3’ corresponding to nucleotides

903-923; human MLK3 siRNA1 (Chadee and Kyriakis, 2004): 5’-

GGGCAGTGACGTCTGGAGTTT-3’ corresponding to nucleotides 1198-1218; human

MLK3 siRNA2: 5’-AAGCGCGAGATCCAGGGTCTC-3’ corresponding to nucleotides

44 1154-1172; non-targeting siRNA that is used as a control (Dharmacon, Lafayette, CO,

USA): 5’-AATCCGGTTGCATAGTTCATC-3’.

2.4 Immunoblotting

Protein expression was studied by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting. For this purpose, 40 ml of 15%

polyacrylamide running gel solution consisting of 20 ml of bis acrylamide (29.8 g

acrylamide and 0.2 g bisacrylamide in 100 ml H2O), 10 ml of 4X Tris-Cl/SDS, pH 8.8

(0.5 M Tris base, 13.87 mM SDS), 160 μl of 50% ammonium persulfate (APS), 16 μl of

TEMED and 10 ml of H2O were prepared. The stacking gels were prepared using 1.3 ml

of 30% acrylamide (29.2 g acrylamide and 0.8 g bisacrylamide in 100 ml H2O), 2.5 ml of

4X Tris-Cl/SDS, pH 6.8, 70 μl of 50% APS, 20 μl of TEMED and 12.2 ml of H2O.

The gels were placed in running buffer (25 mM Tris base, 192 mM glycine and

6.94 mM SDS) and electrophoresis was carried out overnight at 60 volts. Proteins were

then transferred to Immobilon-P Polyvinylidene Fluoride (PVDF) membranes at 57 V for

2 hours using transfer buffer (25 mM Tris base, 192 mM glycine and 20% methanol).

After transfer, membranes were stained with Brilliant Blue R250 solution (40% (v/v)

methanol, 10% (v/v) acetic acid and 0.4 g Brilliant Blue R250) and destained with 70%

(v/v) methanol solution. The membranes were then blocked for an hour on a rotator with

5% (w/v) nonfat dry milk in 1X PBS (prepared from 10X PBS: 137 mM NaCl, 27 mM

KCl, 43 mM KH2PO4 and 14 mM Na2HPO4 pH 7.4). Following blocking, the membranes

were incubated with primary antibodies prepared at a 1:1000 dilution in antibody buffer

that consists of 1X PBS, 0.05% (v/v) (Tween 20, and 5% (w/v) nonfat dry milk. For

45 phosphospecific antibodies, 5% bovine serum albumin (BSA) was used instead of milk in

the antibody buffer. The membranes were then rotated overnight at 4 °C. The next

morning, the membranes were washed in 1X PBS containing 0.05% (v/v) Tween 20, 4

times each for 15 min. Membranes were then incubated in secondary antibody for 2 hours

at room temperature with rotation. The secondary antibody was prepared at a 1:5000

dilution in 5% (w/v) nonfat dry milk dissolved in 1X PBS with 0.05% (v/v) Tween 20

(PBST). The membranes were then washed 4 times in PBST. Protein expression was

detected by developing the membranes with Immobilon-P development solution

(Millipore, Billerica, MA, USA) or with Enhanced Chemiluminescence (ECL) solution.

Immunoblotting was performed with the following antibodies: MLK3 (C-20), p38

(C-20), ERK2 (C-14), JNK (C-17), GST (Z-5), IκBα (C-21), β-Actin (C-4) (Santa Cruz

Biotechnology, Santa Cruz, CA, USA). The MLK4 antibody specifically recognizes the

C-terminus of MLK4β and does not cross-react with MLK4α (Novus Biologicals,

Littleton, CO, USA). Activation-state p-p38 (Thr180/Tyr182), p-ERK (Thr202/Tyr204), p-JNK (Thr183/Tyr185), p-MLK3 (Thr277/Ser281), p-MEK3/MEK6 (Ser189/207) and p-SEK1 (Thr261) antibodies were from Cell Signaling Technology (Beverly, MA, USA).

FLAG antibody was from Stratagene (La Jolla, CA, USA). Anti-MEK6 antibody was from Stressgen Bioreagents Corporation (Victoria, British Columbia, Canada). Anti-HA monoclonal antibody was prepared from 12CA5 B-cells. The secondary antibodies were

Immun-Star Goat Anti-Mouse (GAM) – Horseradish Peroxidase (HRP), Immun-Star

Goat Anti-Rabbit (GAR) – HRP Conjugate (Bio-Rad, Alfred Nobel Drive, CA, USA) and Peroxidase-IgG fraction monoclonal mouse Anti-rabbit IgG, light chain specific

(MAR) (Jackson Immunoresearch Laboratories Inc., Baltimore, PA, USA).

46 2.5 Preparation of whole cell extracts and treatments

Whole cell extracts were prepared by first washing the cells with 1X PBS and

then scraping them into 6X SDS sample buffer (300 μl and 600 μl for 6 cm and 10 cm

dishes, respectively) and transferring them to 1.5 ml microcentrifuge tubes. The sample

buffer consisted of 70% (v/v) of 1M Tris-Cl, pH 6.8, 30% (v/v) glycerol, 346.8 mM SDS

and 5% (v/v) β-mercaptoethanol.

Some experiments involved treating the cells with TNFα (Biosource, Camarillo,

CA, USA) or the ERK or JNK inhibitors. TNFα, treatments were performed at a final concentration of 10 nM or 10 ng/ml and depending on the experiment, the cells were treated for 10 min, 20 min or 6 h. For inhibiting ERK and JNK activities, U0126 and

SP600125 were used, respectively. The inhibitors were added to the cells at a final concentration of 10 μM for U0126 (LC Laboratories, Woburn, MA, USA) and 50 μM for

SP600125 (ENZO Life Sciences, Farmingdale, NY, USA), and cells were incubated with

the inhibitors for 4 h. The same volume of the vehicle, DMSO, was added to one dish as

a negative control.

2.6 Immunoprecipitation

Immunoprecipitation of endogenous or overexpressed proteins was prepared by

lysing the cells on ice with 1 ml of lysis buffer that consists of 20 mM Tris, pH 7.4, 2

mM EGTA, 10 mM MgCl2, 0.1% β-mercaptoethanol, 1% Triton X-100, 100 μM phenylmethane sulphonyl fluoride (PMSF), 2 μM leupeptin and 2 μM pepstatin, 1 μM

aprotinin. The lysed cells were collected in 1.5 ml microcentrifuge tubes and centrifuged

for 10 min at 13,000 rpm. 350 μl of the lysates were set aside for analysis by

47 immunoblotting. In a separate tube, the remaining 650 μl of the cell lysates were added to

35 μl of Protein-G sepharose beads (Pierce, Thermo Fisher Scientific Inc.) and 35 μl of

anti-HA antibody. For GST pull-downs, 650 μl of cell lysates were added to 35 μl of

glutathione-sepharose beads (Pierce, Thermo Fisher Scientific Inc.). Protein samples

were rotated overnight at 4°C. The beads were then pelleted by centrifugation at 13,000

rpm for 2 min and the supernatant was discarded. The beads were then washed once with

high salt wash buffer (20 mM Tris pH 7.4, 2 mM EGTA, 10 mM MgCl2, 0.1 (v/v) % β-

mercaptoethanol, 0.1 % (v/v) Triton X-100, 100 μM PMSF, 1 μM aprotinin, 2 μM

leupeptin, 2 μM pepstatin and 250 mM LiCl) and twice with wash buffer (same as high

salt wash buffer but lacks the LiCl). The beads were then suspended with 35 μl of 6X

SDS sample buffer and boiled for 5 min before analysis SDS-PAGE.

2.7 MLK4 kinase assay

Lysates were prepared from HCT116 cells overexpressing human MLK4 (HA-

MLK4 or FLAG-MLK4), the negative control (pCMV5-FLAG) and the positive controls

(HA-ASK1 or FLAG-MLK3). The lysis buffer consisted of 20 mM Hepes, pH 7.4, 2 mM

EGTA, 50 mM β-glycerophosphate, 1 mM DTT, 1 mM Na3VO4 (prepared from a 20 mM stock that is heat-activated), 1% (v/v) Triton X-100, 10% (v/v) glycerol, 2 μM leupeptin,

1:1000 aprotinin and 400 μM PMSF. The lysates were then centrifuged at 13,000 rpm for

10 min. Three hundred and fifty μl of the lysates were kept for immunoblotting and the

remaining 650 μl were added to 35 μl of Protein-G beads sepharose beads with either 35

μl of anti-HA or 0.5 μl of anti-FLAG antibody. Immunoprecipitation was performed by rotating the samples at 4°C for an hour and a half. The beads were then pelleted by

48 centrifugation for 2 min at 13,000 rpm. The supernatant was then discarded and the beads were washed once with lysis buffer, once with high salt wash buffer (500 mM LiCl, 100 mM Tris-Cl, pH 7.6, 0.1 % (v/v) Triton X-100, 1mM DTT) and three times in assay buffer (20 mM Mops, pH 7.2, 2 mM EGTA, 10 mM MgCl2, 1 mM DTT, 0.1% (v/v)

Triton X-100). After the last wash, beads were resuspended in 40 μl of assay buffer. The substrate (GST-SEK1-KR or the human H2A histone (New England, Biolabs, Ipswich,

MA, USA)) in 18 μl of assay buffer and 15 μl of Mg-ATP mix (50 mM MgCl2 and 0.5 mM ATP) were then added to each tube. For the radioactive kinase assay, 57 μCi of γ32p

[ATP] (PerkinElmer, Waltham, MA, USA) were added to the Mg-ATP mix before addition to each tube. The samples were then incubated for 30 min at 30 °C with shaking at 450 rpm. Twenty μl of 6X SDS buffer was then added to the samples and they were boiled for 5 min and then analyzed by 15% SDS-PAGE. The MLK4 kinase activity was detected by probing with phospho-SEK1 antibody and autoradiography.

2.8 Cell proliferation assay

HCT116 were either transfected with nonspecific or MLK4 siRNA oligo. 18-20 hours post-transfection, 10,000 cells/well, from each sample, were seeded in 24 well plates. The effect of MLK4 on cell proliferation was then studied by using the MTT

(Thiazolyl blue tetrazolium bromide) assay (Sigma-Aldrich, Saint Louis, MO, USA) at days 0, 3 and 4 as described by the manufacturer. Briefly, on the first day after plating, 25

μl of MTT solution (50 mg of MTT reagent dissolved in 10 ml of 1x PBS pH 7.4) was added in each well. The plates were then incubated at 37 °C for 2 hours. 500 μl of MTT buffer (isopropanol with 10% (v/v) Triton X-100 and 0.1 N HCl) was then added to each

49 well and the contents of the wells were mixed thoroughly by repeated pipetting. The

plates were incubated at room temperature on the shaker for half an hour after which

absorbance of MTT formazan, that is generated by live cells, was measured at a

wavelength of 570 nm on an ELISA plate reader. The graphs represent the fold change in

proliferation over time. All experiments were run in triplicate and repeated twice.

Statistical analysis of data was performed by using the student t-test.

2.9 Luciferase assay

The luciferase assay was used to determine the effect of MLK4β on NF-κB

activity in HEK293 cells. In this assay, the reporter plasmid 3XkappaB-Luc (a gift from

Dr. Brian Ashburner, University of Toledo) was used. This reporter contains three copies

of the DNA sequence that encodes for the NF-κB , in the major

histocomptability complex class I gene, upstream of the luciferase reporter gene. Briefly,

cells were seeded in 12-well plates and 24 hours later were transfected with either the

empty vector, 100 ng of FLAG-MLK4β or 200 ng FLAG-MLK4β together with 100 ng

3XkappaB-Luc reporter plasmid and 100 ng β-Gal plasmid using Polyjet as described earlier. 18 h later, the cells were either left untreated or treated with 10 ng/ml of TNF-α for 6 h. The luciferase assays were then performed (Promega Biosciences, CA, USA).

Luciferase activity was normalized to β-Gal activity. The assays were all run in triplicate.

2.10 Invasion assays

Cell invasion assays were performed with 100 μl of 1 mg/ml BD matrigel matrix

(BD Bioscience, San Jose, CA, USA), mixed with serum free media, in 24-well

50 Transwells with 8.0 μm pore size and 24 mm diameter polycarbonate membrane

(Corning, Acton, MA, USA). Cells were washed in DMEM containing 0.5% (v/v) FBS.

10,000 cells were seeded onto the upper chamber. Matrigel and cells remaining in the

upper chamber were removed after 36 h incubation. The cells on the underside surface of

the membrane were fixed using the Diff Quick Stain Kit (Fisher Scientific, Pittsburgh,

PA, USA) and the number of cells per field of view was counted. The average of number

of cells from four different fields of view was calculated. All experiments were run in triplicate and repeated three times.

2.11 Gelatin zymography

MMP-2 and MMP-9 enzymatic activities in the cell culture medium were determined by SDS-PAGE gelatin zymography. Cells were seeded at 5x105 cells per 6

cm dish. Conditioned cell culture medium from each dish was normalized to an equal

amount of protein (100μg), denatured in the sample buffer in the absence of reducing

agents, and subjected to 10% SDS-PAGE containing 0.1% (w/v) gelatin (Acros Organics,

Morris Plains, New Jersey, USA). The gel was incubated in the presence of 2.5% (v/v)

Triton X-100 at room temperature for 2 h and then at 37 °C for 40 h in digestion buffer

containing 1% (v/v) Triton X-100, 10 mM CaCl2, 0.15M NaCl, and 50 mM Tris (pH 7.5).

Gels were stained with Coomassie Blue R-250 and destained with a buffer containing

10% (v/v) methanol and 5% (v/v) acetic acid. Proteolysis was detected as a white band

against a blue background (Schmalfeldt, et al., 2001).

51

Chapter 3

Results

3.1 The role of MLK4β in p38 signaling

p38 is activated in response to different stresses and pro-inflammatory cytokines

(Raingeaud, et al., 1995). Once activated, p38 regulates a variety of cellular responses

including invasion, inflammation and transformation (Hanahan and Weinberg, 2011; Lee,

et al., 1994; Shin, et al., 2005). Multiple MAP3Ks, including MLK3, mediate stimulus- induced p38 activation (Cuenda and Rousseau, 2007; Kim, et al., 2004; Tibbles, et al.,

1996). However, whether MLK4β affects p38 signaling is yet unknown.

3.1.1 The effect of ectopic expression of MLK4β on p38 activation

MLK4β shares high sequence homology with MLK3 (Kashuba et al. 2011).

Therefore, we postulated that MLK4β, similar to MLK3 and other MAP3Ks, is an

upstream activator of the p38 pathway. To test this possibility, MLK4β was ectopically

expressed in HCT116 cells, human colon cancer cells that constitutively express active

Ras, and the levels of active p38 were analyzed using an activation state phospho-specific

p38 antibody (Thr180/Tyr182; phosphorylated-p38 (p-p38)). Surprisingly, cells

overexpressing MLK4β had substantially reduced levels of p-p38 in comparison to that

52 observed in cells transfected with an empty vector, while total p38 levels were unaffected

(Figure 10A, left panel). External osmolarity conditions are critical for cell integrity and

any changes, whether hypo- or hyperosmotic, induce osmotic stress that affects different

cellular processes, including cell survival (Lezama, et al., 2005). The p38 pathway is

activated in response to the hyperosmotic stress inducer, sorbitol (Chan, et al., 2008). To determine whether MLK4β overexpression would affect sorbitol-induced p38 activation,

HCT116 cells were treated with sorbitol. Sorbitol treatment increased the levels of p-p38 in cells transfected with the empty vector (Figure 11A, right panel). MLK4β overexpression, however, completely blocked p38 activation by sorbitol (Figure 10A, right panel). These results show a negative regulatory effect for MLK4β on p38 activity.

p38 signaling is also activated by cytokines (Kyriakis and Avruch, 2001;

Raingeaud, et al., 1995). To test whether MLK4β also affects the activation of p38 by

cytokines, HCT116 cells were treated with the pro-inflammatory cytokine, TNFα, and the effect of MLK4β overexpression on p38 activity was analyzed. Similar to the results obtained with sorbitol, TNF-α treatment elevated the levels of active p38 in cells expressing the empty vector alone (Figure 10B). Overexpressing MLK4β reduced the

levels of TNFα-induced p-p38, while the total levels of p38 remained unchanged (Figure

10B). These results suggest that MLK4β negatively regulates sorbitol- and TNFα –

induced p38 activition.

