PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 KIProj/Acc8-PreslclRC/DateDue.indd

 

NECESSITY AND SUFFICIENCY OF MITOGEN-ACTIVATED
PROTEIN KINASE KINASE SIGNALING PATHWAYS
FOR MELANOMA CELL PROLIFERATION
By

CHIH-SHIA LEE

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILSOPHY
Biochemistry and Molecular Biology

2010

ABSTRACT
NECESSITY AND SUFFICIENCY OF MITOGEN-ACTIVATED PROTEIN
KINASE KINASE SIGNALING PATHWAYS FOR MELANOMA CELL
PROLIFERATION
By

CHIH—SHIA LEE

The mitogen activated protein kinase kinases (MKK or MEK) signaling pathways
are critical for melanoma survival. Although MEK] and MEK2 generally are considered
functionally redundant, a few studies have indicated that they may have distinct
biological functions. However, in assessing the relative contribution of MEK] and
MEK2 towards extracellular signal-regulated kinase (BRIO-mediated biologic response,
investigators have relied on tests of necessity, not sufficiency. Dissecting both necessity
and sufficiency of MEK Signaling pathways is important to understand the interplay of
the signaling network. To do so, I used two complementary approaches to determine the
necessity and sufficiency of MEKI and MEK2 signaling pathways for melanoma cell
proliferation. To determine the necessity, I used isoform-specific siRNA to knock down
either MEK] or MEK2 in human melanoma SK-MEL—28 cells. An effect on
proliferation was Observed only when both MEK] and MEK2 were knocked down,
indicating that neither of the individual isoforms is necessary for SK-MEL—ZS cell
proliferation. To determine the sufficiency, I have developed a novel experimental
system that allows only one MEK/MKK signaling pathway to be present in cells. In this
system, anthrax lethal toxin (LeTx), a pan-MEK/MKK protease, is used to inhibit
multiple endogenous MEK/MKKS in cells, and either MEKl or MEK2 signaling pathway

is simultaneously rescued by the cleavage-resistant form of MEK (MEKcr). Surprisingly,

ERK activity persisted in LeTx-treated cells expressing MEK2cr but not MEK] cr.
Microarray analysis revealed groups Of non-overlapping downstream transcriptional
targets of MEK] and MEK2, and indicated a substantial rescue effect of MEK2cr on
proliferation pathways. Furthermore, LeTx efficiently inhibited the cell proliferation and
anchorage-independent growth of SK-MEL—28 cells expressing MEKl cr but not MEK2cr.
These results indicate MEK2 signaling is sufficient for ERK activation, melanoma cell
proliferation, and anchorage-independent growth, and demonstrate that in SK-MEL-28
cells MEK] and MEK2 signaling pathways are not redundant and interchangeable for

cell proliferation. I conclude that in the absence of other MKK, MEK2 is sufficient for
SK-MEL-28 cell proliferation. MEK] can conditionally compensate for MEK2 only in

the presence of other MKK.

© Copyright by
Chih-Shia Lee
2010

Acknowledgments

Nothing is sufficient to express how grateful I am for the experience of being a
graduate student and to be trained in the Duesbery Lab. 1 am everlastingly indebted to
Dr. Duesbery for his guidance and inexhaustible patience over the past several years.
The debt of gratitude that I owe Dr. Duesbery is so deep that I can only repay it by
keeping Dr. Duesbery’s words in mind. In an E-mail Dr. Duesbery wrote to me:

“Yoshio Masui, my PhD supervisor, told me that my training was conditional on my
agreement to do the same with my students. In turn, this obligation will pass to you.”

I would to thank Drs. Justin McCormick, Lee Kroos, Gregg Howe, and Kathy
Gallo for agreeing to serve as my Guidance Committee members, and for their invaluable
advice throughout my graduate school career. Many thanks are also necessary to Drs. Art
Alberts, Jeff MacKeigan, Cindy Miranti, and Kyle Furge (Investigators at the Van Andel
Research Institute) for their helpful advice and suggestions. I would like to thank Ms.
Elissa Boguslawski for her professional assistance with xenografi experiments, as well as
Mr. Karl Dykema for his help with microarray data analysis. I would also like to thank
the current and former members in the Duesbery Lab for their help, especially to
Dr. Jenn Bromberg-White for her invaluable advice, discussion, and encouragement.
Finally, I would like to thank the Core Services at the Van Andel Research Institute for
their technical help, Mr. David Nadziejka for his professional comments on scientific
writing, and Ms. Laura Holman and Laura Round for their administration support during
the past five years. This research would not be possible without their supportive help.

This dissertation is dedicated to my best-loved family members: my grandmother,

parents, aunt, brother and sister-in-law as well as the expected nephew.

Table of Contents

List of Tables ........................................................................................................... x
List of Figures ........................................................................................................ xi
List of Abbreviations ........................................................................................... xiii
Chapter I. General introduction: MEK signaling pathway and cancer ............... 1
1.1. Introduction to MEK signaling pathway ................................................ 1
1.2. Involvement of MEK Signaling pathway in human cancers .................. 4
1.3. Targeting the Raf-MEK—ERK pathway as a strategy for treating
cancers .................................................................................................... 4
1.3.1 . Small-molecule inhibitors .................................................................. 6
1.3.1.1. PD 098059 .................................................................................... 6
1.3.1.2. U0126 ........................................................................................... 7
1.3.1.3. PD 184352 (CI-1040) ................................................................... 8
1.3.1.4. Derivatives of PD 184352 .......................................................... 11
1.3.1.5. ARRY-142886 (AZD6244) ....................................................... 13
1.3.2. Mechanism of allosteric inhibition .................................................. 15
1.3.3. Biological inhibitors ......................................................................... 18
1.3.3.1. Anthrax lethal toxin ................................................................... 18
1.3.3.2. YopJ ........................................................................................... 22
1.3.4. MEK inhibitors fail tO generate clinical responses .......................... 24
1.4. Introduction to cutaneous melanoma ................................................... 26
1.4.1. Melanoma statistics .......................................................................... 27
1.4.2. Risk factors ...................................................................................... 28
1.4.3. Melanoma progression, the Clark model. ........................................ 28
1.4.4. Melanoma staging ............................................................................ 29
1.4.5. Treatment of melanoma ................................................................... 30
1.4.6. Common genetic abnormalities of melanoma ................................. 31
1.4.6.1. BRAF and NRAS ......................................................................... 31
1.4.6.2. CDKNZA (p161NK4 and p194" ) ............................................... 33
1.4.6.3. PTEN .......................................................................................... 35
1.4.7. Disease models ................................................................................. 37
1 .5. Discussion ............................................................................................ 41
1.6. Tables and figures ................................................................................ 43
Chapter II. Cleavage-resistant MEK proteins; a novel experimental model to
establish MEK sufficiency .............................................................. 52
1 .1 . Introduction .......................................................................................... 52
1.2. Results .................................................................................................. 53
1.2.1. Point mutations at the Pl' site render MEK resistant to LP-
mediated cleavage ............................................................................ 53
1.2.2. The cleavage-resistant mutation does not impair MEK activity. ..... 55

vi

1.2.3. Point mutations at the P1 ’ Site render other MKK members

resistant to LF-mediated cleavage .................................................... 57

1.2.4. MKK4 cleavage in mammalian cells ............................................... 57
1.2.5. MKK7 cleavage in mammalian cells ............................................... 60

1 .3. Discussion ............................................................................................ 61
1.4. Figures .................................................................................................. 66
1.5. Materials and methods ......................................................................... 77
1.5.1. Cell lines and stable cell line establishment ..................................... 77
1.5.2. Chemicals and LeTx ........................................................................ 77
1.5.3. VS-MEKcr construction. .................................................................. 77
1.5.4. Construction of MKK4-V5-His6 and deletion mutants. .................. 78
1.5.5. In-cell MEK cleavage assay. ............................................................ 79

1 .5.6. Irnmunoblotting. ............................................................................... 80
1.5.7. Immunoprecipitation ........................................................................ 81
1.5.8. In vitro kinase assay. ........................................................................ 82

Chapter III. Sufficiency and necessity of MEK signaling pathways for

melanoma cell proliferation ....................................................... 83
1 .1 . Introduction .......................................................................................... 83
1.2. Results .................................................................................................. 85
1.2.1. Necessity of MEK] and MEK2 signaling pathways for
melanoma cell proliferation ............................................................. 85
1.2.1.1. Necessity of MEK] and MEK2 signaling pathways for ERK
activation in melanoma cells ...................................................... 85
1.2.1.2. Necessity of MEK] and MEK2 signaling pathways for
melanoma cell cycle progression ............................................... 86
1.2.2. Sufficiency Of MEKI and MEK2 signaling pathways for
melanoma cell proliferation ............................................................. 87
1.2.2.1. MEK2, but not MEKI, is sufficient to maintain ERK2
activity ........................................................................................ 88
1.2.2.2. Genome-wide cDNA microarray reveals non-overlapping
transcriptional patterns downstream of MEK] and MEK2 ....... 88
1.2.2.3. MEK2cr, but not MEchr, rescued proliferation-related
pathways in LeTx-treated cells .................................................. 90
1.2.2.4. MEK2, but not MEK] , is sufficient for melanoma cell
proliferation in vitro. .................................................................. 90
1.2.2.5. Neither MKK3 nor MKK6 is sufficient for SK-MEL-28 cell
proliferation in vitro. .................................................................. 92
1.2.2.6. MEK2 signaling is sufficient for anchorage-independent
growth ........................................................................................ 92
1.2.3. A xenograft model to test the sufficiency Of MEK] and MEK2
signaling pathways for melanoma tumor growth in vivo ................. 93
1.2.3.1. SK-MEL-28 xenografi tumor growth ........................................ 93
1.2.3.2. LeTx systemic treatment of SK-MEL-28 xenograft tumors ...... 94
1.2.3.3. Sensitivity of MEKcr-expressing xenograft tumors to LeTx
systemic treatment ...................................................................... 95
vii

 

1.2.3.4. Dose-dependent effects of LeTx on SK-MEL-28 xenograft

tumor growth .............................................................................. 96
1.2.3.5. Loss of V5-MEKcr expression in SK-MEL-28 xenograft

tumors ......................................................................................... 97
1.2.3.6. LeTx systemic treatment does not cause MEK cleavage in

tumor cells. ................................................................................. 97

1.2.3.7. Sufficiency of MEK signaling pathways for SK—MEL-28
tumor growth in vivo was not testable due to technical

difficulties. ................................................................................. 98

1 .3. Discussion ............................................................................................ 99
1.4. Tables and figures .............................................................................. 110
1.5. Materials and methods ....................................................................... 143
1.5.1. Cell lines and stable cell line establishment ................................... 143
1.5.2. Chemicals and LeTx ...................................................................... 143
1.5.3. siRNA-mediated MEK knock down. ............................................. 143
1.5.4. In-cell MEK cleavage assay in SK—MEL-28 cells. ........................ 144

1 .5.5. Irnmunoblotting. ............................................................................. 145
1.5.6. Toxicity assay. ............................................................................... 146
1.5.7. SK-MEL-28 tumor xenografi and LeTx systemic treatment ......... 146
1.5.8. Anchorage-independent growth assay. .......................................... 148

1.5.9. Human cDNA microarray and transcriptional signature analysis. 148
1.5.10. Examination of p38 MAPK and JNK activations in SK-MEL-28

cells. ............................................................................................... 149

Chapter IV. General discussion ................................................................... 151
1.1. Experimental tools to study MEK and MK fimctions ..................... 152
1.2. Non-redundant roles of MEK] and MEK2 ........................................ 154
1.3. Complexity of the MEK/MKK signaling network ............................ 156
1.4. Therapeutic implications .................................................................... 159

Appendix I. Transcriptional signatures that are down-regulated by LeTx
treatment and significantly rescued by MEK2cr ............................ 163

Appendix II. Transcriptional signatures that are affected by LeTx treatment
and significantly rescued by both MEK] er and MEK2cr ............. 169

Appendix III. Transcriptional signatures that are not affected by LF treatment
but cannot be rescued by either MEK] or or MEK2cr ................. 182

Appendix IV. Transcriptional signatures that are not affected by LF treatment
but affected by VS-MEKI or expression ...................................... 190

Appendix V. Transcriptional signatures that are not affected by LF treatment
but affected by V5-MBK2cr expression ....................................... 193

viii

Appendix VI. Transcriptional signatures that are not affected by LF treatment
but affected by both V5-MEK1cr and V5—MEK2cr expressions.196

Literature Cited .................................................................................................... 201

ix

List of Tables

Table I. Small-molecule ICso values .................................................................... 44

Table II. Alignment of MEK/MKK amino acid sequences flanking the LF
cleavage sites. .................................................................................... 45

Table III. The Clark model of melanoma progression ......................................... 46

Table IV. Percentage of G1 population of MEK siRNA-treated SK-MEL-28
cells. ................................................................................................. 110

Table V. MEK2cr-rescued transcriptional Signatures that are significantly
affected by LeTx treatment .............................................................. 1 11

List of Figures

Images in this thesis/dissertation are presented in color.

Figure 1.
Figure 2.
Figure 3.
Figure 4.
Figure 5.
Figure 6.
Figure 7.
Figure 8.

Figure 9.

Figure 10.
Figure 11.

Figure 12.

Figure 13.

Figure 14.

Figure 15.

Figure 16.

Figure 17.

Figure 18.

Figure 19.

Figure 20.

Schematic illustration of the Raf-MEK-ERK signaling pathway. ....... 47

Structures of MEK inhibitors. .............................................................. 49
Effect of PD 1843 52 on MEK activation in cells. ................................ 51
Resistance of VS-MEKcr to LF-mediated proteolysis. ........................ 66
Dual translation of VS-MEK expression vectors. ................................ 68
Activity of V5-MEKcr. ........................................................................ 69

Resistance of MKK3cr and MKK6cr to LF-mediated proteolysis ....... 71

Resistance of MKK4cr and MKK7cr to LF-mediated proteolysis. ...... 73

In-cell cleavage of MKK4 by LF. ........................................................ 74
Wild-type MKK7 is not cleaved by LP in cells. ................................ 75
LF does not cleave endogenous MKK7 in mammalian cells. ............ 76
Necessity of MEK] and MEK2 signaling pathways for ERK

activation in SK-MEL-28 cells. ....................................................... 112
Individual MEK signaling in LeTx-treated SK-MEL-28 cells ......... 114
Overlapping and non-overlapping transcriptional targets

downstream Of MEK] and MEK2. .................................................. 116
Expression levels of V5 fusion proteins in SK—MEL-28 cells. ........ 118
Inhibitory effect of LeTx on proliferation of SK-MEL-28 stable
clones. .............................................................................................. 119
Sensitivities of SK-MEL-28 cells to LeTx and MEK inhibitors. ..... 121

LeTx treatment inhibits p38 and JNK MAP kinase activation in
SK-MEL-28 cells. ............................................................................ 123

Sensitivity of V5-MKK3cr- and V5-MKK6cr-expressing cells to
LeTx ................................................................................................. 124

Sufficiency of MEK2 signaling pathway for anchorage-
independent growth of SK-MEL-28 cells ........................................ 126

xi

Figure 21.

Figure 22.

Figure 23.

Figure 24.

Figure 25.

Figure 26.

Figure 27.

Figure 28.

SK-MEL-28 xenograft tumor growth and systemic treatment

with LeTx ......................................................................................... 128
Reproducibility of SK—MEL-28 xenograft tumor growth in
athyrnic nude mice. .......................................................................... 130
Sensitivity of VS-MEchr-expressing xenografi tumors to LeTx.
......................................................................................................... 131
Sensitivity Of V5-MEK2cr-expressing xenograft tumors to LeTx.
......................................................................................................... 133
Does-dependent effect of LeTx on SK-MEL-28 xenograft tumor
growth. ............................................................................................. 135
Loss of V5-MEKcr expression in SK-MEL-28 xenograft tumors.
......................................................................................................... 137
Systemic treatment of LeTx does not cause MEK cleavage in
SK-MEL-28 xenografl tumor cells. ................................................. 139
Unsupervised clustering of gene expression changes in LeTx-
treated cells. ..................................................................................... 141

xii

List of Abbreviations

B-Raf, v—rafmurine sarcoma viral oncogene homolog Bl
CDKN2A, cyclin-dependent kinase inhibitor 2A

cDNA, complementary DNA

CHO K1, Chinese hamster ovary K1

DMSO, dimethyl sulfoxide

DNA, deoxyribonucleic acid

EDTA, ethylenediaminetetraacetic acid

EGF, epidermal growth factor

EGTA, ethylene g1ycol-bis(2-aminoethylether)-N.N,N'. '-tetraacetic acid
ERK, extracellular signal-regulated kinase

GST, glutathione S-transferase

JN K, c-Jun N—terminal kinase

LeTx, lethal toxin (anthrax lethal toxin)

LF, lethal factor (anthrax lethal factor

MAP, mitogen-activated protein

MAPK, mitogen-activated protein kinase

MAPKK, mitogen-activated protein kinase kinase

MEK, mitogen-activated protein kinase/extracellular signal-regulated kinase kinase
MEKcr, cleavage-resistant form of MEK

MKK, mitogen—activated protein kinase kinase

mRN A, messenger RNA

PA, protective antigen

PMA, phorbol 12-myristate 13-acetate

PTEN, phosphatase and Lensin homolog

xiii

Ras, rat sarcoma viral oncogene homolog
RNA, ribonucleic acid

SDS, sodium dodecyl sulfate

shRNA, small hairpin RNA

siRNA, small interfering RNA

SV40, simian virus 40

TPA, 12-O-tetradecanoylphorbol-1 3-acetate
UV, ultraviolet

YopJ, Yersz'nia outer protein J

xiv

Chapter I. General introduction: MEK signaling pathway and cancer

1.1. Introduction to MEK signaling pathway
1.1.1. Exploring the Raf-MEK-ERK signaling pathway

In late 19803, Ray and Sturgill (1987; 1988a) observed that soluble fi'actions of
cell extracts prepared from insulin-treated 3T3 —L1 adipocytes were capable of
phosphorylating the microtubule—associated protein 2 (MAP-2) protein. Initially this
protein kinase was termed microtubule-associated protein (MAP) kinase, as it catalyzed
the phosphorylation of MAP-2 on serine and threonine residues (Sturgill & Ray, 1986).
Later, MAP kinase activity was shown to be stimulated not only by insulin but also by
other mitogenic factors such as serum, insulin-like growth factors, epidermal growth
factor (EGF), and phorbol 12—myristate 13-acetate (PMA; also called TPA) (Rossomando
et al. , 1989; Erickson et al., 1990). Therefore, this enzyme was also named mitogen-
activated protein (MAP) kinase (Rossomando et al., 1989) or extracellular signal-
regulated kinase (ERK) (Boulton et al., 1990). Since growth factor—stimulated MAP-2
phosphorylation was accompanied by DNA synthesis and cell cycle progression (Sato et
al., 1985; Sato et al., 1986), Ray and Sturgill’s seminal discovery Opened the way to
exploring the MAP kinase signal transduction pathway through which these cellular
events were regulated.

Ray and Sturgill (1988b) also showed that ERK was phosphorylated at threonine
and tyrosine residues upon stimulation, and serine/threonine phosphatase as well as
tyrosine phosphatase could completely deactivate ERK (Anderson et al., 1990). These

findings indicated that ERK itself was a phosphoprotein and that activation of ERK

involved phosphorylation by another protein kinase having threonine and tyrosine
specificity. Soon after, the activator for ERK was identified and purified from EGF-
stimulated cells by Ahn (1991) and colleagues (Seger et al., 1992). This dual-specificity
protein kinase (both serine/threonine and tyrosine specificity) (Gomez & Cohen, 1991;
Nakielny et al., 1992; Seger et al., 1992) was called MAP kinase kinase (MAPKK,
MAP2K or MKK) (Gomez & Cohen, 1991) or MAPK/ERK kinase (MEK) (Crews &
Erikson, 1992). Similarly, MEK was characterized as a phosphoprotein that required
serine/threonine phosphorylation for its own kinase activity (Gomez & Cohen, 1991).
Serine/threonine protein phosphatase 2 (PP2A) —- but not tyrosine phosphatases -— was
shown to inactivate MEK activity (Gomez & Cohen, 1991; Nakielny et al., 1992). This
indicated that downstream of the growth factor receptors (which were tyrosine kinases)
there must be at least one serine/threonine protein kinase that functioned as the MEK
activator. Subsequent to this, the Raf family of protein kinases, cellular homologues of
the acutely transforming viral oncogene v-raf(Rapp et al., 1983), were identified as
direct activators of MEK (Dent et al., 1992; Kyriakis et al., 1992). c-Mos has also been
identified as a kinase that activates MEK, though its normal expression is limited to

oocytes (Sagata et al., 1988).

1.1.2. Activation mechanism of the MAP kinase cascade

MAP kinase pathways are evolutionarily conserved signaling pathways that
transduce signals from the cell membrane to the nucleus in eukaryotic cells (reviewed by
English et al., 1999; Chen et al., 2001; Pearson et al., 2001; Johnson & Lapadat, 2002;

Roux & Blenis, 2004). Signal transduction through the Raf-MEK-ERK pathway (Figure

1) begins with the binding of extracellular growth factors to specific transmernbrane
receptors. This results in dimerization and/or conformational changes in the receptors,
leading to a series of cytoplasmic protein recruitment/activation steps that include the Raf
family of proteins. Raf then activates MEK] and MEK2, the only known MEK isoforms,
by phosphorylating them on the conserved serine residues in the activation loop
(Ser218/Ser222 of human MEKl and Ser222/Ser226 of human MEK2) (Alessi et al.,
1994; Zheng & Guan, 1994). Active MEK] and MEK2 then in turn activate ERK1 and
ERK2, the only known substrates of MEK, by phosphorylating the threonine and tyrosine
residues in the conserved T-E-Y motif (Thr202/Tyr204 of human ERK] and
Thr185/Tyr187 of human ERK2) (Payne et al., 1991). After activation, ERK
phosphorylates and activates downstream effectors regulating a variety of cellular events
at the transcriptional and posttranslational levels (reviewed by English et al., 1999; Chen
et al., 2001; Pearson et al., 2001; Johnson & Lapadat, 2002; Roux & Blenis, 2004).

Why does the MAP kinase pathway employ three tiers of protein kinases to
transduce signals instead of one? To simulate MAPK phosphorylation and activation,
Huang and Ferrell (1996) constructed an elegant mathematical model and then tested
predictions Of this model in cytoplasmic extracts oernopus oocytes. They found that
whereas Raf behaved as a typical Michaelis-Menten enzyme, MEK and ERK behaved as
“ultrasensitive” enzymes. In other words, MEK and ERK are relatively less sensitive to
low concentrations of stimuli but above a threshold level stimuli evoke a strikingly rapid
response. This makes MEK and ERK respond to stimuli in an all-or-none fashiOn, like a
switch. Moreover, according tO their model, the ultrasensitivity of this reaction is directly

linked to the dual phosphorylation that is required to activate MEK and then ERK. This

observation has profound functional implications and Should be an important

consideration in designing strategies to inhibit this pathway.

1.2. Involvement of MEK signaling pathway in human cancers

The Raf-MEK-ERK pathway is an important signaling pathway that, in response
to growth factor signals, promotes cell cycle progression and cell proliferation (reviewed
by English et al., 1999; Chen et al., 2001; Pearson et al., 2001; Johnson & Lapadat, 2002;
Roux & Blenis, 2004). Therefore, it is not surprising that deregulated activities of the
kinases in this pathway disrupt growth control, which can lead to cancers. The
hypothesis that elevated activity of the Raf-MEK-ERK signaling pathway may lead to
cancer was tested more than ten years ago when constitutively active MEK-ERK
signaling was shown to be sufficient to transform mammalian cells (Mansour et al.,
1994a). In support of this, screening of somatic mutations in a panel of human cancer
cell lines identified the presence of the constitutively active B-Raf mutant V600E, which
results in a persistent activation of the downstream pathway, in 66% of human
melanomas and in a lower percentage of other types of cancers (Davies et al., 2002).
Based on these findings, targeting the Raf-MEK-ERK signaling pathway became a

prospective strategy for treating human cancers.

1.3. Targeting the Raf-MEK-ERK pathway as a strategy for treating cancers
In principle inhibitors can be designed that will block any step in this pathway
(reviewed by Kohno & Pouyssegur, 2003; Sebolt-Leopold & Herrera, 2004; Kohno &

Pouyssegur, 2006). However, because this is an amplifying signal cascade, it makes

most sense to block that signal as close to the stimulus (receptor tyrosine kinase) as
possible. This is particularly true in this case as MEK and ERK respond to stimuli in an
all-or-none fashion. Therefore, it might be argued that Raf kinases are the preferred point
of intervention. However, in an unexpected twist several investigators now report that B-
Raf inhibitors paradoxically can activate MEK-ERK signaling in cells that express wild-
type B-Raf (Hatzivassiliou et al., 2010; Heidorn et al., 2010; Poulikakos et al., 2010). It
appears that the binding of ATP-competitive B-Raf inhibitors Causes conformational
changes in the kinase that promotes dirnerization and transactivation of non—drug-bound
C-Raf in an active Ras-dependent fashion. So, while it makes sense to target tumors
harboring B-Raf mutations with B-Raf inhibitors, it may actually be disadvantageous to
use these drugs to treat tumors with wild-type B-Raf Since they may enhance
proliferation of tumor cells and may have adverse effects on normal cells with active Ras.
In this case MEK inhibitors may be the most appropriate therapeutic option.

A great deal of effort has gone into looking for highly selective MEK inhibitors.
Many Of these are commonly used in laboratories as powerful tools in the study of the
MEK-ERK signaling pathway, and select inhibitors are currently in cancer clinical trials,
having shown promising anti-cancer activity in preclinical studies. In the following
sections, 1 will review the discoveries, properties, and clinical applications of two
categories of highly selective MEK inhibitors: non—ATP-competitive small-molecule
inhibitors and biological inhibitors. By focusing on their MEK selectivity and inhibition
mechanisms, the following review will provide insights into the potential of these

inhibitors as tools for studying the Raf-MEK-ERK pathway and as anti-tumor agents.

1.3.1. Small-molecule inhibitors

In this section, the best known and commonly used non-ATP-competitive MEK
inhibitors will be reviewed. These inhibitors result in MEK-specific inhibition by an
allosteric mechanism which contributes to highly selective inhibition of MEK without

affecting other protein kinases that have structurally similar ATP binding pockets.

1.3.1.1. PD 098059
By screening a compound library using an in vitro kinase cascade
phosphorylation assay, Dudley et al. (1995) identified PD 098059, 2-(2'-aminO-3'-

methoxyphenyl)-oxanaphthalen-4-one (Figure 2a), as a synthetic inhibitor of the MAPK

pathway. PD 098059 inhibits MEK (IC50 = 10 uM, see Table I) but not other

components of the MAPK pathway. In later studies, PD 098059 showed no significant
effect on a spectrum of other protein kinases, including protein kinase A (PKA), protein

kinase C (PKC), and other kinase members in the MAPK family pathways such as p38

MAP kinase, c-Jun NHz-terminal kinase (JN K), p38 MAP kinase kinase, and MKK4

(Alessi et al., 1995; Davies et al., 2000). Kamakura et al. (1999) showed that PD 098059
had no effect on MEKS (a MEK/MKK family member which is independent from the
MEKl/Z-ERKl/Z pathway) in vitro, but did have an inhibition effect on intracellular
MEKS activity. This was later confirmed by Mody et al. (2001 ), although a high
concentration (100 uM) of PD 098059 was required. PD 098059 inhibits growth factor—
stimulated MAPK activation and DNA synthesis in Swiss 3T3 cells (Dudley et al., 1995),
suggesting the membrane permeability and in-cell inhibitory activity of PD 098059.

Initial observation showed that PD 098059 did not inhibit MEK] that had been

maximally phosphorylated by c-Raf (Alessi et al., 1995; Ahn et al., 2001), raising a
possibility that PD 098059 did not bind to the active site of MEK but instead bound to
another allosteric pocket. The allosteric inhibition mechanism was later confirmed by

using classical Michaelis-Menten methods (Favata etal., 1998).

1.3.1.2. U0126

U0126, 1 ,4-diamino-2,3-dicyano-1,4-bis(2-aminophenylthio)butadiene (Figure
2b), was the second compound identified as an allosteric inhibitor of MEK. Favata et al.
(1998) screened a total of 40,000 compounds for their ability to inhibit activator protein 1
(AP-l)—driven transcription using a luciferase reporter assay. They reported that U0126
inhibited both AP-l—driven gene expression and TPA-induced up-regulation of c-Jun and
c-Fos at the protein and the mRNA levels, suggesting that U0126 targeted the upstream
activators of c—Jun and c-Fos. Indeed, U0126 inhibits MEK (see Table I), but not other
MKK members and MAP kinases (Davies et al., 2000) except for MEKS (Kamakura et
al., 1999; Mody et al., 2001). In addition, U0126 showed a 100-fold greater potency than
PD 098059 (Favata et al., 1998). This was later confirmed by Davies et al. (2000).
Consistent with PD 098059 studies, U0126 has a much lower effect on activated wild-
type MEK than constitutively active MEK (Alessi et al., 1995; Ahn et al., 2001),
suggesting that the conformational change in MEK upon activation interferes with the
binding of these inhibitors. In addition, this implied that these two MEK inhibitors might
not compete with the ATP binding pocket of MEK. Using classical Michaelis-Menten

enzyme modeling, Favata et al. (1998), confirmed the non—ATP-competitive mechanism

of these MEK inhibitors the structural basis of which will be discussed in detail later in
this section.

The results of in vitro kinase assays using purified proteins suggested that PD
098059 and U0126 inhibited the phosphorylation and activation of MEK but not the
ability Of MEK to phosphorylate ERK (Alessi et al., 1995; Ahn et al. , 2001). However,
in later studies, the inhibitory effect of PD 098059 and U0126 on MEK phosphorylation
was not Observed in intact cells when using antibodies specifically against phosphO-MEK
(Ahn et al., 2001). A possible explanation for this controversy is that the action of these
two inhibitors in intact cells is mediated by inhibition Of MEK phosphoryl-transferase
activity, and that both inhibitors may interfere with Raf-dependent MEK phosphorylation
in vitro, only when other intracellular components are absent. This controversy was also
pointed out and discussed by Kohno and Pouyssegur (2003).

PD 098059 and U0126 are widely used in laboratories as powerful tools for
studying the MEK signaling pathway. Both molecules are commercially available and
have thousands of entries in the PubMed database. However, neither of them has entered

clinical trials due to their limited in viva bioactivity.

1.3.1.3. PD 184352 (CI-1040)

PD 1843 52 (Figure 2c), also known as CI—1040, was identified by Sebolt—Leopold
et al. (1999) as a potent MEK inhibitor (IC50 = 17 nM; see Table I) in an in vitro kinase
cascade assay using GST-fusion proteins of MEKl and MAPK. PD 184352 has a higher

potency than PD 098059 and U0126 (Table I), and has no effect on other kinases tested

(Sebolt-Leopold et al., 1999; Davies et al., 2000). Similar to PD 098059 and U0126, PD

1843 52 inhibits intracellular MEKS activity at a much higher concentration than that
required to inhibit the activation of ERK (Mody et al., 2001). PD 1843 52 is not
competitive with ATP or ERK (Sebolt-Leopold et al., 1999), suggesting a similar
allosteric inhibition mechanism for PD 184352 as for PD 098059 and U0126. The

cellular consequences of MEK inhibition by PD 1843 52 include a complete suppression

of platelet-derived growth factor (PDGF)-induced ERK activation, induction of G1 cell

cycle arrest, reversion to untransformed morphology, and inhibition of anchorage-
independent growth in colon cancer cells as well as in other cancer cell lines (Table I)
(Sebolt-Leopold et al., 1999). Interestingly, PD 1843 52 showed a stronger effect on
three colon cancer cell lines that had high ERK activity relative to cancer cell lines with
low ERK activity, suggesting that the inhibitor is more potent for cancer cells that are
addicted to MEK signaling. This was further confirmed by Solit et al. (2006), who
showed that the B-Raf V600E oncogenic mutation in cancer cell lines was associated
with enhanced and selective sensitivity to PD 184352 when compared with B-Raf wild-
type cell lines. In addition to the in vitro effects, Sebolt-Leopold et al. (1999)
successfully showed that PD 1843 52 actively suppressed ERK activation in colon cancer
cell xenografi tumors in mice within one hour after oral administration. More
importantly, this in vivo efficacy was achieved over a wide range of doses (48—200 mg/kg,
two treatments per day for 14 days) without signs of toxicity. This was the first study
demonstrating the in vivo efficacy of MEK inhibition by a small-molecule inhibitor.

A phase I study of PD 184352 was conducted to test the toxicity,
pharmacokinetics, pharmacodynamics, maximum tolerated dose, food effect, and clinical

activity in patients with advanced cancers, including colorectal, non-small cell lung,

pancreatic, and kidney cancers, as well as melanoma (Allen et al., 2003; LoRusso et al.,
2005b). PD 1843 52 was administrated orally in single to multiple daily doses from 100
mg to 1,600 mg for 21 days, followed by 7 days off, for a total of three cycles. After the
maximum tolerated dose profile was determined, the “holiday” week between treatment
cycles was removed to test the safety profile of continuous dosing. Blood and urine
samples were collected throughout the treatment for pharrnacokinetics and biomarker
assessments. Phospho-ERK levels were assessed in blood samples and tumor'biopsies to
determine the clinical targeting activity of PD 1843 52. The study results showed that PD
184352 was well tolerated at a treatment of 800 mg twice daily, with common grade 1 or
2 drug-related adverse events including diarrhea, asthenia, rash, nausea and vomiting.
Although this was a phase I study initially designed only to assess toxicity, some patients
showed promising signs of efficacy: one out of 66 patients had partial response lasting
355 days, and 19 patients (29%) achieved stable disease for 4-17 months with a median
of 5.5 months.

Based on these results, a phase II trial was initiated to assess the antiturnor
activity and safety of PD 1843 52 in patients with advanced breast, colon, non-small-cell
lung, and pancreatic cancers (Rinehart et al., 2004). In this study, PD 1843 52 was given
to patients at 800 mg twice daily and continuously, and was reduced later in ZOO-mg
decrements for patients with grade 2 or greater adverse event severity. Continuous PD
184352 treatment exceeding one year was well tolerated without cumulative toxicities,
and 81% of the patients experienced only grade 1 or 2 adverse events. However, PD

1843 52 was terminated in the phase II trial because no complete or partial responses were

10

observed, although eight patients (12%) achieved stable diseases lasting 4-18 months
with a median of 4.4 months.

The change in ERK activity in the tumors after PD 184352 treatment was not
available from the phase II study as the study was not designed to assess it in tumor
biopsies. However, in the earlier phase I study, ERK activation was inhibited by 46% to
100%, with an average of 73% in tumor biopsies from ten out of ten (100%) patients
(LoRusso et al., 2005b). This suggests that the lack of response to PD 184352 in the
phase II study might not be solely due to incomplete inhibition of ERK activity in the
tumors. In addition, the authors of this study also attributed the lack of response to the
poor metabolic stability of PD 184352. PD 0325901, a PD 184352 analog, was then
developed as a second-generation MEK non—ATP-competitive inhibitor with improved
potency and pharmaceutical properties, and it has Since entered clinical trials (discussed

immediately below).

1.3.1.4. Derivatives of PD 184352

PD 0325901 (Figure 2d), an analog of PD 184352, is also a non—ATP-competitive
inhibitor and possesses high selectivity for MEK (Sebolt-Leopold et al., 2004). PD
0325901 has a 500-fold higher cellular effect on ERK activation than does PD 1843 52
(see Table I). It is believed that this improvement is the result of several factors,
including better solubility and higher metabolic stability. Similar to PD 1843 52, PD
0325901 has higher potency for xenograft tumors harboring the B-Raf V6OOE mutation
than for tumors with wild-type B-Raf (Solit et al., 2006). In recent studies, PD 0325901

showed in vivo potency for oncogenic B-Raf-induced lung adenocarcinoma and

11

melanoma (Dankort et al., 2007; Dankort et al., 2009a). More importantly, in preclinical
studies, PD 0325901 suppressed ERK phosphorylation by more than 50% at a single oral
dose of 25 mg/kg at 24 hours after treatment. In comparison, PD 1843 52 suppressed
ERK phosphorylation at a much higher dose of 150 mg/kg for only eight hours, and the
phosphO-ERK level returned to the control level within 24 hours after treatment (Sebolt-
Leopold et al., 2004).

The improved biological profile and promising preclinical activity of PD 03 25901
moved it into a phase 1-2 clinical study Of patients with advanced cancers, including
breast, colon, and non-small-cell lung cancers as well as melanoma (LoRusso et al.,
2005a; Menon et al., 2005). Based on preclinical studies, cycles of PD 0325901 were
administrated orally from 1 mg once daily to 1, 2, 4, 8, 15, 20 or 30 mg twice daily for 21
days and then repeated every four weeks. The study was designed to focus on
pharmacokinetics, pharmacodynamics, and effectiveness. For this purpose, before and
during treatment, tumors were biopsied to assess ERK activation and proliferation index,
and blood samples were collected to measure the amount of circulating drug and its
metabolite. The results showed that PD 0325901 was well tolerated. In addition to
common toxicities (including rash, diarrhea, nausea, and vomiting), some patients had
visual effects including blurred vision and halos. Strong inhibitory effects of PD
0325901 treatment on phospho-ERK and proliferation marker Ki67 were observed at a
minimum dose of 2 mg twice daily: ERK phosphorylation was consistently suppressed by
an average of 84%, while cell proliferation (measured by Ki67 irnmunostaining) declined
by an average of 60%. Although one partial response (melanoma) and five stable disease

cases (four melanoma and one non-small-cell lung carcinoma) were Observed, this study

12

was terminated prematurely in 2007 mainly due to the safety concerns about the ocular
and neurological toxicity that presented at 10 mg twice daily and higher
(ClinicalTrials. gov Identifier: NCT00147550).

As both PD 1843 52 and PD 0325901 have failed in clinical trials, Pfizer recently
has developed and investigated a series of PD 1843 52 derivatives, among which one
compound has showed relatively superior ERK inhibition and better solubility then PD
1843 52 [compound 12a reported by Warmus et al. (2008)]. More importantly, the

metabolic stability of this compound is much better than that of PD 184352.

1.3.1.5. ARRY-142886 (AZD6244)

ARRY-142886, or 6-(4-bromO-2-chloro-phenylamino)-7-fluorO-3 -methyl-3H-
benzoimidazole-S-carboxylic acid (2-hydroxy-ethoxy)-amide (Figure 2g), also called
AZD6244, was disclosed in 2004 as another second-generation MEK small molecular
inhibitor (Wallace et al., 2004). Like PD 0325901, ARRY-142886 was developed based
on the PD 1843 52 structure, and therefore was highly selective for MEK and

noncompetitive with ATP (Lyssikatos et al., 2004; Yeh et al., 2007). In an in vitro assay,
ARRY-l42886 inhibited the activity of purified MEK with an IC50 of 14.1 nM (Table I),
but had no effect on more than 40 other kinases and a minimal inhibitory effect on MEKS
(Y eh et al., 2007). Similar to other MEK noncompetitive inhibitors, ARRY-l42886
inhibits MEK activity but not the phosphorylation and activation of MEK in intact cells

(Y eh et al., 2004; Yeh et al., 2007). ARRY-142886 has been shown to inhibit

intracellular phosphorylation of ERKl/2 in several cancer cell lines including melanoma,

colon, and pancreatic cancer lines, with an IC50 below 40 nM (Table I), but has no effect

13

on MEKS (Y eh et al., 2007; Haass et al., 2008). Cellular responses to ARRY-142886

include an inhibition in proliferation and cell cycle arrest at the G1/ S transition (Y eh et al.,

2004; Yeh et al., 2007; Haass etal., 2008), whereas prolonged treatment (72 h) results in
only limited apoptosis in some cell lines (Haass et al., 2008). The cytostatic inhibitory
effect of ARRY-l42886 was also observed in a three-dimensional culture model of
melanoma (Haass et al., 2008).

Like an earlier study of PD 184352 (Solit et al., 2006), ARRY-142886 had a
much stronger inhibitory effect on cancer cells harboring Ras or Raf oncogenic mutations
than on those without the mutations or on normal fibroblast cells (Y eh et al., 2004; Yeh
et al., 2007; Haass et al., 2008). Moreover, ARRY-142886 is active when given orally
and shows in vivo efficacy. In mouse xenografi models, ARRY-l42886 not only inhibits
the growth of colon cancer and melanoma in a dose-dependent manner, but also induces
pancreatic xenografi tumor regression (Lee et al., 2004; Yeh et al., 2007; Haass et al.,
2008). Although ARRY- 1 42886 effectively inhibits melanoma xenograft tumor growth,
it does not induce apoptosis in the tumor cells unless combined with microtubule-
stabilizing agents (Haass et al., 2008). Interestingly, xenograft tumor growth recovers
after cessation Of ARRY-l42 886 administration, and the tumors regress again upon re-
administration (Y eh et al., 2007). This supports the idea that the effect of ARRY-142886
is cytostatic as Opposed to cytotoxic and that continuous treatment may be necessary to
maintain the inhibitory effect.

With its promising in viva antiturnor activity, ARRY-142886 entered phase I
clinical trial in 2004, which was initiated to assess the maximum-tolerated dose,

pharmacokinetics, and pharmacodynamics in patients with advanced cancers, including

14

breast cancer, colorectal cancer, and melanoma (Adjei et al., 2008). This study had two
parts. Part A was to determine the maximum-tolerated dose. Patients were orally given
ARRY-142886 at doses Of 50, 100, 200, or 300 mg twice daily for 28-day cycles.
Although the maximum-tolerated dose Obtained from part A was 200 mg twice daily, this
dose was discontinued in the second part of the study (Part B) because a substantial
number of patients in Part A required dose reductions, treatment holidays, or even
termination due to drug-related toxicities including rash, diarrhea, nausea, fatigue, and
blurred vision. Therefore, a dose of 100 mg twice daily was recommended for the Part B
study. The results showed that ARRY-l428 86 was well tolerated at 100 mg twice daily
with grade 1 or 2 adverse events. Promising targeting effects of ARRY-142886 were
observed in post-treatment tumor samples: ERK phosphorylation was strongly inhibited
by an average of 79%, while greater than 50% inhibition of cell proliferation (measured
by Ki67 irnmunostaining) was seen in 25% of the patients. One patient enrolled in this
study with malignant melanoma had a 70% reduction of tumor size after three cycles of
treatment but eventually developed brain metastases. Although no obj ective response
was achieved in the phase I study, nine patients (16%) had stable disease for more than
five months. Based on the promising targeting effect and recommended dosage Obtained
from the phase I study, ARRY-l42886 is now in several phase II studies, including

studies involving combination therapies.

1.3.2. Mechanism of allosteric inhibition
Initially both PD 098059 and U0126 were found to have a stronger effect on less

active MEK than highly active MEK (Alessi et al., 1995 ; Favata et al., 1998), suggesting

15

 

that these kinase inhibitors do not bind to the enzyme active site of MEK. In addition,
Favata et al. (1998) also confirmed the non—ATP-competitive mechanism of these MEK
inhibitors by using classical Michaelis-Menten methods. While these inhibitors do not
compete with ATP or ERK, they do compete with each other, indicating that the
inhibitors share a common binding site independent of the ATP or ERK binding sites
(Favata et al., 1998). Ohren et al. (2004) identified the binding pocket of non—ATP-
competitive MEK inhibitors by co-crystallizing MEK with MgATP and two PD 184352
derivatives [PD 318088 (Figure 2e) for MEKl and PD 334581 (Figure 2f) for MEK2] as
ternary complexes with satisfactory resolutions. The crystal structures indicate that the
MEK inhibitors bind to MEK in a pocket adjacent to the MgATP binding Site. The
inhibitor binding pocket and the MgATP binding pocket are physically separated by the
side chains Of Lys97 and Met143, which are conserved in the MEK active site. This was
the first structural study directly revealing the allosteric binding pocket of these MEK
inhibitors.

Ohren et al. (2004) also showed that the diaryl amine (A ring and B ring, Figure
2c-g) structure of PD 184352 and its derivatives — including PD 0325901, PD 318088,
PD 334581, and ARRY-142886 —— was the key to stabilizing the inhibitors in the
allosteric binding site. The A ring and the B ring form a crucial hydrogen-bond
interaction and numerous non—covalent interactions, respectively, with the hydrophobic
pocket formed by several residues. The binding of the inhibitors results in a series of
conformational changes in MEK, leading to interference with the catalytic residue Lys97
(MEKl) and thus inactivation of MEK activity. In other words, upon inhibitor binding,

MEK adopts a “closed” conformation of an active protein kinase, and the phosphoryl-

16

transferase activity of MEK is inhibited. This is supported by a fact that PD 1843 52
treatment inhibits ERK activation in human melanoma SK-MEL-28 cells without
affecting the activation of MEKl (Figure 3). The structures of MEKl and MEK2, co-
crystallized with MgATP plus the inhibitor, share high homology (Ohren et al., 2004),
indicating that the same mechanism of inhibition is involved in both. Most recently, it
was shown that U0126 binds to MEK in the same pocket as PD 0325901 (Fischmann et
al., 2009). Since the two earliest non—ATP-competitive MEK inhibitors, PD 098059 and
U0126, compete with each other (Favata et al., 1998), it is very likely that all of the MEK
inhibitors discussed above bind to the same pocket and use the same mechanism to
inhibit MEK.

The high selectivity of these allosteric inhibitors for MEK can be explained on the
basis of structure. First, these inhibitors do not compete with ATP, the major
intracellular energy source for all protein kinases; this decreases the binding affinity of
the inhibitors to other kinases. Second, these inhibitors bind to MEK in a unique binding
pocket, which is formed by several residues of MEK and stabilizes the inhibitor binding.
This pocket increases the MEK:inhibitor binding affinity as well as the complexity of the
binding, resulting in a high selectivity for MEK. Since the diaryl amines of these MEK
inhibitors are such a crucial structure for this MEK-specific inhibition mechanism, all the
second-generation MEK allosteric inhibitors have been developed based on this structure

in order to maintain the mechanism (W armus et al., 2008).

17

1.3.3. Biological inhibitors
1.3.3.1. Anthrax lethal toxin

The MEK-specific inhibitory activity of anthrax lethal toxin (LeTx) was revealed
in the National Cancer Institute Antineoplastic Drug Screen project (Duesbery et al.,
1998). In this study, sixty human cancer cell lines (termed NCI-60 cell lines) were tested
for their sensitivity to a panel of anti-neoplastic molecules. It was shown that the
sensitivity profile of the MEK inhibitor PD 098059 was very similar to that of LeTx,
raising the possibility that they sensitized the cells by a similar mechanism. Based on this,
Duesbery et al. (1998) characterized the MEK-specific protease activity of LeTx. About
the same time, Vitale et al. independently demonstrated the same proteolysis of MEK by
LeTx (V itale et al., 1998).

Anthrax lethal toxin is a binary toxin secreted by the bacterium Bacillus anthracis.
It is composed of a binding moiety called protective antigen (PA) and an enzymatic
moiety, lethal factor (LF). During cellular intoxication, PA binds to the anthrax receptors
expressed on the host cell surface and then is cleaved by a furin-like protease. This

cleavage removes a 20-kDa N-terminal fragment and leaves a 63-kDa C-terminal

truncation, PA63, on the cell surface. PA63 then forms an Oligomerized channel to which

the LF binds. After the binding of LF, the complex is internalized into cells through the

endosomal internalization pathway. The acidic environment of the endosome causes a

conformational change in PA63 that results in the formation of pores through which LP is

released into cytosol [reviewed by Singh et al. (2005)].
Since its initial identification as a MEK1/2-specific protease (Duesbery et al.,

1998), additional members of the MAP kinase kinase family, including MKK3, 4, 6 and 7

l8

but not MEKS, have also been found to be substrates of LF [(Vitale et al., 1998;
Pellizzari et al., 1999; Vitale et al., 2000) and discussed by Duesbery et al. (2001)].
However, unexpected results from this dissertation research suggested that MKK7 might
not be a preferred LF substrate. This will be discussed in Chapter 11. LF specifically
cleaves the N-terminus Of MEK and other MKK proteins at the consensus cleavage site:

three to four basic or proline residues followed by three variable residues followed by an

aliphatic residue: (B/P)3_4-(X)3-Al (T able II). The N-termini of this family Of proteins

harbor MAP kinase docking sites that are required for the interaction between
MEK/MKK and MAPK (Tanoue et al., 2000). Not surprisingly, after LF cleavage, the
C-terminal part of MEK/MKK, which contains the kinase domain, loses its affinity for
the downstream MAP kinases (Chopra et al., 2003; Bardwell et al., 2004). In addition,
biochemical evidence has Shown that loss of the amino terminus may destabilize MEK,
leading to the decreased intrinsic kinase activity that is Observed following LF-mediated
proteolysis (Chopra et al., 2003). Thus, cleavage Of MEK/MKK by LF results in a
blockade of not only the ERK pathway but also of the p38 MAPK and JNK pathways.
Because increased MAP kinase activity has been detected in many types of
human cancers, LeTx has potential as a cancer therapeutic. The first evidence for the
anti-tumor activity of LeTx was provided by Friedlander et al. (1998). Duesbery et al.
(2001) then assessed its effects on cancer cells and found that LeTx not only efficiently
inhibited the proliferation and the soft agar colony formation of oncogenic Ras-
transformed NIH-3T 3 cells, but it also caused the cells to revert to a nontransformed
morphology. LeTx treatment also inhibited the growth of Ras-transformed cells

implanted in nude mice, with no overt toxicity. Interestingly, LeTx-treated xenografr

l9

tumors had a pale appearance whereas control tumors were dark red-purple. This study
demonstrated the in viva antitumor activity of LeTx and also suggested that LeTx could
inhibit tumor growth by targeting cell proliferation pathways as well as angiogenesis
pathways.

This finding has been confirmed in several tumor models including melanoma
(KOO et al., 2002; Abi-Habib et al., 2005), Kaposi's sarcoma (Depeille et al., 2007),
fibrosarcoma (Ding et al., 2008), kidney cancer (Huang et al., 2008), neuroblastoma, and
colorectal adenocarcinoma(Rou1eau et al., 2008) cells in vitro, as well as in xenograft
tumor growth in viva. In the earliest xenografi studies, LeTx was not administrated
systemically, but intratumorally. In these studies it was noted that intraturnoral injection
of LeTx had a growth inhibitory effect on distal, uninjected tumors (Duesbery et al., 2001;
Koo et al., 2002). These observations revealed the systemic anticancer potential of LeTx.
Abi-Habib et al.’ (2006a) then determined the in viva potency and safety of systemically
(intraperitoneally) administrated LeTx in athymic nude mice bearing human melanoma
xenografl tumors. In this study, the ratios of PA to LP giving the highest potency were

determined to be 3:1 and 5: 1, and LeTx was well tolerated at a cumulative dosage of 24

pg total LF at a 5:1 ratio of PA to LP. An LDlo for a cumulative dose was estimated at
30—36 pg total LF. Since this preclinical study, a standard LeTx systemic treatment has
been established for human cancer xenograft tumors in nude mice: 10 pg of PA and 2 pg

of LF are pre—mixed and injected intravenously through tail vein every other day, for a

total of six injections in two weeks (a cumulative does of 12 pg total LF, which is less

than half the LD10 dose). At 12 pg, LeTx effectively inhibits the growth of xenografi

tumors without apparent animal toxicity (Depeille et al., 2007; Ding et al., 2008; Huang

20

et al., 2008). In addition LeTx has been modified to selectively target cancer cells
expressing high levels of urokinase plasminogen activators or matrix metalloproteases,
features of metastatic cancer cells (Abi-Habib et al., 2006b; Liu et al., 2008).

Several lines of evidence support the conclusion that LeTx acts in an anti-
angiogenic fashion. First, as noted above, gross inspection reveals that LeTx-treated
tumors have a pale appearance when compared to untreated tumors. Second, the mean
vascular density for LeTx-treated tumors is dramatically reduced (Depeille et al., 2007;
Ding et al., 2008; Huang et al., 2008). Finally, a recent study using high-resolution
ultrasound enhanced with contrast microbubbles revealed the rapid (within 24 hours)
inhibitory effect of LeTx on the perfusion of fibrosarcoma xenografi tumors, but
interestingly not in non-tumor host tissues (Ding et al. , 2008). This suggests that the
inhibitory effect of LeT x on endothelial function is highly selective for tumor-associated
blood vessels.

Although it is clear that LeTx reduces tumor vascularity, the mechanism by which
it does this is uncertain. The initial explanation that LeTx acted directly on tumor cells to
suppress the release of pro-angiogenic factors was supported by in vitro studies
demonstrating that LeTx suppressed release of angiogenic cytokines such as vascular
endothelial growth factor (VEGF), interleukin-6 (IL-6) and basic fibroblast growth factor
(bFGF). However, this hypothesis was questioned after LeTx was demonstrated in
subsequent studies to inhibit the growth of tumors that do not express toxin receptors (Liu
et al., 2008). Moreover, results from this dissertation research show that systemic LeTx
administration can inhibit growth of human melanoma SK-MEL—28 xenograft tumors

without directly targeting the MEK-ERK pathway in tumor cells (presented in Chapter

21

III). These observations indicate that LeTx targets a non-tumor compartment, perhaps
endothelial cells. In support of this, endothelial cells have been reported to be potential
targets of LeTx (Kirby, 2004; Warfel et al., 2005; Alfano et al., 2009), and the MEK
signaling pathway is required for tumor-associated endothelial function (Mavria et al.,
2006). Regardless of the cellular target, these Observations establish LeTx as an anti-
angiogenic agent. However, LeTx appears to be unique in this aspect since an anti-
angiogenic effect of other MEK inhibitors has not been reported. Thus this activity may
be related to its effects on other MKK signaling pathways

In summary, LeTx is a highly selective biological inhibitor of MEK and most
other members in the MK protein family, and it shows very good potential to be a
potent cancer-targeting therapeutic. Although LeTx is unlikely to enter clinical trials
owing to its toxicity, it has become a powerful tool for studying the involvement of
MEK/MKK signaling pathways, not only in cancers but also in other diseases such as
inflammatory diseases (Pellizzari et al., 1999; Park et al., 2002; Agrawal et al., 2003) and

eye diseases (Bromberg-White et al., 2009).

1.3.3.2. YopJ

YopJ is a member of the Yersinia outer protein (Y op) family of effectors proteins.
The Yops are virulence factors produced by Yersinia species including Yersinia pestis,
the bacterial pathogen that caused the bubonic plague or Black Death in the Middle Ages.
Upon infection, Yops are translocated into host cells through a contact-dependent type III
secretion system. Initial studies showed that YopJ was necessary and sufficient for the

pathogen to inhibit multiple eukaryotic signaling pathways, including all the parallel

22

MAPK pathways (Palmer et al. , 1998; Palmer et al., 1999) and the nuclear factor kappa
B (N F -KB) pathway (Ruckdeschel et al., 1998; Schesser et al., 1998). Inhibition of these
multiple signaling pathways by YopJ results in disruption of cytokine secretion (which in
turn disrupts the innate immune response) and an induction of apoptosis in infected cells.
Later, Orth et al. (1999) confirmed the direct and physical interactions of YopJ with
multiple signaling molecules (including IKKB, an NF-ch activator, and MKKl-S), but
not with the members of MKKK or MAPK. This finding suggested that YopJ inhibited
these signaling pathways by a mechanism that required direct binding to these signaling
molecules. Sequence comparisons showed that YopJ has a secondary structure similar to
that of the adenoviral protease (AVP) family. From this, Orth et al. (2000) demonstrated
the protease activity of YopJ and suggested that YopJ proteolytically inhibited the MAPK
and NF-IcB pathways. However, whether proteolysis was the mechanism for YopJ
inhibition of the MAPK and NF -KB pathways was questionable, because the total amount
and molecular weight of MKK or IKKB were not affected by YopJ treatment (Orth et al.,
1999)

The biochemical mechanism for YopJ-mediated inhibition of eukaryotic signaling
pathways was not elucidated until recently, when studies identified YopJ as an O-acetyl
transferase that acetylates the serine/threonine residues in the activation loop Of
MEK/MKK and IKK (Mittal et al., 2006; Mukherjee et al., 2006). Once acetylated,
those residues can no longer be phosphorylated and activated, resulting in inhibition of
the downstream MAPK and NF-ch pathways. A recent mutagenesis study has identified
the conserved G a-helix in the kinase domain of MEK as a YopJ-binding motif (Hao et

al., 2008). In the study, a PCR random mutagenesis library was screened for mutant

23

Psz (MKK yeast homologue) proteins that were able to suppress YopJ-induced growth
inhibition in yeast. The conserved Ile579 residue of Pbs2 (corresponding to Ile317 of
human MEKl) was shown to be required for YopJ to bind and acetylate Psz. This
residue is in the conserved G a-helix in the kinase domain of MEK, and mutation Of
several residues in this region has been shown to prevent interaction with YopJ (Hao et
al., 2008). Interestingly, the same region has been implicated in binding anthrax LF
(Chopra et al., 2003), and the G helix is also a part of the MEK homodimer interface
(Ohren et al., 2004). Although the link between MEK dirnerization and LF/YopJ binding
has not been established, it is tempting to speculate that dimer disruption may be a factor
in MEK inhibition.

Based on the specific inhibitory effects Of YopJ on the MAPK and NF-ICB
signaling pathways, YopJ also has potential as a therapeutic for cancers and
inflammatory diseases. However, unlike LeTx, which can be internalized by host cells,
the delivery of YopJ into host cells is dependent on a type III secretion system that
requires a direct interaction between the pathogen and host cells (Ghosh, 2004).
Although YopJ can be expressed following transfection into mammalian cells, efficient
delivery of YopJ in a clinical setting must be realized in order to develop YopJ as a

therapeutic agent.

1.3.4. MEK inhibitors fail to generate clinical responses
Two categories of MEK inhibitors that inhibit the pathway activity were reviewed
in this chapter: small-molecule inhibitors that bind and inactivate MEK by a non—ATP-

competitive mechanism, and biological inhibitors that are produced by bacteria and

24

inactivate MEK by posttranslational modifications. The MEK inhibitors in each of these
categories have advantages and disadvantages as therapeutic agents. Small-molecule
inhibitors are generally orally active and non-immunogenic, whereas the biological
inhibitors need to be administrated by injection and may cause an undesirable immune
response. On the other hand, small-molecule inhibitors are “washable” and they do not
physically destroy MEK, whereas the posttranslational modifications of MEK by
biological inhibitors are enzyme-mediated and irreversible in cells. Despite their pre-
clinical successes, the MEK small-molecule inhibitors have not performed well in cancer
clinical trials because of issues such as a lack of clinical response, insufficient efficacy,
and safety concerns. Several facts may explain why these MEK inhibitors work well on
xenograft tumors in nude mice but fail to generate clinical responses in human patients.
First, individuals enrolled in cancer clinical trials are generally those patients who do not
respond to conventional treatments and whose diseases have progressed to late stages
with metastasis. The tumor responses in these patients to MEK inhibitors may not be
well represented by subcutaneous xenografis in nude mice. Second, it has been shown
that cancer cells having constitutively active mutations in MEK activators are relatively
sensitive to MEK inhibition, whereas cells without these mutations are resistant (Y eh et
al., 2004; Solit et al., 2006; Yeh et al., 2007; Haass et al., 2008). However, patients
enrolled in most of the cancer clinical trials for MEK inhibitors were not initially selected
for the preferred mutations. Therefore, patients’ responses may not correlate well with
preclinical predictions. In the Phase I study of ARRY-142886, Ras and Raf mutations
were determined in biopsies, and 38% (10 of 26 assessable biopsies) had Ras (n = 9) or

Raf (n = 1) mutations (Adjei et al. , 2008). Although the three tumors showing the

25

strongest reduction in Ki-67 staining were positive for Ras or Raf mutations, no
statistically significant difference in response was observed, possibly due to the small
sample number in this study. Third, along with cancer progression, tumor cells in late-
stage patients acquire more and more genetic alterations, which may activate MEK-
independent survival pathways, leading to a resistance to MEK-dependent growth
inhibition. This could explain why in the clinical studies of PD 184352 and its
derivatives, MEK inhibitors did show desirable ERK inhibition but failed to generate

clinical responses.

1.4. Introduction to cutaneous melanoma

Cutaneous melanoma is a malignancy of melanocytes, which develops in the skin.
Rarely melanomas can also develop in other parts of the body, such as brain, uvea, and
the interior of the body. Cutaneous melanomas count only 4% of all the skin
malignancies. However, 80% Of deaths from skin cancers are melanoma. Two facts
result in the challenging management of melanoma. First, melanomas usually progress
rapidly without Obvious changes on skin surface. Second, metastatic melanomas are very
resistant to radiation and currently available systemic therapies. Therefore, once
melanomas progress to metastatic stages, most patients are refractory to treatments. In
this section, an overview of melanoma will be presented with a focus on molecular
mechanisms of melanoma formation and currently available animal models for melanoma

research.

26

1.4.1. Melanoma statistics
The American Cancer Society (2010) estimates that 68,130 new cutaneous

melanoma cases will be diagnosed in the United States this year. This ranks melanoma

th th . .
the 5 and the 7 most common cancer type in men and women respectively. The

incidence rate Of melanoma has a race-dependent trend. It is more than 20 times higher
in whites than in Afiican Americans (Altekruse et al., 2010). Among whites, the rate is
about 50% higher in men than in women. In the past 35 years, melanoma incidence rates
have been increasing in the population of white adults of 65 years and older (5.1% per
year in men since 1985 and 4.1% per year in women since 1975). Interestingly, a rapid
increasing of the incidence rate has also been observed among young white women aged
15 to 39 years (3.0% per year since 1992).

The American Cancer Society (2010) also estimates that 8,700 deaths from
melanoma will occur in the United States this year. If detected in the earliest stage,
melanoma is also highly curable. The 5-year survival rate for localized melanoma
(confined to primary site) is 98%. Unlike other types of skin cancers, however,
melanoma is more likely to spread to other parts of the body, possibly due to its highly
motile nature inherited from neural-crest progenitors. If melanoma has spread to regional
lymph nodes, the 5-year survival rates range from 78% to 39%, depending on the
numbers of metastatic lymph nodes (Balch et al., 2009). If systemic metastasis has
occurred, the 5-year survival rate drops to 15% (Altekruse et al., 2010) or even lower
than 10% depending on the substages (Balch et al., 2009). Therefore, early detection of

melanoma is necessary for better prognosis.

27

1.4.2. Risk factors

The two major risk factors of melanoma are ultraviolet (UV) exposure
(environmental risk factor) and genetic abnormality (genetic risk factor). Higher
sensitivity to sunburn and/or a history of excessive sun exposure (including both
sunbums and use of tanning booths) increase the risk. Like many other types of cancers,
family history of melanoma is a strong risk factor. Many melanoma patients have
affected family members (Lynch et al., 1983). This led to cytogenetic studies and
mutation screenings in familial melanoma tumor samples. Genetic abnormalities of
several oncogenes and tumor suppressor genes are believed to be risk factors of

melanoma. These will be discussed later in the following sections.

1.4.3. Melanoma progression, the Clark model.

In 1984, Clark et al. reported a study of melanoma tumor progression. This flow
of melanoma progression, known as the Clark model, describes the six steps of
melanoma progression from normal melanocytes to metastatic melanoma. Each of these
steps is accompanied by histological changes. Table III summarizes these six steps. At
the first step, melanocytes may focally proliferate and the lesions are considered benign.
The proliferating melanocytes may follow a programmed differentiation pathway which
leads to disappearance Of a nevus. If the differentiation pathway is not followed the
melanocytes become hyperplastic (the second step). At step 3, random dysplastic
melanocytes with atypical nuclei form. According to Clark’s observations, the vast
majority of dysplastic precursors are terminal lesions that do not progress to the next step

and form melanomas. If the dysplastic precursors do progress, they follow a two-phase

28

growth pattern before they metastasize. The first phase is the radial growth phase (step 4
of the Clark model), which is characterized by an enlargement of the tumor lesion at its
periphery. Tumors at this step grow in the epidermis, and are not metastatic. The second
phase is the vertical growth phase (step 5 of the Clark model). In this step tumors acquire
the competence for metastasis. Tmnors at this step grow vertically into the epidermis in

an expansive manner forming metastases, the last step of the Clark model.

1.4.4. Melanoma staging

The American Joint Committee on Cancer (AJ CC) staging system is most Often
used to describe the stages of melanoma. This system, also referred to as the TNM
staging system, is based on the size of the tumor (T), the extent of spread to the regional
lymph nodes (N), and the presence of distal metastasis (M). Combining these three
categories (T, N, and M), a grouped staging system is described using 0 and Roman
numerals from I to IV for melanoma (Balch et al., 2009; Gershenwald et al., 2010). In
melanoma staging system, subgroups are defined in each of Stages I to 111 depending of
the TNM status.

Stage 0. At this stage, the melanoma is in situ, and stays in the epidermis but has
not spread to the dermis. This stage is equivalent to the first step of the Clark model.

Stage I and Stage II. Melanomas at these two stages are still localized in the
skin and have not spread to regional lymph nodes. Therefore, melanomas at these two
stages may be between the second step and the filth step of the Clark model. The major
criterion distinguishing the two stages is the tumor size. The melanoma with its thickness

smaller than 1 mm is defined as Stage I, while the melanoma with its thickness between 1

29

and 2 mm is defined as Stage II. Depending on the tumor Size and ulceration, each of
these two stages is subclassified to substages (IA, IB, IIA, IIB, and 11C).

Stage III. At this stage, the melanoma has spread to regional lymph nodes, but
distal metastasis has not occurred. Depending on the tumor size, ulceration, and the
numbers of regional lymph nodes with evidence Of melanoma spread, this stage is
subclassified to substages of IIIA, IIIB, and 111C.

Stage IV. If the melanoma has spread beyond the regional lymph nodes and has
metastasized to distal lymph nodes and/or other organs (mostly the lung, liver, or brain),

it is defined as a Stage IV melanoma.

1.4.5. Treatment of melanoma

Surgical removal of the primary growth and surrounding normal tissues is
essential for melanomas at early stages. Sentinel lymph nodes are sometimes biopsied
for stage determination. 1f lymph node metastases are present, more extensive lymph
node surgery may be needed. For melanomas with deep invasions or metastasis to
regional lymph nodes, immunotherapy, chemotherapy, and/or radiation therapy may be
necessary. Unfortunately, none of these therapeutics prolongs survival in patients with
metastatic melanoma. For metastatic melanoma, currently dacarbazine (DTIC) is the
only US Food and Drug Administration (F DA)-approved chemotherapy agent, and high-
dose interleukin—2 (IL-2) is the only F DA-approved irnmunotherapy (Tarhini & Agarwala,
2006). However, DTIC is effective in less than 10% of patients, and IL-2 has a high
toxicity and gives a low response rate. As discussed earlier in this Chapter (section 1.3.1),

several MEK inhibitors have failed in melanoma clinical trials. A number of clinical

30

trials are ongoing to evaluate the clinical efficacy of some other novel agents for

melanomas.

1.4.6. Common genetic abnormalities of melanoma-

The most common genetic abnormalities found in melanomas are those of BRAF,
NRAS, CDKN2A, and PTEN (Miller & Mihm, 2006; Gray-Schopfer et al., 2007). The
involvement of the products of these genes in melanoma formation and progression will

be the focus in the following sections.

1.4.6.1. BRAF and NRAS

As discussed in the earlier sections in this chapter, oncogenic mutations of BRAF
and NRAS lead to a constitutive activation of the MEK-ERK survival pathway, which
may cause cancer. In fact, before the members in the Raf-MEK-ERK pathway were
identified, mutations in NRAS had been linked to cancers. van 't Veer et al. (1989)
screened a panel of melanoma tumor samples and cell lines for NRAS mutation and found
that NRAS mutations were detected only in melanomas from sun-exposed skin. Since
NRAS mutations were also found in four of the ten primary melanomas examined, van ’t
Veer hypothesized that UV exposure causes NRAS mutation in melanocytes which occurs
at a premetastatic stage. This hypothesis was supported by Ball et al. (1994) who
examined 100 melanoma tumor samples and found NRAS mutations in 36 samples (36%).
In addition, they found that NRAS mutations occurred in 19% of the melanoma tumors
that were in the second step of the Clark’s model, and at a higher frequency (56%) in

primary tumors from continuously sun-exposed Skin. In 2000, Tsao et a1. (2000)

31

evaluated 53 cutaneous melanoma cell lines for NRAS mutation, and found activating
mutations in 21% (11 of 53) of the cell lines. Downstream, BRAF mutations were found
in 66% Of melanomas examined and at lower frequency in other types of human cancers
(Davies et al., 2002). Interestingly, Davies found that all of the identified BRAF
mutations were within the kinase domain and the V600E oncogenic mutation accounted
for 80%. Collectively these studies Show oncogenic mutations in NRAS and BRAF play a
critical role in melanoma development.

Based on Davies’ finding, Pollock et al. (2003) examined BRAF status in nevi,
primary melanoma tumors, and melanoma metastases to evaluate the timing of oncogenic
BRAF mutation. Unexpectedly they detected BRAF mutations resulting in V6OOE
substitution in 82% (63 of 77) of nevi. Since most nevi do not progress to malignant
melanoma, this finding suggests that oncogenic BRAF alone is insufficient for melanoma
turnorigenesis. To support this, Michaloglou et al. (2005) expressed B-Raf_V600E in
normal human melanocytes and found that sustained expression of B-Raf_V600E
resulted in cell cycle arrest, which was accompanied with inductions of tumor suppressor
p16 and other senescence markers. This led them conclude that oncogenic B-Raf induced
cell senescence in normal melanocytes. Moreover, Wajapeyee et al. (2008) performed an
shRNA screening and identified the insulin grth factor binding protein 7 (IGFBP7) as
a candidate required for B-Raf_V6OOE-induced melanocyte senescence, and proposed the
therapeutic potential of IGFBP7 for melanoma treatment.

Collectively, data from these studies indicate (1) the RaS-Raf-MEK-ERK

signaling pathway is essential for melanoma formation, (2) elevated activity of this

32

pathway due to mutations is not sufficient for melanoma tumorigenesis, and (3) other

molecular events are required for melanoma development.

1.4.6.2. CDKNZA 0916’MM and 1219‘”)

In 1986, Pedersen et al. reported abnormalities of chromosome 9 in malignant
melanoma tumors (10/10 melanomas derived from 8 patients). This was the first report
of chromosome 9 abnormality in non-cell line melanoma samples. A similar study was
also presented by Limon et al. (1988) showing that chromosome 9 was one of the
chromosomes that were most fi'equently found to be structurally aberrant in melanoma
cells from patients with metastasis. Later studies of dysplastic nevi and malignant
melanomas showed this structural aberrancy of chromosome 9 was a deletion of
chromosome 9p21 in (Cannon-Albright et al., 1992; Fountain et al., 1992; Petty et al.,
1993). A melanoma susceptibility gene was therefore believed to reside in this region.

Using yeast two-hybrid screening, Serran et al. (1993) cloned p16 from a HeLa

cell cDNA library, and characterized p16 as an inhibitor of cyclin-dependent kinase 4

(CDK4). Therefore p16 was also named p1 61NK4 or cyclin-dependent kinase inhibitor

2A (CDKNZA). Within half a year, p16 was mapped to a locus at chromosome 9p21, and
the high frequency of homozygous deletions or mutations ofp16 was confirmed in
melanoma as well as other types of human cancer cell lines (Karnb et al., 1994; Nobori et
al., 1994). Germline mutations Ofp16 were also detected at a high frequency (92%) in
familial melanoma cases and at a lower frequency (30%) in dysplastic nevus cases
(Hussussian et al., 1994). The p16 gene was therefore proposed as a candidate of tumor

suppressor gene for melanoma.

33

Interestingly, when mapping the CDKNZA locus to chromosome 9p21, Stone and
Kamb (Stone et al., 1995) identified a novel form of CDKNZA transcript which contained
a different first exon resulting in an entirely different Open reading frame fi'om the
original p16 transcript that was initially reported in 1994. This alternative first exon of
p16 was later cloned and mapped to chromosome 9p at about 20-kb upstream to the

original p16 exon 1 (Mao et al., 1995). In the same year, Quelle et al. (1995) cloned the
full-length coding sequence of this novel gene and named it p19 (ARF stands for
alternative reading frame) as it encoded a protein with a predicted molecular mass of 19.3

kDa. What was surprising was not only the fact that this dual-utilized DNA sequence

gave rise to two different open reading frames for distinct gene products, but that

. . 6INK4 . .
expressron ofp19 1n p1 -deleted cells could Induce cell cycle arrest, lrke p16 (Quelle
et al., 1995). The mechanism by which p19 induced cell cycle arrest was not clear at that
time.

Since pl 61 NK4 and p19 were cloned and mapped, groups of studies have

reported mutations in the CDKNZA locus which harbors both p1 OJNK4 and p19 , in

familial melanomas in Australia, Europe and the United States (Walker et al., 1995; Borg
et al., 1996; FitzGerald et al., 1996; Flores et al., 1997; Harland et al., 1997; Platz et al.,
1997; Soufir et al., 1998; Aitken et al., 1999). These findings support the link between
loss of p16/p19 tumor suppressor function and melanoma formation. After years of

studies Of these tumor suppressors, the mechanisms by which p16 and p19 suppress
tumor formation are better understood. Both p16 and p19 inhibit G1/S cell cycle
transition through different but related pathways [reviewed by (Lowe & Sherr, 2003;

34

Polager & Ginsberg, 2009)]. The tumor suppressor pl 6 binds to and inhibits cyclin-
dependent kinase 4 (CDK4), the activity of which is required to release the transcription
factor E2F fi'om Rb, and in turn, promote cell cycle progression through the G1/ S check

point. On the other hand, p19 stabilizes p53 (which is also a tumor suppressor inducing
Gl/ S cell cycle transition and/or apoptosis) by binding and inhibiting Mdm2, a p53

negative regulator.

Before p16 and p19 were cloned and mapped, Cowan et al. (1988) reported an
interesting correlation. Loss of one copy of chromosome 9 was detected at a high
frequency in melanomas but at a much lower frequency in nevi. Similarly, germline
mutations Of p16 were detected at a high frequency (92%) in familial melanoma cases,
but at a lower frequency (30%) in dysplastic nevus cases (Hussussian et al., 1994). These
finding suggested that loss Of pl 6/p19 tumor suppressor function is not likely to be
responsible for nevi dysplasia. Instead, it may be a late-stage event during melanoma

development.

1.4.6.3. PTEN

Since Parmiter et al. (1988) presented the first report Of chromosome 10q
abnormality (including break, translocation, and chromosome loss) in melanoma cell
lines and tumor samples, the involvement of a potential tumor suppressor gene residing
on chromosome 10 in melanoma formation had been hypothesized. Following Parmiter’s
initial report, LOH of chromosome 10q22-10qter was found in non-familial melanoma
tumor samples (Herbst et al., 1994). In 1997, the putative tumor suppressor gene

sequence on human chromosome 10q23 was identified by two independent groups almost

35

at the same time (Li et al., 1997; Steck et al., 1997). The gene product was predicted to
contain a protein tyrosine phosphatase domain and a large region with homology to
chicken tensin and bovine auxilin. The gene was therefore named PT EN (phosphatase
and @sin homolog).

After PTEN was mapped and cloned, LOH of the PT EN allele was found in 40-
60% of primary or metastatic melanoma tumor samples (Teng et al. , 1997; Birck et al. ,
2000), and PT EN mutations were found in about 30-50% of examined melanoma cell
lines (Guldberg et al., 1997; Teng et al., 1997; Tsao et al., 1998) as well as 10% of
primary melanomas (Tsao et al., 1998). The tumor suppression role of PTEN in
melanoma formation was then tested and supported by several studies. First, ectopic
expression of PTEN suppressed melanoma tumor cell growth (Robertson et al., 1998).
Second, over-expression of PTEN inhibited melanoma cell colony formation (Tsao et al. ,
2000). Third, Hwang et al. (2001) Showed that PTEN rescue reduced melanoma
tumorigenecity and metastasis. Finally, a similar study demonstrated that PTEN re-
expression in melanoma cells which lacked PTEN protein expression retarded tumor
development in mice (Stahl et al., 2003).

The tumor suppression function of PTEN was also supported by an epigenetic
study. Using immunohistochemistry staining, Zhou et al. (2000) found no to low PTEN
protein expression in 15% and 50% melanoma samples, respectively. Surprisingly, they
found that among the examined melanoma samples which had no PTEN protein
expression, 80% (4/5) of them showed neither mutation of the PTEN gene nor deletion of
the PTEN allele. Their findings not only demonstrated an epigenetic mechanism by

which loss of PTEN function is involved in melanoma formation, but also suggested that

36

the involvement of loss-Of-function of PTEN in melanoma might be underestimated.
This provides an explanation for the fact that in some studies the PTEN mutation rate in
melanomas was very low and no deletion or LOH of PTEN was detected (Wu et al.,

2003)

1.4.7. Disease models

Oncogenic mutations and loss of tumor suppressors have been identified as
critical drivers of melanoma formation and the disease progression. Based on these
findings, great efforts have been made to develop animal models for melanoma research.
These melanoma animal models are developed mainly for two purposes. First, these
models allow testing the requirements of those genetic alterations and involvements in
melanoma tumorigenesis and the disease progression. Second, these melanoma animal
models facilitate the development of melanoma therapeutics as the efficacy of novel
therapeutic agents and treatment strategies can be tested in these models. This section
will discuss these melanoma animal models.

In 1997, Chin et al. (1997) reported a cooperation of oncogenic Ras and

p] 61 NK4a-deficiency for melanoma susceptibility in viva. In this mouse model, the
. G12V . .
oncogemc Ras (H—ras ) was expressed specrfically 1n melanocytes (under the control

of tyrosinase promoter) on a p] 6INK4a-deficient background. These mice (with double

deficiencies) developed spontaneous melanoma after a short latency, indicating the

oncogenic Ras and loss ofp16mK4a allele cooperatively promote melanoma formation.

In addition, two interesting observations were obtained from this work. First, the

37

fi'equency of H—rasGIZ V-transgenic founder production was unusually low (only three

founders from a total of 420 pronuclear injections). This raised the possibility of a

. . . GI 2 V .
selective pressure against oncogemc Ras. Second, before the H-ras -transgen1c

. NK4 . .
founders were crossed wrth the p161 a-deficrent mice, no melanoma or other

malignancy occurred in the three founders (hyper-pigmentation and a proliferating
melanocytic mass were the most relevant phenotypes observed). This indicates that
activation of Ras is not sufficient for melanoma tumorigenesis, and that a second genetic
alternation is required for melanoma tumorigenesis. This was supported by Powell et al.
(1999) using another similar melanoma mouse model in which a mutant H-Ras was
expressed specifically in melanocytes. Powell et al. also found that UV radiation or
chemical carcinogens such as DMBA and TPA were necessary for melanoma formation
in these mice.

Since constitutive activation of Ras may result in low fiequency of transgenic

founder production, in 1999 Chin et al. (1999) developed an inducible melanoma mouse

model, in which the melanocyte-specific H-rasGlzV expression was induced by

doxycycline on a p] 6m lad-null background. Not surprisingly, melanoma developed in

doxycycline-treated mice within 60 days. More interestingly, withdrawal of doxycycline
resulted in a clinical and histological regression of developed tumors, indicating that a
persistent Ras activation is required to maintain melanoma tumor growth.

In 1998, a melanoma mouse model was established by Kelsall et al. to test the
effect of UV exposure on melanoma formation. In this model, the viral oncoprotein

small DNA tumor virus (SV40) early region was engineered to be expressed specifically

38

in melanocytes under the control Of mouse tyrosinase promoter. After UV exposure,
primary melanoma and melanoma metastases occurred without the use of other chemical
carcinogens. This mouse model not only tested the requirement of oncogenic proteins for
melanoma tumorigenesis, but also demonstrated the critical involvement of UV exposure
in melanoma initiation. Since SV40 oncoprotein inactivates both Rb and p53 (Levine,
2009) pathways, this work also demonstrated the involvement Of loss of the tumor
suppressors in UV-induced melanoma formation. More relevant to melanoma-associated
genetic abnormalities, Kannan et al. (2003) evaluated the effect of UV exposure on active
Ras and pl 6- or pl 9-deficient mice. The results fiom this study showed UV treatment
accelerated melanoma formation in active Ras and p19-deficient mice, indicating the
cooperation between UV exposure and genetic alteration in melanoma formation. In
contrast, this UV-induced acceleration of melanoma formation was not Observed in the
active Ras and p16-deficient mice. This finding indicates that the p53 pathway plays a
more critical role in suppressing UV-induced melanoma formation in viva than the Rb
pathway.

To evaluate the functional cooperation between PTEN and p16/p19, a dual
inactivation of which was found in several types of human cancers, You et al. (2002)
established a PTEN/INK4a double deficient mouse model. They found that deficiency of
either PTEN or INK4a was not sufficient to induce melanoma in mice, and double
deficiency of both tumor suppressor genes caused only low frequency of melanoma (7-

10%). Although the contribution Of an oncogene to melanoma tumorigenesis in viva was

not included in this study, they found that an introduction of oncogenic Ras (H-rasGlzv)

increased the anchorage-independent colony formation of the MEFs isolated fi'om the

39

PTEN/INK4a double deficient mice. This also supports that melanoma formation
requires both oncogene activation and loss of tumor suppressers.

In the past decade, Cre recombinase-mediated conditional gene recombination has
been developed as a powerful tool allowing the introduction of desired genetic
alternations in a conditional manner. Combining the tissue-specific transgenic approach,
Bosenberg et. al (2006) recently established a mouse strain in which the expression of

Cre recombinase was inducible by 4-hydroxytamoxifen (OHT) specifically in

. T2 . . . .
melanocytes(desrgnated TyrssCreER ). Usrng thrs transgemc mouse strain, researchers

can conveniently manipulate the genetic context specifically in melanocytes to evaluate
the contribution of a molecular event in melanoma tumorigenesis. Most recently,

Dankort et al. (2009b) demonstrated a great example of this approach by crossing the

TyrirCreERTz mice with another strain of mice in which the mutant B-RafVGOOE

. . . . . V600E
expressron was inducible to create an mducrble B-Raf melanoma mouse model.

Using this Dankort et al. found that, again, constitutive activation of B-Raf was sufficient
to develop benign melanocytic hyperplasia, but was not sufficient for melanoma
formation. Similar to the finding by Chin et al. (1997) discussed earlier in this section,
Dankort et al. concluded that oncogenic B-Raf must cooperate with a loss of tumor
suppressor, PTEN in this example, to induce melanoma formation and lung metastasis.

In addition, using this melanoma mouse model they demonstrated the efficacy Of kinase
pathway inhibitors for melanoma prevention and treatment. This work not only

demonstrated the efficacy of kinase pathway inhibitors in a non-xenograft melanoma

40

animal model, but also provided an example of how a melanoma mouse model may be
used to test therapeutic agents.

Collectively, these studies using melanoma mouse models indicate that both
activation of oncogenes and loss of tumor suppressors are required for melanoma
formation. More importantly, these studies provide a molecular mechanism of melanoma
progression. Acquiring oncogene activation seems to be the earliest event. This is
believed to be the cause of hyperplastic lesions as most benign nevi have Ras or B-Raf
oncogenic mutations (Pollock et al., 2003), and loss of tumor suppressors is not
frequently found in dysplastic nevi (Cowan et al., 1988; Hussussian et al., 1994).
According to the Clark model, most benign nevi at step 1 follow a programmed
differentiation pathway and disappear. Supporting this, previous studies have suggested
that oncogene acquisition is not sufficient for tumor formation, but instead, induces cell
senescence (Michaloglou et al., 2005; Braig & Schmitt, 2006). This also explains why
most benign nevi have Ras/Raf mutations but do not progress to malignant melanoma.
As shown by several melanoma mouse models discussed above, oncogene activations
need to cooperate with loss of tumor suppressors to initiate melanoma. This molecular
mechanism of melanoma progression is in consistent with the Knudson two-hit theory

(Knudson, 1971 ).

1.5. Discussion
The Ras-Raf-MEK—ERK signaling pathway plays a critical role in regulating cell
cycle progression, cell proliferation, and survival. Deregulation of the pathway activity

may cause diseases, including cancers. Somatic mutations of Ras and B-Raf are found in

41

several types of human cancers at a high frequency particularly in melanomas. Using
cell-based in vitro studies as well as in viva mouse models, numerous research groups
have demonstrated the critical role of this oncogenic signaling as well as other tumor

suppressing pathways in melanoma tumorigenesis and survival.

Considerable effort has been made to develop therapeutic agents for treating
cancers. Among these potential therapeutic agents, small-molecule inhibitors targeting
MEKl and MEK2 have shown promising anti-cancer effects in laboratories and pre-
clinical studies. However, none Of them has demonstrated clinical efficacy in cancer
patients and proved as anti-cancer drugs. This indicates that our understating of the roles

of MEK signaling pathways in cancers is incomplete.

MEKl and MEK2 are generally thought to be functionally redundant for the
following reasons. First, MEKl and MEK2 share greater than 85% homology in amino
acid sequences, and almost identical crystal structures (Ohren et al., 2004). Second, both
MEK] and MEK2 phosphorylate ERK1 and ERK2, and ERK] and ERK2 are the only
known substrates of MEK] and MEK2. Regardless the high structural and biochemical
similarities of MEKl and MEK2, however, a handful of studies have suggested that these
two protein isoforms may have non-overlapping biological functions (Giroux et al. , 1999;
Belanger et al., 2003). A better understanding of the difference between MEK] and
MEK2 as well as their distinct biological functions may help developing more efficient
anti-cancer drugs. For this purpose, using two complementary approaches I evaluated the

necessity and sufficiency of MEKl and MEK2 signaling pathways for melanoma cell

42

proliferation. This research provides insights into the complexity of the MEK signaling

pathway. The results are presented in the following chapters.