53 A. B.

Figure 10: MLK4β overexpression inhibits basal and stimulus-induced p38 activation. (A) HCT116 cells were transiently transfected with pCMV5-FLAG or

FLAG-MLK4β and left untreated, or treated with 0.5 M sorbitol for 30 min. Cell lysates were immunoblotted with FLAG, p38, p-p38 and β-Actin antibodies. (B) HCT116 cells were transfected as described in (A) and left untreated or treated with 10 μM TNFα for 20 min. Cell lysated were immunoblotted as described in (A).

3.1.2 The effect of endogenous MLK4 on the activation of p38

The data in Figure 10 were obtained upon ectopic expression of MLK4β.

Therefore, in order to confirm that endogenous MLK4 also inhibits p38 activity, small interfering RNA (siRNA)-mediated knockdown of MLK4 in HCT116 cells was performed and the levels of active p38 were analyzed. MLK4-knockdown cells had elevated levels of p-p38 as compared with cells treated with a non-specific siRNA oligo, while the total p38 levels remained unchanged (Figure 11, left panel). We also tested whether MLK4 knockdown would affect sorbitol-induced p38 activation. Sorbitol

54 treatment increased the basal level of p-p38 (Figure 11, right panel). A further increase in the level of p-p38 was observed in the MLK4-knockdown cells as compared with cells treated with the nonspecific siRNA oligo (Figure 12, right panel). These results, together with the results in Figure 10, indicate that MLK4β negatively regulates basal, sorbitol- and TNFα-induced activation of p38.

Figure 11. Elevated active p38 in MLK4 knockdown cells. HCT116 cells were transfected with nonspecific or MLK4 siRNA. The cells were left untreated (left panel) or treated (right panel) with 0.5 M sorbitol for 30 min. Cell Lysates were then immunoblotted with MLK4β, p-p38, p38 and β-Actin antibodies.

55 3.2 The effect of MLK4 on MEK3/MEK6 activation

MEK3 and MEK6 are MAP2Ks that specifically phosphorylate and activate p38

(Cuenda, et al., 1997; Raingeaud, et al., 1996). To explore whether MLK4 has any effect

on MEK3 and MEK6 activities, MLK4 was knocked down in HCT116 cells and the

levels of active MEK3/MEK6 phosphorylated at serines 187 and 207 were analyzed

using an activation state specific antibody that detects these phosphorylation sites (p-

MEK3/MEK6). The levels of p-MEK3/MEK6 were increased in cells with MLK4

knockdown as compared with cells treated with the nonspecific siRNA (Figure 12A).

This was consistent with the increase in p-p38 levels that was observed in the MLK4-

knockdown cells (Figure 12A). As expected, the levels of total MEK3/MEK6 and p38

did not change (Figure 12A). Similar results were obtained in ovarian epithelial

carcinoma HEY1B cells (Figure 12B), suggesting that the effect observed on p38 and

MEK3/MEK6 activation by MLK4 is not cell-type specific. These data show that MLK4 inhibits not only active p38 but also the activation of its upstream activator,

MEK3/MEK6.

56

Figure 12. MLK4 negatively regulates MEK3/MEK6 activation. (A) HCT116 cells and (B) HEY1B cells were transfected either with a nonspecific oligo or with MLK4 siRNA. Cell lysates were subsequently prepared and immunoblotted with MLK4β, p-p38, p38, p-MEK3/MEK6, MEK3/MEK6 and β-Actin antibodies.

3.3 The role of MLK4β in NF-κB signaling

In addition to activating p38 and JNK, TNFα also activates the NF-κB pathway

(Karin, et al., 2002). As shown earlier, MLK4β reduced TNFα-induced activation of p38.

This prompted us to investigate if MLK4β also affects TNFα-induced NF-κB activation.

NF-κB is sequestered in the cytoplasm in an inactive state by IκBα (Karin, et al., 2002).

In response to TNFα, IκBα is phosphorylated and targeted for degradation, thus leading

57 to NF-κB activation (Karin, et al., 2002). Degradation of IκBα is considered to be an indicator of NF-κB activity. To test whether MLK4β plays a role in the NF-κB signaling pathway, MLK4β was ectopically expressed in HCT116 cells and TNFα-induced IκBα degradation was detected by analyzing the protein levels of IκBα. TNFα treatment activated the NF-κB pathway as evidenced by the complete abolishment of IκBα (Figure

13A). Untreated cells overexpressing MLK4β had elevated levels of IκBα as compared to cells expressing the empty vector. However, MLK4β had no effect on TNFα-induced degradation of IκBα in cells transfected with the empty vector (Figure 13A).

Furthermore, MLK4 knockdown in HCT116 cells reduced the levels of IκBα as

compared with cells treated with a nonspecific siRNA oligo (Figure 13B). IκBα levels

however, were not affected in MLK4-knockdown cells treated with TNFα.

The effect of MLK4β on the NF-κB pathway was further investigated by using an

NF- κB reporter plasmid that was co-transfected into HEK293 cells with the MLK4β

expression vector or the empty vector. TNFα significantly induced expression of the reporter gene and this induction was not affected by MLK4β expression (Figure 13C).

Consistent with the results in Figure 13A, overexpression of MLK4β reduced the basal expression levels of the NF-κB reporter gene observed in the sample expressing the empty vector alone (Figure 13D). This effect however, was not dose-dependent as both concentrations of the MLK4β expression vector used, induced comparable levels of reduction in the expression of the reporter gene (Figure 13D). Collectively, these results show that MLK4β negatively regulates the basal activation of NF-κB but has no effect on

TNFα-induced activation of NF-κB.

58 A. B.

C.

D. E.

Figure 13. MLK4β negatively regulates basal NF-κB activation but has no effect on

TNFα-induced NF-κB signaling. (A) HCT116 cells were transiently transfected with pCMV5-FLAG or FLAG-MLK4β and left untreated or treated with 10 ng/ml TNFα for

10 min. Cell lysates were blotted and immunoblotted with FLAG, IκBα and β-Actin

59 antibodies. (B) HCT116 cells were transfected with nonspecific or MLK4 siRNA. Cell lysates were then immunoblotted with MLK4β, IκBα and β-Actin antibodies. (C)

HEK293 cells were transfected with the NF-κB reporter plasmid and the β-gal control reporter plasmid together with the empty vector, 100 ng or 200 ng of FLAG-MLK4β.

The cells were then left untreated or treated with 10 ng/ml of TNFα for 6 h and the luciferase assay was performed. (D) Untreated samples from (C) are presented in a separate graph with a different scale. Statistically significant differences from the control are indicated (*P < 0.01). (E) A portion of the cells from (C) was lysed and immunoblotted with FLAG and β-Actin antibodies.

3.4 Comparison of the effects of MLK4β and MLK3 on p38 activation

MLK3 activates MEK3 and MEK6, which in turn phosphorylate and activate p38

(Alonso, et al., 2004; Gallo and Johnson, 2002). However, our data thus far indicate that

MLK4β inhibits, rather than activates p38. To confirm that MLK4β, in contrast to MLK3 and other MAP3Ks, is a negative regulator of p38, we compared the effects of MLK3 and

MLK4β expression on p38 signaling. MLK3 or MLK4β were expressed in HCT116 cells and the levels of p-p38 were analyzed by immunoblotting (Figure 14A). Overexpression of MLK3 elevated the basal level of p-p38, whereas MLK4β overexpression reduced p- p38 levels, as compared with cells expressing the empty vector alone (Figure 14A). The total p38 levels were unchanged (Figure 14A). Consistent with these results, while knocking down MLK3 in HCT116 cells had no effect on the basal levels of p-p38,

60 MLK4-knockdown cells displayed higher levels of active p38 as compared with cells

transfected with nonspecific RNA oligo, while the total p38 levels were not affected

(Figure 14B). These data suggest that unlike MLK3, MLK4β is a potent inhibitor of p38

signaling.

Figure 14. MLKβ, unlike MLK3, inhibits activation of p38 (A) HCT116 were

transiently expressed with pCMV5-FLAG, FLAG-MLK3 or FLAG-MLK4β. Cell lysates were prepared and immunoblotted with FLAG, p-p38 and p38 antibodies. (B) HCT116 cells were transfected with nonspecific, MLK3 and MLK4 siRNA oligos. Cell lysates were immunoblotted with MLK4β, MLK3, p-p38, p38 and β-Actin antibodies.

61 3.5 The effects of MLK3 and MLK4 on ERK and JNK activation

MLK3 can mediate multiple cellular processes such as cell proliferation, invasion

and apoptosis, not only by activating p38, but also through the activation of JNK or ERK

(Chadee and Kyriakis, 2004; Chen, et al., 2010). The effect of MLK3 on JNK and ERK

activity was also assessed in this study. MLK3 was ectopically expressed in SKOV3 and

HEY1B and the levels of active ERK and JNK were analyzed using activation state

phospho-specific ERK (Thr202/Tyr204; p-ERK) and JNK (Thr283/Tyr185; p-JNK)

antibodies respectively. As shown in Figure 15, while the total levels of ERK and JNK

remained unchanged upon MLK3 overexpression, the levels of both p-ERK and p-JNK were elevated in both cells lines as compared with cells transfected with the empty vector

(Figure 15). These results support the previous findings of a role for MLK3 as an upstream activator of ERK and JNK.

Figure 15. MLK3 promotes the activation of ERK and JNK in SKOV3 and HEY1B

cells. SKOV3 and HEY1B cells were transfected with pCMV5-FLAG or FLAG-MLK3.

Cell extracts were prepared and immunoblotted with FLAG, p-ERK, p-JNK, ERK and

JNK antibodies.

62 Our data so far showed that MLK4β, unlike MLK3, inhibits p38. However, it was not known whether this inhibitory effect was exclusive for p38 or if it extended to other

MAPKs. Therefore, we set out to test the effect of MLK4 on ERK and JNK. MLK3- or

MLK4-knockdown cells were treated with sorbitol and the levels of the p-JNK and p-

ERK were analyzed. MLK3 knockdown did not affect the sorbitol-induced activity of

any of the three MAPKs tested, whereas MLK4 knockdown increased the levels of not

only p-p38 but also p-JNK and p-ERK, as compared with cells treated with the

nonspecific siRNA (Figure 16A). This suggests that MLK4 also acts as a negative

regulator of ERK and JNK activation.

Knockdown experiments with siRNA can often yield nonspecific effects. To verify

that our data were a result of MLK4 deficiency and not a nonspecific effect of the siRNA

oligo, we repeated the experiments with another siRNA (siRNA2) that targets a sequence

300 bp upstream of the complementary sequence for MLK4 siRNA (1). The second

siRNA oligo was used to knock down MLK4 in HCT116 cells, and the levels of p-p38, p-

ERK and pJNK were analyzed. MLK4-knockdown cells had higher levels of the active

forms of all p38, JNK and ERK as compared with cells transfected with the nonspecific

oligo (Figure 16B). The total levels of p38, ERK and JNK were not changed with the

siRNA treatment (Figure 16B). These results further support a role for MLK4 as a

negative regulator of MAPK signaling.

63

A. B.

Figure 16. MLK4 negatively regulates ERK and JNK. (A) HCT116 cells were transfected with nonspecific, MLK3 or MLK4 siRNA, and then treated with 0.5 M

sorbitol for 30 min. Cell lysates were then immunoblotted with MLK4β, MLK3, p-p38, p38, p-ERK, ERK, p-JNK, JNK and β-Actin antibodies. (B) HCT116 cells were either transfected with nonspecific siRNA or MLK4 siRNA2. Cell lysates were then prepared and immunoblotted as described in (A).

64 3.6 MLK4β is not an upstream activator of MEK4

MAP3Ks, when activated by overexpression or other stimuli, trigger different

cellular responses by directly phosphorylating and activating MAP2Ks (Johnson and

Lapadat, 2002). Therefore, one way by which the kinase activity of MAP3Ks can be studied is by testing their ability to activate MAP2Ks. MLK4β shares more than 70% homology with MLK3 in the kinase domain and, similar to other MAP3Ks, retains the essential conserved sequences of serine/threonine kinases (Kashuba, et al. 2011).

However, our data thus far revealed a negative regulatory role for MLK4β on MAPKs and MEK3/MEK6 and prompted us to investigate whether MLK4β harbors any kinase activity towards MAP2Ks. MEK4, also known as SEK1, is a MAP2K that is directly phosphorylated and activated by multiple MAP3Ks, including ASK1 and MLK3 (Ichijo, et al., 1997; Tibbles, et al., 1996). Therefore, to study the kinase activity of MLK4β, an in vitro kinase assay was carried out using purified GST-SEK1-K129R, or GST-SEK1-KR, as a substrate, and ASK1 or MLK3 as a positive control. Either HA-ASK1 or HA-

MLK4β were ectopically expressed in HCT116 cells, and a kinase assay was performed

(Figure 17A). Cells expressing the empty vector alone were used as a negative control.

Since phosphorylation of Thr261 is required for SEK1 activity, kinase activity was analyzed by immunoblotting using an activation state phospho-specific SEK1 antibody

(Thr261; p-SEK1). Phosphorylated GST-SEK1-KR was only detected in cells expressing

ASK1 but not in cells expressing MLK4β or the negative control. GST-SEK1-KR total protein levels were similar in all samples (Figure 17A). To further investigate the kinase activity of MLK4β, another kinase assay was performed with lysates prepared from

HCT116 cells overexpressing either FLAG-MLK3 or FLAG-MLK4β (Figure 17B). As

65 expected, only with MLK3, phosphorylation of GST-SEK1-KR was observed, and no

phosphorylation of SEK1 was detected in samples with MLK4β or the empty vector

(Figure 17B). These results demonstrate that, unlike ASK1 and MLK3, MLK4β fails to

phosphorylate SEK1 on Thr261, suggesting that MLK4β does not activate SEK1 through

direct phosphorylation of these activation sites.

A. B.

Figure 17. MLK4β does not phosphorylate Thr261 of GST-SEK1-KR. (A) Lysates

from HCT116 cells transiently expressing pCMV5-FLAG, HA-ASK1 or HA-MLK4β

were prepared and immunoprecipitations of HA-ASK1 and HA-MLK4β were performed

with the HA antibody. A kinase assay was then performed using purified GST-SEK1-KR

as a substrate. The protein samples were then separated by SDS-PAGE and immunoblotted with p-SEK1 (Thr261) and GST antibodies. (B) HCT116 cells were transiently transfected with pCMV5-FLAG, FLAG-MLK3 and FLAG-MLK4β and the cell lysates were prepared and FLAG-MLK3 and FLAG-MLK4β were immunoprecipitated with anti-FLAG. The kinase assay and the detection of activity were performed as described in (A). Overexpression of MLK4β and MLK3 was analyzed by immunoblotting the cell lysates with HA antibody in (A) and FLAG antibody in (B).

66 3.7 MLK4β kinase activity

As observed in Figure 17, MLK4β was not able to phosphorylate GST-SEK1-KR

on Thr261, which is required for SEK1 activity. Although these results show that MLK4β

does not activate SEK1, they do not give a clear indication about the kinase activity of

MLK4β. Thus MLK4β kinase activity was further investigated by using a radioactive kinase assay. FLAG-MLK4β and/or FLAG-MLK3 were ectopically expressed in

HEK293 cells and the lysates were subjected to immunoprecipitation. Lysates prepared from cells expressing the empty vector were used as a negative control. The kinase assay was then performed using γ32P [ATP] and GST-SEK1-KR or histone H2A as substrates.