1.6. Tables and figures

43

.933 33 to... 3 3530.5 5.58» u§u§v50w80303 93 6803 5383005 0528.33

.3 3530.5 £3588 <75 A9923 .8380;on :00 .3 3530.5 28380305 :00 v2.22: 030.6 33:00 a

.vEm—énmmosm .233 363:3 wfims 83 80203 was Mam 9 53.3%-: 89a 00253033032”

00—0205 .3360 ME). 8m 95805302 .952 no ~Mm2 ©0502. mam: .3 “00:20.20.“ 0.03 930.3 0350— 9...... :N e

 

38 r... a 3.3 .88 s... a not 0....

.83 r... a not H.3.

 

2: mg... I 8.: 2.. v3... 333%?
Goon r3 .0 flow 68m 5.33.3.5 a. 9236 ”m3. Goon .Bw0mm30m a. 0.335 «0% 3028 Om
2.. 32...... n :5... 2.. :5...
.. .. 38 c... a 8:53
30.0” .20 on - 0 .0 . r . .. .
Goom N I m 32 N 2 cog—hr?” WWW...” .88 Na .0 3.39 .33 Na .0 2303-22.05 $0.”. mmmvw. Om
2.. m... .. 5....
: A.oo~..e
e 0 «>3 H 0
:33 N a 8 “W 1.22”." .0 Ed... mooom :3 .0 355 ”M33 :3 .0 333.5 «0% 8.53
2.. n. u an...
. zoom 4.. a 5.... 608 ..... s 853 33. s...
. e .0 20 n 0 . .. . .. . .
Ago— N fine—ham .0 303m .30— 3 .0 >085 .33 3 .0 307.5 mom omowoo Om
2.. cm 1 n
AN3.303 333.30 .3 3530.5 3:03 33:00 03.233 ch...» 2.. .3 60530.5 .3380 M92 952550

 

.2... 80.

 

0.05:3 ch— ciao—eaéafim A 0335

2500.

.wam a. 80:00:09 :0me :3 8.30035 .3 .000: 80.020 05 :0 B33 3:303 03 33.. E 03 mm- 35 0:03:00: 035350 0.3. a
003200: 050.35 .33 603:0 0.8 00:0 03.320
05 8 8:: 00:23: 0:28: :0 RU 0233 BE. 00:0 0330—0 "3 05 :0 :_ 60300900 .3.: 03 835 L 5360: 0330.0 0308:: 05

3 2.3 00.6.00: 0.33%? .335 0030—0 ”5 0:03.: :oummo: 05 8 “003:0 03 00:00 53070.»:2 :m :3030 000:0:30 200 05:3. 0

 

H4 \ x M N 012w... \ m. muamgmm 0:950:00
n... .H. 0 a 0. 0 m a \ .050 a z 0 m m .0. 4 CHISEEM:
w w a z m A m .H \ 3.30 q 0. m m a m. 0 $0330.53
0 m m m m M m H \ .0H.M q o A z m M M 0MM2.
e m z m z a B h \ .002: .4 e 0 M, m 3 m .mmmummm.0MMz
m m m z 0 m z A \ GEM q 4 .M m M o o .00...-00M3MVE
m m M 0 z 0 0 H \ .0m.m : a M M M w 0 anM2
H a m 2 H a q d \ .OH.m : > m M m m : mMmz
: m m m z a o H \ .0.0 e m m m M A 2 HMmz
.0 .0 .0 .0 .0 .m .N .H H m m 0 m 0 h 0 "0.0. 00000000: 000000350

n.

 

ammunmsvmm UHUM ocHad

0000000350 :4

 

.0030 03.5.0.0 ha 05 was: 00055—000 30: 05:3 MESS—HE no 0:08:33. .= 030%

45

 

:00:«0 0:26.202

6626208 :0: 00:80:58 5:? :062 «30:00::«0 9.2::

66220:: :0: 00:20:88 3056» :062 30:00:.«0 586:

A2923 820:: 5:3

6:00 :0 00:«:«0::« Ea :0::«u:0:0.m€ ::«::0n«v 2&3 0E0_0$0 E205:

w:« «6290:»: 06:00:22: 30:23::2 E0668: 5:3 30: 02308202

.A:0_:«t:0:0.m€ ::«::0n«v «6290:»:
05:00:20.: 6:022:0— 00:6:: :5 ::066:0: 5:3 30: 0230:2022

6:060: :wE0£
603005—08 _«:::0: b23036 0:8 30: 0:50:22: 00:33:: 3:05:00

«80:20:: 0320202 .0

«60:20:: E86: :0 062:: 539% _«0_::0> .m

«80:20:: S86: :0 08:: 538w 203: .v

«62:93 2308202 .m

«629093 0230:2032 .N

250: 0:50:23): 2

 

80::6060Q

6:06 200:: 2:20

 

.:06»0:w0:: «30:20:: «0 360:: 2:20 0:: a: 030:.

46

Figure 1. Schematic illustration of the Raf-MEK—ERK signaling pathway.
Ras-mediated Raf activation involves a complex series of events, including
protein recruitment to the cell membrane and phosphorylation at numerous residues
[reviewed by Chong et al. (2003)]. Active Raf then activates MEK by phosphorylating
the two conserved serine residues in the activation loop. Active MEK then in turn
activates ERK by phosphorylating the threonine and tyrosine residues in the conserved T-
E-Y motif. After its activation, ERK phosphorylates and activates downstream effectors
that regulate a variety of cellular events, including cell cycle progression and cell
proliferation at the transcriptional and posttranslational levels. The circled P symbols
denote phosphate groups added on the critical serine residues and threonine/tyrosine

residues of MEK and ERK, respectively.

47

Figure 1. Schematic illustration of the Raf-MEK-ERK signaling pathway.

_
lRasI

H

 

 

6'6

@

@3

Cell cycle progression
Cell proliferation

48

Figure 2. Structures of MEK inhibitors.
(a). PD 098015. (b). U0126. (c). PD 184352. ((1). PD 0325901. (3). PD 318088.
(f). PD 334581. (g). ARRY-142886. (e-g). The A and B rings of the diaryl amines of PD

1843 52 and its derivatives are denoted.

49

Figure 2. Structures of MEK inhibitors.

NH2 CN NH2 0
NHZCN NH2

 

PD 098059 U0126
O,NH o dl-IO’\/\O NH 0
NH ‘\ NH ‘\
PD 184352 PDF 0325901
e f.
QCNXNHN
O,NH 0
El he
PD 318088 PDF 334531

3””fi
Br

ARRY-142886

50

Figure 3. Effect of PD 184352 on MEK activation in cells.

Human melanoma SK—MEL-28 cells were cultured as described in the Materials
and methods section of Chapter 11 (page 7 7). Cells were treated with PD 184352 at the
concentrations indicated for 24h. Total cell lysates [prepared as described in the
Materials and methods section of Chapter 11 (page 79)] were subjected to immunoblotting
probed with antibodies against phospho-MEKl/Z, phospho-ERKl/Z, total ERKl/Z, and

GAPDH.

[PD184352](nM): o 10 1001000

E - Q q MEK1I2-P04

...—._.' ___ ERK1I2-PO4

 

 

 

1m... up ‘ nun-l ERK1I2

 

 

GAPDH

 

51

Chapter II. Cleavage-resistant MEK proteins; a novel experimental model to

establish MEK sufficiency

1.1. Introduction

In order to determine the sufficiency of a MEK/MKK signaling pathway for a
cellular function, the experimental system used will have to inhibit the pathways of
multiple MEK/MKK members, but leave only one active. It is not expected to be
efficiently achievable by transfecting several interference RNA molecules into cultured
cells to target multiple protein molecules. In this chapter, the development of a novel
experimental model achieving this challenging task will be presented. In this system,
multiple endogenous MEK/MKK proteins and the downstream signaling pathways are
inhibited by treating cells with LeTx, a bacterial toxin with MEK/MKK-specific
proteolytic activity (see Chapter I section 1.3.3.1 for the introduction). Simultaneously, a
mutant MEK/MKK protein that is resistant to LF-mediated cleavage is expressed in cells
to preserve the specific MEK/MKK signaling pathway. This achieves a special cellular
context in which only one MEK/MKK signaling pathway is present. If a MEK/MKK
pathway is sufficient for a cellular function, the cleavage-resistant form of the
corresponding MEK/MKK should rescue this cellular function when the cells are treated

with LeTx.

52

1.2. Results
1.2.1. Point mutations at the Pl' site render MEK resistant to LF—mediated

cleavage

An alignment of all the MEK and MK amino acid sequences (Table II) revealed

a consensus sequence for LF cleavage: (B/P)3.4X3/Al, where B represents basic residues,

P represents proline, X represents variable residues, and Al represents aliphatic residues.
The aliphatic residue at the Pl' position is fully conserved in all the MEK/MKK proteins,
and has been shown to be critical for the cleavage by LF (Park et al., 2002; Chopra et al.,
2003). Therefore, to make MEK resistant to LF-mediated cleavage, I introduced an
aliphatic-to-aspartic acid mutation in this position of human MEK and MK proteins.
These cleavage-resistant (or) mutant proteins are named MEKcr or MKKcr in this

dissertation. For the subsequent study, a V5 tag was fused to the amino termini of wild-

type and the cleavage resistant mutants of MEK/MKK proteins. This NHz-terminal V5

fusion was included to examine the cleavage of the proteins by LF at the NHZ—termini

(explained below).

To confirm the cleavage resistance of MEKcr, an in-cell LF cleavage assay was
developed. In this assay, wild-type VS-MEK or VS-MEKcr proteins were transiently
expressed in CHO K1 cells which were then treated with PA alone (1 uyml PA as a
control) or LeTx (1 ug/ml PA and 0.1 ug/ml LF) for 12h. As transient transfection of
plasmid DNA can efficiently over-express exogenous proteins in CKO Kl cells, MEK
cleavage by LF may not be clearly observed if an excessive amount of VS-MEK is

produced during LF cleavage. To overcome this, cells were co-treated with 10 ug/ml

53

cycloheximide to block de novo protein synthesis during PA or LeTx treatment. To
ensure successful internalization of LF and in-cell proteolytic activity of LeTx, cleavage
of endogenous MEKI was first examined by immunoblotting probed with antibodies that
specifically recognize the amino terminus of MEKl (LeTx causes loss of epitope) or the

carboxyl-terminus of MEKl (LeTx causes mobility shift). As shown in Figure 4A, LeTx

treatment resulted in a loss of the NHz-terminal epitope (Figure 4A, second panel) and an

increase in the electrophoretic mobility of endogenous MEKl (Figure 4A, third panel) in
control cells as well as in cells transfected with VS-lacZ control plasmid. This result

indicates LF-mediated MEK cleavage occurs in CHO Kl cells. Under the same

conditions, LF-mediated proteolysis resulted in a loss of NHz-terminal V5 epitope only

in CHO K1 cells expressing wild-type VS-MEKI or V5-MEK2 but not in cells
expressing VS-MEKI or or V5-MEK2cr (Figure 4B and C, top panels). To exclude the
possibility that the loss of V5 epitope of wild-type VS-MEKI and VS-MEKZ was a result
of failure in expressing the exogenous proteins, and to further confirm the cleavage of
wilt-type VS-MEK proteins, antibodies recognizing carboxyl termini of MEK] or MEK2
were used to detect the electrophoretic mobility shift of cleaved VS-MEKI and V5-
MEKZ. As shown in the second panels of Figure 4B and C, bands with slightly higher
electrophoretic mobility were detected by immunoblotting with antibodies recognizing
carboxyl-terminus of MEK] or MEK2 in cells transfected with wild-type VS-MEKI or
VS-MEKZ, but not in cells transfected with VS-MEchr or V5-MEK20r. These results
demonstrate that VS-MEchr and V5-MEK20r were resistant to LF-mediated proteolysis.
In this experiment, a substantial increase of non-tagged MEK expression along

with VS-MEK transfection was noticed (Figure 4B middle panel, compare none-

54

transfected cells with V5-MEK1 transfected cells). This was caused by internal
translation fiom the original translation initiation codon of the MEK sequence cloned in
the V5-expression vector. This was evidenced by the following observation. The
original translation initiation codons that were inserted together with MEK sequences

into the V5 expression vector were removed from the plasmid by doing a site-directed
mutagenesis. The modified expression vectors were then transfected into CHO K1 cells.
After this manipulation, the increase of non-tagged MEK expression was no longer
observed (Figure 5). These experiments demonstrated that the increased non-tag MEK
level after VS-MEK plasmid transfection was due to exogenous expression and not an
up-regulation of the endogenous MEK. This dual translation pattern was also observed in

VS-MEK-transfected SK-MEK-28 stable cell lines (will be presented in Chapter 111).

1.2.2. The cleavage-resistant mutation does not impair MEK activity.

Cleavage by LF removes an amino terminal docking domain that is required for
interaction with MAPK (Tanoue et al., 2000; Chopra et al., 2003; Bardwell et al., 2004).
To confirm that neither the addition of the V5 tag nor the introduction of the LF
cleavage-resistant mutation interferes with MEK’s ability to interact with and
phosphorylate ERK, the activity of VS-MEchr and V5-MEK20r was examined. To do
this, clonal lines of human melanoma SK-MEL-28 cells stably expressing VS-MEKI or or
V5-MEK2cr, or their wild-type counterparts were established. Similar to most
melanomas, SK-MEL-28 cells harbor the B-Raf__V600E mutation (Davies et al., 2002),
which constitutively phosphorylates and activates MEKl and MEK2. Therefore, this cell

line allows a convenient examination of the status of cellular MEK] and MEK2. When

55

an antibody that only recognized activated MEKl and MEK2 was used in the
immunoblotting, the activation status of both VS-MEchr and V5-MEK2cr were found

to be comparable to that of their wild-type counterparts (Figure 6A). This indicates that

(1) neither the NHz-terminal V5 fusion nor the CR mutation impairs MEK activation and

(2) VS-MEchr and VS-MEKZCr are activated in SK-MEL-28 cells.

To further examine the ability of VS-MEKI or and V5-MEK2cr to phosphorylate
their only known substrate, ERK, immunoprecipitation-kinase (IP-kinase) assays were
performed. To do this, VS-MEchr and V5-MEK20r were immunoprecipitated from
SK-MEL-28 cells and used as the kinase source for ERK in the kinase assay. Since
saturating the kinase in a kinase assay may result in a false positivity, the optimum
amount of wild-type VS-MEK immunoprecipitates in this IP-kinase assay was
determined first. A fixed amount of recombinant ERK2 (400 ng) was used as the
substrate and mixed with a decreasing amount of the VS-immunoprecipitates in the
presence of active B-Raf to determine the minimum amount of the V5-
irnmunoprecipitaes required for the kinase reaction. As shown in Figure 6B, a minimum
of 5 ul of VS-MEKI or V5-MEK2 immunoprecipitates was required to fully
phosphorylate 400 ng of ERK2 in the kinase reaction. Based on this optimization

experiment, a full set of the IP-kinase assays, in which the VS-MEK immunoprecipitates
V were mixed with or without active B-Raf and recombinant ERK2, was performed based
on the optimized condition to test the kinase activity of VS-MEchr and V5-MEK2cr.
As shown in Figure 6C, after considering the amount of the V5 immunoprecipitates used
for each assay (Figure 6C, lower panel), VS-MEchr and V5-MEK20r were as capable

as their wild-type counterparts in phosphorylating ERK2. These results indicate that

56

neither the addition of the V5 tag nor the introduction of the cleavage-resistant mutation

alters the ability of MEK to interact with and phosphorylate ERK.

1.2.3. Point mutations at the Pl’ site render other MKK members resistant to LF-
mediated cleavage

A similar strategy was used to generate cleavage-resistant forms of MKK3,
MKK4, MKK6 and MKK7. As expected, the aliphatic-to-aspartic mutation at the P1 ’
position rendered MKK3 and MKK6 resistant to LP cleavage (Figure 7). As MKK4 and
MKK7 were previously reported to harbor two LF cleavage sites (V itale et al., 2000),
two cleavage-resistant mutants with the aliphatic-to-aspartic mutation introduced to one
of the cleavage sites were generated for each of MKK4 and MKK7: V5-MKK4cr_46 and
V5—MKK4cr_59 for MKK4, and V5-MKK7cr_45 and V5-MKK7cr_77 for MKK7. LF

completely cleaved wild-type V5-MKK4 in CHO K1 cells (Figure 8A). Unexpectedly,

the aliphatic-to-aspartic mutation introduced to Leu46 of MKK4 was sufficient to make

. . . 59
MKK4 re51stant to the cleavage, but the same mutation introduced to Phe was not. On

the other hand, V5-MKK7cr_45 and V5-MKK7cr_77 were both resistant to the cleavage
in CHO Kl cells (Figure SB). However, under the same condition, a convincing

cleavage of wild-type MKK7 was not detected (Figure 8B).

1.2.4. MKK4 cleavage in mammalian cells
Further experiments were performed to determine why V5-MKK4cr_46 was not

cleaved by LP in CHO Kl cells. As shown in Figure 8A, LeTx treatment of CHO-K1

cells expressing wild-type V5-MKK4 resulted in a complete loss of the V5 epitope,

57

demonstrating the cleavage of MKK4 by LF. Since two LF cleavage sites were identified
on MKK4, a mutant MKK4 with the cleavage-resistant mutation introduced at only one
of the cleavage sites would be expected to be still cleaved by LF at the other cleavage site
which remained unchanged. As expected, V5-MKK4cr_59 was still sensitive to LF-

mediated proteolysis (Figure 8A). Presumably LF cleaved this mutant MKK4 at the

Lys45-Leu46 position. Unexpectedly, however, the aliphatic-to-aspartic mutation
introduced to Leu46 of human MKK4 was sufficient to render MKK4 resistant to
cleavage, even if the Phe59 residue was remained unchanged (Figure 8A). Two
possibilities can explain this observation. First, LF cleaves MKK4 only at the Lys45-

Leu46 position but not at the Argsg-Phe59 position (i. e., Argsg-Phe59 of MKK4 is not a

real LF cleavage site). Alternatively, LF cleaves MKK4 at both positions in a processive

manner (1'. e. , cleavage at ArgSB-Phe59 requires a prior cleavage at Lys45-Leu46).

To test these two possibilities in an unambiguous way, wild-type human MKK4
fused with a V5 tag followed by a 6xHis tag at the carboxyl terminus was constructed
(denoted as MKK4-V5-His6 in this dissertation). The carboxy-terminal fusion allowed
more readily determining the molecular weight of the carboxyl terminus of MKK4 after
LF cleavage. In addition, two MKK4-V5-His6 deletion mutants were constructed as
molecular weight indicators: MKK4 with a deletion of amino acid residues 1-45 (denoted
as MKK4_d45-V5-His6), and MKK4 with a deletion of amino acid residues 1-58
(denoted as MKK4_d58-V5-His6). The MKK4_d45-V5-His6 deletion mutant also

allowed testing the second possibility that LF cleaves MKK4 at both sites in a processive

58

manner. The set of these proteins was then expressed in CHO Kl cells and in-cell
cleavage assays were performed.

As shown in Figure 9, LeTx treatment on CHO Kl cells expressing wild-type
MKK4—V5-His6 resulted in a complete loss of the V5 epitope, indicating MKK4
cleavage by LF. However, under the same conditions I could not detect the LF-cleaved
carboxyl terminal fragment of MKK4 using an anti-V5 antibody (Figure 9, lane 2). An
explanation for this is that after LF cleavage, the carboxyl terminus of MKK4 is degaded,
as previously observed for MEKl (Duesbery et al., 1998) and MEK2. To test this, MG-
132, a proteosome inhibitor, was included in the in-cell cleavage assay. As shown in
Figure 9 (lane 4), MG-132 treatment rendered the LF-cleaved carboxyl terminus of
MKK4 recognizable by V5 antibody in the immunoblotting, indicating that after LF
cleavage MKK4 is degraded through a proteosome-dependent pathway.

The two MKK4 deletion mutants, MKK4_d45-V5-His6 and MKK4_d58-V5-His6,
were distinguishable by the differential electrophoretic mobility in an SDS-PAGE gel
(Figure 9, land 5-8 and lane 9—12). After LF cleavage, the carboxyl terminus of MKK4-
VS-His6 (Figure 9, lane 4) had the same electrophoretic mobility as that of MKK4_d45-

V5-His6 (Figure 9, land 5-8) but not MKK4_d58-V5-His6 (Figure 9, lane 9-12). This

result demonstrates that wild-type MKK4 is cleaved by LF only at the Lys45-Leu46

position but not the Arg58-Phe59 position in cells. In addition, co-treatment of LeTx and

MG-132 did not result in an electrophoretic mobility shift of the MKK4_d45-VS-His6

deletion mutant. This indicates that MKK4_d45-V5-His6 was not cleaved by LF and that

the Argsg-Phe59 position of MKK4 was not an LP cleavage site. Taken together, these

59

data show that MKK4 has only one LF cleavage site, the Lys45-Leu46 position, as

opposed to two LF cleavage sites as previously observed and reported by Vitale et al.

(2000) in an in vitro assay.

1.2.5. MKK7 cleavage in mammalian cells

As shown in Figure 8B, LeTx treatment resulted in a subtle decrease of the NH2-

terminal V5 epitope of wild-type MKK7 (the second lane of the top panel in Figure 8B).
One possible explanation for this is that a limited amount of LF could not cleave the
excess of V5-MKK7 that was expressed in the transfected cells. To test this, a decreasing
amount of the plasmid DNA encoding for wild-type V5-MKK7 was transfected into
CHO K1 cells and the cleavability of MKK7 was tested in cells in the presence of
cyclohexirnide. Under this condition, the expression level of wild-type V5-MKK7 was
decreased proportionate with the decreasing amount of the plasmid DNA transfected
(Figure 10, upper panel). However, LF was still not able to cleave wild-type MKK7 even
when the expression of V5-MKK7 was decreased to a barely-detectable level. The
uncleavability of wild-type V5-MKK7 in this experiment was not due to experimental
failure as a complete cleavage of endogenous MEKl by LF was readily detected in the
same cell lysates (Figure 10, middle panel). This result indicates that MKK7 might not
be an enzymatic substrate of LP in mammalian cells. To further test this, 293FT cells (a
human embryonic kidney 293-derived cell line) were treated with LeTx for 24, 48, and

72 hours, and whole cell extracts were prepared for immunoblotting to examine MKK7

cleavage by LF. If LF cleaves MKK7 at Gln44-Leu45 and Gln76-Leu77 positions as

60

previously reported by Vitale et al. (2000), then a decrease of NHz-terminal signal and an

electrophoretic mobility shift of MKK7 should be detected following cleavage. As
shown in Figure 11A, LeTx treatment did not result in a loss or a decrease of the
immunoblotting signal of endogenous MKK7 in 293FT cells, even if the cells were
treated with LeTx for 72h. In addition, an antibody recognizing the last 20 amino acid of
MKK7 failed to detect an electrophoretic mobility shift of MKK7. This result clearly
shows that MKK7 is not a preferred LF substrate in mammalian cells, and argues the

MKK7 cleavage by LF observed by Vitale et al. (2000) may be an in vitro artifact.

1.3. Discussion

In this chapter, the design and the engineering of cleavage-resistant forms of
MEK and MK (MEKcr and MKKcr) were presented. This set of MEKcr were designed
to be used as a tool to study the sufficiency of MEKl and MEK2 signaling pathways for
melanoma cell proliferation, which will be presented in Chapter III. Cleavage resistant
forms of MEK/MKK proteins have been previously reported for MEKl (Duesbery et al.,
1998), MKK3 (Park et al., 2002), and MKK6 (Park et al., 2002; Chopra et al., 2003).
Although the mutations were introduced at slightly different positions on each of MEKI
(P1), MKK3 (P1 and P1 '), and MKK6 (P1 and P1 ', or Pl' alone), all of these studies
suggest that hydrophobicity and surface charge surrounding the LF cleavage position are
critical for the cleavage. In this dissertation project, aliphatic-to-aspartic acid mutations
were introduced only to the Pl' positions of MKK1-4, 6 and 7, and the produced MEKcr
and MKKcr were all resistant to LF-mediated proteolysis. These results extend previous

observation (Chopra et al. , 2003) showing that the aliphatic residue adjacent to the LF

6]

cleavage site are critical for LF-mediated proteolysis.

When VS-MEchr and V5-MEK20r were stably expressed in human melanoma
SK-MEL-28 cells, which harbor B-Raf_V600E constitutive activation mutation, these
cleavage-resistant MEK mutants were phosphorylated and activated in the cells
(examined by immunoblotting with an antibody that specifically recognizes MEK] and
MEK2 with phosphate groups at the activation loops). These MEKcr proteins were
phosphorylated to a comparable level as their wild-type counterparts, indicating that
introducing the aliphatic-to-aspartic acid mutation to MEKl and MEK2 does not alter
their ability to be activated. Moreover, these VS-MEKcr proteins could be
immunoprecipitated from SK-MBL-28 cells and used as kinase source in an in vitro
kinase assay. These V5-MEKcr proteins were as capable as their wild-type counterparts
in phosphorylating their substrate, ERK2, in the assay. Interestingly, immunoprecipitated
VS-MEK and VS-MEKcr proteins were not able to phosphorylate ERK2 in the absence
of active B-Raf, although they were activated in SK-MEL—28 cells. A few possibilities
can explain this. First, during cell lysis and immunoprecipitation, these proteins may
have been de-phosphorylated by phosphatases, the activity of which was not fully
inhibited by phosphatase inhibitors added in the lysis buffer. Second, in SK-MEL—28
cells, the B-Raf_V600E may have only activated a small portion of VS-MEK, yet this
was sufficient to be detected by the phospho-MEK1/2 antibody by immunoblotting.

After immunoprecipitation, the vast majority of the immunoprecipitated VS-MEK may
not have been activated by B-Raf_V600E and was not able to phosphorylate ERK2 in the

kinase assay until after the addition of active B-Raf. Finally, it is also possible that active

62

VS-MEK may be sequestered in a protein complex that may mask the epitope for
antibody recognition. As a result of this, only less active VS-MEK could be

immunoprecipitated.

Vitale et al. (2000) reported that in addition to MEK] and MEK2, most members
in the MK family were enzymatic substrates of LF with an exception of MEKS. For the
most part the results repeated here support this observation. However, my data do not
wholly support their findings. Both MKK4 and MKK7 were previously found to harbor

two LF cleavage sites. The data presented above in this chapter show that in mammalian

cells LF cleaves MKK4 only at the Lys45-Leu46 position but not the Argsg-Phe59

position, and that LF does not cleave MKK7 in cells. A comparison of the LF cleavage

site consensus sequence [(B/P)3_4X3/Al] and the amino acid sequences of the two

reported MKK4 cleavage sites shows that the amino acid sequence of the first LF

cleavage site of MKK4, QGKRKALK45L46, matches the consensus LF cleavage site
. 58 59 .
whereas the second srte, PPFKSTAR F , does not (Table II in Chapter I, see page 45).
. 58 59 . . . .
This supports that Arg -Phe of MKK4 is not a real LF cleavage srte in mammalian

cells, and that the cleavage at the Argsg-Phe59 of MKK4 observed by Vitale et al. was

possibly an artifact resulted from the in vitro experimental system.

Since Vitale’s initial report ten years ago, there have been no further reports of

LF-mediated MKK7 cleavage published in the literature. The in-cell cleavage assays of

63

endogenous MKK7 by LF presented in this chapter were carried out in two different
mammalian cell lines, Chinese hamster ovary cells (CHO K1) and human embryonic
kidney 293 -derived cells (293 FT). The cleavability of either a low expression level of
exogenous MKK7 or of endogenous MKK7 by LF was examined in these two
mammalian cell lines, and no MKK7 cleavage was observed. The data strongly suggest
that MKK7 is not a preferred LF substrate in mammalian cells. Moreover, on careful
analysis of Vitale’s initial report, it is not clear whether MKK7 is cleaved by LP in vitro
[the MKK7 panel of Figure 4 in the report by Vitale et al. (2000)]. In their in vitro
cleavage experiments, LF resulted in a noticeable decrease of GST::MKK3, 4, and 6
Coomassie Blue staining signals. However, LF did not result in any decrease of the
GST::MKK7 signal. Both reported MKK7 cleavage sites were identified by isolating
two very weak Coomassie Blue bands with higher electrophoretic mobility in the gel.
These two truncated MKK7 may alternatively be explained by protein degradation, which
is commonly observed in GST-fusion protein purification. Collectively, the results

presented in this chapter argue against the cleavability of MKK7 by LF.

Unlike MKK4, the amino acid sequences of both cleavage sites of MKK7 match
the consensus sequence of LF cleavage site perfectly (Table II in Chapter I, see page 45).
If the amino acid sequences match the LF cleavage site consensus sequence and MKK7
also contains the conserved the LF-interacting region (Chopra et al., 2003), then why is
MKK7 not cleaved by LF in cells? Perhaps the sequence of the LF cleavage site gives an
answer to this question. The consensus sequence of LF cleavage site is generalized based

on an alignment of all the reported MEK/MKK cleavage sites, including the two of

64

MKK7. If MKK7 is not an LP substrate, the two reported cleavage sites of MKK7
should not be included in the alignment, and the consensus of LF cleavage site may be
different. In other words, if the two reported LF cleavage sites of MKK7 are excluded

from Table 11, then a better consensus sequence of LF cleavage site would be

(B)3X2_4(B/P)1/Al, where B represents basic residues, X represents variable residues, P

represents proline, and Al represents aliphatic residues. In this new consensus sequence,
a proline residue or a basic residue at the Pl position is conserved in MEK] , MEK2,
MK3, 4, and 6, but not in MKK7. Therefore, it is very possible that the amino acid
residue at the Pl position of MEK/MKK is also critical for LP cleavage. Two previous
studies support this possibility. First, Duesbery et al. (1998) have demonstrated a mutant
MEK] with an alanine substitution for this critical proline residue was resistant to LP
cleavage. Second, the critical role of the amino acid at P1 position of MEK/MKK for LF
cleavage is supported by a crystal structure study of LF. Pannifer et al. (2001) have
shown that the side chain of proline or arginine at the P1 position of MEK/MKK sits in

the SI substrate-binding pocket of LP to generate a productive cleavage complex.

This chapter describes the development of a novel experimental system that
allows evaluating sufficiency of an individual MEK/MKK signaling pathway when
multiple other parallel family pathways are inhibited by LeTx. This system was then
utilized and coupled with isoform-specific MEK knockdown to determine both
sufficiency and necessity of MEK] and MEK2 signaling pathways for melanoma cell

proliferation. This will be presented in Chapter III.

65

1.4. Figures

Figure 4. Resistance of VS-MEKcr to LF-mediated proteolysis.

(A) Non-transfected (None), mock-transfected (Mock), and VS-lacZ-transfected
CHO K1 cells, as well as cells transfected with (B) wild-type VS-MEKI or VS-MEchr,
or (C) wild-type V5-MEK2 or V5-MEK2cr, were treated with PA alone control (PA) or
LeTx (LT) in the presence of cycloheximide for 12 h as described in Materials and
methods. Total cell lysates were then harvested and subjected to immunoblotting probed

with the antibodies indicated on the right of each panel. V5 antibody (V 5) and an

antibody against NHz-terminus of MEK] (MEKl (NH2)) were used to detect loss of

NHz-terminal epitopes following LF-mediated proteolysis. Antibodies against the

carboxyl terminus of MEK] (MEKI (COOH)) and MEK2 (MEK2 (COOH)) were used
to detect cleaved MEK] and MEK2, respectively, with increasing electrophoretic

mobility. Antibodies against a-tubulin and B-tubulin were used as loading controls.

66

Figure 4. Resistance of VS-MEKcr to LF-mediated proteolysis.

A None Mock V5-IacZ

PA LT PA LT PA LT
[ - -] vs
E -- --- J MEK1(NH2)
F- : :- ...;. A; q MEK1 (coou)
M a-tubulin

 

 

 

 

 

 

 

None V5-MEK1 V5-MEK1cr V5-MEK2 V5-MEK2cr
PA LT PA LT PA LT PA LT

PA
1 f" -iV5 l- .2...ng
.--

M B-tubulin

 

 

 

 

 

 

 

 

MEK1 (coon)

 

 

 

 

 

 

67

Figure 5. Dual translation of VS-MEK expression vectors.
CHO K1 cells were transfected with plasmids encoding for VS-MEKl or V5-

MEchr, or the modified plasmids in which the original translation initiation codons of
MEK sequences were removed (V 5-MEK1_dATG and V5-MEK1cr_dATG). Non-
transfected cells (None) or cells transfected with VS-IacZ were used as controls. After
transfection, cells were treated with PA alone (PA) or LeTx (LT) in the presence of

cycloheximide for 12 h as described in Materials and methods. Total cell lysates were

then harvested and subjected to immunoblotting probed with antibodies against NH2-

terminus of MEK] (top panel) and GAPDH (bottom panel). VS-MEKI and non-tagged

MEK] are indicated. Lane numbers are labeled at the bottom of the blots.

VS-MEK‘I VS-IIEKt
None V5-lacZ V5-MEK1 V5-MEK1cr dATG an:

 

 

 

 

 

 

PA LT PA LT PA LT PA LT PA LT PA LT

V5-MEK1

: MEK1

 

 

 

   

GAPDH

--~“~~“~~- H”

 

123456789101112

68

Figure 6. Activity of VS-MEKcr.
(A) Phosphorylation and activation of VS-MEKcr in cells. SK-MEL-28 cells

stably expressing VS-MEK or VS-MEKcr were treated with PA alone (PA) or LeTx (LT)
for 24 h. Whole cell lysates were prepared and immunoblotted with anti-
phosphoMEKl/Z antibody to examined the activation status of cellular V5-MEchr and
V5-MEK2cr (upper panel), or with anti-GAPDH antibody as a loading control (lower
panel). (B) Optimization of V5-MEK IP-kinase assay. VS-lacZ, VS-MEKI, V5-
MEchr, VS-MEKZ, and V5-MEK2cr were immunoprecipitated from SK-MEL—28 cells
as described in Material and methods. A decreasing amount (5, 0.5, and 0.05 ul) of the
VS-MEK immunoprecipitates was added into in vitro kinase assay reactions which were
then carried out as described in Material and methods. (C) Kinase activity of VS-MEKcr.
Anti-V5 immunoprecipitates were prepared from SK-MEL—28 cells as described in
Material and methods, and then used for in vitro kinase assays (top panel) in the presence
or absence of active B-Raf and/or recombinant ERK2. Equal amounts of V5
immunoprecipitates were subjected to V5 immunoblotting (bottom panel) as the V5-

MEK input control.

69

Figure 6. Activity of VS-MEKcr.

A ca‘ vs“ 0" 0
“94°“ «6"? «6’6 46"“, sf)"

LT PA LT PA LT PA LT PA LT

 

1!!! Mafia- O ...... O in MEKW-Pqi

 

 

E-cfiu-—--—.--—-“]GAPDH

 

B ‘3'

A“ V5-MEK1 V5-MEK1cr V5-MEK2 V5-MEK2cr

5 5 0.5 0.05 5 0.5 0.05 5 0.5 0.05 5 0.5 0.05

I . . . . 14- ERK2

 

 

 

 

 

 

 

   

 

 

 

 

 

 

Q9
31" V5-MEK1 V5-MEK1cr V5-Iacz V5-MEK2 V5-MEK2cr
B-RafT-+.+.+.+ .+.+-+-+.+-+
ERK2+--++.-++ .-++..++..++
.. 'TTW‘Trf . . . ,4.
l” n. 4ERK2
r..- 2 75 m 7 L I“, __-:_,
-E 3
xxgg
Emmmm
w¥¥¥¥
mmmmm
>>>>>
up..-

 

 

 

70

Figure 7. Resistance of MKK3cr and MKK6cr to LF-mediated proteolysis.
(A) Human melanoma SK-MEL-28 cells stably expressing V5-MKK3 or V5-

MKKBcr were treated with LeTx in an LP concentration-dependent manner (indicated)
for 24h. Total cell lysates were harvested and subjected to immunoblotting probed with
antibodies against V5 epitope (top panel), carboxyl terminal of MKK3 (middle panel),
and a-tubulin (bottom panel). (B) CHO Kl cells were transfected with V5-MKK6, or
V5-MKK6cr. In-cell cleavage assays were performed as described in Materials and
methods. Total cell lysates were then collected and irnmunoblotted with antibodies
against V5 epitope (top panel), carboxyl terminus of M6 (middle panel) and a-tubulin

(bottom panel).

71

Figure 7. Resistance of MKK3cr and MKK6cr to LF-mediated proteolysis.

A V5-MKK3 V5-MKK3cr

 

 

[LF](ngImI): o 1 10 100 o 1 10 100

E 1,, ~‘flfl]<v5-MKK3

. N

H,

‘ _\ _ _ .
I
. illll .. _.1 1,,
, 4 Al.” - 4... ._. L. .

B V5-MKK6 VS-MKKBcr
, PA LT PA LT

[~. . -“I<V5-MKK6

W - vsms

R .. —— < MKK6
. E71 ‘ Cleaved MKK6

 

his is an < V5-MKK3
< MKK3
‘ < Cleaved MKK3

  
 

   

 

 

< a-tubulin

 

 

 

 
 
 

 

1...

 

 

 

 

< a-tubulin

 

72

Figure 8. Resistance of MKK4cr and MKK7cr to LF-mediated proteolysis.
CHO Kl cells were transiently transfected with (A) V5-MKK4, V5-MKK4cr_46,

or V5-MKK4cr_59, or (B) V5-MKK7, V5-MKK7cr_45, or V5-MKK7cr_77. In-cell
cleavage assays were performed as described in Materials and methods. Total cell lysates
were then collected and immunoblotted with antibodies against V5 epitope (top panels),

a-tubulin and B-actin (indicated).

A V5-MKK4 V5-MKK4cr_46 V5-MKK4cr_59
PA LeTx PA LeTx PA LeTx

 

 

firm r—m vme- g V5-MKK

 

 

 

 

— - -- -4 a-tubulin

 

B V5-MKK7 V5-MKK7cr_45 V5-MKK76r_77
PA LeTx PA LeTx PA LeTx

I- ii- ” a a ”I V5-MKK

 

 

 

 

I” 111—. ’ — m -1 a-tubulin

I..- «.u.‘ B-actin

 

 

73

Figure 9. In-cell cleavage of MKK4 by LF.
CHO K1 cells were transfected with MKK4-V5-His6, MKK4_d45-V5-His6, or

MKK4_d58-V5-His6 plasmids. After transfection, cells were split to four dished and
cultured for 12h. Cells in each dish were treated with PA alone (1 ug/ml PA) or LeTx (1
jig/ml PA plus 0.1 ug/ml LF) in the presence of 0.1% DMSO or 10 ug/ml MG-132 for
24h. Total cell lysates were collected and immunoblotted with antibodies against V5

epitope (top panel) and GAPDH (bottom panel).

HKK4-V5-HIOG MKK4_d45-V5-H186 MKK4_d58-V5-Hi88
DMSO M6432 DMSO M6432 DMSO M6432

 

 

 

 

 

 

PA LeTx PA LeTx PA LeTx PA LeTx PA LeTx PA LeTx

  
 
  
 
   

 

, ". «d— MKK4—venue
‘ 1 “3* 4— MKK4_d45-vs-Hiss
« MKK4_dss-vs-Hiss

_ , . . t,
a. .P' I _‘. ' .1'.'{Lt “v

, , , . . *"i ,1 .
wt, «w --_ I‘::.-’v"/3£‘~'-n-!‘-

m.~.
1
.1
- n.

r. .. ‘p
..... ... ~
" " f1.

 

123456789101112

74

Figure 10. Wild-type MKK7 is not cleaved by LF in cells.

Non-transfected (None) CHO K1 cells and cells transfected with a decreasing
amount of V5-MKK7 plasmids (8, 4, 2, and 1 ug as indicated) were treated with PA
alone (PA) or LeTx (LT) for 12h in the presence of cycloheximide as described in

Material and methods. Total cell lysates were collected and immunoblotted with
antibodies against V5 epitope (top panel), NHz—terminus of MEKl (middle panel), and [3-

actin (bottom panel).

V5-MKK7

None Bus 4:19 2mg 1119
mumumumumu

m

 

 

 

 

 

 

   
 

V5-MKK7

 

i MEK1_NT

 

 

- ‘ B-actin

 

 

75

Figure 11. LF does not cleave endogenous MKK7 in mammalian cells.
293FT cells were treated with PA alone (PA) or LeTx (LT) for 24, 48, and 72h

(indicated). Total cell lysates were collected and immunoblotted with antibodies against
the last 20 amino acids of human MKK7 (top panel), NHz-terminus of MEK] (middle

panel), and fi—actin (bottom panel).

 

 

 

76

1.5. Materials and methods
1.5.1. Cell lines and stable cell line establishment.

Chinese hamster ovary (CHO) Kl cells were purchased fi'om ATCC and grown in
Dulbecco’s minimal essential medium (DMEM) supplemented with 10% fetal bovine
serum (FBS) and 50 units/m1 penicillin/streptomycin (Invitrogen). SK-MEL-28 cells
were obtained from CeeTox Inc. and grown in RPMI 1640 medium supplemented with
5% FBS and 50 units/ml penicillin/streptomycin. 293FT cells were purchased from
Invitrogen and grown in DMEM medium supplemented with 10% FBS and 50 units/ml

penicillin/streptomycin (Invitrogen). All the cells were cultured at 37°C in a humidified

5% C02 incubator.

1.5.2. Chemicals and LeTx

Cycloheximide (Sigma) and MG-132 (Calbiochem) were dissolved in dimethyl
sulfoxide (DMSO) (Sigma) at a stock concentration of 100 mg/ml and 10 mg/ml,
respectively. PA and LF were expressed in an attenuated strain of Bacillus anthracis
(BH445) and purified by fast pressure liquid chromatography as described (Bromberg-

White & Duesbery, 2008).

1.5.3. VS-MEKcr construction.

All the plasmids used for transient transfection were prepared by using the
QIAprep® Spin Miniprep Kit (Qiagen). Human MEK] (N M_002755.2) sequence was
PCR-amplified from the pREST-A/MKKI vector, which was a generous gift from
Natalie Ahn (Mansour et al., 1994b). Human MEK2 (NM_030662.2), MKK3

77

 

(NM_145109.2), MKK4 (NM_003010.2), and MKK6 (NM_002758.2) sequences were
PCR-amplified from the following I.M.A.G.E. clones (Open Biosystems). MEK2, clone
#2961198; MKK3, clone #5215093; MKK4, clone #5272439; MKK6, clone #4499772.