MLK4β and MLK3 autophosphorylation, and phosphorylation of the GST-SEK1-KR

substrate were then analyzed. Consistent with the results obtained in Figure 17, 32P-GST-

SEK1-KR was only detected in the sample expressing FLAG-MLK3 (Figure 18). These

results indicate that MLK4β fails to phosphorylate GST-SEK1-KR (Figure 18).

Interestingly, 32P-MLK4β was detected in samples overexpressing MLK4β which may

indicate autophosphorylation of MLK4β, but could also be due to another phospho-

protein that was present in the immunoprecipitation (Figure 18). Autophosphorylation of

MLK3 was also observed (Figure 18). Furthermore, immune complexes prepared with the

FLAG antibody from cells overexpressing FLAG-MLK4β were also able to phosphorylate H2A (Figure 18). Although this observation could indicate that MLK4β has kinase activity towards H2A, it is possible other kinases that may have precipitated with MLK4β could have also phosphorylated H2A in these samples. Expression of both

MLK3 and MLK4β in these samples was analyzed by immunoblotting with anti-FLAG antibody (Figure 18). The sample where MLK3 and MLK4β were co-expressed showed

67 reduced expression of both kinases as compared to the other samples (Figure 18). Hence,

a conclusion regarding the effect of MLK4β on MLK3 kinase activity cannot be drawn from this experiment. These findings demonstrate that MLK4β could undergo

autophosphorylation. In addition, the data reveal that MLK4β, unlike other MAP3Ks,

does not phosphorylate and/or activate SEK1, but may harbor kinase activity towards

other substrates.

Figure 18. MLK4β: autophosphorylation and substrate specificity. HEK293 cells

were ectopically expressed with the empty vector, FLAG-MLK3 and/or FLAG-MLK4β.

A kinase assay was performed with γ32P [ATP], with GST-SEK1-KR and histone H2A being added as substrates in the first four lanes and the last lane, respectively.

Autophosphorylation and substrate phosphorylation were detected using autoradiography.

Samples were then immunoblotted with anti-FLAG and anti-GST antibodies to check for overexpression and the presence of the GST-SEK1-KR substrate.

68 3.8 The effect of MLK4β on MLK3 activation

MLK3 is an activator of p38, JNK and ERK signaling pathways (Brancho, et al.,

2005; Chadee and Kyriakis, 2004; Gallo and Johnson, 2002; Rana, et al., 1996). Our results showed that MLK4β negatively regulates MAPK signaling. Thus, we postulated that MLK4β might negatively regulate MAPK signaling by inhibiting MLK3 activation.

To test this possibility, MLK4β and MLK3 were ectopically expressed in HCT116 cells either each alone or together and the level of active, phosphorylated MLK3 (p-MLK3) was analyzed. Interestingly, ectopically expressed MLK3 was active and MLK4β overexpression with MLK3 completely abolished the levels of p-MLK3, suggesting that

MLK4β inhibits MLK3 activation (Figure 19).

Figure 19. MLK4β inhibits induced MLK3 activation. pCMV5-FLAG, FLAG-MLK3 or FLAG-MLK4β were ectopically expressed in HCT116 cells. Cell lysates were prepared and immunoblotted with p-MLK3 and FLAG antibodies.

69 The effect of MLK4β on MLK3 activation was also analyzed in another cell line,

SKOV3 cells. SKOV3 cells have a detectable level of active MLK3 (Figure 20A).

Ectopic expression of MLK4β, however, substantially reduced the levels of p-MLK3 that were observed in cells expressing the empty vector alone, without affecting total MLK3 levels (Figure 20A). This negative effect of MLK4β on MLK3 activation was also confirmed in SKOV3 MLK4-knockdown cells (Figure 20B). Silencing MLK4 elevated the levels of p-MLK3 as compared with cells treated with the nonspecific siRNA, whereas the levels of total MLK3 remained unchanged (Figure 20B). Moreover, consistent with the results obtained for HCT116 cells, MLK4β expression also reduced p-

p38 levels in SKOV3 cells (Figure 20A), whereas MLK4 knockdown elevated the basal

levels of active p38 (Figure 20B). The effect of MLK4 on MLK3 and p38 activations was

also studied using MLK4 siRNA2. Knocking down MLK4 in HCT116 cells using the

second oligo also led to elevated levels of both p-MLK3 and p-p38 as compared with

cells treated with the nonspecific oligo without affecting the total levels of either protein

(Figure 20C). The data from both the HCT116 and SKOV3 cell lines reveal an unknown

inhibitory role for MLK4β in MLK3 activation.

70

Figure 20. MLK4 inhibits the basal activation of MLK3 in SKOV3 cells. (A) SKOV3 cells transiently transfected with pCMV5-FLAG or FLAG-MLK4β. Cell lysates were

then prepared and immunoblotted with MLK4β, p-MLK3, MLK3, p-p38 and p38

antibodies. (B) SKOV3 cells were transfected with nonspecific or MLK4 siRNA and cell

lysates were immunoblotted with MLK4β, p-MLK3, MLK3, p-p38, p38 and β-Actin

antibodies. (C) SKOV3 cells were transfected with nonspecific oligo or with MLK4 siRNA2. Cell lysates were immunoblotted as described in (B).

71 3.9 The correlation between MLK4β expression and active MLK3 in different cell lines

MLK3 expression and activity are elevated in breast and ovarian cancer cells

(Chen, et al., 2010; Zhan, et al., 2012). MLK4β, as our data suggest, inhibits MLK3 activity. Thus, we were interested in testing if there is a correlation between the expression level of MLK4β and MLK3 activity. We analyzed the expression levels of

MLK4β and MLK3, as well as the levels of p-MLK3 in the immortalized epithelial ovarian cell line, T80; the colon cancer cell lines, HCT116 and HT-29; the ovarian cancer cell lines, HEY1B and SKOV3; and the glioblastoma cell line, T98G. HCT116, HT-29 and SKOV3 cells displayed higher expression levels of MLK4β as compared to HEY1B cells which had very low level of expression (Figure 21). T80 and T98G cells expressed barely detectable levels of MLK4β (Figure 21). No correlation between the expression levels of MLK4β and MLK3 was observed (Figure 21). Interestingly, low MLK4β expression was associated with an increased level of p-MLK3 in T80, HEY1B and T98G cells; and high MLK4β expression was associated with low MLK3 activity in HCT116,

SKOV3 and HT-29 cells (Figure 21). The levels of p-ERK were also analyzed in these cell lines. Higher levels of p-ERK were detected in the cell lines expressing a high level of p-MLK3 and a low level of MLK4β (Figure 21). However, high p-ERK was observed in HEY1B cells as compared to T98G cells, although the latter express higher levels of active MLK3 (Figure 21). This could be attributed to other upstream activators of ERK that could be expressed more in HEY1B. Among the cancer cell lines, T98G cells that expressed the lowest levels of MLK4β harbored the highest levels of active MLK3

(Figure 21). Overall these results suggest that there may be an inverse correlation between the levels of active MLK3 and MLK4β expression in the cell lines tested.

72

S.E 0.2 L.E 0.15 β 0.1

of MLK4 0.05

0 Relative protein level level protein Relative

1.5 1 0.8 1 ERK - 0.6 0.4 0.5

0.2 MLK3 - level of p of level p

Relative protein protein Relative 0 0 Relative protein level of of level protein Relative

0.8 0.5 0.4 0.6 0.3 0.4 0.2

MLK3 0.2

of ERK 0.1 0 0 Relative protein level of of level protein Relative Relative protein level level protein Relative

73 Figure 21. Correlation between MLK4β expression and active MLK3. Protein

extracts were prepared from T80, HCT116, HT-29, HEY1B, SKOV3 and T98G cells.

Cell lysates were then immunoblotted with MLK4β, p-MLK3, MLK3, p-ERK, ERK and

β-Actin antibodies (upper, left panel). Quantification of MLK4β (upper, right panel), p-

MLK3, MLK3, p-ERK and ERK levels (lower panels) was determined using Image J software (NIH, Bethesda, Maryland, USA). MLK4β, p-MLK3, MLK3, p-ERK or ERK levels were normalized to the levels of β-Actin in each sample and the values presented represent the relative active levels (p-ERK and p-MLK3) or the relative protein expression levels (MLK4β, MLK3 and ERK) between the different cell lines tested. S.E stands for short exposure and L.E for long exposure. Quantification of p-MLK3 levels was done using the L.E p-MLK3 blot.

74 3.10 MLK4β associates with MLK3

MLK4β and MLK3 share > 65% sequence homology in all of the structural domains, including the leucine zipper domain, while the C-terminal regions are more divergent (Kashuba et al. 2011). Thus, we postulated that these two proteins may be associated in cells. To test if MLK4β interacts with MLK3, (GST)-tagged MLK3 and

HA-tagged MLK4β were expressed in HCT116 cells and co-immunoprecipitation experiments were performed. In cells expressing both GST-MLK3 and HA-MLK4β, co- immunoprecipitation of GST-MLK3 and HA-MLK4β was observed (Figure 22A).

Moreover, MLK3 was not detected in lysates from untransfected cells, cells expressing the empty vector alone, or cells expressing only GST-MLK3 or HA-MLK4β (Figure

22A). The interaction between MLK3 and MLK4β was also observed in reciprocal experiments where GST-MLK3 was pulled down and immunoblotting was performed to detect HA-MLK4β (Figure 22B). Again, MLK4β was not detected in untransfected or empty vector controls or cells expressing GST-MLK3 or HA-MLK4β (Figure 22B).

These results suggest that GST-MLK3 and HA-MLK4β are associated in cells.

The association observed between overexpressed MLK3 and overexpressed

MLK4β does not necessarily mean that the endogenous proteins interact. To address this possibility, endogenous MLK4β was immunoprecipitated from HCT116 cells and the immunoprecipitates were analyzed by immunoblotting (Figure 22C). Indeed, co- immunoprecipitation of endogenous MLK3 and MLK4β was detected (Figure 22C).

Collectively, these results demonstrate that endogenous MLK4β and MLK3 are associated in cells.

75 A. B.

C.

Figure 22. MLK4β associates with MLK3. (A) HCT116 cells were transfected with

pCMV5 vector, GST-MLK3, HA-MLK4β or both GST-MLK3 and HA-MLK4β. HA-

MLK4β was immunoprecipitated from cell lysates with HA antibody. Co-

immunoprecipitation of GST-MLK3 was assessed by immunoblotting with MLK3 antibody. HA-MLK4β was immunoprecipitated and immunoblotted with HA antibody to verify MLK4β expression. (B) GST pull-downs were performed with lysates from the cells as described in A. The presence of HA-MLK4β and GST-MLK3 in the GST pull- downs was determined by immunoblotting with HA and GST antibodies, respectively.

(C) Endogenous MLK4β was immunoprecipitated from HCT116 cell lysates and associated endogenous MLK3 was detected by immunoblotting with MLK3 antibody.

The level of endogenous MLK4β in the immunoprecipitation was assessed by immunoblotting with MLK4 antibody. Control immunoprecipitations were performed with no antibody or with rabbit IgG.

76 3.11 The effect of MLK4 on cell proliferation

MAP3Ks, including MLK3, through activation of MAPKs, have been reported to play a role in proliferation (Brancho, et al., 2005; Chadee and Kyriakis, 2004; Dhillon, et al., 2007). Based on the inhibitory role of MLK4β on MLK3 activity and MAPK signaling, we postulated that MLK4β would have a negative effect on proliferation. As shown in Figure 21, HCT116 cells expressed the highest levels of MLK4β among the cell lines tested. Therefore, to test if MLK4 has any effect on cell proliferation, an MTT assay was performed with MLK4 knockdown HCT116 cells. MLK4 knockdown HCT116 cells, similar to cells transfected with the nonspecific oligo, had approximately a 7 fold increase in cell number 3 days post-seeding. After 4 days, the MLK4 knockdown cells exhibited approximately a 14 fold increase, which was comparable to that of cells treated with the nonspecific oligo (Figure 23, upper panel). The extent of MLK4 knockdown was detected by immunoblotting (Figure 23, lower panel). These results show that MLK4 does not affect HCT116 cell proliferation. Stable expression of MLK4β in cells, that normally have low MLKβ and a high level of active MLK3, such as T98G cells, would be a better approach to elucidate any role for MLK4β in cell proliferation.

77

18

16

14

12

10 Nonspecific oligo 8 MLK4 siRNA (2) 6

4

2 Fold increase in cell number cell in increase Fold 0 Day 0 Day 3 Day 4

Figure 23. MLK4 has no effect on HCT116 cell proliferation. HCT116 cells were transfected with either the MLK4 siRNA or the nonspecific oligo. Transfected cells were then seeded in 24-well plates and cell proliferation was assessed using the MTT assay at days 0, 3 and 4 (upper panel). MLK4 knockdown was detected by immunoblotting using

MLK4β and β-Actin antibodies (lower panel). Statistical analysis was performed using the Student t-test.

78 3.12 MLK3 is required for cell invasion in ovarian cancer cells

MLK3, in addition to cell proliferation, is involved in numerous other biological functions in which a role for MLK4β could be implicated. Cell invasion represents one of these functions. For instance, MLK3 was shown to be required for cell invasion in multiple cancer systems, such as breast, gastric and lung cancer cell lines (Chen, et al.,

2010; Mishra, et al., 2010; Swenson-Fields, et al., 2008). The invasive ovarian cancer cell lines, SKOV3 and HEY1B, express basal levels of active MLK3 which suggests that

MLK3 might play a crucial role in promoting invasion in these cells. To test this possibility, MLK3 was knocked down in both SKOV3 (Figure 24A) and HEY1B (Figure

24B) cells, and the capacity to invade through the Matrigel in vitro was analyzed. In

MLK3-knockdown SKOV3 cells, invasion was reduced by approximately 66% with siRNA oligo 1 and 82% with siRNA oligo 2 as compared with cells transfected with the nonspecific oligo (Figure 24A). Similarly, HEY1B cells with MLK3 knockdown exhibited approximately 70% (siRNA oligo 1) and 80% (siRNA oligo 2) reduction in cell invasion as compared with cells treated with the nonspecific oligo (Figure 24B). These results clearly demonstrate a specific requirement for MLK3 in the invasion of the ovarian cancer cell lines, SKOV3 and HEY1B.

79

Figure 24. MLK3 is essential for SKOV3 and HEY1B cell invasion. SKOV3 (A) and

HEY1B (B) cells were transfected with nonspecific oligo or MLK3 siRNA oligo 1. Cell invasion was analyzed using Transwell chambers containing Matrigel. Cells that crossed the membrane were stained and counted (left panel). The experiment was repeated three times. Each bar represents the mean of data obtained from three independent experiments and the error bars represent deviation from the mean. Statistically significant differences from the control are indicated (*P < 0.01). A portion of the cells were lysed and immunoblotted with MLK3 and β-Actin antibodies (right panel). (The invasion assay for

SKOV3 cells using the siRNA oligo 1 in panel A was performed by Nidhi Modi).

80 3.13 MLK4β inhibits SKOV3 cell invasion

Given the role of MLK3 in cell invasion and our results that MLK4β suppresses

MLK3 activity, we were interested in testing the effect of MLK4β on cell invasion. For this purpose, the invasive SKOV3 cells were transiently transfected with MLK4β and the effect on cell invasion was tested in vitro by analyzing the capacity of cells to invade through the Matrigel (Figure 24). Cells overexpressing MLK4β exhibited a significant reduction in cell invasion as compared with cells expressing the empty vector alone

(Figure 25, left panel). MLK4β overexpression was detected by immunoblotting (Figure

25, right panel). These results show that MLK4β reduces SKOV3 cell invasion.

81 Figure 25. MLK4β reduces the invasion of SKOV3 cells. SKOV3 cells were

transfected with pCMV5-FLAG or FLAG-MLK4β. Cell invasion was analyzed using

Transwell chambers containing Matrigel. Cells that crossed the membrane were stained and counted (left panel). The experiment was repeated three times. Each bar represents the mean of data obtained from three independent experiments and the error bars represent standard deviation from the mean. Statistically significant differences from the control are indicated (*P < 0.01). A portion of the cells was lysed and immunoblotted with FLAG and β-Actin antibodies (right panel).

3.14 MLK3 regulates MMP-2 and MMP-9 enzyme activity

MMPs are proteolytic enzymes that promote tumor cell invasion by breaking

down components of the extracellular matrix (Nagase and Woessner, 1999; Schmalfeldt,

et al., 2001). Such processes are associated with an increase in the expression and activity

of MMP-2 and MMP-9 (Schmalfeldt, et al., 2001). To test whether MLK3 regulates

MMP activity which may in turn impact cell invasion, MLK3 was knocked down in

SKOV3 and HEY1B cells and the effect on MMP-2 and MMP-9 enzyme activities was

analyzed using gelatin zymography (Figure 26A). Culture media collected from SKOV3

and HEY1B siRNA (1 or 2) knockdown cells exhibited a substantial reduction in the

enzyme activities of both MMP-2 and MMP-9 as compared to the MMP activities in the culture media of cells transfected with the nonspecific siRNA oligo (Figure 26A). These results indicate that MLK3 is required for MMP-2 and -9 enzyme activities in SKOV3

82 and HEY1B ovarian cancer cells. Possibly, the regulation of MMP-2 and MMP-9

activities is a potential mechanism by which MLK3 promotes invasion in these ovarian

cell lines.

As shown in Figure 15, MLK3 activates JNK and ERK in SKOV3 and HEY1B

cells. Moreover, both ERK and JNK have previously been shown to regulate the

activation of MMP-2 and MMP-9 (Kim, et al., 2010). Thus, we proposed that ERK and

JNK could mediate MLK3-dependent regulation of MMPs. To test this, SKOV3 and

HEY1B cells were ectopically transfected with FLAG-MLK3 and the cells then treated with pharmacological inhibitors of ERK (U0126) or JNK (SP600125), and MMP-2 and

MMP-9 activation was assessed using the gelatinase assay. ERK and JNK activities were

inhibited by U0126 and SP600125, respectively (Figure 26B). MLK3 overexpression

resulted in an increase in MMP-2 and MMP-9 gelatinase activities in SKOV3 and

HEY1B cells (Figure 26C). Inhibition of either ERK or JNK blocked MLK3-induced activation of MMP-2 and MMP-9 (Figure 26C). These results suggest a role for both

ERK and JNK in mediating MLK3-dependent activation of MMP-2 and MMP-9.

83

A.

C. B.

Figure 26. MLK3 mediates MMP-2 and MMP-9 activation in SKOV3 and HEY1B

cells by a mechanism that involves ERK and JNK. (A) SKOV3 and HEY1B cells were

transfected with nonspecific (NS) siRNA or MLK3 siRNA oligo 1 or 2. Cell culture

media was collected and MMP-2 and -9 enzyme activities were studied by subjecting the collected media to gelatin zymography analysis (upper panel). Cell extracts were immunoblotted with MLK3 and β-Actin antibodies (lower panel) to check for MLK3

knockdown. (The gelatinase assay for SKOV3 cells using MLK3 siRNA (1) was

performed by Yu Zhan). (B) SKOV3 cells were either treated with U0126 or SP600125

for 4 hours and cell extracts were immunoblotted with p-ERK, ERK, p-JNK, JNK

antibodies. (C) SKOV3 and HEY1B cells were transfected with pCMV5 vector or

FLAG-MLK3. Cells expressing FLAG-MLK3 were treated with U0126 or SP600125.

Cell culture medium was subjected to gelatin zymography analysis (upper two panels).

Cell extracts were immunoblotted with anti-FLAG antibody (lower panel).

84 3.15 MLK4β reduces MMP-9 activity in SKOV3 cells

Our results indicate that MLK3 is required for MMP-2 and MMP-9 activities in ovarian cancer cells. Since we also observed that MLK4β inhibits MLK3 activity, we postulated that MLK4β might negatively regulate the enzyme activities of MMP-2 and

MMP-9. Therefore, to investigate if MLK4β affects MMP-2 and MMP-9 enzyme activities, the levels of active MMP-2 and MMP-9 were analyzed in SKOV3 cells ectopically expressing MLK4β by gel zymography (Figure 27). Although overexpression of MLK4β had little effect on MMP-2 gelatinase activity, it decreased the gelatinase activity of MMP-9 (Figure 27, upper panel). These results suggest that MLK4β may regulate MMP-9 activity.

Gelatinase Assay

FLAG Immunoblot Actin

Figure 27. MLK4β reduces MMP-9 enzyme activity. SKOV3 cells were transfected with pCMV5-FLAG or FLAG-MLK4β. Culture media was then collected and subjected to gelatin zymography analysis (upper panel). Cell extracts were immunoblotted with

FLAG and β-Actin antibodies (lower panel) (duplicated from Figure 25).

85 Chapter 4

Discussion

MAPKs play important roles in a plethora of cellular responses including proliferation, transformation, invasion, differentiation and apoptosis (Dhanasekaran and

Johnson, 2007; Dhillon, et al., 2007; Raman, et al., 2007). The specific response(s) triggered by the MAPKs depends on the stimulus, cell type and degree of MAPK activation (Dhanasekaran and Johnson, 2007; Dhillon, et al., 2007). Controlled regulation of MAPK signaling is crucial to maintaining normal cellular function. Since deregulated

MAPK signaling can lead to cancer-related processes and cell death, MAP3Ks, as main upstream activators of MAPKs, are critical players that modulate MAPK pathways to generate specific biological responses. Enhanced basal activities of MAP3Ks can promote constitutive activation of one or multiple MAPKs, which can in turn lead to different pathologies. Indeed, mutations in B-Raf, which result in constitutive activation of the ERK pathway, occur at high frequencies in multiple cancers (Beeram, et al., 2005;

Davies, et al., 2002). Thus, elucidating the functions of MAP3Ks is important to better understand the regulation of MAPK signaling and to present MAP3Ks as potential therapeutic targets for treating cancer and other pathologies.

MLK3 activates JNK, ERK and p38 and is involved in numerous cellular responses including apoptosis, transformation, proliferation and invasion (Chadee and

86 Kyriakis, 2004; Chen, et al., 2010; Gallo and Johnson, 2002; Hartkamp, et al., 1999).

However, functional characterization of the MLK subfamily members other than MLK3 has so far been limited. Unraveling the cellular function(s) of MLKs 1, 2 and 4 is necessary to identify redundancies and uncover unique functions of the MLK enzymes.

Activating mutations in MLK4 have been reported in colon cancer (Bardelli, et al., 2003).

These mutations might implicate a role for MLK4 in cancer. Therefore, in this study, we investigated the role of MLK4β in MAPK signaling. Given that MLK4β and MLK3 share

75% homology in their catalytic kinase domains, it might be assumed that MLK4β activates MAPK signaling (Kashuba et al. 2011). However, our findings indicate that

MLK4β is a negative regulator of MAPK signaling, and MLK4β inhibits both basal and sorbitol- and TNFα-induced, p38 activity (Figures 10 and 11). Consistent with this, a recent study reported that MLK4β overexpression failed to activate co-expressed p38 and that MLK4β overexpression reduced the levels of LPS-induced p38 activation 15 min post treatment (Seit-Nebi, et al., 2012). These observations support our findings that

MLK4β functions to inhibit, rather than to activate, p38 activity.

Our results demonstrate an elevation in the basal and sorbitol-induced levels of active p38, JNK and ERK when mlk4 is silenced, suggesting that the inhibitory role of

MLK4 is not specific to p38, but extends to JNK and ERK as well (Figure 16). These results are in accordance with similar findings by Seit-Nebi et al. that showed that MLK4 inhibits LPS-induced activation of the JNK and ERK pathways (Seit-Nebi, et al., 2012).

Collectively, our results demonstrate that MLK4β, unlike MLK3 and other MAP3Ks, acts as a negative regulator of MAPK signaling (Figure 28).

87 MLK4β shares >70% sequence homology with MLK3 in the catalytic domain

(Kashuba et al. 2011). Moreover, MLK4β retains the conserved serine/threonine

sequences present in MLK3 and other serine/threonine kinases. For instance, the kinase

domain of MLK4β contains the conserved KAAR sequence (Lys 151- Arg 154) that is

required for ATP binding and the conserved TTXXS motif, found within the DFG-APE

sequences, that is phosphorylated in many kinases, at the second threonine and serine

residues (residues Thr300 and Ser303 in MLK4β), and is required for activation of the

kinase. Since MLK3 and other MAP3Ks were shown to directly phosphorylate and

activate MAP2Ks, it might be assumed that MLK4β would also phosphorylate and

activate MAP2Ks. However, our in vitro kinase assay data show that unlike MLK3 and

ASK1, MLK4β fails to phosphorylate GST-SEK1-KR at Thr261, a phosphorylation event that is required for SEK1 activation (Figure 17). MLK4β also failed to phosphorylate other residues of GST-SEK1-KR (Figure 18). These results demonstrate that MLK4β does not phosphorylate and/or activate SEK1. Furthermore, MLK4β undergoes autophosphorylation and phosphorylates histone H2A (Figure 18). These results demonstrate that although MLK4β, similar to MLK3, can autophosphorylate,

MLK4β and MLK3 exhibit different substrate specificities towards SEK1, which may suggest why these two enzymes have such different biological effects.

MLK3 signaling promotes the activation of multiple MAPK pathways (Chadee and Kyriakis, 2004; Gallo and Johnson, 2002). Thus, we proposed that MLK4β might exert its inhibitory effect on MAPK signaling by inhibiting MLK3 activity. Indeed, we found that MLK4β expression reduces the level of active MLK3 in HCT116 and SKOV3 cells, while silencing mlk4 elevates the level of p-MLK3 (Figures 19 and 20).

88 Interestingly, comparison of tumor cell lines and normal cell lines revealed an inverse

correlation between the level of MLK4β and the level of active MLK3 (Figure 21).

Moreover, low MLK4β protein levels are associated with an elevation in the levels of active ERK (Figure 21). Our results, however, revealed that the level of active ERK do not always correlate with the levels of p-MLK3 (Figure 21). For instance, although T98G cells have a higher level of active MLK3 than HEY1B cells, higher levels of p-ERK are present in HEY1B cells (Figure 21). MAPKs can be activated by numerous upstream activators (Raman, et al., 2007). Therefore, it is possible that HEY1B cells express high levels of other upstream activators of ERK, in addition to MLK3. Collectively, these

results show that MLK4β inhibits MAPK signaling and MLK3 activition (Figure 28).

Co-immunoprecipitation experiments revealed an association between overexpressed MLK4β and MLK3, and endogenous MLK4β and MLK3 in HCT116 cells

(Figure 22), which to our knowledge is the first demonstration of an association between two MLK subfamily members. However, the nature of the interaction between MLK4β and MLK3, whether direct or indirect, and the domains involved remains to be determined. Because the leucine zipper domains of MLK4β and MLK3 do share 66% identity (Kashuba et al. 2011), heterodimerization between these two proteins remains a possibility.

MLK3 activity is autoinhibited by the N-terminal SH3 domain and autoinhibition is relieved upon binding of active Cdc42 to the MLK3 CRIB domain (Du, et al., 2005;

Gallo and Johnson, 2002). An association between MLK4β and Cdc42 was not observed

(Seit-Nebi, et al., 2012). Therefore, it is still possible that the association between

89 MLK4β and MLK3 blocks MLK3 activation by interfering with Cdc42 binding to the

MLK3 CRIB domain.

p38 can promote cell survival by activating the NF-κB pathway (Maulik, et al.,

1998; Zechner, et al., 1998). Given that MLK4β overexpression reduced basal p38 activity (Figure 10), it might be assumed that MLK4β would act as a negative regulator of the NF-κB pathway. Indeed, MLK4β overexpression elevated the basal level of IκBα,

while MLK4β knockdown reduced the level of IκBα (Figure 13). Moreover, MLK4β

overexpression significantly reduced NF-κB luciferase activity in HEK293 cells. These results demonstrate a negative regulatory effect for MLK4β on basal NF-κB activity.

Knocking down MLK3 was shown to elevate the levels of active IKK and reduce IκBα protein levels, suggesting an inhibitory function for MLK3 in the NF-κB pathway (Cole, et al., 2009). Considering the inhibitory effect for MLK4β on MLK3 activition, it would be expected that MLK4β would either enhance NF-κB activity or not affect it at all.

MAPKs and NF-κB signaling, however can be triggered by different upstream targets other than MLK3, including TLR4 (Kawai and Akira, 2007; Zbinden-Foncea, et al.,

2012). Interestingly, MLK4 was found to interact with and inhibit TLR4 (Seit-Nebi, et al., 2012). Therefore, it is possible that MLK4β negatively regulates basal NF-κB

pathway by inhibiting TLR4. MLK4β, however, fails to block TNFα-induced NF-κB

activation as evidenced by its inability to prevent IκBα degradation in response to TNFα

treatment (Figure 13). This finding is consistent with the observation that MLK4 does not

affect LPS-induced IκB degradation (Seit-Nebi, et al., 2012). Hence, MLK4β negatively

regulates basal but not TNFα- or LPS-induced NF-κB activation. Possibly, in response to

90 stimuli, such as TNFα and LPS, multiple upstream activators of the NF-κB pathway are activated which can then overcome the inhibitory effect of MLK4β.

HCT116 cells express high levels of MLK4β (Figure 21). Knock down of MLK4 in HCT116 cells, did not affect cell proliferation even four days after siRNA transfection

(Figure 23). Moreover, MLK4-knockdown cells were viable and did not exhibit any overt signs of stress or changes in morphology (data not shown). Thus, MLK4 is not vital for cell survival or proliferation in HCT116 cells. In a recent study, MCF7 breast cancer cells stably expressing MLK4β failed to proliferate and a role for MLK4 in apoptosis was proposed (Kashuba et al. 2011). Nonetheless, in the same study, transgenic mice expressing human pCMV-MLK4β were generated and the mice were normal, healthy and showed no phenotypic abnormalities (Kashuba et al. 2011). A more thorough study is thus needed to clarify the role of MLK4β in cell proliferation.

Ovarian cancer is one of the leading types of cancer among women and is characterized by a poor prognosis and a high mortality rate primarily due to the lack of early detection (Dutta, et al., 2010). Therefore, unraveling the early events involved in ovarian cancer provides insights for new therapeutic targets that can yield novel treatments to counteract this disease. Cell invasion is an essential process in the transition of tumors to a malignant state (Hanahan and Weinberg, 2011). Hence, identifying the key players in ovarian cancer cell invasion is crucial to the development of new therapeutic options. Recent studies have shown a requirement for MLK3 in the invasion of mammary epithelial tumor cells and in the migration of gastric and A459 lung tumor cells (Chen, et al., 2010; Mishra, et al., 2010; Swenson-Fields, et al., 2008). In addition, we have demonstrated that MLK3-knockdown SKOV3 and HEY1B cells exhibit impaired cell

91 invasion (Figure 24). These results suggest that MLK3 has a critical function in ovarian

cancer cell invasion. Considering the inhibitory effect of MLK4β on MLK3, we

postulated that MLK4β might hinder the invasion of ovarian cancer cells. As expected, overexpressing MLK4β reduced SKOV3 cell invasion (Figures 25 and 28), which suggests a novel role for MLK4β in ovarian cancer cell invasion.