Human MKK7 (NM_145185.2) sequences were PCR-amplified from the

TM
pcDNA3/MKK7 vector. Amplified sequences were then inserted into pENTR

. . ® . . . .
Directional TOPO vector (Invrtrogen) according to the manufacturer’s instructions. To

generate MEKcr/MKKcr, aspartic acid residues were introduced into the Pl' position of

LF cleavage sites (the 9th, 11th, 27th and 15th amino acid of MEK], MEK2, MKK3 and
. th th . . th th . .
MKK6, respectively; the 46 and 59 amino acrds of MKK4; 45 and 77 amino acrds

of MKK7.) by using QuikChange® Site-Directed Mutagenesis (Stratagene). All the

wild-type MEK/MKK and MEKcr/MKKcr sequences were verified by DNA sequencing.
V5-MEK/MKK mammalian expression vectors were created by performing LR
recombination reactions to move MEK/MKK sequences from the Entry vectors to the
pcDNA3.1/nV5-DEST Destination vectors (Invitrogen) according to the manufacturer’s

instructions.

1.5.4. Construction of MKK4-V5-His6 and deletion mutants.

To make MKK4 carboxyl terminal fusion, the stop codon of MKK4 sequence was

removed from the MKK4 Entry vector by using QuikChange® Site-Directed Mutagenesis

(Stratagene) with modified PCR reaction in which 10% DMSO was included to solve

secondary structure. A second run of modified Site-Directed Mutagenesis reaction was

78

 

performed to remove amino acid 2-45 and 2-58 of MKK4 from the modified Entry vector
in order to generate the deletion mutant MKK4_d45 and MKK4_d58, respectively. The
coding regions of MKK4 on all the Entry vectors were verified by DNA sequencing.
MKK4-V5-His6 mammalian expression vectors were created by performing LR
recombination reactions to move MKK4 sequences from the modified Entry vectors to
the pcDNA-DEST40 Destination vectors (Invitrogen) according to the manufacturer’s

instructions.

1.5.5. In-cell MEK cleavage assay.
To examine the cleavage of wild-type MEK/MKK or the cleavage resistance of

MEKcr/MKKcr in CHO Kl cells, cells were transfected with V5-MEK/MKK expression

TM
vectors or V5-lacZ (as a control) by using Lipofectarnine 2000 (Invitrogen) according

to the manufacturer’s instructions. After transfection, cells were trypsinized and split into
two dishes, and cells in each dish were treated with either PA alone control (1 ug/ml PA)
or LeTx (1 ug/ml PA plus 0.1 ug/ml LF) in the presence of 10 ug/ml cycloheximide

(Sigma) for 12h. Total cell lysates were collected on ice in RIPA lysis buffer [50 mM
Tris-HCl pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 2 mM Na3VO4, 20 mM

sodium pyrophosphate, 1% sodium deoxycholate, 1% Triton X-100, 0.1% SDS, and l X

EDTA-free protease inhibitor cocktail (Roche)] and homogenized by sonication in ice

TM
bath. Protein concentrations were determined by BCA Protein Assay Kit (Pierce)

according to the manufacturer’s instructions. Total cell lysates were then prepared in

1><SDS sample buffer [47.5% Laemmli Sample Buffer (Bio-Rad) and 2.5% B-

79

 

mercaptoethanol]. Five micrograms of total cell lysates were subjected to
immunoblotting to detect LF-mediated cleavage as described in the results.

To examine the cleavage of MEK], MEK2, and MKK3 in SK-MEL-28 cells
(presented in Figure 6A and Figure 7A), clonal lines of cells stably expressing V5-
MEchr, V5-MEK20r, V5-MKK3cr, or their wild-type counterparts were established
(described in the Material and methods of Chapter 111, see page 143). For examining
cleavages of MEK] and MEK2 (Figure 6A), cells were treated with PA alone control (1
rig/ml PA) or LeTx (1 rig/ml PA plus 0.1 [1ng LF) for 24h. For examining the cleavage
of MKK3 (Figure 7A), cells were treated with LeTx in a LF concentration-dependent
manner (1 pig/ml PA plus 0, 1, 10, 100 rig/ml LF) for 24h. Total cell lysates were

collected and prepared for immunoblottings.

1.5.6. Immunoblotting.

Five micrograms of total cell lysates were separated in 10% Novex® Pre-Cast

Tris-Glycine Gels (Invitrogen). To examine cleavage of endogenous MKK7 in 293FT
cells (Figure 11), 4—20% gradient gels were used. After electrophoresis, proteins were
electro-transferred onto polyvinylidene fluoride (PVDF) membranes (Millipore)
according to the manufacturers’ instructions. Membranes were then soaked in 5% non-
fat milk for 30 min and hybridized with primary antibodies against V5 epitope (Bethyl

Laboratories), NHz-terminus of MEKl (Upstate, #07-641), COOH-terminus of MEKl

(Santa Cruz Biotechnology, SC-219), COOH-terminus of MEK2 (Santa Cruz
Biotechnology, SC-525), phospho-MEK1/2 (Cell Signaling #9154), COOH-terminus of

MKK3 (Santa Cruz Biotechnology, SC-961), COOH-terminus of MKK6 (Epitomics,

80

1 821-1), COOH-terminus of MKK7 (Epitomics, 1949-1), a-tubulin (Sigma, T9026), [3-

tubulin (Sigma, T5201), B-actin (Sigma, A1978), or GAPDH (Cell Signaling, #2118) at
4°C for overnight. Conditions for primary antibody hybridizations were followed
according to the antibody datasheets. After primary antibody hybridization, membranes
were washed three times in TBST buffer (50 mM Tris, 150 mM NaCl, and 0.1% Tween-
20), hybridized with HRP-conjugated secondary antibodies (Kirkegaard & Perry

Laboratories) according to the instructions of the antibodies, and then washed three times

TM
in TBST buffer. Irnmunoblotting signals were then detected by LumiGLO Reagent

and Peroxide (Cell Signaling) and exposed to X-Ray files (Kodak).

1.5.7. Immunoprecipitation
To immunoprecipitate V5-MEK from SK-MEL—28 cells, 5><106 cells were

washed with ice-code PBS (Invitro gen) three times. Whole cell extracts were prepared

on ice in 1 ml of lysis buffer (20 mM Tris-HCl pH 7.5, 150 mM NaCl, 1.5 mM MgC12, 2

mM EGTA, 1% Triton X-100, 2 mM DTT, 10 mM NaF, 1 mM Na3VO4, and 12 mM [3-

glycerophosphate), and then homogenized by sonication in ice bath. Soluble fraction of

cell extracts was isolated by centiifugation at 12,000g at 4°C for 10 min. Protein

TM
concentrations were determined by BCA Protein Assay Kit (Pierce) according to the

manufacturer’s instructions. Two hundred mg of clear extracts were incubated with 25 ul
of agarose-immobilized V5 antibody (Bethyl Laboratories) in a total volume of 500 ill of

lysis buffer on a rotator at 4°C for 12b. The precipitates were then washed twice with the

81

lysis buffer and twice with kinase assay buffer (25 mM Tris-HCl pH 7.5, 5 mM [3-

glycerophosphate, 2 mM DTT, 0.1 mM Na3VO4, and 10 mM MgClz), and resuspended

in 50 ul of kinase assay buffer as the kinase source for the following kinase assay.

1.5.8. In vitro kinase assay.
Five rnicrolitters of prepared V5 immunoprecipitates were incubated on ice with
or without 0.05 ug of active B-Raf (Upstate Biotechnology) and 400 ng of recombinant

ERK2 in a total volume of 2 ul of assay dilution buffer (20 mM MOPS pH 7.2, 25 mM [3-

glycerophosphate, 5 mM EGTA, 1 mM Na3VO4 and 1 mM DTT), and then 3 ul of ATP
mixture [0.5 ul of [y-3ZP]ATP (Amersham; 10 mCi/ml, 3000 mCi/mmole) in 250 uM

ATP and 37.5 mM MgC12] was added. In vitro phosphorylation reaction was carried out

by incubating the reactions in 30°C water bath for 30 min, and then stopped by addition
of 10 ul of 2><SDS sample buffer. Proteins in the reactions were then separated in 10%
Tris-Glycine gels, and ERK2 phosphorylation was visualized by using a FLA-5000
Phospholrnager (Fuji). The input of 5 ul of V5-fusion proteins was shown by

immunoblotting using V5 antibody.

82

Chapter III. Sufficiency and necessity of MEK signaling pathways for melanoma

cell proliferation

1.1. Introduction

MEKl and MEK2 share greater than 85% amino acid homology in their kinase
domains. Therefore, currently available MEK small-molecule inhibitors target both
MEKs. Since the only known substrates of both MEK 1 and 2 are ERK] and ERK2, it
has generally been assumed that MEKl and MEK2 functions are overlapping and
redundant. However, studies of mouse knockouts as well as recent studies in cultured
cells indicate these molecules may have distinct biological functions. Whereas MEKI
deficiency results in embryonic lethality and a reduction in placental vascularization
(Giroux et al., 1999), MEK2 knock-out mice are viable and fertile (Belanger et al., 2003).
In addition, in vitro studies indicate MEK] and MEK2 have non-redundant roles in colon
cancer cell proliferation (Voisin et al., 2008) and epidermal neoplasia (Scholl et al. ,
2009b).

The B-Raf/MEK/ERK pathway is constitutively activated in many cancers
including melanoma. Indeed, more than eighty percent of human melanomas harbor
somatic B-Raf and N-Ras mutations (Tsao et al., 2000; Davies et al., 2002) that cause
constitutive activation of MEK 1 and 2. Elevated MEK 1 and 2 activities promote
melanoma tumorigenesis, angiogenesis, and progression (Cohen et al., 2002;
Govindarajan et al., 2003; Tanami et al., 2004; Wellbrock et al., 2004; Trisciuoglio et al.,
2005). Conversely, biological and small molecule inhibitors targeting MEK 1 and 2
inhibit melanoma cell proliferation and xenograft tumor growth as well as metastasis

(Koo et al., 2002; Collisson et al., 2003; Abi-Habib et al., 2005; Solit et al., 2006).
' 83

Consequently, a great deal of effort has been placed on developing inhibitors of MEK]
and 2 (Roberts & Der, 2007). Despite this, MEK1/2 inhibitors have failed to meet
expectations in clinical trials owing to poor efficacy and unexpected toxicities (LoRusso
et al., 2005b; Eisen et al., 2006; Lee & Duesbery, 2010). This indicates we do not yet
fully appreciate the role of MEK signaling in melanoma. A better understanding of the
relative roles of MEK] and MEK2 may aid in the design of more effective therapies for
treating melanoma and other MEK-dependent cancers.

To test the hypothesis that the functions of MEKl and MEK2 are critical and
interchangeable for melanoma cell proliferation, two series of experiments were
performed. In the first approach, the necessity of MEK] and MEK2 signaling pathway
for SK—MEL-28 cell proliferation was determined using MEK-specific siRNA. In this
experiment, either MEK] or MEK2 was knocked-down by the specific siRNA while the
other MEK isoform and other MKK family proteins were present. In the second
approach, the novel experimental system presented in Chapter II was used to evaluate the
sufficiency of MEK signaling pathways for cell proliferation by making either MEK] or
MEK2 active in SK-MEL—28 cells. Briefly, this system takes advantage of the
MEK/MKK-specific proteolytic activity of anthrax lethal toxin (LeTx). LeTx is a binary
toxin composed of protective antigen (PA) and lethal factor (LF). PA binds to the
anthrax toxin receptors on cell surface to form a complex, which mediates the
internalization of LF into cells (reviewed by Singh et al., 2005). LF proteolytically
inactivates MEK 1 and 2, (Duesbery et al., 1998; Vitale et al., 1998), MKK 3-4, 6 and 7
(V itale et al., 1998; Pellizzari et al., 1999; Vitale et al., 2000) at the amino-terminal

consensus cleavage site, although results presented in Chapter 11 show that MKK7 may

84

not be a preferred LF substrate in marmnalian cells (Chapter 11, section 1.2.5, page 60).
As a result of the cleavage, MEK loses a critical docking domain which is required for
interacting with the downstream substrate ERK (Chopra et al., 2003; Bardwell et al.,
2004). As a consequence MEK signaling is blocked (Duesbery et al., 1998; Park et al.,
2002; Depeille et al., 2007; Ding et al., 2008). Taking advantage of this, LeTx was used
as a global MEK/MKK inhibitor to inhibit both of MEK] and MEK2 as well as most
other MKK signaling pathways in SK-MEL-28 cells. Simultaneously, a cleavage-
resistant form of MEK (MEKl or or MEKZCr) was expressed to rescue the individual
MEK signaling pathway in cells. Sufficiency of MEKl and MEK2 signaling pathways

for melanoma cell proliferation was then evaluated.

1.2. Results
1.2.1. Necessity of MEK] and MEK2 signaling pathways for melanoma cell

proliferation

1.2.1.1. Necessity of MEK] and MEK2 signaling pathways for ERK activation
in melanoma cells

To test the hypothesis that MEKl and MEK2 are redundant and interchangeable,
the necessity of MEK] and MEK2 signaling pathways for ERK activation in melanoma
cells was evaluated first by performing siRNA-mediated MEK knock-down. To do this,
human melanoma SK—MEL-28 cells were transfected with individual siRNA or pools of
siRNAs that specifically target either MEK] or MEK2. These siRNAs were designed

and validated by QIAGEN Inc. (Valencia, CA). The knock-down efficiency of each

85

siRN A was estimated to be 75% or greater based on immunoblotting and the effects of
each siRN A appeared to alter the targeted MEK without affecting the other (Figure 12).
Individual knock-down of either MEKl or MEK2 had no effect on ERK activation in
SK-MEL-28 cells. In contrast, however, pooled siRNA targeting both MEKl and MEK2
did cause a decrease in ERK phosphorylation (Figure 12). This indicates that neither
MEK] nor MEK2 is necessary for ERK activation, and that MEK] and MEK2 can

compensate for each other to activate ERK in SK-MEL—28 cells.

1.2.1.2. Necessity of MEK] and MEK2 signaling pathways for melanoma cell
cycle progression
The necessity of MEK] and MEK2 signaling for cell cycle progression was then

further determined in SK—MEL-28 cells using fluorescence-activated cell sorting (F ACS).

Under normal cell culture conditions, about 70% of SK-MEL-28 cells are in G0/Gl phase.

Whereas knockdown of either MEK] or MEK2 had no effect on cell cycle progression,

the combined knockdown of MEK] and MEK2 resulted in a significant 10% increase in

cells at GO/Gl (p < 0.0001) (Table IV). These data indicate that neither MEKl nor

MEK2 signaling is necessary for ERK activation or cell cycle progression in SK-MEL-28
cells, and that MEKI and MEK2 can compensate for each other to activate ERK and to
promote cell cycle progression in SK—MEL—28 cells. This supports the hypothesis that

MEK] and MEK2 are interchangeable for melanoma cell proliferation.

86

1.2.2. Sufficiency of MEKl and MEK2 signaling pathways for melanoma cell
proliferation

To confirm the feasibility of the novel experimental system (presented in Chapter
II) in SK-MEL-28 cells, V5-MEK1cr, V5-MEK2cr, or their wild-type counterparts (as
controls) were stably transfected into SK—MEL-28 cells. These cells were then treated
with LeTx in a LF concentration-dependent manner (1 ug of PA plus 0, l, 10, and 100

ng/ml of LF) for 24h, and the cellular levels of MEK] and MEK2 were then examined

using immunoblotting. LF-mediated proteolysis caused loss of the NHz-terminal V5

epitope in cells expressing wild-type VS—MEKl and V5-MEK2 within 24 hours, even
when the cells were treated with low concentrations of LF (1 ng/ml). This loss of V5
epitope did not appear to be a result of decreasing cytomegalovirus (CMV) promoter
activity (due to the down regulation of c-jun and c-fos by LeTx treatment) as the
expression level of V5-lacZ in SK-MEL—28 cells remained unchanged in the presence of
LF (Figure 138). In contrast, VS-MEKI or and V5-MEK2cr were much more resistant
and showed only partial cleavage at the highest LF concentration used (100 rig/m1)
(Figure 13A, top panels). These data confirm the earlier results showing that V5-

MEKl or and V5-MEK20r remained intact in V5-MEKcr-expressing CHO K1 cells after
with LeTx treatment (presented in Chapter II). In addition, following LeTx treatment,
MEK] and MEK2 were not detectable in V5-MEK2cr-expressing cells and VS-MEchr-
expressing cells, respectively (Figure 13A, the second panels for MEK] , and the third
panels for MEK2). It was noted that in SK-MEL—28 cells, the V5-MEK expression
vectors also gave the same dual-translation pattern which was described earlier in

Chapter II (see the text in page 54 and Figure 5). Taken together, these results validate

87

observations in CHO K1 cells and confirm MEK] or MEK2 expression can be stably

maintained in the absence of other MKK.

1.2.2.1. MEK2, but not MEK], is sufficient to maintain ERK2 activity

ERK1 and ERK2 are the only identified enzymatic substrates of MEK] and
MEK2. To evaluate the sufficiency of MEKl and MEK2 to phosphorylate ERK in
melanoma cells, I examined the status of ERK phosphorylation when SK-MEL—28 cells
expressed MEKl or MEK2. As expected, ERK2 was more abundant than ERKl in SK-
MEL—28 cells (Figure 13A, the fifth panels), and ERK2 activation was dramatically
inhibited by LeTx in SK-MEL-28 parental cells as well as in cells expressing wild-type
V5-MEK1 or wild-type VS-MEKZ (Figure 13A, the fourth panels). In contrast, V5-
MEKZcr maintained ERK2 phosphorylation in the presence of LeTx. Surprisingly,
ERK2 activation was not maintained in VS-MEchr-expressing cells treated with LeTx,
although V5-MEK1cr was activated in SK-MEL-28 cells and the kinase activity of V5-
MEKl or was confirmed by the earlier IP-kinase assay result. These results indicate that
in SK—MEL-28 melanoma cells, MEK2 alone is sufficient to maintain ERK2 activity in
the presence of LF, but MEK] is not, and that MEK] and MEK2 are not interchangeable

for ERK2 activation in this cellular context.

1.2.2.2. Genome-wide cDN A microarray reveals non-overlapping
transcriptional patterns downstream of MEK] and MEK2
In response to extracellular signals, MAP kinase pathways regulate gene

expression at the translational and/or post-translational level. Therefore, if MEK] and

88

MEK2 are redundant and interchangeable, transcriptional targets downstream of MEKl
and MEK2 should overlap. In contrast, if MEK] and MEK2 are not redundant, they
should have non-overlapping transcriptional effects. To test this, SK-MEL-28 cells were
treated with LeTx, so that (i) VS-IacZ-expressing cells had neither MEKl nor MEK2 (as
both were cleaved by LeTx); (ii) V5-MEK1cr—expressing cells had only MEK] (V 5-
MEchr) but not MEK2; and (iii) V5-MEK2cr-expressing cells had only MEK2 (V5-
MEK20r) but not MEKl (Figure 14A). Under these conditions, endogenous MEK and
other MKK proteins were cleaved and inactivated by LP in LeTx-treated cells (data not
shown). After LeTx treatment, total RNA was collected from the cells, and global gene
expression change was analyzed by using the Agilent 60—mer Whole Human Genome
Microarrays. The microarray data analysis was performed in collaboration with Mr. Karl
Dykema from Dr. Kyle Furge’s laboratory at the Van Andel Research Institute.
Comparing the gene expression patterns in these cells, we found that relative to control
treatment (PA plus LF_E687C, an inactive LF mutant) in V5-lacZ-expressing cells, LeTx
treatment resulted in statistically significant changes in expression of 2,560 genes out of
the 18,359 genes represented on the microarray. Of the 2,560 genes, 268 were rescued in
VS-MEchr-expressing cells, and 1,214 were rescued in V5-MEK2cr-expressing cells.
Surprisingly, when we compared the LeTx effect on these two types of cells, we found
that the rescued genes were incompletely overlapping (Figure 14B). This indicates that
MEKl and MEK2 have both overlapping and non-overlapping downstream targets, and
that MEK] and MEK2 are not absolutely redundant and interchangeable with respect to

gene expression in melanoma cells.

89

1.2.2.3. MEK2cr, but not MEchr, rescued proliferation-related pathways in
LeTx-treated cells

We examined the pathway outputs in LeTx-treated cells by performing a gene set
enrichment analysis (Furge et al., 2007a; Furge et al., 2007b). We found that in SK-
MEL—28 cells LeTx treatment resulted in down-regulation of several transcriptional
signatures, most of which were associated with cell proliferation, such as those of E2F
transcription factor, RNA processing, DNA replication and recombination, and cell cycle
progression, as well as oncogenic pathways. This supports previous findings that LeTx
inhibits melanoma cell proliferation in vitro (Koo et al., 2002; Abi-Habib et al., 2005).
Interestingly, 49 of these signatures were uniquely rescued in V5-MEK20r-expressing
cells whereas no signatures appeared to be uniquely rescued by VS-MEKl cr-expressing
cells (see Table V for representative pathways, and Appendix I for the full pathway list).
This finding indicates that MEKl and MEK2 signaling pathways are not equally
sufficient for gene expression associated with melanoma cell proliferation and suggests

that MEK2 has a more substantial role in driving melanoma cell proliferation.

1.2.2.4. MEK2, but not MEK], is sufficient for melanoma cell proliferation in
vitro.

To test whether MEK2 and not MEK] is sufficient to drive SK-MEL—28
melanoma cell proliferation, additional clonal SK-MEL-28 cells stably expressing the
V5-fusion proteins were established and included in the following experiments. After
SK-MEL-28 cells were transfected with wild-type V5-MEK or VS-MEKcr expression

vectors, three independent stable clones displaying different levels of V5-fusion protein

90

expression were selected (Figure 15). SK-MEL-28 cells stably expressing V5-lacZ were
established as a control. In this set of stably transfected cell lines, expression levels of
V5-MEKcr were comparable to their wild-type counterparts, and expression levels of V5-
MEKI were comparable to V5-MEK2 (Figure 15). These stable cell lines were then
tested for their sensitivity to LeTx in vitro by performing a LeTx toxicity assay. To do
this, cells were cultured in the presence of PA and various concentrations of LF for 72h,
and relative viability (compared to no LF) of LeTx-treated cells was then determined

(Figure 16). Whereas parental SK-MEL—28 cells and the cells expressing V5-lacZ, V5-

MEKI , VS-MEchr, and V5-MEK2 had similar sensitivities to LeTx (with IC50 values

around 0.5 11ng of LF), cells expressing V5-MEK2cr showed a significantly lower
sensitivity to LeTx-induced proliferation inhibition in vitro (Figure 16 and Figure 17).

Importantly, the resistance of V5-MEK20r-expressing cells to LeTx positively correlated

with expression levels of V5-MEK2cr. The IC50 of LF on SK-MEL—28 cells expressing

low, moderate, and high levels of V5-MEK2cr was 3-, 8-, and 21-fold higher than that for
parental cells, respectively (Figure 16 and Figure 17). Furthermore, if resistance to LP is
strictly dependent upon expression of MEK2cr then these cells should retain their
sensitivity to MEK small-molecule inhibitors. To test this, cells were tested for their
sensitivity to two MEK inhibitors, U0126 and PD 1843 52, in a dose-dependent manner.
Consistent with the hypothesis, V5-MEK20r-expressing cells were as sensitive as other
V5-MEK-expressing cells to the MEK inhibitors U0126 and PD184352 (Figure 17).
Taken together, these experiments demonstrate that SK-MEL-28 cells expressing V5-

MEK2cr are relatively resistant to LeTx-induced proliferation inhibition in vitro, and

91

indicate that the MEK2, but not MEK] , signaling pathway is sufficient for SK-MEL-28

melanoma cell proliferation in vitro.

1.2.2.5. Neither MKK3 nor MKK6 is sufficient for SK-MEL-28 cell
proliferation in vitro.

It was recently reported that cross talk between ERK and p38 MAPK stimulates
melanoma proliferation (Estrada et al. , 2009). Despite the low basal level of p38 MAPK
activity in SK-MEL-28 cells under the cell culture conditions that were used (Figure 18),
the sufficiency of MKK3 or MKK6 signaling pathways for SK-MEL-28 cell proliferation
was tested. When SK-MEL-28 cells stably expressing V5-MKK30r or V5-MKK6cr were
tested for their sensitivities to LeTx-induced proliferation inhibition in vitro, none of
these stable cell lines were resistant to LeTx (Figure 19). This result indicates that

neither MKK3 nor MKK6 is sufficient for melanoma cell proliferation in vitro.

1.2.2.6. MEK2 signaling is sufficient for anchorage-independent growth

The results presented above show the sufficiency of the MEK2 signaling pathway
for melanoma SK-MEL-28 cell proliferation in vitro. Next I attempted to determine the
sufficiency of MEKl and MEK2 signaling pathways for melanoma tumor growth in vivo
using a xenograft mouse model. However, as shown in section 1.2.3 of this chapter, I
was not able to answer this question due to technical limitations of the xenograft model.
Despite this, the xeno graft experiment results revealed valuable information regarding the

in viva targeting of LeTx. This part of the data is presented in the next section.

92

Using a soft agar colony forming assay as an alternative strategy, I evaluated the
ability of these cells to grow in an anchorage-independent fashion. Parental SK-MEL28
as well as SK-MEL28 cells expressing VS-lacZ, V5-MEK and VS-MEKcr readily
formed colonies (>50 uni diameter) within 21 days (Figure 20A, top panels). However,
in the presence of LeTx, colony formation of all cell types tested was inhibited, except
for V5-MEK2cr-expressing cells (Figure 20A, lower panels). When these results were
quantified, it was noted that expression of V5-MKK2cr in SK—MEL—28 cells resulted in a
significant rescue of 70% of the colony formation in the presence of LeTx (Figure 20B).
This result, in agreement with the proliferation results presented earlier, demonstrates that
the MEK2 signaling pathway alone is sufficient for anchorage-independent growth of

SK-MEL-28 cells.

1.2.3. A xenograft model to test the sufficiency of MEK] and MEK2 signaling
pathways for melanoma tumor growth in viva

As discussed in the General Introduction (Chapter I), LeTx possesses a potent
inhibitory activity on tumor growth in vivo in different cancer xenograft models
(Duesbery et al., 2001; Koo et al., 2002; Depeille et al., 2007; Ding et al., 2008). A
similar approach was used in this dissertation project to determine the sufficiency of

MEK signaling pathways for melanoma tumor growth in vivo.

1.2.3.1. SK-MEL—28 xenograft tumor growth
To grow SK-MEL-28 xenograft tumors in nude mice, cells were subcutaneously

injected into the right side of the dorsalateral area of a group of athymic nude mice.

93

Tumor volumes were measured every two days by using a caliper. Figure 21 shows a
representative SK—MEL—28 xenograft tumor growth curve. Shortly after subcutaneous
inoculation, lesions were measurable due to swelling. The swellings generally
disappeared within two weeks and the volume of the lesions remained not measurable for
the next five weeks. Seven weeks after cell inoculation, xenograft tumors started
growmg in an exponential manner (Figure 21). Although considerable variation of tumor
size between mice was normally observed in this xenograft model, the experimental
reproducibility was revealed when tumor growth curves from multiple independent

experiments were overlaid on the same plot (Figure 22).

1.2.3.2. LeTx systemic treatment of SK—MEL—28 xenograft tumors
In this dissertation project, an established approach for LeTx systemic treatment

of cancer xenograft tumors in nude mice was used. Xenogratt tumors were allowed to

3 . . . . . .
grow to an average volume of 50 mm , at which time mice were divrded into two groups.

Mice in each group were then intravenously injected with LeTx (PA plus LP) or control
(PA plus LF_E687C) at a dosage of one standard dose (1 XSD equals 10 ug PA plus 2 ug
LP or LF_E687C) every two days for a total of six injections (6 standard doses).
Previous studies have shown that this treatment has no obvious adverse effects on the
health of athymic nude mice while xenograft tumor growth is inhibited (Abi-Habib et al.,
2006a; Depeille et al., 2007; Ding et al., 2008; Huang et al., 2008). As shown in Figure
21, SK-MEL-28 xenograft tumor growth was also inhibited by the same systemic LeTx

treatment.

94

1.2.3.3. Sensitivity of MEKcr-expressing xenograft tumors to LeTx systemic
treatment

The inhibitory effect of LeTx on SK-MEL-28 xenograft tumor growth suggests
that MEK signaling pathways are required for melanoma tumor growth in viva. Based on
this, sufficiency of MEK signaling pathways for SK-MEL-28 xenograft tumor growth
should be testable by expressing MEKcr proteins in xenograft tumors. If the MEK] or
MEK2 signaling pathway individually is sufficient for SK-MEL-28 xenograft tumor
growth in viva, tumors expressing VS-MEchr or V5-MEK2cr are expected be resistant
to the growth inhibition effect of LeTx systemic treatment. To test this, SK-MEL-28
parental cells or cells expressing high levels of V5-MEK1 , VS-MEchr, V5-MEK2, or
V5-MEK20r were subcutaneously injected into groups of athymic nude mice to establish

SK-MEL-28 xenograft tumors expressing each of these VS-fiision proteins. After these

tumors reached an average volume of 50 m3, mice received LeTx or control (PA plus

LF_E687C) systemic treatment as described. As shown in Figure 23, VS-MEchr-
exressing tumors were still sensitive to LeTx although a slight decrease in sensitivity was
observed when compared with parental and wild-type V5-MEK1-expressing tumors.
Similarly, tumors expressing V5-MEK2cr were also sensitive to LeTx systemic treatment
of LeTx (Figure 24). These results suggest that although the MEK2 signaling pathway
alone was sufficient to drive SK-MEL-28 cell proliferation in vitro, neither MEKl nor

MEK2 was sufficient for SK-MEL—28 xenograft tumor growth in viva.

95

1.2.3.4. Dose-dependent effects of LeTx on SK-MEL-28 xenograft tumor
growth

When the MEKcr proteins were tested for their resistance to LF-mediated
proteolysis in SK-MEL—28 cells, restored cleavage of MEchr and MEK2cr by LF was
observed at the highest concentration of LF (Figure 13A, top panels). Therefore, it was
possible that elevated LeTx levels in SK-MEL-28 xenograft tumors were sufficient to
overcome the cleavage resistance of MEKcr and masked their sufficiency for tumor
growth. To test this, a LeTx dose-dependency xenograft study was conducted. In this
experiment, cells were inoculated into groups of 20 mice. After tumors were established,
mice were divided into four groups, and mice in each group were intravenously injected
with either l><SD control (PA plus LF_E687C) or a decreasing dose of LeTx (1 XSD,
0.5XSD, and 0.25XSD). Systemic treatment with LeTx inhibited SK-MEL-28 parental
xenograft tumor growth in a dose-dependent manner (Figure 25, parental tumors),
indicating that the one standard dose of LeTx used in the previous experiments was not
an over dosage. Under these conditions, no obvious therapeutic resistance of V5-
MEKZcr-expressing tumors was observed when compared with parental or wild-type V5-
MEK2-expressing tumors (Figure 25). This result excluded the possibility that an
exposure of high concentration of LeTx proteolytically inactivated VS-MEKcr and
masked the potential resistance of V5-MEK20r-expressing tumors, and confirmed that
MEKI or MEK2 signaling pathway individually was not sufficient for melanoma tumor

growth in viva.

96

1.2.3.5. Loss of VS-MEKcr expression in SK-MEL—28 xenograft tumors

To confirm the finding that MEKl or MEK2 signaling pathway was not sufficient
for SK-MEL-28 xenograft tumor growth in viva as presented above, MEK cleavage as
well as the cleavage-resistance of MEKcr in tumor cells were examined. To do this, at
the end of the xenograft experiment, mice were euthanized and the xenograft tumors were
dissected. Tumor lysates were then prepared and subjected to immunoblotting.
Surprisingly, a loss of VS-MEKcr expression, both of V5-MEK1cr and V5-MEK2cr, in
the xenograft tumor cells was observed at the end of the experiment (about three months
from cell inoculation) (Figure 26). The loss of V5-MEKcr expression was independent
from LeTx systemic treatment as it was also observed in control-treated tumor cells.
Before I could accept this as an explanation for why V5-MEKcr-expressing tumors were
not resistant to LeTx, I needed to confirm that MEK proteins were indeed cleaved in

LeTx-treated tumors.

1.2.3.6. LeTx systemic treatment does not cause MEK cleavage in tumor cells.
When ERK activation status was further examined in the xenograft tumor lysates,
an unexpected result was observed: ERK activation was maintained in LeTx-treated
tumor cells even if the treated tumors were responding (Figure 26). In addition, LeTx
treatment did not result in a loss of V5 epitope immunoblotting signal in the lysates
prepared from tumor cells expressing wild-type V5-MEKs (Figure 26). This observation
raised a possibility that for some unknown reason LF did not cleave MEK in SK-MEL—28
tumor cells, and led to a further examination of cleavage of endogenous MEK in LeTx-

treated tumors. Surprisingly, no cleavage of endogenous MEK was observed in LeTx-

97

treated SK-MEL-28 parental xenograft tumor cells (Figure 27). Two possibilities could
explain this surprising result. First, the endogenous MEK in SK-MEL-28 xenograft
tumor cells became biochemically uncleavable by LF due to an unknown reason. Second,

LF was not delivered into the tumor cells when LeTx was administrated intravenously.

To test these two possibilities, the 6th LeTx or control injection for half of the mice in the

xenograft experiment was switched fi'om intravenous injection to intratumoral injection
by which tumor cells would be directly exposed to LeTx. One day after the last injection,

tumors were dissected and tumor lysates were prepared to examine endogenous MEK

cleavage by immunoblotting. As shown in Figure 27, a loss of endogenous MEKl NH2-

terminal epitope and a slight electrophoretic mobility shift of cleaved MEK] (indicating
MEKl cleavage by LF) were observed in tumor cells only when the last LeTx injection
was administrated intratumorally. In addition, in those tumor cells whose endogenous
MEK was cleaved, ERK activation was inhibited. This result indicates that MEK did not
become uncleavable by LF in xenograft tumors, but rather that systemic injection of
LeTx is not able to deliver LF into tumor cells. Therefore, the inhibitory effect of LeTx

on xenograft tumor growth must be mediated by a non-tumor compartment.

1.2.3.7. Sufficiency of MEK signaling pathways for SK-MEL-28 tumor
growth in viva was not testable due to technical difficulties.

As shown above, after three-month growth in nude mice, SK-MEL-28 xenograft
tumors cells lost V5-MEKcr expression. In addition to this, it was also found that LeTx
was not delivered into tumor cells if injected intravenously. This makes the sufficiency

of MEK signaling pathways for melanoma tumor growth in viva non-testable. However,

98

clinical-valuable information was also obtained from this unexpected observation. This

will be discussed later in this chapter.

1.3. Discussion

Kinase inhibitors have been extensively used as tools to study signal transduction
pathways. However, most kinases share high sequence homology and similar structures
with their family members, making it difficult to discriminate between kinase isoforms
using small molecule inhibitors. Although RNA interference can specifically target one
of the protein isoforms, concerns about incomplete protein knockdown can confound
interpretation, especially when these proteins are highly expressed or slowly turned over
in cancer cells. In addition, in order to study the sufficiency of an individual protein
belonging to a family which has several members and isoforms (e.g. the MEK/MKK
protein family), multiple protein members in the family need to be simultaneously
inhibited or knocked down. Combining several siRNA molecules decreases the
efficiency of RNA interference and increases the likelihood of off-target effects. Due to
these limitations, studies of the function of a protein are restricted to evaluating necessity
but not sufficiency. This impedes a better understanding of individual protein functions,

especially for those involved in complicated signaling networks.

In this Chapter, 3 novel experimental system which allows one to determine the
sufficiency of MEK/MKK signaling pathways for a specific phenotype or cellular

function was presented. This system utilizes the specific proteolytic activity of anthrax

99

lethal toxin (to globally inhibit multiple MEK/MKK signaling pathways) and a set of
mutant MEK/MKK proteins that are resistant to LF-mediated proteolysis (to rescue one
specific MEK or MKK signaling pathway in cells treated with LeTx). This system is
applicable to most mammalian cells because anthrax toxin receptors are ubiquitously
expressed in many different types of cells (Bradley et al., 2001; Scobie et al., 2003).
With an appropriate phenotype as a readout (e. g., proliferation in this study), this system
is a powerful tool allowing dissection of the roles of each MEK/MKK signaling pathway.
Since LF efficiently cleaves and inactivates multiple members in the MK family and
inactivates the three canonical MAPK pathways, this experimental approach eliminates
the potential cross talk between MKK family proteins. To the best of my knowledge, this
is the only available approach for evaluating the sufficiency of individual MEK/MKK

signaling pathways.

Combining RNA interference technology with this novel experimental system,
both necessity and sufficiency of MEKl and MEK2 signaling pathways for melanoma
cell proliferation were determined in order to test the hypothesis that MEKl and MEK2
signaling pathways are interchangeable and redundant for melanoma cell proliferation.

In the first series of experiments, MEK-directed siRNA were used to selectively knock-
down either MEKl or MEK2, or both, in human melanoma SK-MEL-28 cells. Doing
this I found that only simultaneous knock-down of both MEK] and MEK2 could block
ERK activation and cell cycle progression. This indicates that neither MEK] nor MEK2
is necessary for melanoma cell proliferation, and that they can compensate for each other.

These results are consistent with the hypothesis that MEK] and MEK2 are

100

interchangeable and redundant for melanoma cell proliferation. Other studies support
this conclusion. For example, Scholl et al show that stepwise deletion of MEK 1 or MEK
2 alleles reduce Ras-induced epidermal hyperplasia to a similar extent (Scholl et al.,

2009a).

In the second series of experiments, VS-MEKcr constructs were stably expressed
in SK-MEL28 cells so that when the cells were treated with LeTx, multiple endogenous
MEK and MK were inactivated with the exception of the cleavage-resistant mutants
which remained intact and capable of signaling. Doing this, I found first that MEKl and
MEK2 are not equally sufficient for ERK activation in SK-MEL-28 cells. MEK2
signaling pathway alone was sufficient for ERK activation, but MEK] was not. This led
me to use whole genome cDNA microarrays to examine the transcriptional patterns
downstream of MEK] and MEK2. This approach revealed the non-overlapping
transcriptional effect of MEKl and MEK2. Further, collaborating with the
bioinformatics group at the Van Andel Research Institute, we performed a gene set
enrichment analysis to look for transcriptional signatures that can be rescued by V5-
MEKl or and V5-MEK20r. We found that LeTx treatment down-regulated many
proliferation-associated transcriptional signatures, many of which were able to be rescued
only by V5-MEK2cr but not by V5-MEK1 cr (Appendix I). Although another group of
signatures was rescued by both of V5-MEK1cr and V5-MEK2cr (Appendix II), in vitro
assays confirmed MEK2 signaling was sufficient for SK-MEL-28 cell proliferation but
MEKl signaling was not, suggesting that (1) rescuing this group of transcriptional

activities was not sufficient to rescue cell proliferation and/or (2) these transcriptional

101

activities were rescued by VS-MEchr to a less extent. Consistent with the in vitro
proliferation assay, soft agar assays showed that MEK2 signaling was sufficient for the
anchorage-independent growth of SK-MEL-28 cells, and V5-MEchr expression was not
sufficient to rescue colony formation of SK-MEL-28 cells in the presence of LeTx.
Collectively, these findings lead me to conclude that (1) MEK2 signaling pathway is
sufficient for ERK activation, in vitro proliferation, and anchorage independent growth of
SK-MEL—28 melanoma cells, whereas MEK] is not, and (2) though they are critical for
SK-MEL28 cell proliferation, MEK l and MEK 2 are not interchangeable. Additional
studies provide compelling evidence that under some conditions MEKl and MEK2 are
not functionally interchangeable. During mouse embryonic development, loss of MEKl
expression is lethal, likely because of reduced blood vessels in the labyrinthine layer of
the placenta (Giroux et al., 1999). In contrast, MEK2 is dispensable as MEK2 deficient
mice are viable and fertile (Belanger et al., 2003). Voisin et al (2008) have shown that
although both of the constitutively active MEKl and MEK2 isoforms individually are
sufficient to transform normal intestinal epithelial cells, shRNA-mediated MEK2
knockdown has much stronger inhibitory effect on colon cancer cell proliferation than
MEK] knockdown does, leading them to conclude that MEKl and MEK2 may be
differentially regulated or may target distinct effector pathways in certain cellular and/ or
genetic contexts. Scholl et al (2009b) demonstrated that MEKl but not MEK2 is
required for DMBA/'1‘ PA-induced benign epidermal papilloma formation. Also,
Catalanotti et a1 (2009) have recently reported that MEK] and MEK2 form heterodirner

complexes in which MEKl serves as a negative regulator of MEK2.

102

These two sets of experiments led to apparently contradictory conclusions. In the
first set (the necessity study) MEK] and MEK can compensate for each other, whereas in
the second MEKl cannot compensate for loss of MEK2. In fact, these are not truly
contradictory as the two experimental approaches are fundamentally different. In the
siRNA experiments, only one of the MEK/MKK signaling pathways was inhibited,
whereas in the LeTx plus MEKcr experiments, multiple MEK and MK pathways were
blocked. These two experimental strategies resulted in different cellular backgrounds.
Taking this into consideration, these results can be explained by formulating a new
hypothesis: while MEK2 is sufficient for SK-MEL28 melanoma tumor growth, MEKl
can compensate for loss of MEK2 only in the presence of an as yet unidentified factor.
This factor is likely an MKK or is regulated by an MKK since the sufficiency of MEK2
is revealed only in the presence of LF, an MKK-specific protease. However, this factor
likely is not p38 MAPK or JN K since the basal activities of these kinases are too low to
be detected in SK-MEL-28 cells by immunoblotting (Figure 18). Identification of this
LF-sensitive factor may have important clinical consequences since its targeted
inactivation would block the ability of MEKl to compensate for loss of MEK2 activity

and as a consequence increase the dependency on signaling through MEK2 for survival.