MMPs, particularly MMP-2 and MMP-9, play important functions in promoting cell invasion by breaking down the ECM and mediating the EMT process (Hua, et al.,

2011; Roy, et al., 2009; Tester, et al., 2007). We proposed that MLK3 may mediate ovarian cancer cell invasion by regulating MMP-2 and MMP-9 activities. Indeed, using siRNA-mediated silencing of mlk3, we determined a reduction in the levels of active

MMP-2 and MMP-9 in SKOV3 and HEY1B cells lacking MLK3 (Figure 26A). This suggests a requirement for MLK3 in the activation of these enzymes. p38, ERK and JNK can mediate cell invasion by mechanisms that depend on MMP-2 and/or MMP-9

(Spallarossa, et al., 2006; Zhou, et al., 2007). Other than activating transcription factors that induce MMP gene expression, the ERK and JNK pathways have also been shown to directly regulate MMP-2 and MMP-9 activation (Kim, et al., 2010). For instance, TIMP2 which plays a role in MMP-2 activation, is activated by ERK and JNK, and in turn MMP-

2 activates MMP-9 (Butler, et al., 1998; Kim, et al., 2010; Toth, et al., 2003). MLK3 is required for ERK and JNK activation in colon and lung fibroblasts and also in SKOV3 cells (Chadee and Kyriakis, 2004; Zhan, et al., 2012). We postulated that MLK3 may induce MMP activity and ovarian cancer invasion through activation of ERK and JNK.

This was demonstrated in HEY1B and SKOV3 cells whereby ERK or JNK inhibition potently reduced MLK3-dependent MMP-2 and MMP-9 gelatinase activities (Figure 26

92 C). Being a negative regulator of MLK3 and MAPK signaling, we predicted that MLK4β

would inhibit MMP activity. Indeed, we found that MLK4β expression reduces MMP-9

proteolytic activity (Figure 27).

In this study we demonstrated unexpected differences in the signaling between

MLK4β and MLK3. The MLK subfamily members are highly diverse in their C-terminal domains (Gallo and Johnson, 2002). The fact that MLK3 and MLK4β, despite sharing >

65% homology in all domains but the C-terminal region, have opposing functions

suggests that the C-terminus may have a critical role in regulating MLK functions.

Identification of the proteins that interact with the MLK C-termini is an interesting direction for future investigation. Our results indicated that wild-type MLK4β has no kinase activity towards SEK1. Consistent with this, wild type MLK4 exhibited low kinase activity that was comparable to that of the kinase dead variant of MLK4 (Martini et al. 2013). Interestingly, in a recent study, mutations in MLK4, H261Y, G291E, R470C and R555STP were found to substantially increase MLK4 kinase activity (Martini et al.

2013). Moreover, these mutants, unlike wild-type MLK4, significantly enhanced the transformational capacity of cells expressing the G12V Ras mutant, which readily triggers cellular transformation (Martini et al. 2013). Possibly these transformation mutations, allow MLK4β to adopt a tertiary structure that enhances the activation of the enzyme. Considering that MLK4β associates with and inhibits MLK3 and TLR4, MLK4β

may have a scaffold function that mediates inhibition of these proteins. Hence, the

transforming mutations might induce a conformational change of the MLK4 enzyme that

favors a kinase-directed over a scaffold function. Thus, MLK4β harboring one of the

activating mutation might function as a MAP3K that activates MAPK signaling. The

93 colon cancer cell line, LoVo, harbors one of the activating mutations, R470C (Bardelli, et

al., 2003). Hence, for future studies, it will be interesting to study the role of MLK4β in

MAPK signaling in LoVo cells as compared to other cell lines. Moreover, analysis of the

effect of MLK4 activating mutations on MAPK signaling and cell invasion is important

to better understand the biochemical functions of MLK4β.

Figure 28. Schematic diagram illustrating the role of MLK4β in MAPK signaling.

MLK3, upon activation, activates the p38, JNK and ERK pathways leading to different

cellular responses, including invasion. Our results indicate that MLK4β negatively

regulates MLK3 activation and the ERK, JNK and p38 MAPK-signaling pathways.

Therefore, MLK4β-dependent suppression of MAPK signaling may be mediated through inhibition of MLK3 activation.

94

References

A. N. Abell and G. L. Johnson. 2005. MEKK4 is an effector of the embryonic TRAF4 for JNK activation. J Biol Chem, 280(43):35793-6.

A. N. Abell, J. A. Rivera-Perez, B. D. Cuevas, M. T. Uhlik, S. Sather, et al. 2005. Ablation of MEKK4 kinase activity causes neurulation and skeletal patterning defects in the mouse embryo. Mol Cell Biol, 25(20):8948-59.

N. G. Ahn, R. Seger, R. L. Bratlien, C. D. Diltz, N. K. Tonks, et al. 1991. Multiple components in an epidermal growth factor-stimulated protein kinase cascade. In vitro activation of a myelin basic protein/microtubule-associated protein 2 kinase. J Biol Chem, 266(7):4220-7.

A. Alonso, J. Sasin, N. Bottini, I. Friedberg, A. Osterman, et al. 2004. Protein tyrosine phosphatases in the . Cell, 117(6):699-711.

E. Alvarez, I. C. Northwood, F. A. Gonzalez, D. A. Latour, A. Seth, et al. 1991. Pro-Leu- Ser/Thr-Pro is a consensus primary sequence for substrate protein phosphorylation. Characterization of the phosphorylation of c-myc and c-jun proteins by an epidermal growth factor receptor threonine 669 protein kinase. J Biol Chem, 266(23):15277-85.

N. G. Anderson, J. L. Maller, N. K. Tonks and T. W. Sturgill. 1990. Requirement for integration of signals from two distinct phosphorylation pathways for activation of MAP kinase. Nature, 343(6259):651-3.

A. Ashworth, S. Nakielny, P. Cohen and C. Marshall. 1992. The amino acid sequence of a mammalian MAP kinase kinase. Oncogene, 7(12):2555-6.

B. Baliga and S. Kumar. 2003. Apaf-1/cytochrome c apoptosome: an essential initiator of caspase activation or just a sideshow? Cell Death Differ, 10(1):16-8.

A. Bardelli, D. W. Parsons, N. Silliman, J. Ptak, S. Szabo, et al. 2003. Mutational analysis of the tyrosine kinome in colorectal cancers. Science, 300(5621):949.

M. K. Barthwal, P. Sathyanarayana, C. N. Kundu, B. Rana, A. Pradeep, et al. 2003. Negative regulation of mixed lineage kinase 3 by protein kinase B/AKT leads to cell survival. J Biol Chem, 278(6):3897-902.

95 B. Bauvois. 2012. New facets of matrix metalloproteinases MMP-2 and MMP-9 as cell surface transducers: outside-in signaling and relationship to tumor progression. Biochim Biophys Acta, 1825(1):29-36.

M. Beeram, A. Patnaik and E. K. Rowinsky. 2005. Raf: a strategic target for therapeutic development against cancer. J Clin Oncol, 23(27):6771-90.

R. Ben-Levy, S. Hooper, R. Wilson, H. F. Paterson and C. J. Marshall. 1998. Nuclear export of the stress-activated protein kinase p38 mediated by its substrate MAPKAP kinase-2. Curr Biol, 8(19):1049-57.

M. M. Benjamin and R. A. Khalil. 2012. Matrix metalloproteinase inhibitors as investigative tools in the pathogenesis and management of vascular disease. EXS, 103:209-79.

N. Bisson, M. Tremblay, F. Robinson, D. R. Kaplan, S. P. Trusko, et al. 2008. Mice lacking both mixed-lineage kinase genes Mlk1 and Mlk2 retain a wild type phenotype. Cell Cycle, 7(7):909-16.

J. L. Blank, P. Gerwins, E. M. Elliott, S. Sather and G. L. Johnson. 1996. Molecular cloning of mitogen-activated protein/ERK kinase kinases (MEKK) 2 and 3. Regulation of sequential phosphorylation pathways involving mitogen-activated protein kinase and c- Jun kinase. J Biol Chem, 271(10):5361-8.

B. C. Bock, P. O. Vacratsis, E. Qamirani and K. A. Gallo. 2000. Cdc42-induced activation of the mixed-lineage kinase SPRK in vivo. Requirement of the Cdc42/Rac interactive binding motif and changes in phosphorylation. J Biol Chem, 275(19):14231- 41.

T. G. Boulton, S. H. Nye, D. J. Robbins, N. Y. Ip, E. Radziejewska, et al. 1991. ERKs: a family of protein-serine/threonine kinases that are activated and tyrosine phosphorylated in response to insulin and NGF. Cell, 65(4):663-75.

D. Brancho, N. Tanaka, A. Jaeschke, J. J. Ventura, N. Kelkar, et al. 2003. Mechanism of p38 MAP kinase activation in vivo. Genes Dev, 17(16):1969-78.

D. Brancho, J. J. Ventura, A. Jaeschke, B. Doran, R. A. Flavell, et al. 2005. Role of MLK3 in the regulation of mitogen-activated protein kinase signaling cascades. Mol Cell Biol, 25(9):3670-81.

D. G. Breckenridge and D. Xue. 2004. Regulation of mitochondrial membrane permeabilization by BCL-2 family proteins and caspases. Curr Opin Cell Biol, 16(6):647- 52.

L. Buday and J. Downward. 2008. Many faces of Ras activation. Biochim Biophys Acta, 1786(2):178-87.

96

P. D. Burbelo, D. Drechsel and A. Hall. 1995. A conserved binding motif defines numerous candidate target proteins for both Cdc42 and Rac GTPases. J Biol Chem, 270(49):29071-4.

M. Buschbeck, S. Hofbauer, L. Di Croce, G. Keri and A. Ullrich. 2005. Abl-kinase- sensitive levels of ERK5 and its intrinsic basal activity contribute to leukaemia cell survival. EMBO Rep, 6(1):63-9.

G. S. Butler, M. J. Butler, S. J. Atkinson, H. Will, T. Tamura, et al. 1998. The TIMP2 membrane type 1 metalloproteinase "receptor" regulates the concentration and efficient activation of progelatinase A. A kinetic study. J Biol Chem, 273(2):871-80.

C. Caelles, A. Helmberg and M. Karin. 1994. p53-dependent apoptosis in the absence of transcriptional activation of p53-target genes. Nature, 370(6486):220-3.

B. J. Canagarajah, A. Khokhlatchev, M. H. Cobb and E. J. Goldsmith. 1997. Activation mechanism of the MAP kinase ERK2 by dual phosphorylation. Cell, 90(5):859-69.

D. N. Chadee and J. M. Kyriakis. 2004. MLK3 is required for mitogen activation of B- Raf, ERK and cell proliferation. Nat Cell Biol, 6(8):770-6.

D. N. Chadee and J. M. Kyriakis. 2004. A novel role for mixed lineage kinase 3 (MLK3) in B-Raf activation and cell proliferation. Cell Cycle, 3(10):1227-9.

P. M. Chan, L. Lim and E. Manser. 2008. PAK is regulated by PI3K, PIX, CDC42, and PP2Calpha and mediates focal adhesion turnover in the hyperosmotic stress-induced p38 pathway. J Biol Chem, 283(36):24949-61.

J. Chen, E. M. Miller and K. A. Gallo. 2010. MLK3 is critical for breast cancer cell migration and promotes a malignant phenotype in mammary epithelial cells. Oncogene, 29(31):4399-411.

M. Chen and J. A. Cooper. 1995. Ser-3 is important for regulating Mos interaction with and stimulation of mitogen-activated protein kinase kinase. Mol Cell Biol, 15(9):4727- 34.

W. Chen, M. A. White and M. H. Cobb. 2002. Stimulus-specific requirements for MAP3 kinases in activating the JNK pathway. J Biol Chem, 277(51):49105-10.

Z. Chen, T. B. Gibson, F. Robinson, L. Silvestro, G. Pearson, et al. 2001. MAP kinases. Chem Rev, 101(8):2449-76.

P. C. Cheung, D. G. Campbell, A. R. Nebreda and P. Cohen. 2003. Feedback control of the protein kinase TAK1 by SAPK2a/p38alpha. EMBO J, 22(21):5793-805.

97 H. Chi, S. P. Barry, R. J. Roth, J. J. Wu, E. A. Jones, et al. 2006. Dynamic regulation of pro- and anti-inflammatory cytokines by MAPK phosphatase 1 (MKP-1) in innate immune responses. Proc Natl Acad Sci U S A, 103(7):2274-9.

H. Chi, M. R. Sarkisian, P. Rakic and R. A. Flavell. 2005. Loss of mitogen-activated protein kinase kinase kinase 4 (MEKK4) results in enhanced apoptosis and defective neural tube development. Proc Natl Acad Sci U S A, 102(10):3846-51.

M. Chinkers. 2001. Protein phosphatase 5 in signal transduction. Trends Endocrinol Metab, 12(1):28-32.

A. Cho, J. Graves and M. A. Reidy. 2000. Mitogen-activated protein kinases mediate matrix metalloproteinase-9 expression in vascular smooth muscle cells. Arterioscler Thromb Vasc Biol, 20(12):2527-32.

J. Cho, M. Melnick, G. P. Solidakis and P. N. Tsichlis. 2005. Tpl2 (tumor progression locus 2) phosphorylation at Thr290 is induced by lipopolysaccharide via an Ikappa-B Kinase-beta-dependent pathway and is required for Tpl2 activation by external signals. J Biol Chem, 280(21):20442-8.

Y. Y. Cho, A. M. Bode, H. Mizuno, B. Y. Choi, H. S. Choi, et al. 2004. A novel role for mixed-lineage kinase-like mitogen-activated protein triple kinase alpha in neoplastic cell transformation and tumor development. Cancer Res, 64(11):3855-64.

M. Christe, N. Jin, X. Wang, K. E. Gould, P. W. Iversen, et al. 2004. Transgenic mice with cardiac-specific over-expression of MLK7 have increased mortality when exposed to chronic beta-adrenergic stimulation. J Mol Cell Cardiol, 37(3):705-15.

E. T. Cole, Y. Zhan, W. F. Abi Saab, A. C. Korchnak, B. P. Ashburner, et al. 2009. Mixed lineage kinase 3 negatively regulates IKK activity and enhances etoposide- induced cell death. Biochim Biophys Acta, 1793(12):1811-8.

M. Costa, M. Marchi, F. Cardarelli, A. Roy, F. Beltram, et al. 2006. Dynamic regulation of ERK2 nuclear translocation and mobility in living cells. J Cell Sci, 119(Pt 23):4952- 63.

P. Coulombe, G. Rodier, S. Pelletier, J. Pellerin and S. Meloche. 2003. Rapid turnover of extracellular signal-regulated kinase 3 by the ubiquitin-proteasome pathway defines a novel paradigm of mitogen-activated protein kinase regulation during cellular differentiation. Mol Cell Biol, 23(13):4542-58.

N. W. Court, C. G. dos Remedios, J. Cordell and M. A. Bogoyevitch. 2002. Cardiac expression and subcellular localization of the p38 mitogen-activated protein kinase member, stress-activated protein kinase-3 (SAPK3). J Mol Cell Cardiol, 34(4):413-26.

98 E. A. Craig, M. V. Stevens, R. R. Vaillancourt and T. D. Camenisch. 2008. MAP3Ks as central regulators of cell fate during development. Dev Dyn, 237(11):3102-14.

C. M. Crews, A. Alessandrini and R. L. Erikson. 1992. The primary structure of MEK, a protein kinase that phosphorylates the ERK gene product. Science, 258(5081):478-80.

M. R. Cronan, K. Nakamura, N. L. Johnson, D. A. Granger, B. D. Cuevas, et al. 2012. Defining MAP3 kinases required for MDA-MB-231 cell tumor growth and metastasis. Oncogene, 31(34):3889-900.

A. Cuenda, P. Cohen, V. Buee-Scherrer and M. Goedert. 1997. Activation of stress- activated protein kinase-3 (SAPK3) by cytokines and cellular stresses is mediated via SAPKK3 (MKK6); comparison of the specificities of SAPK3 and SAPK2 (RK/p38). EMBO J, 16(2):295-305.

A. Cuenda and S. Rousseau. 2007. p38 MAP-kinases pathway regulation, function and role in human diseases. Biochim Biophys Acta, 1773(8):1358-75.

B. D. Cuevas, A. N. Abell and G. L. Johnson. 2007. Role of mitogen-activated protein kinase kinase kinases in signal integration. Oncogene, 26(22):3159-71.