The novel experimental system presented in this dissertation relies on MEKcr
rescue. Therefore, it is important to make sure that the aliphatic-to-aspartic acid mutation
does not alter the activity of MEK. Thus far, the only known biochemical activity of
MEK is its kinase activity. In addition, in order to be an active kinase, MEK needs to be

phosphorylated and activated by a MEK kinase such as B-Raf. The validation

103

experiments (Figure 6) presented in Chapter 11 indicate that (1) both MEK] er and
MEK2cr are phosphorylated in SK-MEL—28 cells, and (2) both MEchr and MEK2cr are
capable of phosphorylating ERK2 in an IP-kinase assay. These results indicate that the
aliphatic-to-aspartic acid mutation introduced to MEKl and MEK2 neither impairs
MEK’s kinase activity nor interferes with the RaszEK and MEKzERK interactions.
However, it may be argued that the in vitro interaction between MEKcr and ERK does
not accurately represent the physical interaction occurring in cells, although the amount
of V5-MEK immunoprecipitates added to the in vitro kinase assay reactions was not
saturating (as shown in Figure 6B). 1 reasoned that if the CR mutation made MEK
incapable of interacting and phosphorylating ERK in cells, then (1) expressing V5-
MEKcr should be the same as expressing V5-lacZ, and (2) the transcriptional changes
caused by LeTx treatment in V5-MEKcr-expressing cells should be very similar to that in
VS-lacZ-expressing cells. To test this, we performed a clustering analysis on the
microarray data (see section 1.2.2.2 in page 88) to compare the global transcriptional
changes in LeTx-treated SK—MEL-28 cells expressing V5-lacZ, V5-MEchr, or V5-
MEK2cr. The clustering analysis revealed a higher similarity between LeTx-treated V5-
MEchr-expressing cells and V5-MEK2cr-expressing cells than VS-MEchr-expressing
cells and VS-IacZ-expressing cells (Figure 28), indicating that the introduced cleavage-
resistant mutation did not disrupt MEK] function. This result argues against the
possibility that the CR mutation disrupts the interaction of MEK with ERK, and supports
the interpretation presented in section 1.2.2.1 (page 88) that MEK] individually is not

sufficient to phosphorylate ERK in SK-MEL-28 cells.

104

As discussed in Chapter I, LeTx treatment has shown very promising anti-cancer
activities in cultured tumor cell lines as well as xenograft models (Duesbery et al., 2001;
Koo et al., 2002; Depeille et al. , 2007; Ding et al., 2008; Huang et al., 2008; Rouleau et
al., 2008). In addition, it has become a very powerful tool for studying MEK/MKK
signaling pathways and their roles in human diseases. As a global MEK/MKK protease,
LF cleaves multiple MEK/WK proteins and LeTx treatment results in a global
inhibition of the ERK, p38 MAPK, and JN K pathways. In this chapter, I presented an
effort to combine LeTx systemic treatment and V5-MEKcr-melanoma xenograft model
for evaluating the sufficiency of MEKl and MEK2 signaling pathways for melanoma
tumor growth in viva. Although this was eventually not testable due to technical
limitations, the series of xenograft experiments have provided valuable information. This

is discussed as follows.

As presented in this chapter, it took about 3-4 months to complete an SK—MEL-28
xenograft experiment. This was due to the 7-to-8-week dormancy before tumors started
growing. This relatively long-term tumor dormancy was not commonly observed in the
literature and it does not seem to be mouse strain- or animal facility-specific as other
cancer cell-derived xeno graft tumors grown in the same strain of athymic nude mice at
the Van Andel Research Institute do not require this long-term tumor dormancy.
Therefore, it appears to be cell line-specific. On the other hand, however, this long-term
tumor dormancy was not observed in other reports when SK-MEL-28 cells were
xenotransplanted (Rouleau et al., 2008). It seems that this tumor dormancy is specific to

the SK-MEL-28 cells that we obtained. Although the molecular mechanism of this tumor

105

dormancy is not clear, it has been hypothesized that during this period of time tumor cells
are switching to an angiogenic phenotype, which is essential for tumor growth [reviewed

by Naumove et al. (2006)].

When VS-MEK-expressing cells were xenografted, a few interesting observations
were obtained. First, V5-MEK-expressing xenograft tumors required shorter terms of
dormancy (Figure 23 and Figure 24). This was particularly obvious for V5-MEchr-
expressing tumors which bypassed the dormancy and directly grow up after tumor
inoculation (Figure 23). Second, whereas VS-MEchr-expressing tumors required a
shorter term of dormancy than wild-type VS-MEKl-expressing tumors (Figure 23), V5-
MEK2cr-expressing tumors required a longer term of dormancy than wild-type V5-
MEK2-expressing tumors (Figure 24). Third, these differences appear to be only in
tumor dormancy, not tumor growth rate, as these different xenograft tumors had similar
exponential growth rates after they broke dormancy. It is possible that MEK
overexpression reduced the period of tumor dormancy (i. e. , the MEK signaling pathway
positively regulates the development of the tumor microenvironment), and MEKl and
MEK2 might have different effects on this process. However, since V5-MEKcr
expression was lost during the course of xenograft tumor growth, it is not clear if this
shorter period of tumor dormancy was due directly to MEK expression level. Besides,
since each SK-MEL-28 stable clone was isolated from a single cell colony by selection
with antibiotics, the possibility of clonal effects cannot be excluded. To directly address

this question, VS-IacZ-expressing cells should be xenografted as a control, and other

106

moderate— and low-MEK expressors should be included to test a correlation with MEK

expression levels.

As shown in Figure 26, the expression levels of V5-MEKcr were dramatically
decreased in xenograft tumors during the course of subcutaneous growth and the
expression level of V5-MEKcr was not comparable to its wild-type counterpart. This is
one of the reasons why it was not possible to test the sufficiency MEK signaling
pathways for SK—MEL-28 xenograft tumor growth in viva. Although it was not clear
when the tumors started losing the VS-MEKcr expression, this loss of V5-MEK
expression appears to be MEKcr-specific. Possibilities to explain this include (1) the
aliphatic-to-aspartic mutation makes MEK less stable, and (2) wild-type MEK and
MEKcr have different transcriptional effects which have an autoregulatory feedback
effect. Both of these possibilities appear to be culture environment-specific since this
loss of VS-MEKcr expression was not observed in cultured cells. On the other hand, the
possibility of clonal effects cannot be excluded since only one (the high expressor) of the
three stable clones in each set was inoculated for xenograft tumor establishment and it
was not clear if this loss of V5-MEKcr expression in xenograft tumors would also be
observed if the moderate- or low-expressors were also inoculated. To address this, other
stable clones should be included in this study and expression levels of VS-MEK and V5-
MEKcr in tumor cells should be examined. In addition, in order to understand when the
VS-MEKcr expression was lost in tumor cells, more nude mice should be included in the

study and tumors should be dissected at different time points during the study.

107

It was assumed that systemic LeTx treatment via tail vein injection delivered
LeTx to xenograft tumors, LF was internalized into tumor cells, and LF cleaved
MEK/MKK proteins and blocked multiple MEK/MKK signaling pathways in tumor cells.
It was further assumed that as a result of this, tumor cell proliferation was inhibited, and
in turn, tumor growth was inhibited. However, a surprising observation fi'om the SK-
MEL—28 xenograft model and LeTx systemic treatment was that the endogenous MEK
was still intact in LeTx-treated tumor cells even if the tumor growth was inhibited (Figure
26 and Figure 27). This resistance of MEK to LF cleavage was not due to a change in
MEK cleavability per se as the endogenous MEK was still sensitive to LF cleavage if
LeTx was directly injected into tumor cells via intratumoral injection (Figure 27). If
endogenous MEK proteins were not cleaved by LeTx in tumor cells, then the special
cellular context in which only one individual MEK pathway was active in tumor cells
was not achievable. This was another reason why I was unable to determine the
sufficiency of MEK signaling pathways for melanoma tumor grth in viva. However,
this observation provided valuable information. This result clearly indicates that LeTx
systemic treatment inhibits xenograft tumor growth without targeting tumor cells. If
LeTx does not target tumor cells, why can it inhibit tumor growth? It is possible that
LeTx systemic treatment inhibits xenograft tumor growth by inhibiting tumor-associated
endothelial function, considering that endothelial cells are potential targets of LeTx
(Kirby, 2004; Warfel et al., 2005; Alfano et al., 2009) and that LeTx treatment decreases
inicrovessel density in xenograft tumors (Duesbery et al., 2001; Depeille et al., 2007;
Ding et al., 2008). This gives valuable clinical information: targeting tumor-associated

endothelial cells may be sufficient to inhibit xenograft tumor growth, and targeting

108

melanoma-associated endothelium could be another strategy to treat melanoma. In
addition, this raises an alternative hypothesis that MEK signaling pathways in melanoma-

associated endothelium are essential for tumor survival in viva.

In summary, using complementary experimental approaches 1 have tested the
redundancy of MEKl and MEK2 signaling in the in vitro proliferation of SK-MEL-28
melanoma cells. I observed that (1) neither MEKl nor MEK2 is necessary for melanoma
cell proliferation and (2) MEK2, but not MEKl, is sufficient for melanoma cell
proliferation. These results lead to an alternative hypothesis that while MEK2 can
compensate for loss of MEKl , MEKl can compensate for the loss of MEK2 function
only when an unidentified factor is present. This study provides novel insight into the
complicated interplay between MEK] and MEK2 that may have significant clinical

impact.

109

1.4. Tables and figures

Table IV. Percentage of G1 population of MEK siRNA-treated SK-MEL-28 cells.

 

 

SiRNA experiments G1 population (% i S.D.)*
None 69.45 :i: 0.91
Lipid only 69.68 i 1.21
Non-silencing 70.49 i 0.16
MEKl siRNA-7 70.45 i 3.27
MEK] siRNA-8 71.30 i: 3.07
MEKlsiRNA-ll 70.43 i 1.83
MEK] siRNA-7811 72.92 i 1.71
MEK2 siRNA-9 70.65 i 0.48
MEK2 siRNA-l4 71.21 :1: 0.86
MEK2 siRNA-914 71.53 i: 1.24

P1 (MEKl siRNA-811 + MEK2 siRNA-914) 81.33 i 1.48 **
P2 (MEKl siRNA-7811 + MEK2 siRNA-914) 78.24 a 3.41 "

 

* Data were obtained from at least three independent experiments.
** p < 0.0001 , determined by one-way ANOVA followed by Bonferroni post-hoe
analysis.

110

.Euo.>wo_ofioacow.333\\fitn ”Sac—3:0 080V 00 am.n.nohambcwmmahomwhco.Bfiumfipmofi.333\\uat: ”flow 050 memmE c

 

 

 

.256 55 532w 839
a. 96: £8 mammoaxo.8_v~m2-m> B 888 05 was $585on 3508:: “cacaomowfi 8:: Bow 3:350 083 85:83 a
H 595%? E 338on mm madamcwmm anzntomcab mo am: 03388 23. a
$3556 8m 250 mean: Rs 9: 5m- 8228 296 =8 £6
bfiuoalomfiommgbgvtoflo: Z
”am 280 mama:
x88 :3 a being «am new on. mam- Sage scagaéaoaeaz
o L 308 N00 32m mm; cod- coumfinaooou <75
$248 moo ow. m Se 8%. 288on neaoem
3808 N00 3am mmd vim- wEmmooofi SAM“
SIEQENmézsgmmooom ease
com 0:00 mama—z amd 84 £6. 53:8qu EM .3an .Emm
GSA :3 3 EB «am as N? E H- as:
9 855.8% Sam: .8552 Ncu~-m> a meow oaagmxaspfimfimobdgwmm
-m> -m>
a 85:82

 

4:25.35 what— »: wauooba banana—Mme. 93 «a5 non—5.3»? Buewntcmgh cognates—HE .> 93.;

111

Figure 12. Necessity of MEK] and MEK2 signaling pathways for ERK activation in
SK—MEL—28 cells.

SK—MEL—28 cells were transfected with nothing (N); lipid only (L); non-silencing
control siRNA (NS); MEK-specific siRNA (MEKl siRNA 7, 8, or 11, and MEK2 siRNA
9 or 14), or pools of MEK-specific siRNA (MEKI siRNA 7+8+11 or MEK2 siRNA
9+14) as described in Materials and methods. As controls, cells were transfected with
combinations of both MEKl- and MEK2-specific siRNA (P1, MEKl siRNA 8+11 plus
MEK2 siRNA 9+14; P2, MEK] siRNA 7+8+11 plus MEK2 siRNA 9+14). After
transfection, cells were trypsinized and split to two dishes for cell lysate collection and
for cell cycle analysis (results shown in Table IV). Seventy-two hours later, whole cell
lysates were harvested and immunoblotted. The efficiency of siRNA-mediated MEK
knock-down was examined by immunoblottings with antibodies against MEK] or MEK2.
ERK activation was detected by antibodies specifically against phospho-ERK. Total
ERK expression was detected by ERK antibody as a control. Antibody against GAPDH

was used as an equal loading control.

' 112

Figure 12. Necessity of MEK] and MEK2 signaling pathways for ERK activation in

SK—MEL-28 cells.

MEK1 siRNA +
MEK1 siRNA MEK2 siRNA MEK2 siRNA

 

L NS 7 8 1178119 14 914 P1 P2

[11" .1... 1......

--..:“’"‘ .. JMEKZ

 

 

 

 

3'” , ..

 

I--..----Q- - ~ mu] ERK1I2-PO4

z .'3;= 3 33.] ERK1I2

 

 

 

 

 

 

 

113

Figure 13. Individual MEK signaling in LeTx-treated SK-MEL-28 cells.
SK-MEL—28 parental cells and cells stably expressing (A) VS-MEK or V5-
MEKcr, or (B) V5-lacZ were treated with LeTx (1 ug/ml PA plus 0, 1, 10, or 100 11ng
LF) for 24 h. Total cell lysates were then harvested and immunoblotted with antibodies
against the V5 epitope to confirm the cleavage resistance of VS-MEKcr (top panels).
Antibodies against the carboxyl terminus of MEK] (second panel) and the carboxyl
terminus of MEK2 (third panel) were used to demonstrate individual MEK expression in
LeTx-treated cells. Antibodies against phospho-ERKl/Z (fourth panel) and total ERKl/Z
(fifih panel) were used to examine ERK activation, and an antibody against B-actin and B-

tubulin were used as a loading control (bottom panels).

114

Figure 13. Individual MEK signaling in LeTx-treated SK-MEL-28 cells.

 

 

A

V5-MEK1 V5-MEK1cr Parental V5-MEK2 V5-MEK2cr

0 1 10100 0 1 10100 0 1 10100 0 1 10100 O 1 10 100 <LF(ngIml)
Walt” ' " ** m «vs-MEK

   

 

«d V5-MEK1

]< MEK1

’ «- VS-MEKZ

 

 

 

< MEK2

  
 

 

 

 

 

' < ERK1-PO4
< ERK2-P04

l“ '4 JP" 0"" _ —
[....._.1] L..- ¢.,,,;-- 222:;
loom-fiend quc-nuupq< B-actin

 

 

 

Parental V5-Iacz
1 10100 0 1 10100 <LF(nglml)

‘—-———

1, “a." t-~--V5-IacZ

Woman"

 

 

 

115

Figure 14. Overlapping and non-overlapping transcriptional targets downstream of
MEK] and MEK2.

Two sets of SK—MEL—28 parental cells and the cells stably expressing VS-lacZ,
VS-MEKI, VS-MEchr, VS-MEKZ, or V5-MEK2cr were treated with PA plus
LF_E687C control (BC) or LeTx (LF) for 24h as described in Material and methods. (A)

Total cell lysates were collected from cells in one of the sets for immunoblotting probed

with antibodies against V5 epitope (top panel), NHz-terminus of MEK] (the second

panel), NHz-terminus of MEK2 (the third panel), phospho-ERK1/2 (the fourth panel),

total ERKl/Z (the fifth panel), and GAPDH (bottom panel). (B) After control or LeTx
treatment, total RNA samples were collected from the other set of cells, and subjected to
cDNA microarray hybridization and data analysis. The gene expression profile in
control-treated V5-lacZ-expressing cells was compared to that in LeTx-treated cells to
generate a list of genes that were significantly changed by LeTx treatment (shaded circle).
The lists of genes that were significantly rescued by MEK] (solid circle) or MEK2
(dashed circle) were generated by comparing the gene expression profile in LeTx-treated
VS-lacZ-expressing cells with LeTx-treated VS-MEKI cr-expressing cells or LeTx-
treated V5-MEK2cr-expressing cells, respectively. The numbers of genes that were

significantly changed are labeled.

116

Figure 14. Overlapping and non-overlapping transcriptional targets downstream of
MEKl and MEK2.

‘
A o‘ .,1. 46 are" 13
0
9°” ‘1 “9 4‘)” ~16" q 4

--- -“'IV5-MEK

_

3.". ---":— IMEK1
. o O O C-JMEK2

- .- - - .:] ERK1I2-PO4

—_

n1,-
-

T

l 1

b333é33388-a| ERK1I2

[Wu] GAPDH

 

B Rescued by V5-MEK2cr:
970 + 244 = 1,214 genes

      
 

LeTx effect:
2,560 genes

a‘—---‘-
0|
N
OI

24

Rescued by V5-MEK1cr:
24 + 244 = 268 genes

117

Figure 15. Expression levels of V5 fusion proteins in SK-MEL-28 cells.

Clonal SK-MEL—ZS cells stably expressing V5-lacZ (Z) and cells expressing high
(H), moderate (M), or low (L) expression levels of wild-type VS-MEK or VS-MEKcr
. were established as described. Parental SK-MEL-28 cells (P) were used as a control.
Total cell lysates were harvested and immunoblotted with antibody against V5 epitope to
detect expression levels of V5 fusion proteins (upper panel) and antibody against a

tubulin for equal loading control (lower panel).

V5-MEK1 V5-MEK1cr V5-MEK2 V5-MEK2cr
P ZLMHLMHLMHLMH

 

 

180 kDa-
115 kDa-

 

- 4 V5-lacZ

82 kDa-

 

64 kDa-

4skoa. ~ - *mmgggglws-Mex

 

 

 

i ~ 1 5;? “'1 1.
l. llLr-~~ 111

< a-tubulin

   

118

Figure 16. Inhibitory effect of LeTx on proliferation of SK-MEL-28 stable clones.

SK-MEL-28 parental cells (A, solid line) and the cells stably expressing VS-lacZ
(A, dashed line), VS-MEKI (B, green lines), V5-MEchr (B, red lines), VS-MEKZ (C,
green lines) or V5-MEK2cr (C, red lines) with different VS-fusion protein expression
levels: low (B and C, dashed lines), moderate (B and C, thin solid lines) or high (B and C,
thick solid lines) were tested for their sensitive to LeTx by doing a toxicity assay as
described. The x axis represents the concentration of LF. The y axis represents relative
viability normalized by PA alone control. Data presented in this figure is a representative
of three independent experiments. Error bars represent standard divisions of triplicate

wells in the assay. Images in this figure are presented in color.

119

Figure 16. Inhibitory effect of LeTx on proliferation of SK-MEL-28 stable clones.

A.

Wability (% to PA alone) Viability (% to PA alone)

Viability (% to PA alone)

 

 

 

 

 

 

o A;
100 A) {LN + Parental
--0--V5-lacZ

75% \

50% \

25% x“

0% '
0.01 0.1 1 10 100 1000 10000

100%

[LF} (”Q/ml)

 

--e--V5-MEK1 (L)
-e—V5-MEK1 (M)

 

75% "

-e—V5-MEK1 (H)
--e--V5-MEK1cr (L)

 

50%

—e—V5-MEK1cr (M)
—0—V5-MEK1cr (H)

 

25%

 

0%

100%

75%

 

0.1 1 10 100 100010000
[LFl(ng/m|)

 

--I--V5-MEK2 (L)
+V5-MEK2 (M)

 

50%

25%

 

0%
0.01

 

-l—V5-MEK2 (H)

--I-- V5-MEK2cr (L)
+ V5-MEK2cr (M)
—l—V5-MEK2cr (H)

0.1 1 10 100 100010000
[LF](ng/m|)

120

Figure 17. Sensitivities of SK-MEL-28 cells to LeTx and MEK inhibitors.
SK—MEL-28 parental cells and the cells stably expressing VS-lacZ, or expressing

low (L), moderate (M), or high (H) levels of wild-type VS-MEK or VS-MEKcr, were

tested for their sensitivities to LeTx, U0126, or PD184352 by use of toxicity assays as

described in Materials and methods. (A set of representative proliferation curves is

shown in Figure 16). The IC50 values for each stable cell line were normalized to the

IC50 of parental cells. Results are presented as an average of the change of at least three

independent experiments :t S.D. Statistical significance was determined by one-way
ANOVA (http://www.physics.csbsju.edu/stats/anova_NGROUP_NMAX_form.htrnl)
followed by post-hoc analysis (http://graphpad.com/quickcalcs/posttest1.cfm). P values

were greater than 0.05 except for (*) p < 0.005, and (**) p < 0.00001.

121

Figure 17. Sensitivities of SK—MEL—28 cells to LeTx and MEK inhibitors.

35

 

**%

30

25 20 15 10

j l l 1

Relative I050 (fold to parental cells)

122

 

Parental
V5-Iacz

V5-MEK1 (L)
V5-MEK1 (M)
V5-MEK1 (H)
V5-MEK1cr (L)
V5-MEK1cr (M)
V5—MEK1cr (H)

V5-MEK2 (L)
V5-MEK2 (M)
V5-MEK2 (H)
V5-MEK2cr (L)
V5-MEK2cr (M)
V5-MEK2cr (H)

Parental
V5-MEK2 (H)
V5-MEK2cr (H)

Parental
V5-MEK2 (H)
V5-MEK2cr (H)

Figure 18. LeTx treatment inhibits p38 and JN K MAP kinase activation in SK-
MEL—28 cells.

SK-MEL-28 cells were treated with PA alone control or LeTx for 24 h followed
by UV treatment as described in Materials and methods. Whole cell extracts were
prepared and subjected to immunoblotting as described in Materials and methods.
Activities of p38 MAPK (top panel) and JN K (middle panel) were detected by antibodies
against phospho-p38 MAPK and phospho-JN K respectively. GAPDH (bottom panel)

expression levels were detected as the equal loading control.

Treatment: PA PA LeTx
UV: - + +

 

 

l Phospho-p38 MAPK

 

'; ‘- Phospho-JNKp54
~ «a- Phospho-JNK p46

 

 

GAPDH

 

123

Figure 19. Sensitivity of V5-MKK3cr- and V5—MKK6cr-expressing cells to LeTx.
SK-MEL—28 parental cells and the cells stably expressing VS-lacZ, or expressing
low (L), moderate (M), or high (H) levels of V5-MKK3, V5-MKK3cr, V5-MKK6, or

V5-MKK6cr were tested for their sensitivities to LeTx as described in Material and

methods and the legend of Figure 17. The IC50 values for each stable cell line were

normalized to the IC50 of parental cells. Results are presented as an average of the

change of at least three independent experiments i S.D. Statistical significance was
determined by one-way ANOVA
(http://www.physics.csbsju.edu/stats/anova_NGROUP_NMAX_form.html) followed by
post-hoe analysis (http://graphpad.com/quickcalcs/posttest1.cfrn). P values were greater

than 0.05 except for (*) p < 0.05. Image in this figure is presented in color.

124

Figure 19. Sensitivity of V5—MKK3cr— and V5-MKK6cr—expressing cells to LeTx.

Parental

V5-lacZ

V5-MKK3(L)
V5-MKK3(M)
V5-MKK3(H)
V5-MKK3cr(L)
V5-MKK3cr(M)
V5-MKK3cr(H)

V5-MKK6(L)
V5-MKK6(M)
h—EP V5-MKK6(H)
* i——E—'-'—_—ij_”jb V5-MKK6cr(L)
145‘]h V5-MKK6cr(M)
'11.] V5-MKK6cr(H)

 

 

- - u r I I u

35 30 25 20 15 10 5 0
Relative IC50 (fold to parental cells)

125

Figure 20. Sufficiency of MEK2 signaling pathway for anchorage-independent
growth of SK-MEL-28 cells.

SK-MEL—28 parental cells and cells stably expressing VS-IacZ or high levels of
wild-type VS-MEK or VS-MEKcr were seeded as single-cell suspensions in soft agar and
then treated with PA alone control or LeTx, as described in Materials and methods. After
treatment (21 days), colonies were fixed and stained with 1% crystal violet prepared in
10% ethanol. (A) Representative images of colonies grown in soft agar in the presence of
PA alone or LeTx. Bars, 1mm. (B) Anchorage-independent growth was quantified, and
the colony density was determined by dividing the number of colonies by the area and
then normalizing to the colony density of the parental cells treated with PA alone. One-
way ANOVA followed by post-hoe analysis were performed to determine statistical

significance. P values were greater than 0.05 except for (*), p < 0.00001.

126

Figure 20. Sufficiency of MEK2 signaling pathway for anchorage-independent

growth of SK-MEL—28 cells.

A Parental V5-IacZ V5-MEK1 V5-M EK1cr V5-MEK2 V5-ME.K20r
rfiai?’ We ~-.-.- ....---- r01 '1'.- a -:.
. " O

'.-'.‘ a} '

 

 

 

 

 

 

“i... ‘d .Qa... . 1’ $ t 1,‘ ’ ‘
‘\a' 1“:J’P'??€. '9 ‘3! :1 (fi’r'hz': ‘o ’ : .2..~'.3‘3‘ '
PAalone 1",; , .1 "Q1359; . n: . 6 01.9.3". 3.3”. i, 3"...
mar}. -..‘-'r'--‘—': '41—'231 104-:5.-
‘. 1- ,3
.. ..- ‘0 ”'-
LOTX ‘ . ' .;.._..'
n— — _ rm — _ ‘A‘. g;
B 100%
E i.
3... 75% .
o
a 3
>2
8 fl
0
.05 50,6
0 e
U
Eat
div
2 25%
O
(I:
0% _ T I t 1

 

Parental V5-lacZ V5-MEK1 V5-MEK1cr V5-MEK2 V5-MEK26r

127

Figure 21. SK-MEL—28 xenograft tumor growth and systemic treatment with LeTx.

SK-MEL-28 xenograft tumors were established as described in Material and

methods. Briefly, cells were subcutaneously injected at a number of 107 cells in 100 pl

of Hanks' balanced salt solution (HBSS) into the right side of the dorsalateral area of

groups of ten athymic nude mice. Tumor volmnes were measured every two days by

using a caliper. After the average volume of tumors in the group reached to 50 m3,

mice were divided to two groups. Five mice in each group were intravenously injected
with either LeTx (PA plus LF) or control (PA plus LF_E687C, an inactive form of LF)
prepared in 50 pl of HBSS at a dosage of one standard dose (1 XSD) every two days for a
total of six injections (6XSD total). One standard dose equals 10 pg of PA plus 2 pg of
LF or LF_E687C. Xenografi tumor growth is presented by average tumor volume (y axis)
as a function of time (x axis). The thick solid line represents the tumor growth before
treatment. The dashed line represents the growth of control-treated tumors. The thin

solid line represents the growth of LeTx-treated tumors. Error bars represent standard

deviations of tumor volumes.

128

Figure 21. SK—MEL-28 xenograft tumor growth and systemic treatment with LeTx.

 

 

100
asp-Before treatment ‘
- oA- . Control 3
+Le'rx '
75 A}

5. .11.,

A

 

.I

25 FT ,. .‘ 111..

Mb-

0 71421 28 35 424956 63 70 778491 98105
Time (Days)

 

 

 

Average Tumor Volume (mm‘3)

 

129

 

Figure 22. Reproducibility of SK-MEL-28 xenograft tumor growth in athymic nude
mice.

SK-MEL-28 xenograft tumors were established as described in Material and
methods and the legend of Figure 21. Xenograft tumor growth is presented by average
tumor volume (y axis) as a function of time (x axis). Tumor growth curves from five
independent experiments are presented as lines with different colors and indicated by ND
xenograft project numbers. Total numbers (n) of mice used in each project are indicated.

Image in this figure is presented in color.

 

 

 

 

  

100
+ N086, n=14
+ N088, n=14
a i- ND89, n=11
< ~9- N091, n=14
E + N093, n=20
E 75 I,
9
a /
3 x
B .-/ i, '
> W
h 50 all i g '9
O (I? '
E ['1 3. ' /
a
l— 3;",
0 .9/ /.
g 25
> . ’
< ,
/‘ /J
BIC-I...

 

   

2835424956637077849198
Days after cell line injection

130

Figure 23. Sensitivity of VS-MEKI cr-expressing xenograft tumors to LeTx.
SK-MEL-28 parental xenograft tumors (black lines) and tumors expressing V5-

MEKl (green lines) or VS-MEchr (red lines) were established as described in Material

and methods and the legend of Figure 21. After the average volume of tumors in each

group reached to 75 m3, mice were divided to two groups. Five mice in each group

were intravenously injected with either LeTx (PA plus LF) or control G’A plus
LF_E687C) at a dosage of one standard dose (1 XSD) every two days for a total of six
injections (6XSD total). One standard dose equals 10 pg of PA plus 2 pg of LP or
LF_E687C. Xenografi tumor growth is presented by average tumor volume (y axis) as a
fimction of time (x axis). The thick solid lines represent the tumor growth before
treatment. The dashed lines represent the growth of control-treated tumors. The thin
solid lines represent the growth of LeTx-treated tumors. Error bars represent standard

deviations of tumor volumes. Image in this figure is presented in color.

131

Figure 23. Sensitivity of VS-MEchr-expressing xenograft tumors to LeTx.

 

 

 

 
 
 
 
  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

15° -a—Parenta, before treatment

3 . oe- - Parental, control V5-MEK1cr Parental
< 125 -_ —0—Parental, LeTx 9
E -|s-V5-MEK1, before treatment A V5-MEK1 .. Q
.E, o o» - V5—MEK1, control ' :
0 +V5—MEK1, LeTx 0
§ 100 " -a—V5-MEK1cr, before treatment f
3 . 0A0 - V5-MEK1cr, control '
2 -0—V5-MEK1cr, LeTx ‘ x?
o 75
E
'3 .
é so
5
3

25

0 ..‘—“-Ii;393...ss.‘..u..mm...
0 7 14 21 28 35 424956 63 70 778491 98105

Time (Days)

132

Figure 24. Sensitivity of V5-MEK2cr-expressing xenograft tumors to LeTx.
SK-MEL-28 parental xenograft tumors (black lines) and tumors expressing V5-

MEK2 (green lines) or V5-MEK2cr (red lines) were established as described in Material

and methods and the legend of Figure 21. After the average volume of tumors in each

group reached to 50-70 mm3, mice were divided to two groups. Five mice in each group

were intravenously injected with LeTx (PA plus LF) or control (PA plus LF_E687C) at a
dosage of one standard dose (1 XSD) every two days for a total of six injections (6XSD
total). One standard dose equals 10 pg of PA plus 2 pg of LP or LF_E687C. Xenografi
tumor growth is presented by average tumor volume (y axis) as a function of time (x axis).
The thick solid lines represent the tumor growth before treatment. The dashed lines
represent the growth of control-treated tumors. The thin solid lines represent the growth
of LeTx-treated tumors. Error bars represent standard deviations of tumor volumes.

Image in this figure is presented in color.

133

Figure 24. Sensitivity of VS-MEKZcr-expressing xenograft tumors to LeTx.

Average Tumor Volume (mm‘3)

 

175

 

 

-A-Parental, before treatment

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. on - Parental, control V5-MEK2
150 a -e—Parental, LeTx t
-a-V5-MEK2, before treatment .' Parental
125 “ ovo-VS-MEKZ, control 4 v5.MEKgcr A
-e—V5-MEK2, LeTx ' .K
-a-V5-MEK2cr, before treatment ‘ f .-
100 ll °-A--V5-MEK2cr, control 1 ,5, '
-e—V5-MEK2cr, LeTx . i A
A ,0 "
75 ’ ,-
A
so "
25
0 .2 "‘ seams-.ta'u'd‘l'a's'a‘fi... as...“
0 714 21 28 35 42 49 56 63 70 77 84 91 98105

Time (Days)

134

Figure 25. Does-dependent effect of LeTx on SK-MEL-28 xenograft tumor growth.
SK-MEL—28 parental xenograft tumors (black lines),and tumors expressing V5-
MEKZ (green lines) or V5-MEK2cr (red lines) were established in 20 athymic nude mice

as described in Material and methods and the legend of Figure 21. After the average

volume of tumors in each group reached to 50-70 mm3, mice were divided to four groups.

Five mice in each group were intravenously injected with PA plus LF_E687C control
(solid triangles) at a dosage of one standard dose, or LeTx (PA plus LF) at a dosage of 1
(open diamonds), 0.5 (open squares), or 0.25 (closed circles) standard dose every two
days for a total of six injections. One standard dose equals 10 pg of PA plus 2 pg of LF
or LF_E687C. Xenograft tumor growth is presented by average tumor volume (y axis) as
a function of time (x axis). The thickest solid lines with open triangle data points
represent the tumor growth before treatment. Error bars represent standard deviations of

tumor volumes. Image in this figure is presented in color.

135

Figure 25. Does-dependent effect of LeTx on SK—MEL—28 xenograft tumor growth.

Average Tumor Volume (mm‘3)

175

A

0|

0
1

.3
M
(II

100

N
0|

01
O

N
0|

0

 

 

-a—Parenta|, before treatment
0 0A0 0 Parental, control 1 SD
l-o—Parental, LeTx 1 $0
-—a— Parental, LeTx 0.5 SD

. ------ .,....- Parental, LeTx 0.25 SD
-a—V5-MEK2, before treatment

. 0A0 - V5-MEK2, control 1 SD
--+—V5-MEK2, LeTx 1 $0

—a— V5-MEK2, LeTx 0.5 SD
~----------~eV5-MEK2, LeTx 0.25 SD
h-Is—VS-MEKZcr, before treatment
. vo - V5-MEK2cr, control 1 SD
.—o—V5-MEK20r, LeTx 1 $0
—a— V5-MEK20r, LeTx 0.5 SD
wow-- V5-MEK20r, LeTx 0.25 SD

 

 

 

 

 

 

 

0 71421283542

"-"_"___-_-v~ - 7.7.1.4673. . . .
u uuuvuuuruuuvuuuruu

your“

49 56 63 70 77 84 91 98105

 

V5-MEK2
A
: Parental
5 V5-MEK2cr A
.x . f
1‘ A

 
 
 
 
 
 
 
  
  

 

 

Time (Days)

136

.353 Sauce mange es. .

.983 28:5 $92923on .288 no: 82.8 m> is“... 863:5 5E 8:28.556. 85 so; 835:: 8:83.02

SEE .3 3836:: mpg—9:3 83%— .883 382% was Banana ob? 8:33 .583. .mEoEtomxo 05 we can 05 Ban 8:: Bot

Buses 203 22:5 8858.. a A“: Ba <3 £3 .5 650ml": 33 «6 89:8 55 38.3. 3.8835 203 8952
-m> 8 $52-3 .8Qm2-m> .Em2-m> message £25: 8 £85 canoes. wNJmEQm were... 8:: 82. case...

.9353 «.3.—ween autumz¢nm E agencies gov—qumNr me 364 6N 0.5»;

137

 

 

 

 

 

 

 

 

 

 

SWLOON

966 'I-CSZIV 99,BI~ZSI-clvg

0|»?

 

 

 

5.0.. 3.5000 5.0.. 3.5000 5.0.— 3.5000 5.04 38:00 02.0.. 3.8000 5.0.. 33:00

 

 

35.5265 35565 32.25.. 5035265 53265 .0395“.

.9535 ”.3.-wen: aim—Z05 5 5:29.93 nova—2.3» be $94 6N 9:.me

138

Figure 27. Systemic treatment of LeTx does not cause MEK cleavage in SK—MEL-
28 xenograft tumor cells.
SK-MEL-28 parental xenograft tumors were established in athymic nude mice

and LeTx was administrated intravenously as described with the following minor

. . th . . . . th
modification. Two days after the 5 intravenous injection, the 6 LeTx or control

treatment (1 SD) was switched to intratumoral injection (it) on selected mice. Other mice

still received the 6th intravenous injection (iv). Twenty-four hours after the last

administration, tumors (indicated by tumor identification numbers) were dissected for

tumor lysates preparation. Immunoblotting was performed using antibodies against NHZ-

terminus of MEKl (top panel), carboxyl terminus of MEK] (the second panel), phospho-

ERKl/Z (the third panel), total ERK1/2 (the fourth panel), and GAPDH (bottom panel).

139

 

Figure 27. Systemic treatment of LeTx does not cause MEK cleavage in SK—MEL—

28 xenograft tumor cells.

N093 Parental

 

Control 1 SD LeTx 1 80
iv it iv it

 

 

 

 

 

 

 

 

 

 

N h P n
§§ss§§ss§
8
g A_nt__lbody
[§--fl¥ 9°? 4”:] MKK1 (NT)
[M “.M.. J p-ERK‘I’Z
;- .. . 5:- 7“: '"" I; I: w_ serum total

 

GAPDH

 

 

140

 

Figure 28. Unsupervised clustering of gene expression changes in LeTx-treated cells.
SK-MEL—28 parental cells and the cells stably expressing V5-lacZ, VS-MEKl ,
VS-MEchr, V5-MEK2, or V5-MEK2cr were treated with PA plus LF_E687C control
(BC) or LeTx (LF) for 24h as described in Material and methods. Total RNA samples
were collected and subjected to cDNA microarray hybridization and data processing as
described in Material and methods. Three independent experiments (indicated by _A, _B,
and _C) were performed to generate statistical significance. Total RNA isolated from
control V5-IacZ-expressing cells (treated with PA plus LF_E687C) were used as a gene
expression reference control to normalize the changes in LeTx-treated V5-lacZ-, V5-
MEKl cr-, and V5-MEK2cr-expressing cells. Normalized gene expression changes were

organized by hierarchical clustering.

141

 

Figure 28. Unsupervised clustering of gene expression changes in LeTx-treated cells.

 

_l oufifl 55.92-?

 

_l <ufi3t2xm2a>

 

 

.I mrfifltflxwze>

 

l. mrfifi town—26>

 

 

 

.| 3%.. \ 82%-?

 

 

I 093459ng

m 03 \ N87?
1H 3.8.. . New»?
r .

 

 

5.0.. \ Nom.-m>

 

254

. _ q . q q
5 0 5 0 5 O
5 5 4 4 3 3

142

1.5. Materials and methods

1.5.1. Cell lines and stable cell line establishment.
SK-MEL-28 cells were obtained from CeeTox Inc. and grown in RPMI 1640

medium supplemented with 5% FBS and 50 units/m1 penicillin/streptomycin. Cells were

cultured at 37°C in a humidified 5% C02 incubator. To establish stable cell lines, SK-
MEL-28 cells were transfected with V5-MEK, VS-MEKcr or V5-IacZ control expression

TM
vectors by using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s

. . . . . ®
instructions. Clonal stable transfectants were selected against 1 mg/ml Genetlcm .

Expression levels of V5-fusion proteins in each stable clone were determined by

immunoblotting. After stable cell lines were isolated, cells were maintained in medium

containing 0.5 mg/ml Geneticin®. One day before subsequent assays, stable cells were

. . . . . ®
subcultured once, and cultured 1n culture medium w1thout Genetlcm .

1.5.2. Chemicals and LeTx

U0126 (Calbiochem) and PD 1843 52 (U SBiological) were dissolved in dimethyl
sulfoxide (DMSO) (Sigma) at a stock concentration of 100 myml. PA and LF were
expressed in an attenuated strain of Bacillus anthracis (BH445) and purified by fast

pressure liquid chromatography as described (Bromberg-White & Duesbery, 2008).

1.5.3. siRNA-mediated MEK knock down.

143

Multiple siRNAs specifically against human MEK] or MEK2 and AllStars non-

silencing control siRNA were purchased from Qiagen. To deliver siRN A into SK-MEL—

TM
28 cells, the siLentFect Lipid (Bio-Rad) was used as the transfection reagent. To do

this, SK-MEL-28 cells (3 X 105 cells) were seeded and cultured in 60-mm dishes. The

next day, cells were washed three times with PBS and covered with 1.6 ml of OPTI-

MEM®I (Invitrogen), to which siRNAzLipid complexes (prepared as described below)

were then added. To prepare siRNAzLipid complexes, 10 ul of lipid was added into 30 p1

of OPTI-MEM®I and incubate at room temperature for five minutes. The prepared lipid

was then added into 360 pl of OPTI—MEM®I containing 100 pmole of each control or

MEK siRNAs plus 100 pmole of AlexaFluor488-conjugated non-silencing siRN A, and
incubated at room temperature for 20 minutes to allow siRNAzLipid complexes form.
Eight hours after addition of siRNAzLipid complexes, cells were then trypsinized and
split to two 60-mm dishes and culture in DMEM containing 10 % FBS and 50 units/ml
penicillin/ streptomycin for 72 hours. At this time point, AlexaFluor488 fluorescence
could be observed in greater than 95% of cells under a fluorescence microscope (data not
shown), indicating a high efficiency of siRNA transfection. Cells were then harvested for

immunoblotting and cell cycle analysis.

1.5.4. In-cell MEK cleavage assay in SK—MEL-28 cells.

SK-MEL—28 parental cells and cells stably expressing VS-fiision proteins were
treated with LeTx in a LF concentration-dependent manner (1 ug/ml PA plus 0, 1, 10,
100 11ng LF) for 24h. Total cell lysates were collected on ice in RIPA lysis buffer [50

144

 

mM Tris-HCl pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 2 mM Na3VO4, 20

mM sodium pyrophosphate, 1% sodium deoxycholate, 1% Triton X-100, 0.1% SDS, and

l X EDTA-free protease inhibitor cocktail (Roche)] and homogenized by sonication in ice

TM
bath. Protein concentrations were determined by BCA Protein Assay Kit (Pierce)

according to the manufacturer’s instructions. Lysates were then prepared in SDS sample
buffer [47.5% Laemmli Sample Buffer (Bio-Rad) and 2.5% B-mercaptoethanol (Sigma)].
Five micrograms of total cell lysates were subjected to immunoblotting to detect LF-

mediated cleavage and ERK activation as described in the results.

1.5.5. Immunoblotting.

Five micrograms of total cell lysates were separated in 10% Novex® Pre-Cast

Tris-Glycine Gels (Invitrogen) and then electro-transferred onto polyvinylidene fluoride
(PVDF) membranes (Millipore) according to the manufacturers’ instructions.

Membranes were then soaked in 5% non-fat milk for 1 hour and hybridized with primary

antibodies against V5 epitope (Bethyl Laboratories), NHz-terminus of MEK] (Upstate,

#07-641), COOH-terminus of MEK] (Santa Cruz Biotechnology, SC-219), NHZ-

terminus of MEK2 (Santa Cruz Biotechnology, SC-524), COOH-terminus of MEK2
Santa Cruz Biotechnology, SC-525), phospho-ERK (Cell Signaling, #9106), ERK (Cell
Signaling, #9102), phospho-p38 MAPK (Cell Signaling #4631), phospho-JNK (Cell
Signaling #9255), a-tubulin (Sigma, T9026), B-tubulin (Sigma, T5201), B-actin (Sigma,
A1978), or GAPDH (Cell Signaling, #2118) at 4°C for overnight. Conditions for

primary antibody hybridizations were followed according to the antibody datasheets.

145

After primary antibody hybridization, membranes were washed three times in TBST
buffer (50 mM Tris, 150 mM NaCl, and 0.1% Tween-20), hybridized with HRP-
conjugated secondary antibodies according to the instructions of the antibodies, and then

washed three times in TBST buffer. Immunoblotting signals were then detected by

TM
LumiGLO Reagent and Peroxide (Cell Signaling).

1.5.6. Toxicity assay.

For LeTx toxicity assay, SK-MEL-28 cells (1,500 cells) were cultured in 96—well
plates for 24 hours, and then treated with control (1 ug/ml PA) or LeTx (l ug/ml PA plus
0.01-10,000 ng/ml LF) for 72h. For U0126 toxicity assay, cells were treated with DMSO
control or 1-100,000 nM U0126 for 72h. For PD 1843 52 toxicity assay, cells were

treated with DMSO control or 0.01-10,000 nM PD 1843 52 for 72h. Cell viability was

determined by using CellTiter 96® Aqueous Non-Radioactive Cell Proliferation Assay

(Promega) according to the manufacturer’s instructions. LeTx toxicity was presented by

plotting the relative viability (normalized by no LF control) against LF concentrations.

IC50 values were determined by SigmaPlot software.

1.5.7. SK—MEL—28 tumor xenograft and LeTx systemic treatment
SK-MEL-28 parental cells or cells stably expressing VS-MEK or VS-MEKcr

were cultured in exponential growth phase and then collected in Hanks' balanced salt

solution (HBSS) (Invitrogen) at a density of 108 cells/ml. Cells were then

subcutaneously injected at a number of 107 cells (in 100 pl of HBSS) into the right side

146

 

of the dorsalateral area of groups of ten athymic nude mice (Charles River Laboratories).
Tumor volumes were measured every two days by using a caliper. Animal health and

tumor volume were monitored till the end of the experiment. After the average volume

of tumors in the group reached to 50 m3, mice were divided to two groups. Five mice

in each group were intravenously injected with either LeTx (PA plus LF) or control (PA
plus LF_E687C, an inactive form of LF) prepared in 50 pl of HBSS at a dosage of one
standard dose (1 ><SD) every two days for a total of six injections (6XSD total). One
standard dose equals 10 ug of PA plus 2 pg of LP or LF_E687C. Two hours after the last
injection, mice were euthanized and tumors were dissected for subsequent analysis. To
prepare tumor lysates for immunoblotting, tumors were homogenized in RIPA lysis
buffer on ice, and clear tumor lysates were harvested by centrifugation.

To treat tumors in a LeTx does-dependent manner (Figure 25), cells were
inoculated in 20 mice. After tumors established, mice were divided into four groups, and
mice in each group were intravenously injected with l><SD control (PA plus LF_E687C)
or 1, 0.5, or 0.25XSD LeTx (PA plus LF) for a total of six injections.

To test if MEK in xenograft tumor cells was still cleavable by LF (Figure 27), two

th . . . . th .
days afier the 5 intravenous Injection, the 6 control and LeTx treatment were swrtched
to intratumoral injection on selected mice. Other mice still received the 6 intravenous

injection. Twenty-four hours after the 6 injection, rmce were euthamzed and tumors

were dissected for subsequent analysis.

147

 

1.5.8. Anchorage-independent growth assay.

Anchorage-independent growth assays were performed as previously described
(Ding et al., 2008). Briefly, 10,000 cells were prepared as single cell suspensions in
0.35% agar prepared in a volume of 1 ml of culture media, and then seeded onto 1 ml of
hard agar (1% agar prepared in culture media) in 12-well plate. One day after cell
seeding, 1 ml of culture media containing PA along control (1 jig/ml PA) or LeTx (1

jig/ml PA and 0.01 jig/ml LF) were added on the top of cell-containing agar layer. Cells

were then incubated at 37°C in a humidified 5% C02 incubator. The PA- or LeTx-

containing culture media was replaced every 2-3 days for a total of 21 days. Colonies
were then fixed and stained with 1% crystal violet prepared in 10% ethanol for
observation. Images were then taken under a dissecting microscope. Colonies were
quantified using Imagine software (Ding et al., 2008). Unsaturated area was selected,
and colonies larger than 50 pm in the area were counted by the software. Colony density

was determined by normalizing colony numbers by area size.

1.5.9. Human cDNA microarray and transcriptional signature analysis.

Two sets of SK-MEL-28 parental cells and cells stably expressing V5-lacZ, V5-
MEKl, V5-MEchr, V5-MEK2, or V5-MEK2cr were treated with LeTx (1 ug/ml PA
plus 10 ng/ml LF) or control (1 ug/ml PA plus 10 ng/ml LF_E687C) for 24 h. Totally
cell lysates were collected fiom one of the two sets and subjected to immunoblotting to

confirm MEK status and LeTx treatment. Total RNA samples were prepared from the

TM
other set of cells using a mirVana RNA isolation kit (Ambion). Four RNA samples

(control-treated V5-lacZ-expressing cells, LeTx-treated V5-IacZ-, V5-MEchr, and V5-

148

 

MEK2cr-expressing cells) were submitted to the Microarray Core Laboratory at the Van
Andel Research Institute for microarray hybridization. A total of three independent
experiments were performed. Gene expression profiles (n=12) were generated using the
Agilent 60-mer Whole Human Genome 44k Microarray platform (Agilent) according to
the manufacturer’s instructions. All subsequent analysis was performed using the
Bioconductor software environment. Microarray data was processed and normalized 3
using the limma Bioconductor package (Gentleman et al., 2004; Patterson et al., 2006).
Gene set enrichment analysis was then performed using the PGSEA Bioconductor

package (Furge et al., 2007a) with only the most significant results shown. A signature

 

was determined to be uniquely rescued by VS-MEchr or V5-MEK20r when it satisfied
two criteria. First, the signature in V5-lacZ-expressing cells had to have a highly
significant (p < 0.005) positive or negative response to LeTx treatment while in cells
expressing V5-MEK1cr or V5-MEK2cr showed a highly significant (p < 0.005) response
to LeTx treatment in the opposite direction. Second, the signature in cells expressing the
other VS-MEKcr isoform did not demonstrate a highly significant (p < 0.005) to LeTx
treatment. Differentially expressed genes were identified using a moderated t-statistic as
implemented in the limma Bioconductor package. Significance values were adjusted

using the false discovery rate (FDR) method to compensate for multiple testing.

1.5.10. Examination of p38 MAPK and JNK activations in SK-MEL-28 cells.
Two sets of SK-MEL-28 parental cells were cultured and treated with PA alone (1
jig/ml PA) or LeTx (1 jig/ml PA plus 0.1 ug/ml LF) for 24h as described. One set of the

cells were then subjected to UV treatment as positive controls for detecting p38 MAPK

149

and JN K activation using immunoblotting. The other set of the cells were used to detect
basal activities of p38 MAPK and INK. To treat cells with UV, the toxin-containing
culture media were collected into a sterile tube. Culture dishes were placed in a UV
Stratalinker (Stratagene) with the lid removed. Cells were then exposed to UV for 1 min
(time mode). Control cells were exposed to air for 1 min as the non-UV-treated control.
After UV or control treatment, the toxin-containing media were added back to the culture
dishes, and cells were cultured for 2h. Total cell lysates were then harvested, and

immunoblotting was performed to detect p38 MAPK and JN K activities.

150

 

Chapter IV. General discussion

To better understand the critical role of MEK signaling pathways in melanoma
biology, it is important to determine both necessity and sufficiency of MEK signaling
pathways for melanoma cell proliferation. For this purpose, two complementary
approaches were used. First, MEK-specific siRN A was used to specifically knock down
either MEK] or MEK2 in human melanoma SK-MEL—28 cells and to determine the
necessity of MEK] and MEK2 signaling pathways for melanoma cell proliferation.
Second, a novel experimental system was developed for determining the sufficiency of
MEK] and MEK2 signaling pathways for SK-MEL-28 cell proliferation. This
experimental system utilizes the proteolytic activity of anthrax lethal toxin to inhibit
multiple MEK/MKK signaling pathways in SK-MEL-28 cells. Simultaneously, either
MEK] or MEK2 was rescued by expressing a mutant MEK] or MEK2 which is resistant
to LF-mediated cleavage. Combining the siRNA technology and the novel experimental
system, I determined both necessity and sufficiency of MEK] and MEK2 signaling
pathways for SK—MEL-28 melanoma cell proliferation, and made the following
conclusions. First, neither MEK] nor MEK2 signaling pathway is necessary for SK-
MEL—28 cell proliferation as loss of one of the MEKs can be compensated by the other
isoform. Second, the MEK2 signaling pathway alone is sufficient to drive SK-MEL-28
cell proliferation, but MEK] is not. Third, while MEK2 can compensate for loss of
MEK] , MEKl can compensate for loss of MEK2 only when an as yet identified factor is
present. From this work, several interesting observations and valuable information were

obtained. These will be discussed in this Chapter.

151

1.1.Experimental tools to study MEK and MK functions
I have used three distinct experimental tools to study MEK/MKK functions;
small-molecule inhibitors, RNA interference, and LeTx. Each of these has its strengths
and weaknesses depending on the needs of a study. This will be discussed as follows.
Small-molecule inhibitors are easy handled and delivered. Treating cultured cells
with small-molecule inhibitors does not require transfection, and animals can often be
treated with these inhibitors simply by oral gavage. In addition, small-molecule

inhibitors directly target the proteins as opposed to the mRNAs. This generally gives a

 

better efficiency compared to using siRNAs because expression of relatively stable
proteins cannot be completely inhibited by targeting their coding mRNAs. The
specificities of small-molecule inhibitors is always a cause for concern (Cohen, 1999). If
the MEKl/2-ERK pathways need to be targeted, the non—ATP-competitive MEK
inhibitors, such as PD 1843 52 and its derivatives, are the best agents because of their
high selectivity. However, highly selective non—ATP-competitive inhibitors targeting
MKK3/6 or MKK4/ 7 have not been identified, and ATP-competitive inhibitors have
broad off-target effects as the ATP binding site is conserved in other protein kinases. On
the other hand, due to the high homology shared between MEKI and MEK2, the non—
ATP-competitive MEK inhibitors do not discriminate one from the other, so that they are

not suitable for dissecting the necessity of a single isoform.

If evaluating the necessity of a specific single MEK or MKK is the goal of a study,

RNA interference is the best choice as it has the highest isoform specificity. For this

152

purpose, siRNA may be used to knockdown the expression of only one of the family
proteins in cultured cells. In addition, RNA interference also has a great advantage of
being amenable to rescue experiments. A targeted mRNA can be rescued by expressing
an mRN A with one or more wobble mutations introduced into the targeting site. This
strategy makes the rescuing mRN A resistant to RNA interference without changing the

primary protein sequence. However, delivering interfering RNA can be a challenging

__ _____ i‘q

task. AlthOUgh reagents giving high transfection efficiency of siRNA are commercially
available, they are best suited for transient transfection in cell culture. If knocking down

a specific gene in xenotransplanted tumors is desired, cell lines stably expressing small

 

hairpin RNA (shRNA) need to be established first. If the targeted gene product is
required for cell proliferation, an inducible system will have to be included for stable cell
line establishment. In this case, leakiness of shRNA expression is another technical
problem that needs to be overcome. In addition, as mentioned above, interfering RNA
targets expression at the mRN A level. If the targeted gene product is relatively stable,
residual protein may be a concern especially for signaling molecules that are generally
very sensitive to their activators. Finally, if multiple MEK/MKK family proteins need to
be simultaneously knocked down, RNA interference is not suitable. Transfecting many
siRNA molecules into cell culture to target multiple MEK/MKK proteins is not expected
to be an efficient method because it may decrease the efficiency of RNA interference and

increase off-target effects.

LeTx is a highly selective biological inhibitor targeting MEK1/2 and most other

members in the MK protein family. As a pan MEK/MKK protease, LeTx

153

proteolytically inhibits multiple MEK/MKK proteins in mammalian cells, and in turn,
blocks the three canonical MAPK pathways. Compared to small-molecule inhibitors and
siRNA, LeTx has both advantages and disadvantages depending on the purpose of a
study. Delivering LeTx is relatively easy and efficient. It does not require transfection
because LeTx mediates its own entry into cells, and the toxin receptors are ubiquitously
expressed in many different types of cells (Bradley et al., 2001; Scobie et al., 2003). In
addition, the proteolytic inactivation of MEK/MKK proteins by LF is an irreversible

reaction in cells, yielding an efficient durable effect. In terms of experimental design, if

 

studying the sufficiency of a single MEK/MKK signaling pathway is the goal, LeTx plus

the MEKcr/MKKcr is the only available approach. Because of these strengths, LeTx has
become a powerful tool for studying the involvement of MEK/MKK signaling pathways,
not only in cancers but also in other MAPK pathway-involved diseases such as
inflammatory diseases (Pellizzari et al., 1999; Park et al., 2002; Agrawal et al., 2003) and
eye diseases (Bromberg-White et al., 2009). On the other hand, however, LeTx does not
give an objective evaluation for necessity of individual MEK/MKK signaling as LeTx
targets multiple MEK/MKK. For this purpose, RNA interference technology may be

included as complementary approach, as presented in this dissertation.

1.2.Non—redundant roles of MEKl and MEK2

MEKI and MEK2 are generally assumed to be functionally redundant as they
share the same upstream activators and downstream substrates. However, a few previous
studies as well as the results presented in Chapter III indicate MEK] and MEK2 have

distinct and non-overlapping biological functions. During mouse embryonic

154

development, loss of MEK] expression is lethal due to placental deficiency, which is a
result of reduced blood vessels in the labyrinthine layer of the placenta (Giroux et al.,
1999). In contrast, MEK2 is dispensable as MEK2 deficient mice are viable and fertile
(Belanger et al., 2003). These knock-out studies indicate that MEKI is required for
blood vessel formation during mouse embryo development, and MEK2 is not required for
this process. However, they do not indicate whether MEK2 can compensate for MEKI

as the developmental patterns of expression of these kinases may be different. Scholl et
al (2009b) recently demonstrated that MEK] is required for DMBA/TPA-induced benign
epidermal papilloma formation, but MEK2 is not required. On the other hand, Voisin et
al (2008) have shown that shRNA-mediated MEK2 knockdown has much stronger
inhibitory effect on colon cancer cell proliferation than MEK] knockdown does, leading
to the conclusion that MEK2 is more important for colon cancer cell proliferation than
MEKl. Finally, I demonstrate here that MEK2 signaling is sufficient for melanoma cell
proliferation, but MEKl signaling is not. Collectively these results lead to the conclusion
that MEKl and MEK2 are functionally distinct and that their relative importance varies
by cell type and activity. MEK] signaling seems to be more critical at early stages of
embryo development and the initiation step of cancer cell transformation, while MEK2
signaling is more important to maintain cell proliferation and survival after the initiation
steps. Further study is needed in order to fully understand the differential roles of MEK]

and MEK2.

155

1.3.Complexity of the MEK/MKK signaling network

For historic reasons MEK/MKK pathways have been artificially separated into
modules of three tiered kinase cascades, each of which comprises an MKK kinase
(MKKK), a MEK or MKK, and a MAP kinase. However, it is becoming apparent that
substantial cross-talk between these pathways may coordinately regulate cellular
responses to stimuli. For example, Xia et al. (1995) have suggested that apoptosis in

NGF-differentiated PC-12 cells is regulated by opposing activities of ERK and p38

.' 'K‘— m

MAPK/IN K. Similar results also have been reported by MacKeigan et al. (2000)

 

showing that enhanced apoptosis in paclitaxel (Taxol)-treated cells is regulated by

Pfiu x

opposing activities of ERK and IN K. More recently, Estrada et al. (2009) reported the

cross talk between ERK and p38 MAPK for melanoma proliferation.

In Chapter III, I used two complementary approaches to test the hypothesis that
the functions of MEK] and MEK2 are critical and interchangeable for melanoma cell
proliferation. In the first series of experiments MEK-directed siRNAs were used to
selectively knock-down either MEK] or MEK2, or both, in human melanoma SK-MEL-
28 cells to determine the necessity of MEK signaling pathways for melanoma cell
proliferation. In these experiments I found that only simultaneous knock-down of both
MEK] and MEK2 was capable of inhibiting ERK activation and blocking cell cycle
progression. This indicates that neither MEK] nor MEK2 is necessary for these activities
in melanoma cells and supports the initial hypothesis that MEK] and MEK2 are
interchangeable for melanoma cell proliferation. In the second series of experiments, the

novel experimental system utilizing the specific proteolytic activity of anthrax lethal

156

toxin and a rescue by MEK] cr or MEK2cr was used to evaluate the sufficiency of
MEK/MKK signaling pathways for SK-MEL-28 cell proliferation. Using this approach I
found the MEK2 signaling pathway, but not MEKI , was sufficient to drive proliferation
of SK-MEL-28 cells. This indicates that MEK] and MEK2 are not functionally

redundant. Results from these two complementary approaches lead me to formulate a

 

 

new hypothesis: while MEK2 is sufficient for SK-MEL28 melanoma cell proliferation, F
MEKl can compensate for loss of MEK2 only in the presence of an as yet unidentified

factor. Although it is not known what this critical factor is, it compensates for the loss of

MEK2 by cooperating with MEKI to drive cell proliferation. Since activity of this factor E

is repressed by LeTx, this factor is likely to be another MKK or is regulated by another
MKK. This finding not only supports the previous findings that cross-talk between MEK
and other MKK pathways may coordinately regulate cellular functions, but also provides
a better appreciation of the complexity of the MEK/MKK network and the alternative

mechanisms by which cancer cells maintain their survival.

A related observation arose when we analyzed gene expression profiles in LeTx-
treated V5-MEchr- or V5-MEK2cr-expressing cells (Chapter III section 1.2.2.2). First,
we observed that not all the genes that were affected by LeTx treatment can be rescued
by wither MEK] cr or MEK2cr (Figure 14 the gray area that is not covered by the solid
circle and the dashed circle). Using gene set enrichment analysis we noted that the
signatures in this non-rescued area are not associated with cell proliferation or cell
survival (see Appendix III for the full list of the signatures). It is very likely that other

MKK3 are responsible for the activities of these signatures. Further analysis using other

157

VS-MKKcr-expressing cells should be performed to address this question. Second, we
observed that in addition to the non-overlapping rescue by VS-MEchr and V5-MEK2cr,
this analysis also revealed that MEK2- and, to a less degree, MEKl-induced
transcriptional activities that were not wholly contained within the LeTx-transcriptional

footprint (Figure 14, the clear areas in the V5-MEKcr rescue circles but not overlapped

 

with the LeTx effect circle. See Appendix IV, Appendix V, and Appendix VI for the full F-
lists of the signatures). Since LeTx is a pan MEK/MKK inhibitor, I expected that the

result of MEKcr expression would relieve a sub-set of LeTx-induced transcriptional 3
changes. However, I found that MEK2cr, and to a lesser extent MEKI or altered the E

expression of genes that were unaffected by LeTx. Although it is possible that the
change in expression of this group of genes is a result of forced expression of exogenous
V5-MEKcr, I cannot exclude the explanation that this group of genes is subject to
antagonistic regulation by MEK and at least one of the other MKKs. Changes in
expression of these genes are revealed when this dynamic balance is disturbed. This
observation suggests an underlying complexity of the MEK/MKK signaling that has
previously not been widely appreciated. There are examples of MKK-related
antagonistic co-regulation in the literature. It has been demonstrated that the survival of
differentiated rat PC-12 pheochromocytoma cells in culture is dependent on the presence
of nerve growth factor (N GF ), and removal of NGF from the medium causes an increase
in p38 MAPK and JNK activities and a decrease in ERK activity, indicating that these
changes are necessary and sufficient to induce apoptosis. Interestingly, expressing
constitutively activated MEKl prevents apoptosis induced by NGF withdrawal (Xia et al.,

1995). These results indicate that apoptosis in NGF-differentiated PC-12 cells is

158

regulated by opposing activities of ERK and p38 MAPK/INK. Similar results also have
been obtained by MacKeigan et al. (2000) showing that enhanced apoptosis in paclitaxel
(T axol)-treated cells is regulated by opposing activities of ERK and JN K. Collectively,
these results demonstrate that there is cross-talk between MEK/MKK pathways. Thus,

the true function of the pathways cannot be fully appreciated if they are only considered

 

in isolation.
1.4.Therapeutic implications :7
Since Ray and Sturgill’s (1987; 1988a) initial discovery, the regulators of ERK L

activity have been recognized as playing a key role in a variety of cellular processes
including tumorigenesis and cancer survival, and inhibiting the Raf-MEK-ERK pathway
has become a viable therapeutic strategy against cancers. Despite their pre-clinical
successes, MEK small-molecule inhibitors have not performed well in cancer clinical
trials because of issues such as a lack of clinical response, insufficient efficacy, and
safety concerns. As discussed earlier in Chapter 1, (section 1.3.4, page 24), several facts
may explain why these MEK inhibitors work well on xenograft tumors in nude mice but
fail to generate clinical responses in human patients. Nevertheless I believe targeting
MEK remains a viable strategy against cancers, as the Raf-MEK-ERK signaling pathway
plays a pivotal role in cancer cell survival. Designing novel inhibitors with the desired
selectivity for MEK may increase their potency. Alternatively, new approaches may be
required in order to increase clinical response. My research and other studies have
suggested that targeting multiple MEK/MKK pathways may be a better strategy for

treating cancers as LeTx, a pan MEK/MKK inhibitor, inhibits not only tumor cell

159

proliferation, but also tumor-associated endothelial functions. This is discussed as

follows.

F riedlander et al. (1998) provided the first evidence of the anti-tumor activity of
LeTx. Following this, Duesbery et al. (2001) then reported its anti-cancer effects not
only on cancer cell proliferation and anchorage-independent growth, but also on the E“
growth of xenograft tumors in viva. This anti-cancer effect was also observed on other

types of human cancer cell xenografis (Koo et al., 2002; Depeille et al., 2007; Ding et al.,

 

2008; Huang et al., 2008; Rouleau et al., 2008). Moreover, LeTx also has been modified i
to selectively target cancer cells that have potential for metastasis (Abi-Habib et al.,

2006b; Liu et al., 2008), demonstrating the potential of using LeTx to treat late-stage

cancers. In addition to cancers, LeTx has been used in other disease models (Pellizzari et

al., 1999; Park et al., 2002; Agrawal et al., 2003; Bromberg-White et al., 2009).

However, LeTx indeed has toxicity for animals, and PA and LF may cause an immune

response as they are proteins. These obstacles make it difficult to advance LeTx to

clinical trials. Nevertheless, LeTx has been successfully used as a powerful tool to study

the involvement of MEK/MKK pathways in human diseases.

LeTx has been shown to possess anti-cancer activities by inhibiting not only
tumor cell proliferation but also tumor angiogenic processes. The first evidence was
provided by Duesbery et al. (2001) showing that LeTx-treated tumors have a pale
appearance when compared to untreated tumors. Evidence supporting this anti-

angiogenic activity of LeTx includes that the mean vascular density for LeTx-treated

160

 

tumors is dramatically reduced (Duesbery et al., 2001; Depeille et al., 2007; Ding et al.,
2008; Huang et al., 2008). Furthermore, a recent study demonstrated the rapid inhibitory
effect of LeTx on the perfusion of fibrosarcoma xenograft tumors but not in non-tumor
host tissues (Ding et al., 2008), suggesting that the inhibitory effect of LeTx on

endothelial function is highly selective for tumor-associated blood vessels. However, the

 

mechanism by which LeTx selectively targets trunor-associated blood vessels is uncertain. E—
Initially it was hypothesized that systemic LeTx treatment via tail vein injection delivered .
LeTx into tumor cells and LeTx blocked MEK/MKK signaling pathways, as a result of

which the release of pro-angiogenic factors from tumor cells was suppressed. This was L;

supported by in vitro studies demonstrating that LeTx suppressed release of angiogenic
cytokines such as vascular endothelial growth factor (V EGF), interleukin-6 (IL-6) and
basic fibroblast growth factor (bFGF) (Depeille et al., 2007; Ding et al., 2008). However,
this hypothesis is challenged by the surprising observation fi'om the SK-MEL—28
xenograft study (presented in Chapter III of this dissertation) showing that systemic
administration did not deliver LeTx into tumor cells and the endogenous MEK in the
tumor cells was still intact. Similarly, Liu et al. (2008) reported that the growth of
xenograft tumors derived fi'om CHO cells lacking anthrax toxin receptors (i.e., these cells

are resistant to LeTx) was still inhibited by LeTx.

This leads me to propose an alternative hypothesis. Systemic treatment with
LeTx inhibits xenograft tumor growth by targeting a non-tumor compartment, which is
critical for melanoma-associated endothelial function, such as endothelial cells. This is

supported by several lines of evidence. First, endothelial cells are a potential target of

161

 

LeTx (Kirby, 2004; Warfel et al., 2005; Alfano et al., 2009). As discussed earlier in this
section, LeTx systemic treatment reduces microvessel density in xenograft tumors, and
also inhibits acute fibrosarcoma xenograft tumor perfusion within 24h. These studies
have shown that LeTx reduces endothelial cell viability and endothelial function. Second,
tumor-associated endothelial cells express elevated levels of the tumor endothelial marker
8 (TEMS), one of the identified anthrax toxin receptors (St Croix et al., 2000), potentially
making the cells more susceptible to LeTx. Finally, Mavria et al. (2006) have reported
that expression of dominant-negative MEKl specifically in tumor-associated
endothelium inhibited tumor growth, indicating the MEK signaling pathway is required
for tumor-associated endothelial function. Alternatively, LeTx may target another non-
tumor compartment that is essential for vascular function such as macrophages or

pericytes.

Regardless of which cell type LeTx is targeting, it is clear that inhibition of more
than one MEK/MKK pathway is an effective strategy for inhibiting tumor growth and
vascularization. Based on my study, I have proposed that MEK] cooperates with a not-
yet-identified factor to compensate for MEK2 function and to drive cell proliferation.
This factor is very likely to be another MKK protein or an MKK-regulated protein.
However, since the basal activities of p38 MAPK and JNK are very low in SK-MEL-28
cells (Figure 18), one cannot exclude the possibility that this factor may be a not-yet-
identified LF substrate. Initial attempts to identify this factor have proved unsuccessful.
However, its identification should be a priority since its function may be critical for

melanoma survival.

162

 

 

 

Appendix I. Transcriptional signatures that are down-regulated by LeTx treatment
and significantly rescued by MEK2cr

163

 

 

 

 

 

 

mEOmOmofimm 8m 250 mama: ”$3898 8." and mom- 5388 was? $28
causal/v.52 sum 25c mama: 85 one was..- mafia. 5.28
3325 new a; 86 _ S- mammoooa <58
Eewofl
volummw>ueo<<233<oo com 080 mama: mg mg mg. coverage 8.8%2 828 8?
m=8 SEE «6865
gauge 5 E2555 nous—>883
98 souaificfi mo eoumfinaoo
S|<o>o|mem§o§muo<n<me 2: .3 83329. 8858;
”sum 88 mama: ”2. w? an. Baggage? reap a? 880
clump? ”am 25c mama: 3N who N3- cocoa Sequence mm:
Siam? ”am 88 mama: $.m N? w _ .v. cocoa 835mg "am
magnum
Simmaammgugmmeoom 8m use mama: mam B; was- coinage Ex .3an .ES
Goom re a 2me com a? $5 m3- 88a coverage. 88
Slocuzmmm> ”am use mama: 3m :2 as. See coinage Emm
815.20g? new use mama: 8a o3 :é. x shoe 83683-0»: use a:
Goom :3 a. 2m: use m3. «2 is- a»:
a 88$; .5sz a Em: NSE> a3 afieiaafiaeafiaaa
-m> -m>

 

e 85:83

 

~0ng .3 6253.. 52:85:»? 6:: 2.25.3.5 who..— .3 wouiwouugee 0.3 :2: 3.53:»? _anetatomaauh A €=25<

164

 

 

 

 

83898 83. an m _ .m- 33:3 <75 82.3335
EZec<umm§mmng§eeoBopz

”com 080 mama—2 x88 :3 3 553$ 3m am a? 3m- 883. 3335-338352
EmnnofiemEamznegmlofiooém

com 080 mama: x88 :3 3 323% 3m 83 Ed Re. 8333908 gage

Emnnogemzlmaegm ”com 080 mama: N3. of 331 E3338 cause»;

«$8860 32 :2 ”3. Hana <75

eEn<ze_mnme<2<nu<zn ”am use mawnmz mam S; 33. 038% <23
202225285629

”am 2.8 mama: x88 :3 3 bespfifi 3m gm of 8m- garages»: <zo

2825 do own a? 83. 5353682 <75

5.53032 new 080 mama: ”2588 do 33” :2 one- 383:2

2325 do 83 32 32- 3832
xmnmzoolémSEOmnoDzoEm

”am 230 3%: x88 :3 e 56:33 3m 23 one on? 5388 aaoaooeéfia

$838 do own 5o 83. caches 08882

S325 Moo RN «2 9%. 388a 233308 <5:

3823 do 35. m3 3%- mamaoea <2”:

3 358.3% BmWME .5 ”Wm/HE Noe~-m> flow 0503>a35afimo§m-wmm

 

a 8:333

 

165

 

 

 

2.8 552 a .8833 32833

 

A88 3334 38m .3 3 333d .33 33 one 3.3.- 8% 33332.33 33 35 380
£3 3388 233 3230 x38. :3 3 333$ .33 3.». 32 8.3. 33 33338 23» 3230
m2 3332: 23» 3230 x38 :3 3 3338 .33 33.3 NS 8.3. m2 3332: 353 3230
8 2:32: 23» 3230 x38 .3 3 3339 .33 23 _E 33.3- S 33336 233 3230
manuzhflmae?

3m 230 mama: x88 ..3 3 38$ .33 mom one 2.3- 38» 3333832333
zauqq<ueufizlomz<§mm 1543-3
3m 230 mama: x88 3 3 8.38% 3m 33.3 33; 3.3. 32 a 333332-338 3:33 8,3
mowmsqobafi
n I 3.33 3.33 a 83383 9:
Zn. 32.3 3:3sz 3m 230 mama: N3 :._ £3- 8% 33338.338 33 35 830
zauaogguozfi 3233
33m 230 mama: x88 :3 3 38$ .3”. R3 33 Sm- 3383833333 sagagé
£8 5on
083553 58:: EBEBEEEE.
§|0mmn<§<mo<fi<mm 3m 230 mama: m3 m: 8.3- s 33333233 33 35 380
238900 3.3 a: 2.3.- 338322
23555 3m 236 mama: R3 33.“ $3- 83353 333 =8 $6
a 8:05.31: .8952 .8 SEE N633m> flow ocoOR>33€afimo§3—&mm

-m> -m>

 

3 85383

 

166

 

 

98% b.2558 0308 3.595
$159932; 3m 250 mama: SN. 23 an». «a 5 8332.9. as as 880
awmxézfio’xvofi 8m 250 mama: 8a. 8.? N3 “3-23:8 <zmeo€mo mamas
29mg 3%
gag—wot: ~83 298800 229::
Sla<xoz<§m2m> 8m 080 mafia: 35. 33. SN Ba 3585 3. x8880:
58898 EN. woo- ova mews; ooamoaofimoé
zalmmooupmfimomemlzammm 8:535.“ spam
pom 250 mama: x38 :3 3 328 «um 8+. 8. T 9% E 833?-..38 8“ as 850
high ogwoono mg
zgnzuw ”am 080 832 NE- 3o- o3 space .732 mo coofionfisz
gab:
O 093 SE88 63332:”
0521220 new 080 mama: 3a. 93. as 2:85 53% .6 Bofiofiwsz
manézomuéfiéa 8m 050 mawaz G3 33 SN- 8:% Baawoasaiaize
m=oo FADE S #58895 Eon 2888
soon £83 68m :3 3 send mom SN 84 SN- 8% 8332.9. 2“ 35 880
£8 E02
5 Beaumon— 3353333 829% ‘
EON £63 ”88. :3 3 Quad gem RN w _ A man. 8% 8838-598 03 as 880
a 885$ 63m: aim: N335 38 Sooigagmofigwa
-m> -m>

 

85:33

 

167

 

._Efido3wofil>nlom38£Bo§oB§zBB.Eomfiamdutch—chug: ”33on 09% 5050
”\wl—oxwo—ofioocomékrzrkugn ”came—8:0 33$ 00 Em_..aouwom\nnw_m8\somw\wuo335530~D§BB§at: ”flow 050 memmE a

.256 55 83on mos—g .

Q 96: £3 wfimmofixo.8~v~m2-m> 5 880m 05 was figs—tong 3580:: E09589: 08% Sch @0538 0.53 moumuflmé a

 

 

Room :3 3 56 «oz cod- wed- mwé 85% NEE
Cbom r3 8 Emfiév mom OWN- 3;. 9mm méz
a 858mom ~0ng .5552 Noc~-m> flow oaowhxmafimfimohsmzwmm

-m> -m>

 

a 8:383

 

168

 

Appendix II. Transcriptional signatures that are affected by LeTx treatment and
significantly rescued by both MEchr and MEK2cr

 

 

169

 

 

 

 

 

znuavfimISHZm 8m 25a mag: m; «3 3m- znuovfimlfinzm

m DEUCE 08w .5280 ~N.O~ V0.0 . mwdu m 24608 2.5m .5280

mm 07608 08w .5050 ©w.© 056 mod- mm DEUCE 05% .5050

Svozmufizo com 250 8%: 8.0 ”3 mod- Seozmufizo
<20m|~mzo com 250 3%: a; man 2d- <zomu~mzo
351$sz 8m 250 mama: m: can ad- _éufizo

M: 0388 use» 50:5 Rd 86 «2- M: 2358 8% 5050
za;Sonuo._<z<2 sum 250 mama: 8.5 can on? zausvnoimuoizg
2535:363an ”3m 080 Ema: 2.0 :6 end. 2535:5831on
_quoumfiufimzmg 8m 250 mama: owe mum o3- _quunmmmudmzmg
w; :3 3.2- 88» Bane? mm

.8882 ”e 8295 9; a: 8.2- .5 8m
505448:memomemnzammm m8

”6m 080 8%: a; one 32. Edmuuemfimommanzammm

em 0385 as» 5050 8:2 N; Sm 7 a 2358 8% H356

Q ggbwod how”: BMW—M2 Nufi~um> 90m 0G00\m%m>>5mm\mo§m&mw

 

a 83333

 

.8ng
was hos—HE 52— .3 1283.. .5.—35:»: E... 2.25.3.5 who:— .3 683:: 0.3 2:: 3:52—me iueflntomnauh .= anomad‘

 

170

 

 

”2 0258 25» beau 3m 5m 8s. m: 0388 08» beau

339; ”B 8295 w? «2. N3- .5 52%?

2:33: ”a 3295 3m 9% 8.»- n5 ”Bade:

Saddam? 8m 080 8me 26 gm :3. Snoodmmm>

m2 2:88 8% 5050 an own 3.”- g 0388 25w 588

w 2:on 25w 5280 a; v3“ vow- w 2358 25m .50ch

S 2358 08» beau one 0? 3w- 3 0388 25» 825

53.220 3m 080 mama: 8.0 «.3 v3- voéummzo

«Eozlfizo 8m 089 mama: 86 8+ m3- $621820

mm 0258 25m .850 R6 25 9;- Ma 0:62: 08w 5050
m§<zoa|22mooozod§

8m 080 832 as 86 3;- mmB<szdEmwoozo|m§

w: 2358 one» .5030 was go as- w: 0:62: 25» 525

_meummzo pom 250 mama: 3.0 33 m3- _meummzo
maumngmmmnzbmwmuozéo

sum 250 mama: Rs Sm mg- Summzonammlzammmuoimo

2 0358 25w 5050 3.9 8+ 2.x- 2 2:88 8% 8050

a REP—aux HONuWM—E .5 WWW”: Nou~um> 30m 0§U\m%m35wm\wo§mflmmm

 

a 89353

 

171

 

 

513%? ”am 080 mama: :3 Rd mi. 54293
2.228.320 8m 250 mama: :6 a? 8.». 2.2281320

SN 0358 25m 505.0 mg 8.9 NE. SN 2:88 08» beau

5. 2:85 8% 525 3.0 8.0 N2- 5 2258 8% .5030
$289.20 ”5m 250 8%: 03 a3 3. ~<zou|~220
m20eo<mmuzofi<oimmmn<zo m
”am 250 mama: 26 3m 3%. Eoeo<mmuzofi<oimmmu<zn

83:2 “e 8295 2+ 22 82.. 238 2.3on

88%— ”B 8282 9% am Re. a: 2522

50.21%sz 8m 250 3%: 82 2% was- 302.9220
maummfimoelmmq pom 250 mama: 3.0 2% Es- 318382.134
£282 ”9 8223 NE 36 m5- owsfio 35
mmqmfigmuzmmZmuzgtom am
new 080 mama: «E 9% 0:- qm<pu2mn=mm2mu22<zmmom

SN 0388 25w 525 m; N: ”5. N8 0358 25m beau
«2338;222:254 8m 250 mama: 3m 8N £5- Ememauumaauae.
mm 03.58 25m .850 one as 2:. mm 0388 8% 50.30

a 3:80.23— .BWWME SWHME Nou~-m> 30m ocochxmafiafimogacwmm

 

a 8:383

 

172

 

 

 

 

220:220 80 080 0000: 00+ 00.0 2.0- 220:220

800:220 “00 080 000202 3.0 8+ +00- 800:220

20200310220022 0.00 080 00062 :0 .2 +00- Su+0000|02<|2002<m

2010200<0|Hmamm|xoo ”000 080 000202 E0 8+ 2.0. 20:02020Qmammlxon

0058002 ”B 00202 00.0 2+ 2.0- <2 20300220005202

0200.220 ”000 080 mama: 00.0 2.0 2.0- 02001220

8:22? 000. 080 0000.2 :+ 00.~ 00.0- 8.22?

00021000029420 000 080 mama: 8.0 SN _00- 0002200061300
20n>mzezugum>uq<2215

000 0000 0005: 5% 00.0 2.0- 20u>mzo§|EIm>J<02|E

+2 03020 080 0850 5+ 00.0 5.0. +2 0302: 080 0880

000>0J020n2 _3<mm ”000 0:00 00022 00.0 00.... 2.0- 0005:2812???

2010221220 000 080 000202 2+ 20 80. 206021220

5006:4000 000 080 0005.2 8.0 2.0 8.0- 3006:1500

Rm 0302: 080 5800 3+ 0.0 8.0- 20 2:008 080 08:00

02 03008 080 5050 20 +2.0 0:- 02 03008 050 5800

20208 “00 00+ 20 02 .0. 80808.. <20 080882520

0 8:80.202 8w”: .8 ”WE N00~-m> 300 0000\020350Qmoggwmm

 

0 0020280-“

 

173

 

 

 

 

000001220 000 080 000002 :0 0++ ++.0- 000001220

002001220 000 0000 0000: 00+ 00+ 00.0- 002001220

02-01220 000 080 050002 00+ 00.0 00.0- 00.01220

0000000 ”00 00.0 00.0 00.0- 080 0008002

23022201330000.0000; 000 080 050002 00.0 0+.0 00.0- 2302220133555;

00+ 20008 0000 .8000 00+ 00.0 00.0- 00+ 03008 0000 0850

000000: ”B 000,300 00+ 00.0 00.0- .5 20 00. .22: 0580080000

2010023000102 000 080 050002 00.0 00+ 00.0- 2012<10001200

0000000 ”00 00+ 00 00.0- 0308 020005.00

0020000210020me 000 050 000002 +0+ 00+ 00.0- 00200021002200

00201001002202 000 080 000002 00.0 00+ 00.0- 00200100100280:

0012300002100 000 0:00 05000.2 _00 00.0 +0.0- 001250002100

000000: “B 002000 00.0 00.0 00.0- .5 2: 0 200228
00100005010210.05020000

000 080 05000.2 30 00.0 00.0- 001020.001021005020000

22221220 000 0000 050002 00.0 00+ 000. 02221220

0 000000-002 Hum-WW: BmW—ME N00~10> 0000 ogwhmaafimfimoganwmm

 

0 002000000

 

174

EL.

r70.-

 

 

 

00100002012-01002-00 000 080 000002 0+.0 2+ 051 001000020600100200
00 0 03008 050 .8000 00+ 00.0 00+- 000 20008 080 0050

000000: ”9 002000 00.0 00.0 00+. 00 28 0 00020232050800

020001220 000 080 0500: 00+ 00.0 00+. 020201220
0.0000002012055000 000 080 0500: 00+ ++.0 00+- 00000000000120.2300
0200201221202 000 0000 000002 +0.0 00.0 00+: 0200201221202
00:00: ”B 000,200 _00 00.0 00+- 00 00.055020

22001220 000 080 0000: 00+ 00.0 00+. 0002001220

0010001000 000 080 050002 00.0 0+0 00+. 0010001000
0200020020 000 080 050002 00.0 00.0 20. 0200020220
03020001222002010-0100200002 00002

000 0000 0000: 00+ 00.0 00.0- 00012200000210.0100200002

00221220 000 050 0500: 00+ 00.0 00.0- . 02221220
03000010010000001020 000 080 0000: 00+ 0+0 00.0- 0+000001o01m-00001dm0
1 1 <2

0000000_ ”B 0020.5 00+ 00.0 00.0- 0+E.E§000§.0000 0000 080.
00150210530222":02012025 0015.021
000 0000 050002 00.0 00.0 0+0- 00.03020020020120050

0 000000.002 bmWME 8 ”WE N00~1m> $00 ocowkxaBfimfimoggwmm

 

0 002000000

 

175

 

 

 

+00 03008 080 08000 00.0 00.0 00+. +00 03008 080 08000

50001220 000 080 000002 0++ 00.0 00+. 02001220

00000001000510.0000 000 080 000002 +0+ 00.0 00.? 00.200201000061020
0020010005 0002100021000

000 0:00 000002 00.0 00.0 +++- 00.20010030000010-002120

20001220 000 080 000002 00.0 00.0 0++- 20001220

00100<10025<0 000 080 0000: 00+ 00.0 0+.+- 00100<1o0200<0

201>z<102 000 080 000002 00.0 00.0 0+.+- 201>2<102

000000: ”00 002000 00.0 00.0 0+? 00 2: 0 222: 20000800

1 1 200005000201 1 1

00.20000 02200100 0220.50 000+0<02 2000500020 000.2002 02

000 0000 000002 +_+ 00.0 00+- 20.501022030100035:

02212200020020 000 080 000002 00.0 0.0 00+- . 221-2200020020

000000010012000.20000210000 0000

000 080 000002 :0 00.0 00+. 00010012022052.3000

300201220 000 050 000002 00.0 _00 00.01 500201220

0200200312100200000 000 080 0000: 0+ _0.0 00+- 02520031810020.0002

3030013000000: 000 080 0000: 00+ 00.0 20+- 30500130000002

0 Dog-5.00% 8w”: .5 WWW: NUQN1W> flow 0G00\m%035«m\m0§wcwmm

 

0 000000000

 

176

H...
LJ

 

 

 

 

000000: ”00 002000 00+ 00.0 00+. 2300 2: 00 00000
2.020.31002.2002201201200050002 1 15000-0300
000 0000 000002 00.0 00.0 00+- 502025:- 20 20250000
0030000 ”0. 002000 00+ 00.0 00+- 00 0.002
0 _ 0+000 U00 00+ 00.0 +0 .01 00,08 008302
001E1§1S§1§02102<12<<220> 1 1 1 1 “510.310
000 080 000002 +2.0 00.0 0_.+- > 52 :202 00,2 2<<220>
0000000 ”00 2+ 00.0 023 00202
000000 ”00 3+ 0.0 0 _ .+. 2000 =8 00220 00 00000 2
0000000 “00 00.0 00.0 0:1 00,000 0820 5380
002021220 000 080 00002 2+ 0+.0 00+. 002021220
E0200<100<22100120E<00002 1 1 1 E20.
000 0000 000002 00.0 00.0 00+- 0< 00302 .00 29.050000

'- I I I t I
0>F0< 00<2§ 220.02 00 200050002 1 1 21020-3 00<2
000 0:00 00002 00.0 0.0 00+- 2 200002 .00 29.2.5002
00120210000100.0020 000 080 000002 00.0 0+.0 00+. 00120210000100.0020
00<>1xom00<102<020 000 080 000002 00.0 00.0 00+- 03001202030252:
0000000 ”00 00+ 0.0 00+. 2000 08 000002
0 8020000 50002 6202 020-00, 003. 00000003500850.0000

-0> -0>

 

0 00200800

 

177

 

 

 

 

00000000 ”00 002000 00.0 00.0 00.0- 2300 2: 00 000008580
0010.2020010001005020020

000 080 000002 00.0 00+ 00.0- 001<2000010001005020000

00213001+0001002122<2002 200

000 050 0000: 00+ 000 00.0- 013001+0001002122<2002

0000000 ”00 _+.+ 00.0 00.0- 000008 =8 00 00003000

201-20.02 000 080 0000: 00+ 0++ 00.0- 201000.02

0000000 ”00 00+ 00.0 00. 0008000058 0000.