B. D. Cuevas, A. N. Abell, J. A. Witowsky, T. Yujiri, N. L. Johnson, et al. 2003. MEKK1 regulates calpain-dependent proteolysis of focal adhesion proteins for rear-end detachment of migrating fibroblasts. EMBO J, 22(13):3346-55.

B. D. Cuevas, M. T. Uhlik, T. P. Garrington and G. L. Johnson. 2005. MEKK1 regulates the AP-1 dimer repertoire via control of JunB transcription and Fra-2 protein stability. Oncogene, 24(5):801-9.

S. C. Cunningham, E. Gallmeier, T. Hucl, D. A. Dezentje, E. S. Calhoun, et al. 2006. Targeted deletion of MKK4 in cancer cells: a detrimental phenotype manifests as decreased experimental metastasis and suggests a counterweight to the evolution of tumor-suppressor loss. Cancer Res, 66(11):5560-4.

S. R. D'Mello, C. Galli, T. Ciotti and P. Calissano. 1993. Induction of apoptosis in cerebellar granule neurons by low potassium: inhibition of death by insulin-like growth factor I and cAMP. Proc Natl Acad Sci U S A, 90(23):10989-93.

H. Davies, G. R. Bignell, C. Cox, P. Stephens, S. Edkins, et al. 2002. Mutations of the BRAF gene in human cancer. Nature, 417(6892):949-54.

D. N. Dhanasekaran and G. L. Johnson. 2007. MAPKs: function, regulation, role in cancer and therapeutic targeting. Oncogene, 26(22):3097-9.

D. N. Dhanasekaran and E. P. Reddy. 2008. JNK signaling in apoptosis. Oncogene, 27(48):6245-51.

99

A. S. Dhillon, S. Hagan, O. Rath and W. Kolch. 2007. MAP kinase signalling pathways in cancer. Oncogene, 26(22):3279-90.

J. DiDonato, F. Mercurio, C. Rosette, J. Wu-Li, H. Suyang, et al. 1996. Mapping of the inducible IkappaB phosphorylation sites that signal its ubiquitination and degradation. Mol Cell Biol, 16(4):1295-304.

J. Downward. 1996. Control of ras activation. Cancer Surv, 27:87-100.

Y. Du, B. C. Bock, K. A. Schachter, M. Chao and K. A. Gallo. 2005. Cdc42 induces activation loop phosphorylation and membrane targeting of mixed lineage kinase 3. J Biol Chem, 280(52):42984-93.

C. D. Dumitru, J. D. Ceci, C. Tsatsanis, D. Kontoyiannis, K. Stamatakis, et al. 2000. TNF-alpha induction by LPS is regulated posttranscriptionally via a Tpl2/ERK- dependent pathway. Cell, 103(7):1071-83.

K. L. Dunn, P. S. Espino, B. Drobic, S. He and J. R. Davie. 2005. The Ras-MAPK signal transduction pathway, cancer and chromatin remodeling. Biochem Cell Biol, 83(1):1-14.

S. Dutta, F. Q. Wang, A. Phalen and D. A. Fishman. 2010. Biomarkers for ovarian cancer detection and therapy. Cancer Biol Ther, 9(9):668-77.

A. G. Eliopoulos, C. C. Wang, C. D. Dumitru and P. N. Tsichlis. 2003. Tpl2 transduces CD40 and TNF signals that activate ERK and regulates IgE induction by CD40. EMBO J, 22(15):3855-64.

S. Estus, W. J. Zaks, R. S. Freeman, M. Gruda, R. Bravo, et al. 1994. Altered gene expression in neurons during : identification of c-jun as necessary for neuronal apoptosis. J Cell Biol, 127(6 Pt 1):1717-27.

G. R. Fanger, N. L. Johnson and G. L. Johnson. 1997. MEK kinases are regulated by EGF and selectively interact with Rac/Cdc42. EMBO J, 16(16):4961-72.

M. Faris, N. Kokot, K. Latinis, S. Kasibhatla, D. R. Green, et al. 1998. The c-Jun N- terminal kinase cascade plays a role in stress-induced apoptosis in Jurkat cells by up- regulating Fas ligand expression. J Immunol, 160(1):134-44.

C. Feijoo, D. G. Campbell, R. Jakes, M. Goedert and A. Cuenda. 2005. Evidence that phosphorylation of the microtubule-associated protein Tau by SAPK4/p38delta at Thr50 promotes microtubule assembly. J Cell Sci, 118(Pt 2):397-408.

A. Frandsen and A. Schousboe. 1990. Development of excitatory amino acid induced cytotoxicity in cultured neurons. Int J Dev Neurosci, 8(2):209-16.

100 T. Furuchi and R. G. Anderson. 1998. Cholesterol depletion of caveolae causes hyperactivation of extracellular signal-related kinase (ERK). J Biol Chem, 273(33):21099-104.

K. A. Gallo and G. L. Johnson. 2002. Mixed-lineage kinase control of JNK and p38 MAPK pathways. Nat Rev Mol Cell Biol, 3(9):663-72.

K. A. Gallo, M. R. Mark, D. T. Scadden, Z. Wang, Q. Gu, et al. 1994. Identification and characterization of SPRK, a novel src-homology 3 domain-containing proline-rich kinase with serine/threonine kinase activity. J Biol Chem, 269(21):15092-100.

M. J. Garnett and R. Marais. 2004. Guilty as charged: B-RAF is a human oncogene. Cancer Cell, 6(4):313-9.

M. J. Garnett, S. Rana, H. Paterson, D. Barford and R. Marais. 2005. Wild-type and mutant B-RAF activate C-RAF through distinct mechanisms involving heterodimerization. Mol Cell, 20(6):963-9.

T. P. Garrington, T. Ishizuka, P. J. Papst, K. Chayama, S. Webb, et al. 2000. MEKK2 gene disruption causes loss of cytokine production in response to IgE and c-Kit ligand stimulation of ES cell-derived mast cells. EMBO J, 19(20):5387-95.

B. Ge, H. Gram, F. Di Padova, B. Huang, L. New, et al. 2002. MAPKK-independent activation of p38alpha mediated by TAB1-dependent autophosphorylation of p38alpha. Science, 295(5558):1291-4.

P. Gerwins, J. L. Blank and G. L. Johnson. 1997. Cloning of a novel mitogen-activated protein kinase kinase kinase, MEKK4, that selectively regulates the c-Jun amino terminal kinase pathway. J Biol Chem, 272(13):8288-95.

S. Ghosh. 1999. Regulation of inducible gene expression by the transcription factor NF- kappaB. Immunol Res, 19(2-3):183-9.

S. Ghosh and M. Karin. 2002. Missing pieces in the NF-kappaB puzzle. Cell, 109 Suppl:S81-96.

M. Goedert, A. Cuenda, M. Craxton, R. Jakes and P. Cohen. 1997. Activation of the novel stress-activated protein kinase SAPK4 by cytokines and cellular stresses is mediated by SKK3 (MKK6); comparison of its substrate specificity with that of other SAP kinases. EMBO J, 16(12):3563-71.

I. Gotoh, M. Adachi and E. Nishida. 2001. Identification and characterization of a novel MAP kinase kinase kinase, MLTK. J Biol Chem, 276(6):4276-86.

M. E. Guicciardi and G. J. Gores. 2009. Life and death by death receptors. FASEB J, 23(6):1625-37.

101

D. Hanahan and R. A. Weinberg. 2011. Hallmarks of cancer: the next generation. Cell, 144(5):646-74.

C. A. Harris, M. Deshmukh, B. Tsui-Pierchala, A. C. Maroney and E. M. Johnson, Jr. 2002. Inhibition of the c-Jun N-terminal kinase signaling pathway by the mixed lineage kinase inhibitor CEP-1347 (KT7515) preserves metabolism and growth of trophic factor- deprived neurons. J Neurosci, 22(1):103-13.

J. Hartkamp, J. Troppmair and U. R. Rapp. 1999. The JNK/SAPK activator mixed lineage kinase 3 (MLK3) transforms NIH 3T3 cells in a MEK-dependent fashion. Cancer Res, 59(9):2195-202.

M. Hasegawa, A. Cuenda, M. G. Spillantini, G. M. Thomas, V. Buee-Scherrer, et al. 1999. Stress-activated protein kinase-3 interacts with the PDZ domain of alpha1- syntrophin. A mechanism for specific substrate recognition. J Biol Chem, 274(18):12626- 31.

T. Hayakawa, A. Matsuzawa, T. Noguchi, K. Takeda and H. Ichijo. 2006. The ASK1- MAP kinase pathways in immune and stress responses. Microbes Infect, 8(4):1098-107.

T. A. Haystead, P. Dent, J. Wu, C. M. Haystead and T. W. Sturgill. 1992. Ordered phosphorylation of p42mapk by MAP kinase kinase. FEBS Lett, 306(1):17-22.

S. Hirai, F. Cui de, T. Miyata, M. Ogawa, H. Kiyonari, et al. 2006. The c-Jun N-terminal kinase activator dual leucine zipper kinase regulates axon growth and neuronal migration in the developing cerebral cortex. J Neurosci, 26(46):11992-2002.

S. Hirai, M. Katoh, M. Terada, J. M. Kyriakis, L. I. Zon, et al. 1997. MST/MLK2, a member of the mixed lineage kinase family, directly phosphorylates and activates SEK1, an activator of c-Jun N-terminal kinase/stress-activated protein kinase. J Biol Chem, 272(24):15167-73.

M. Hirano, S. Osada, T. Aoki, S. Hirai, M. Hosaka, et al. 1996. MEK kinase is involved in tumor necrosis factor alpha-induced NF-kappaB activation and degradation of IkappaB-alpha. J Biol Chem, 271(22):13234-8.

K. P. Hoeflich, M. T. Eby, W. F. Forrest, D. C. Gray, J. Y. Tien, et al. 2006. Regulation of ERK3/MAPK6 expression by BRAF. Int J Oncol, 29(4):839-49.

H. Hua, M. Li, T. Luo, Y. Yin and Y. Jiang. 2011. Matrix metalloproteinases in tumorigenesis: an evolving paradigm. Cell Mol Life Sci, 68(23):3853-68.

Q. Huang, J. Yang, Y. Lin, C. Walker, J. Cheng, et al. 2004. Differential regulation of interleukin 1 receptor and Toll-like receptor signaling by MEKK3. Nat Immunol, 5(1):98-103.

102

W. C. Huangfu, E. Omori, S. Akira, K. Matsumoto and J. Ninomiya-Tsuji. 2006. Osmotic stress activates the TAK1-JNK pathway while blocking TAK1-mediated NF- kappaB activation: TAO2 regulates TAK1 pathways. J Biol Chem, 281(39):28802-10.

M. Hutchison, K. S. Berman and M. H. Cobb. 1998. Isolation of TAO1, a protein kinase that activates MEKs in stress-activated protein kinase cascades. J Biol Chem, 273(44):28625-32.

H. Ichijo, E. Nishida, K. Irie, P. ten Dijke, M. Saitoh, et al. 1997. Induction of apoptosis by ASK1, a mammalian MAPKKK that activates SAPK/JNK and p38 signaling pathways. Science, 275(5296):90-4.

A. Ikeda, K. Hasegawa, M. Masaki, T. Moriguchi, E. Nishida, et al. 2001. Mixed lineage kinase LZK forms a functional signaling complex with JIP-1, a scaffold protein of the c- Jun NH(2)-terminal kinase pathway. J Biochem, 130(6):773-81.

Y. T. Ip and R. J. Davis. 1998. Signal transduction by the c-Jun N-terminal kinase (JNK)- -from inflammation to development. Curr Opin Cell Biol, 10(2):205-19.

T. Irie, T. Muta and K. Takeshige. 2000. TAK1 mediates an activation signal from toll- like receptor(s) to nuclear factor-kappaB in lipopolysaccharide-stimulated macrophages. FEBS Lett, 467(2-3):160-4.

S. Ishibe, D. Joly, Z. X. Liu and L. G. Cantley. 2004. Paxillin serves as an ERK-regulated scaffold for coordinating FAK and Rac activation in epithelial morphogenesis. Mol Cell, 16(2):257-67.

E. Ispanovic and T. L. Haas. 2006. JNK and PI3K differentially regulate MMP-2 and MT1-MMP mRNA and protein in response to actin cytoskeleton reorganization in endothelial cells. Am J Physiol Cell Physiol, 291(4):C579-88.

J. L. Jadrich, M. B. O'Connor and E. Coucouvanis. 2006. The TGF beta activated kinase TAK1 regulates vascular development in vivo. Development, 133(8):1529-41.

A. Jaeschke and R. J. Davis. 2007. Metabolic stress signaling mediated by mixed-lineage kinases. Mol Cell, 27(3):498-508.

S. A. Jesch, T. S. Lewis, N. G. Ahn and A. D. Linstedt. 2001. Mitotic phosphorylation of Golgi reassembly stacking protein 55 by mitogen-activated protein kinase ERK2. Mol Biol Cell, 12(6):1811-7.

Y. Jiang, C. Chen, Z. Li, W. Guo, J. A. Gegner, et al. 1996. Characterization of the structure and function of a new mitogen-activated protein kinase (p38beta). J Biol Chem, 271(30):17920-6.

103 G. L. Johnson and R. Lapadat. 2002. Mitogen-activated protein kinase pathways mediated by ERK, JNK, and p38 protein kinases. Science, 298(5600):1911-2.

R. Johnson, B. Spiegelman, D. Hanahan and R. Wisdom. 1996. Cellular transformation and malignancy induced by ras require c-jun. Mol Cell Biol, 16(8):4504-11.

T. Kanamoto, M. Mota, K. Takeda, L. L. Rubin, K. Miyazono, et al. 2000. Role of apoptosis signal-regulating kinase in regulation of the c-Jun N-terminal kinase pathway and apoptosis in sympathetic neurons. Mol Cell Biol, 20(1):196-204.

M. Karin, Y. Cao, F. R. Greten and Z. W. Li. 2002. NF-kappaB in cancer: from innocent bystander to major culprit. Nat Rev Cancer, 2(4):301-10.

V.I. Kashuba, E.V. Grigorieva, S.M. Kvasha, T.V. Pavlova, V. Grigoriev, et al. 2011. Cloning and initial functional characterization of MLK4α and MLK4β. Genomic Insights, 4:1-12.

Y. Kato, R. I. Tapping, S. Huang, M. H. Watson, R. J. Ulevitch, et al. 1998. Bmk1/Erk5 is required for cell proliferation induced by epidermal growth factor. Nature, 395(6703):713-6.

T. Kawai and S. Akira. 2007. Signaling to NF-kappaB by Toll-like receptors. Trends Mol Med, 13(11):460-9.

S. Kebache, J. Ash, M. G. Annis, J. Hagan, M. Huber, et al. 2007. Grb10 and active Raf- 1 kinase promote Bad-dependent cell survival. J Biol Chem, 282(30):21873-83.

K. Kesavan, K. Lobel-Rice, W. Sun, R. Lapadat, S. Webb, et al. 2004. MEKK2 regulates the coordinate activation of ERK5 and JNK in response to FGF-2 in fibroblasts. J Cell Physiol, 199(1):140-8.

S. M. Keyse. 2000. Protein phosphatases and the regulation of mitogen-activated protein kinase signalling. Curr Opin Cell Biol, 12(2):186-92.

K. Y. Kim, B. C. Kim, Z. Xu and S. J. Kim. 2004. Mixed lineage kinase 3 (MLK3)- activated p38 MAP kinase mediates transforming growth factor-beta-induced apoptosis in hepatoma cells. J Biol Chem, 279(28):29478-84.

R. Kim. 2005. Unknotting the roles of Bcl-2 and Bcl-xL in cell death. Biochem Biophys Res Commun, 333(2):336-43.

S. Kim, J. H. Oh, Y. Lee, J. Lee, K. H. Cho, et al. 2010. Induction of tissue inhibitor of matrix metalloproteinase-2 by cholesterol depletion leads to the conversion of proMMP-2 into active MMP-2 in human dermal fibroblasts. Exp Mol Med, 42(1):38-46.

104 P. L. Kovalenko, L. Kunovska, J. Chen, K. A. Gallo and M. D. Basson. 2012. Loss of MLK3 signaling impedes ulcer healing by modulating MAPK signaling in mouse intestinal mucosa. Am J Physiol Gastrointest Liver Physiol, 303(8):G951-60.

S. F. Krens, H. P. Spaink and B. E. Snaar-Jagalska. 2006. Functions of the MAPK family in vertebrate-development. FEBS Lett, 580(21):4984-90.

J. M. Kyriakis and J. Avruch. 2001. Mammalian mitogen-activated protein kinase signal transduction pathways activated by stress and inflammation. Physiol Rev, 81(2):807-69.

J. M. Kyriakis, P. Banerjee, E. Nikolakaki, T. Dai, E. A. Rubie, et al. 1994. The stress- activated protein kinase subfamily of c-Jun kinases. Nature, 369(6476):156-60.

T. Lamark, M. Perander, H. Outzen, K. Kristiansen, A. Overvatn, et al. 2003. Interaction codes within the family of mammalian Phox and Bem1p domain-containing proteins. J Biol Chem, 278(36):34568-81.

S. Lawler, Y. Fleming, M. Goedert and P. Cohen. 1998. Synergistic activation of SAPK1/JNK1 by two MAP kinase kinases in vitro. Curr Biol, 8(25):1387-90.

C. Lechner, M. A. Zahalka, J. F. Giot, N. P. Moller and A. Ullrich. 1996. ERK6, a mitogen-activated protein kinase involved in C2C12 myoblast differentiation. Proc Natl Acad Sci U S A, 93(9):4355-9.

J. C. Lee, J. T. Laydon, P. C. McDonnell, T. F. Gallagher, S. Kumar, et al. 1994. A protein kinase involved in the regulation of inflammatory cytokine biosynthesis. Nature, 372(6508):739-46.

I. W. Leung and N. Lassam. 2001. The kinase activation loop is the key to mixed lineage kinase-3 activation via both autophosphorylation and hematopoietic progenitor kinase 1 phosphorylation. J Biol Chem, 276(3):1961-7.

R. Lezama, A. Diaz-Tellez, G. Ramos-Mandujano, L. Oropeza and H. Pasantes-Morales. 2005. Epidermal growth factor receptor is a common element in the signaling pathways activated by cell volume changes in isosmotic, hyposmotic or hyperosmotic conditions. Neurochem Res, 30(12):1589-97.

J. Liu, G. Yang, J. A. Thompson-Lanza, A. Glassman, K. Hayes, et al. 2004. A genetically defined model for human ovarian cancer. Cancer Res, 64(5):1655-63.

G. Lu, Y. J. Kang, J. Han, H. R. Herschman, E. Stefani, et al. 2006. TAB-1 modulates intracellular localization of p38 MAP kinase and downstream signaling. J Biol Chem, 281(9):6087-95.

105 Z. Lu, S. Xu, C. Joazeiro, M. H. Cobb and T. Hunter. 2002. The PHD domain of MEKK1 acts as an E3 ubiquitin ligase and mediates ubiquitination and degradation of ERK1/2. Mol Cell, 9(5):945-56.

J. C. Luckett, M. B. Huser, N. Giagtzoglou, J. E. Brown and C. A. Pritchard. 2000. Expression of the A-raf proto-oncogene in the normal adult and embryonic mouse. Cell Growth Differ, 11(3):163-71.

M. Maeda and R. A. Firtel. 1997. Activation of the mitogen-activated protein kinase ERK2 by the chemoattractant folic acid in Dictyostelium. J Biol Chem, 272(38):23690-5.

M. Marchi, A. D'Antoni, I. Formentini, R. Parra, R. Brambilla, et al. 2008. The N- terminal domain of ERK1 accounts for the functional differences with ERK2. PLoS One, 3(12):e3873.

E. Marcora, K. Gowan and J. E. Lee. 2003. Stimulation of NeuroD activity by huntingtin and huntingtin-associated proteins HAP1 and MLK2. Proc Natl Acad Sci U S A, 100(16):9578-83.

M. Martini, M. Russo, S. Lamba, E. Vitiello, E. Hannah Crowley, et al. 2013. Mixed lineage kinase MLK4 is activated in colorectal cancers where it synergestically cooperates with activated Ras signaling in driving tumorigenesis. Cancer Research, DOI:10.1158/0008-5472.CAN-12-3074.

M. Mata, S. E. Merritt, G. Fan, G. G. Yu and L. B. Holzman. 1996. Characterization of dual leucine zipper-bearing kinase, a mixed lineage kinase present in synaptic terminals whose phosphorylation state is regulated by membrane depolarization via calcineurin. J Biol Chem, 271(28):16888-96.

A. Matsuzawa, K. Saegusa, T. Noguchi, C. Sadamitsu, H. Nishitoh, et al. 2005. ROS- dependent activation of the TRAF6-ASK1-p38 pathway is selectively required for TLR4- mediated innate immunity. Nat Immunol, 6(6):587-92.

N. Maulik, M. Sato, B. D. Price and D. K. Das. 1998. An essential role of NFkappaB in tyrosine kinase signaling of p38 MAP kinase regulation of myocardial adaptation to ischemia. FEBS Lett, 429(3):365-9.

P. B. Mehta, B. L. Jenkins, L. McCarthy, L. Thilak, C. N. Robson, et al. 2003. MEK5 overexpression is associated with metastatic prostate cancer, and stimulates proliferation, MMP-9 expression and invasion. Oncogene, 22(9):1381-9.

M. Mikula, M. Schreiber, Z. Husak, L. Kucerova, J. Ruth, et al. 2001. Embryonic lethality and fetal liver apoptosis in mice lacking the c-raf-1 gene. EMBO J, 20(8):1952- 62.

106 P. Mishra, S. Senthivinayagam, V. Rangasamy, G. Sondarva and B. Rana. 2010. Mixed lineage kinase-3/JNK1 axis promotes migration of human gastric cancer cells following gastrin stimulation. Mol Endocrinol, 24(3):598-607.

H. Mita, J. Tsutsui, M. Takekawa, E. A. Witten and H. Saito. 2002. Regulation of MTK1/MEKK4 kinase activity by its N-terminal autoinhibitory domain and GADD45 binding. Mol Cell Biol, 22(13):4544-55.

T. Miyashita and J. C. Reed. 1995. Tumor suppressor p53 is a direct transcriptional activator of the human bax gene. Cell, 80(2):293-9.

J. C. Montero, A. Ocana, M. Abad, M. J. Ortiz-Ruiz, A. Pandiella, et al. 2009. Expression of Erk5 in early stage breast cancer and association with disease free survival identifies this kinase as a potential therapeutic target. PLoS One, 4(5):e5565.

E. Morgunova, A. Tuuttila, U. Bergmann, M. Isupov, Y. Lindqvist, et al. 1999. Structure of human pro-matrix metalloproteinase-2: activation mechanism revealed. Science, 284(5420):1667-70.

T. Moriguchi, F. Toyoshima, N. Masuyama, H. Hanafusa, Y. Gotoh, et al. 1997. A novel SAPK/JNK kinase, MKK7, stimulated by TNFalpha and cellular stresses. EMBO J, 16(23):7045-53.

D. K. Morrison and R. J. Davis. 2003. Regulation of MAP kinase signaling modules by scaffold proteins in mammals. Annu Rev Cell Dev Biol, 19:91-118.

M. Mota, M. Reeder, J. Chernoff and C. E. Bazenet. 2001. Evidence for a role of mixed lineage kinases in neuronal apoptosis. J Neurosci, 21(14):4949-57.

H. R. Mott, J. W. Carpenter, S. Zhong, S. Ghosh, R. M. Bell, et al. 1996. The solution structure of the Raf-1 cysteine-rich domain: a novel ras and phospholipid binding site. Proc Natl Acad Sci U S A, 93(16):8312-7.

H. Nagase and J. F. Woessner, Jr. 1999. Matrix metalloproteinases. J Biol Chem, 274(31):21491-4.

K. Nakamura and G. L. Johnson. 2003. PB1 domains of MEKK2 and MEKK3 interact with the MEK5 PB1 domain for activation of the ERK5 pathway. J Biol Chem, 278(39):36989-92.

K. Nakamura, M. T. Uhlik, N. L. Johnson, K. M. Hahn and G. L. Johnson. 2006. PB1 domain-dependent signaling complex is required for extracellular signal-regulated kinase 5 activation. Mol Cell Biol, 26(6):2065-79.

D. Nihalani, S. Merritt and L. B. Holzman. 2000. Identification of structural and functional domains in mixed lineage kinase dual leucine zipper-bearing kinase required

107 for complex formation and stress-activated protein kinase activation. J Biol Chem, 275(10):7273-9.

D. Nihalani, D. Meyer, S. Pajni and L. B. Holzman. 2001. Mixed lineage kinase- dependent JNK activation is governed by interactions of scaffold protein JIP with MAPK module components. EMBO J, 20(13):3447-58.

K. Okazaki and N. Sagata. 1995. The Mos/MAP kinase pathway stabilizes c-Fos by phosphorylation and augments its transforming activity in NIH 3T3 cells. EMBO J, 14(20):5048-59.

E. Ortner and K. Moelling. 2007. Heteromeric complex formation of ASK2 and ASK1 regulates stress-induced signaling. Biochem Biophys Res Commun, 362(2):454-9.

C. M. Overall and C. Lopez-Otin. 2002. Strategies for MMP inhibition in cancer: innovations for the post-trial era. Nat Rev Cancer, 2(9):657-72.

D. M. Owens and S. M. Keyse. 2007. Differential regulation of MAP kinase signalling by dual-specificity protein phosphatases. Oncogene, 26(22):3203-13.

J. Pan, Q. G. Zhang and G. Y. Zhang. 2005. The neuroprotective effects of K252a through inhibiting MLK3/MKK7/JNK3 signaling pathway on ischemic brain injury in rat hippocampal CA1 region. Neuroscience, 131(1):147-59.

C. G. Parker, J. Hunt, K. Diener, M. McGinley, B. Soriano, et al. 1998. Identification of stathmin as a novel substrate for p38 delta. Biochem Biophys Res Commun, 249(3):791- 6.

G. Pearson, F. Robinson, T. Beers Gibson, B. E. Xu, M. Karandikar, et al. 2001. Mitogen-activated protein (MAP) kinase pathways: regulation and physiological functions. Endocr Rev, 22(2):153-83.

N. Pedraza, M. Rafel, I. Navarro, M. Encinas, M. Aldea, et al. 2009. Mixed lineage kinase phosphorylates transcription factor E47 and inhibits TrkB expression to link neuronal death and survival pathways. J Biol Chem, 284(47):32980-8.

D. R. Phelan, G. Price, Y. F. Liu and D. S. Dorow. 2001. Activated JNK phosphorylates the c-terminal domain of MLK2 that is required for MLK2-induced apoptosis. J Biol Chem, 276(14):10801-10.

L. Poitras, S. Jean, N. Islam and T. Moss. 2003. PAK interacts with NCK and MLK2 to regulate the activation of jun N-terminal kinase. FEBS Lett, 543(1-3):129-35.

J. Raingeaud, S. Gupta, J. S. Rogers, M. Dickens, J. Han, et al. 1995. Pro-inflammatory cytokines and environmental stress cause p38 mitogen-activated protein kinase activation by dual phosphorylation on tyrosine and threonine. J Biol Chem, 270(13):7420-6.

108

J. Raingeaud, A. J. Whitmarsh, T. Barrett, B. Derijard and R. J. Davis. 1996. MKK3- and MKK6-regulated gene expression is mediated by the p38 mitogen-activated protein kinase signal transduction pathway. Mol Cell Biol, 16(3):1247-55.

M. Raman, W. Chen and M. H. Cobb. 2007. Differential regulation and properties of MAPKs. Oncogene, 26(22):3100-12.

M. Raman, S. Earnest, K. Zhang, Y. Zhao and M. H. Cobb. 2007. TAO kinases mediate activation of p38 in response to DNA damage. EMBO J, 26(8):2005-14.

A. Rana, K. Gallo, P. Godowski, S. Hirai, S. Ohno, et al. 1996. The mixed lineage kinase SPRK phosphorylates and activates the stress-activated protein kinase activator, SEK-1. J Biol Chem, 271(32):19025-8.

G. Remy, A. M. Risco, F. A. Inesta-Vaquera, B. Gonzalez-Teran, G. Sabio, et al. 2010. Differential activation of p38MAPK isoforms by MKK6 and MKK3. Cell Signal, 22(4):660-7.

A. A. Reszka, R. Seger, C. D. Diltz, E. G. Krebs and E. H. Fischer. 1995. Association of mitogen-activated protein kinase with the microtubule cytoskeleton. Proc Natl Acad Sci U S A, 92(19):8881-5.

R. Roy, J. Yang and M. A. Moses. 2009. Matrix metalloproteinases as novel biomarkers and potential therapeutic targets in human cancer. J Clin Oncol, 27(31):5287-97.

L. K. Rushworth, A. D. Hindley, E. O'Neill and W. Kolch. 2006. Regulation and role of Raf-1/B-Raf heterodimerization. Mol Cell Biol, 26(6):2262-72.

G. Sabio, J. S. Arthur, Y. Kuma, M. Peggie, J. Carr, et al. 2005. p38gamma regulates the localisation of SAP97 in the cytoskeleton by modulating its interaction with GKAP. EMBO J, 24(6):1134-45.

G. Sabio, S. Reuver, C. Feijoo, M. Hasegawa, G. M. Thomas, et al. 2004. Stress- and mitogen-induced phosphorylation of the synapse-associated protein SAP90/PSD-95 by activation of SAPK3/p38gamma and ERK1/ERK2. Biochem J, 380(Pt 1):19-30.

H. Sakuma, A. Ikeda, S. Oka, Y. Kozutsumi, J. P. Zanetta, et al. 1997. Molecular cloning and functional expression of a cDNA encoding a new member of mixed lineage protein kinase from human brain. J Biol Chem, 272(45):28622-9.

H. Sakurai, N. Shigemori, K. Hasegawa and T. Sugita. 1998. TGF-beta-activated kinase 1 stimulates NF-kappa B activation by an NF-kappa B-inducing kinase-independent mechanism. Biochem Biophys Res Commun, 243(2):545-9.

109 J. M. Salvador, P. R. Mittelstadt, G. I. Belova, A. J. Fornace, Jr. and J. D. Ashwell. 2005. The autoimmune suppressor Gadd45alpha inhibits the T cell alternative p38 activation pathway. Nat Immunol, 6(4):396-402.

K. A. Sauter, E. A. Magun, M. S. Iordanov and B. E. Magun. 2010. ZAK is required for doxorubicin, a novel ribotoxic stressor, to induce SAPK activation and apoptosis in HaCaT cells. Cancer Biol Ther, 10(3):258-66.

K. A. Schachter, Y. Du, A. Lin and K. A. Gallo. 2006. Dynamic positive feedback phosphorylation of mixed lineage kinase 3 by JNK reversibly regulates its distribution to Triton-soluble domains. J Biol Chem, 281(28):19134-44.

B. Schmalfeldt, D. Prechtel, K. Harting, K. Spathe, S. Rutke, et al. 2001. Increased expression of matrix metalloproteinases (MMP)-2, MMP-9, and the urokinase-type plasminogen activator is associated with progression from benign to advanced ovarian cancer. Clin Cancer Res, 7(8):2396-404.

R. Seger and E. G. Krebs. 1995. The MAPK signaling cascade. FASEB J, 9(9):726-35.

A. Seit-Nebi, W. Cheng, H. Xu and J. Han. 2012. MLK4 has negative effect on TLR4 signaling. Cell Mol Immunol, 9(1):27-33.