002012 000 080 000002 000 00.0 00.0- 003212

20350120231250: 0.00 0:00 000002 00.0 00.0 00.0- 20020510000000.1253

0000000 ”00 00.0 00.0 00.0- 0509.800 2000 :00

02025120022012? 000 080 000002 00.0 00.0 000- 00000201200322.1050

00000010000102.0002 000 080 00002 +0 00.0 00.0- 00051000010202:

0010-2100070 0.00 080 00002 00.0 00.0 00.0- 0010-21007:

00000301000103 000 050 0000: 00+ 000 0+. 02000301200103

00201000061020 000 080 000002 00.0 00.0 00+- 00201000061020

2010262002000 000 080 0000: 00+ 20 00+- 2010215002000

0 oocobmom pom-WM} bmwwmrzz N0001m> 0000 0000\000350Q00030020m

 

a momumflwwmuu

 

 

178

 

ringing-1.53.! 0.1... . .. .y

 

 

29002310201022.0000

 

000 0000 000002 00.0 00+ 00.0- 20000231020103.2000

E>F0<100<00002 000 050 000002 +0.0 00.0 00.0- 3030310020002

0.000020100001200 000 0000 000002 00.0 000 00.0- 02005100001200

20000001200103 000 080 0000: 00+ 000 00.0- 0.000000010000103

++0 0000 ”00 00+ 0++ +0 .0- 08802300 232803,

000 0000 ”00 0+ 00+ 000- 05802200 0002, 0020

00000002 ”00 002000 00+ 00.0 _00- 2300 20 000 0000008030

0000000 ”00 0+ 00.0 00.0- 08802300 000.0

\>om.:E.000.&0\\”&E 000w0 $000002 00000020000

- 0020... 00.0 002 600 020800 23800 00+ 00.0 00.0- ”02530002 0. 0+000

0010000021200 000 2.00 0000: 00+ 00.0 0+0- 001000202120.
20102080210212?me

000 050 000002 00.0 00.0 0+0- 201-02000021021000.5000
00010220<000<>1<0000<2

000 0000 000002 00.0 2+ 0+0- 00010020<200<>1<0000<2

2010+00210+0<1><0<00< 000 050 000002 00.0 00+ 00.0- 2010202105955? ,
20003010062102 000 800 0000: 00+ 00 00.0- 00000000010015.0109
0 000000002 BN-WME Hog-WM} N00~-m> 0000 0000\000350Q000300wmm

 

U momumflaumna

 

 

179

 

3:080:0>:0000:0<>_350:00:

00:00

 

000 0000 00002 00.0- 00.0. 00.0 00:0>:0000:0<>_>000:00m

020000 ”00 000. 00.0. 00.0. 2303 09.08-00.020

000005

000 _ 000 ”00 00.0 00.0 00.0- 000802008 00.80 00 00020000

0 03008 050 0800 00.0 ~00 00.0- 0 03008 0000 00000

0 330 000 200 00302 00.0 00.0 00.0- 0:: 000

000080 3:5ng

002000 ”00 00.0 00.0 00.0- 0308208 00 0002002 020000

0002050000020 000 200 00002 02.0 00.0 00.0- 006205000002":
00 20008 0000 50000 00.0 0 2 .0 00.0- 00 20008 080 00000

_00 20008 080 0200 00.0 2.0 00.0- am 03008 200 00000
20:x0000<:02<:a 000 200 0000: $0 00.0 00.0- 20:200034503
00:0000:>20::<2,0 000 200 000002 02.0 00.0 20.0- Edy—$920002”:
002000 ”00 00.0 00.0 :2. 0000005030 00080

0:850:20: 000 050 0000: 00:0. 3.0 00.0- 0:050:20

0000000 ”00 00.0 00.0 00.0- 0008000 =8 00 00003032

5:025:03 000 0000 000202 00.0 00.0 00.0- 5:025:03

a 0000.600M BWWME BMW—2 N007m> 0000 0000\020350E00500wmm

 

0 00005000..“

 

180

.Eb:.00_038I00I0038£000§0\0§002\500.0.8.053000:30:}9E ”001608 000w 000000

{waxwofiaoocowéghdta ”Ago—8:0 050V 00 E00.500000\€w_08\000w\m800300050000930.0030BE ”000m 0000 mama—2
.086 :05 088% 0030M
& 0%: 0:00 3800000095 ~vm=20> E 000000 05 0:0 000055020 00.58028 0000:0009: 00.05 80¢ 0000050 0.003 000000000-“ 0

181

Appendix III. Transcriptional signatures that are not affected by LF treatment but
cannot be rescued by either MEchr or MEK2cr

182

 

 

 

 

183

 

0000000 ”00 00.0 00.0 00.0- 000,000 008000 <20
20.003000200000000 000 200 00002 00.0 00.0 20. 20:0.0<0:0002000000::0
002000 “00 00.0 00.0 000- 0000000 <20E

0000000 ”00 00.0 000 00.0- 0000282 <20

002000 ”00 00.0 000 00.0- 00000080. <20 00 08600000000

0000000 ”00 00.0 000 00.0- 0000000 <20

0000000 ”00 00.0 00.0 00.0- 0000000 000 00002

00000.& 800000

0000000 ”00 00.0 00.2 00.0- 000800 00 0002002 30000

202552525 000 200 000002 00.0 00.0- 00.0- 2920902525
0_ _0000 ”00 00.0- 00.0 00.0- 0000>000 :8 ,0

0.80055 200 320:: 0:0

.000 _00 “00 0: 02 00.0- 0008000 0208000 00008002

00 00000 H00 02 00.0 00.0- 000,08 008000

0000002 US 002000 00.2 00.0 00.0- <2 000500.080.

0 000000030 BMWME 5 ”Wm/dz N00~:m> 0000 0000\000350E00030&_m

 

0 0030390-»

 

you—HE
.3 005—82 00500 00. 60:000.. 0.. 00:58 :5 «no—500.5 rad 03 68000.3 00: 9:0 :5» 00032.30 13030—20050..." 4: 0:50.30.

 

 

 

 

0500:000220000003000: : :05
000 0000 0:00:02 :00 0:0- 00.0- 00 00000200000>00<m 00:
03000002200000.03000: : @002
000 0:00 0.0002 :00 000. 00.0- :0 0002200000550 00:
0000000 ”00 00.0- 00.0- 00.0- 0000>000 :00 00 0000:0000 30000
00000600
0000000 ”00 00.0- 00.0- 00.0- 0000000200 000000000 030000
0003:000-002000000030d0: : :50
000 0000 000002 00.0 :0- 00.0-- 0< 00022000005020 00:
00: 03000. 0000 000000 00.0 :0- 00.0- 00: 0.0008 0000 00000
: : : <2: amt-000
0000000: ”a 002000 00.0- 00.0 00.0- 000:0 :0 00000000 00 00500.0
000000
0000000 ”00 00.0- 00.0- 00.0- 00000050 000008080000 00:2
20:00<-002:02200 000 0:00 00002 00.0 _0. _- 00.0- 20300502020000
0000000 ”00 00.: :0 00.0- 0000>000 :00 00 00003000
0000000 ”00 00.: 00.0 00.0- 00002000 000000320 00000000
00:03.00
0000000 ”00 00.0 :0- 00.0- =00 :0 0000:0000 020000
a 00:000-00M .8sz B~Mm2 N00~:m> 0000 0:00\0>03£0m\00030:w_m
-0> -0>

 

0 0000:0000

 

 

184

 

 

 

 

 

00Zm0unm0<000m0120000oz<az0>300 I 00:000-005-
000 0000 0:00:02 00.: 00.0 00.0- 00000 20000020. 2:500:00
201000120000400000 000 0000 000002 :00 00.0- 00.0- 20100023004300
050800 30800-00030
5300032305: Bob 0:05 0.800008
05:55 .00 0030:5889: 0:088
0000000 ”00 00.0 :00 00.0- 00 00000 00000000 000.000 300000-
00000500000002 000 0000 0:00:02 00.0 00.0 00.0- 0.0000002010005007:
000.5808 0
0000000 ”00 00.: :00 00.0- 000000 0000000 0:380 003000000:
2300
0000000: ”B 00200.: 0:: 00.0 00.0- :20 00: :22: 0000000000000
:000000 H00 00.: 00.0 00.0- 000000008 <20
0020010000000200001300000100200,: I 000000100
000 0000 0:00:02 00.: 00.0 00.0- 000002-000 $0000: 0020:):
£000.00: 000 0000 0:00:02 00.: 00.0 00.0- :0000u0002
303000 0080-0055—0508
0000000 “00 00.0 :00 00.0- 00000000000200.0000000000
2000010000,: 000 0000 00000:): :00 :00 00.0- 20000000:
0000000 Ho0 00.0 :00- 00.0- 000,000 0.000000 00000003000
0 8000.000 00000: 00 :00: 000:-0> 0:00 0000000300000000000
-0> -0>

 

0 00000000-“

 

185

 

 

 

 

 

$I2:<00:I0z:00< ”:00 0000 0:00:02 0:0- 0:0 0:.0 0::I2:<00I02:m0<
0000000: ”0: 002000 00.0- 00.0 0:.0 0300 000300500
>000I0<00002 000 0000 05000,: 00.0 0:.0- 00.0- >000I0<m0002
0000:00000<I020200200I0<000000 50:20
000 0000 0:00:02 00.: 00.0 00.0.. 00<I020200200I0<000000
00I:-:-<000> 000 0000 0:00:02 00.0 00.0 00.0.. 00I:-0<000>
E<00000I0EO0000< 000 0000 0:00:02 00.: . 0:: :00- 2<00000I00000000<
0000m>20Im00mz<000 000 0000 0:00:02 0:0- 00.0- 00.0- 00000020000020.0000
00000520 000 0000 0:00:02 0:0- 00.0- 00.0- 00000520
000:000 “00 000 00.0- 00.0- 0000>000 000000;:00 0000:0000

0000000 ”00 :0: 00.0 00.0- 00000 00000000002002 m
00 : 0000 ”00 0:0- 00.0- 00.0- 00,000 0.0000000000000002
I I >002000<I02<I0: I I I000
0020020: 00000200 E000000000020000 02000< az< 0:0-0200000: 00
000 0000 0:00:02 00.: 0:0- 00.0- 002002000000000020000
0000000 ”00 0:: 00.0 :00- 0.000000 000000500 <20
0300
000 0000 ”00 0:: 00.0 :00- 000000500 <20 0000000<zn
0 comebmum How-WM} 8 ”Wm/=2 N00~-m> 300 0500000350385035

 

0 00000800

 

 

 

 

0000000: ”0: 0002000 00.:- 0:.0 :00 00 :.00m>
000-0020002030 000 0000 00000:): 00.0- 00.: 00.0 000-002.0353
:0I00I:000> 000 0000 0:00:02 0:.0 00.0- 00.0 :0I00I:000>
:00:000 ”00 00.0- 00.0 00.0 =00 0: 00000200 00000000002

0000000 ”00 00.0 00.0 00.0 00000000000 000<

0000000 H00 00.0 00.0 00.0 000000 00 00000000000
000002<00I000<0002Im020m> 0.00
000 0000 0:00:02 00. :- 00.0- 00.0 002<00I000S002Im00000>

0000000: :0: 002000 00. :- :00- 00.0 0300 00.0>.0<00

:00 0:000:0 0000 000000 00. :- 00. :- 00.0 :00 030000 0000 000000

003800

0: :0000 ”00 :00- 00.0 00.0 00:00:00 :08 0005:0000 000:-0<

20:000I002I0000 ”:00 0000 000:0: 00. :- 00.0 00.0 20:0:0I002I0000
20:0000000I2000I0200I00000<0= 20:00
000 0000 0:00:02 00.0- :0 00.0 00000I2000I0200I00000<00

00I0<m00220 000 0000 0:00:02 00.0 :00- 00.0 0000-00022:
00I00>0> 000 0000 0000000,: 00. :- 00.0- 00.0 00I00>0>

0 000000-030 ENHWME BMW/=2 N373, 0000 00000000350000.0300me

 

0 00000500

 

187

 

 

 

 

00::0oo0> 000 0000 05000,: 00.0- 0:.0- 00.0 004,000?

0000000 000 00.0- 00.0- 00.0 00000000: 000000

0: :0000 do 00.0- 00.0- 00.0 0000 00000800,:

5.00000, 000 0000 05000,: 00.:- 00.0- 00.0 8100:0000

0000000: ”0:: 002000 00.:- 0:.0- 00.0 0300 0o.0>.0<0::

00'00m0> 000 0000 050:0: 00.0- 00.0- 00.0 00:00m0>

0000000 How 00.0- 00.0- 00.0 000:: 20000000:

0000000: 00:: 002000 :0.:- 0:.0 :0.0 7:300: 00:

0000000: ”0:: 00:200.: :0. :- 0:.0 :00 0000:0 50020

0: :0000 000 00.0- 00.0- 00.0 :08 00006

000:000 ”00 :0. :- 0 : .0 00.0 00:00:00 00 00000000:

0000000 How 00.0 00.0- 0: .0 :00 000:: 0000000000

8505024000590 000 0000 0:90:02 00. :- 00.0 :00 51000505005520;
\>ow.£::._o:.&_0\\x§: Emma warm—5100

- 000000 0:00: :07: 0000 0000600 0000000 00.0- 0:.0 00.0 0:20 2:000:00 0.00:000
€6.52.“

0000 :00 How 00.0 00.0- 00.0 0: 00000000 00000 00 0000:0000:

a 3:85.03: ENHWME BMW/:2 Nuu~-m> 080 ocoO\0>0350m\0o§00:wmm

 

0 00:00:30-:

 

188

._EE.0o3:02:50plo0300£Bo§o\0§Bz\=wo68.090000033:5090: :02:on 008w 08:80

<w8$w£3c85w§33<dtn 63285 00ng OD ”0:00.300000wa00aho0wh833505389333005: ”000m 050 mama—2
.086 :05 030on 002000“
R 020:: 0:8 w500oaxo-8 :Mm2-m> a: 00.800 2:: 98 000580598 000.5098? 0:353:05 025 89a 35030 0.83 00:00:80-: 0

189

Appendix IV. Transcriptional signatures that are not affected by LF treatment but
affected by VS-MEchr expression

190

 

 

 

 

 

 

 

20:00:00.: 000 0000 0:00:02 0:. :- 00.0- 00.0 20100105
00|0MB0IE<0010020 000 0000 0:00:02 00.: 00.0 00.0 00:00:30u2:<00:|0:0:00
0000000 000 0000 0:00:02 00.: 00.0 :00 00000 0:0
0030000500:
0000000 ”00 00.0 :00 :00 00000 00 0000:0000 0000007:
00:00.5; 0.00 0000 0:00:02 00.: 00.0 00.0 0010:0000
00:00.00?an: 00 0000 0:00:02 00.: 00.0 00.0 0010.230120:
00 0:0000: 0000 000000 00.0- 00.0- 00.0- 00 20:00.0 0000 08000
00105005002062.0300 000 0000 00002 00.: 00.0 00.0- 00.030050020620100
292003.300 000 0000 0:00:02 00.: 00.0 00.0- 20:05:21,500
05:00:me 000 0000 0:00:02 :0.: 0:.0 00.0- 0510:1000
001020.000 0.00 0000 0:00:02 00.: 00.0 00.0- 00:020..me

I I 000000003000 u
20200000000 0:0 20:05:00.2 05:20.02 0000000 0002020000050
000 0000 0:00:02 00.: 00.0 00.:- 100:7:0:0<,:000:0um>:,0<0mz
0 00008003: BMW”: 00 ”Wm/=2 N007m> 0000 0000000350m00§0§mm

 

0050580-:
0 . .

 

1]!

003000500 355—20> 03 6390b: :5 0:0—500.5 ha— »: cowuoba «0: 0.3 :05 09:50:90 _auoflntoauuh .>— 0:051:30

191

 

u .
‘-
F .04: 01... o .00.... 0'—

._EE.003008..»nl00300é00000800802300000.00.000.00000Q0t0g: ”00:60:: 000m 000000

”\w00xwo—0000000wégtfit: A3285 0:09 00 a030000033008000m\w00.0085050005.333:”BE ”000m 0000 mawmmz
. wood :05 000.00% 0020m
00 020:: 0:00 wE000ax0¢3MmE$> E 000000 05 0:0 0000800000 0000:00me 0000000005 00:5 80¢ 005030 0.53 00:00:30-0 0

192

Appendix V. Transcriptional signatures that are not affected by LF treatment but
affected by V5-MEK2cr expression

193

 

 

 

$088 ”00 $0. 3.0 8.0- 00.68 0:55 805..

00:E§> 00m 080 0006: SN 5.0 08- 00:E§>

2 0388 250 0850 3.0 ~00 8.0- a 2:88 25» 08:00

£088 ”00 RN 8.0 as- 8080 0328

£305 0%. 250 0005: 6+ 35 So- 3.: .50
8058 ”00 2.0 8.0 3.0.. 0505.0 50800

gingham? 00m 080 0006: 00m 2.0 3.0- 0013:0030,
£305 00m 080 0005: $0 03 8.0- $3 .0000
300000 0200000000

0888 ”00 02 NM: NM. 0- 800% 80?. 26080 08 8005

23:20 0% 250 000%: 0.3 3.0- 2.0- SE: .20
@388 ”00 8.0 2.0 mm. 0- as: 2,0080% 3 38080

$00000 00000050000

8: So ”00 02 3: mm. 0- 529a 00 800302 96080

5608 ”00 85 02 am. 0- 00.58 030108000
zaummmamélmmmfimxo 00m 250 00052 03 3.0 8.0- 20000050001300.0me
0 00000003— BmWME BMW—2 N00~-m> 30m 0000\30350Qm00300w0m

 

0 85:80-0

 

000800900 0085200» .3 000250 :5 000050000 ma .3 000000.00 00: 000 0000 0000000000 3000000000000. .> 00000007..

194

 

 

 

.0000:.00300000I>0100300£000§00§001300000-000000.000000%9E ”00300000 000w 000000

mh0oxw0000000000w§3§>900 ”ammo—00:0 00000 00 am20000033008000w\w00.003000500000.333\\uat: “000m 0000 meEE
.300 005 000.00% 830m
A 023 0:00 w00mm00qx0.000v~m2-m> 00 000000 08 05 5000000000900 0000000000000 0000000000000 0005 800-0 000000000 0003 0000000300 0

 

 

00.0 03008 080 08000 00.0 00.0 <2 0% 03008 080 08000

50000 ”00 00 .0- 00. 0- 00.0 E088 8000 020050

0.0 00.008 050 0850 00.0- 00. 0- 00.0 90 03008 250 08:00

02.0000 H00 00.0- 00.0- 00.0 0000088000

200:6 000 250 000002 00.0. 00.0 00.0 2000 .000
02-002.000-0020-00005 000 250 000002 00.0- 00.0- 00.0 02-002.000-002.0-0000<0
001000.000? 000 250 000002 00.0 00.0 00.0 00180-00000»
00000000008 0>00300£00000

.0000 .00 0000005000 00000000000000.0000

0000 _00 ”00 00.0- 00.0- 2.0 e 020:8 .0308 800-5.

0000 _00 ”00 00.0 00.0 00.0 05005 5835

Q oofiobmom BMW”: .5 “WM—2 NUQNum> meow oaomv\m%m>>5am\moéafiwmm

 

0 0000000800

 

195

Appendix VI. Transcriptional signatures that are not affected by LF treatment but
affected by both V5-MEchr and VS-MEKZcr expressions

196

 

 

 

 

00I00I0§00I0002I<050<0

 

000 080 000002 00.0 00.0 00. _- 00I00I0§00I0002I<0mm<0
00902900620000. 0%. 080 000002 00.0 :0 00. 0- 00902000062000.6-
00000000I20I<00Ixe 000 2000 0000: 0.0.0 00.0 2.0- 00000000I20I<00Ix00
zgozmzaIB0FzzzZzzm<<<>m I 2302
000 0000 000002 00.0 0.0.0 2.0- 0075 30007022223:
00I00<00000I0xmmm>04<0mm I I I 00
000 0000 000002 00.0 00.0 00.0- 0050000 000000-50 2000
0000000 ”00 00.0 00.0 mm. 0- 0500085 055800
0DImmmvamoiIEEBIEEDmIqum I I I SHEEN
000 050 000002 00.0 00.0 00.0- 00000 52.3 20.200 0000
20I0002020>Im§d<0§<mm I I z
000 0000 00000: 00.0 00.0 00.0- 0 00020.20.» 000-020-2500

I I I I 2000-E I
00000.6 20,000. 00 200050000 00003.0: I I 2000-000000050 20.00
000 080 000002 00.0 :0 00.0- < 00 2000-50000 0000.00.00
00I0>2Immq 000 080 000002 00.0. 02 00. 0- 00I0>2Immq
0 000000.030 003502 00 _anz N00~Im> 000m 0000\003350Q0003000w0m

.9 .3

 

0 00000000000

 

000000000900 00052
Im> 0:0 00 “Mm—20> 500 .3 00000000 0:: 0000050000 0: .3 0200b: 0000 000 005 00030500 .00050—0000000H .—> 05:00.50-

197

 

 

 

20I0§00I000000020I>200I>20 I I anImm:
000 250 000002 00.0 0.3. 00.0- 00 0000000200 >20: >20
00000000 ”B 82000 00.0 00.0 00.0- 2300 :00
0 03008 800 08000 00.0 00.0 00.0- 0 20008 080 08000
00 03008 050 08000 00.0 00.0 :0- 00 20008 080 000000
00600000000600.2002 000 080 000002 00.0 00.0 00.0- 00ISx00>mI00<z<2
20020202284080.0230: I I 2000
000 080 00.002 00.0 2.0 R0- <00ZD00200 0000 00200.00
20I0m<00200<Im0m§m§m2 I I
000 0000 000002 00.0 00.0 00.0- 20 002002.09. 000500.002
00.0.3020 000 080 000002 00.0 00.0 00.0- 00050020
02020000>00I002mmaam 000 080 000002 00.0. 00.0 00.0- 02020000>00Im02mmemm
00I00202000Imsmm-000 000 080 000002 00.0 00.0 00.0- 00Impz020~mIm<000000
000:5 000 080 000002 :01 00.0- 00.0- 0.00000
20I2<Imm§00m§0I>20mI>20 I I IzmI-0
000 00an0 000002 00.0 :0 00.0- 0< 0000000200 >200 >20
0000.2? 000 080 00052 00.0 00.0. 00.0- 00020?
000d 9500me 00260505 I I I I
000 080 00002 00.0 03 00.0- 00 00 0500: 0.52 0.00005-
0 8800000 003002 030002 Nos-g 03 080\m00350&mo§0&_m
-m> -0>

 

0 00000000000

 

198

".0
r4hll 5.4: ‘3I.EIJ.-.1

 

 

 

 

00m 20008 250 0850 00.0 00.0 00.0 000 03008 250 08:00

0:0-056000000 000 080 000002 00.0 00.0 00.0 000206000000-

0: 03008 080 0850 00.0 00.0 00.0 E 03005 250 08000

0085m0£ 000

00 00000 ”00 00.0 00.0 00.0 80305000 0.508 0000:8300

0.00 20020 080 08000 0.0.0 00.0. 00.0 «mm 20008 250 03000

N00 03008 250 08000 00.0 ~00 00.0 000 03008 080 000000

mad-20 000 080 000002 03.. 00.0 :0- _ mad-20

00ImMBNIz0§mI00000 000 080 000002 00.0 00.0 00.0- 00ImMB~Iz0§mI00000

2000§Im<00200§§m 000 050 000002 3.0 00.... 00.0- 200000Im<03000§0xm0

000000002Izo_mm:0<I-200I00 €00.00 000000-00

000 0000 000002 00.0 00.0 8.0. 2Izoamma<Idm0I2$0<mm

\>0w.£0.000.&0\\x€0_ wofig 0000800 ”020000.253

- 888 00.0 002 as... 0000600 8800000 00.0 00.0 00.0- 8050000.: 0.80:
00I0E0m0mI>szMI0z000<

000 250 000002 00.0 :0 00.0- §I0E0000I>szMI0Em0<

00I020Immoz0-E0m0 000 250 000002 00.0 00.0 00.0. 00I02mI-00020-E000

mm 03005 080 08000 00.0 00.0 00.0- 00 03005 080 0850

0 000000-030 BmWME bmW—ME N00~Im> 0000 0000320350Em0§00w0m

 

a moflmmumumnu

 

199

 

.Eng.mo_svofilxnlomzrthSogohgszB.3853.mouonetuntn ”mo—=88 28m .5230

”\muoxwo—ofioonowighfitn ”came—9:0 0:09 00 EmEouwomivwfifihommEho.Baumfiuaofi.333\\uat: ”30m 250 mama—2
.mood 55 083% $3.9m
m 03E mzoo wimmoaxo.8_v_m2-m> E 8.58 05 was .358598 08086:: 259532: 8:: 800m @0390 0003 85:83 a

 

 

200

SM 0386 25» 8080 3mm 30 N3 3 0388 8% 6050

$330 How :3. 8.0.- 3o :8 382on

3335 do :2. 8.0.- 3o Ba @3803

maua¢m0mu02<m gum 250 mawaz SN 3v 3o mbufifimgmnzm

o? 2:85 250. .850 93 2+ So o? 2:85 8% 5050

8320 new 25a mama: :6. 0mm- :3 8320

$28 ”oo 22 03 one macs; 86¢ 5320

ad 0255 25m .525 2.0 N: Rd as 03on 08m 588

$350 now 250 mama: 00.0.- V3- 85 8an
zo_mmma<lq<oon_uo_ 332%

”am 250 mama: Em 23 one zoammauiooflemgém

2 0388 gm 5050 03. o? one Q 2258 8% 5050

a 855.00% SNWME bmwxm/az Nuu~-m> 30m oaogmxfisfiafimoggwmm

 

a moumtfimé

 

Literature Cited

201

 

'C

 

 

Literature Cited
Abi-Habib RJ, Singh R, Leppla SH, Greene JJ, Ding Y, Berghuis B, Duesbery NS ,
Frankel AE (2006a). Systemic Anthrax Lethal Toxin Therapy Produces
Regressions of Subcutaneous Human Melanoma Tumors in Athymic Nude Mice.
Clin Cancer Res. 12, 7437-7443.

Abi-Habib RJ, Singh R, Liu S, Bugge TH, Leppla SH , Frankel AE (2006b). A
urokinase-activated recombinant anthrax toxin is selectively cytotoxic to many
human tumor cell types. Mol Cancer Ther. 5, 2556-2562.

Abi-Habib RJ, Urieto J 0, Liu S, Leppla SH, Duesbery NS , Frankel AE (2005). BRAF
status and mitogen-activated protein/extracellular signal-regulated kinase kinase
1/2 activity indicate sensitivity of melanoma cells to anthrax lethal toxin. Mol
Cancer Ther. 4, 1303-1310.

Adjei AA, Cohen RB, Franklin W, Morris C, Wilson D, Molina JR, Hanson LJ, Gore L,
Chow L, Leong S, Maloney L, Gordon G, Simmons H, Marlow A, Litwiler K, ,.
Brown S, Poch G, Kane K, Haney J , Eckhardt SG (2008). Phase I H
Pharmacokinetic and Pharrnacodynamic Study of the Oral, Small-Molecule
Mitogen-Activated Protein Kinase Kinase 1/2 Inhibitor AZD6244 (ARRY-

142886) in Patients With Advanced Cancers. J Clin Oncol. 26, 2139-2146.

 

Agrawal A, Lingappa J, Leppla SH, Agrawal S, Jabbar A, Quinn C , Pulendran B (2003).
Impairment of dendritic cells and adaptive immunity by anthrax lethal toxin.
Nature. 424, 329-334.

Ahn NG, Seger R, Bratlien RL, Diltz CD, Tonks NK , Krebs EG (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.
Journal of Biological Chemishy. 266, 4220—4227.

Ahn NG, Stines Nahreini T, Tolwinski NS, Resing KA , W.E. Balch CJDaAH (2001).
Pharmacologic Inhibitors of MK] and MKKZ. In Methods in Enzymology):
Academic Press, pp. 417-431.

Aitken J, Welch J, Duffy D, Milligan A, Green A, Martin N , Hayward N (1999).
CDKNZA variants in a population-based sample of Queensland families with
melanoma. J Natl Cancer Inst. 91, 446-452.

Alessi DR, Cuenda A, Cohen P, Dudley DT , Saltiel AR (1995). PD 098059 Is a Specific
Inhibitor of the Activation of Mitogen-activated Protein Kinase Kinase in Vitro
and in Vivo. J. Biol. Chem. 270, 27489-27494.

Alessi DR, Saito Y, Campbell DG, Cohen P, Sithanandam G, Rapp U, Ashworth A,

Marshall CJ , Cowley S (1994). Identification of the sites in MAP kinase kinase-
] phosphorylated by p74raf-l. EMBO J. 13, 1610-1619.

202

Alfano RW, Leppla SH, Liu S, Bugge TH, Meininger CJ, Lairmore TC, Mulne AF,
Davis SH, Duesbery NS , Frankel AE (2009). Matrix metalloproteinase—activated
anthrax lethal toxin inhibits endothelial invasion and neovasculature formation
during in vitro morphogenesis. Mol Cancer Res. 7, 452-461.

Allen LF, Sebolt-Leopold J , Meyer MB (2003). CI-1040 (PD184352), a targeted signal
transduction inhibitor of MEK (MAPKK). Seminars in Oncology. 30, 105-116.

Altekruse SF, Kosary CL, Krapcho M, Neyman N, Aminou R, Waldron W, Ruhl J,
Howlader N, Tatalovich Z, Cho H, Mariotto A, Eisner MP, Lewis DR, Cronin K,
Chen HS, Feuer EJ, Stinchcomb DG , Edwards BK: SEER Cancer Statistics
Review, 1975-2007, based on November 2009 SEER data submission, posted to
the SEER web site, 2010. (Accessed June 19, 2010)

American Cancer Society (2010). Cancer Facts & Figures 2010edAeds). Atlanta:
American Cancer Society.

Anderson NG, Maller J L, Tonks NK , Sturgill TW (1990). Requirement for integration
of signals from two distinct phosphorylation pathways for activation of MAP
kinase. Nature. 343, 651-653.

Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, Davis AP, Dolinski
K, Dwight SS, Eppig JT, Harris MA, Hill DP, Issel-Tarver L, Kasarskis A, Lewis
S, Matese J C, Richardson J E, Ringwald M, Rubin GM , Sherlock G (2000). Gene
ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat
Genet. 25, 25-29.

Astier AL, Xu R, Svoboda M, Hinds E, Munoz 0, de Beaumont R, Crean CD, Gabig T ,
Freedman AS (2003). Temporal gene expression profile of human precursor B
leukemia cells induced by adhesion receptor: identification of pathways
regulating B-cell survival. Blood. 101, 1118-1127.

Balch CM, Gershenwald J E, Soong SJ, Thompson JF, Atkins MB, Byrd DR, Buzaid AC,
Cochran AJ, Coit DG, Ding S, Eggermont AM, F laherty KT, Gimotty PA,
Kirkwood JM, McMasters KM, Mihm MC, Jr., Morton DL, Ross MI, Sober AJ ,
Sondak VK (2009). Final version of 2009 AJ CC melanoma staging and
classification. J Clin Oncol. 27, 6199-6206.

Ball NJ, Yohn JJ, Morelli JG, Norris DA, Golitz LE , Hoefiler JP (1994). Ras mutations
in human melanoma: a marker of malignant progression. J Invest Dermatol. 102,
285-290.

Bardwell AJ, Abdollahi M , Bardwell L (2004). Anthrax lethal factor-cleavage products

of MAPK (mitogen—activated protein kinase) kinases exhibit reduced binding to
their cognate MAPKs. Biochem. J. 378, 569-577.

203

 

 

 

Belanger L-F, Roy S, Tremblay M, Brott B, Steff A-M, Mourad W, Hugo P, Erikson R ,
Charron J (2003). Mek2 Is Dispensable for Mouse Growth and Development. Mol.
Cell. Biol. 23, 4778-4787.

Bild AH, Yao G, Chang JT, Wang Q, Potti A, Chasse D, Joshi MB, Harpole D, Lancaster
JM, Berchuck A, Olson JA, Jr., Marks JR, Dressman HK, West M , Nevins JR
(2006). Oncogenic pathway signatures in human cancers as a guide to targeted
therapies. Nature. 439, 353-357.

Birck A, Ahrenkiel V, Zeuthen J, Hon-Jensen K , Guldberg P (2000). Mutation and
allelic loss of the PTEN/MMACI gene in primary and metastatic melanoma
biopsies. J Invest Dermatol. 114, 277-280.

Borg A, Johannsson U, Johannsson O, Hakansson S, Westerdahl J, Masback A, Olsson
H , Ingvar C (1996). Novel gerrnline p16 mutation in familial malignant
melanoma in southern Sweden. Cancer Res. 56, 2497-2500.

Bosenberg M, Muthusamy V, Curley DP, Wang Z, Hobbs C, Nelson B, Nogueira C,
Homer J W, 2nd, Depinho R , Chin L (2006). Characterization of melanocyte-
specific inducible Cre recombinase transgenic mice. genesis. 44, 262-267.

Boulton TG, Yancopoulos GD, Gregory J S, Slaughter C, Moomaw C, Hsu J , Cobb MH
(1990). An insulin-stimulated protein kinase similar to yeast kinases involved in
cell cycle control. Science. 249, 64-67.

Bradley KA, Mogridge J, Mourez M, Collier RJ , Young JA (2001). Identification of the
cellular receptor for anthrax toxin. Nature. 414, 225-229.

Brai g M , Schmitt CA (2006). Oncogene-induced senescence: putting the brakes on
tumor development. Cancer Res. 66, 2881 -2884.

Bromberg-White J L, Boguslawski E , Duesbery NS (2009). Perturbation of Mouse
Retinal Vascular Morphogenesis by Anthrax Lethal Toxin. PLoS ONE. 4, e6956.

Bromberg-White J L , Duesbery NS (2008). Biological and biochemical characterization
of anthrax lethal factor, a proteolytic inhibitor of MEK signaling pathways.
Methods Enzymol. 438, 355-365.

Cannon-Albright LA, Goldgar DE, Meyer LJ, Lewis CM, Anderson DE, Fountain JW,
Hegi ME, Wiseman RW, Petty EM, Bale AE , et a1. (1992). Assignment of a
locus for familial melanoma, MLM, to chromosome 9p13-p22. Science. 258,
1 148-1 152.

204

 

 

Catalanotti F, Reyes G, J esenberger V, Galabova-Kovacs G, de Matos Simoes R, Carugo
O , Baccarini M (2009). A Mekl-Mek2 heterodimer determines the strength and
duration of the Erk signal. Nat Struct Mol Biol. 16, 294-303.

Chang HY, Sneddon J B, Alizadeh AA, Sood R, West RB, Montgomery K, Chi JT, van
de Rijn M, Botstein D , Brown PO (2004). Gene expression signature of
fibroblast serum response predicts human cancer progression: similarities between
tumors and wounds. PLoS Biol. 2, E7.

Chen Z, Gibson TB, Robinson F, Silvestro L, Pearson G, Xu B-e, Wright A, Vanderbilt
C , Cobb MH (2001). MAP Kinases. Chemical Reviews. 101, 2449-2476.

Chin L, Pomerantz J, Polsky D, Jacobson M, Cohen C, Cordon-Cardo C, Homer JW,
2nd , DePinho RA (1997). Cooperative effects of IN K4a and ras in melanoma
susceptibility in vivo. Genes Dev. 11, 2822-2834.

Chin L, Tam A, Pomerantz J, Wong M, Holash J, Bardeesy N, Shen Q, O'Hagan R,
Pantginis J, Zhou H, Homer JW, 2nd, Cordon-Cardo C, Yancopoulos GD ,
DePinho RA (1999). Essential role for oncogenic Ras in tumour maintenance.
Nature. 400, 468-472.

Chong H, Vikis HG , Guan KL (2003). Mechanisms of regulating the Raf kinase family.
Cell Signal. 15, 463-469.

Chopra AP, Boone SA, Liang X , Duesbery NS (2003). Anthrax Lethal Factor
Proteolysis and Inactivation of MAPK Kinase. J. Biol. Chem. 278, 9402-9406.

Clark WH, Jr., Elder DE, Guerry Dt, Epstein MN, Greene MH , Van Horn M (1984). A
study of tumor progression: the precursor lesions of superficial spreading and
nodular melanoma. Hum Pathol. 15, 1147-1165.

ClinicalTrials.gov Identifier: NCT00147550 , US. National Institutes of Health. A
service of the US. National Institutes of Health: ClinicalTrials. gov Identifier:
NCT00147 550. MEK Inhibitor PD-325901 To Treat Advanced Breast Cancer,
Colon Cancer, And Melanoma. http://clinicaltrials. gov/ct2/show/N CT00147 550
(Accessed August 12, 2009)

Cohen C, Zavala-Pompa A, Sequeira JH, Shoji M, Sexton DG, Cotsonis G, Cerimele F,
Govindarajan B, Macaron N , Arbiser J L (2002). Mitogen-actived Protein Kinase
Activation Is an Early Event in Melanoma Progression. Clin Cancer Res. 8, 3728-
3733.

Cohen P (1999). The development and therapeutic potential of protein kinase inhibitors.
Current Opinion in Chemical Biology. 3, 459-465.

205

 

 

 

Collisson EA, De A, Suzuki H, Gambhir SS , Kolodney MS (2003). Treatment of
Metastatic Melanoma with an Orally Available Inhibitor of the Ras-Raf-MAPK
Cascade. Cancer Res. 63, 5669-5673.

Cowan JM, Halaban R , Francke U (1988). Cytogenetic analysis of melanocytes from
premalignant nevi and melanomas. J Natl Cancer Inst. 80, 1159-1164.

Crews CM , Erikson RL (1992). Purification of a murine protein-tyrosine/threonine
kinase that phosphorylates and activates the Erk-1 gene product: relationship to
the fission yeast byrl gene product. Proc Natl Acad Sci U S A. 89, 8205-8209.

Dankort D, Curley DP, Cartlidge RA, Nelson B, Karnezis AN, Damsky Jr WE, You MJ,
DePinho RA, McMahon M , Bosenberg M (20093). BratV600E cooperates with
Pten loss to induce metastatic melanoma. Nat Genet. 41, 544-552.

Dankort D, Curley DP, Cartlidge RA, Nelson B, Karnezis AN, Damsky WE, Jr., You MJ,
DePinho RA, McMahon M , Bosenberg M (2009b). Braf(V600E) cooperates
with Pten loss to induce metastatic melanoma. Nat Genet. 41, 544-552.

Dankort D, Filenova E, Collado M, Serrano M, Jones K , McMahon M (2007). A new
mouse model to explore the initiation, progression, and therapy of BRAFV600E-
induced lung tumors. Genes & Development. 21, 379-384.

Davies H, Bignell GR, Cox C, Stephens P, Edkins S, Clegg S, Teague J, Woffendin H,
Garnett MJ, Bottomley W, Davis N, Dicks E, Ewing R, Floyd Y, Gray K, Hall S,
Hawes R, Hughes J, Kosrnidou V, Menzies A, Mould C, Parker A, Stevens C,
Watt S, Hooper S, Wilson R, J ayatilake H, Gusterson BA, Cooper C, Shipley J,
Hargrave D, Pritchard-J ones K, Maitland N, Chenevix-Trench G, Riggins GJ,
Bigner DD, Palmieri G, Cossu A, Flanagan A, Nicholson A, Ho JWC, Leung SY,
Yuen ST, Weber BL, Seigler HF, Darrow TL, Paterson H, Marais R, Marshall CJ,
Wooster R, Stratton MR , Futreal PA (2002). Mutations of the BRAF gene in
human cancer. Nature. 417, 949-954.

Davies SP, Reddy H, Caivano M , Cohen P (2000). Specificity and mechanism of action
of some commonly used protein kinase inhibitors. Biochem. J. 351, 95-105.

Dent P, Haser W, Haystead TA, Vincent LA, Roberts TM , Sturgill TW (1992).
Activation of mitogen-activated protein kinase kinase by v-Raf in NIH 3T3 cells .
and in vitro. Science. 257, 1404-1407.

Depeille P, Young JJ, Boguslawski EA, Berghuis BD, Kort EJ, Resau J H, Frankel AE ,
Duesbery NS (2007). Anthrax Lethal Toxin Inhibits Growth of and Vascular
Endothelial Growth Factor Release fi'om Endothelial Cells Expressing the Human
Herpes Virus 8 Viral G Protein Coupled Receptor. Clin Cancer Res. 13, 5926-
5934.

206

 

 

Ding Y, Boguslawski EA, Berghuis BD, Young JJ, Zhang Z, Hardy K, Furge K, Kort E,
Frankel AE, Hay RV, Resau J H , Duesbery NS (2008). Mitogen-activated protein
kinase kinase signaling promotes growth and vascularization of fibrosarcoma.
Mol Cancer Ther. 7, 648-658.

Dudley DT, Pang L, Decker SJ, Bridges AJ , Saltiel AR (1995). A synthetic inhibitor of
the mitogen-activated protein kinase cascade. Proceedings of the National
Academy of Sciences of the United States of America. 92, 7686-7689.

Duesbery NS, Resau J, Webb CP, Koochekpour S, Koo HM, Leppla SH , Vande Woude
GF (2001). Suppression of ras-mediated transformation and inhibition of tumor
growth and angiogenesis by anthrax lethal factor, a proteolytic inhibitor of
multiple MEK pathways. Proceedings of the National Academy of Sciences of the
United States of America. 98, 4089-4094.

Duesbery NS, Webb CP, Leppla SH, Gordon VM, Klirnpel KR, Copeland TD, Ahn NG,
Oskarsson MK, Fukasawa K, Paull KD , Vande Woude GF (1998). Proteolytic
Inactivation of MAP-Kinase-Kinase by Anthrax Lethal Factor. Science. 280, 734-
737.

Eisen T, Ahmad T, Flaherty KT, Gore M, Kaye S, Marais R, Gibbens I, Hackett S, James
M, Schuchter LM, Nathanson KL, Xia C, Simantov R, Schwartz B, Poulin-
Costello M, O'Dwyer PJ , Ratain MJ (2006). Sorafenib in advanced melanoma: a
Phase II randomised discontinuation trial analysis. Br J Cancer. 95, 581-586.

English J, Pearson G, Wilsbacher J, Swantek J, Karandikar M, Xu S , Cobb MH (1999).
New Insights into the Control of MAP Kinase Pathways. Experimental Cell
Research. 253, 255-270.

Erickson AK, Ray LB , Sturgill TW (1990). Microtubule-associated protein 1A is the
fibroblast HMW MAP undergoing mitogen-stimulated serine phosphorylation.
Biochem Biophys Res Commun. 166, 827-832.

Estrada Y, Dong J , Ossowski L (2009). Positive crosstalk between ERK and p38 in

melanoma stimulates migration and in vivo proliferation. Pigment Cell Melanoma
Res. 22, 66-76.

Favata MF, Horiuchi KY, Manos EJ, Daulerio AJ, Stradley DA, Feeser WS, Van Dyk
DE, Pitts WJ, Earl RA, Hobbs F, Copeland RA, Magolda RL, Scherle PA ,
Trzaskos J M (1998). Identification of a Novel Inhibitor of Mitogen-activated
Protein Kinase Kinase. .1. Biol. Chem. 273, 18623-18632.

Ferrando AA, Armstrong SA, Neuberg DS, Sallan SE, Silverrnan LB, Korsmeyer SJ ,
Look AT (2003). Gene expression signatures in MLL-rearranged T-lineage and
B-precursor acute leukemias: dominance of HOX dysregulation. Blood. 102, 262-
268.

207

 

 

Fischrnann TO, Smith CK, Mayhood TW, Myers J E, Reichert P, Mannarino A, Carr D,
Zhu H, Wong J, Yang R-S, Le HV , Madison VS (2009). Crystal Structures of
MEK] Binary and Ternary Complexes with Nucleotides and Inhibitors.
Biochemistry. 48, 2661-2674.

FitzGerald MG, Harkin DP, Silva-Arrieta S, MacDonald DJ, Lucchina LC, Unsal H,
O'Neill E, Koh J, Finkelstein DM, Isselbacher KJ, Sober AJ , Haber DA (1996).
Prevalence of gerrn-line mutations in p16, p19ARF, and CDK4 in familial
melanoma: analysis of a clinic-based population. Proc Natl Acad Sci U S A. 93,
8541-8545.

Flores JF, Pollock PM, Walker GJ, Glendening JM, Lin AH, Palmer J M, Walters MK,
Hayward NK , Fountain JW (1997). Analysis of the CDKN 2A, CDKN2B and
CDK4 genes in 48 Australian melanoma kindreds. Oncogene. 15, 2999-3005.

Fountain JW, Karayiorgou M, Emstoff MS, Kirkwood J M, Vlock DR, Titus-Ernstoff L,
Bouchard B, Vijayasaradhi S, Houghton AN, Lahti J , et al. (1992). Homozygous
deletions within human chromosome band 9p21 in melanoma. Proc Natl Acad Sci
USA. 89, 10557-10561.

Friday BB, Yu C, Dy GK, Smith PD, Wang L, Thibodeau SN , Adjei AA (2008). BRAF
V600E Disrupts AZD6244-Induced Abrogation of Negative Feedback Pathways
between Extracellular Signal-Regulated Kinase and Raf Proteins. Cancer Res. 68,
6145-6153.

Friedlander AM (1998). Antitumor effects of anthrax lethal toxin. In Third International
Conference on AnthraxedAeds). University of Plymouth, pp. 104.

Furge KA, Chen J, Koeman J, Swiatek P, Dykema K, Lucin K, Kahnoski R, Yang XJ ,
Teh BT (2007a). Detection of DNA copy number changes and oncogenic
signaling abnormalities from gene expression data reveals MYC activation in
hi gh- grade papillary renal cell carcinoma. Cancer Res. 67, 3171-3176.

Furge KA, Tan MH, Dykema K, Kort E, Stadler W, Yao X, Zhou M , Teh BT (2007b).
Identification of deregulated oncogenic pathways in renal cell carcinoma: an

integrated onco genomic approach based on gene expression profiling. Oncogene.
26, 1346-1350.

Gentleman RC, Carey VJ, Bates DM, Bolstad B, Dettling M, Dudoit S, Ellis B, Gautier L,
Ge Y, Gentry J, Homik K, Hothorn T, Huber W, Iacus S, Irizarry R, Leisch F, Li
C, Maechler M, Rossini AJ, Sawitzki G, Smith C, Smyth G, Tierney L, Yang JY ,
Zhang J (2004). Bioconductor: open software development for computational
biology and bioinformatics. Genome Biol. 5, R80.

208

 

 

 

Gershenwald JE, Soong SJ , Balch CM (2010). 2010 TNM staging system for cutaneous
melanoma...and beyond. Ann Surg Oncol. 17, 1475-1477.

Ghosh P (2004). Process of Protein Transport by the Type III Secretion System.
Microbial. Mal. Biol. Rev. 68, 771-795.

Giroux S, Tremblay M, Bernard D, Cardin-Girard J F, Aubry S, Larouche L, Rousseau S,
Huot J, Landry J, J eannotte L , Charron J (1999). Embryonic death of Mekl-
deficient mice reveals a role for this kinase in angiogenesis in the labyrinthine
region of the placenta. Current Biology. 9, 369.

Gomez N , Cohen P (1991). Dissection of the protein kinase cascade by which nerve
growth factor activates MAP kinases. Nature. 353, 170-173.

Govindarajan B, Bai X, Cohen C, Zhong H, Kilroy S, Louis G, Moses M , Arbiser JL
(2003). Malignant Transformation of Melanocytes to Melanoma by Constitutive
Activation of Mitogen-activated Protein Kinase Kinase (MAPKK) Signaling. J.
Biol. Chem. 278, 9790-9795.

 

Gray-Schopfer V, Wellbrock C , Marais R (2007). Melanoma biology and new targeted
therapy. Nature. 445, 851-857.

Guldberg P, thor Straten P, Birck A, Ahrenkiel V, Kirkin AF , Zeuthen J (1997).
Disruption of the MMACl/PTEN gene by deletion or mutation is a frequent event
in malignant melanoma. Cancer Res. 57, 3660-3663.

Haass NK, Sproesser K, Nguyen TK, Contractor R, Medina CA, Nathanson KL, Herlyn
M , Smalley KSM (2008). The Mitogen-Activated Protein/Extracellular Signal-
Regulated Kinase Kinase Inhibitor AZD6244 (ARRY-142886) Induces Growth
Arrest in Melanoma Cells and Tumor Regression When Combined with
Docetaxel. Clin Cancer Res. 14, 230-239.

Hao Y-H, Wang Y, Burdette D, Mukherjee S, Keitany G, Goldsmith E , Orth K (2008).
Structural requirements for Yersinia YopJ inhibition of MAP kinase pathways.
PLoS ONE. 3, el375.

Harland M, Meloni R, Gruis N, Pinney E, Brookes S, Spurr NK, Frischauf AM, Bataille
V, Peters G, Cuzick J, Selby P, Bishop DT , Bishop JN (1997). Germline
mutations of the CDKN2 gene in UK melanoma families. Hum Mol Genet. 6,
2061-2067.

Hatzivassiliou G, Song K, Yen I, Brandhuber B], Anderson DJ, Alvarado R, Ludlam MJ,
Stokoe D, Gloor SL, Vigers G, Morales T, Aliagas 1, Lin B, Sideris S, Hoeflich
KP, Jaiswal BS, Seshagiri S, Koeppen H, Belvin M, Friedman LS , Malek S
(2010). RAF inhibitors prime wild-type RAF to activate the MAPK pathway and
enhance growth. Nature. 464, 431-435.

209

 

 

Heidom SJ, Milagre C, Whittaker S, Nourry A, Niculescu-Duvas I, Dhomen N, Hussain J,
Reis-Filho J S, Springer CJ, Pritchard C , Marais R (2010). Kinase-dead BRAF
and oncogenic RAS cooperate to drive tumor progression through CRAF. Cell.
140, 209-221.

Herbst RA, Weiss J, Ehnis A, Cavenee WK , Arden KC (1994). Loss of heterozygosity
for 10q22-10qter in malignant melanoma progression. Cancer Res. 54, 31 11-3114.

Huang CY , Ferrell J E, Jr. (1996). Ultrasensitivity in the mitogen-activated protein
kinase cascade. Proc Natl Acad Sci U S A. 93, 10078-10083.

Huang D, Ding Y, Luo W-M, Bender S, Qian C-N, Kort B, Zhang Z-F, VandenBeldt K,
Duesbery NS, Resau JH , Teh BT (2008). Inhibition of MAPK Kinase Signaling
Pathways Suppressed Renal Cell Carcinoma Growth and Angiogenesis In vivo.
Cancer Res. 68, 81 -88.

 

Hussussian CJ, Struewing JP, Goldstein AM, Higgins PA, Ally DS, Sheahan MI), Clark
WH, Jr., Tucker MA , Dracopoli NC (1994). Germline p16 mutations in familial
melanoma. Nat Genet. 8, 15-21.

Hwang PH, Yi HK, Kim DS, Nam SY, Kim JS , Lee DY (2001). Suppression of
tumorigenicity and metastasis in B16F 10 cells by PTEN/MMACl/TEPI gene.
Cancer Lett. 172, 83-91.

Johnson GL , Lapadat R (2002). Mitogen-Activated Protein Kinase Pathways Mediated
by ERK, JNK, and p38 Protein Kinases. Science. 298, 1911-1912.

Kamakura S, Moriguchi T , Nishida E (1999). Activation of the Protein Kinase
ERKS/BMKI by Receptor Tyrosine Kinases. IDENTIFICATION AND
CHARACTERIZATION OF A SIGNALING PATHWAY TO THE NUCLEUS.
J. Biol. Chem. 274, 26563-26571.

Kamb A, Gruis NA, Weaver-Feldhaus J, Liu Q, Harshman K, Tavtigian SV, Stockert E,
Day RS, 3rd, Johnson BE , Skolnick MH (1994). A cell cycle regulator
potentially involved in genesis of many tumor types. Science. 264, 436-440.

Kanehisa M, Araki M, Goto S, Hattori M, Hirakawa M, Itoh M, Katayama T, Kawashima
S, Okuda S, Tokimatsu T , Yamanishi Y (2008). KEGG for linking genomes to
life and the environment. Nucleic Acids Res. 36, D480-484.

Kannan K, Sharpless NE, Xu J, O'Hagan RC, Bosenberg M , Chin L (2003).

Components of the Rb pathway are critical targets of UV mutagenesis in a murine
melanoma model. Proc Natl Acad Sci U S A. 100, 1221-1225.

210

Kelsall SR , Mintz B (1998). Metastatic cutaneous melanoma promoted by ultraviolet

radiation in mice with transgene-initiated low melanoma susceptibility. Cancer
Res. 58, 4061-4065.

Kirby J E (2004). Anthrax Lethal Toxin Induces Human Endothelial Cell Apoptosis.
Infect. Immun. 72, 430-439.

Knudson AG, Jr. (1971). Mutation and cancer: statistical study of retinoblastoma. Proc
Natl Acad Sci U S A. 68, 820-823.

Kohno M , Pouyssegur J (2003). Pharmacological inhibitors of the ERK signaling
pathway: application as anticancer drugs. Prog Cell Cycle Res. 5, 219-224.

Kohno M , Pouyssegur J (2006). Targeting the ERK signaling pathway in cancer therapy.

Ann Med. 38, 200-211.

Koo H-M, VanBrocklin M, McWilliams MJ, Leppla SH, Duesbery NS , Woude GFV
(2002). Apoptosis and melanogenesis in human melanoma cells induced by
anthrax lethal factor inactivation of mitogen-activated protein kinase kinase.

Proceedings of the National Academy of Sciences of the United States of America.
99, 3052-3057.

Kyriakis JM, App H, Zhang XF, Banerjee P, Brautigan DL, Rapp UR , Avruch J (1992).
Raf-1 activates MAP kinase-kinase. Nature. 358, 417-421.

Lamb J (2007). The Connectivity Map: a new tool for biomedical research. Nat Rev
Cancer. 7, 54-60.

Lamb J, Crawford ED, Peck D, Modell JW, Blat IC, Wrobel MJ, Lerner J, Brunet JP,
Subramanian A, Ross KN, Reich M, Hieronymus H, Wei G, Armstrong SA,
Haggarty SJ, Clemons PA, Wei R, Carr SA, Lander ES , Golub TR (2006). The
Connectivity Map: using gene-expression signatures to connect small molecules,
genes, and disease. Science. 313, 1929-1935.

Lee CS , Duesbery NS (2010). Highly Selective MEK inhibitors. Current Enzyme
Inhibition (in press).

Lee P, Wallace E, Yeh T, Poch G, Litwiler K, Pheneger T, Lyssikatos J , Winkler J
(2004). ARRY-142886, a potent and selective MEK inhibitor: III) Efficacy in
murine xenograft models correlates with decreased ERK phosphorylation. In

Annual Meeting of the American Association for Cancer ResearchedAeds), pp.
897.

Levine AJ (2009). The common mechanisms of transformation by the small DNA tumor

viruses: The inactivation of tumor suppressor gene products: p53. Virology. 384,
285-293.

211

 

P

 

Li J, Yen C, Liaw D, Podsypanina K, Bose S, Wang SI, Puc J, Miliaresis C, Rodgers L,
McCombie R, Bigner SH, Giovanella BC, Ittrnann M, Tycko B, Hibshoosh H,
Wigler MH , Parsons R (1997). PTEN, a putative protein tyrosine phosphatase
gene mutated in human brain, breast, and prostate cancer. Science. 275, 1943-
1947.

Limon J, Dal Cin P, Sait SN, Karakousis C , Sandberg AA (1988). Chromosome changes
in metastatic human melanoma. Cancer Genet Cytogenet. 30, 201-211.

Liu S, Wang H, Currie BM, Molinolo A, Leung HJ, Moayeri M, Basile JR, Alfano RW,
Gutkind J S, Frankel AE, Bugge TH , Leppla SH (2008). Matrix
Metalloproteinase—activated Anthrax Lethal Toxin Demonstrates High Potency in
Targeting Tumor Vasculature. J. Biol. Chem. 283, 529-540.

LoRusso P, Krishnarnurthi S, Rinehart JR, Nabell L, Croghan G, Varterasian M, Sadis SS,
Menon SS, Leopold J , Meyer MB (2005a). A phase 1-2 clinical study of a
second generation oral MEK inhibitor, PD 0325901 in patients with advanced

cancer. In American Society of Clinical Oncology Annual MeetingedAeds), pp.
301 1.

LoRusso PM, Adjei AA, Varterasian M, Gadgeel S, Reid J, Mitchell DY, Hanson L,
DeLuca P, Bruzek L, Piens J, Asbury P, Van Becelaere K, Herrera R, Sebolt-
Leopold J , Meyer MB (2005b). Phase I and Pharmacodynamic Study of the Oral
MEK Inhibitor CI-1040 in Patients With Advanced Malignancies. J Clin Oncol.
23, 5281-5293.

Lowe SW , Sherr CJ (2003). Tumor suppression by Ink4a-Arf: progress and puzzles.
Curr Opin Genet Dev. 13, 77-83.

Lynch HT, Fusaro RM, Danes BS, Kimberling WJ , Lynch JF (1983). A review of
hereditary malignant melanoma including biomarkers in familial atypical multiple
mole melanoma syndrome. Cancer Genet Cytogenet. 8, 325-358.

Lyssikatos J, Yeh T, Wallace E, Marsh V, Bemat B, Gross S, Evans R, Colwell H, Parry
J, Baker S, Ballard J, Morales T, Smith D, Brandhuber B , Winkler] (2004).
ARRY-l42886, a potent and selective MEK inhibitor: I) ATP-independent
inhibition results in high enzymatic and cellular selectivity. In Annual Meeting of
the American Association for Cancer ResearchedAeds), pp. 896-b.

MacKeigan JP, Collins TS , Ting JP (2000). MEK inhibition enhances paclitaxel-
induced tumor apoptosis. J Biol Chem. 275, 38953-38956.

Mansour SJ, Matten WT, Hermann AS, Candia J M, Rong S, Fukasawa K, Vande Woude

GF , Ahn NG (1994a). Transformation of mammalian cells by constitutively
active MAP kinase kinase. Science. 265, 966-970.

212

 

 

Mansour SJ, Resing KA, Candi J M, Hermann AS, Gloor JW, Herskind KR, Wartmann M,
Davis RJ , Ahn NG (1994b). Mitogen-activated protein (MAP) kinase
phosphorylation of MAP kinase kinase: determination of phosphorylation sites by
mass spectrometry and site-directed mutagenesis. J Biochem. 116, 304-314.

Mao L, Merlo A, Bedi G, Shapiro GI, Edwards CD, Rollins BJ , Sidransky D (1995). A
novel p16INK4A transcript. Cancer Res. 55, 2995-2997.

Mavria G, Vercoulen Y, Yeo M, Paterson H, Karasarides M, Marais R, Bird D ,
Marshall CJ (2006). ERK-MAPK signaling opposes Rho-kinase to promote
endothelial cell survival and sprouting during angiogenesis. Cancer Cell. 9, 33-44.

Menon SS, Whitfield LR, Sadis S, Meyer MB, Leopold J, Lorusso PM, Krishnamurthi S,
Rinehart JR, Nabell L , Croghan G (2005). Pharmacokinetics (PK) and
pharmacodynamics (PD) of PD 0325901, a second generation MEK inhibitor
after multiple oral doses of PD 0325901 to advanced cancer patients. In American
Society of Clinical Oncology Annual MeetingedAeds), pp. 3066.

Michaloglou C, Vredeveld LC, Soengas MS, Denoyelle C, Kuilrnan T, van der Horst CM,
Majoor DM, Shay JW, Mooi WJ , Peeper DS (2005). BRAFE600-associated
senescence-like cell cycle arrest of human naevi. Nature. 436, 720-724.

Miller AJ , Mihm MC, Jr. (2006). Melanoma. N Engl J Med. 355, 51-65.

Mittal R, Peak-Chew S-Y , McMahon HT (2006). Acetylation of MEK2 and IicB kinase
(IKK) activation loop residues by YopJ inhibits signaling. Proceedings of the
National Academy of Sciences. 103, 18574-18579.

Mody N, Leitch J, Armstrong C, Dixon J , Cohen P (2001). Effects of MAP kinase
cascade inhibitors on the MKKS/ERKS pathway. FEBS letters. 502, 21-24.

Mukherjee S, Keitany G, Li Y, Wang Y, Ball HL, Goldsmith EJ , Orth K (2006).
Yersinia YopJ Acetylates and Inhibits Kinase Activation by Blocking
Phosphorylation. Science. 312, 1211-1214.

Nakielny S, Cohen P, Wu J , Sturgill T (1992). MAP kinase activator from insulin-
stimulated skeletal muscle is a protein threonine/tyrosine kinase. EMBO J. 11,
2123-2129.

Naumov GN, Akslen LA , Folkrnan J (2006). Role of angiogenesis in human tumor
dormancy: animal models of the angiogenic switch. Cell Cycle. 5, 1779-1787.

Nobori T, Miura K, Wu DJ, Lois A, Takabayashi K , Carson DA (1994). Deletions of

the cyclin-dependent kinase-4 inhibitor gene in multiple human cancers. Nature.
368, 753-756.

213

 

 

Ohren ] F, Chen H, Pavlovsky A, Whitehead C, Zhang E, Kuffa P, Yan C, McConnell P,
Spessard C, Banotai C, Mueller WT, Delaney A, Omer C, Sebolt-Leopold J,
Dudley DT, Leung IK, Flamme C, Warmus J, Kaufman M, Barrett S, Tecle H ,
Hasemann CA (2004). Structures of human MAP kinase kinase 1 (MEKl) and
MEK2 describe novel noncompetitive kinase inhibition. Nat Struct Mol Biol. 11,
1192-1197.

Orth K, Palmer LE, Bao ZQ, Stewart S, Rudolph AE, Bliska JB , Dixon JE (1999).
Inhibition of the Mitogen-Activated Protein Kinase Kinase Superfamily by a
Yersinia Effector. Science. 285, 1920-1923.

Orth K, Xu Z, Mudgett MB, Bao ZQ, Palmer LE, Bliska JB, Mange] WF, Staskawicz B ,
Dixon J E (2000). Disruption of Signaling by Yersinia Effector YopJ, a Ubiquitin-
Like Protein Protease. Science. 290, 1594-1597.

Palmer LE, Hobbie S, Galan J E , Bliska JB (1998). YopJ of Yersinia pseudotuberculosis
is required for the inhibition of macrophage TNF-alpha production and
downregulation of the MAP kinases p38 and JNK. Mal Microbial. 27, 953-965.

Palmer LE, Pancetti AR, Greenberg S , Bliska J B (1999). YopJ of Yersinia spp. Is
Sufficient To Cause Downregulation of Multiple Mitogen-Activated Protein
Kinases in Eukaryotic Cells. Infect. Immun. 67, 708-716.

Pannifer AD, Wong TY, Schwarzenbacher R, Renatus M, Petosa C, Bienkowska J, Lacy
DB, Collier RJ, Park S, Leppla SH, Hanna P , Liddington RC (2001). Crystal
structure of the anthrax lethal factor. Nature. 414, 229-233.

Park J M, Greten FR, Li Z-W , Karin M (2002). Macrophage Apoptosis by Anthrax
Lethal Factor Through p38 MAP Kinase Inhibition. Science. 297, 2048-2051.

Parmiter AH, Balaban G, Clark WH, Jr. , Nowell PC (1988). Possible involvement of the
chromosome region 10q24----q26 in early stages of melanocytic neoplasia.
Cancer Genet Cytogenet. 30, 313-317.

Patterson TA, Lobenhofer EK, Fulmer-Smentek SB, Collins PJ, Chu TM, Bao W, Fang
H, Kawasaki ES, Hager J, Tikhonova IR, Walker SJ, Zhang L, Hurban P, de
Longueville F, Fuscoe J C, Tong W, Shi L , Wolfinger RD (2006). Performance
comparison of one-color and two-color platforms within the MicroArray Quality
Control (MAQC) project. Nat Biotechnol. 24, 1 140-1150.

Payne DM, Rossomando AJ, Martino P, Erickson AK, Her J H, Shabanowitz J, Hunt DF,
Weber M] , Sturgill TW (1991). Identification of the regulatory phosphorylation
sites in pp42/mitogen-activated protein kinase (MAP kinase). EMBO J. 10, 885-
892.

214

Pearson G, Robinson F, Beers Gibson T, Xu BE, Karandikar M, Berrnan K , Cobb MH
(2001). Mitogen-activated protein (MAP) kinase pathways: regulation and
physiological functions. Endocr Rev. 22, 153-183.

Pedersen MI, Bennett JW , Wang N (1986). Nonrandom chromosome structural
aberrations and oncogene loci in human malignant melanoma. Cancer Genet
Cytogenet. 20, 11-27.

Pellizzari R, Guidi-Rontani C, Vitale G, Mock M , Montecucco C (1999). Anthrax lethal
factor cleaves MKK3 in macmphages and inhibits the LPS/IFNgamma-induced
release of NO and TNFalpha. FEBS Lett. 462, 199-204.

Peng T, Golub TR , Sabatini DM (2002). The immunosuppressant rapamycin mimics a
starvation-like signal distinct from amino acid and glucose deprivation. Mol Cell
Biol. 22, 5575-5584.

Petty EM, Gibson LH, Fountain JW, Bolognia J L, Yang-Peng TL, Housman DE , Bale
AE (1993). Molecular definition of a chromosome 9p21 germ-line deletion in a
woman with multiple melanomas and a plexiform neurofibroma: implications for
9p turnor-suppressor gene(s). Am J Hum Genet. 53, 96-104.

Platz A, Hansson J, Mansson-Brahme E, Lagerlof B, Linder S, Lundqvist E, Sevigny P,
Inganas M , Ringborg U (1997). Screening of gerrnline mutations in the
CDKN2A and CDKN2B genes in Swedish families with hereditary cutaneous
melanoma. J Natl Cancer Inst. 89, 697-702.

Polager S , Ginsberg D (2009). p53 and E2f: partners in life and death. Nat Rev Cancer.
9, 73 8-748.

Pollock PM, Harper UL, Hansen KS, Yudt LM, Stark M, Robbins CM, Moses TY,
Hostetter G, Wagner U, Kakareka J, Salem G, Pohida T, Heenan P, Duray P,
Kallioniemi O, Hayward NK, Trent J M , Meltzer PS (2003). High frequency of
BRAF mutations in nevi. Nat Genet. 33, 19-20.

Poulikakos PI, Zhang C, Bollag G, Shokat KM , Rosen N (2010). RAF inhibitors
transactivate RAF dimers and ERK signalling in cells with wild-type BRAF.
Nature. 464, 427-430.

Powell MB, Gause PR, Hyman P, Gregus J, Lluria-Prevatt M, Nagle R , Bowden GT
(1999). Induction of melanoma in TPras transgenic mice. Carcinogenesis. 20,
1 7 47 -1 753.

Quelle DE, Zindy F, Ashmun RA , Sherr CJ (1995). Alternative reading frames of the

INK4a tumor suppressor gene encode two unrelated proteins capable of inducing
cell cycle arrest. Cell. 83, 993-1000.

215

 

Rapp UR, Goldsborough MD, Mark GE, Bonner TI, Groffen J, Reynolds FH, Jr. ,
Stephenson JR (1983). Structure and biological activity of v-raf, a unique
' oncogene transduced by a retrovirus. Proc Natl Acad Sci U S A. 80, 4218-4222.

Ray LB , Sturgill TW (1987). Rapid stimulation by insulin of a serine/threonine kinase
in 3T3-L1 adipocytes that phosphorylates microtubule-associated protein 2 in
vitro. Proc Natl Acad Sci U S A. 84, 1502-1506.

Ray LB , Sturgill TW (1988a). Characterization of insulin-stimulated microtubule-
associated protein kinase. Rapid isolation and stabilization of a novel
serine/threonine kinase from 3T3-L1 cells. J Biol Chem. 263, 12721-12727.

Ray LB , Sturgill TW (1988b). Insulin-stimulated microtubule-associated protein kinase
is phosphorylated on tyrosine and threonine in vivo. Proc Natl Acad Sci U S A. 85,
3753-3757.

Rinehart J, Adjei AA, LoRusso PM, Waterhouse D, Hecht JR, Natale RB, Hamid O,
Varterasian M, Asbury P, Kaldjian EP, Gulyas S, Mitchell DY, Herrera R, Sebolt-
Leopold J S , Meyer MB (2004). Multicenter Phase II Study of the Oral MEK
Inhibitor, CI-1040, in Patients With Advanced Non-Small-Cell Lung, Breast,
Colon, and Pancreatic Cancer. J Clin Oncol. 22, 4456-4462.

Roberts PJ , Der CJ (2007). Targeting the Raf-MEK-ERK mitogen-activated protein
kinase cascade for the treatment of cancer. Oncogene. 26, 3291-3310.

Robertson GP, Furnari FB, Miele ME, Glendening MJ, Welch DR, Fountain JW, Lugo
TG, Huang HJ , Cavenee WK (1998). In vitro loss of heterozygosity targets the
PTEN/MMACl gene in melanoma. Proc Natl Acad Sci U S A. 95, 9418-9423.

Rossomando AJ, Payne DM, Weber MJ , Sturgill TW (1989). Evidence that pp42, a
major tyrosine kinase target protein, is a mitogen—activated serine/threonine
protein kinase. Proc Natl Acad Sci U S A. 86, 6940-6943.

Rouleau C, Menon K, Boutin P, Guyre C, Yoshida H, Kataoka S, Perricone M, Shankara
S, Frankel AE, Duesbery NS, Vande Woude G, Biemann HP , Teicher BA
(2008). The systemic administration of lethal toxin achieves a growth delay of
human melanoma and neuroblastoma xenografts: assessment of receptor
contribution. IntJ Oncol. 32, 739-748.

Roux PP , Blenis J (2004). ERK and p38 MAPK-Activated Protein Kinases: a Family of
Protein Kinases with Diverse Biological Functions. Microbial. Mal. Biol. Rev. 68,
320-344.

Ruckdeschel K, Harb S, Roggenkarnp A, Hornef M, Zumbihl R, Kohler S, Heesemann J ,

Rouot B (1998). Yersinia enterocolitica Irnpairs Activation of Transcription
Factor NF-kappa B: Involvement in the Induction of Programmed Cell Death and

216

in the Suppression of the Macrophage Tumor Necrosis Factor alpha roduction.
J. Exp. Med. 187, 1069-1079.

Sagata N, Oskarsson M, Copeland T, Brumbaugh J , Vande Woude GF (1988). Function
of c-mos proto-onco gene product in meiotic maturation in Xenopus oocytes.
Nature. 335, 519-525.

Sato C, Nishizawa K, Nakayama T , Kobayashi T (1985). Effect upon mitogenic
stimulation of calcium-dependent phosphorylation of cytoskeleton—associated
350,000- and 80,000-mol-wt polypeptides in quiescent 3Y1 cells. J. Cell Biol. 100,
748-753.

Sato C, Nishizawa K, Nakayama T, Nose K, Takasaki Y, Hirose S , Nakamura H (1986).
Intranuclear appearance of the phosphorylated form of cytoskeleton-associated

350-kDa proteins in Ul-ribonucleoprotein regions after growth stimulation of
fibroblasts. Proc Natl Acad Sci U S A. 83, 7287-7291.

Schechter I , Berger A (1967). On the size of the active site in proteases. I. Papain.
Biochem Biophys Res Commun. 27, 157-162. _

Schesser K, Spiik A-K, Dulcuzumuremyi J -M, Neurath MF, Pettersson S , Wolf-Watz H
(1998). The yopJ locus is required for Yersinia-mediated inhibition of NF-kappaB
activation and cytokine expression: YopJ contains a eukaryotic SH2-like domain
that is essential for its repressive activity. Molecular Microbiology. 28, 1067-1079.

Scholl FA, Dumesic PA, Barragan DI, Charron J , Khavari PA (2009a). Mekl/2 gene
dosage determines tissue response to oncogenic Ras signaling in the skin.
Oncogene. 28, 1485-1495.

Scholl FA, Dumesic PA, Barragan DI, Harada K, Charron J , Khavari PA (2009b).
Selective Role for Mekl but not Mek2 in the Induction of Epidermal Neoplasia.
Cancer Res. 69, 3772-3778.

Scobie HM, Rainey GJ, Bradley KA , Young JA (2003). Human capillary
morphogenesis protein 2 functions as an anthrax toxin receptor. Proc Natl Acad
Sci USA. 100, 5170-5174.

Sebolt-Leopold J S, Dudley DT, Herrera R, Becelaere KV, Wiland A, Gowan RC, Tecle
H, Barrett SD, Bridges A, Przybranowski S, Leopold WR , Saltiel AR (1999).
Blockade of the MAP kinase pathway suppresses growth of colon tumors in vivo.
Nat Med. 5, 810-816.

Sebolt-Leopold J S , Herrera R (2004). Targeting the mitogen-activated protein kinase
cascade to treat cancer. Nat Rev Cancer. 4, 937-947.

217

Sebolt-Leopold J S, Merriman R, Omer C, Tecle H, Bridges A, Klohs W, Loi C-M, Valik
H, Przybranowski S, Meyer M , Leopold WR (2004). The biological profile of
PD 0325901: A second generation analog of CI-1040 with improved
pharmaceutical potential. In Annual Meeting of the American Association for
Cancer ResearchedAeds), pp. 925.

Segal E, Friedman N, Koller D , Regev A (2004). A module map showing conditional
activity of expression modules in cancer. Nat Genet. 36, 1090-1098.

Seger R, Ahn NG, Posada J, Munar ES, Jensen AM, Cooper JA, Cobb MH , Krebs EG
(1992). Purification and characterization of mitogen-activated protein kinase
activator(s) from epidermal growth factor-stimulated A431 cells. Journal of
Biological Chemistry. 267, 14373-14381.

Serrano M, Hannon GJ , Beach D (1993). A new regulatory motif in cell-cycle control
causing specific inhibition of cyclin D/CDK4. Nature. 366, 704-707.

Singh Y, Liang X , Duesbery NS (2005). Pathogenesis of Bacillus anthracis: the role of
Anthrax toxins. In Microbial Toxins. Molecular and Cellular Biology. (T Profi,
ed). Norfolk, England: Horizon Bioscience, pp. 285-312.

Solit DB, Garraway LA, Pratilas CA, Sawai A, Getz G, Basso A, Ye Q, Lobo JM, She Y,
Osman I, Golub TR, Sebolt-Leopold J, Sellers WR , Rosen N (2006). BRAF
mutation predicts sensitivity to MEK inhibition. Nature. 439, 358.

Soufir N, Avril MF, Chompret A, Demenais F, Bombled J, Spatz A, Stoppa-Lyonnet D,
Benard J , Bressac-de Paillerets B (1998). Prevalence of p16 and CDK4 germline

mutations in 48 melanoma-prone families in France. The French Familial
Melanoma Study Group. Hum Mol Genet. 7, 209-216.

St Croix B, Rago C, Velculescu V, Traverso G, Romans KE, Montgomery E, Lal A,
Riggins GJ, Lengauer C, Vogelstein B , Kinzler KW (2000). Genes expressed in
human tumor endothelium. Science. 289, 1197-1202.

Stahl J M, Cheung M, Sharma A, Trivedi NR, Shanmugam S , Robertson GP (2003).
Loss of PTEN promotes tumor development in malignant melanoma. Cancer Res.
63, 2881-2890.

Steck PA, Pershouse MA, J asser SA, Yung WK, Lin H, Ligon AH, Langford LA,
Baumgard ML, Hattier T, Davis T, Frye C, Hu R, Swedlund B, Teng DH ,
Tavtigian SV (1997). Identification of a candidate tumour suppressor gene,
MMACI , at chromosome 10q23.3 that is mutated in multiple advanced cancers.
Nat Genet. 15, 356-362.

218

 

Stone S, Jiang P, Dayananth P, Tavtigian SV, Katcher H, Parry D, Peters G , Kamb A
(1995). Complex structure and regulation of the P16 (MTSI) locus. Cancer Res.
55, 2988-2994.

Sturgill TW , Ray LB (1986). Muscle proteins related to microtubule associated protein-
2 are substrates for an insulin-stimulatable kinase. Biochem Biophys Res Commun.
134, 565-571.

Tanami H, Irnoto I, Hirasawa A, Yuki Y, Sonoda I, Inoue J, Yasui K, Misawa-Furihata A,
Kawakami Y , Inazawa J (2004). Involvement of overexpressed wild-type BRAF
in the growth of malignant melanoma cell lines. Oncogene. 23, 8796.

Tanoue T, Adachi M, Moriguchi T , Nishida E (2000). A conserved docking motif in
MAP kinases common to substrates, activators and regulators. Nat Cell Biol. 2,
110-116.

Tarhini AA , Agarwala SS (2006). Cutaneous melanoma: available therapy for
metastatic disease. Dermatol Ther. 19, 19-25.

Teng DH, Hu R, Lin H, Davis T, Iliev D, Frye C, Swedlund B, Hansen KL, Vinson VL,
Gumpper KL, Ellis L, El-Naggar A, Frazier M, J asser S, Langford LA, Lee J,
Mills GB, Pershouse MA, Pollack RE, Tomos C, Troncoso P, Yung WK, Fujii G,
Berson A, Steck PA , et al. (1997). MMACl/PTEN mutations in primary tumor
specimens and tumor cell lines. Cancer Res. 57, 5221-5225.

Tian B, Nowak DE, J amaluddin M, Wang S , Brasier AR (2005). Identification of direct
genomic targets downstream of the nuclear factor-kappaB transcription factor
mediating tumor necrosis factor signaling. J Biol Chem. 280, 17435-17448.

Trisciuoglio D, Iervolino A, Zupi G , Del Bufalo D (2005). Involvement of PI3K and
MAPK Signaling in bcl-2-induced Vascular Endothelial Growth Factor
Expression in Melanoma Cells. Mol. Biol. Cell. 16, 4153-4162.

Tsao H, Zhang X, Benoit E , Haluska PG (1998). Identification of PTEN/MMACI
alterations in uncultured melanomas and melanoma cell lines. Oncogene. 16,
3397-3402.

Tsao H, Zhang X, Fowlkes K , Haluska F G (2000). Relative Reciprocity of NRAS and
PTEN/MMACI Alterations in Cutaneous Melanoma Cell Lines. Cancer Res. 60,
1 800-1 804.

van 't Veer LJ, Burgering BM, Versteeg R, Boot AJ, Ruiter DJ, Osanto S, Schrier PI ,

Bos J L (1989). N-ras mutations in human cutaneous melanoma from sun-exposed
body sites. Mol Cell Biol. 9, 3114-3116.

219

Vitale G, Bemardi L, Napolitani G, Mock M , Montecucco C (2000). Susceptibility of
mitogen-activated protein kinase kinase family members to proteolysis by anthrax
lethal factor. Biochem. J. 352, 739-745.

Vitale G, Pellizzari R, Recchi C, Napolitani G, Mock M , Montecucco C (1998).
Anthrax Lethal Factor Cleaves the N-Terrninus of MAPKKs and Induces
Tyrosine/Threonine Phosphorylation of MAPKs in Cultured Macrophages.
Biochemical and Biophysical Research Communications. 248, 706-711.

Voisin L, Julien C, Duharnel S, Gopalbhai K, Claveau I, Saba-El-Leil M, Rodrigue-
Gervais I, Gaboury L, Lamarre D, Basik M , Meloche S (2008). Activation of
MEKI or MEK2 isoform is sufficient to fully transform intestinal epithelial cells
and induce the formation of metastatic tumors. BMC Cancer. 8, 337.

Wajapeyee N, Serra RW, Zhu X, Mahalingam M , Green MR (2008). Oncogenic BRAF
induces senescence and apoptosis through pathways mediated by the secreted
protein IGFBP7. Cell. 132, 363-374.

Walker GJ, Hussussian CJ, Flores JF, Glendening JM, Haluska FG, Dracopoli NC,
Hayward NK , Fountain JW (1995). Mutations of the CDKN2/p16INK4 gene in
Australian melanoma kindreds. Hum Mol Genet. 4, 1845-1852.

Wallace E, Yeh T, Lyssikatos J, Winkler J, Lee P, Marlow A, Hurley B, Marsh V, Bemat
B, Evans R, Colwell H, Ballard J, Morales T, Smith D, Brandhuber B, Gross S,
Poch G, Litwiler K, Hingorani G, Otten J, Sullivan F, Blake J, Pheneger T,
Goyette M , Koch K (2004). Preclinical development of ARRY-142886, a potent
and selective MEK inhibitor. In Annual Meeting of the American Association for
Cancer ResearchedAeds), pp. 897-a.

Warfel JM, Steele AD , D'Agnillo F (2005). Anthrax Lethal Toxin Induces Endothelial
Barrier Dysfimction. Am J Pathol. 166, 1871-1881.

Warmus J S, Flamme C, Zhang LY, Barrett S, Bridges A, Chen H, Gowan R, Kaufman M,
Sebolt-Leopold J, Leopold W, Merriman R, Ohren J, Pavlovsky A,
Przybranowski S, Tecle H, Valik H, Whitehead C , Zhang B (2008). 2-
Alkylamino- and alkoxy-substituted 2-amino-1,3,4-oxadiazoles--O-Alkyl
benzohydroxamate esters replacements retain the desired inhibition and selectivity
against MEK (MAP ERK kinase). Bioarganic & Medicinal Chemistry Letters. 18,
6171-6174.

Wellbrock C, Ogilvie L, Hedley D, Karasarides M, Martin J, Niculescu-Duvaz D,

Springer CJ , Marais R (2004). V599EB-RAF is an Oncogene in Melanocytes.
Cancer Res. 64, 2338-2342.

220

 

Whitwarn T, Vanbrocklin MW, Russo ME, Haak PT, Bilgili D, Resau JH, Koo HM ,
Holmen SL (2007). Differential oncogenic potential of activated RAS isoforms in
melanocytes. Oncogene. 26, 4563-4570. '

Wu H, Goel V , Haluska FG (2003). PTEN signaling pathways in melanoma. Oncogene.
22, 3113-3122.

Xia Z, Dickens M, Raingeaud J, Davis RJ , Greenberg ME (1995). Opposing effects of
ERK and JNK-p38 MAP kinases on apoptosis. Science. 270, 1326-1331.

Yeh T, Wallace E, Lyssikatos J , Winkler J (2004). ARRY-142886, a potent and
selective MEK inhibitor: II) Potency against cellular MEK leads to inhibition of
cellular proliferation and induction of apoptosis in cell lines with mutant Ras or
B-Raf. In Annual Meeting of the American Association for Cancer

ResearchedAeds), pp. 896-c-897.

Yeh TC, Marsh V, Bemat BA, Ballard J, Colwell H, Evans RJ, Parry J, Smith D,
Brandhuber BJ, Gross S, Marlow A, Hurley B, Lyssikatos J, Lee PA, Winkler JD,
Koch K , Wallace E (2007). Biological Characterization of ARRY-142886
(AZD6244), a Potent, Highly Selective Mitogen-Activated Protein Kinase Kinase
1/2 Inhibitor. Clin Cancer Res. 13, 1576-1583.

You MJ, Castrillon DH, Bastian BC, O'Hagan RC, Bosenberg MW, Parsons R, Chin L ,
DePinho RA (2002). Genetic analysis of Pten and Ink4a/Arf interactions in the
suppression of tumorigenesis in mice. Proc Natl Acad Sci U S A. 99, 1455-1460.

Zheng CF , Guan KL (1994). Activation of MEK family kinases requires
phosphorylation of two conserved Ser/Thr residues. EMBO J. 13, 1 123-1131.

Zhou XP, Girnm O, Hampel H, Niemann T, Walker MJ , Eng C (2000). Epigenetic

PTEN silencing in malignant melanomas without PTEN mutation. Am J Pathol.
157, 1123-1128.

221

irritant

030

3 1293