O. M. Seternes, B. Johansen, B. Hegge, M. Johannessen, S. M. Keyse, et al. 2002. Both binding and activation of p38 mitogen-activated protein kinase (MAPK) play essential roles in regulation of the nucleocytoplasmic distribution of MAPK-activated protein kinase 5 by cellular stress. Mol Cell Biol, 22(20):6931-45.

R. X. Shao, N. Kato, L. J. Lin, R. Muroyama, M. Moriyama, et al. 2007. Absence of tyrosine kinase mutations in Japanese colorectal cancer patients. Oncogene, 26(14):2133- 5.

A. D. Sharrocks, S. H. Yang and A. Galanis. 2000. Docking domains and substrate- specificity determination for MAP kinases. Trends Biochem Sci, 25(9):448-53.

J. H. Shim, C. Xiao, A. E. Paschal, S. T. Bailey, P. Rao, et al. 2005. TAK1, but not TAB1 or TAB2, plays an essential role in multiple signaling pathways in vivo. Genes Dev, 19(22):2668-81.

I. Shin, S. Kim, H. Song, H. R. Kim and A. Moon. 2005. H-Ras-specific activation of Rac-MKK3/6-p38 pathway: its critical role in invasion and migration of breast epithelial cells. J Biol Chem, 280(15):14675-83.

T. Smeal, B. Binetruy, D. A. Mercola, M. Birrer and M. Karin. 1991. Oncogenic and transcriptional cooperation with Ha-Ras requires phosphorylation of c-Jun on serines 63 and 73. Nature, 354(6353):494-6.

110 Y. H. Soung, J. W. Lee, S. Y. Kim, S. W. Nam, W. S. Park, et al. 2006. Kinase domain mutation of MLK4 gene is uncommon in gastric and hepatocellular carcinomas. Dig Liver Dis, 38(4):283.

P. Spallarossa, P. Altieri, S. Garibaldi, G. Ghigliotti, C. Barisione, et al. 2006. Matrix metalloproteinase-2 and -9 are induced differently by doxorubicin in H9c2 cells: The role of MAP kinases and NAD(P)H oxidase. Cardiovasc Res, 69(3):736-45.

E. B. Springman, E. L. Angleton, H. Birkedal-Hansen and H. E. Van Wart. 1990. Multiple modes of activation of latent human fibroblast collagenase: evidence for the role of a Cys73 active-site zinc complex in latency and a "cysteine switch" mechanism for activation. Proc Natl Acad Sci U S A, 87(1):364-8.

S. S. Sridhar, D. Hedley and L. L. Siu. 2005. Raf kinase as a target for anticancer therapeutics. Mol Cancer Ther, 4(4):677-85.

A. Strasser, P. J. Jost and S. Nagata. 2009. The many roles of signaling in the immune system. Immunity, 30(2):180-92.

W. Sun, K. Kesavan, B. C. Schaefer, T. P. Garrington, M. Ware, et al. 2001. MEKK2 associates with the adapter protein Lad/RIBP and regulates the MEK5-BMK1/ERK5 pathway. J Biol Chem, 276(7):5093-100.

K. I. Swenson-Fields, J. C. Sandquist, J. Rossol-Allison, I. C. Blat, K. Wennerberg, et al. 2008. MLK3 limits activated Galphaq signaling to Rho by binding to p63RhoGEF. Mol Cell, 32(1):43-56.

G. Takaesu, R. M. Surabhi, K. J. Park, J. Ninomiya-Tsuji, K. Matsumoto, et al. 2003. TAK1 is critical for IkappaB kinase-mediated activation of the NF-kappaB pathway. J Mol Biol, 326(1):105-15.

K. Takeda, R. Shimozono, T. Noguchi, T. Umeda, Y. Morimoto, et al. 2007. Apoptosis signal-regulating kinase (ASK) 2 functions as a mitogen-activated protein kinase kinase kinase in a heteromeric complex with ASK1. J Biol Chem, 282(10):7522-31.

M. Takekawa, M. Adachi, A. Nakahata, I. Nakayama, F. Itoh, et al. 2000. p53-inducible wip1 phosphatase mediates a negative feedback regulation of p38 MAPK-p53 signaling in response to UV radiation. EMBO J, 19(23):6517-26.

M. Takekawa, T. Maeda and H. Saito. 1998. Protein phosphatase 2Calpha inhibits the human stress-responsive p38 and JNK MAPK pathways. EMBO J, 17(16):4744-52.

M. Takekawa and H. Saito. 1998. A family of stress-inducible GADD45-like proteins mediate activation of the stress-responsive MTK1/MEKK4 MAPKKK. Cell, 95(4):521- 30.

111 M. Takekawa, K. Tatebayashi and H. Saito. 2005. Conserved docking site is essential for activation of mammalian MAP kinase kinases by specific MAP kinase kinase kinases. Mol Cell, 18(3):295-306.

R. X. Tang, F. Y. Kong, B. F. Fan, X. M. Liu, H. J. You, et al. 2012. HBx activates FasL and mediates HepG2 cell apoptosis through MLK3-MKK7-JNKs signal module. World J Gastroenterol, 18(13):1485-95.

T. Tanoue and E. Nishida. 2003. Molecular recognitions in the MAP kinase cascades. Cell Signal, 15(5):455-62.

H. Teramoto, O. A. Coso, H. Miyata, T. Igishi, T. Miki, et al. 1996. Signaling from the small GTP-binding proteins Rac1 and Cdc42 to the c-Jun N-terminal kinase/stress- activated protein kinase pathway. A role for mixed lineage kinase 3/protein-tyrosine kinase 1, a novel member of the mixed lineage kinase family. J Biol Chem, 271(44):27225-8.

A. M. Tester, J. H. Cox, A. R. Connor, A. E. Starr, R. A. Dean, et al. 2007. LPS responsiveness and neutrophil chemotaxis in vivo require PMN MMP-8 activity. PLoS One, 2(3):e312.

J. P. Thiery, H. Acloque, R. Y. Huang and M. A. Nieto. 2009. Epithelial-mesenchymal transitions in development and disease. Cell, 139(5):871-90.

L. A. Tibbles, Y. L. Ing, F. Kiefer, J. Chan, N. Iscove, et al. 1996. MLK-3 activates the SAPK/JNK and p38/RK pathways via SEK1 and MKK3/6. EMBO J, 15(24):7026-35.

K. Tobiume, T. Inage, K. Takeda, S. Enomoto, K. Miyazono, et al. 1997. Molecular cloning and characterization of the mouse apoptosis signal-regulating kinase 1. Biochem Biophys Res Commun, 239(3):905-10.

M. Toth, I. Chvyrkova, M. M. Bernardo, S. Hernandez-Barrantes and R. Fridman. 2003. Pro-MMP-9 activation by the MT1-MMP/MMP-2 axis and MMP-3: role of TIMP-2 and plasma membranes. Biochem Biophys Res Commun, 308(2):386-95.

T. Turpeenniemi-Hujanen. 2005. Gelatinases (MMP-2 and -9) and their natural inhibitors as prognostic indicators in solid cancers. Biochimie, 87(3-4):287-97.

M. T. Uhlik, A. N. Abell, B. D. Cuevas, K. Nakamura and G. L. Johnson. 2004. Wiring diagrams of MAPK regulation by MEKK1, 2, and 3. Biochem Cell Biol, 82(6):658-63.

M. T. Uhlik, A. N. Abell, N. L. Johnson, W. Sun, B. D. Cuevas, et al. 2003. Rac- MEKK3-MKK3 scaffolding for p38 MAPK activation during hyperosmotic shock. Nat Cell Biol, 5(12):1104-10.

112 A. Umar, J. L. Viner, E. Richmond, W. F. Anderson and E. T. Hawk. 2002. Chemoprevention of colorectal carcinogenesis. Int J Clin Oncol, 7(1):2-26.

P. O. Vacratsis and K. A. Gallo. 2000. Zipper-mediated oligomerization of the mixed lineage kinase SPRK/MLK-3 is not required for its activation by the GTPase cdc 42 but Is necessary for its activation of the JNK pathway. Monomeric SPRK L410P does not catalyze the activating phosphorylation of Thr258 of murine MITOGEN-ACTIVATED protein kinase kinase 4. J Biol Chem, 275(36):27893-900.

P. O. Vacratsis, B. S. Phinney, D. A. Gage and K. A. Gallo. 2002. Identification of in vivo phosphorylation sites of MLK3 by mass spectrometry and phosphopeptide mapping. Biochemistry, 41(17):5613-24.

A. B. Vojtek, S. M. Hollenberg and J. A. Cooper. 1993. Mammalian Ras interacts directly with the serine/threonine kinase Raf. Cell, 74(1):205-14.

M. Wagner and M. A. Siddiqui. 2007. Signal transduction in early heart development (I): cardiogenic induction and heart tube formation. Exp Biol Med (Maywood), 232(7):852- 65.

P. T. Wan, M. J. Garnett, S. M. Roe, S. Lee, D. Niculescu-Duvaz, et al. 2004. Mechanism of activation of the RAF-ERK signaling pathway by oncogenic mutations of B-RAF. Cell, 116(6):855-67.

C. Wang, L. Deng, M. Hong, G. R. Akkaraju, J. Inoue, et al. 2001. TAK1 is a ubiquitin- dependent kinase of MKK and IKK. Nature, 412(6844):346-51.

L. Wang, Y. Pan and J. L. Dai. 2004. Evidence of MKK4 pro-oncogenic activity in breast and pancreatic tumors. Oncogene, 23(35):5978-85.

X. Wang, A. Destrument and C. Tournier. 2007. Physiological roles of MKK4 and MKK7: insights from animal models. Biochim Biophys Acta, 1773(8):1349-57.

X. Wang, M. M. Mader, J. E. Toth, X. Yu, N. Jin, et al. 2005. Complete inhibition of anisomycin and UV radiation but not cytokine induced JNK and p38 activation by an aryl-substituted dihydropyrrolopyrazole quinoline and mixed lineage kinase 7 small interfering RNA. J Biol Chem, 280(19):19298-305.

C. B. Weldon, A. B. Scandurro, K. W. Rolfe, J. L. Clayton, S. Elliott, et al. 2002. Identification of mitogen-activated protein kinase kinase as a chemoresistant pathway in MCF-7 cells by using gene expression microarray. Surgery, 132(2):293-301.

C. Wellbrock, M. Karasarides and R. Marais. 2004. The RAF proteins take centre stage. Nat Rev Mol Cell Biol, 5(11):875-85.

113 J. Westermarck and V. M. Kahari. 1999. Regulation of matrix metalloproteinase expression in tumor invasion. FASEB J, 13(8):781-92.

C. Widmann, S. Gibson, M. B. Jarpe and G. L. Johnson. 1999. Mitogen-activated protein kinase: conservation of a three-kinase module from yeast to human. Physiol Rev, 79(1):143-80.

A. Wittinghofer, K. Scheffzek and M. R. Ahmadian. 1997. The interaction of Ras with GTPase-activating proteins. FEBS Lett, 410(1):63-7.

X. Wu, S. J. Noh, G. Zhou, J. E. Dixon and K. L. Guan. 1996. Selective activation of MEK1 but not MEK2 by A-Raf from epidermal growth factor-stimulated Hela cells. J Biol Chem, 271(6):3265-71.

Y. Xia, C. Makris, B. Su, E. Li, J. Yang, et al. 2000. MEK kinase 1 is critically required for c-Jun N-terminal kinase activation by proinflammatory stimuli and growth factor- induced cell migration. Proc Natl Acad Sci U S A, 97(10):5243-8.

J. Xu, S. Lamouille and R. Derynck. 2009. TGF-beta-induced epithelial to mesenchymal transition. Cell Res, 19(2):156-72.

L. G. Xu, L. Y. Li and H. B. Shu. 2004. TRAF7 potentiates MEKK3-induced AP1 and CHOP activation and induces apoptosis. J Biol Chem, 279(17):17278-82.

Z. Xu, A. C. Maroney, P. Dobrzanski, N. V. Kukekov and L. A. Greene. 2001. The MLK family mediates c-Jun N-terminal kinase activation in neuronal apoptosis. Mol Cell Biol, 21(14):4713-24.

O. Yamaguchi, T. Watanabe, K. Nishida, K. Kashiwase, Y. Higuchi, et al. 2004. Cardiac- specific disruption of the c-raf-1 gene induces cardiac dysfunction and apoptosis. J Clin Invest, 114(7):937-43.

J. Yang, M. Boerm, M. McCarty, C. Bucana, I. J. Fidler, et al. 2000. Mekk3 is essential for early embryonic cardiovascular development. Nat Genet, 24(3):309-13.

J. J. Yang, Y. J. Lee, H. H. Hung, W. P. Tseng, C. C. Tu, et al. 2010. ZAK inhibits human lung cancer cell growth via ERK and JNK activation in an AP-1-dependent manner. Cancer Sci, 101(6):1374-81.

S. Yoon and R. Seger. 2006. The extracellular signal-regulated kinase: multiple substrates regulate diverse cellular functions. Growth Factors, 24(1):21-44.

T. Yujiri, S. Sather, G. R. Fanger and G. L. Johnson. 1998. Role of MEKK1 in cell survival and activation of JNK and ERK pathways defined by targeted gene disruption. Science, 282(5395):1911-4.

114 T. Zarubin and J. Han. 2005. Activation and signaling of the p38 MAP kinase pathway. Cell Res, 15(1):11-8.

H. Zbinden-Foncea, L. Deldicque, N. Pierre, M. Francaux and J. M. Raymackers. 2012. TLR2 and TLR4 activation induces p38 MAPK-dependent phosphorylation of S6 kinase 1 in C2C12 myotubes. Cell Biol Int, 36(12):1107-13.

D. Zechner, R. Craig, D. S. Hanford, P. M. McDonough, R. A. Sabbadini, et al. 1998. MKK6 activates myocardial cell NF-kappaB and inhibits apoptosis in a p38 mitogen- activated protein kinase-dependent manner. J Biol Chem, 273(14):8232-9.

Y. Zhan, W. F. Abi Saab, N. Modi, A. M. Stewart, J. Liu, et al. 2012. Mixed lineage kinase 3 is required for matrix metalloproteinase expression and invasion in ovarian cancer cells. Exp Cell Res, 318(14):1641-8.

Y. Zhan, N. Modi, A. M. Stewart, R. I. Hieronimus, J. Liu, et al. 2011. Regulation of mixed lineage kinase 3 is required for Neurofibromatosis-2-mediated growth suppression in human cancer. Oncogene, 30(7):781-9.

H. Zhang and K. A. Gallo. 2001. Autoinhibition of mixed lineage kinase 3 through its Src homology 3 domain. J Biol Chem, 276(49):45598-603.

J. Zhang, B. Zhou, C. F. Zheng and Z. Y. Zhang. 2003. A bipartite mechanism for ERK2 recognition by its cognate regulators and substrates. J Biol Chem, 278(32):29901-12.

C. F. Zheng and K. L. Guan. 1993. Cloning and characterization of two distinct human extracellular signal-regulated kinase activator kinases, MEK1 and MEK2. J Biol Chem, 268(15):11435-9.

G. Zhou, T. Golden, I. V. Aragon and R. E. Honkanen. 2004. Ser/Thr protein phosphatase 5 inactivates hypoxia-induced activation of an apoptosis signal-regulating kinase 1/MKK-4/JNK signaling cascade. J Biol Chem, 279(45):46595-605.

H. Y. Zhou, Y. L. Pon and A. S. Wong. 2007. Synergistic effects of epidermal growth factor and hepatocyte growth factor on human ovarian cancer cell invasion and migration: role of extracellular signal-regulated kinase 1/2 and p38 mitogen-activated protein kinase. Endocrinology, 148(11):5195-208.

W. Zou, M. B. Greenblatt, J. H. Shim, S. Kant, B. Zhai, et al. 2011. MLK3 regulates bone development downstream of the faciogenital dysplasia protein FGD1 in mice. J Clin Invest, 121(11):4383-92.

115